In the hyphal tip of Candida albicans we have made detailed quantitative measurements of (i) exocyst components, (ii) Rho1, the regulatory subunit of (1,3)-β-glucan synthase, (iii) Rom2, the specialized guanine-nucleotide exchange factor (GEF) of Rho1, and (iv) actin cortical patches, the sites of endocytosis. We use the resulting data to construct and test a quantitative 3-dimensional model of fungal hyphal growth based on the proposition that vesicles fuse with the hyphal tip at a rate determined by the local density of exocyst components. Enzymes such as (1,3)-β-glucan synthase thus embedded in the plasma membrane continue to synthesize the cell wall until they are removed by endocytosis. The model successfully predicts the shape and dimensions of the hyphae, provided that endocytosis acts to remove cell wall-synthesizing enzymes at the subapical bands of actin patches. Moreover, a key prediction of the model is that the distribution of the synthase is substantially broader than the area occupied by the exocyst. This prediction is borne out by our quantitative measurements. Thus, although the model highlights detailed issues that require further investigation, in general terms the pattern of tip growth of fungal hyphae can be satisfactorily explained by a simple but quantitative model rooted within the known molecular processes of polarized growth. Moreover, the methodology can be readily adapted to model other forms of polarized growth, such as that which occurs in plant pollen tubes.
INTRODUCTION
The capacity of the human fungal pathogen Candida albicans to grow in a true hyphal mode is a key virulence attribute allowing invasion of mucosal surfaces to establish both superficial infections in otherwise healthy individuals and life-threatening disseminated infections in vulnerable, immunocompromised patients. C. albicans hyphae grow in a highly polarized fashion from their tips in the same way as other fungal hyphae (1). Recently, there have been considerable advances in our understanding of the mechanism of polarized growth of C. albicans hyphae (reviewed in reference 2), providing an opportunity to use C. albicans as a model to facilitate understanding of the general process of polarized growth in fungal hyphae.
There have typically been two approaches to trying to understand the mechanism of such extreme polarized growth in fungal hyphae (which shares some common features with the process of bud formation in yeasts). In the “bottom-up” approaches, a wealth of information has been accumulated at the detailed molecular level; in the “top-down” approaches, investigators have tended to consider growth in terms of overall physical properties, without knowledge of the detailed molecular cellular biological processes. As noted by Harold (3) and later by Slaughter and Li (4), there has been relatively little progress in bridging the gap between these two levels of description. One very successful model, the vesicle supply center (VSC) model, has gone some way toward combining detailed concepts of cell biology and 3-dimensional (3D) physical calculations (5–7). However, this model predates much of the present knowledge of the details of molecular processes that has emerged from studies of model fungi such as Saccharomyces cerevisiae, Neurospora crassa, or Schizosaccharomyces pombe.
In this paper, we present steps toward reconciling these two approaches by a relatively simple mathematical treatment of the process of hyphal growth that incorporates three well-established molecular cell biological processes: the fundamental role of the spatial pattern of cell wall synthesis in determining hyphal form, the role of the exocyst in determining where the cell wall synthesis enzymes are delivered to the plasma membrane, and the role of endocytic patches in ensuring that these enzymes stay only in a discrete zone of the plasma membrane. We apply the model to the growth of hyphae in the human fungal pathogen C. albicans.
Dissection of the secretory pathway in S. cerevisiae established that polarized growth depends on the delivery of secretory vesicles to the sites of polarized growth (8). The fusion of a vesicle with the plasma membrane achieves three things. First, the membrane of the vesicle is incorporated into the plasma membrane, thus allowing expansion of the plasma membrane. Second, the contents of the vesicles, such as glycosylphosphatidylinositol (GPI)-linked mannoproteins, are released into the extracellular space (9). Third (and crucially), enzymes such as (1,3)-β-glucan synthase (GS), which catalyzes the biosynthesis of (1,3)-β-glucan from UDP-glucose monomers, are embedded in the plasma membrane and are thought to extrude (1,3)-β-glucan polymers into the extracellular space (10, 11). (1,3)-β-Glucan is the major load-bearing polymer of the cell wall (along with chitin) and forms a scaffold for the attachment of (1,6)-β-glucan and GPI-anchored mannoproteins (10, 11). Once extruded into the extracellular space, (1,3)-β-glucan is cross-linked to itself, to chitin, and to (1,6)-β-glucan, forming a complex 3-dimensional mesh. It is often assumed that this process of cross-linking is the reason why the cell wall becomes less deformable in subapical regions as it ages.
It must be emphasized that while tip extension in hyphae requires both the expansion of the plasma membrane and of the cell wall, it is the pattern of cell wall deposition that primarily controls the shape of the hyphal tip. The components of lipid membranes can flow, and thus, the lipid membrane can readily undergo large-scale deformations under the influence of forces, whereas, in contrast, the branched and cross-linked nature of the polysaccharide cell wall means that it is much less readily deformable and that it determines the shape of the cell (10). It is also very important to realize that vesicles, in the main, deliver the wall-building enzymes and not the wall material itself: to use a strong analogy, they deliver the bricklayers and not the bricks. In the context of glucan synthase, the bricks are UDP-glucose monomers. UDP-glucose is formed from glucose-1-phosphate and UTP by the action of UTP-glucose-1-phosphate uridylyltransferase (EC 2.7.7.9), encoded by UGP1 (12). In S. cerevisiae, this enzyme is cytoplasmic (13), so presumably glucan synthase takes up UDP glucose from the cytoplasm and extrudes the glucan polymer into the extracellular space. Interestingly, a reduction in Ugp1 activity to 10% of normal, resulting in a 50% drop in UDP-glucose levels, had no effect on the growth rate, suggesting that the supply of bricks is not rate-limiting (12). This paper focuses on the consequences of the vesicles delivering new cell wall synthesizing capacity.
Vesicles dock with the exocyst before fusion with the plasma membrane. The exocyst is a multiprotein complex composed of eight evolutionarily conserved proteins: Sec3, Sec5, Sec6, Sec8, Sec10, Sec15, Exo70, and Exo84 (14–17). In S. cerevisiae, temperature-sensitive mutations in exocyst components result in an accumulation of post-Golgi vesicles that are unable to fuse with the plasma membrane (8). Deletion of C. albicans SEC3 is not lethal but leads to an accumulation of vesicles in yeast buds, and upon hyphal induction the hyphal tip reverts to nonpolarized growth after the first septum forms, showing that an intact exocyst complex is required for the polarized growth of hyphae (18). It is thought that in S. cerevisiae, Sec3 and part of the Exo70 pool are located independently of the secretory pathway at sites of polarized growth, while the remaining components are carried on vesicles (14, 19). As vesicles arrive at sites of polarized growth, they are tethered to the plasma membrane by the formation of the complete eight-unit complex as the vesicle-associated exocyst components interact with Exo70 and Sec3 located on the plasma membrane. Thus, it follows that the spatial distribution and rate of vesicle fusion will be determined by the location and density of Exo70 and Sec3. We have shown that in C. albicans, all exocyst components locate to a surface crescent, while vesicle-associated proteins, such as Sec4, Sec2, and Mlc1, localize to a subapical spot (20). Moreover, the vesicle-associated proteins were much more dynamic in FRAP (fluorescence recovery after photobleaching) experiments than the exocyst components. Finally we showed that Sec4 disperses very rapidly upon disruption of the actin cytoskeleton, while the exocyst components are, again, much more stable. Based on these observations, we suggested that in C. albicans it is likely that not all vesicles carry their own complement of exocyst subunits. Rather, while exocyst components might originally be transported to the plasma membrane on vesicles, once delivered, they remain in place to tether other vesicles that do not carry exocyst components.
Enzymes, such as (1,3)-β-glucan synthase, that synthesize the main components of the cell wall are delivered to the plasma membrane by vesicles and continue to synthesize the cell wall until they are removed or negatively regulated. Thus, it is necessary to consider how cell wall expansion will lead to the movement of the enzymes away from the region of vesicle fusion and how the molecules are ultimately removed from the membrane. The likely mechanism for this removal is endocytosis, which is thought to occur at the actin cortical patches (21). In C. albicans, as in other fungi that form hyphae, the patches form a subapical collar (22–25). Endocytosis is required for hyphal growth, as evidenced by the fact that a variety of different mutations that interfere with endocytosis either abolish hyphal growth completely or result in abnormal hyphae with swollen tips (26–31). Similarly, treatment of hyphae with the actin-disrupting drug cytochalasin A results in hyphal tip swelling (23, 32).
It is assumed that in S. cerevisiae, the location of Sec3 and Exo70 is determined by the distribution of active GTP-bound Cdc42, and there has been much interest recently in how this Cdc42 distribution is established in the absence of external cues such as the bud site selection pathway (33–36). In this paper, we are not concerned with how the sites of polarized growth are established; instead, we take as our starting point the empirical fact that Exo70 and Sec3 localize to a crescent at the hyphal tip, where we can measure their intensity distribution. We present a model of hyphal tip growth in which secretory vesicles fuse to the plasma membrane at a rate proportional to the local concentration of Exo70 and Sec3, which, based on concurrently available evidence in both S. cerevisiae and C. albicans, determines the distribution of vesicle docking. Because Exo70 and Sec3 localize to the plasma membrane independently of vesicles in S. cerevisiae, we monitored these exocyst components as reporters of where vesicles dock in C. albicans. However, based on our previous work, we suspect that the other exocyst components may also be located at the cell surface rather than being carried on incoming vesicles (20). In our model, the cell wall-synthesizing enzymes, such as glucan synthase, that are consequently inserted continue to generate new cell wall material until they are internalized by endocytosis at the subapical bands of actin cortical patches.
To make the treatment quantitative, we divide the hyphal surface into a set of concentric annular regions. Within each annular region, we calculate the amount of synthase that will be deposited in the plasma membrane during a small interval of time, as secretory vesicles fuse at a rate determined by the local exocyst concentration. We also calculate the enlargement of the cell wall area that will occur during the same interval of time due to the activity of synthase molecules. Based on simple geometry, we calculate the resulting change in the shape of the hypha, assuming that it remains inflated by turgor pressure. We iterate this process forward in time and record the development of the size of each annulus and of the hypha as a whole.
By focusing on the development of annular regions as they develop at the tip, enlarge, and eventually become fixed in size in the parallel walls of the hyphal tube, we make it possible to maintain a close intuitive link between the mathematics of the calculations and the underlying physical and biological phenomena. The direct computational (i.e., numerical) method allows great flexibility in the choice of the rules that determine the rate of change of the cell surface area and prevents the mathematics from being highly specialized. Simplifying assumptions can remain obscured in more complex treatments, and a clear understanding of the existence and nature of these assumptions can be key to deciding where future experimentation should be directed. The equations can be used to calculate the development of the hyphal form (both in time and in 3D space) in a very straightforward manner. We have used the Python programming language (www.python.org), which is freely available across many different computing platforms and is highly regarded for its accessibility to nonspecialists. The simple nature of the computing method also facilitates the addition in the future of extra details, or modifications, as pertinent experimental data become available. Moreover, the methodology can be readily adapted to model other forms of polarized growth, such as those occurring in plant pollen tubes, animal cells, and axons. However, care must be taken to recognize that there may be fundamental differences between these systems. In fungal growth, we have emphasized here that vesicles deliver the capacity to synthesize the cell wall, whereas in pollen tubes, the vesicles deliver the cell wall itself in the form of pectin. We have made careful measurements in living C. albicans cells of the density of fluorescence from exocyst proteins fused to yellow fluorescent protein (YFP). We show that when these measurements are incorporated into the model, the predicted forms of the hyphae are in good accord with experimental observations only if a mechanism exists to remove synthase from the membrane (or downregulate its activity) at a position that is in good accord with the locations of the experimentally observed collars of actin patches. We further show that the distribution of active GS, as reported by green fluorescent protein (GFP) fusions with Rho1, the positive regulatory subunit of GS, and with its guanine-nucleotide exchange factor (GEF) Rom2, is much more extensive than the distribution of exocyst components, to a degree that is well predicted by the model.
MATERIALS AND METHODSStrains.
All strains were derived from BWP17. C-terminal fusions were generated as described previously using pFA-XFP plasmids carrying the appropriate URA3, ARG4, or HIS1 genes (37). An N-terminal GFP fusion of Rho1 was generated as described previously using the pFA-HIS1 pMAL-GFP plasmids (38). Full genotypes of the strains are provided in Table 1, and the oligonucleotides used are listed in Table S1 in the supplemental material.
Hyphal growth was induced by growing yeast cells at 30°C overnight to saturation in YEPD (2% glucose, 2% Bacto peptone [Difco], and 1% Bacto yeast extract [Difco] plus 80 mg uridine liter−1). The stationary-phase culture was washed with distilled water, inoculated at a 1:20 dilution into synthetic defined (SD) medium (consisting of 0.67% yeast nitrogen base [Difco], 2% [wt/vol] glucose, 80 mg of uridine liter−1, and 40 mg arginine and histidine liter−1 plus 10% calf serum [Sigma-Aldrich]), and incubated at 37°C. For live-cell imaging experiments, cells were grown on agar pads incorporating SD medium as described previously (39).
Microscopy and image processing.
Images of exocyst components fused to YFP, and of Rom2-GFP and GFP-Rho1, were made using a DeltaVision Spectris 4.0 RT microscope (Applied Precision Instruments, Seattle, WA) with an Olympus 100× UPlanApo lens (numerical aperture [NA], 1.35; Olympus, Tokyo, Japan). Images were acquired with Softworx software, which was also used to deconvolve z-stacks where specified.
Time lapse videos of exocyst components fused to YFP and of GFP-Rho1 were captured in a single z-plane over 10 min, during which images were taken every 30 s. Abp1-YFP fluorescence was captured in a z-stack of 2 μm in which sections were 0.2 μm deep, and the z-stack was then deconvolved. All images were taken with an exposure time of 0.3 s except for images of Rom2-GFP, where a single 10-s exposure was used.
Further image processing was performed with a combination of NIH ImageJ (40) and in-house Python programs (utilizing the Python Imaging Library [PIL] module). To bring cells into the standard vertical alignment, the position of the apex of the tip and the angle of the local hyphal axis at the tip were determined visually using ImageJ. For the time series, the positions of the apex in the first and last frames were used to determine the degree of translation required per image, assuming a constant rate of growth. Images were extracted from proprietary formats using ImageJ. Python programs were then used to extract, rotate, and translate a region (typically ∼3 by 3 μm around the apex) of each image and also to generate an average image via a simple pixel-by-pixel summation process.
To create final images, and to lay out multiple images, Python programs were used in conjunction with an in-house PostScript-based plotting system, jplot. Selected threshold levels were linearly mapped onto the full RGB (red-green-blue) intensity range in the output images. For the overlays, an opaque overlay method was used to avoid distortion. All other graphs were also made with jplot. ImageMagick (www.ImageMagick.org) or Ghostscript was used to convert PostScript images to bitmaps.
Image quantitation.
To measure the variability of the protein distribution in the time series, an automated tracing method was implemented. A point along the hyphal axis approximately one hyphal width back from the apex was chosen as an origin. A line was then traced out from the origin at an angle θ to the axis, and the point of maximum fluorescence intensity was found. The angle θ was then swept from −120° to 120° to produce the graphs in Fig. S1 in the supplemental material.
In order to allow a direct comparison between measurements made in different experiments where small variations in hyphal width occur (and to facilitate comparison with other organisms), we normalize all measurement of σ to the width of the hypha in which the measurement was made, and we express the results in hyphal width units (hwu). To obtain protein distributions as a function of arc length from the tip, for use in calculations, a trace was made around the average images in ImageJ. The coordinates of the trace were exported, and Python programs were used for extracting the intensity profile. The profiles were then fitted to a Gaussian function, allowing the intensity, width, and baseline offset of the function to be varied. Fitting was performed using the Levenberg Marquardt algorithm implemented in an in-house fitting program using Numerical Recipes routines.
Model calculations.
All hyphal growth calculations were performed using Python programs (the code used is presented in Table S2 in the supplemental material). For practical reasons, length units were considered to be in μm and were converted to hwu following the calculations. Final production calculations were run for 5,000 steps, initialized with 1,200 annular regions initially, with a central disc with a radius of 0.05 μm, and with equal spacing of 0.001 μm. The other parameters were σ (0.46 μm [measured from hyphae with radii of 1.7 μm; thus, σ is equal to 0.27 hwu]), γ (0.01), and ε (0.005). sendo and ϕ were set as described below. A new annular region was shed from the central region every 5 steps, maintaining ri/li at >50. For the illustrative examples in Fig. 4, a much coarser division into regions was used. These calculations were run with 30 annular regions initially, with a central disc with a radius of 0.2 μm, initial annular spacing of 0.1 μm, a σ of 0.4 μm, a γ of 0.01, and an ε of 0.005. For Fig. 4B, the value of sendo was 0.85 μm and the ϕ value was 0.005. A new annular region was shed from the central region every 200 steps.
In the model, ε is a scaling constant that subsumes various factors such as the number of synthase molecules per vesicle and the relationships between the unit of Ai and the arbitrary units of Ei and Si. γ is a scaling constant that subsumes factors such as the rate of synthesis per synthase molecule, the cell wall thickness, and conversion factors involving the units of various quantities. In the calculations, the combined factor εγ simply scales the amount of time represented by a time step and does not affect the form of the hypha. For example, the identical form is created if ε is doubled and γ is halved. Thus, it is important that the factor εγ is small enough that the amount of growth per time step is small. In our calculations, we set the time step δt to 1 (in arbitrary units). Both Ei and Si are best thought of as proportional to the numbers of molecules per unit area, although it will never be necessary (or possible) to specify their absolute units in this work, since such units could equally be subsumed into the parameters ε and γ. Because ϕ is defined as a value per time step, its interpretation is linked to the issue of the definition of the meaning of the time step, and hence to the particular choice of ε and γ. Therefore, it is best interpreted in terms of the range of arc length over which a significant drop in synthase density occurs due to endocytosis.
Raytracing.
Raytraced images were made using POV-Ray (www.povray.org). The output files from calculation time points were converted to POV-Ray primitive objects (cone and torus) using Python programs. The coloring of the primitive objects was based on the appropriate protein intensity (normalized to 1 at its maximum) linearly mapped onto 0 to 240° of the hue range of the HSL (hue, saturation, and lightness) representation of RGB color space.
RESULTSDetermination of protein distributions.
In order to parameterize and test our model, it was necessary to measure the distributions of a number of fluorescently tagged proteins in the tip.
Exocyst components.
In our first observations of GFP-tagged exocyst components made on rather short time scales (<0.2 s), we frequently observed that the distribution of exocyst density was not symmetrically arranged around the growth axis and that there was also a variation in the width of the distribution in different images. In order to test whether the asymmetry and variation in exocyst distribution were static and variable between cells, or were dynamic and more uniform between cells when time-averaged, we recorded time lapse videos, collecting frames every 30 s over a period of 10 min. We then analyzed data for cells that grew straight during this period; a representative cell is shown in Fig. 1A. For each such data set, we applied a rotation to the images to bring the hypha into a standard vertical orientation. We also used the initial and final tip positions to apply a translation (linearly scaled with time) to each image. This simple method was found sufficient to bring all images from a series into alignment during a time in which each hypha grows by approximately 2 hyphal diameters. The full data sets (see Fig. S1 in the supplemental material) confirm that the position of the maximum exocyst density is variable in time, a behavior we have observed repeatedly in all data sets collected from cells expressing Sec3-YFP, Sec8-YFP, Exo70-YFP, and Exo84-YFP.
Quantitation of protein distributions. (A) Differential interference contrast images showing the growth of a C. albicans hypha from time zero (left) to 10 min (right). The marker spots are at the same positions in the two images and show that the mother cell body does not move significantly in this time, while the tip moves through approximately 2 hyphal diameters. Bar, 5 μm. (B) Fluorescence images and quantitation of fluorescence intensity in live cells. Data are shown from time series for three cells expressing tagged exocyst components (Exo70, Sec3) and two cells expressing tagged Rho1 and from a collection of cells expressing Rom2-GFP. Leftmost panels show representative single images either from the time series (Exo70, Sec3, Rho1) or from a single cell (Rom2), with the minimum intensity set to the background outside the cell (plus 1 standard deviation). Central and right panels show average images from the time series (Exo70, Sec3, Rho1) or from a collection of 42 cell images (Rom2) with two threshold settings. In the central panels, the minimum intensity is set at the background outside the cell (plus 1 standard deviation), whereas in the panels on the right, the minimum intensity is set at the background cytoplasmic level away from the apex. Two thresholds are shown because the cytoplasmic background is much higher in the Rho1 and Rom2 images, obscuring the crescent at the tip. Bars, 1 μm. The graph on the right shows the fluorescence intensity traced around the hyphal tip (black dots) and the fit of these data to Gaussian curves (solid curves), plotted against the arc length from the tip of the cell. The value of σ for each fitted curve is shown. The fluorescence data were normalized to 1 at the maximum and to zero far from the apex. The units for arc length and σ are hyphal width units (hwu), in which the width of the hypha is 1 unit (see the text).
To provide a measure of exocyst density over a time period in which the hypha grows significantly, the images within each series were therefore averaged. Resulting average images are shown in Fig. 1B for Exo70-YFP and for two separate experiments using Sec3-YFP. The distribution of intensity was traced around each average image and was found to fit adequately to a Gaussian function. In order to allow a direct comparison between measurements made in different experiments where small variations in hyphal width occur (and to facilitate comparison with other organisms), we normalized all measurements of σ to the width of the hypha in which the measurement was made. In such hyphal width units (hwu), the fits for the three images for exocyst components shown in Fig. 1B give σ in a range of 0.26 to 0.28 hwu. As an example, a value for σ of 0.27 hwu corresponds to a σ of 0.46 μm in a hypha with a width of 1.7 μm.
GS and Rho1.
To measure the distribution of GS activity, we first attempted to determine the distribution of the localization of Gsl21-GFP. Gsl21 is one of three proteins encoded in the C. albicans genome homologous to S. cerevisiae genes encoding GS. However, we found that that its localization was anomalous, since it appeared be trapped in an intracellular membranous compartment.
GS is activated by a GTPase, Rho1 (41). In secretory vesicles, GS is complexed with Rho1-GDP, and Rho1 is converted to the active GTP-bound form by its GEF Rom2 only after fusion with the plasma membrane (42). In turn, Rom2 interacts with the cell wall stress sensors Mid2 and Wsc1 to Wsc3 (43). Rom2 is one of two GEFs for Rho1, along with Tus1. However, Rom2 has been shown to be the specialized GEF for Rho1 in its GS-activating role (44). We therefore used GFP-Rho1 as a surrogate marker of GS localization. The data for GFP-Rho1 were collected in the same fashion as those for the exocyst components. The distributions for GFP-Rho1 are considerably broader than those for the exocyst components, with a σ of ∼0.4 hwu.
Rom2.
We were unable to delete the second copy of Rho1 in the RHO1/GFP-RHO1 strain, raising doubts as to whether GFP-Rho1 is fully functional. Moreover, GS is active only when Rho1 is in its GTP-bound form, for which GFP-Rho1 is not informative. Because of these limitations of GFP-Rho1, we also measured the fluorescence of a functional Rom2-GFP strain. The fluorescence signal of Rom2-GFP was faint; nevertheless, upon long exposure, a broad crescent was evident (Fig. 1B). These long exposures bleached the GFP signal, precluding the repeated exposures necessary for time lapse movies, which we used to quantitate the fluorescence for the exocyst proteins. The data for Rom2-GFP were therefore obtained by averaging the images of a number of individual cells (n = 42). The distribution is even broader than that for GFP-Rho1, with a σ of ∼0.5 hwu.
Growth model.
Our treatment of hyphal geometry is shown in Fig. 2, and the set of associated variables is listed in Table 2. This treatment comes largely from the strong analogy between an inflated hypha and a Chinese paper lantern or lampshade (Fig. 2C). Such a lantern, when supplied flat (Fig. 2C, right), is wrinkled because the arc lengths (li) of the annular regions between the hoops are larger than the differences in the radii (ri) of successive hoops (Fig. 2A). The 3-dimensional shape that the lantern takes when pulled out makes each panel taut and thus extended to the maximum allowed by li. The wire or wood hoops in the lantern would not be necessary if its ends were sealed and it were inflated with gas.
Hyphal geometry. (A) Geometry. We treat the hypha as a set of concentric annular regions. Provided the hypha remains inflated by turgor pressure and the subdivision into annuli is sufficiently fine, each annular region approximates well to the form of a truncated conical shell with open circular ends, which is most conveniently characterized by two variables, the radius, ri, and the arc length, li. Conversely, if the values of ri and li are specified for all of the annular regions, the entire shape of the hypha can be calculated. The height of the truncated cone, and hence the contribution to the extension of the hypha, is readily calculated by applying Pythagoras' theorem to the shaded triangle: hi=li2−(ri−ri−1)2. The distance of each annular region i from the tip can then be calculated as di=∑j≤ihj, the arc length round to an annular region is calculated as si=∑j≤ilj, and the surface area of one face of the shell, Ai, is given by 2πliri (where li is ≪ri). The very apex of the tip is described by a small circular region. (B) A narrow annular region can be freely rotated. An annular region where li is ≪ri can be freely deformed as shown. Where li is ≪ri, the fractional changes in area or in the inner and outer radii of the annulus are negligible during such a deformation. (C) Analogy with a Chinese paper lampshade, shown extended (left) and in a flattened, packaged form (right). (D) Schematic representation of two time points during a calculation. A hyphal tip is shown with one annular region highlighted at time t (left) and at an arbitrary time Δt later (right). Synthase molecules deposited up to and including time t are represented by black circles. A synthase molecule deposited during the time Δt is depicted by a white circle. Assumption ii asserts that all molecules found in the annular region at time t are also found in the annular region at time t + Δt (unless endocytosis is active). The zigzag lines represent the extrusion of a polymer into the cell wall. Since this polymer is envisaged as enmeshed with the existing polymer, the synthase molecules are unable to diffuse relative to the local cell wall.
Summary of variables used in the model
Variable
Description
Fundamental geometrical entities
ri
Radius of annular region i
li
Arc length of annular region i
P
Point on which the annuli are centered
Derived geometrical variables
hi
Height contributed by annular region i
di
Distance of annular region i from tip
si
Arc length from P to annular region i
Ai
Surface area of one face of annular region i
Protein densities
Ei
Exocyst density in annular region i
Si
Synthase density in annular region i
Parameters
σ
Standard deviation of Gaussian distribution for E
ε
Parameter linking exocyst density to rate of synthase exocytosis
γ
Parameter linking synthase density to rate of surface area growth
sendo
Arc length from tip at which endocytosis starts (model using equation 3b)
ϕ
Fraction of synthase that undergoes endocytosis during one time step (model using equation 3b)
δt
Time step for updating the variables; acts simply to scale ε, γ, and ϕ and is therefore implicitly set to 1 in calculations
The disposition of paper in a lantern is fixed. A hypha is a much more complex structure—it is as if a lantern were having paper continually added to its panels. Therefore, in addition to these geometrical variables, each annular region is described in our model by two further variables, the local exocyst density (Ei) and the synthase density (Si). The computational process for updating the variables in the model, and hence “growing” a hypha, is defined in Fig. 3, and the corresponding equations used to update the variables are given in Table 3. [Note: we use the term synthase here in a generic way to represent all enzymes that contribute to the synthesis of new cell wall material. The major component of the cell wall that determines its shape is (1,3)-β-glucan synthase, whose distribution we address experimentally. In this paper we use the term synthase in theoretical modeling of the cell wall, and we use the abbreviation GS when we refer to the experimentally determined distribution of (1,3)-β-glucan synthase.]
Flow chart showing the steps for updating the variables describing hyphal geometry. Given a starting set of variables, the calculations in the box are first carried out for each annulus. From the new set of ri and li the new overall form of the hypha can be calculated (Fig. 2A). This process makes use of the fact that for an annular region where li is ≪ri, there is freedom to alter the degree of slope of the wall so that the radius matches up with that of the preceding ring (Fig. 2B). The process can then be iterated for many time steps.
Equations of the growth model
Throughout these equations superscript t or t + δt is used to indicate the value of the parameter at time t or t + δt. Equation 1 expresses the Gaussian distribution of exocyst density used in the model, which is parameterized from the experimental observation for σ. In equation 2, the new area of an annular region at time t + δt is determined by calculating the amount of cell wall synthesis in the small time δt (∝Sit Ait δt) and converting it to a change in area using the parameter γ. The two forms of equation 3 express the method by which the new density of synthase at time t + δt is calculated and differ only by whether endocytosis is included (equation 3b) or is neglected (equation 3a). In both forms, the amount of synthase added to an annular region in time δt is ∝Eit Ait δt and is converted to units of Si via the parameter ε. This amount is added to the preexisting amount of synthase and is then scaled down by the new area. In this way, the decrease in synthase density due to an increase in cell wall area is included. In equation 3b, if s is >sendo, a fraction of synthase (ϕ) is removed per time step due to endocytosis. In equations 4, the new radii and arc lengths of annular regions are calculated assuming an isotropic mode of growth in which the radius and arc length of an annular region grow by the same factor. The parameters ε and γ do not affect the form of the hypha; they affect only the amount of growth per time step (see Materials and Methods). The very apex of the tip is represented by a small circular region. This region grows according to the same growth model as the annular regions except that it is described only by a radius and not by an arc length. After a set number of steps, the region is replaced by a combination of a new annular region and the circular region with its radius reset to its initial value.
The assumptions inherent in the equations in Table 3 are as follows. (i) The only route of entry of the cell wall-building synthase into the membrane is via the docking of vesicles, and the rate of docking of vesicles with the plasma membrane is proportional to the exocyst density (equations 3a and 3b). The exocyst density is assumed to remain constant and is approximated in the model by a Gaussian distribution (equation 1).
(ii) The synthase deposited in an annular region remains located in the same annular region (equations 3a and 3b). (Equation 3b additionally allows for the removal of synthase by endocytosis.) We propose that upon deposition in the plasma membrane, synthase molecules immediately start to synthesize cell wall polymers in the existing cell wall matrix and thereby become “locked to” the local cell wall by the enmeshing of newly synthesized cell wall material (Fig. 2D).
(iii) The synthases work at a constant activity, so that the rate of cell wall material synthesis is simply proportional to the synthase density (equation 2). This is a strong statement about the primary regulation of cell wall growth, which we consider further in Discussion.
(iv) The cell wall thickness is constant, and so the change in area is proportional to the amount of cell wall material synthesized (equation 2). This is related to assumption iii and likewise is discussed further below.
(v) The expansion of the cell wall is isotropic (equations 4). The manner in which the cell wall deforms locally upon the deposition of new material has been considered in several previous models (5, 7, 45–47). In principle, it is relatively straightforward to calculate the internal forces in the cell wall, but it is much less straightforward to calculate how the cell wall will deform, because this depends, for instance, on whether the cell wall material is considered to deform as a plastic solid or as a viscous fluid. Given the uncertainty in this area, we assume the simplest case, that the expansion is isotropic (i.e., locally the same in all directions).
(vi) We also assume that any temporary imbalance between the membrane area and the cell wall area (due to exocytosis) will be rectified by lipid flow and endocytosis and that the cell shape is dictated by the fully inflated cell wall. From assumption ii, the distribution of synthase will be unaffected by this lipid movement.
We first consider calculations using the model incorporating equation 3a, in which no endocytosis occurs. The calculations are initialized from a flat cell wall (Fig. 4Ai) with sufficient equally spaced annular regions that the initial area of the cell wall is large compared to the cross-sectional area of the hypha that will emerge. The concentration of synthase is initialized to be zero throughout the cell wall. In the initial few time steps, the synthase concentration begins to build up in annular regions centered on point P in Fig. 2A, where the exocyst concentration is significant (equation 3a) (Fig. 4Aii). As the synthase levels rise, the areas of regions close to P increase, due to the deposition of cell wall material, and ri and li increase according to equations 2 and 4. In contrast, at a large distance from P, where Ei is small, synthase does not build up significantly and the annuli retain a constant radius and arc length. The only way that the enlarged radii and arc lengths of the annuli nearer P can be accommodated is by deformation of the planar cell wall, and in the presence of excess pressure on one side of the cell wall, the cell wall in the vicinity of P will bulge out (Fig. 4Aiii). The resulting height profile of the cell wall is readily calculated via the method described in Fig. 2.
Progress of hyphal growth calculations. For the purposes of making this figure, calculations were performed with a considerably coarser subdivision of the hyphal surface than that used for regular calculations (see Materials and Methods for details). The images in this figure (and in Fig. 5C) were produced by raytracing of the hyphal forms generated by the model, as described in Materials and Methods. The hyphae are colored on a scale of zero to 1, as shown on the key cylinder. In each hypha, the maximum protein density (exocyst or synthase density according to the part of the figure) is normalized to 1. (A and B) Illustrative time courses of calculations excluding (A) or including (B) endocytosis. Hyphal forms (colored according to synthase density) before any time steps (i) and after 20 (ii), 200 (iii), 500 (iv), 1,000 (v), and 1,500 (Avi) or 2,000 (Bvi) time steps are shown. In panel A, the calculations start from a flat cell wall with no synthase in the membrane (i). After a short time (ii), the synthase density has increased due to exocytosis, and at a later time (iii), the cell wall has started to grow locally and bulge. The synthase density remains nonzero far back from the tip, and so the hypha continues to expand in diameter (iv to vi). The necking at the base is pronounced, because in the early part of the calculation, the distribution of Si is determined by the relatively narrow distribution of Ei, whereas in later parts of the calculation, the distribution of Si becomes much broader. As shown in panel B, while the early parts of the calculation including endocytosis yield a hyphal form very similar to that in panel A, the distribution of synthase remains relatively narrow and settles down to a constant form (as a function of the arc length from the tip), thus giving rise to the parallel-sided hypha. A small degree of necking occurs in panel B, since the width of the distribution of Si approximately doubles from time ii to time v. (C) Illustrative positions of the hyphal tip showing the development of the size and position of annular regions and of the corresponding protein densities. The tip of the hypha is shown at time t and at a time Δt later. The tips are colored according to exocyst density or synthase density as marked. The arrows show the correspondence between the annular regions at the two times, showing how the geometry of the regions varies in time. At the later time, the tip colored according to synthase density is shown in order to emphasize the distinction between the distributions of exocyst density and synthase density that arises in the model. The time points shown are from a calculation using the model with endocytosis.
Subsequent steps of the calculation are shown in Fig. 4Aiv to vi. Although a somewhat tubular hypha is grown, it is always thicker at the base (once an initial constriction due to the initial conditions is overcome), and it continues to grow in width without limit. This is because synthase remains in the membrane far back from the tip. It appears, therefore, that it is essential to have a mechanism for active removal of the synthase enzymes from the membrane. A role for endocytosis in the recycling of polarity proteins has been proposed previously (24, 25, 48). However, so far as we are aware, a critical role of endocytosis in removing cell wall-synthesizing enzymes has not been proposed previously. The key experimentally observed features of the actin patch distribution, as visualized by Abp1-YFP, are an absence of endocytosis close to the apex of the tip and an extensive band farther back from the tip (Fig. 5A).
Actin patches and zones of endocytosis. (A) Fluorescence and differential interference contrast images of typical “collars” of actin patches in live C. albicans cells. Images of 5 cells are shown, with the fluorescence of Abp1-GFP overlaid onto the differential interference contrast image. For maximum contrast, the overlay is transparent for values below the minimum threshold and opaque for values above the threshold. In each image, the intense spots of fluorescence are interpreted as actin patches at sites of endocytosis events. The patches localize predominantly to a collar, although a few patches are also observed closer to the apex (e.g., in the left-hand image). Bars, 1 μm. (B) Four possible endocytosis scenarios yielding unit hyphal width. The form of the hypha (visualized as half cross-sections) is shown for calculations run with a σ of 0.27 hwu for four pairs of sendo and ϕ values yielding a self-consistent hyphal width of 1.0 hwu. From left to right, the values for sendo and ϕ are 0.60 hwu and 0.050, 0.48 hwu and 0.010, 0.36 hwu and 0.006, and 0.24 hwu and 0.0045, respectively. The region in green extends from the point where s is equal to sendo to the point where the synthase density drops to 5% of its value when s is equal to sendo. Tick marks represent intervals of 0.25 hwu. (C) Raytraced 3D forms of hyphae generated with a σ of 0.27 hwu, an sendo of 0.36 hwu, and a ϕ of 0.006. The same form is shown in each part, colored according to (i) the synthase activity (Si) predicted by the model, (ii) the observed GS activity (as inferred from Rho1 distribution and approximated to the experimental data by a Gaussian σexp of 0.41 hwu), or (iii) the observed exocyst distribution (as approximated to the experimental data by a Gaussian σexp of 0.27 hwu). The same color scheme is used as for Fig. 4.
We introduced a very simple model for such endocytosis by modifying equation 3a to the form in equation 3b, where sendo is the arc length from the tip at which the band of endocytosis starts and ϕ is the fraction of synthase that is removed from the membrane by endocytosis during each time step. When ϕ is small, the zone where Si decays significantly will be broad. As ϕ is increased, the zone will become narrower. With this simple adaptation, the model generates a form of growth in which, starting from a flat membrane, the geometry close to the tip of the bulge varies in time as it distorts, but the tip rapidly takes on a form that translates without deformation as the tip moves away from the membrane (Fig. 4B). A corollary of this is that the radii of annuli as they leave the growth region become constant, and a parallel-sided tube results. Details of the way in which annular regions develop near the tip are shown in Fig. 4C.
We next examined the dependence of the predicted hyphal width on the parameters sendo and ϕ. We used the experimentally observed value of 0.27 hwu for σ (the average of the three values determined for the distribution of the exocyst in Fig. 1) and ran a number of separate calculations with sendo set to 0.24, 0.36, 0.48, and 0.60 hwu. For each value of sendo, we iteratively adjusted ϕ to give a self-consistent hyphal width of 1.0. The corresponding zones where Si decays significantly are shown in Fig. 5B. Assuming that such zones correspond to the zone of endocytosis as visualized by Abp1-YFP, a simple visual comparison indicates that the best agreement of the model with the distribution of actin patches is for an sendo of ∼0.36 hwu and a ϕ of 0.006. With this choice of sendo and ϕ, we can visualize the results by mapping the predicted distribution of synthase from the model onto the calculated 3D form of the hypha (Fig. 5Ci). This can then be compared with the observed distribution of active GS (as inferred from the distribution of Rho1) mapped onto the same 3D form (Fig. 5Cii), together with the observed distribution of the exocyst components (Fig. 5Ciii). It is clear that the model predicts a broader distribution of synthase than of the exocyst, in agreement with that observed. Comparison of Fig. 5A and B shows that the choice of an sendo of ∼0.36 hwu is only a very approximate estimate and that a broad range of sendo values is compatible with the observed distribution of Abp1. However, values of sendo in the range of 0.4 to 0.8 give broadly similar degrees of agreement between the predicted synthase and observed GS distributions. As sendo is decreased and the zone of endocytosis increases, the hyphal tip becomes more pointed (Fig. 5B). The precise forms of the hyphal tips observed experimentally are, however, quite variable and dynamic, so we do not believe that any attempt to distinguish further between these endocytosis models on the basis of current observations is warranted.
DISCUSSION
Work endeavoring to understand the overall growth of hyphae dates back at least a century. Consequently, numerous theories have been propounded, sometimes with overlapping concepts and confusing changes of nomenclature. The reviews (and work) of Harold, Wessels, and Koch (3, 49, 50) are very useful summaries of the body of thought that predates the development of GFP-tagging methodologies. These reviews contain explanations of the soft spot, surface stress, steady-state, and vesicle supply center models. The molecular detail that has emerged from the ability to visualize the locations of specific proteins has been brought into this context by the review of Slaughter and Li (4).
In their various guises, the soft spot, surface stress, and steady-state models place as the key controlling factor in tip growth the fact that newly synthesized cell wall is more readily deformable than older cell wall. Theoretical treatments of material properties (3, 46, 47, 51) have followed from these ideas. All these models place the physical properties of the cell wall at center stage, for instance, relying on the aging (generally considered to be cross-linking over time) of the cell wall material to control the final fixing of the hyphal tube shape.
In contrast, the vesicle supply center model remains the most developed and most widely discussed quantitative model, which arguably first brought a strong cell biological focus to the control of hyphal growth. The key distinction from the surface stress model (and related models) is that while in the VSC models the deformation of the hyphal wall also arises from the effect of turgor pressure and the plasticity of the hyphal wall, this process is regulated primarily by biological control. In the VSC model, this pattern is determined by the locations of vesicle fusion events, which are calculated on the basis of the distance of the membrane from the vesicle supply center, and such events are assumed to deliver cell wall material directly.
We have presented here a model that, for the first time, incorporates the knowledge that vesicles must dock with the exocyst before fusion with the plasma membrane and that it is primarily the cell wall-synthesizing enzymes that are delivered by such events, and not the cell wall material itself. Crucially, the synthase enzymes so inserted will remain in the membrane until they are removed by endocytosis. Using the experimentally measured distribution of exocyst subunits and actin cortical patches, our model successfully predicts hyphal form. Moreover, the key prediction that the area occupied by synthase is significantly greater than that of exocyst subunits is consistent with the observed distributions of GFP-Rho1 and Rom2-GFP.
We have demonstrated that a dynamic model based on decomposition of the hypha into annular regions can be readily constructed. This methodology allows alternative biological mechanisms to be evaluated. For example, if the exocyst does not determine the location of vesicle docking, then presumably the exocyst is arriving on vesicles, and the observed distribution of the exocyst on the cell surface marks the region where vesicles are docking. Note that in such a case in the presence of diffusion, the region of docking is probably even narrower than that observed. Furthermore, an implicit or explicit assumption of many treatments of polarized growth in fungi is that the vesicles deliver new cell wall material rather the capacity to synthesize this material. A model where vesicles only deliver new cell wall material to the region marked by exocyst distribution will generate a hyphal shape, but crucially, the width of the hypha is narrower than that observed by a factor of 0.7 (see Fig. S2 in the supplemental material). To recapitulate a hypha of the width observed, the area occupied by the exocyst would have to be much more extensive than that observed and would be equivalent to the distribution of GS in Fig. 5Cii. The concrete yet mathematically simple nature of our treatment allows us to avoid descriptions that either are qualitative or are framed in the language of complex applied mathematics. For example, in Fig. 2D we use the consideration of an explicit annular region to clarify the meaning of our assumption ii, that synthase molecules remained locked to the local cell wall. The observed distributions of GFP-Rho1 and Rom2-GFP are consistent with the model incorporating this assumption; however, we cannot currently rule out the alternative possibility that synthase is able to move relative to the local cell wall. This could occur via diffusion in the membrane and/or via movement with the local membrane as it is reorganized due to the addition of new membrane following the fusion of secretory vesicles. This would require us to incorporate into the model the ability of molecules to move from one annular region to a neighboring region. This cannot be modeled without knowledge of the rates of synthase enzyme insertion at the exocyst and endocytosis at the actin cortical patches, the diffusion constant of synthase within the membrane, and the rate of lipid insertion into the membrane. At present, all of these parameters are unknown.
The maintenance of a constant cell wall thickness (assumption iv) is one of the least understood aspects of hyphal growth. It is an implicit assumption in most models of hyphal growth, including the vesicle supply model, and appears to be supported by the available experimental data. The maintenance of a constant cell wall thickness requires a delicate balance between cell wall deposition, which will thicken the wall, and cell wall deformation, which will tend to thin the wall (49). The topic is intricately bound up with the question of the regulation (or lack of regulation) of synthase and also of enzymes that can modify the physical properties of the cell wall, such as the glucanases and cross-linking enzymes.
We interpret the observation of a readily detectable presence of Rom2 with a distribution broader than that of Rho1 as evidence that all the Rho1/GS complexes are activated. Therefore, in our model, the local rate of cell wall material deposition is determined solely by the local density of synthase (assumption iii). In such a scenario, the local cell wall deformation rate must be tuned to this rate. Tuning of the local cell wall deformation rate could be achieved by appropriate control of glucanases and cross-linking enzymes.
The alternative scenario is that the local cell wall synthesis rate is tuned to the local rate of cell wall deformation. We have argued from our observations of Rom2 that our current data suggest that there is little spatial control of GS activity. This does not rule out a role for global regulation of the synthesis rate, via, for instance, regulation of the overall GS levels at a transcriptional level. In S. cerevisiae, the cell wall sensors Wsc1 and Mid2 activate the cell wall integrity (CWI) mitogen-activated protein (MAP) kinase pathway to upregulate the expression of cell wall-synthesizing enzymes after cell wall damage. In S. cerevisiae, β-1,3-glucan synthase is encoded by two genes, FKS1 and FKS2 (43). During unstressed growth, FKS1 is constitutively active and is only weakly subject to CWI regulation. In contrast, FKS2 is upregulated by the CWI pathway upon cell wall damage. This suggests that the CWI pathway may be important only under conditions of cell wall damage and not in normal cell growth (43). The interaction of Wsc1 and Mid2 with Rom2, in order to recruit Rom2 to sites of cell wall stress, may likewise be relevant only under conditions of cell wall damage. Our computational framework is very flexible and will allow us to incorporate new knowledge of these mechanisms as it becomes available.
Without a cutoff imposed by endocytosis to finally remove the synthase from the membrane, the simulations predicted hyphal swelling. We showed that a range of endocytosis parameters consistent with the observed actin patch distribution could generate hyphae of the appropriate form, although we lack sufficient information on the rates of endocytosis for specific membrane components to make the model more specific. As described above, there is considerable experimental support for a key role of endocytosis in the polarized growth of hyphae. A role for endocytosis in the recycling of the polarized growth proteins has been proposed previously (24, 25, 48). Proteins such as v-SNARES, t-SNARES, and other polarity proteins will need to be recycled, and endocytosis has been proposed to do this. However, according to our model, a key role of endocytosis is to remove cell wall-synthesizing enzymes so as to prevent continued cell wall synthesis, which would lead to hyphal swelling rather than to polarized growth. Defining this key function of endocytosis does not mean that endocytosis does not also carry out the recycling roles suggested previously. Indeed, the internalized cell wall-synthesizing enzymes could be recycled to the tip, and this could be essential for the rapid hyphal tip extension observed in many fungi, as suggested previously (48). The recycling of endocytic vesicles carrying polarity markers and cell wall-synthesizing enzymes into exocytic vesicles (52) could lead to the accumulation of vesicles that constitute the Spitzenkörper and thus explain the key role that the Spitzenkörper plays in the extremely polarized growth characteristic of fungal hyphae.
An insight provided by these simulations is that polarized outgrowth from a flat surface without endocytosis generates a structure that closely resembles the constricted neck between a mother and a bud of a yeast cell (Fig. 4A). The intuitive explanation is that in the very first stages of the outgrowth, cell wall-synthesizing enzymes are located at the site of insertion into the plasma membrane, leading to polarized growth. As this growth continues, these enzymes become located progressively farther back, leading to a more swollen structure. The initial period of polarized growth forms the narrow neck, while the later stages form the subsequent swelling at the base of the round bud. This explanation is reminiscent of an argument proposed previously (4). This raises the possibility that the variety of structures observed in the growth forms of C. albicans, such as yeast, pseudohyphal, and hyphal growth, may result from different balances between endocytosis and polarized growth. In support of this notion, we have observed that the lifetime of actin patches is shorter in hyphal cells than in yeast cells (D. Caballero-Lima and P. E. Sudbery, unpublished data), an observation that has also been reported by others (53). This may indicate that endocytosis is more active at hyphal tips than at the tips of pseudohyphal buds and buds of small yeast cells. This example illustrates how the open nature of our model allows it to be applied to situations other than the tip growth of hyphae. As discussed above, the treatment also allows one to ask clear-cut questions about which there is little knowledge at present. In particular, what is the nature of the regulation that allows cell wall synthesis and deformation to act in closely tuned harmony? If it is shown that negative regulation of synthase operates and the pattern of this regulation is experimentally determined, then equation 3b can be modified by adding a term that quantifies the regulation, in the same way in which a term was added to equation 3a to construct equation 3b, which accounts for the removal of synthase by endocytosis. Thus, as new knowledge becomes available, it can be readily incorporated into our model.
Supplementary Material
Supplemental material
Published ahead of print 10 May 2013
Supplemental material for this article may be found at http://dx.doi.org/10.1128/EC.00085-13.
ACKNOWLEDGMENTS
This work was supported by BBSRC research grant BB/J002305/1. S.W. and I.K. were supported by BBSRC doctoral training grants. DeltaVision microscope images were obtained using the facilities of the Sheffield University Light Microscope Facility (LMF), supported by Wellcome Trust grant GR077544AIA.
REFERENCESSollDRHermanMAStaebellMA1985The involvement of cell wall expansion in the two modes of mycelium formation of Candida albicans.
131:2367–23753906042SudberyPE2011Growth of Candida albicans hyphae.
9:737–74821844880HaroldFM2002Force and compliance: rethinking morphogenesis in walled cells.
37:271–28212431461SlaughterBLiR2006Toward a molecular interpretation of the surface stress theory for yeast morphogenesis.
18:47–5316337116Bartnicki-GarciaSHergertFGierzG1989Computer-simulation of fungal morphogenesis and the mathematical basis for hyphal (tip) growth.
153:46–57GierzGBartnicki-GarciaS2001A three-dimensional model of fungal morphogenesis based on the vesicle supply center concept.
208:151–16411162061TindemansSHKernNMulderBM2006The diffusive vesicle supply center model for tip growth in fungal hyphae.
238:937–94816105670NovickPFerroSSchekmanR1981Order of events in the yeast secretory pathway.
25:461–4697026045EsmonBNovickPSchekmanR1981Compartmentalized assembly of oligosaccharides on exported glycoproteins in yeast.
25:451–4607026044KlisFMBoorsmaADe GrootPWJ2006Cell wall construction in Saccharomyces cerevisiae.
23:185–202.10.1002/yea.134916498706LesageGBusseyH2006Cell wall assembly in Saccharomyces cerevisiae.
70:317–34316760306DaranJMDalliesNThines-SempouxDPaquetVFrancoisJ1995Genetic and biochemical characterization of the Ugp1 gene encoding the UDP-glucose pyrophosphorylase from Saccharomyces cerevisiae.
233:520–5307588797HuhWKFalvoJVGerkeLCCarrollASHowsonRWWeissmanJSO'SheaEK2003Global analysis of protein localization in budding yeast.
425:686–69114562095HeBGuoW2009The exocyst complex in polarized exocytosis.
21:537–54219473826HeBXiFZhangJTerBushDZhangXGuoW2007Exo70p mediates the secretion of specific exocytic vesicles at early stages of the cell cycle for polarized cell growth.
176:771–77717339375TerBushDRNovickP1995Sec6, Sec8, and Sec15 are components of a multisubunit complex which localizes to small bud tips in Saccharomyces cerevisiae.
130:299–3127615633TerBushDRMauriceTRothDNovickP1996The exocyst is a multi-protein complex required for exocytosis in Saccharomyces cerevisiae.
15:6483–64948978675LiCRLeeRT-HWangYMZhengXDWangY2007Candida albicans hyphal morphogenesis occurs in Sec3p-independent and Sec3p-dependent phases separated by septin ring formation.
120:1898–190717504812BoydCHughesTPypaertMNovickP2004Vesicles carry most exocyst subunits to exocytic sites marked by the remaining two subunits, Sec3p and Exo70p.
167:889–90115583031JonesLASudberyPE2010Spitzenkorper, exocyst, and polarisome components in Candida albicans hyphae show different patterns of localization and have distinct dynamic properties.
9:1455–146520693302KaksonenMToretCPDrubinDG2005A modular design for the clathrin- and actin-mediated endocytosis machinery.
123:305–32016239147AndersonJSollDR1986Differences in actin localization during bud and hypha formation in the yeast Candida albicans.
132:2035–20473540192CrampinHFinleyKGerami-NejadMCourtHGaleCBermanJSudberyPE2005Candida albicans hyphae have a Spitzenkorper that is distinct from the polarisome found in yeast and pseudohyphae.
118:2935–294715976451Taheri-TaleshNHorioTAraujo-BazanLDouXEspesoAPenalvaMAOsmaniSAOakleyBR2008The tip growth apparatus of Aspergillus nidulans.
19:1439–144918216285UpadhyaySShawBD2008The role of actin, fimbrin and endocytosis in growth of hyphae in Aspergillus nidulans.
68:690–705.10.1111/j.1365-2958.2008.06178.x18331474AslesonCMBensenESGaleCAMelmsASKurischkoCBermanJ2001Candida albicans INT1-induced filamentation in Saccharomyces cerevisiae depends on Sla2p.
21:1272–128411158313BorthNWaltherAReijnstPJordeSSchaubYWendlandJ2010Candida albicans Vrp1 is required for polarized morphogenesis and interacts with Wal1 and Myo5.
156:2962–296920656786MartinRHellwigDSchaubYBauerJWaltherAWendlandJ2007Functional analysis of Candida albicans genes whose Saccharomyces cerevisiae homologues are involved in endocytosis.
24:511–52217431925OberholzerUIoukTLThomasDYWhitewayM2004Functional characterization of myosin I tail regions in Candida albicans.
3:1272–128615470256ReijnstPWaltherAWendlandJ2010Functional analysis of Candida albicans genes encoding SH3-domain-containing proteins.
10:452–46120402797WaltherAWendlandJ2004Polarized hyphal growth in Candida albicans requires the Wiskott-Aldrich syndrome protein homolog Wal1p.
3:471–48215075276AkashiTKanbeTTanakaK1994The role of the cytoskeleton in the polarized growth of the germ tube in Candida albicans.
140:271–2808180692JohnsonJMJinMLewDJ2011Symmetry breaking and the establishment of cell polarity in budding yeast.
21:740–74621955794JoseMTollisSNairDSibaritaJBMcCuskerD2013Robust polarity establishment occurs via an endocytosis-based cortical corralling mechanism.
200:407–41823401000SavageNSLaytonATLewDJ2012Mechanistic mathematical model of polarity in yeast.
23:1998–201322438587Wedlich-SoldnerRAltschulerSWuLLiR2003Spontaneous cell polarization through actomyosin-based delivery of the Cdc42 GTPase.
299:1231–123512560471GolaSMartinRWaltherADunklerAWendlandJ2003New modules for PCR-based gene targeting in Candida albicans: rapid and efficient gene targeting using 100 bp of flanking homology region.
20:1339–134714663826SchaubYDunklerAWaltherAWendlandJ2006New pFA-cassettes for PCR-based gene manipulation in Candida albicans.
46:416–42917009297BishopALaneRBenistonRLazoBSmytheCSudberyP2010Hyphal growth in Candida albicans requires the phosphorylation of Sec2 by the Cdc28-Ccn1/Hgc1 kinase.
29:2930–294220639857SchneiderCARasbandWSEliceiriKW2012NIH Image to ImageJ: 25 years of image analysis.
9:671–67522930834QadotaHPythonCPInoueSBArisawaMAnrakuYZhengYWatanabeTLevinDEOhyaY1996Identification of yeast Rho1p GTPase as a regulatory subunit of 1,3-beta-glucan synthase.
272:279–2818602515AbeMQadotaHHirataAOhyaY2003Lack of GTP-bound Rho1p in secretory vesicles of Saccharomyces cerevisiae.
162:85–9712847085LevinDE2011Regulation of cell wall biogenesis in Saccharomyces cerevisiae: the cell wall integrity signaling pathway.
189:1145–117522174182KrauseSACundellMJPoonPPMcGhieJJohnstonGCPriceCGrayJV2012Functional specialisation of yeast Rho1 GTP exchange factors.
125:2721–273122344253Bartnicki-GarciaSBrackerCEGierzGLópez-FrancoRLuH2000Mapping the growth of fungal hyphae: orthogonal cell wall expansion during tip growth and the role of turgor.
79:2382–2390.10.1016/S0006-3495(00)76483-611053117CampàsOMahadevanL2009Shape and dynamics of tip-growing cells.
.10.1016/j.cub.2009.10.075DumaisJLongSRShawSL2004The mechanics of surface expansion anisotropy in Medicago truncatula root hairs.
136:3266–327515448192ShawBDChungDWWangCLQuintanillaLAUpadhyayS2011A role for endocytic recycling in hyphal growth.
115:541–54621640317WesselsJGH1993Wall growth, protein excretion and morphogenesis in fungi.
123:397–413KochL1994The problem of hyphal growth in streptomycetes and fungi.
171:137–150GorielyATaborM2008Mathematical modeling of hyphal tip growth.
22:77–83.10.1016/j.fbr.2008.05.001ReadNDHickeyPC2001The vesicle trafficking network and tip growth in fungal hyphae, p 137–147InGeitmanACrestiMHeathIB (ed),
.
IOS Press,
Amsterdam, NetherlandsZengGWangYMWangY2012Cdc28-Cln3 phosphorylation of SlaI regulates actin patch dynamics in different modes of fungal growth.
23:3485–349722787279WilsonBDavisDMitchellAP1999Rapid hypothesis testing in Candida albicans through gene disruption with short homology regions.
181:1868–187410074081oai:pubmedcentral.nih.gov:36974682013-07-02eukcellpmc-openEukaryot CellEukaryotic CelleukcelleukcellEUKCELLEukaryotic Cell1535-97781535-9786American Society for Microbiology1752 N St., N.W., Washington, DCPMC3697468PMC36974683697468236871152368711500096-1310.1128/EC.00096-13ArticlesA SAS-6-Like Protein Suggests that the Toxoplasma Conoid Complex Evolved from Flagellar Componentsde LeonJessica CruzaScheumannNicolebBeattyWandycBeckJosh R.dTranJohnson Q.aYauCandaceaBradleyPeter J.dGullKeithbWicksteadBillbeMorrissetteNaomi S.aDepartment of Molecular Biology and Biochemistry, University of California, Irvine, Irvine, California, USASir William Dunn School of Pathology, University of Oxford, Oxford, United KingdomDepartment of Molecular Microbiology, Washington University School of Medicine, St. Louis, Missouri, USADepartment of Microbiology, Immunology and Molecular Genetics, University of California, Los Angeles, Los Angeles, California, USACentre for Genetics and Genomics, University of Nottingham, Nottingham, United KingdomAddress correspondence to Naomi Morrissette, nmorriss@uci.edu, or Bill Wickstead, bill.wickstead@nottingham.ac.uk.
SAS-6 is required for centriole biogenesis in diverse eukaryotes. Here, we describe a novel family of SAS-6-like (SAS6L) proteins that share an N-terminal domain with SAS-6 but lack coiled-coil tails. SAS6L proteins are found in a subset of eukaryotes that contain SAS-6, including diverse protozoa and green algae. In the apicomplexan parasite Toxoplasma gondii, SAS-6 localizes to the centriole but SAS6L is found above the conoid, an enigmatic tubulin-containing structure found at the apex of a subset of alveolate organisms. Loss of SAS6L causes reduced fitness in Toxoplasma. The Trypanosoma brucei homolog of SAS6L localizes to the basal-plate region, the site in the axoneme where the central-pair microtubules are nucleated. When endogenous SAS6L is overexpressed in Toxoplasma tachyzoites or Trypanosoma trypomastigotes, it forms prominent filaments that extend through the cell cytoplasm, indicating that it retains a capacity to form higher-order structures despite lacking a coiled-coil domain. We conclude that although SAS6L proteins share a conserved domain with SAS-6, they are a functionally distinct family that predates the last common ancestor of eukaryotes. Moreover, the distinct localization of the SAS6L protein in Trypanosoma and Toxoplasma adds weight to the hypothesis that the conoid complex evolved from flagellar components.
INTRODUCTION
Centrioles and basal bodies are microtubule-based structures found in many eukaryotic lineages, including animals, lower plants, and diverse unicellular organisms (1–3). Basal bodies are distinguished from centrioles by association with a flagellar axoneme that is templated from an extension to the centriole known as a transition zone (1, 4). The widespread occurrence of centrioles and basal bodies shows that they have an ancient origin in eukaryotes (5–8), but lineages such as flowering plants and most fungi have lost the ability to build these structures. Centrioles and basal bodies typically contain nine triplet microtubules organized with radial symmetry, although they can rarely be built of nine doublet or singlet microtubules (e.g., those in Drosophila embryos and Caenorhabditis elegans testes, respectively) or have likely 6-fold symmetry (9, 10). In spite of their presence at spindle poles, centrioles are dispensable for bipolar spindle formation in several lineages (11–15). However, all species that build centrioles/basal bodies have flagella at some stage of their life cycle; this observation likely underlies the true evolutionary imperative of these structures (2, 3).
Many protozoan organisms, including apicomplexan and kinetoplastid parasites, use spatially separated and morphologically distinct microtubule-organizing centers (MTOCs) to organize individual microtubule populations (16). Kinetoplastid parasites of the genera Trypanosoma and Leishmania cause human infectious diseases, including African sleeping sickness, Chagas disease, and kala azar (17). The shape of these organisms is maintained by an array of ∼100 densely packed corset microtubules that underlie the plasma membrane (18). The single flagellum is nucleated by a membrane-docked basal body that is distant from the nucleus (18–22). Flagella are not disassembled during division, and in replicating cells, basal bodies do not contribute to the organization of the poles of the intranuclear spindle (23). Apicomplexan parasites also cause a variety of medically significant diseases, including malaria, toxoplasmosis, and cryptosporidiosis (17). Apicomplexans typically have a complex life cycle that involves both asexual and sexual replication, and they alter microtubule architecture between the sexual and asexual life cycle stages. Microgamete motility is required for fertilization of macrogametes and is powered by flagella that originate from apical basal bodies (24). Asexual stages lack flagella and use a characteristic actin-and-myosin-based gliding motility to invade host cells (25, 26). These asexual forms (e.g., merozoites, tachyzoites) have two microtubule populations; spindle microtubules coordinate chromosome segregation during mitosis, and subpellicular microtubules subtend the pellicle to impose an elongated cell shape and cell polarity (27–29). Each microtubule population is associated with a distinct MTOC; subpellicular microtubules radiate from the apical polar ring (APR), an MTOC unique to apicomplexan organisms (28, 30, 31), whereas spindle microtubules originate near a specialized region of the nuclear envelope termed the centrocone (32–35).
Toxoplasma gondii is a member of the Coccidia—a subclass of apicomplexan parasites that build two tubulin-based structures in addition to the spindle and subpellicular microtubules, i.e., the conoid and centrioles. The conoid is an apical organelle constructed of comma-shaped tubulin sheets that spiral to form a cone-shaped structure (36–38). Two preconoidal rings surmount the conoid, and when extended, the conoid and preconoidal ring complex reside above the APR (28, 39). The conoid and preconoidal rings can also retract through the APR to be surrounded by the subpellicular microtubules. The conoid is permanently retracted in intracellular parasites, but extracellular tachyzoites extend and retract this structure, a probing behavior that is believed to facilitate host cell invasion (36–38). Two short, closely apposed microtubules are located at the center of the conoid. In contrast to the apical conoid, Toxoplasma centrioles are found in the cytoplasm proximal to the nucleus and duplicated centrioles are found at the spindle poles in the region of the centrocone structures. Coccidian centrioles are composed of nine singlet microtubules and a central tubule and are organized in a parallel rather than an orthogonal configuration (27, 33, 40).
SAS-6 is required for centriole biogenesis in eukaryotes ranging from protozoa to vertebrates (41–44). Chlamydomonas reinhardtii mutants deficient in SAS-6 (bld12) produce basal bodies with variable numbers of triplet blades (45). Studies of nematodes and Drosophila indicate that SAS-6 is needed to form the “cartwheel hub” in nascent centrioles (45–49). The structural properties of SAS-6 elegantly dovetail with its ability to template a centriole (50, 51). SAS-6 assembles into homodimers that contain a globular head domain at the N terminus and an extended coiled-coil rod. Remarkably, SAS-6 homodimers can self-assemble into a structure with a central hub and nine rods, akin to the cartwheel observed in structural studies of centrioles. In this paper, we describe for the first time the properties of a family of SAS-6-like proteins that are related to SAS-6 but structurally distinct. We investigate the function of SAS-6-like proteins in two distantly related protozoa and discuss their evolutionary relationship to conventional SAS-6.
MATERIALS AND METHODSCulture of parasites.
Toxoplasma lines, including the previously described RNG1-yellow fluorescent protein (YFP) line (31, 52), were grown in confluent monolayers of human foreskin fibroblast (HFF) cells as previously described (53). Trypanosoma brucei procyclic forms were cultured at 28°C in SDM-79 medium supplemented with 10% fetal calf serum (54).
Generation of fusion protein lines.
To generate in-frame C-terminal chimeras with proteins of interest, regions of TGGT1_040920 (TgSAS-6), TGGT1_059860 (TgSAS6L), Tb927.10.7920 (TbSAS6L), and Tb11.02.5550 (TbWDR16) were amplified and cloned into LIC (52), pLew-MH-TAP-eYFP (55), and pEnT6P-Y (56) vectors by standard methods. TgSAS-6 and TgSAS6L constructs in the pYFP.LIC.DHFR vector were transfected into ku80-null Toxoplasma parasites as previously described, and stable lines were isolated by selection in 1 μM pyrimethamine and single cell cloned (52). The coding sequence of TgSAS6L was validated by amplification of cDNA encoding the complete open reading frame with primers described in Table 1. A pre-existing GRASP55-YFP construct (57) was modified to replace the GRASP55 coding sequence with SAS6L to drive the expression of TgSAS6L-YFP from the α1-tubulin promoter, and stable lines were isolated by selection for chloramphenicol resistance. Linearized plasmid DNA was used to transfect exponentially growing cultures of procyclic-form T. brucei by electroporation (three 100-μs pulses of 4.25 kV/cm). Transfected cells were selected by the addition of 1 μg/ml puromycin (pEnT6P-TbSAS6L1-Y) or 10 μg/ml blasticidin (pEnT6B-WDR16-mCherry). Inducible overexpression of TbSAS6L was achieved by cloning TbSAS6L into the pLew-MH-TAP-eYFP vector with primers described in Table 1. Transfectants were selected in 5 μg/ml phleomycin, and overexpression of TbSAS6L1-YFP was induced by the addition of 1 μg/ml doxycycline to the culture medium.
Primer pairs used for molecular biology constructs in this study
Primer
Sequence
TbSAS6L CDS pEnT6P 5′
GTTGCGGCCGCAAAGAACCCCAACGCTTTTT
TbSAS6L CDS pEnT6P 3′
GTCTACTAGTTCGCGTTTTTCCAAGTGTG
TbSAS6L UTR pEnT6P 5′
GTTAAGCTTGAGTGCAAATTCTTCAGTGCG
TbSAS6L UTR pEnT6P 3′
GTTGCGGCCGCCGACATTAGCTCCGGTTTCT
WDR16 CDS pEnT6P 5′
CGTTCTAGAGCAGCGAAGGAGTACGAAA
WDR16 CDS pEnT6P 3′
TCGAGCACGGCCAGGTAGCTGAT
WDR16 UTR pEnT6P 5′
GCTCGAGCTTTCGCAAACCCATTCG
WDR16 UTR pEnT6P 3′
GTAGGATCCAACGGAAGGGTGGCAGTT
5′ TbSAS6L HindIII pLew
ATCAAGCTTATGGATCGCATAGAAATATACTACCAG
3′ TbSAS6L1 XhoI pLew
GATCTCGAGTCGCGTTTTTCCAAGTGTGACAGTAGC
SAS6L LIC vector 5′
TACTTCCAATCCAATTTAATGCACAGACGGAAATGCTCTCC
SAS6L LIC vector 3′
TCCTCCACTTCCAATTTTAGCGAGGAACCGAGTGGATGC
SAS6L MBP plasmid 5′
ATGGCGACAAACTTCGGCTTTGG
SAS6L MBP plasmid 3′
TATAAAGCTTTCAGAGGAACCGAGTGGATGCGCC
ptub-SAS6L-YFP 5′
CAAAGATCTATGGCGACAAACTTCGGCTTTGG
ptub-SAS6L-YFP 3′
TATACCTAGGGAGGAACCGAGTGGATGCGCC
SAS6L KO 5′ flank 5′
AAGGTACCCCCTCTCTTCACAGTCGAAGACC
SAS6L KO 5′ flank 3′
TTGGGCCCGTTATTCTGTTCGAACCCGGGG
SAS6L KO 3′ flank 5′
AAACTAGTCGGTGGGAGGTCTCAAGCG
SAS6L KO 3′ flank 3′
GCGCGGCCGCCACGTAATCACACAATCCGAGCGTATAG
Toxoplasma SAS6L gene knockout.
We constructed a knockout vector with the pmini.GFP.HPT plasmid (58). A 4.9-kb region upstream of the TGGT1_059860 coding sequence and a 5-kb downstream region were amplified with primers listed in Table 1. These inserts were cloned upstream and downstream of the hypoxanthine xanthine guanine phosphoribosyl transferase (HPT) gene. We electroporated ku80Δ hpt mutant RH tachyzoites with the construct and selected for transformants in 50 μg/ml mycophenolic acid and 50 μg/ml xanthine. Single-knockout clones were isolated from two independent transfections by single-cell cloning. We also isolated green fluorescent protein (GFP)-expressing transformants with a nonhomologous integration of the knockout vector. These were used in competition assays with the knockout line (see below).
Escherichia coli expression and purification of TgSAS6L.
The cDNA sequence of TgSAS6L was inserted into the c2x pMAL vector (NEB) in order to express a maltose-binding protein (MBP)-TgSAS6L fusion in E. coli. After induction for 2 h with 300 μM isopropyl-β-d-thiogalactopyranoside (IPTG), the protein was purified on an amylose column and cleaved away from MBP with Factor Xa. Dialyzed TgSAS6L was used as an immunogen to generate mouse polyclonal antiserum.
Toxoplasma conoid extrusion.
Conoid extrusion was induced in freshly lysed extracellular parasites by treatment with HEPES-buffered saline supplemented with 5 μM ionomycin (Sigma) and 5 mM CaCl2 (Sigma) (36–38), and samples were fixed as previously described (31).
Immunofluorescence staining and fluorescence microscopy.
YFP- or mCherry-tagged intracellular Toxoplasma parasites were fixed, permeabilized, and stained as previously described (27). Extracellular parasites were filtered through a 3-μm polycarbonate filter (GE Water & Process Technologies), centrifuged at 1,000 × g for 20 min at 4°C, and suspended in a small volume of phosphate-buffered saline (PBS). Detergent-extracted parasites were generated and stained as previously described (31). The antibodies used for immunofluorescence assays included a mouse polyclonal serum against TgSAS6L, a mouse monoclonal antibody against GFP (Roche), and a rabbit Toxoplasma-specific tubulin serum (27) detected with Alexa 594-, Alexa 488-, and Cascade blue-conjugated secondary antibodies (Invitrogen). DNA was visualized by 4′,6-diamidino-2-phenylindole (DAPI) staining. Toxoplasma samples were imaged on a Zeiss Axiovert 200M microscope with the AxioVision system. Trypanosomes were settled onto glass slides and either fixed with 1% formaldehyde for whole-cell preparations or extracted with 1% Nonidet P-40 in PEME [100 mM 100 mM piperazine-N,N′-bis(2-ethanesulfonic acid) (PIPES, pH 6.9), 2 mM EGTA, 1 mM MgSO4, 0.1 mM EDTA] for cytoskeleton preparations. Cells were labeled with BBA4 or YL1/2 antibodies, which were visualized with anti-mouse IgM–Cy5 and anti-rat IgG–tetramethyl rhodamine isothiocyanate (Jackson ImmunoResearch). Cells were mounted in 90% glycerol–50 mM sodium phosphate (pH 8) supplemented with 1% 1,4-diazabicyclo[2.2.2]octane and 0.4 μg ml−1 DAPI.
Immunoblot assays.
Protein lysates containing ∼5 × 106Toxoplasma tachyzoites per lane were resolved by 15% SDS-acrylamide gel electrophoresis and transferred to nitrocellulose. After blocking for 1 h in 5% milk in PBS-Tween, the blot was probed with primary antibodies (1:300 anti-TgSAS6L and 1:750 anti-ISP3 in 5% milk) for 1 h, washed in PBS-Tween, probed with a secondary antibody (1:2,000 horseradish peroxidase-conjugated goat anti-mouse antibody; Invitrogen) for 1 h, and washed in PBS-Tween prior to detection by chemiluminescence assay.
Electron microscopy (EM).
Freshly lysed Toxoplasma tachyzoites from stable TgSAS6L-YFP lines were isolated by filtration. Parasites were fixed and cryosectioned as previously described (31), with an anti-GFP antibody (Abcam). Trypanosomes were settled onto Formvar-coated nickel finder grids and extracted with 1% NP-40–PEME. For Ca2+ preparations, cells were additionally incubated with 65 mM Ca2+–PIPES, pH 6.9. Preparations were fixed with 2.5% glutaraldehyde–PEME for 10 min and then neutralized with 1% glycine–PEME. Grids were mounted in H2O, and native fluorescence was analyzed with a Leica DM5500 microscope. Grids were negatively staining with 1% aurothioglucose (USP) and viewed on an FEI Tecnai-F12 electron microscope.
Toxoplasma competition assays.
T25 flasks with confluent HFF cells were inoculated with a 1:1 ratio of sas6l-null parasites and GFP-expressing clones from the same transfection that harbor a nonhomologous integration of the knockout vector (107 parasites of each line) as described in a previously established competition assay (59). The relative numbers of GFP-positive (control) and GFP-negative (sas6l-null) parasites were determined by flow cytometry (Apogee Flow Systems).
Alignments and phylogenetic analysis.
SAS6 and SAS-6-like proteins form a natural set on the basis of a BLASTp clustering approach (60). This clustering results from the presence of a conserved domain of ∼200 aa covered by a Pfam-B domain (Pfam_B_2528) of unknown function. To define the family of SAS6 and SAS-6-like proteins, the hidden Markov model defining Pfam_B_2528 was used to search the complete predicted proteomes of 44 diverse eukaryotes (HMMER3.0; hmmer.janelia.org). All 69 proteins containing the domain (e-value, <0.001) were extracted, trimmed to the conserved domain, and aligned by MAFFT v6.811b by adopting the E-INS-i strategy (61). Bayesian phylogenies were inferred from four runs of the metropolis-coupled Markov chain Monte Carlo method implemented in the program MrBayes3.1.2 (62) (WAG substitution matrix; GTR+Γ+I with Γ distribution approximated to four discrete categories and shape parameters estimated from the data). Maximum-likelihood support for the inferred topology was generated from 100 bootstrap replicates of the data with PhyML3.0 (63) (LG matrix; GTR+Γ+I, as described above).
RESULTSSAS-6 and SAS6L localize to distinct structures in Toxoplasma tachyzoites.
SAS-6 is an essential and universal component of centrioles that creates the “cartwheel hub” found at the proximal end of the centriole barrel (41, 42, 45). We came upon the hypothetical protein TGGT1_059860 during a genome database search (www.ToxoDB.org) for the Toxoplasma homolog of SAS-6. Since TGGT1_059860 has similarity to SAS-6, we named it SAS-6-like (SAS6L). When the Toxoplasma SAS-6 homolog (TGGT1_040920) was tagged by fusion of the C terminus to YFP, the label localized exclusively to the centriole, coincident with the centriolar marker centrin (Fig. 1A), as is expected from the localization of SAS-6 in other systems. In contrast, TgSAS6L localizes exclusively to a region at the apex of the parasites (Fig. 1B). Since TgSAS6L labeled a small spot in the apex of Toxoplasma tachyzoites, we induced conoid extrusion in extracellular parasites and used a marker for the APR (RNG1) (31) to localize TgSAS6L more accurately. TgSAS6L is found at the apical tip of extended conoids, above the RNG1 signal (Fig. 1C). In parasites with retracted conoids, labeling is coincident with the RNG1 marker (Fig. 1D). TgSAS6L remains associated with the cytoskeleton following treatment with deoxycholate (DOC), which extracts much of the inner membrane complex-associated cytoskeleton and frees the subpellicular microtubules. In DOC-extracted samples, TgSAS6L labeling is at the apical tip of the conoid (Fig. 1E). Immunogold labeling of TgSAS6L-YFP parasites with an anti-GFP antibody indicates that the protein is located at the apex of the conoid, in a region consistent with the preconoidal rings (Fig. 1F).
TgSAS6 and TgSAS6L localize to distinct regions in Toxoplasma tachyzoites. (A) The Toxoplasma homolog of SAS6 (green) localizes to centrioles along with the centriole marker centrin (red). The overlap between these proteins is not complete, which likely reflects distinct localization within centrioles; while centrin is typically located throughout the centriole, SAS-6 is a marker of the proximal end of centrioles. (B) TgSAS6L (red, anti-SAS6L antiserum) is found at the apex of Toxoplasma tachyzoites and does not colocalize with SAS-6 in the centrioles (green, YFP tag). (C) Ionomycin triggers conoid extrusion in most extracellular parasites. In these cells, SAS6L-mCherry labeling (red) is above the APR (RNG1-YFP, green). (D) When the conoid is retracted, TgSAS6L and RNG1-YFP partially colocalize (the preconoidal region is smaller than the APR). (E) Extraction of extracellular tachyzoites with DOC frees the subpellicular microtubules and conoid (blue) and APR (RNG1-YFP, green) from other cellular material. TgSAS6L is located at the apical end of the conoid, and RNG1-YFP is beneath the extended conoid at the APR. The tubulin channel alone permits identification of the conoid (arrows) above the subpellicular microtubules. (F) Immunogold labeling indicates that SAS6L-YFP localizes to the extreme apical region of Toxoplasma tachyzoites, above the conoid in both longitudinal (arrows, top) and transverse (arrows, bottom) sections of the apical region. Scale bars, 2 μm (A to E) or 250 nm (F).
SAS6L-null Toxoplasma are less fit.
We deleted the SAS6L gene from Toxoplasma tachyzoites by targeted integration of the HPT gene (Fig. 2A) and confirmed the knockout by the absence of SAS6L protein (Fig. 2B and C). Since there are no specific markers for the conoid fibers, we induced conoid extrusion and stained with a tubulin antibody to ascertain that the conoid is still present in null parasites (not shown). TgICMAP1 is a microtubule-associated protein that specifically localizes to the two intraconoid microtubules that run down the center of the conoid (64). We used an established expression construct (pmin-eGFP-TgICMAP1 [64]) to verify that the eGFP-TgICMAP1 fusion protein correctly localizes to the region of the intraconoid microtubules in the absence of SAS6L (not shown).
Loss of TgSAS6L reduces Toxoplasma fitness. (A) Knockouts were generated by homologous recombination of the HPT gene into the TgSAS6L locus. In the case that the vector integrates nonhomologously, both the GFP and HPT genes are integrated, generating fluorescent parasites. (B) sas6lΔ mutants were confirmed by loss of immunofluorescence labeling for TgSAS6L (red). All parasites (1 and 2) are labeled for tubulin (green). Nonhomologous integration of the knockout vector (parasite 1) leads to parasites that retain TgSAS6L and also have the GFP signal. (C) Immunoblotting with both anti-TgSAS6L and anti-ISP3 antisera also demonstrates the complete absence of TgSAS6L from sas6lΔ mutant knockout (KO) cells relative to the parental line (P). The values to the left are molecular sizes in kilodaltons. (D) Two independently derived sas6l-null parasite lines have reduced growth relative to that of GFP-expressing clones with a nonhomologous integration of the knockout vector.
During creation of the sas6lΔ mutant lines, we also cloned lines that had nonhomologous integration of the knockout vector for use as a matched control in growth competition assays. We coinfected host cells with an equivalent number of each clonal line and followed the relative rates of replication and growth over serial passage by flow cytometry (Fig. 2D). Two independent knockout lines were outcompeted by a matched control at a rate of 0.07 day−1, demonstrating a reduction in fitness. The reduced sas6lΔ mutant parasite growth observed may reflect a specific defect in a number of cellular processes required for invasion and intracellular growth (for example, motility, attachment, invasion, doubling time, replication fidelity, egress, and extracellular viability). In order to assess specific defects in these processes, we measured gliding motility, performed plaque formation assays (for invasion and extracellular survival), assessed Ca2+ ionophore-mediated egress, and quantified replication rates and the frequency of replication defects. Within the resolution of these assays, no significant differences between two independent sas6l-null lines and the matched control were detected (data not shown). However, the expected size of a specific defect in assays such as the plaque assay (∼10%, given the rate of outcompetition) was below the assay's resolving power because of the relative variation between biological replicates. Hence, it is likely that the direct fitness assay detected a defect that was below the sensitivity of the other assays in our experiments.
SAS6L is an ancient protein found in diverse simple eukaryotes.
Similarity searches suggested the presence of SAS6L homologs in diverse eukaryotes. All SAS6L proteins contain an ∼200-amino-acid (aa) conserved domain that is shared with SAS-6 homologs (see Fig. S1 in the supplemental material) and includes the ∼50-aa PISA motif previously described in SAS-6 (42). This domain is described by the automatically generated Pfam-B domain, Pfam_B_2528. We used Pfam_B_2528 to identify proteins of the SAS-6/SAS6L family encoded by the genomes of 29 of 44 diverse eukaryotes (Fig. 3A). These proteins share an average of 24% identity (46% similarity) across the conserved domain. A similar set was defined by an alternative approach based on clustering by BLASTp score (data not shown).
SAS-6 and SAS6L proteins are found in a subset of organisms that form cilia or flagella. (A) Presence (filled circles) or absence (empty circles) of identifiable homologues of SAS-6 and SAS6L in the predicted proteomes of 44 diverse eukaryotes. SAS6L protein is present in most organisms that also possess SAS-6 but absent from the Holozoa. (B) The SAS-6 and SAS6L protein families share a conserved domain of ∼200 aa but can be robustly separated by phylogenetic analysis and have distinct architectures. A Bayesian phylogeny of the conserved domain is shown with topology support from Bayesian posterior probabilities and maximum likelihood bootstrap replicates (PP/BS). Architectures show predictions of Pfam-A and -B domains (89) and coiled-coil regions (90).
Phylogenetic analysis based on the conserved domain robustly supported the existence of two subfamilies comprising SAS-6 and SAS6L (Fig. 3B). Genes encoding SAS-6 are found in the genomes of eukaryotes that are known to build a centriole or basal body and also in the microsporidian Encephalitozoon cuniculi and Ostreococcus tauri as previously described (42–45, 60). With the exception of Giardia lamblia, which lacks an identifiable SAS-6 ortholog (60), SAS6L is found only in lineages that also contain SAS-6 (Fig. 3A). In addition to Toxoplasma, SAS6L is found in multiple protistan lineages (including other apicomplexans, such as Plasmodium and Cryptosporidium species). However, SAS6L appears to be absent from metazoan and choanoflagellate genomes (Holozoa). The presence of SAS6L in a basal flagellate fungus (Batrachochytrium dendrobatidis) indicates that the absence of SAS6L from animals is likely the result of a specific loss near the root of Holozoa.
Although only the conserved domain was used to define the phylogenetic families, the SAS-6 and SAS6L proteins have clearly distinct architectures (Fig. 3B). SAS-6 sequences are generally ∼600 aa in length and comprise the conserved domain near the N terminus, followed by ∼300 aa of sequence rich in predicted coiled coils. In contrast, SAS6L sequences contain little sequence outside the conserved domain and appear to lack the C-terminal tail entirely. The conserved domain essentially matches the globular head domain of SAS-6, the structure of which has been solved (50, 51). Homology modeling suggests that SAS6L proteins can adopt structures very similar to those seen in the head of SAS-6 (see Fig. S2 in the supplemental material). The apicomplexan SAS-6 proteins are unusual in possessing long N-terminal extensions before the conserved domain (Fig. 3B).
TbSAS6L localizes to the axonemal basal plate in trypanosomes.
Antiserum against TgSAS6L labels a spot at the apex of the related parasite Neospora caninum (not shown). Unfortunately, our antiserum does not label SAS6L proteins in more distantly related organisms, such as Tetrahymena thermophila (data not shown). Since the majority of SAS6L-producing eukaryotes do not have a conoid, we tested the localization of a SAS6L ortholog in a distantly related organism. In Trypanosoma brucei trypomastigotes constitutively expressing TbSAS6L-YFP from the endogenous locus, TbSAS6L-YFP localizes near the base of the flagellum (Fig. 4A to C). Label appears as one or two prominent dots, depending on the cell cycle stage, and there is also weaker labeling in the proximal region of the axoneme, in the region of the flagellar pocket collar. TbSAS6L-YFP labeling is distal to markers of the proximal (BBA4), middle (tyrosinated α-tubulin, YL1/2), and distal (WDR16) parts of the basal body. Quantification of the location of TbSAS6L-YFP relative to these markers indicates that this protein is located in a fixed position within the axoneme (Fig. 4D), in a region consistent with the location of the basal plate (the area where the central-pair microtubules of a flagellar axoneme are nucleated). Ablation of TbSAS6L by RNA interference does not alter the overall morphology of the flagellum, as viewed by EM, suggesting that SAS6L has a nonessential role in flagellum formation (not shown). This is not entirely surprising, given the plasticity of flagellum formation, but it is also, of course, extremely difficult to detect subtle alterations in this complex structure.
TbSAS6L-YFP localizes to a prominent focus near the base of the flagellum in T. brucei. (A) TbSAS6L-YFP (green) is distal to the basal body, which is marked by tyrosinated α-tubulin (YL1/2, red, arrowheads). A lower-intensity SAS6L-YFP signal is also associated with the developing probasal body. SAS6L-YFP is also distal to the basal-body markers BBA4 (B) and WDR16 (C) (arrowheads). (D) Distance measurements in relation to YL1/2 indicate that SAS6L-YFP localizes to a region near the axoneme basal plate (20 measurements for each marker). Scale bars, 5 μm.
Overexpression of SAS6L-YFP causes ectopic filament formation.
To investigate the in vivo properties of SAS6L, we expressed tagged SAS6L in both Toxoplasma and Trypanosoma over endogenous protein. In Toxoplasma, when TgSAS6L-YFP expression was driven from the α1-tubulin promoter, excess protein assembled into filaments in a subset of the sibling Toxoplasma parasites within a single parasitophorous vacuole (Fig. 5A). Filaments sometimes deformed the parasite's shape and may interfere with cell division, as in the parasites shown in Fig. 5B, where a large filament extends between daughters that have begun a new round of replication without completing scission. TgSAS6L-YFP filaments form when intracellular parasites are treated with oryzalin, which inhibits microtubule polymerization (not shown). Ectopic TgSAS6L-YFP filaments are retained after extraction with 1% Triton X-100 but are sensitive to 10 mg/ml DOC, which removed TgSAS6L-YFP filaments without disrupting the apical TgSAS6L signal (Fig. 5C). Immunogold staining of EM cryosections of whole cells showed that TgSAS6L-YFP was associated with continuous electron-dense structures in the cytoplasm (Fig. 5D).
Overexpression of SAS6L-YFP leads to filament formation. (A) Intracellular Toxoplasma tachyzoites show prominent TgSAS6L-YFP filaments as follows: Toxoplasma tubulin, red; SAS6L-YFP, green; DNA, blue. (B) TgSAS6L-YFP filaments appear to interfere with the completion of division by some tachyzoites. (C) SAS6L-YFP filaments are insensitive to 1% Triton X-100 (TX-100), but 10 mg/ml DOC causes filament depolymerization although apical SAS6L is retained. (D) EM of immunogold-labeled cryosections of Toxoplasma overexpressing TgSAS6L-YFP show electron-dense structures of variable width within the cytoplasm. (E) Overexpression of the TbSAS6L protein in trypanosomes also leads to ectopic filaments. (F) Correlative light microscopy and EM of TbSAS6L-YFP filaments in trypanosomes (arrowheads). The same cell cytoskeleton is shown by phase-contrast microscopy (F1) and negatively stained transmission EM (F2 to -4). (G) Correlative microscopy of Ca2+-treated cytoskeleton preparations shows that TbSAS6L-YFP filaments are resistant to Ca2+ (arrows). Scale bars: A and B, 2 μm; D, 250 nm; E, 2 μm; F4, 200 nm; G4 inset, 200 nm.
Overexpression of TbSAS6L-YFP in trypanosomes also resulted in the accumulation of ectopic filaments (Fig. 5E). The formation of these structures has no effect on the morphology or duplication time of the cells (data not shown). The filaments are insensitive to extraction with nonionic detergent (Nonidet P-40) and are not labeled with anti-tubulin antibodies (not shown), indicating that they are independent of microtubules. TbSAS6L-YFP filaments are insensitive to treatment with Ca2+, which depolymerizes T. brucei subpellicular microtubules (Fig. 5G). Correlative microscopy of detergent-extracted trypomastigotes was used to unambiguously determine the ultrastructure of TbSAS6L-YFP filaments by first identifying filaments by fluorescence microscopy and subsequently imaging them by EM (Fig. 5F and G). TbSAS6L filaments are ∼20 nm wide but have a structure distinct from that of microtubules (Fig. 5G).
DISCUSSION
We describe a novel conserved protein that is related to the centriole protein SAS-6, which we have hence named the SAS-6-like (SAS6L) protein. The existence of SAS6L produced by diverse organisms suggests that this protein appeared early in eukaryote evolution and most likely before the last common ancestor of all extant eukaryotes (65). SAS6L is found in simple eukaryotes that contain SAS-6 and centrioles/basal bodies, but the protein has been lost from the holozoan lineage. SAS6L proteins consist of a conserved domain that also comprises the N-terminal head domain of SAS-6. Recent data have shown that SAS-6 forms homodimers through interactions between their coiled-coil tail domains and that these homodimers interact via other parts of the head domain to form higher-order structures (in particular, rings) (50, 51). Since SAS6L proteins lack a SAS-6 tail, they cannot dimerize by this means. However, the opposite face of the protein is available for interaction and is predicted to have a structure very similar to that of SAS-6. The observation that overexpression of SAS6L proteins induces SAS6L filament formation indicates that SAS6L proteins do, indeed, have the ability to form higher-order structures. However, the distinct localization of SAS-6 and SAS6L proteins in Toxoplasma and Trypanosoma suggests that these proteins do not form heterodimers in vivo. The decreased stability of the ectopic SAS6L filaments after DOC extraction suggests that additional interactions stabilize the native preconoidal ring structure and do not occur in the cytoplasmic filaments.
Consistent with its location in other diverse eukaryotes, SAS-6 localizes to the basal body in Trypanosoma and the centriole Toxoplasma tachyzoites. In contrast, SAS6L localizes to the region of the basal plate in Trypanosoma brucei—the electron-dense region at the distal end of the transition zone from which the central-pair microtubules are nucleated. Unlike T. brucei (where flagella are present in all life cycle forms), only the microgamete forms of Toxoplasma or other apicomplexan organisms build flagella (24, 32, 66–69). Comparative genomics indicate that the Toxoplasma and Plasmodium genomes lack a surprising number of conserved basal-body components (60), and while studies of Plasmodium microgamete development show de novo formation of basal bodies (66), it is unclear whether Toxoplasma microgamete basal bodies form de novo or if the atypical centriole found in the asexual tachyzoite stage serves as a template for basal-body formation. We anticipate that SAS6L will also localize to the basal plate region in Toxoplasma microgametes, but because of the difficulty in obtaining this stage from cat intestinal epithelium, we cannot verify this prediction. In agreement with this hypothesis, there is evidence of upregulation of the Plasmodium falciparum ortholog of SAS6L (PF3D7_1316400) in gametocyte development (70).
The Toxoplasma tachyzoite stage causes most of the pathology associated with parasite infection. Despite lacking flagella, tachyzoites are motile and employ an unusual substrate-dependent gliding motility that requires actin and myosin (71–73). The apical region of tachyzoites is vital to host cell invasion and contains specialized secretory organelles (micronemes and rhoptries) and a complex and unusual organization of the microtubule cytoskeleton (28, 39, 74, 75). The elongated cell shape of tachyzoites is maintained by a corset of 22 subpellicular microtubules that radiate out of the APR. An additional tubulin-based structure, the conoid, can be extended through the APR or retracted into it. In intracellular parasites, the conoid is retracted and immotile, while in extracellular parasites, the conoid can extend and retract as tachyzoites glide and invade host cells (36). Extension of the conoid is associated with the elongation and narrowing of the apical end of tachyzoites, and it is believed that the conoid has a mechanical role in parasite invasion of host cells.
Apicomplexans share ancestry with ciliates and dinoflagellates; collectively, these organisms are classified as alveolates (76, 77). It is likely that the last common ancestor of apicomplexans and dinoflagellates had an open-sided “incomplete” conoid that was modified into a closed conoid in apicomplexans and lost from dinoflagellates (77, 78). Within the Apicomplexa, the conoid was again lost from noncoccidian lineages. Although “true” conoid structures are found only in coccidian apicomplexans, alveolates such as Colpodella vorax and Rastrimonas subtilis (previously known as Cryptophagus subtilis) have “pseudoconoid” or “incomplete conoid” structures consisting of a set of 5 to 14 interlinked microtubules located at the cell apex but lack the adjacent APR structure observed in coccidians (78–81). Colpodella and Rastrimonas also contain apical secretory organelles reminiscent of apicomplexan micronemes and rhoptries that are used along with a pseudoconoid to partially or completely invade other unicellular protists. Colpodella attaches to and aspirates the cytoplasmic contents of unicellular flagellates (76, 78, 80, 81), while Rastrimonas is an intracellular parasite of unicellular algae (79, 80). The precise phylogenetic position of Colpodella and Rastrimonas is unclear, as these alveolates have been variously described as dinoflagellates, as early-branching apicomplexans, as Perkinsozoa, or within sister lineages of dinoflagellates or apicomplexans (76, 78, 80, 81). What is clear is that these organisms are poised between predator and parasite states and provide tantalizing clues to the evolution of the intracellular apicomplexan lifestyle. Both Colpodella and Rastrimonas simultaneously build a pseudoconoid structure and adjacent flagella (Fig. 6A and B). In contrast, the flagellar axoneme and “true” conoid are mutually exclusive structures in coccidian parasites; microgametes have apically located basal bodies that template flagellar axonemes, while asexual proliferative-stage coccidians have juxtanuclear centrioles and apical conoid structures (Fig. 6A and B). This study describes SAS6L, a novel conserved protein that is located in the region of the basal plate of the axoneme in T. brucei and above the conoid in Toxoplasma tachyzoites, a life cycle stage that lacks flagella (Fig. 6C). Rastrimonas absorbs its flagella during intracellular growth, and this feature may have become more pronounced during the evolution of apicomplexan parasites, ultimately leading to the loss of flagella in all stages other than microgametes. Our results emphasize this relationship between the conoid and flagellum, providing molecular evidence of a link between these organelles.
The Toxoplasma conoid incorporates flagellar components, and its appearance and that of flagella are mutually exclusive during the parasite life cycle. (A) Toxoplasma tachyzoites (part 1) have juxtanuclear centrioles and a specialized apical region used to invade host cells. Rhoptries (red) secrete components used during invasion, while the conoid (green) can extend from the APR to cause elongation of this region. Toxoplasma tachyzoites have a closed conoid (CC) that is topped by preconoidal rings (PCR), the location of SAS6L (orange). Toxoplasma microgametes (part 2) lack a conoid but have apically located basal bodies (BB) that template two flagella that extend toward the posterior. Alveolate organisms such as Colpodella and Rastrimonas (part 3) contain rhoptry-like secretory organelles (red) and an incomplete conoid (IC) structure consisting of a set of interlinked microtubules at the cell apex. They build flagella from adjacent basal bodies (BB) but lack the conoid-associated APR structure observed in coccidians. (B) The Toxoplasma tachyzoite conoid consists of 10 to 14 tubulin-containing filaments that spiral to form a closed cone that is narrower at the apex than at the base (box 1). The conoid may be extended from or retracted into the APR, which also serves to nucleate 22 subpellicular microtubules (not shown). The SAS6L protein is located at the extreme tip of the conoid (orange) in the region of the PCR. The juxtanuclear centrioles (box 2) are located at a distance from the conoid in tachyzoites. Toxoplasma microgametes (box 3) have apical centrioles that build a transition zone (TZ) and template flagellar axonemes. We predict that SAS6L localizes to the basal plate (BP) in this stage (orange). Alveolates such as Colpodella (box 4) have apical secretory organelles (red) and an incomplete conoid (IC) structure (green) that does not fully encircle the rhoptry necks and may be as simple as a set of four to six interlinked microtubules. Adjacent apical centrioles build a transition zone (TZ) to template flagellar axonemes with a basal-plate (BP) structure. These organisms lack the conoid-associated APR structure observed in Toxoplasma. (C, left) The basal body (BB) templates the flagellar axoneme (AX). The transition from the triplet microtubule organization of the basal body to the doublet microtubule organization of the axoneme occurs at the basal plate (BP), which is also where the central-pair microtubules of the axoneme begin. SAS6L in trypanosomes localizes to the region of the flagellar plate (orange). (C, right) The Toxoplasma conoid is formed from 10 to 14 C-shaped and spiraling tubulin filaments. Two 13-protofilament microtubules extend through the center of the conoid, reminiscent of the central-pair microtubules of the axoneme. The conoid ends in two preconoidal rings, and Toxoplasma SAS6L localizes to this region (orange). The conoid can be extended beyond or retracted into the APR (red), which serves as an MTOC for the subpellicular microtubules.
The observations presented here are reinforced by recent work on the basal-body-associated protein SFA (striated fiber assemblin) in Toxoplasma. SFA proteins were first described in green algae, where flagellar basal bodies are embedded in a “cage” of flagellar rootlet fibers. Rootlets consist of distinct populations of centrin-containing contractile fibers (82), sinister fibers (83), and SFA-containing noncontractile striated fibers (84, 85). Genes for SFA homologs are found in apicomplexan genomes (86, 87). Remarkably, two SFA proteins (TgSFA2 and TgSFA3) form a filament that links the upper edge of the conoid to the juxtanuclear centrioles during tachyzoite replication (88). The SFAs appear shortly after centriole duplication (87), colocalizing in a fiber-like structure that emerges from the centrioles and extends apically to tether the conoid and APR to the centrioles (88). Loss of either SFA protein inhibits parasite replication by blocking the formation of daughter APRs and the subsequent construction of daughter buds. Nuclear division is not inhibited, and multiple nuclei accumulate in these aberrant parasites. These results suggest that although the asexual stages of apicomplexans have dispensed with building flagella, they retain SFAs to ensure the reliable inheritance of apical organelles.
Given the observation that the SFAs and SAS6L are retained in aflagellate tachyzoites, an intriguing possibility is that SAS6L is one of the components that tether the SFA filament to the conoid. SAS6L cannot be the sole factor anchoring SFA to the apex, as our results indicate that SAS6L is not essential (although its loss is associated with a fitness defect) while loss of either SFA2 or SFA3 leads to a lethal arrest at the time of daughter cell budding. In light of our data on SAS6L and the observations on SFA2 and SFA3, we propose that the complete conoid evolved from an ancestral pseudoconoid in part by the incorporation of components from a vestigial apical flagellar apparatus. These components may have been from flagellum-associated structures, such as flagellar rootlets (as seen for the small portion of Trypanosoma SAS6L that is associated with the flagellar collar region), or could represent the direct subsumption of the flagellum into the conoid. The former premise suggests that flagellar components are retained because they are required for an essential cellular process, such as replication. In this context, SAS6L may have been retained along with the basal-body-associated SFA proteins in order to coordinate organelle inheritance in aflagellar tachyzoites (88). The latter hypothesis implies direct homology between the central-pair microtubules of the axoneme and the microtubule pair at the center of the conoid and suggests that other orthologs of basal-plate and central-apparatus components may play a role in invasion or replication by coccidia.
Supplementary Material
Supplemental material
Published ahead of print 17 May 2013
Supplemental material for this article may be found at http://dx.doi.org/10.1128/EC.00096-13.
ACKNOWLEDGMENTS
We thank Catherine Li and Ni Yao for technical assistance, David Coder for guidance with flow cytometry, and David Krueger for advice on Illustrator. Ke Hu (Indiana University) kindly provided us with the pmin-eGFP-TgICMAP1 construct. David Ferguson (Oxford) was a critically important sounding board for ideas about the apicomplexan basal body, and Susan Dutcher (Washington University) critically commented on the manuscript.
N.S. was supported by a graduate studentship from the EPA Trust, and research was funded by NIH grant AI067981 to N.S.M., a Wellcome Trust grant to K.G., and BBSRC New Investigator grant BB/J01477X/1 to B.W.
REFERENCESDutcherSK2004Dissection of basal body and centriole function in the unicellular green alga Chlamydomonas reinhardtii,
p 71–92InNiggEA (ed),
.
Wiley-VCH, Weinheim, GermanyAzimzadehJMarshallWF2010Building the centriole.
20:
R816–
R82520869612MarshallWF2009Centriole evolution.
21:
14–
1919196504KobayashiTDynlachtBD2011Regulating the transition from centriole to basal body.
193:
435–
44421536747AzimzadehJBornensM2007Structure and duplication of the centrosome.
120:
2139–
214217591686BornensM2008Organelle positioning and cell polarity.
9:
874–
88618946476PaolettiABornensM1997Organisation and functional regulation of the centrosome in animal cells.
3:
285–
2999552423ChapmanMJDolanMFMargulisL2000Centrioles and kinetosomes: form, function, and evolution.
75:
409–
42911125699SchrevelJBesseC1975A functional flagella with a 6 + 0 pattern.
66:
492–
507169268PrebleAMGiddingsTMJrDutcherSK2000Basal bodies and centrioles: their function and structure.
49:
207–
23311005020WatersJCSalmonE1997Pathways of spindle assembly.
9:
37–
439013671MarshallWFVucicaYRosenbaumJL2001Kinetics and regulation of de novo centriole assembly. Implications for the mechanism of centriole duplication.
11:
308–
31711267867BastoRLauJVinogradovaTGardiolAWoodsCGKhodjakovARaffJW2006Flies without centrioles.
125:
1375–
138616814722HinchcliffeEHMillerFJChamMKhodjakovASluderG2001Requirement of a centrosomal activity for cell cycle progression through G1 into S phase.
291:
1547–
155011222860PielMNordbergJEuteneuerUBornensM2001Centrosome-dependent exit of cytokinesis in animal cells.
291:
1550–
155311222861MignotJ-P1996The centrosomal big bang: from a unique central organelle towards a constellation of MTOCs.
86:
81–
91DespommierDRGHotezPKnirschC2005Parasitic diseases, fifth edition.
Apple Trees Productions, LLC,
New York, NYKohlLGullK1998Molecular architecture of the trypanosome cytoskeleton.
93:
1–
99662023BriggsLJDavidgeJAWicksteadBGingerMLGullK2004More than one way to build a flagellum: comparative genomics of parasitic protozoa.
14:
R611–
R61215296774GullKBriggsLVaughanS2004Basal bodies and microtubule organization in pathogenic protozoa,
p 401–423InNiggEA (ed),
.
Wiley-VCH,
Weinheim, GermanyGullK1999The cytoskeleton of trypanosomatid parasites.
53:
629–
65510547703SeebackTHemphillALawsonD1990The cytoskeleton of trypanosomes.
6:
49–
5215463295ScottVSherwinTGullK1997Gamma-tubulin in trypanosomes: molecular characterisation and localisation to multiple and diverse microtubule organising centres.
110(Pt 2):157–1689044046FergusonDJPHutchisonWMDunachieJFSiimJC1974Ultrastructural study of early stages of asexual multiplication and microgametogony of Toxoplasma gondii in the small intestine of the cat.
82B:
167–
1814528099FrixioneEMondragonRMezaI1996Kinematic analysis of Toxoplasma gondii motility.
34:
152–
1638769726HåkanssonSMorisakiHHeuserJSibleyLD1999Time-lapse video microscopy of gliding motility in Toxoplasma gondii reveals a novel, biphasic mechanism of cell locomotion.
10:
3539–
354710564254MorrissetteNSSibleyLD2002Disruption of microtubules uncouples budding and nuclear division in Toxoplasma gondii.
115:
1017–
102511870220NicholsBAChiappinoML1987Cytoskeleton of Toxoplasma gondii.
34:
217–
2263585817ShawMKComptonHLRoosDSTilneyLG2000Microtubules, but not actin filaments, drive daughter cell budding and cell division in Toxoplasma gondii.
113:
1241–
125410704375RussellDGBurnsRG1984The polar ring of coccidian sporozoites: a unique microtubule-organizing centre.
65:
193–
2076715423TranJQde LeonJCLiCHuynhMHBeattyWMorrissetteNS2010RNG1 is a late marker of the apical polar ring in Toxoplasma gondii.
67:
586–
59820658557SibertGJSpeerCA1981Fine structure of nuclear division and microgametogony of Eimeria nieschulzi Dieben, 1924.
66:
179–
1897324546DubremetzJFElsnerYY1979Ultrastructural study of schizogony of Eimeria bovis in cell cultures.
26:
367–
376536929GubbelsMJVaishnavaSBootNDubremetzJFStriepenB2006A MORN-repeat protein is a dynamic component of the Toxoplasma gondii cell division apparatus.
119:
2236–
224516684814BrooksCFFranciaMEGissotMCrokenMMKimKStriepenB2011Toxoplasma gondii sequesters centromeres to a specific nuclear region throughout the cell cycle.
108:
3767–
377221321216Del CarmenMGMondragonMGonzalezSMondragonR2009Induction and regulation of conoid extrusion in Toxoplasma gondii.
11:
967–
98219416276MondragonRFrixioneE1996Ca(2+)-dependence of conoid extrusion in Toxoplasma gondii tachyzoites.
43:
120–
1278720941MonteiroVGde MeloEJAttiasMde SouzaW2001Morphological changes during conoid extrusion in Toxoplasma gondii tachyzoites treated with calcium ionophore.
136:
181–
18912051898MorrissetteNSMurrayJMRoosDS1997Subpellicular microtubules associate with an intramembranous particle lattice in the protozoan parasite Toxoplasma gondii.
110(Pt 1):35–429010782SheffieldHG1966Electron microscope study of the proliferative form of Besnoitia jellisoni.
52:
583–
5944957507CulverBPMeehlJBGiddingsTHJrWineyM2009The two SAS-6 homologs in Tetrahymena thermophila have distinct functions in basal body assembly.
20:
1865–
187719158390LeidelSDelattreMCeruttiLBaumerKGonczyP2005SAS-6 defines a protein family required for centrosome duplication in C. elegans and in human cells.
7:
115–
12515665853PelletierLO'TooleESchwagerAHymanAAMuller-ReichertT2006Centriole assembly in Caenorhabditis elegans.
444:
619–
62317136092DammermannAMuller-ReichertTPelletierLHabermannBDesaiAOegemaK2004Centriole assembly requires both centriolar and pericentriolar material proteins.
7:
815–
82915572125NakazawaYHirakiMKamiyaRHironoM2007SAS-6 is a cartwheel protein that establishes the 9-fold symmetry of the centriole.
17:
2169–
217418082404StevensNRRoqueHRaffJW2010DSas-6 and Ana2 coassemble into tubules to promote centriole duplication and engagement.
19:
913–
91921145506BlachonSCaiXRobertsKAYangKPolyanovskyAChurchAAvidor-ReissT2009A proximal centriole-like structure is present in Drosophila spermatids and can serve as a model to study centriole duplication.
182:
133–
14419293139Rodrigues-MartinsARiparbelliMCallainiGGloverDMBettencourt-DiasM2008From centriole biogenesis to cellular function: centrioles are essential for cell division at critical developmental stages.
7:
11–
1618196975Rodrigues-MartinsABettencourt-DiasMRiparbelliMFerreiraCFerreiraICallainiGGloverDM2007DSAS-6 organizes a tube-like centriole precursor, and its absence suggests modularity in centriole assembly.
17:
1465–
147217689959KitagawaDVakonakisIOliericNHilbertMKellerDOliericVBortfeldMEratMCFluckigerIGonczyPSteinmetzMO2011Structural basis of the 9-fold symmetry of centrioles.
144:
364–
37521277013van BreugelMHironoMAndreevaAYanagisawaHAYamaguchiSNakazawaYMorgnerNPetrovichMEbongIORobinsonCVJohnsonCMVeprintsevDZuberB2011Structures of SAS-6 suggest its organization in centrioles.
331:
1196–
119921273447HuynhMHCarruthersVB2009Tagging of endogenous genes in a Toxoplasma gondii strain lacking Ku80.
8:
530–
53919218426RoosDSDonaldRGMorrissetteNSMoultonAL1994Molecular tools for genetic dissection of the protozoan parasite Toxoplasma gondii.
45:
27–
637707991BrunRSchonenberger1979Cultivation and in vitro cloning or procyclic culture forms of Trypanosoma brucei in a semi-defined medium.
36:
289–
292WirtzELealSOchattCCrossGA1999A tightly regulated inducible expression system for conditional gene knock-outs and dominant-negative genetics in Trypanosoma brucei.
99:
89–
10110215027KellySReedJKramerSEllisLWebbHSunterJSaljeJMarinsekNGullKWicksteadBCarringtonM2007Functional genomics in Trypanosoma brucei: a collection of vectors for the expression of tagged proteins from endogenous and ectopic gene loci.
154:
103–
10917512617HartmannJHuKHeCYPelletierLRoosDSWarrenG2006Golgi and centrosome cycles in Toxoplasma gondii.
145:
125–
12716266757GilbertLARavindranSTuretzkyJMBoothroydJCBradleyPJ2007Toxoplasma gondii targets a protein phosphatase 2C to the nuclei of infected host cells.
6:
73–
8317085638MaCTranJLiCGanesanLWoodDMorrissetteN2008Secondary mutations correct fitness defects in Toxoplasma gondii with dinitroaniline resistance mutations.
180:
845–
85618780736HodgesMEScheumannNWicksteadBLangdaleJAGullK2010Reconstructing the evolutionary history of the centriole from protein components.
123:
1407–
141320388734KatohKKumaKTohHMiyataT2005MAFFT version 5: improvement in accuracy of multiple sequence alignment.
33:
511–
51815661851RonquistFHuelsenbeckJP2003MrBayes 3: Bayesian phylogenetic inference under mixed models.
19:
1572–
157412912839GuindonSGascuelO2003A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood.
52:
696–
70414530136HeaslipATEms-McClungSCHuK2009TgICMAP1 is a novel microtubule binding protein in Toxoplasma gondii.
4:
e7406.10.1371/journal.pone.000740619823689StechmannACavalier-SmithT2002Rooting the eukaryote tree by using a derived gene fusion.
297:89–
9112098695SindenRTalmanAMarquesSWassMSternbergM2010The flagellum in malarial parasites.
13:
491–
50020566299MuellerBEDesserSSHaberkornA1981Ultrastructure of developing gamonts of Eimeria contorta Haberkorn, 1971 (Protozoa, Sporozoa) with emphasis on the host-parasite interface.
67:
487–
4957264836FergusonDJPBirch-AndersenAHutchisonWMSiimJC1980Ultrastructural observations on microgametogenesis and the structure of the microgamete of Isospora felis.
88B:
151–
159ScholtyseckEMehlhornHHammondDM1972Electron microscope studies of microgametogenesis in coccidia and related groups.
38:
95–
1314622927YoungJAFivelmanQLBlairPLde la VegaPLe RochKGZhouYCarucciDJBakerDAWinzelerEA2005The Plasmodium falciparum sexual development transcriptome: a microarray analysis using ontology-based pattern identification.
143:
67–
7916005087KeeleyASoldatiD2004The glideosome: a molecular machine powering motility and host-cell invasion by Apicomplexa.
14:
528–
53215450974SibleyLD2004Intracellular parasite invasion strategies.
304:
248–
25315073368BaumJRichardDHealerJRugMKrnajskiZGilbergerTWGreenJLHolderAACowmanAF2006A conserved molecular motor drives cell invasion and gliding motility across malaria life cycle stages and other apicomplexan parasites.
281:
5197–
520816321976MorrissetteNSSibleyLD2002Cytoskeleton of apicomplexan parasites.
66:
21–
3811875126HuKRoosDSMurrayJM2002A novel polymer of tubulin forms the conoid of Toxoplasma gondii.
156:
1039–
105011901169Cavalier-SmithTChaoEE2004Protalveolate phylogeny and systematics and the origins of Sporozoa and dinoflagellates (phylum Myzozoa nom. nov.).
40:
185–
212LeanderBSKeelingPJ2003Morphostasis in alveolate evolution.
18:
395–
402LeanderBSKuvardinaONAleshinVVMylnikovAPKeelingPJ2003Molecular phylogeny and surface morphology of Colpodella edax (Alveolata): insights into the phagotrophic ancestry of apicomplexans.
50:
334–
34014563171BrugerolleG2002Cryptophagus subtilis: a new parasite of cryptophytes affiliated with the Perkinsozoa lineage.
37:
379–
390BrugerolleG2002Colpodella vorax: ultrastructure, predation, life-cycle, mitosis, and phylogenetic relationships.
38:
113–
125SimpsonAPattersonD1996Ultrastructure and identification of the predatory flagellate Colpodella pugnax Cienkowski (Apicomplexa) with a description of Colpodella turpis n. sp. and a review of the genus.
33:
187–
198SalisburyJL1995Centrin, centrosomes, and mitotic spindle poles.
7:
39–
457755988GeimerSCleesJMelkonianMLechtreckKF1998A novel 95-kD protein is located in a linker between cytoplasmic microtubules and basal bodies in a green flagellate and forms striated filaments in vitro.
140:
1149–
11589490727LechtreckKFFrinsSBilskiJTeltenkotterAWeberKMelkonianM1996The cruciated microtubule-associated fibers of the green alga Dunaliella bioculata consist of a 31 kDa SF-assemblin.
109(Pt 4):827–8358718674LechtreckKFMelkonianM1991Striated microtubule-associated fibers: identification of assemblin, a novel 34-kD protein that forms paracrystals of 2-nm filaments in vitro.
115:
705–
7161918160HarperJDThuetJLechtreckKFHardhamAR2009Proteins related to green algal striated fiber assemblin are present in stramenopiles and alveolates.
236:
97–
10119333716LechtreckKF2003Striated fiber assemblin in apicomplexan parasites.
128:
95–
9912706803FranciaMEJordanCNPatelJDSheinerLDemerlyJLFellowsJDde LeonJCMorrissetteNSDubremetzJFStriepenB2012Cell division in apicomplexan parasites is organized by a homolog of the striated rootlet fiber of algal flagella.
10:
e1001444.10.1371/journal.pbio.100144423239939FinnRDMistryJTateJCoggillPHegerAPollingtonJEGavinOLGunasekaranPCericGForslundKHolmLSonnhammerELEddySRBatemanA2010The Pfam protein families database.
38:D211–
D22219920124LupasAVan DykeMStockJ1991Predicting coiled coils from protein sequences.
252:
1162–
11642031185oai:pubmedcentral.nih.gov:36974722013-07-02eukcellpmc-openEukaryot CellEukaryotic CelleukcelleukcellEUKCELLEukaryotic Cell1535-97781535-9786American Society for Microbiology1752 N St., N.W., Washington, DCPMC3697472PMC36974723697472236500872365008700040-1310.1128/EC.00040-13ArticlesDiscovery of a Sexual Cycle in Aspergillus lentulus, a Close Relative of A. fumigatusSwilaimanSameira S.aO'GormanCéline M.aBalajeeS. ArunmozhibDyerPaul S.aSchool of Biology, University of Nottingham, University Park, Nottingham, United KingdomCenter for Global Health, Centers for Disease Control and Prevention, Atlanta, Georgia, USA
S.S.S. and C.M.O. contributed equally to this article.
Aspergillus lentulus was described in 2005 as a new species within the A. fumigatus sensu lato complex. It is an opportunistic human pathogen causing invasive aspergillosis with high mortality rates, and it has been isolated from clinical and environmental sources. The species is morphologically nearly identical to A. fumigatus sensu stricto, and this similarity has resulted in their frequent misidentification. Comparative studies show that A. lentulus has some distinguishing growth features and decreased in vitro susceptibility to several antifungal agents, including amphotericin B and caspofungin. Similar to the once-presumed-asexual A. fumigatus, it has only been known to reproduce mitotically. However, we now show that A. lentulus has a heterothallic sexual breeding system. A PCR-based mating-type diagnostic detected isolates of either the MAT1-1 or MAT1-2 genotype, and examination of 26 worldwide clinical and environmental isolates revealed similar ratios of the two mating types (38% versus 62%, respectively). MAT1-1 and MAT1-2 idiomorph regions were analyzed, revealing the presence of characteristic alpha and high-mobility-group (HMG) domain genes, together with other more unusual features such as a MAT1-2-4 gene. We then demonstrated that A. lentulus possesses a functional sexual cycle with mature cleistothecia, containing heat-resistant ascospores, being produced after 3 weeks of incubation. Recombination was confirmed using molecular markers. However, isolates of A. lentulus failed to cross with highly fertile strains of A. fumigatus, demonstrating reproductive isolation between these sibling species. The discovery of the A. lentulus sexual stage has significant implications for the management of drug resistance and control of invasive aspergillosis associated with this emerging fungal pathogen.
INTRODUCTION
Aspergillus lentulus is an opportunistic human pathogenic fungus from the Aspergillus section Fumigati (1) that can cause invasive aspergillosis in immunocompromised patients (2). It was first discovered in 2004, as an unexpected result from a U.S. study examining the susceptibility of isolates of A. fumigatus to azole antifungal drugs (3). Four A. lentulus isolates that caused fatal infections in hematopoietic stem cell transplant patients between the years 1995 and 2000 had been misidentified as poorly sporulating variants of A. fumigatus because their 18S rRNA sequences matched that of A. fumigatus. However, all of the isolates had distinct random amplified polymorphic DNA (RAPD) fingerprint patterns and mitochondrial cytochrome b sequences, and several exhibited in vitro antifungal susceptibility profiles with significantly higher resistance than normal for A. fumigatus. A follow-up study examined the phylogenetic relationships of these isolates to A. fumigatus using multilocus sequence typing (2). The analysis revealed that the four isolates formed a separate clade clearly distant from A. fumigatus, and it was concluded that these represented a new species, which was named A. lentulus (2). A subsequent phylogenetic study of Japanese clinical isolates showed that along with A. fumisynnematus, A. lentulus forms a sister clade to A. fumigatus (4).
Similar to other members of the section Fumigati that are closely related to A. fumigatus, such as A. fumigatiaffinis, many of the phenotypic characters of A. lentulus overlap with A. fumigatus. This makes identification difficult when based solely on morphological grounds, which often leads to misdiagnoses in clinical laboratories (5). However, a variety of methods have since been found that can differentiate between the two species. Phenotypic dissimilarities include their conidial ornamentation (4), conidiophore architecture, growth characteristics (1), and mycotoxin profiles—most notably the inability of A. lentulus to produce gliotoxin (1, 6). The most obvious growth characteristics that distinguish A. lentulus from A. fumigatus are its delayed onset of sporulation and inability to grow at 48°C, although it should be noted that these two features are common to several other species in the section Fumigati (7). Species-specific molecular methods are also available to identify A. lentulus and include the use of a microsphere-based Luminex assay (8), RAPD patterns (1), a multiplex PCR assay (9), and restriction fragment length polymorphisms (10).
Despite the wide geographic distribution and presence of the species in common environmental niches (3–5, 7, 11–13), opportunistic A. lentulus infections appear rare. To date, only six studies have reported cases of invasive aspergillosis in which A. lentulus was confirmed as the probable or causal agent (3, 11, 12, 14–16). Two further cases involved the colonization of cystic fibrosis patients (13, 17). This apparent low incidence rate may be due to the misidentification problems described previously (5), because A. lentulus isolates have been recovered in several retrospective studies of clinical fungal samples (4, 7).
The pathology of invasive infections caused by A. lentulus appears to mirror that of A. fumigatus (3). However, of medical significance is the fact that most isolates exhibit an increased natural resistance to several antifungal agents currently in clinical use compared to A. fumigatus, namely, itraconazole, voriconazole, caspofungin, and amphotericin B (3, 4, 7). The molecular mechanisms underlying these resistance mechanisms have recently started to be elucidated (18, 19), but further work is required to fully understand the genetic basis of resistance. Of particular interest are the mechanisms governing echinocandin resistance, as A. lentulus is highly unusual in being simultaneously resistant to caspofungin yet highly sensitive to anidulafungin and micafungin (7, 19, 20). The discovery of A. lentulus has also proven beneficial to industry, as an A. lentulus isolate has been found with activity as a biosorbent for the removal of toxic compounds. A. lentulus strain AML05, which was recovered from industrial textile effluent in India by a chromium enrichment process, can very successfully remove Cr(VI) from electroplating industry effluent (21, 22) and dyes from textile effluent (23, 24).
A. lentulus is currently known to reproduce only by asexual means, through the production of conidia (2). Hong and coworkers (1) previously tried unsuccessfully to mate strains of A. lentulus, but crucially, this was before recent reports of the discovery of a sexual cycle in A. fumigatus (teleomorph Neosartorya fumigata) (25) and other related Aspergillus and Penicillium species which were previously considered asexual (26–31). In most of these cases, the discovery of a sexual state was preceded by the identification of mating-type (MAT) genes within the species, such genes acting as key regulators of sexual identity in filamentous ascomycete fungi (32, 33). In heterothallic (obligate outbreeding) species, these MAT genes are contained within a region of the genome termed the MAT locus, with highly divergent forms of this locus, known as “idiomorphs,” present in isolates of sexually compatible MAT1-1 and MAT1-2 genotypes. In contrast, homothallic (self-fertile) species may contain one or more MAT loci within the same genome rather than dissimilar idiomorphs (33, 34).
The close phylogenetic relationship of A. lentulus to A. fumigatus suggested that it might be possible to induce sexual reproduction in A. lentulus using conditions similar to those required for A. fumigatus. The aims of this study were, therefore, first, to investigate the presence of MAT genes in A. lentulus to see if isolates with putative sexual compatibility could be identified; second, to determine if a sexual cycle could be induced in A. lentulus; third, to use molecular analysis of offspring to determine if the breeding system was heterothallic or homothallic in nature; and, finally, to see if any gene flow was possible between A. lentulus and A. fumigatus via sexual crossing. The ability to perform sexual crosses in A. lentulus would provide a valuable tool for the genetic analysis of traits relating to pathogenicity, antifungal drug resistance, and industrial processes in this species.
MATERIALS AND METHODSStrains, growth conditions, and DNA extraction.
Twenty-six isolates of A. lentulus from clinical and environmental sources were used in the study, comprising most isolates of the species previously reported in the scientific literature (see Table S1 in the supplemental material). Sixteen of the clinical isolates were kindly donated by Kieren Marr and Edmond Byrnes (John Hopkins School of Medicine, Maryland). Strains were maintained on Aspergillus complete medium (ACM) (35) at 28°C and are stored in 10% glycerol under liquid nitrogen at the School of Biology, University of Nottingham, United Kingdom (BDUN [Botany Department, University of Nottingham] culture collection). Genomic DNA was extracted using a DNeasy plant minikit (Qiagen) in accordance with the manufacturer's instructions. Cultures were grown in liquid ACM at 28°C for 5 days. The resulting mycelia were harvested, flash frozen, and ground under liquid nitrogen prior to DNA extraction.
Multiplex mating-type PCR assay and PCR amplification of the mating-type idiomorphs.
The mating-type genotype of all A. lentulus isolates and ascospore progeny was determined with the A. fumigatus multiplex mating-type PCR diagnostic of Paoletti et al. (35), using primers AFM1, AFM2, and AFM3 (see Table S2 in the supplemental material). Attempts were then made to amplify the entire MAT idiomorph regions of A. lentulus isolates 78-2 and 78-3 with the primer pair Loc1 and AFM3 (see Table S2) using the conditions described by Paoletti et al. (35); these primers had previously been used to successfully amplify the idiomorph regions of A. fumigatus (35). Resulting amplicons were purified using a Geneflow Q-Spin PCR purification kit (according to the manufacturer's instructions) and were sequenced using primers AL31-AL34 and AL51-AL53 (see Table S2) at the DNA Sequencing Facility of the School of Biomedical Sciences, University of Nottingham, United Kingdom. Arising sequences were analyzed and aligned using MacVector 11 (MacVector Inc.).
Sexual crosses.
Sixteen representative A. lentulus isolates of MAT1-1 and MAT1-2 genotype (see Table S1 in the supplemental material) were chosen for sexual crossing experiments and inoculated in all possible pairwise combinations by following the protocol of O'Gorman et al. (25). Briefly, crosses were set up on oatmeal agar (pinhead oatmeal, Odlums, Ireland [36]) in triplicate, sealed with one layer of Nescofilm, and incubated at 25, 28, or 30°C in the dark. Four A. fumigatus crosses known to reliably produce cleistothecia and ascospores (AfRB2 × AfIR928, AfRB2 × AfIR964, AfIR974 × AfIR928, and AfIR974 × AfIR964) (25) were set up in parallel as controls. Control “selfed” A. lentulus crosses were also tested on oatmeal agar at 28°C using one representative isolate of each mating type (78-2 [MAT1-2] and 78-3 [MAT1-1]). In addition, crosses were set up on oatmeal agar at 28°C and 30°C between representative A. lentulusMAT1-1 (78-3, 78-6, and 78-20) and MAT1-2 (78-2 and 78-8) isolates (see Table S1 in the supplemental material) and known highly fertile isolates of A. fumigatus (AFB62 [MAT1-1] and AfIR928 [MAT1-2]) (37) to assess reproductive isolation. All crosses were examined periodically for the presence of cleistothecia for up to 7 weeks using a Nikon-SMZ-2B dissection microscope and then, finally, after 4 and 12 months of incubation where cleistothecia were not detected in the initial growth period.
Preparation of single-ascospore cultures.
Mature cleistothecia from the cross 78-2 × 78-3 were removed and cleaned as described previously (25), with the modification that 4% water agar was used to clean the cleistothecia instead of a drop of sterile water. Five cleistothecia were added to 50 μl of 0.05% Tween 80 (BDH Chemicals) under sterile conditions and ruptured by squashing with a needle tip. The solution was brought up to 500 μl with 0.05% Tween 80 and vortex mixed for 1 min to release the ascospores. The suspension was then heat treated at 80°C for 30 min, this temperature being sufficient to kill any contaminating conidia without damaging the ascospores (data not shown). One hundred microliters of a 5 × 105-ascospore ml−1 heat-treated suspension was spread inoculated on three defined areas of an ACM plate. Triplicates were prepared and incubated at 37°C for 14 h. Single-spore cultures were established on ACM by transferring individual germinating ascospores with a LaRue lens cutter attached to a Nikon-Optiphot microscope.
Analysis of recombination.
The segregation of five genetic markers (four RAPD bands and the mating-type genotype) in the ascospore progeny was examined for evidence of recombination. RAPD-PCR fingerprinting was performed by following the protocol of O'Gorman et al. (25). Four primers (OMT1, R108, R151, and OPWO8 [see Table S2 in the supplemental material]) from an initial screen of 12 were found to yield suitable polymorphisms for genotyping.
SEM.
Cleistothecia were collected from 6-month-old crosses of 78-2 × 78-3 that had been incubated at 28°C. Representative intact and crushed cleistothecia were transferred onto 0.2-μm filter discs (Whatman) and fixed in 2% osmium tetroxide (Sigma) for 2 h at room temperature. The fixed samples were then mounted onto aluminum stubs, dried at 37°C overnight, and sputter coated with gold. Scanning electron microscopy (SEM) micrographs were taken using a JSM-840 JEOL scanning electron microscope.
Statistical analysis.
The hypothesis of a 1:1 ratio of mating types in the worldwide sample population and ascospore progeny was tested using χ2 and contingency χ2 tests. Where expected frequencies were <5, Fisher's exact test was used instead (38).
Accession numbers.
The sequences for the two mating-type loci amplified from isolates 78-3 (MAT1-1) and 78-2 (MAT1-2) have been deposited in GenBank under accession numbers KC876046 and KC876047. The diagnosis of the Aspergillus lentulus (neosartorya-morph) has been deposited in MycoBank under accession number MB356679 (see the supplemental material).
RESULTS AND DISCUSSION
Sex is thought to have evolved in early eukaryotic microbes and is now widespread throughout the Eukaryota (39). The ability to undergo sexual reproduction is considered to be of major importance given the many benefits it confers. These include the potential to purge harmful mutations and improve the fitness of offspring, which, in turns, allows them to better resist adverse environmental conditions (40–43). It is therefore surprising that the kingdom Fungi appears to contain a disproportionally large number of supposedly asexual species, with an estimated 20% having no known sexual stage. Many are members of the phylum Ascomycota that are of medical or economic significance (44). Some, such as A. oryzae, have been shown to possess MAT genes and other “sexual machinery,” yet their sexual cycle remains elusive (45). While there are clear advantages to asexual over sexual reproduction, such as the relatively lower metabolic cost and ability to produce spores under a wider range of environmental conditions, the rewards of sexual reproduction appear to be much greater (46).
The discovery of a heterothallic sexual cycle in A. fumigatus, which had long been considered to be reproduce purely by mitotic means, confirmed suspicions that at least some of these supposed “asexuals” do in fact have the potential to reproduce sexually (25, 43). Significantly, a “sexual revolution” has since followed, with the reporting of functional sexual cycles in several other related filamentous fungi (26–31, 47). The aim of this work was to determine whether it was possible to induce a sexual cycle in A. lentulus, given its close phylogenetic relationship to A. fumigatus (1) and its increasing importance as a human pathogen with resistance to drugs in several antifungal classes relative to A. fumigatus (2).
Presence, distribution, and characterization of the MAT idiomorphs.
The A. lentulus genome was first examined for the presence of the master regulatory MAT genes that are transcription factors common to all heterothallic fungi and which determine cell sexual identity (32). It was found that the previously described multiplex PCR mating type diagnostic for A. fumigatus (35) produced corresponding amplicons of the predicted size (ca. 834 bp for MAT1-1 and 438 bp for MAT1-2) in different isolates of A. lentulus (see Fig. S1 in the supplemental material). This indicated a heterothallic (obligate outcrossing) arrangement in the species and confirmed the phylogenetic affinity with A. fumigatus, given that this was designed as a species-specific diagnostic test. The worldwide collection of A. lentulus isolates was then screened to determine the ratio of complementary MAT1-1 and MAT1-2 genotypes (see Table S1 in the supplemental material). Amplicons were generated for all isolates, and the overall mating-type distribution did not deviate significantly from a 1:1 ratio (38.5% MAT1-1 and 61.5% MAT1-2; χ2 = 1.38; n = 26), consistent with a sexually reproducing species. When isolates were grouped according to geographic origin, there was also no significant difference in the MAT distribution (data not shown).
It was then found that the entire MAT1-1 and MAT1-2 idiomorph regions of A. lentulus could be amplified using the primers Loc1 and AMF3, which had previously been used to amplify MAT regions from A. fumigatus (35). This yielded amplicons of 2,524 and 2,731 bp, respectively. Figure 1 shows the complete sequenced idiomorph structure of the two mating-type loci amplified from isolates 78-3 (MAT1-1) and 78-2 (MAT1-2). The idiomorphs showed the same overall structural organization as those previously reported from A. fumigatus (35). The MAT1-1 idiomorph of 78-3 contains a 1,157-bp putative open reading frame (ORF), predicted to encode a 368-amino-acid protein with a characteristic α1 domain, and was therefore termed MAT1-1-1 (33). The ORF is interrupted by a 50-bp intron in the conserved position found in other ascomycetes, and the overall protein and α1 domain region share 93% and 98% amino acid identity to A. fumigatus, respectively (see Fig. S2A in the supplemental material). The MAT1-2 idiomorph of 78-2 contains two putative ORFs. The first is a 1,077-bp ORF predicted to encode a 322-amino-acid protein with a characteristic high-mobility-group (HMG) box, which was therefore termed MAT1-2-1 (33). This ORF contains two introns (53 and 55 bp), and the overall protein and HMG domain share 89% and 92% amino acid identity to A. fumigatus, respectively (see Fig. S2B). The second is a putative ORF of 903 bp with three introns (46, 46, and 58 bp), predicted to encode a 242-amino-acid product which was found to share 95% and 87% amino acid identity to the putative MAT1-2-4 proteins from A. fumigatus and Neosartorya fischeri, respectively (see Fig. S2C) (48, 49; C. Eagle and P. S. Dyer, unpublished data). A related MAT1-2-4 family gene has also been described for Talaromyces (Penicillium) marneffei (50). However, similar MAT1-2-4 genes are absent from many other aspergilli, and their expression and possible functionality remain to be determined. A MAT1-2-4 gene has also been reported for Coccidioides immitis and C. posadasii (51, 52), but this shares little sequence conservation with the MAT1-2-4 proteins from the aspergilli and is absent from the MAT loci of many other heterothallic eurotiomycete species, such as those recently described for Blastomyces dermatitidis (53).
MAT locus of Aspergillus lentulus. The arrangement of the A. lentulus idiomorph region shows the difference in organization between isolates 78-3 (MAT1-1) and 78-2 (MAT1-2). Colored block arrows indicate MAT1-1-1 (red), MAT1-2-1 (green), and MAT1-2-4 (white) sequences. Colored boxes indicate the α1 domain (salmon pink), HMG domain (light green), and nearly identical flanking regions (gray). Introns are represented by black boxes; lines extending between boxes and arrows represent idiomorph sequence. Smaller red and green segments represent regions between 10 and 29 bp in length with ≥70% MAT1-1-1 and MAT1-2-1 nucleotide conservation, respectively. Black arrowheads (direction indicates 5′ to 3′ sequence) show the positions of primers (see Table S2 in the supplemental material) used for amplification of the idiomorph region.
Similar to the previous report for A. fumigatus (35), although the MAT1-2-1 gene commenced within the MAT1-2 idiomorph, a terminal 374-bp region was found to lie within the flanking sequence bordering both idiomorphs (Fig. 1). However, the fragment bordering the MAT1-1 idiomorph appeared nonfunctional, as it lacked the HMG domain region and any start codon and contained a deletion, giving rise to a frameshift mutation. Therefore, the fragment was termed dMAT1-1-1 in recognition of the disabled ORF (48). Intriguingly, further analysis of the MAT1-1 idiomorph revealed an additional 14 regions between 10 and 15 bp in length with ≥70% nucleotide conservation compared to the A. lentulus MAT1-2-1 gene. These extended directly upstream from the 374-bp dMAT1-2-1 fragment, with seven of the regions present within the predicted HMG domain (Fig. 1). Similarly, further analysis of the MAT1-2 idiomorph revealed 18 regions between 10 and 29 bp in length within the idiomorph which exhibited ≥70% nucleotide conservation compared to A. lentulus MAT1-1-1 gene, including 13 within the MAT1-2-4 gene itself (Fig. 1). These data indicate a complex evolutionary history for the idiomorphs, possibly signifying the presence of an ancestral homothallic MAT locus containing both MAT1-1-1 and MAT1-2-1 genes, which has since undergone accelerated mutation and evolution in the transition to heterothallism (54). It has been suggested that homothallism might be the ancestral state within the Aspergillus section Fumigati, with the overwhelming majority of teleomorphic Neosartorya species exhibiting homothallic breeding systems (55, 56). It can also be speculated that the MAT1-2-4 gene, whose origins are obscure, might have arisen from sequence derived from the MAT1-1-1 gene.
Sexual crosses.
Crosses were set up between isolates of complementary mating type using the conditions that had successfully induced sex in A. fumigatus (25), with one modification. Both 25°C and 28°C were tested in addition to the 30°C required for A. fumigatus mating, given that A. lentulus has a lower growth temperature range than A. fumigatus (1). Nine isolates of clinical and environmental origin from North America and Asia (see Table S1 in the supplemental material) were crossed in all pairwise combinations (Table 1). Significantly, after 3 weeks of incubation at both 28°C and 30°C, certain crosses were found to be fertile, producing cleistothecia that contained viable ascospores (Table 1 and Fig. 2; see supplemental material for species diagnosis). In contrast, no cleistothecia were observed in crosses incubated at 25°C. Cleistothecia also failed to develop on single MAT cultures, indicating that A. lentulus is a heterothallic species.
Mean numbers of cleistothecia produced by 14 Aspergillus lentulus crosses on oatmeal agar at 28°C or 30°C in the dark after 3 weeksa
MAT1-2 strain
Score for cleistothecium production of cross with MAT1-1 strain
28°C
30°C
78-3
78-6
78-3
78-6
78-1
−
−
−
−
78-2
>
+
++++
+++
78-4
−
−
−
−
78-5
−
−
−
−
78-7
−
−
−
−
78-8
+
+
+
−
78-9
−
−
−
−
Ratings indicate the mean number of cleistothecia produced from three replicate crosses on oatmeal agar in 9-cm-diameter petri dishes after incubation in the dark, as follows: −, none; +, 1 to 19; +++, 40 to 59; ++++, 80 to 100; >, more than 100.
Sexual reproduction in Aspergillus lentulus. (A) Paired cultures of isolates 78-2 (MAT1-2) × 78-3 (MAT1-1) on oatmeal agar (9-cm-diameter petri dish) with cleistothecia (arrows) along the barrage zones following 3 weeks of incubation at 28°C. (B) Close-up of a barrage zone showing pale yellow cleistothecia. Scale bar, 400 μm. (C) SEM micrograph of a cleistothecium showing the interwoven hyphae that form the peridial wall. Scale bar, 100 μm. (D) SEM micrograph of lenticular ascospores (white arrow) and smaller globose conidia (black arrow). Scale bar, 10 μm.
In crosses where cleistothecia were produced, they formed along the barrage zones between isolates of opposite mating type (Fig. 2A) and were pale yellow-orange in color (Fig. 2B). All crosses forming cleistothecia produced viable ascospores. The ascospores were heat resistant (57), capable of surviving at 80°C for at least 30 min, as is typical for other members of the genus Neosartorya (56). This might reflect selection in a common ancestor of Neosartorya for survival in ecological niches where high temperatures might be encountered, such as composting vegetation (58). Ascospore progeny from a cross between isolates 78-2 and 78-3 were then assessed for evidence of recombination. Distinct segregation patterns were clearly observed between four RAPD-PCR markers and the MAT genotype in 12 ascospore progeny, with Fisher's exact test confirming 1:1 Mendelian segregation of the markers due to independent assortment, confirming a heterothallic sexual breeding system (Table 2 and Fig. 3). Unique genotypes were found in 92% of the progeny, with only one of the offspring identical to its parent (based on the markers examined).
Genotypes of parental isolates and 12 ascospore progeny from a cross between A. lentulus isolates 78-2 and 78-3, based on mating type and RAPD-PCR bands
Isolate
Mating type
RAPD banda
Genotypeb
OMT1
R108
OPW08
R151
78-3
MAT1-1
−
+
+
−
P1
78-2
MAT1-2
+
−
−
+
P2
3-2-1
MAT1-2
−
−
−
−
A
3-2-2
MAT1-2
−
−
+
+
B
3-2-3
MAT1-1
−
+
+
−
P1
3-2-4
MAT1-1
−
−
+
−
C
3-2-6
MAT1-1
−
−
−
−
D
3-2-7
MAT1-2
−
−
+
−
E
3-2-8
MAT1-2
−
+
+
−
F
3-2-9
MAT1-1
−
−
−
+
G
3-2-12
MAT1-2
−
+
−
−
H
3-2-13
MAT1-2
−
−
+
−
E
3-2-15
MAT1-1
+
−
+
−
I
3-2-16
MAT1-1
+
−
+
+
J
RAPD-PCR bands amplified using primers OMT1, R108, OPW08, and R151. “+” and “−” denote presence and absence, respectively, of a particular amplicon. P values (two-tailed) for OMT1, R108, OPW08, and R151 were 0.06, 0.24, 0.57, and 0.24, respectively. Fisher's exact test was conducted to check for deviation from the null hypothesis of independent assortment of mating-type and RAPD markers in the progeny (i.e., a 1:1:1:1 MAT1-1+:MAT1-1−:MAT1-2+:MAT1-2− ratio for each RAPD marker). Fisher's exact test was used instead of the χ2 test because the expected frequencies were <5. A contingency χ2 test was conducted to check for deviation from the null hypothesis of independent assortment of mating-type and RAPD markers in the progeny (i.e., an overall 1:1:1:1 MAT1-1+:MAT1-1−:MAT1-2+:MAT1-2− ratio for the sum of the RAPD markers). It showed a value of 0.375 with 1 degree of freedom.
The genotype of each progeny isolate, defined by unique combinations of mating-type and RAPD markers as distinct from those of the parental isolates (designated P1 and P2), is identified by a different letter of the alphabet.
Evidence for meiotic recombination. The gel shows segregation patterns of a RAPD-PCR amplicon in A. lentulus parental isolates (P1 and P2) and 12 ascospore progeny (lanes 1 to 12) from the cross 78-3 (P1) × 78-2 (P2), using primer OPW08. MM, molecular size marker; C, water control. Arrow indicates the diagnostic RAPD band. Red and green lane headings indicate MAT1-1 and MAT1-2 genotypes, respectively.
In accordance with the “One Fungus = One Name” proposal (59), the newly discovered sexual state (teleomorph) of A. lentulus will not be assigned a separate Latin name, as was formerly the case under “dual nomenclature” (60). This follows the recent taxonomic move to simplify the naming of pleomorphic fungi. In the future, the teleomorph of A. lentulus should be referred to as A. lentulus (neosartorya-morph) where appropriate (59), given the phylogenetic link between the teleomorph genus Neosartorya and the Aspergillus section Fumigati (47, 56).
As shown in Table 1, three of the four fertile crosses (fertility is here defined as the production of cleistothecia with viable ascospores) yielded sexual offspring at both 28°C and 30°C, suggesting that A. lentulus may not be as fastidious as A. fumigatus in its temperature requirement for mating (61). Crosses were then reincubated for a further 4 weeks and reexamined for the presence of cleistothecia (see Table S3 in the supplemental material). The longer incubation period resulted in a further 7 and 14% of crosses reaching sexual maturity at 28°C and 30°C, respectively. Thus, a total of 35% of the crosses were fertile at both temperatures. A second series of 20 crosses was then set up at 28°C for 3 weeks to test the fertility of an additional seven isolates (see Table S4 in the supplemental material). Six of the seven isolates were fertile with at least one mating partner, although overall fertility was still only 35%. These results illustrate the importance of having multiple isolates of opposite mating type in close proximity in the environment, to ensure compatibility with at least one complementary strain.
Finally, crosses were attempted between representative MAT1-1 and MAT1-2 isolates of A. lentulus and highly fertile “supermater” isolates of A. fumigatus (37). Hyphal aggregations resembling immature cleistothecia were formed very occasionally in such crosses (see Fig. S3 in the supplemental material). However, despite prolonged incubation, for up to 12 months, at both 28°C and 30°C, these never matured to form ascospores. This indicates that the species are true sibling species, being phylogenetically closely related but without gene flow via sexual means (62). This has important implications for the evolution of resistance to antifungal drugs in A. fumigatus, as it suggests, fortunately, that transmission of genes conferring such resistance is unlikely to occur between A. lentulus and A. fumigatus. This result is consistent with the phylogenetic divergence reported between the species (2, 4), although it is cautioned that crosses were attempted only between a subset of A. lentulus and A. fumigatus isolates. It has also been reported that abortive cleistothecia can be produced in crosses between Neosartorya fennelliae and A. fumigatus (63). These crossing results compare with data for some other ascomycete species, such as within the genus Neurospora, where interspecies crossing is possible, albeit with reduced fertility and ascospore viability (64, 65).
The large variation in fertility of A. lentulus isolates depending on the mating partner is similar to observations reported for A. fumigatus by both O'Gorman et al. (25) and Sugui et al. (37) and for Neosartorya udagawae by Sugui et al. (66). For example, the A. lentulus cross 78-3 × 78-8 produced fewer than 20 cleistothecia per plate (Table 1). However, when 78-3 was crossed with 78-2 it was the most fertile pairing at both temperatures, consistently producing ≥80 cleistothecia per plate (Table 1). The pairing of 78-3 × 78-2 is therefore recommended for community use. It should be noted that these two isolates came from agricultural soils within 50 mi of each other in the Republic of Korea. Their similar geographic origins may hint at a close genetic relationship and lack of reproductive barriers, this being one of a variety of factors suggested to influence fertility of fungal sexual crosses (67–69). Genome relatedness is thought to be an important factor for fertility, as illustrated for A. fumigatus by the “supermater” pair (37). These isolates share 99% genome similarity based on a CGH (comparative genomic hybridization) analysis, and the cross is the most successful A. fumigatus pairing to date.
It is important to note that only 25% of the A. lentulus crosses at both 28°C and 30°C were fertile after 3 weeks. This figure is extremely low in comparison to that for A. fumigatus, for which the study by Sugui et al. (37) found that 80% (n = 50) of crosses were fertile after 4 weeks. O'Gorman et al. (25) reported 94% of their A. fumigatus crosses to be fertile (n = 36), but this was after 6 months of incubation, which can be considered to represent maximum maturity. The large disparity in fertility between A. lentulus and A. fumigatus is surprising given their nearly identical MAT idiomorph structures (Fig. 1) and conditions required for the sexual cycle. It is conceivable that either part or all of the global A. lentulus population is undergoing a slow decline in fertility (44). Alternatively, it could suggest a difference in the population dynamics of the two species due to an unknown underlying fertility or incompatibility mechanism in A. lentulus. Natural A. fumigatus populations have been well studied and have yielded evidence of recombination, confirming that sexual reproduction is taking or has recently taken place (35, 70, 71). Similar studies have yet to be conducted for A. lentulus, for which many fewer isolates are available for study. Defining its population structure will inform future studies to determine how widespread sexual reproduction is in nature and whether the low fertility seen in this study is representative of the global population.
Conclusions.
The discovery of a sexual cycle in A. lentulus is important both for the biology of the species and for future efforts to control this pathogen. It also shows that yet another supposedly asexual pathogenic fungus possesses a functional sexual cycle, thereby harboring the potential to evolve rapidly in the face of selective pressures (43). Sexuality in its close relative A. fumigatus can now be directly compared to that of A. lentulus, and future studies may shed light on their different evolutionary paths. However, of most significance is the fact that the sexual cycle will provide an invaluable tool for classical genetic analysis to facilitate research into the genetic basis of pathogenicity and drug resistance in this emerging agent of aspergillosis.
Supplementary Material
Supplemental material
Published ahead of print 6 May 2013
Supplemental material for this article may be found at http://dx.doi.org/10.1128/EC.00040-13.
ACKNOWLEDGMENTS
We thank Kieren Marr and Edmond Byrnes for providing A. lentulus isolates and Tim Smith for technical assistance with scanning electron microscopy.
S.S.S. was supported by a Government of Iraq PhD scholarship. C.M.O. and P.S.D. are funded by The Wellcome Trust (United Kingdom).
In the bloodstream of mammalian hosts, the sleeping sickness parasite, Trypanosoma brucei, exists as a proliferative slender form or a nonproliferative, transmissible, stumpy form. The transition between these developmental forms is controlled by a density-dependent mechanism that is important for the parasite's infection dynamics, immune evasion via ordered antigenic variation, and disease transmissibility. However, stumpy formation has been lost in most laboratory-adapted trypanosome lines, generating monomorphic parasites that proliferate uncontrolled as slender forms in vitro and in vivo. Nonetheless, these forms are readily amenable to cell culture and high-throughput screening for trypanocidal lead compounds. Here, we have developed and exploited a high-throughput screen for developmental phenotypes using a transgenic monomorphic cell line expressing a reporter under the regulation of gene control signals from the stumpy-specific molecule PAD1. Using a whole-cell fluorescence-based assay to screen over 6,000 small molecules from a kinase-focused compound library, small molecules able to activate stumpy-specific gene expression and proliferation arrest were assayed in a rapid assay format. Independent follow-up validation identified one hit able to induce modest, yet specific, changes in mRNA expression indicative of a partial differentiation to stumpy forms in monomorphs. Further, in pleomorphs this compound induced a stumpy-like phenotype, entailing growth arrest, morphological changes, PAD1 expression, and enhanced differentiation to procyclic forms. This not only provides a potential tool compound for the further understanding of stumpy formation but also demonstrates the use of high-throughput screening in the identification of compounds able to induce specific phenotypes, such as differentiation, in African trypanosomes.
INTRODUCTION
Members of the order Kinetoplastida comprise a highly divergent group of eukaryotic organisms, with several representatives causing life-threatening and debilitating human diseases in a human-human or zoonotic transmission cycle. Trypanosoma brucei sp. is a blood-dwelling parasitic organism of this order, transmitted by tsetse flies, that causes both human and animal African trypanosomiasis (HAT and AAT, respectively) in sub-Saharan Africa. These diseases cause morbidity and then death when untreated in humans (1) and significantly contribute to poverty and restrict economic development in afflicted regions. Due to the poor safety and efficacy of current drugs, new treatments for African trypanosomiasis are urgently required (2, 3), and, consequently, considerable efforts are being made to search for new trypanocidal compounds. One approach has been through whole-cell-based high-throughput screening (4–10) where thousands of small molecules are rapidly analyzed for antiparasitic activity, identifying classes of lead compounds with potential for further development as antitrypanosomal drugs (6, 7, 9, 10). In addition to the use of high-throughput chemical inhibitor screens for the identification of compounds with antiparasitic properties, similar approaches have been used in Plasmodium falciparum and Toxoplasma gondii to search for small molecules able to induce specific cytological phenotypes (11–13), providing potential novel therapeutic approaches and tool compounds for biological research (14).
In the bloodstream of their mammalian host, African trypanosomes differentiate from the proliferative slender form to the transmissible stumpy form (15), a transition that is key to the within-host infection dynamics and transmissibility of the parasite (16). In naturally occurring pleomorphic cell lines, stumpy formation is triggered by parasite density, as sensed through the accumulation of an unidentified stumpy induction factor (SIF) (17). Although laboratory-adapted monomorphic bloodstream forms have lost the ability to differentiate to stumpy forms in response to population density in vitro or in vivo, these cell lines do have the capacity for differentiation to a “stumpy-like” state when treated with cell-permeable cyclic AMP (8-pCPT-cAMP) and, more potently, the hydrolysis products of these compounds and cell-permeable AMP, 8-pCPT-2′-O-methyl 5′-AMP (16–19). In response to these stimuli, monomorphic cells undergo cell cycle arrest, demonstrate an increased capacity for differentiation to the procyclic stage, and undergo changes in gene expression associated with stumpy formation (16–18). It has been hypothesized that cAMP analogues act to mimic the induction of stumpy formation, bypassing the initial step(s) of environmental density sensing and triggering at least part of the stumpy induction signaling cascade within the cell. Until recently, very little was known regarding the molecules involved in stumpy formation, with only negative regulators TbTOR4 kinase (Tb927.1.1930), which prevents stumpy formation in monomorphic slender forms (20), and two further inhibitory kinases, mitogen-activated protein kinase 5 (MAPK5) (Tb927.6.4220) (21) and zinc finger kinase (ZFK) (Tb927.11.9270) (22), having been identified. However, recently, promoters of stumpy formation have also been identified based on the selection of monomorphic RNA interference (RNAi) libraries for resistance to cell-permeable cAMP/AMP analogues, and these promoters have been further validated in vivo using pleomorphs for SIF responsiveness. Hence, a cohort of molecules representing many steps in the signal response leading to stumpy formation have now been identified, with small-molecule drivers of the stumpy-like response having been central to identification of the pathway components. Clearly, this knowledge of the stumpy induction pathway could also lead to novel therapeutic approaches since molecular inhibitors of stumpy formation could be targeted to induce premature development in the bloodstream, reducing parasite virulence or reducing abundance below a transmission threshold. Alternatively, molecular drivers of stumpy formation could be activated to achieve the same therapeutic outcome. Hence, compounds promoting developmental arrest in the parasite have value as biological tool compounds as well as offering novel approaches to disease control (20–22).
We have previously demonstrated that transgenic cell lines that utilize reporter genes (12) coupled to the 3′ untranslated region (UTR) of the PAD1 gene (Tb927.7.5930), a functional molecular marker for stumpy forms (23), can report on the response of monomorphic slender cells to conditions that promote the production of stumpy-like forms (16). Here, we have built on this system to develop a simple high-throughput assay for the detection of stumpy-like induction in monomorphic cell lines. This assay has been used to screen over 6,000 kinase-focused inhibitors for their ability to induce PAD1 3′ UTR-regulated reporter expression as a proxy for the induction of stumpy formation. This led to the identification of two structurally similar compounds that caused an unspecific increase in reporter expression as well as one chemically distinct compound that not only caused an upregulation of PAD1 mRNA expression but also generated consistent changes in mRNA expression for a small set of genes that are representative of partial stumpy formation in monomorphs and that also generate a stumpy-like phenotype in pleomorphic cell lines. This demonstrates the validity of a reporter screen for stumpy formation and of the use of high-throughput screening for the identification of small compounds that induce not only the cytocidal or cytostatic outcomes routinely analyzed but also specific phenotypic responses useful for the molecular analysis of trypanosome biology.
MATERIALS AND METHODSTrypanosoma brucei cell lines and culturing.
Lister 427 cells were transfected with the pHD617 glucuronidase (GUS)-PAD1 3′ UTR reporter construct modified from pHD617 chloramphenicol acetyltransferase (CAT)-PAD1 3′ UTR (16, 24) to generate the cell line 427 GUS-PAD1 3′ UTR used for compound screening. The previously described (16) 427 CAT-PAD1 3′ UTR GUS-Const 3′ UTR cell line was utilized for follow-up analysis (16). Trypanosomes were grown in HMI-9 medium with 20% fetal calf serum (FCS) at 37°C in 5% CO2 (25).
Differentiation experiments and assays of stumpy formation.
Differentiation of bloodstream forms to procyclic forms was induced by addition of 6 mM cis-aconitate and a temperature reduction from 37°C to 27°C. Differentiation capacity was determined by EP procyclin expression, measured by flow cytometry. Cell cycle analysis was carried out by 4′,6′-diamidino-2-phenylindole (DAPI) staining of fixed cells (100 ng/ml) and analysis via flow cytometry (26). For phase-contrast/DAPI microscopy, cells were fixed in ice-cold methanol and stained with 1 μg/ml DAPI. PAD1 protein expression was measured either by immunofluorescence microscopy with cells fixed for 10 min in 3% ice-cold paraformaldehyde (PFA) and stained using an anti-PAD1 antibody (23), by staining fixed cells for PAD1 protein and analyzing via flow cytometry, or by Western blot, carried out as previously described (23), with protein detected using Pierce ECL Western blotting substrate. 8-pCPT-cAMP was purchased from Sigma-Aldrich (United Kingdom).
Compound screening.
The compound screening was carried out at the Drug Discovery Unit at the University of Dundee using a kinase-focused compound library from the Scottish Hit Discovery Facility. A kinase inhibitor library was chosen for two reasons: first, kinases play an important role in many cell processes, including stumpy formation (20–22), and, second, kinases have been proposed as good drug targets in trypanosomes (27–29). The 427 GUS-PAD1 3′ UTR cell line was plated into 96-well plates at 5 × 105 cells/ml at 100 μl/well. Cells were treated for 48 h with test compounds at 50 μM in single cultures during primary screening and at 5.6 μM, 16.7 μM, and 50 μM in duplicate cultures during secondary screening in 0.5% (vol/vol) dimethyl sulfoxide (DMSO). This initial 50 μM concentration was chosen for primary screening based on standard screening procedures at the Drug Discovery Unit and because the positive control for induction of a stumpy-like form in vitro is routinely used at 100 μM 8-pCPT-cAMP (16) (Sigma, United Kingdom). Secondary screening was carried out in a titration series to assess hit potency although for tool compound development this is less important than phenotypic specificity. Screen performance was determined via internal controls (see Fig. 2), with a Z′ score (based on GUS enzyme assay) of 0.61 ± 0.13 for primary screening and 0.58 ± 0.12 for secondary screening (mean ± standard deviation). The purity and molecular mass of all hit compounds were determined by liquid chromatography-mass spectrometry.
GUS enzyme activity was measured via the addition of an equal volume of 4-methylumbelliferyl-β-glucopyranosiduronic acid (MUG; Sigma, United Kingdom) in lysis buffer (1 mM MUG, 0.82 M Tris-HCl, pH 8, 0.6% SDS, 0.3 mg/ml bovine serum albumin [BSA]) and measuring fluorescence at an excitation wavelength (λex) of 340 nm and emission wavelength (λem) of 460 nm after a 2-h incubation. Hit identification was based on results from a GUS enzyme assay only; however, for a measure of cell growth, 10% (vol/vol) alamarBlue was also added to the cells, and fluorescence was measured at λex of 530 nm and λem of 590 nm after a 4-h incubation (30). The alamarBlue assay was carried out immediately prior to the GUS assay, and no cross talk was observed. Fluorescence measured in GUS assays carried out during follow-up analysis was corrected for cell number. Addition of pCPT-cAMP, compound DDD00070762, or compound DDD00015314 to HMI-9 medium did not cause any increase in background fluorescence compared to DMSO-treated HMI-9 medium.
Validation analysis using a CAT-based reporter was also used. Here, CAT protein levels were determined by CAT enzyme-linked immunosorbent assay (ELISA; Roche) according to the manufacturer's instructions, with absorbance being measured using a BioTek ELx808 Absorbance Microplate Reader. Statistical analysis was carried out using Minitab, version 16, with data analyzed using general linear models (GLMs). Transformation of data was carried out to satisfy assumptions of normality. P values of less than 0.05 were considered significant.
RNA sequencing.
RNA samples for analysis by high-throughput sequencing of RNA transcripts (RNA-Seq) were harvested from CAT-PAD1 3′ UTR GUS-Const 3′ UTR cells after treatment with 50 μM DDD00015314 or 0.5% (vol/vol) DMSO, in duplicate, for 24 h. Library construction and sequencing were carried out by BGI-Hong Kong using an Illumina TruSeq RNA Sample Preparation Kit. The quality of raw sequence data was assessed using FastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/). Paired-end sequences were aligned to the Trypanosoma brucei brucei genome (obtained from ftp.sanger.ac.uk/pub4/pathogens/Trypanosoma/brucei/Latest_Whole_Genome_Sequence/Tb927_WGS_24_08_2012/chromosomes/) using Bowtie2 (parameters, very-sensitive-local; version 2.0.2); the outputs were filtered to remove alignments flagged with the XS Sequence Alignment/Map (SAM) flag, yielding an alignment data set that included only reads that map to a single location in the reference genome. These data were subsequently sorted and indexed using samtools. The annotated Trypanosoma brucei brucei genome was viewed using Artemis software (http://www.sanger.ac.uk/resources/software/artemis/ and http://ukpmc.ac.uk/abstract/MED/11120685), and coding segment (CDS) region coordinates were extracted prior to conversion to bedtools bedfile format. Bedtools (version 2.15.0; parameters, multicov, -bams) was used to generate coverages for each CDS for each sample replicate. Reads per kilobase per million reads mapped (RPKM) values were calculated by dividing coverage by CDS size using only reads that mapped to a single location in the reference genome (RPKM-like). Statistical analyses of the two sample groups were undertaken in the R environment (www.R-project.org) using Bioconductor (www.bioconductor.org) packages. Differential expression was explored using linear models and empirical Bayes methods, using the limma package (31). RPKM-like values were offset by 1, logged, and quantile normalized prior to group-wise comparison. Postcomparison, data were filtered to remove loci whose normalized mean RPKM-like values were below the 20% quantile for all samples.
Meta-analysis of the log2 expression data from Capewell et al. (32) and Jensen et al. (33) required identification of overlapping gene sets, achievable using systematic gene names. However, as there have been many improvements in both the sequence and annotation of the Trypanosoma brucei brucei 927 genome, some loci had been renamed, with the previous names retained in their annotation. Using simple parsing scripts, a table of current and previous gene names was generated, thereby facilitating data set comparisons.
Nucleotide sequence accession number.
RNA sequence data have been deposited in the NCBI Gene Expression Omnibus (GEO) repository under accession number GSE46483.
RESULTSValidation of the reporter system.
The 3′ UTR of the PAD1 gene has been shown, at least in part, to control its stumpy-specific gene expression (16). By coupling the PAD1 3′ UTR to a reporter gene, it is possible to monitor the increase in expression of reporter activity as an indicator of stumpy-like formation or the induction of stumpy-enriched mRNA expression. This method has been used previously for the analysis of known chemical inducers of stumpy-like formation via a CAT reporter gene (16), where treatment with cell-permeable cAMP (8-pCPT-cAMP) or a cAMP analogue (8-pCPT-2′-O-methyl-cAMP) caused an increase in CAT mRNA and protein expression as well as cell growth arrest and more efficient in vitro differentiation to procyclic forms after exposure to 6 mM cis-aconitate. Hence, these analogues drive the expression of several characteristics associated with physiological stumpy formation, there being an intersection between the molecules responsible for response to these analogues and the response to SIF itself (26). Nonetheless, it should be noted that while cell-permeable cAMP is able to induce upregulation of protein levels of reporter genes coupled to the PAD1 3′ UTR in monomorphic cell lines, it does not induce upregulation of the PAD1 protein itself (see Fig. S1 in the supplemental material), suggesting that there are additional levels of expression control operating on PAD1 in these cells.
High-throughput screening for reporter activation.
For the purposes of high-throughput screening, it was necessary to use a reporter that could be monitored quickly and cost effectively on a large scale. Hence, a β-glucuronidase reporter able to convert a nonfluorescent substrate into a fluorescent product was selected such that reporter activity could be measured in a rapid, simple, and economical assay. For the assay, a substrate/lysis buffer was added to whole cells, and fluorescence was measured, providing compatibility with a 96-well plate format. An expression vector was created that coupled the GUS reporter gene to the PAD1 3′ UTR (Fig. 1A), and this was transfected into T. brucei s427 monomorphic cells to create the 427 GUS-PAD1 3′ UTR cell line. Upon induction with 8-pCPT-2′-O-Me-cAMP (16, 18), these cells upregulated GUS reporter activity, confirming their ability to respond appropriately in the phenotypic screen (Fig. 1B). This also confirmed the ability of the PAD1 3′ UTR to respond to a stumpy trigger regardless of the reporter gene used.
Regulation of GUS reporter gene expression by the PAD1 3′ UTR. (A) Schematic representation of the pHD617 GUS-PAD1 3′ UTR reporter construct used to create the 427 GUS-PAD1 3′ UTR cell line used for compound screening. This vector was modified from those described in MacGregor et al. and Biebinger et al. (16, 24). Also shown are the pHD617 CAT-PAD1 3′ UTR and pHD617 GUS-Const 3′ UTR reporter constructs previously used to create the 427 CAT-PAD1 3′ UTR GUS-Const 3′ UTR cells (16), which are utilized in this study for follow-up analysis. The vector integrates into the ribosomal spacer locus via a ribosomal integration site sequence (RITS). (B) Monomorphic bloodstream form T. brucei cells transfected with the pHD617 GUS-PAD1 3′ UTR construct (427 GUS-PAD1 3′ UTR cells) showed an increase in GUS reporter activity when treated with 10 μM 8-pCPT-2′-O-Me-cAMP. Treatment with 0.5% (vol/vol) DMSO had no effect on reporter activity. Error bars represent standard errors of the means (n = 3).Puro, puromycin; Hygro, hygromycin.
The 427 GUS-PAD1 3′ UTR cell line was used to carry out a pilot screen of small molecules with the aim of identifying novel chemical inducers of stumpy formation. The screen was carried out at the Drug Discovery Unit at the University of Dundee using a kinase inhibitor-focused small-molecule library composed of 6,764 molecules (Fig. 2), according to the work flow outlined in Fig. 3A. First, in a primary screen, 100 μl of 5 × 105 cells/well of the 427 GUS-PAD1 3′ UTR cell line was incubated with a 50 μM concentration of each of the compounds from the library for 48 h in 96-well plates (Fig. 2). Prior to monitoring for GUS reporter activity, an alamarBlue assay (30) was carried out in order to provide a simple measure of cell proliferation; as stumpy formation is associated with cell cycle arrest (15), it was expected that any compound causing differentiation would also cause a decrease in proliferation of the population (data not shown). On all plates, wells treated with 100 μM 8-pCPT-cAMP were included as a positive control and as a benchmark for reporter activation (16, 17), with activation of the reporter and inhibition of cell proliferation induced by the 8-pCPT-cAMP set at 100% change. Given that this reports on a developmental transition (as opposed to enzymatic activity alone), physiologically relevant hit compounds would not be expected to generate a response significantly greater than that of the positive control (Fig. 2A). As such, compounds able to generate >53% GUS reporter activity, as determined through primary screening (Fig. 2B), were selected for secondary screening (mean Z′, 0.61) (Fig. 2A). This cutoff was selected to include the top ∼200 hits, a number suitable for secondary screening.
Scattergram of the results of primary screen. During primary screening, the 427 GUS-PAD1 3′ UTR cell line was treated with a 50 μM concentration of 6,764 compounds for 48 h. (A) In 96-well plates, eight wells on each plate contained 100 μM 8-pCPT-cAMP (positive control), and eight wells contained 0.5% (vol/vol) DMSO (negative control). The percent activation of the GUS reporter activity across all control wells during primary screening is shown, with the mean of all wells for 8-pCPT-cAMP treatment set at 100%. (B) Results of primary screening with 198 compounds; those showing >53% activation of GUS reporter activity were classified as possible hits.
Three hit compounds were identified through screening for the ability to induce stumpy-specific reporter gene expression. (A) Overview of workflow used to identify three hit compounds from a library of 6,764 small molecules. (B) Structures of three hit compounds. (C) Percent GUS reporter activity after 48 h of treatment as measured during secondary screening for each of the three hit compounds at 5.6 μM, 16.7 μM, and 50 μM, where 100% activation is that caused by 100 μM 8-pCPT-cAMP. (D) Percent inhibition of alamarBlue activity after 48 h of treatment as measured during secondary screening for each of the three hit compounds at 5.6 μM, 16.7 μM, and 50 μM, where 100% inhibition is that caused by 100 μM 8-pCPT-cAMP. Data points represent the mean of two replicates, with error bars representing the range.
The selected compounds were next subjected to secondary screening whereby each compound was tested in duplicate at 5.6 μM, 16.7 μM, and 50 μM (see Data Set S1 in the supplemental material). From secondary screening, 162 compounds generated fluorescence with a mean of >53% at 50 μM, providing a signal confirmation rate of the assay of 81.8% (mean Z′, 0.58). All 162 compounds were examined via mass spectrometry to confirm purity and identity. Compounds that failed this analysis were removed, leaving 115 compounds. Since many of these generated signals far greater than 8-pCPT-cAMP, it was anticipated that biologically irrelevant hits would comprise a significant proportion of these. Therefore, these 115 compounds were tested for autofluorescence in the absence of trypanosomes. This step determined that 112 of the 115 compounds exhibited autofluorescence under the conditions used in the screen (λex of 340 nm and λem of 460 nm) and, hence, likely represented false-positive results (see Data Set S1). To confirm that this group did not contain a significant proportion of biologically active autofluorescent compounds, a small selection of these 112 compounds were analyzed for their ability to induce stumpy reporter activation using an independent, nonfluorescent CAT reporter-based assay, with each exhibiting no activity although some of the compounds did affect cell growth (data not shown; see Fig. S2 in the supplemental material). Thus, to stringently select for compounds that represent biologically relevant hits with respect to the phenotypic screen used, all 112 autofluorescent compounds were excluded from further analysis, identifying three hits from a primary screen of 6,764 molecules (Fig. 3B). This apparently low hit rate of 0.04% highlights the selectivity of the screen for a differentiation-specific response; a far higher number of compounds (37% in the primary screen) caused trypanocidal or trypanostatic growth effects (as measured by alamarBlue assay), probably through multiple specific and nonspecific cellular mechanisms.
Reporter and phenotypic confirmation of screen outputs.
Relative to 100 μM 8-pCPT-cAMP, the three hit compounds identified were able to activate the GUS reporter gene at levels between 58% and 77% at 50 μM (Fig. 3C). Each compound also caused a decrease in alamarBlue activation, indicative of a reduction in population growth of between 31% and 123%, where 100% inhibition is that caused by 100 μM 8-pCPT-cAMP (Fig. 3D). Of the three compounds, DDD00070762 and DDD00070827 were structurally related (Fig. 3B), and the third was chemically distinct.
Of the two related compounds identified, compound DDD00070762 was chosen for validation and follow-up analysis. This compound demonstrated a 77% activation level of the GUS reporter activity during screening (Fig. 3C) and also caused reduced population growth as measured by an alamarBlue assay (Fig. 3D). For validation, 427 CAT-PAD1 3′ UTR GUS-Const 3′ UTR cells were used that have the CAT reporter gene under PAD1 3′ UTR control linked to a constitutively expressed GUS reporter (this provides a measure of constitutive gene expression in the presence of the compound in addition to stumpy-specific expression changes within the same cell line) (16). Treatment of these cells with 50 μM compound DDD00070762 stopped population growth within 24 h (Fig. 4A) and caused upregulation of CAT reporter gene expression after 24 and 48 h (Fig. 4B). This result confirmed that DDD00070762 promoted stumpy-specific reporter gene expression in an independent, non-fluorescence-based assay, validating the results of the high-throughput screen. It was also observed, however, that the control GUS reporter gene expression was highly variable between analyses (data not shown) and that within 24 h of treatment, >80% of the cells had become multinucleated (Fig. 4C). This contrasted with the small number of cells (<5%) that showed aberrant cell cycle types after 24 h of 8-pCPT-cAMP treatment, where multinucleated cells were not frequently observed. Indeed, while 8-pCPT-cAMP treatment causes monomorphic cells to display a somewhat atypical morphology after 48 h (Fig. 4D), this was distinct from the misshapen multinucleated cells that arose after only 24 h of DDD00070762 treatment. Hence, although the stumpy-linked reporter gene was upregulated by DDD00070762, this was not associated with other hallmarks of a physiological differentiation to stumpy-like forms. Rather, the compound appeared to cause considerable cell cycle and morphological defects unrelated to stumpy formation, and therefore DDD00070762 and the structurally related DDD00070827 were excluded from further analysis.
Compound DDD00070762 causes upregulation of stumpy reporter gene expression but not stumpy formation. 427 CAT-PAD1 3′ UTR GUS-Const 3′ UTR cells were treated with 50 μM DDD00070762. Negative-control cells were treated with 0.5% (vol/vol) DMSO; positive-control cells were treated with 100 μM 8-pCPT-cAMP. (A) Population growth was monitored at 24 and 48 h. Cells treated with DDD00070762 did not grow after 24 or 48 h. (B) CAT reporter gene expression, under the control of the PAD1 3′ UTR, increased after 24 and 48 h of treatment. GUS reporter gene expression, under the control of a constitutive 3′ UTR, showed variable expression levels between experiments after treatment (data not shown). (C) After 24 h of treatment, >80% of cells had an abnormal composition of DNA-containing organelles (nucleus [N] and kinetoplast [K]), from 250 cells per sample counted, in contrast with cells treated with 100 μM 8-pCPT-cAMP, where <5% of cells had aberrant DNA content after 24 h. (D) Examples of cells treated with DMSO, 8-pCPT-cAMP, or DDD00070762 for 24 h and 48 h are shown. Cells were stained with DAPI. Scale bar, 10 μm. Error bars in all graphs represent standard errors of the means (n = 3).
Compound DDD00015314, N-(3-bromophenyl)-3-[(3-oxo-4H-1,4-benzoxazin-6-yl)sulfonyl] propanamide, did not show any structural similarities to either of the other two hits from the screen (Fig. 3B). This compound demonstrated a 68% activation of the GUS reporter activity during screening (Fig. 3C) and also caused a modest reduction in population growth, as measured by the alamarBlue assay (Fig. 3D). For validation, 427 CAT-PAD1 3′ UTR GUS-Const 3′ UTR cells were treated with a 50 μM concentration of the compound. As in the primary screen, this caused a modest reduction in population growth (Fig. 5A) to 43.6% of that of negative controls over 48 h. Also, a 2.5-fold increase in stumpy-specific reporter gene expression was reproducibly observed after 24 h of treatment (Fig. 5B) (DMSO-treated cells versus DDD00015314-treated cells at 24 h, F1,4 = 104.51, P = 0.001), which was not accompanied by any corresponding increase in the constitutive GUS reporter (Fig. 5C) (DMSO-treated cells versus DDD00015314-treated cells at 24 h, F1,4 = 4.92, P = 0.091). This is distinct from 8-pCPT-cAMP treatment, which results in reduced levels of GUS reporter expression, likely reflecting cellular quiescence of the treated parasites (34). This result demonstrated that DDD00015314 generated specific activation of the stumpy reporter although no morphological changes in the treated cells were observed by microscopy (data not shown). Hence, the overall response appeared to be less robust than with 8-pCPT-cAMP, but the compound nonetheless generated specific and reproducible activation of the reporter expression in monomorphic cells.
Compound DDD00015314 specifically causes a modest upregulation of stumpy-specific reporter gene expression without causing upregulation of constitutive reporter gene expression in monomorphic cells. 427 CAT-PAD1 3′ UTR GUS-Const 3′ UTR cells were treated with 50 μM DDD00015314. Negative-control cells were treated with 0.5% (vol/vol) DMSO; positive-control cells were treated with 100 μM 8-pCPT-cAMP. (A) Population growth was monitored at 24 and 48 h. Cells treated with DDD00015314 showed reduced growth compared to DMSO-treated controls. (B) CAT reporter gene expression, under the control of the PAD1 3′ UTR, increased after 24 and 48 h of treatment. (C) GUS reporter gene expression, under the control of a constitutive 3′ UTR, was slightly reduced after 24 h of treatment with DDD00015314; however, at 48 h GUS reporter gene expression was again equivalent to that of DMSO-treated controls. Error bars in all graphs represent standard errors of the means (n = 3).
The specific increase in stumpy reporter gene expression induced by DDD00015314 suggested that the compound promoted some aspects of development to stumpy forms or stumpy-enriched gene expression. To investigate this, transcriptome analysis via RNA-Seq was carried out on treated and control cells to monitor global changes in gene expression in response to compound exposure. RNA samples were harvested from cells after treatment, in duplicate, for 24 h with 50 μM DDD00015314 or 0.5% (vol/vol) DMSO. The resulting total RNA was then subjected to RNA-Seq analysis following Illumina TruSeq RNA Sample Preparation (BGI, Hong Kong). Over 13 million 90-bp reads were generated per sample, with a quality score of 20 (Q20; i.e., error rate of ≤1%) in >95.5% of all samples. Raw sequence data were aligned to the Trypanosoma brucei brucei 927 genome, and reads per kilobase per million mapped reads (RPKM) were calculated for each CDS for each sample replicate, including only reads that map to a single location in the reference genome.
Pairwise comparison between the populations treated with DDD00015314 or 0.5% (vol/vol) DMSO revealed a correlation of 0.995 (Fig. 6A), suggesting that treatment with 50 μM DDD00015314 caused little general perturbation of gene expression, despite the slowed growth of the treated cells, and revealing considerable consistency in the respective sample replicates. Nevertheless, 95 genes were identified as upregulated, and 71 genes were downregulated, with a ≥1.25-fold change after DDD00015314 treatment (see Data Set S2 in the supplemental material) and a P of <0.05 before adjustment for multiple testing, with no genes showing a significant change after adjustment for multiple testing. Table 1 shows a truncated list of genes up- and downregulated after DDD00015314 treatment, with all hypothetical (47/95 upregulated, 27/71 downregulated), variant surface glycoprotein (VSG; 3/95 upregulated, 6/71 downregulated), and retrotransposon hot spot protein (5/95 upregulated) genes removed for clarity.
RNA-Seq on DDD00015314-treated cells reveals modest but specific changes in gene expression. (A) Scatter plot of log-normalized mean reads per kilobase per million mapped reads (RPKM) for control and DDD00015314-treated cells based on two replicates, with values below the 20% quartile removed. The PAD1 transcript is highlighted with a diamond. (B) Histogram showing absolute fold change of mRNA levels for selected genes that were up- and downregulated after DDD00015314 treatment. PCF, procyclic form.
A truncated list of genes upregulated and downregulated in response to DDD00015314 treatmenta
Gene
Descriptionb
Groupc
Fold change in expression
Fold change in slender vs stumpy forms according to:d
Zinc finger protein family member, putative (ZC3H36)
RNA binding protein
−1.40
−3.16
1.32
Tb927.9.11760
DNA repair protein RAD2, putative
Other
−1.30
1.38
1.07
Tb927.7.6220
Protein kinase, putative
Other
−1.38
−2.29
1.11
Tb927.9.10170
Predicted zinc finger protein (homolog of Tbg APC11)e
Other
−1.33
−1.08
1.12
After 24 h of DDD00015314 treatment, 95 genes were upregulated, and 73 genes were downregulated greater than 1.2- fold. Further, the CAT reporter gene coupled to the PAD1 3′ UTR showed a 1.47-fold increase in mRNA expression, similar to the 1.44-fold increase in PAD1 mRNA expression (highlighted in boldface). Here, hypothetical, VSG, and retrotransposon hot spot protein genes have been removed for clarity. A full list of genes (with associated statistics) up- and downregulated after DDD00015314 treatment is provided in Data Set S2 in the supplemental material. Product assignments were based on annotations and comments at TriTrypDB.org.
ESAG, expression site-associated gene; PAD, protein associated with differentiation; CoA, coenzyme A; PSSA, procyclic form surface phosphoprotein; LRRP, leucine-rich repeat protein; PGKC, phosphoglycerate kinase.
±, with or without.
The absolute fold change in transcript abundance during complete differentiation from slender to stumpy forms as observed in two independent analyses (32, 33) is also shown for each gene. In Capewell et al. (32) PAD1 was not distinguishable from other PAD family transcripts. NA, not available.
Tbg, Trypanosoma brucei gambiense.
As expected, complete differentiation from slender to stumpy forms did not occur upon DDD00015314 treatment; differentiation to stumpy forms is associated with large-scale changes in gene expression (32, 33, 35, 36), whereas DDD00015314 treatment caused changes in mRNA expression only for a limited number of genes. Nonetheless, of the changes that did occur, many were associated with transcript changes known to occur during stumpy formation, and clear expression trends were observed with functionally coordinated transcripts being coregulated (Table 1 and Fig. 6B). Further, while these changes appear modest (i.e., <2-fold change between treated and untreated cells), the changes observed during complete physiological differentiation from slender to stumpy forms are also in this range (Table 1, columns 5 and 6; see also Table S1 in the supplemental material) (32, 33). As the cell line used for this transcript analysis expressed the CAT reporter gene coupled to the PAD1 3′ UTR, it was also possible to monitor the mRNA expression levels of the CAT reporter gene as well as the endogenous PAD1 gene itself. This revealed good agreement between the two transcripts, with a 1.47-fold upregulation of CAT mRNA in response to DDD00015314 treatment and a 1.44-fold increase in PAD1 mRNA expression (Table 1). A corresponding upregulation of the PAD1 protein was not expected since the expression of this protein appears to be restricted in monomorphic cells (see Fig. S1 in the supplemental material). Supporting specificity of the observed changes, the mRNA expression level of the constitutively expressed GUS reporter in the same cell lines was unchanged compared to that in untreated cells (1.06-fold difference).
Analyzing the regulated changes after exposure of cells to DDD00015314 demonstrated the elevation of several stumpy-enriched transcripts. These included PAD2, known to be upregulated in stumpy and procyclic forms (23) (upregulated 1.37-fold), and two ESAG9 genes (Tb927.7.170 and Tb927.9.7340), also known to be enriched in stumpy forms (37). Similarly, the mRNA levels of two procyclic form surface proteins (Tb927.6.150 and Tb927.10.11220) were also increased, consistent with their elevation in stumpy forms in preparation for expression upon differentiation to procyclic forms in the tsetse midgut. Indeed, GPEET is the predominant surface protein found on early procyclic forms (38), and multiple genome-wide analyses have shown that mRNA expression of Tb927.6.150, a GPEET 2 precursor, is upregulated in stumpy forms compared to levels in slender forms (33, 35, 36). Similarly, Tb927.10.11220, or PSSA-2, is a procyclic-form phosphoprotein that has been shown by Northern blotting to have greatly increased mRNA expression in procyclic forms compared to bloodstream slender forms (39), which is preceded by its upregulation also in stumpy forms compared to slender forms (33, 35, 36).
Most notably, a number of genes associated with metabolism were seen to have altered mRNA expression after DDD00015314 treatment (Fig. 6B). Slender bloodstream form trypanosomes have a repressed mitochondrion and generate ATP solely from glucose through glycolysis, with most of this pathway localized within the glycosomes (40–42). Stumpy forms, however, have a partially elaborated mitochondrion in preparation for differentiation to the procyclic form (43), which has a fully developed mitochondrion and can metabolize amino acids, particularly proline, as well as glucose. Notably, DDD00015314 treatment caused upregulation of 10 mitochondrial proteins and five proteins involved in amino acid metabolism or transport, including proline dehydrogenase and glutamate dehydrogenase (Tb927.7.120 and Tb927.9.5900, 1.30-fold and 1.32-fold upregulated, respectively). Conversely, there was a downregulation of 10 glucose transporter gene transcripts and 13 glycosomal proteins, including proteins that play a role in glycolysis (Table 1 and Fig. 6B). Thus, DDD00015314 treatment appears to cause a shift in the expression of several metabolic genes indicative of early steps in differentiation in monomorphic cells.
DDD00015314 promotes the development to stumpy forms in pleomorphic parasites.
Given that DDD00015314 was able to induce a partial differentiation phenotype in monomorphic trypanosomes, we investigated whether this compound would have a similar, or more pronounced, effect on pleomorphic cells naturally competent for stumpy formation. Therefore, a pleomorphic T. brucei brucei AnTat 1.1 cell line was treated with 50, 10, or 5 μM DDD00015314 for up to 48 h. At all concentrations tested, DDD00015314 caused decreased population growth (Fig. 7A) and accumulation in G1/G0 (Fig. 7B) equivalent to that induced by 100 μM 8-pCPT-cAMP. Since stumpy forms have an increased capacity for differentiation to procyclic forms compared to slender forms, cells that had been compound treated for 24 h were then induced to differentiate to procyclic forms by the addition of 6 mM cis-aconitate and a temperature reduction to 27°C. Figure 7C shows that DDD00015314-treated cells showed an increased level of differentiation after 6 h, with a mean differentiation after 50 μM, 10 μM, and 5 μM DDD00015314 treatment of 49.3%, 57.4%, and 59.7%, respectively, compared to 29.2% in cells treated with DMSO alone (Fig. 7C). This was not, however, as effective as 8-pCPT-cAMP exposure, nor did DDD00015314 treatment produce the levels of differentiation observed in in vivo-derived stumpy forms, which had a mean differentiation of 69.8% and 88.2%, respectively. Moreover, by 24 h postinduction of differentiation, DMSO-treated controls exhibited a higher level of EP procyclin-positive parasites than DDD00015314-treated cells, perhaps indicating that the treatment allows cells to embark on the differentiation pathway but not sustain a proliferative procyclic population.
Compound DDD00015314 causes a stumpy-like phenotype in pleomorphic bloodstream form cells. AnTat1.1 90:13 cells were treated with 50 μM, 10 μM, or 5 μM DDD00015314. Negative-control cells were treated with 0.5% (vol/vol) DMSO; positive-control cells were treated with 100 μM 8-pCPT-cAMP. (A) Population growth was monitored at 24 and 48 h. Cells treated with DDD00015314 or 8-pCPT-cAMP showed reduced growth compared to DMSO-treated controls. (B) Treatment with DDD00015314 or 8-pCPT-cAMP caused an accumulation of cells in G1/G0 after 24 and 48 h, as determined by flow cytometry of DAPI-stained populations. (C) After 24 h of treatment, cells were induced to differentiate to procyclic forms by the addition of 6 mM cis-aconitate and a temperature change from 37°C to 27°C. Cells treated with DDD00015314 or 8-pCPT-cAMP differentiated more efficiently than DMSO-treated cells by 6 h postinduction, as determined by expression of EP procyclin. At 24 h however, cells treated with DDD00015314 had differentiated less well than DMSO-treated cells. CCA, cis-aconitate. (D) Treated cells were fixed and stained for PAD1 protein expression and analyzed by flow cytometry. Cells treated with DDD00015314 or 8-pCPT-cAMP showed upregulation of PAD1 protein expression compared to negative controls after 24 and 48 h although to a lesser degree than that observed from a stumpy form positive control. FITC, fluorescein isothiocyanate. (E) Immunofluorescence microscopy of cells after 24 h of treatment confirms upregulation of PAD1 protein expression in DDD00015314- or 8-pCPT-cAMP-treated cells. Two examples are shown per treatment. In each pair, the left panel shows a phase-contrast with DAPI image and the right panel shows the same cells stained for PAD1 protein. 1° Ab, primary antibody.
As a final measure of the development to stumpy forms induced by DDD00015314, we analyzed the expression of the stumpy-specific marker protein, PAD1, by both flow cytometry (Fig. 7D) and immunofluorescence microscopy (Fig. 7E). With both methods, an increase in PAD1 protein expression in the population was observed after treatment with 8-pCPT-cAMP and at all concentrations of DDD00015314 tested. Further, microscopy revealed the presence of cells with intermediate and stumpy-like morphology after 24 and 48 h of treatment with either 8-pCPT-cAMP or DDD00015314 (Fig. 7E and Fig. 8). After 48 h of 50 μM DDD00015314 treatment, however, cells began to look misshapen. Thus, although the effects of DDD00015314 are modest in monomorphic cells, treatment of pleomorphic cells with this compound has substantial effects, generating an accumulation in G1/G0, expression of the PAD1 protein, increased capacity for the initiation of differentiation to procyclic forms, and morphological changes associated with stumpy formation.
Compound DDD00015314 causes a stumpy-like morphology in pleomorphic bloodstream form cells. Microscope images of treated cells reveal that DDD00015314 or 8-pCPT-cAMP treatment induces a stumpy-like phenotype after 24 h. This morphology remains apparent in cells treated with 5 or 10 μM DDD00015314 or with 8-pCPT-cAMP after 48 h, but many cells treated with 50 μM DDD00015314 have an aberrant morphology after 48 h.
In conclusion, from a pilot screen of over 6,000 compounds from a kinase-targeted library, a compound has been identified that is able to specifically activate a reporter for stumpy-enriched gene expression in monomorphic cell lines and induce a stumpy-like phenotype in pleomorphic cell lines, as characterized by a number of diagnostic assays. Moreover, an analysis of the transcriptome of treated monomorphic cells revealed a modest but consistent elevation of a number of transcripts known to be regulated during the development to stumpy forms in vivo. This study, therefore, validates high-throughput screening as an approach to identify developmental regulators in trypanosome parasites. Moreover, a novel tool compound applicable for dissection of components of this developmental pathway is identified.
DISCUSSION
Whole-cell-based high-throughput screening has been successfully utilized in Trypanosoma brucei research in the search for compounds with trypanocidal or trypanostatic activity, with the aim of identifying new lead compounds for the development of novel therapeutics against African trypanosomiasis (4–10). Here, we explored the potential for using a similar screening method for the identification of compounds able to induce other trypanosome phenotypes, namely, differentiation between life stages, providing possible tool compounds of value in research, as well as therapeutics exploiting the irreversible proliferation arrest associated with the transition to stumpy forms. The fluorescence-based approach developed was a rapid, inexpensive, and simple assay able to be carried out in 96-well plate format, and autofluorescent library compounds could be rapidly excluded in a cell-free assay pre- or postscreening to identify cytologically relevant hits. As such, over 6,000 compounds could be screened on fewer than 100 plates, requiring less than 1 liter of trypanosome culture. Smaller-scale (and hence lower-cost) validation screening using a non-fluorescence-based reporter that also allowed reporter-specific effects to be distinguished from general gene expression effects was then exploited to unambiguously identify compounds able to drive the developmental response under test.
From a starting point of 6,764 compounds, this screening method successfully identified three compounds able to induce upregulation of PAD1 reporter gene expression as a proxy for stumpy formation, and these were validated through follow-up analysis. This confirmed the efficacy of the approach for detecting specific cytological phenotypes distinct from generalized cytostatic and cytotoxic effects likely to be elicited by a much broader range of hit compounds. Two of the three hit compounds were structurally very similar; the identification of both of these molecules from a large compound set illustrates the consistency and reproducibility of the method. Follow-up analysis confirmed that the increase in PAD1 reporter gene expression induced by one of these compounds (compound DDD00070762) was consistent and reproducible. However, it did not represent a specific physiological progression to a stumpy or stumpy-like form since the constitutive reporter employed in the secondary screen exhibited similar (though inconsistent) elevation and since the exposed cells rapidly became multinucleated. Hence, these compounds may target one or more processes in the cell that generate phenotypes distinct from, or in addition to, a specific developmental response. In contrast, the third compound identified, DDD00015314, did cause specific upregulation of both PAD1 reporter gene activity and endogenous PAD1 mRNA levels, without activation of the constitutive reporter in monomorphic lines. Moreover, upon treatment with DDD00015314, the cells underwent a change in expression of a restricted set of genes, many of which are also elevated upon stumpy formation in vivo. In particular, several genes relating to energy metabolism were elevated (mitochondrial genes and enzymes involved in proline metabolism), whereas the mRNA levels of a number of genes involved in glucose transport and glycolysis were downregulated. These responses were limited in extent (albeit similar in scale to those seen in the natural developmental process) but reproducible and indicative of a progression toward the stumpy phenotype in the monomorphic lines used in the high-throughput screen. Treatment of pleomorphic cells with the same compound had a more pronounced effect, even at lower concentrations: cells displayed a stumpy-like phenotype with an accumulation of cells in G1/G0, expression of PAD1 protein, and the presence of cells with an intermediate or stumpy morphology within the population. Although the response was not entirely physiological since the elevation of EP procyclin expression at 6 h after exposure to cis-aconitate was diminished after 24 h, the data indicate that compound DDD00015314 is able to induce partial stumpy formation or activate/inhibit some components of the stumpy formation pathway.
Cell-permeable hydrolysable cAMP and AMP analogues have been utilized for the induction of some, albeit not all, stumpy characteristics in monomorphic cells, providing reagents to analyze some events of stumpy formation. Recently, a whole-genome RNAi screen was designed to identify genes required for responsiveness to cAMP and AMP analogues in monomorphs, resulting in the identification of multiple genes involved in physiological stumpy formation in vivo. These represented the first known positive inducers of stumpy formation encompassing molecules from throughout the signaling cascade, including kinases, phosphatases, and at least one predicted RNA binding protein (26). Clearly, the outputs from high-throughput small-compound screens such as the screen described here could provide reagents to further dissect this pathway or other uncharacterized signaling pathways, for example, by assisting in the ordering of components following further genome-wide RNAi screens. Indeed, using multiple and distinct chemical inducers of a full or partial stumpy-like phenotype would allow iterative RNAi screens, with resistance phenotypes being observed only for molecules falling downstream of where the compounds act on the signaling pathway. As compound DDD00015314 appears to induce partial differentiation toward stumpy formation, it could be targeting only some aspects of the SIF signaling pathway, such that a whole-genome RNAi screen with this compound could identify genes involved in a specific branch of a stumpy formation cascade as well as those shared with the response to cell-permeable cAMP/AMP. Currently, the limited availability of compound DDD00015314 and its expense preclude such analysis with a monomorphic RNAi library (44); however, the enhanced sensitivity of pleomorphic cells could permit this analysis if RNAi libraries could be generated at sufficient coverage in these cell lines. Alternatively, DDD00015314 could be chemically refined to increase its potency and specificity in order to potentially achieve activity at lower concentrations. This provides a starting point, and proof of principle, for the development of further tools for the study of trypanosome differentiation and developmental gene regulation.
Compounds driving complete or incomplete stumpy formation also offer potential as a complementary or adjunctive approach to trypanocidal therapeutics, generating a reduction of parasite virulence or transmission between hosts. Hence, accelerated differentiation of slender forms to stumpy forms within the bloodstream could reduce transmission in epidemic situations by reducing the density of stumpy forms below that needed for effective tsetse passage. Perhaps more importantly, where a complete differentiation of parasites into stumpy forms could be achieved in the bloodstream, a trypanostatic response (stumpy formation) would inevitably lead to a trypanocidal outcome through the action of the immune response (45), clearing the infection. Unlike conventional therapies, this approach could also be evolution resistant since emergent resistant parasites, while more virulent in the originating host, would have reduced transmissibility through their reduced rate of stumpy formation, preventing the propagation of resistance in a population.
Supplementary Material
Supplemental material
Published ahead of print 17 January 2014
Supplemental material for this article may be found at http://dx.doi.org/10.1128/EC.00335-13.
ACKNOWLEDGMENTS
This work was funded by a high-throughput screening grant from the Scottish Universities Life Science Alliance and a Wellcome Trust Programme grant (088293MA) to K.R.M. and by a Wellcome Trust strategic award (095831MA) to the Centre for Immunity, Infection, and Evolution.
Compound screening was carried out at the Drug Discovery Unit at the University of Dundee using a compound library from the Scottish Hit Discovery Facility (http://www.sulsa.ac.uk/facilities/SHDF).
Temperature is a ubiquitous environmental variable which can profoundly influence the physiology of living cells as it changes over time and space. When yeast cells are exposed to a sublethal heat shock, normal metabolic functions become repressed and the heat shock transcription factor Hsf1 is activated, inducing heat shock proteins (HSPs). Candida albicans, the most prevalent human fungal pathogen, is an opportunistic pathogen that has evolved as a relatively harmless commensal of healthy individuals. Even though C. albicans occupies thermally buffered niches, it has retained the classic heat shock response, activating Hsf1 during slow thermal transitions such as the increases in temperature suffered by febrile patients. However, the mechanism of temperature sensing in fungal pathogens remains enigmatic. A few studies with Saccharomyces cerevisiae suggest that thermal stress is transduced into a cellular signal at the level of the membrane. In this study, we manipulated the fluidity of C. albicans membrane to dissect mechanisms of temperature sensing. We determined that in response to elevated temperature, levels of OLE1, encoding a fatty acid desaturase, decrease. Subsequently, loss of OLE1 triggers expression of FAS2, encoding a fatty acid synthase. Furthermore, depletion of OLE1 prevents full activation of Hsf1, thereby reducing HSP expression in response to heat shock. This reduction in Hsf1 activation is attributable to the E3 ubiquitin ligase Rsp5, which regulates OLE1 expression. To our knowledge, this is the first study to define a molecular link between fatty acid synthesis and the heat shock response in the fungal kingdom.
INTRODUCTION
Microorganisms inhabit dynamic environments in which they are continually exposed to environmental stimuli and stresses. Survival depends upon effective environmental response strategies that have been uniquely tuned over evolutionary time. By reacting to environmental changes via a sense and respond logic, cells continuously monitor their environment and coordinate appropriate cellular responses to specific stimuli (1). These cellular response strategies have been intensively studied for a variety of model organisms (2–5). Fundamentally, organisms utilize a myriad of signaling pathways that drive physiological adaptation to diverse environmental stresses, including temperature fluctuations and osmotic, oxidative, and weak acid stresses, as well as nutrient limitation (6, 7).
Temperature is a ubiquitous environmental variable, which can profoundly influence the physiology of living cells as it changes over time and space. When yeast cells are exposed to a sublethal heat shock, normal metabolic functions become repressed, and genes encoding heat shock proteins (HSPs) are induced (8). This induction occurs through the evolutionarily conserved heat shock transcription factor Hsf1. How the cell senses changes in ambient temperature, thus triggering Hsf1 activation, was thought to be driven by the accumulation of denatured proteins (9). However, a sustained heat shock would give rise to ongoing protein denaturation, leading to continual activation of HSPs. Instead, the transient nature of the heat shock response suggests that the thermal sensor becomes desensitized, allowing cells to adapt to a new basal level of activity (10, 11).
The cellular membrane serves as a direct sensor of its organism's environment. Its lipid structure is key to determining the physicochemical environment of the membrane, with the molecular packing of lipids acting as a direct determinant of membrane fluidity. At physiological temperatures, the fluidity of the membrane is increased through disorganized, unpacked unsaturated fatty acids, whereas saturated fatty acids remain tightly packed and retain a higher melting temperature (12, 13). Consequently, lipid saturation is regulated in response to changes in temperature. The fluidity of the membrane is governed by an intricate balance between saturated and unsaturated fatty acids. In the model yeast Saccharomyces cerevisiae, after the initial reaction of fatty acid synthesis, elongation of the carbon chain is catalyzed by the fatty acid synthases Fas1 and Fas2 to produce the saturated fatty acids palmitic acid (16:0) and stearic acid (18:0) (14). These are then converted to the monounsaturated fatty acids palmitoleic acid (16:1) and oleic acid (18:1) by the fatty acid Δ9 desaturase (Ole1), located at the surface of the endoplasmic reticulum (ER), by introducing double bonds at carbon 9 in the carbon chains (15).
The degree of unsaturation in the S. cerevisiae cellular membrane is highest at 15°C and lowest at 37°C, with fatty acid chains becoming longer with increasing temperature (16). But how does this affect the heat shock response? Studies on the fungal pathogen Histoplasma capsulatum established that addition of increasing concentrations of palmitic acid, a saturated fatty acid, paired with a temperature upshift upregulated transcription of the heat shock protein HSP82; conversely, addition of oleic acid, an unsaturated fatty acid, decreased heat shock gene transcription upon a rise in temperature (17). Furthermore, an S. cerevisiae mutant that lacks the fatty acid desaturase gene OLE1 was complemented using constructs with the native (S. cerevisiae) OLE1 promoter or the OLE1 promoter from a temperature-tolerant or temperature-sensitive H. capsulatum strain, with the resulting strains each displaying differential expression of OLE1 (18). Depending on the promoter used, complemented strains adjusted the physiology of the membrane by modifying the ratio of saturated to unsaturated fatty acids. Subsequently, each mutant displayed a different threshold temperature of heat shock gene expression. For instance, the temperature-sensitive strain complemented with the OLE1 promoter that upregulates OLE1 expression displayed a dramatic decrease in palmitic acid, as well as HSP expression (17).
Candida albicans, one of the most prevalent human fungal pathogens, is an opportunistic pathogen that has evolved as a relatively harmless commensal of the mucous membranes and digestive tracts of healthy individuals (19, 20). C. albicans frequently causes infections of mucosal membranes (thrush), and in immunocompromised patients this yeast can cause life-threatening systemic infections (19, 21). Even though C. albicans is obligately associated with warm-blooded mammals and occupies thermally buffered niches, it has retained the classic heat shock response (22). Indeed, Hsf1 is essential for viability in C. albicans, reflecting the fundamental importance of heat shock adaptation in all organisms (22–24). Our recent exploration of the dynamic regulation of Hsf1 during thermal adaptation has suggested that Hsf1 is activated even during slow thermal transitions such as the increases in temperature suffered by febrile patients (11). However, the mechanisms by which C. albicans senses changes in ambient temperature and activates Hsf1 remain an enigma.
In this study, we explore whether C. albicans utilizes membrane fluidity as a direct sensor of temperature. We determined that cells alter the levels of the fatty acid desaturase Ole1 in response to increased temperature and that loss of OLE1 induces expression of FAS2, encoding a fatty acid synthase. We show that depletion of OLE1 prevents full activation of Hsf1, contributing to a reduction in HSP expression in response to heat shock, and that this reduction in Hsf1 activation is attributable to the E3 ubiquitin ligase Rsp5, which regulates OLE1 through the transcription factor Spt23. Therefore, we have established for the first time in the fungal kingdom a molecular link between membrane fluidity and the heat shock response.
MATERIALS AND METHODSStrains and growth conditions.
All strains used are listed in Table 1. Strains were grown in YPD (1% yeast extract, 2% Bacto peptone, 2% glucose) (25). To impose an instant heat shock of 30°C to 42°C, cells were grown in YPD at 30°C to exponential phase and mixed with an equal volume of medium that had been prewarmed to 54°C in flasks that had been prewarmed at 42°C. Cells were grown at 42°C for the times indicated below. Doxycycline was added to YPD medium at a concentration of 1 μg/ml or 20 μg/ml.
To regulate oleic acid levels, one copy of the OLE1 gene was deleted, and the other was placed under the control of the tetO promoter. Briefly, the NAT (nourseothricin) flipper cassette (pLC49) was amplified with oLC2798/oLC2799 containing regions of homology upstream and downstream of OLE1. The PCR product was transformed into the wild-type strain SN95 (CaLC239), and NAT-resistant transformants were PCR tested with oLC275/oLC2802 and oLC274/oLC2803 to verify integration of the cassette. The NAT cassette was then excised to create CaLC2851. The tetracycline-repressible transactivator, the tetO promoter, and the NAT flipper cassette were PCR amplified from pLC605 using oligonucleotides oLC2801 and oLC2798. The PCR product was transformed into CaLC2851. Correct upstream and downstream integration was verified by amplifying across both junctions by colony PCR using primer pairs oLC2802/oLC275 and oLC2804/oLC300, respectively. Loss of the wild-type band was verified using primer pair oLC2802/oLC2804. Hsf1 was then tagged in this strain as described below to create CaLC3032.
To regulate RSP5, one copy of the RSP5 gene was deleted, and the other was placed under the control of the tetO promoter. Briefy, the NAT flipper cassette (pLC49) was amplified with oLC3359/oLC3360 containing regions of homology upstream and downstream of RSP5. The PCR product was transformed into the wild-type strain SN95 (CaLC239), and NAT-resistant transformants were PCR tested with oLC275/oLC3361 and oLC274/oLC3362 to verify integration of the cassette. The NAT cassette was then excised to create CaLC3032. The tetracycline-repressible transactivator, the tetO promoter, and the NAT flipper cassette were PCR amplified from pLC605 using oligonucleotides oLC3357/oLC3388. The PCR product was transformed into CaLC3302, and NAT-resistant transformants were PCR tested with oLC534/oLC3361 (upstream), oLC300/oLC3389 (downstream), and oLC3361/oLC3389 (wild type) to verify integration of the cassette, the NAT cassette was then excised. Hsf1 was then tagged in this strain as described below to create CaLC3366.
To determine Hsf1 phosphorylation status, Hsf1 was tagged with the TAP (tandem affinity purification) tag at its C terminus in the wild-type strain SN95 (creating CaLC2993) (Table 1) and in the tetO-HSF1/hsf1Δ strain to confirm functionality of the tagged allele, using a PCR-based strategy as described previously (26). Briefly, the tag and a selectable marker (ARG4) were PCR amplified from pLC573 (pFA-TAP-ARG4 [27]) using oligonucleotides oLC2950/2922 (see Table S1 in the supplemental material). Fifty microliters of PCR product was transformed into C. albicans. Correct genomic integration was verified using appropriate primer pairs that anneal ∼500 bp upstream (oLC1597) or downstream (oLC1598) from both insertion junctions together with oLC1593 (TAP-R) and oLC1594 (ARG4-F), which target the TAP and the selectable marker (see Table S1).
qRT-PCR.
To monitor gene expression changes in response to OLE1 or RSP5 depletion, strains SN95 (CaLC2993), CaLC3032 (tetO-OLE1/ole1Δ), and CaLC3366 (tetO-RSP5/rsp5Δ) were grown overnight at 30°C in YPD, with shaking at 200 rpm. Stationary-phase cultures were split and adjusted to an optical density at 600 nm (OD600) of 0.1; one culture was treated with doxycycline (BD Biosciences), while the other was left untreated. Cells were grown for 4 h at 30°C. To monitor gene expression changes in response to heat shock, wild-type (CaLC2993) and tetO-OLE1/ole1Δ (CaLC3032) cells were grown to mid-log phase and subjected to a 30°C to 42°C heat shock, and 50 ml was harvested from each culture at the specified time, centrifuged at 3,000 rpm for 2 min at 4°C, and washed once with distilled water (dH2O) before being frozen at −80°C. RNA was then isolated using the Qiagen RNeasy kit, and cDNA synthesis was performed using the AffinityScript cDNA synthesis kit (Stratagene). PCR was carried out using the SYBR green JumpStart Taq ReadyMix (Sigma-Aldrich) under the following cycle conditions: 95°C for 3 min, 95°C for 10 s, and 60°C for 30 s for 39 rounds, 95°C for 10 s, and 65°C for 5 s. All reactions were done in triplicate using the following primer pairs: for HSP104, oLC1620/oLC1621; for HSP21, oLC3217/oLC3218; for OLE1, oLC2805/oLC2806; for RSP5, oLC3408/oLC3409; and for FAS2, oLC3168/oLC3169. Transcript levels were normalized to ACT1 (oLC2285/oLC2286) (see Table S1 in the supplemental material). Data were analyzed using Bio-Rad CFX Manager software, version 3.1 (Bio-Rad).
Western blotting.
Total soluble protein was extracted and subjected to Western blotting using published protocols (28, 29). Briefly, mid-log-phase cells were pelleted by centrifugation, washed with sterile water, and resuspended in lysis buffer (0.1 M Tris-HCl [pH 8.0], 10% glycerol, 1 mM dithiothreitol [DTT], phenylmethylsulfonyl fluoride [PMSF], and protease inhibitor cocktail). An equal volume of 0.5-mm acid-washed beads was added to each tube. Cells were mechanically disrupted on a BioSpec (Bartlesville, OK) Mini-Beadbeater for six 30-s periods, with 1 min on ice between each cycle. The lysate was pelleted by high-speed centrifugation and the supernatant removed for analysis. Protein concentration was determined using a Bradford reagent (Sigma-Aldrich) assay. Protein samples were mixed with one-sixth volume of 6× sample buffer (0.35 M Tris-HCl, 10% [wt/wt] SDS, 36% glycerol, 5% β-mercaptoethanol, and 0.012% bromophenol blue). Between 2 μg and 30 μg of protein was loaded in wells of a 6% SDS-PAGE gel. Separated proteins were transferred to a polyvinylidene difluoride (PVDF) membrane for 1 h at 100 V at 4°C. Membranes were blocked in 5% milk in phosphate-buffered saline (PBS) containing 0.1% Tween 20 (PBS-T) at room temperature for 1 h and subsequently incubated in primary antibody as follows. All primary antibodies were left on the membrane for 1 h at room temperature. Membranes were washed with 1× PBS-T and probed for 1 h with secondary antibody dissolved in 1× PBS-T and 5% milk. Membranes were washed in PBS-T and signals detected using an ECL Western blotting kit as per the manufacturer's instructions (Pierce).
TAP-tagged Hsf1 was detected using a 1:5,000 dilution of anti-TAP tag rabbit polyclonal antibody (Thermo Scientific; CAB1001) in PBS-T plus 5% milk. To detect Act1, an anti-Act1 antibody was used (Santa Cruz Biotechnology; sc47778) at a 1:1,000 dilution in PBS-T plus 5% milk. To detect Hsp90, a 1:10,000 dilution of anti-Hsp90 antibody was used (courtesy of Bryan Larson) in PBS-T plus 5% milk.
RESULTSOLE1 is regulated by temperature.
Mechanisms of temperature sensing in fungal pathogens remain an enigma, but evidence obtained with the benign yeast S. cerevisiae implicates membrane fluidity as a primary sensor of temperature (17). To establish whether temperature sensing and membrane fluidity are linked in the fungal pathogen C. albicans, which is obligately associated with warm-blooded mammals, we aimed to determine if OLE1 expression is affected by temperature. As oleic acid is an unsaturated fatty acid, we hypothesized that during high-temperature growth, OLE1 would be downregulated, promoting a less fluid membrane. Wild-type cells (CaLC2993) were grown at 30°C, 37°C, or 42°C for 4 h or subjected to a 30°C to 42°C heat shock before being harvested and snap-frozen. RNA was extracted, and expression of OLE1 was determined relative to that of the housekeeping gene ACT1 (Fig. 1A). As expected, levels of OLE1 were significantly depleted when cells were heat shocked or grown at 42°C versus 30°C and 37°C (Fig. 1A). These data reinforce the notion that temperature affects fluidity of the membrane.
Membrane fluidity is tightly regulated in response to temperature and fatty acid ratios. (A) OLE1 transcript levels decrease upon exposure to high temperatures. C. albicans wild-type (WT) cells were grown to exponential phase at 30°C, 37°C, or 42°C or subjected to a 15-min 30°C to 42°C heat shock (HS), and OLE1 transcript levels were measured and normalized to the ACT1 loading control. *, P < 0.05 compared to 30°C growth (Student t test). (B) OLE1 transcript levels decrease following depletion in tetO-OLE1/ole1Δ cells. WT (CaLC2993 [Table 1]) and tetO-OLE1/ole1Δ cells (CaLC3032) were treated or not with 1 μg/ml of doxycycline (Dox) for 4 h, and OLE1 transcript levels were measured by qRT-PCR and normalized to the ACT1 loading control. *, P < 0.05 compared to the value for the wild type (Student t test). (C) Depleting Ole1 leads to an increase in FAS2 transcript levels. WT and tetO-OLE1/ole1Δ cells were treated or not with 1 μg/ml of doxycycline for 4 h, and FAS2 transcript levels were measured and normalized to the ACT1 loading control. *, P < 0.05 compared to the value for the wild type (Student t test).
To elucidate the mechanisms of temperature sensing through membrane fluidity further, we created a conditional mutant with doxycycline-repressible expression of the fatty acid desaturase gene OLE1, which is known for its essentiality (30). One allele of OLE1 was deleted, and the other was placed under the control of a tetracycline-repressible promoter. The addition of 1 μg/ml of the tetracycline analog doxycycline halted growth of the tetO-OLE1/ole1Δ strain after 6 h but had no effect on wild-type cells (data not shown). Expression of OLE1 was measured by quantitative reverse transcription-PCR (qRT-PCR) after 4 h of growth in the absence or presence of 1 μg/ml of doxycycline, when cells were in mid-log phase (OD600 = 0.6) (Fig. 1B). The addition of doxycycline had no effect on OLE1 expression in wild-type cells but caused a significant decrease in OLE1 expression in the tetO-OLE1/ole1Δ strain (Fig. 1B).
Ole1 is responsible for introducing the double bond into saturated fatty acetyl coenzyme A (acetyl-CoA) substrates to produce monounsaturated fatty acids. After the initial reaction of fatty acid synthesis, which involves the incorporation of acetyl-CoA with CO2 to generate malonyl-CoA (31), elongation of the carbon chain is catalyzed by the fatty acid synthases Fas1 and Fas2 to produce long-chain saturated fatty acids such as palmitic and stearic acids (14); these act as the precursors for the subsequent desaturation reactions to produce monounsaturated fatty acids. We postulated that depletion of Ole1 would act as a signal for the production of saturated fatty acid precursors, essentially leading to an increase in the levels of the fatty acid synthases. To test this, we depleted OLE1 as previously described and determined expression of the fatty acid synthase gene FAS2 by qRT-PCR (Fig. 1C). As expected, loss of OLE1 triggered a 5-fold increase in FAS2 levels. Therefore, the cell responds to temperature by intricately regulating the levels of saturated and unsaturated fatty acids.
Ole1 regulates components of the heat shock response.
To determine whether a link exists between membrane fluidity and temperature sensing in C. albicans, we looked to components of the heat shock response. The heat shock transcription factor Hsf1 is rapidly activated by phosphorylation following a heat shock, leading to the upregulation of heat shock proteins (22). We hypothesized that if the membrane does indeed sense temperature, depletion of oleic acid would misregulate Hsf1 activation. Wild-type and tetO-OLE1/ole1Δ cells were grown in the absence or presence of 1 μg/ml of doxycycline for 4 h to deplete OLE1. Cells were either left untreated (30°C) or treated with a 15-min 30°C to 42°C heat shock. Protein was extracted, and Hsf1-TAP was visualized using an anti-TAP antibody (Fig. 2A). Wild-type cells in the absence and presence of doxycycline respond similarly, activating Hsf1 by phosphorylation, as seen by an upshift and broadening of the band corresponding to Hsf1 upon the thermal shock. The tetO-OLE1/ole1Δ strain also activates Hsf1 similarly to wild-type cells in the absence of doxycycline. However, in the presence of doxycycline, we observed a dramatic decrease not only in the phosphorylation status of Hsf1 but also in its protein levels. Based on this, we hypothesized that levels of the heat shock protein Hsp90, which is regulated by Hsf1, would be reduced. However, when we probed the blot with an anti-Hsp90 antibody, Hsp90 levels remained constant. This is most likely due to the stability of Hsp90 (M. D. Leach, unpublished data).
Depletion of Ole1 inhibits full activation of Hsf1, leading to a loss in HSP expression during high-temperature growth and heat shock. (A) Phosphorylation of C. albicans Hsf1 during a 30°C to 42°C heat shock, revealed by Western analysis. Exponentially growing WT (wild type, CaLC2993 [Table 1]) and tetO-OLE1/ole1Δ cells (CaLC3032) were treated or not with1 μg/ml of doxycycline for 4 h and subjected to a 15-min 30°C to 42°C heat shock. Protein was extracted immediately, and Hsf1 phosphorylation was monitored using an anti-TAP antibody recognizing HSF1-TAP. Membranes were stripped and reprobed for Hsp90, and actin served as an internal loading control. (B to D) Expression of HSF1 (B), HSP21 (C), and HSP104 (D) during growth at 30°C, 37°C, or 42°C or after a 15-min 30°C to 42°C heat shock (HS), as determined by qRT-PCR of the corresponding transcripts relative to the internal ACT1 mRNA control. *, P < 0.05 compared to the value for the wild type at the corresponding temperature (Student t test).
Next, we aimed to determine if the impact of OLE1 depletion on Hsf1 levels manifests exclusively at the translational level, or if it is also evident at the transcriptional level. Expression of HSF1 was measured in wild-type cells and the tetO-OLE1/ole1Δ cells in the presence of 1 μg/ml of doxycycline after growth at 30°C, 37°C, or 42°C or upon a 15-min 30°C to 42°C heat shock (Fig. 2B). In wild-type cells, HSF1 levels increased up to 8-fold with increasing temperature and rapidly increased upon heat shock, equivalent to levels of cells grown at 42°C. However, loss of OLE1 hampers HSF1 expression during high-temperature growth, with levels increasing only 3-fold at 42°C. Additionally, when cells are exposed to a heat shock, HSF1 expression is only half that of wild-type cells. Doxycycline has no effect on HSF1 expression (see Fig. S1 in the supplemental material). These data provided the first molecular link between membrane fluidity and temperature sensing in fungi.
Our next goal was to ascertain if misregulation of Hsf1 upon OLE1 depletion affects expression of key heat shock proteins involved in the heat shock response. Under the same conditions as used to determine HSF1 levels, we monitored expression of HSP104 and HSP21, which encode two heat shock proteins essential for thermal adaptation in C. albicans (22, 32). We made three observations. First, there was an increase in both HSP104 and HSP21 under basal conditions upon OLE1 depletion (Fig. 2C and D). This increase remained at 37°C, with both genes presenting significantly higher expression in the tetO-OLE1/ole1Δ strain than in its wild-type counterpart. Therefore, depletion of OLE1 promotes the induction of HSP expression in the absence of any heat stress. Second, depletion of OLE1 had little impact on expression of the HSPs at 42°C. This could be due to a balance between the initial upregulation of these HSP genes in response to exposure to 42°C, before full depletion of Ole1, and stability of the HSP genes during high-temperature growth. Third, upon treatment with a 30°C to 42°C heat shock, HSP expression increased 100-fold for HSP21 (Fig. 2C) and 35-fold for HSP104 (Fig. 2D) in wild-type cells but reached only half of wild-type levels when OLE1 was depleted. Again, addition of doxycycline to wild-type cells had no effect on HSP expression (Fig. S1). In summary, depletion of OLE1 reduces Hsf1 protein levels and Hsf1 activation upon heat shock. In addition, HSF1 and HSP levels are misregulated at the transcriptional level.
Rsp5 regulates expression of OLE1 and activation of Hsf1.
The promoter of S. cerevisiae OLE1 contains multiple transcriptional regulatory elements, allowing for the elaborate control of the levels of unsaturated fatty acids. Transcription can be repressed through the addition of unsaturated fatty acids and activated in response to low temperature and hypoxia via the ER-bound transcription factors ScSpt23p and ScMga2p (33). Both factors are, in turn, activated by ubiquitin/proteasome-dependent ER-associated degradation, through the E3 ubiqutin ligase ScRsp5 (34). C. albicans retains only one homolog of the S. cerevisiae functionally redundant gene, CaSPT23. Repression of SPT23 in C. albicans blocks expression of OLE1 (35), suggesting that Rsp5 could be the upstream element that transduces signals pertaining to temperature changes sensed at the membrane.
We hypothesized that Rsp5 regulates OLE1 expression and hence Hsf1 activation in C. albicans. To test this, we constructed a tetracycline-repressible RSP5 conditional expression strain, in which RSP5 expression is repressed upon the addition of doxycycline. Addition of 20 μg/ml of doxycycline for prolonged periods did not affect the viability of the tetO-RSP5/rsp5Δ strain, suggesting that RSP5 is not an essential gene (data not shown). To ensure full depletion of RSP5, tetO-RSP5/rsp5Δ cells were subcultured into YPD in the presence or absence of 20 μg/ml of doxycycline for 18 h and then subcultured again the following morning under the same conditions and grown to mid-log phase (approximately 5 h). RNA was extracted for qRT-PCR analysis of RSP5 transcript levels (Fig. 3A). As expected, RSP5 was significantly decreased upon addition of 20 μg/ml of doxycycline. We also noted that in the absence of doxycycline, RSP5 levels in the tetO-RSP5/rsp5Δ strain were significantly higher than in the wild-type counterpart, suggesting that the tetO promoter is stronger than the native promoter. To determine if Rsp5 regulates OLE1 expression, we depleted RSP5 at 30°C overnight and until mid-log phase and examined OLE1 expression by qRT-PCR. Notably, OLE1 levels were 4-fold lower upon RSP5 depletion than in wild-type cells (Fig. 3B), indicating that Rsp5 regulates OLE1, likely through the transcription factor Spt23 (35). Even though RSP5 expression was significantly increased in the absence of doxycycline (Fig. 3A), OLE1 expression remained the same as in wild-type cells (see Fig. S2A in the supplemental material).
Rsp5 regulates Hsf1 in part through OLE1. (A) RSP5 transcript levels are upregulated in tetO-RSP5/rsp5Δ strain in the absence of doxycycline and depleted in the presence of doxycycline. WT (wild type, CaLC2993 [Table 1]) and tetO-RSP5/rsp5Δ (CaLC3366) cells were treated or not with 20 μg/ml of doxycycline overnight for 18 h and then subcultured the following morning and grown for a further 5 h under the same conditions. RSP5 transcript levels were normalized to the ACT1 loading control. (B) Depleting Rsp5 causes a decrease in OLE1 transcript levels. WT and tetO-RSP5/rsp5Δ cells were treated or not with 20 μg/ml of doxycycline as stated above and OLE1 transcript levels were measured and normalized to the ACT1 loading control. (C) Rsp5 is required for Hsf1 expression and full activation. WT and tetO-RSP5/rsp5Δ cells were treated or not with 20 μg/ml of doxycycline as stated above, exposed to a 15-min 30°C to 42°C heat shock, and subjected to Western analysis. Decreased Hsf1-TAP levels were observed in the tetO-RSP5/rsp5Δ strain. Membranes were stripped and reprobed for actin, which served as the internal loading control. (D) Expression of HSP21 and HSP104 during growth at 30°C or after a 15-min 30°C to 42°C heat shock (HS), as determined by qRT-PCR relative to the internal ACT1 mRNA control. *, P < 0.05 in tetO-RSP5/rsp5Δ cells compared to wild-type cells (Student t test).
Given that we can manipulate RSP5 expression and thereby reduce levels of OLE1, we wanted to examine the effect of RSP5 depletion on Hsf1 activity in response to heat shock. Wild-type cells and tetO-RSP5/rsp5Δ cells were grown in the absence or presence of 20 μg/ml of doxycycline as previously described and either kept at 30°C or exposed to a 15-min 30°C to 42°C heat shock. Proteins were extracted and subjected to Western analysis to visualize Hsf1 activation (Fig. 3C). As predicted, doxycycline-mediated depletion of RSP5 led to reduced levels of Hsf1 activation, as indicated by a decrease in phosphorylation with a narrower band and loss of a band shift. We also noted that in the absence of doxycycline, although a visible shift is seen in the presence of a heat shock in the tetO-RSP5/rsp5Δ strain, the levels of Hsf1 are significantly lower than in the wild-type counterpart, suggesting that the misregulation of RSP5 affects Hsf1 protein levels.
In order to confirm that Hsf1 activation upon depletion of RSP5 is not equivalent to that seen in wild-type cells upon a heat shock, we monitored expression of HSP104 and HSP21 under the same conditions (Fig. 3D). As expected, expression of both HSPs increased significantly upon heat shock; however, this increase was reduced when RSP5 was depleted. Expression of HSP genes was not affected by doxycycline in the wild-type strain (see Fig. S2B and C in the supplemental material). Although expression of HSP104 and HSP21 was significantly lower upon depletion of RSP5, it was not reduced to the same extent as that observed upon depletion of OLE1 (Fig. 3C and D). In part, this may be because depletion of RSP5 does not fully deplete OLE1 (Fig. 3A), even in the presence of heat shock (Fig. S2D). Therefore, depletion of RSP5 phenocopies loss of OLE1.
DISCUSSION
In this paper, we have addressed one of the most fundamental questions in biology: how do cells sense temperature? The ability to sense the surrounding temperature is key for virulence. Indeed, many fungal pathogens inhabit a remarkable diversity of environments; for example, Cryptococcus neoformans, Histoplasma capsulatum, and Aspergillus fumigatus are found in diverse environments such as pigeon excreta and soil, but each retains the ability to grow at 37°C. The loss of genes necessary for high-temperature growth in these pathogens results in attenuated virulence and at times even death (36–38), suggesting the importance of temperature sensing for pathogenesis. Besides governing virulence, temperature regulates numerous cellular processes that often contribute to pathogenesis. For example, morphological transitions in C. albicans are controlled by temperature. At ambient temperature, this organism favors the yeast form, while elevated temperatures induce filamentous growth, enabling penetration of the epithelium (39). C. albicans can also switch from a white to opaque cellular state in host niches with lower temperatures, such as skin, facilitating mating (40).
By manipulating C. albicans membrane composition, we have begun to understand the ways in which fungal pathogens sense temperature. Using a tetracycline-repressible promoter, we were able to deplete levels of the fatty acid desaturase gene OLE1, which regulates synthesis of the monounsaturated fatty acid oleic acid. In doing so, we triggered a 5-fold increase in the fatty acid synthase gene FAS2 (Fig. 1C). These data suggest that the cell compensates for the loss of unsaturated fatty acids by increasing the levels of FAS2 in an attempt to increase levels of saturated fatty acids, which are the precursors for unsaturated fatty acids. Furthermore, heat shock or high-temperature growth caused a significant decrease in OLE1 expression (Fig. 1A). This is consistent with previous studies with S. cerevisiae illustrating a 4- to 6-fold increase in OLE1 mRNA when cells were rapidly shifted from 30°C to 10°C (41). These data reinforce the notions that temperature affects fluidity of the membrane and that the cell continually monitors the ratio of saturated to unsaturated fatty acids.
Previous studies with the fungal pathogen H. capsulatum and the model yeast S. cerevisiae have suggested a role for the membrane in sensing the surrounding temperature. Indeed, Carratù and colleagues observed a strong increase in HSP expression in H. capsulatum when cells were heat shocked at 37°C upon the addition of external saturated fatty acids into the media. Addition of unsaturated fatty acids, however, reduced HSP expression during the same heat shock (17). Furthermore, Chatterjee and colleagues demonstrated a 9°C increase in the optimal activation of the heat shock response upon supplementation of S. cerevisiae with unsaturated fatty acids (42). Therefore, it would appear that the ratio of saturated to unsaturated fatty acids of the membrane is a key determinant in the perception of rapid temperature changes. Consistent with these data, we discovered that loss of OLE1 caused upregulated expression of key components of the heat shock response during growth at 30°C and 37°C, but this upregulation was lost upon high-temperature growth or exposure to a short heat shock (Fig. 2B to D). Indeed, based on our findings and those of others, it would appear that upon loss of OLE1, levels of unsaturated fatty acids are severely depleted within the membrane, prompting the cell to assume that the temperature has increased. In turn, HSP expression is increased at 30°C and 37°C. However, if the ratio of fatty acids is acting as the primary sensor of temperature, upon subjecting the cell to a drastic heat shock, the temperature change in a membrane that is depleted of unsaturated fatty acids is not properly recognized, thereby preventing full activation of Hsf1 and reducing upregulation of HSP genes (Fig. 2).
One of the conundrums surrounding temperature sensing is the remarkable speed at which cells are able to sense and respond to the surrounding environment. C. albicans is able to activate Hsf1 within 1 min of exposure to a heat shock (11), suggesting that the sensor itself may reside in the plasma membrane, becoming activated upon slight changes in membrane fluidity. This would then be transduced into a signal that could further regulate fatty acid expression. Based on this notion, we examined regulators of fatty acid synthesis. In S. cerevisiae, OLE1 expression is regulated via the ER-bound transcription factors Spt23 and Mga2 (33), which are activated by the E3 ubiquitin ligase Rsp5 (34). C. albicans has retained only one homolog of the S. cerevisiae functionally redundant genes, CaSPT23. Oh and colleagues found that upon repression of SPT23 in C. albicans, expression of OLE1 was blocked (35). Therefore, we postulated that Rsp5 regulates expression of OLE1, through Spt23, thereby acting as a downstream element that is activated upon changes that occur in the plasma membrane and subsequently regulating activation of the heat shock response. By regulating the levels of Rsp5 using the tetracycline-repressible promoter (Fig. 3A), we noted a significant decrease in OLE1 expression upon loss of RSP5 (Fig. 3B). This translated to a reduction in Hsf1 activation during a short heat shock (Fig. 3C), similar to that seen upon loss of OLE1 (Fig. 2A). Furthermore, this decrease in Hsf1 activation contributed to a decrease in HSP21 and HSP104 expression upon a heat shock (Fig. 3D).
We also observed that in the absence of doxycycline in the tetO-RSP5/rsp5Δ strain, RSP5 levels are significantly higher than in wild-type cells (Fig. 3A; see also Fig. S2E in the supplemental material). When we looked at Hsf1 activation under these conditions, although Hsf1 is activated and HSP expression remains similar to that observed in wild-type cells (see Fig. S2B and C), Hsf1 levels are much lower than in the wild-type counterpart and similar to those observed upon Rsp5 depletion. For S. cerevisiae, two recent studies by Haitani and colleagues show that an Rsp5 mutant, rsp5A410E, exhibits a decrease in Hsf1 levels compared to that for wild-type cells but relatively similar HSF1 mRNA levels. Further investigation showed accumulation of HSF1 mRNA in the nucleus, suggesting that Rsp5 is required for HSF1 nuclear export (43, 44). Our data suggest that a similar mechanism operates in C. albicans, whereby misregulation of Rsp5 prevents proper export of HSF1 mRNA. However, depletion of OLE1 does reduce HSF1 expression, suggesting more complex circuitry connecting Rsp5-Ole1-Hsf1.
These data provide evidence for the first inhibitor of Hsf1 activation in response to a heat shock. The loss of the unsaturated fatty acid desaturase gene OLE1 drastically increases the fatty acid synthase gene FAS2 (Fig. 1C), leads to an upregulation of HSP genes in the absence of heat shock but inhibits full HSP expression upon heat shock (Fig. 2C and D). This inhibition is likely due to a reduction in Hsf1 activation (Fig. 2A). This suggests that membrane fluidity is a key sensor of temperature; however, before changes can occur in levels of fatty acids, a signal must be transduced. OLE1 is regulated by the transcription factor Spt23, which is regulated by the E3 ubiquitin ligase Rsp5. We discovered that loss of RSP5 phenocopies loss of OLE1, suggesting that Rsp5 could be one of the early sensors of temperature, which coordinately regulates fatty acid synthesis and activation of the heat shock response.
Understanding circuitry governing temperature sensing offers therapeutic opportunities for crippling diverse microbial pathogens. In the context of Candida species, many are human commensals and able to disseminate into the bloodstream, causing systemic infections. Inhibiting fatty acid biosynthesis would be an ideal way to eradicate Candida species from the mycobiome prior to immunosuppressive treatments that render patients vulnerable to infection. Importantly, fungal Ole1 is comprised of two domains: the N and C domains, containing the catalytic motif for the desaturase activity and the cytochrome b5 activity, respectively. In contrast, humans have two separate enzymes to carry out these activities, making fungal Ole1 an ideal antifungal target (33). Furthermore, recent studies illustrated that components of fatty acid biosynthesis from different Candida species are essential for establishing and maintaining infection in a mouse model of candidiasis (45–47). Xu and colleagues demonstrated that C. albicans Ole1 is necessary for virulence, and Nguyen and colleagues obtained similar results with C. parapsilosis. This is consistent with our findings that loss of Ole1 leads to a significant decrease in Hsf1 activation (Fig. 2A) and a previous study showing that activation of Hsf1 is necessary for virulence in C. albicans (48). Finally, Krishnamurthy and colleagues discovered that depletion of OLE1 blocks one of the key virulence traits in C. albicans, hyphal development, when cells are grown under aerobic conditions (30). This study provides us with the first mechanistic link between fatty acid synthesis and the heat shock response in the fungal kingdom, highlighting Ole1 as an attractive antifungal. Beyond fungal pathogens, fatty acid biosynthesis is emerging as a powerful target for the development of antibacterial agents, with small-molecule inhibitors currently in clinical use to treat tuberculosis and as ubiquitous consumer antimicrobials (49).
Supplementary Material
Supplemental material
Published ahead of print 20 June 2014
Supplemental material for this article may be found at http://dx.doi.org/10.1128/EC.00138-14.
ACKNOWLEDGMENTS
We thank Grant Hatch and Fred Xu for help with data analysis. We also thank Joe Heitman and Al Brown for inspiring scientific discussions.
M.D.L. is supported by a Sir Henry Wellcome Postdoctoral Fellowship (Wellcome Trust grant 096072), and L.E.C. is supported by a Canada Research Chair in Microbial Genomics and Infectious Disease and by Canadian Institutes of Health Research GrantMOP-119520.
REFERENCESKussellELeiblerS2005Phenotypic diversity, population growth, and information in fluctuating environments.
309:2075–2078. 10.1126/science.111438316123265TiroshIWeinbergerACarmiMBarkaiN2006A genetic signature of interspecies variations in gene expression.
38:830–834. 10.1038/ng181916783381GaschAPSpellmanPTKaoCMCarmel-HarelOEisenMBStorzGBotsteinDBrownPO2000Genomic expression programs in the response of yeast cells to environmental changes.
11:4241–4257. 10.1091/mbc.11.12.424111102521EnjalbertBNantelAWhitewayM2003Stress-induced gene expression in Candida albicans: absence of a general stress response.
14:1460–1467. 10.1091/mbc.E02-08-054612686601CaustonHCRenBKohSSHarbisonCTKaninEJenningsEGLeeTITrueHLLanderESYoungRA2001Remodeling of yeast genome expression in response to environmental changes.
12:323–337. 10.1091/mbc.12.2.32311179418EstruchF2000Stress-controlled transcription factors, stress-induced genes and stress tolerance in budding yeast.
24:469–486. 10.1111/j.1574-6976.2000.tb00551.x10978547de DiosCHRomanEMongeRAPlaJ2010The role of MAPK signal transduction pathways in the response to oxidative stress in the fungal pathogen Candida albicans: implications in virulence.
11:693–703. 10.2174/13892031079455765521235505MagerWHDe KruijffAJ1995Stress-induced transcriptional activation.
59:506–5317565416ParsellDALindquistS1993The function of heat-shock proteins in stress tolerance: degradation and reactivation of damaged proteins.
27:437–496. 10.1146/annurev.ge.27.120193.0022538122909SorgerPK1991Heat shock factor and the heat shock response.
65:363–366. 10.1016/0092-8674(91)90452-52018972LeachMDTycKMBrownAJPKlippE2012Modelling the regulation of thermal adaptation in Candida albicans, a major fungal pathogen of humans.
7:e32467. 10.1371/journal.pone.003246722448221LeachMDCowenLE2013Surviving the heat of the moment: a fungal pathogens perspective.
9:e1003163. 10.1371/journal.ppat.100316323505364LeachMDCowenLE2014To sense or die: mechanisms of temperature sensing in fungal pathogens.
8:185–191. 10.1007/s12281-014-0182-1SchweizerEHofmannJ2004Microbial type I fatty acid synthases (FAS): major players in a network of cellular FAS systems.
68:501–517. 10.1128/MMBR.68.3.501-517.200415353567StukeyJEMcDonoughVMMartinCE1989Isolation and characterization of OLE1, a gene affecting fatty acid desaturation from Saccharomyces cerevisiae.
264:16537–165442674136KloseCSurmaMAGerlMJMeyenhoferFShevchenkoASimonsK2012Flexibility of a eukaryotic lipidome—insights from yeast lipidomics.
7:e35063. 10.1371/journal.pone.003506322529973CarratùLFranceschelliSPardiniCLKobayashiGSHorvathIVighLMarescaB1996Membrane lipid perturbation modifies the set point of the temperature of heat shock response in yeast.
93:3870–3875. 10.1073/pnas.93.9.38708632982GarganoSDi LalloGKobayashiGSMarescaB1995A temperature-sensitive strain of Histoplasma capsulatum has an altered Δ9-fatty acid desaturase gene.
30:899–906. 10.1007/BF025374808538376OddsFC1988
, 2nd ed.Bailliere Tindall, London, United KingdomPiispanenAEHoganDA2008PEPped up: induction of Candida albicans virulence by bacterial cell wall fragments.
4:1–2. 10.1016/j.chom.2008.06.00518621003CalderoneRA2002
. ASM Press, Washington, DCNichollsSLeachMDPriestCLBrownAJ2009Role of the heat shock transcription factor, Hsf1, in a major fungal pathogen that is obligately associated with warm-blooded animals.
74:844–861. 10.1111/j.1365-2958.2009.06883.x19818013LeachMDBudgeSWalkerLMunroCCowenLEBrownAJP2012Hsp90 orchestrates transcriptional regulation by Hsf1 and cell wall remodelling by MAPK signalling during thermal adaptation in a pathogenic yeast.
8:e1003069. 10.1371/journal.ppat.100306923300438LeachMDKlippECowenLEBrownAJ2012Fungal Hsp90: a biological transistor that tunes cellular outputs to thermal inputs.
10:693–704. 10.1038/nrmicro287522976491ShermanF1991Getting started with yeast.
194:3–21. 10.1016/0076-6879(91)94004-V2005794DiezmannSMichautMShapiroRSBaderGDCowenLE2012Mapping the Hsp90 genetic interaction network in Candida albicans reveals environmental contingency and rewired circuitry.
8:e1002562. 10.1371/journal.pgen.100256222438817LavoieHSellamAAskewCNantelAWhitewayM2008A toolbox for epitope-tagging and genome-wide location analysis in Candida albicans.
9:578. 10.1186/1471-2164-9-57819055720LeachMDSteadDAArgoEMacCallumDMBrownAJ2011Molecular and proteomic analyses highlight the importance of ubiquitination for the stress resistance, metabolic adaptation, morphogenetic regulation and virulence of Candida albicans.
79:1574–1593. 10.1111/j.1365-2958.2011.07542.x21269335LeachMDSteadDAArgoEBrownAJ2011Identification of sumoylation targets, combined with inactivation of SMT3, reveals the impact of sumoylation upon growth, morphology, and stress resistance in the pathogen Candida albicans.
22:687–702. 10.1091/mbc.E10-07-063221209325KrishnamurthySPlaineAAlbertJPrasadTPrasadRErnstJF2004Dosage-dependent functions of fatty acid desaturase Ole1p in growth and morphogenesis of Candida albicans.
150:1991–2003. 10.1099/mic.0.27029-015184585SchneiterRGuerraCELamplMGoggGKohlweinSDKleinHL1999The Saccharomyces cerevisiae hyperrecombination mutant hpr1delta is synthetically lethal with two conditional alleles of the acetyl coenzyme A carboxylase gene and causes a defect in nuclear export of polyadenylated RNA.
19:3415–342210207065MayerFLWilsonDJacobsenIDMiramonPSlesionaSBohovychIMBrownAJHubeB2012Small but crucial: the novel small heat shock protein Hsp21 mediates stress adaptation and virulence in Candida albicans.
7:e38584. 10.1371/journal.pone.003858422685587MartinCEOhCSJiangY2007Regulation of long chain unsaturated fatty acid synthesis in yeast.
1771:271–285. 10.1016/j.bbalip.2006.06.01016920014HoppeTMatuschewskiKRapeMSchlenkerSUlrichHDJentschS2000Activation of a membrane-bound transcription factor by regulated ubiquitin/proteasome-dependent processing.
102:577–586. 10.1016/S0092-8674(00)00080-511007476OhCSMartinCE2006Candida albicans Spt23p controls the expression of the Ole1p Δ9 fatty acid desaturase and regulates unsaturated fatty acid biosynthesis.
281:7030–7039. 10.1074/jbc.M51074620016415349BhabhraRMileyMDMylonakisEBoettnerDFortwendelJPanepintoJCPostowMRhodesJCAskewDS2004Disruption of the Aspergillus fumigatus gene encoding nucleolar protein CgrA impairs thermotolerant growth and reduces virulence.
72:4731–4740. 10.1128/IAI.72.8.4731-4740.200415271935LamothFJuvvadiPRFortwendelJRSteinbachWJ2012Heat shock protein 90 (Hsp90) is required for conidiation and cell wall integrity in Aspergillus fumigatus.
11:1324–1332. 10.1128/EC.00032-1222822234OdomAMuirSLimEToffalettiDLPerfectJHeitmanJ1997Calcineurin is required for virulence of Cryptococcus neoformans.
16:2576–2589. 10.1093/emboj/16.10.25769184205GowNABrownAJOddsFC2002Fungal morphogenesis and host invasion.
5:366–371. 10.1016/S1369-5274(02)00338-712160854LachkeSALockhartSRDanielsKJSollDR2003Skin facilitates Candida albicans mating.
71:4970–4976. 10.1128/IAI.71.9.4970-4976.200312933839NakagawaYSakumotoNKanekoYHarashimaS2002Mga2p is a putative sensor for low temperature and oxygen to induce OLE1 transcription in Saccharomyces cerevisiae.
291:707–713. 10.1006/bbrc.2002.650711855848ChatterjeeMTKhalawanSACurranBPG1997Alterations in cellular lipids may be responsible for the transient nature of the yeast heat shock response.
143:3063–3068. 10.1099/00221287-143-9-30639308188HaitaniYShimoiHTakagiH2006Rsp5 regulates expression of stress proteins via post-translational modification of Hsf1 and Msn4 in Saccharomyces cerevisiae.
580:3433–3438. 10.1016/j.febslet.2006.05.01616713599HaitaniYTakagiH2008Rsp5 is required for the nuclear export of mRNA of HSF1 and MSN2/4 under stress conditions in Saccharomyces cerevisiae.
13:105–116. 10.1111/j.1365-2443.2007.01154.x18233954XuDSillaotsSDavisonJHuWJiangBKauffmanSMartelNOcampoPOhCTrosokSVeilletteKWangHYangMZhangLBeckerJMartinCERoemerT2009Chemical genetic profiling and characterization of small-molecule compounds that affect the biosynthesis of unsaturated fatty acids in Candida albicans.
284:19754–19764. 10.1074/jbc.M109.01987719487691NguyenLNTrofaDNosanchukJD2009Fatty acid synthase impacts the pathobiology of Candida parapsilosis in vitro and during mammalian infection.
4:e8421. 10.1371/journal.pone.000842120027295NguyenLNGacserANosanchukJD2011The stearoyl-coenzyme A desaturase 1 is essential for virulence and membrane stress in Candida parapsilosis through unsaturated fatty acid production.
79:136–145. 10.1128/IAI.00753-1020974817NichollsSMacCallumDMKaffarnikFASelwayLPeckSCBrownAJ2011Activation of the heat shock transcription factor Hsf1 is essential for the full virulence of the fungal pathogen Candida albicans.
48:297–305. 10.1016/j.fgb.2010.08.01020817114HeathRJWhiteSWRockCO2002Inhibitors of fatty acid synthesis as antimicrobial chemotherapeutics.
58:695–703. 10.1007/s00253-001-0918-z12021787NobleSMJohnsonAD2005Strains and strategies for large-scale gene deletion studies of the diploid human fungal pathogen Candida albicans.
4:298–309. 10.1128/EC.4.2.298-309.200515701792oai:pubmedcentral.nih.gov:44866692015-07-29eukcellpmc-openEukaryot CellEukaryotic CelleukcelleukcellEUKCELLEukaryotic Cell1535-97781535-9786American Society for Microbiology1752 N St., N.W., Washington, DCPMC4486669PMC44866694486669260027182600271800078-1510.1128/EC.00078-15ArticlesCsr1/Zap1 Maintains Zinc Homeostasis and Influences Virulence in Candida dubliniensis but Is Not Coupled to MorphogenesisZinc Homeostasis in C. dubliniensisBöttcher et al.BöttcherBettinaaPaligeKatjabJacobsenIlse D.cdHubeBernhardacBrunkeSaschaaeDepartment of Microbial Pathogenicity Mechanisms, Leibniz Institute for Natural Product Research and Infection Biology—Hans Knoell Institute Jena, Jena, GermanyMicrofluidic ChipShop GmbH, Jena, GermanyFriedrich Schiller University, Jena, GermanyResearch Group Microbial Immunology, Leibniz Institute for Natural Product Research and Infection Biology—Hans Knoell Institute Jena, Jena, GermanyCenter for Sepsis Control and Care, Jena, GermanyAddress correspondence to Sascha Brunke, sascha.brunke@leibniz-hki.de.
Citation Böttcher B, Palige K, Jacobsen ID, Hube B, Brunke S. 2015. Csr1/Zap1 maintains zinc homeostasis and influences virulence in Candida dubliniensis but is not coupled to morphogenesis. Eukaryot Cell 14:661–670. doi:10.1128/EC.00078-15.
The supply and intracellular homeostasis of trace metals are essential for every living organism. Therefore, the struggle for micronutrients between a pathogen and its host is an important determinant in the infection process. In this work, we focus on the acquisition of zinc by Candida dubliniensis, an emerging pathogen closely related to Candida albicans. We show that the transcription factor Csr1 is essential for C. dubliniensis to regulate zinc uptake mechanisms under zinc limitation: it governs the expression of the zinc transporter genes ZRT1, ZRT2, and ZRT3 and of the zincophore gene PRA1. Exclusively, artificial overexpression of ZRT2 partially rescued the growth defect of a csr1Δ/Δ mutant in a zinc-restricted environment. Importantly, we found that, in contrast to what is seen in C. albicans, Csr1 (also called Zap1) is not a major regulator of dimorphism in C. dubliniensis. However, although a csr1Δ/Δ strain showed normal germ tube formation, we detected a clear attenuation in virulence using an embryonated chicken egg infection model. We conclude that, unlike in C. albicans, Csr1 seems to be a virulence factor of C. dubliniensis that is not coupled to filamentation but is strongly linked to zinc acquisition during pathogenesis.
INTRODUCTION
Access to zinc is essential for organisms throughout the three domains of life. It is the only metal that occurs as a cofactor in all six classes of enzymes, from oxido-reductases to lyases (1), and the average proportion of enzymes containing zinc is 8.8% in eukaryotic proteomes (2). In pathogens, virulence-associated proteins frequently bind zinc for structural stability or catalytic activity; e.g., the Ser/Thr-protein kinase PrkC of Bacillus anthracis, which is essential for its pathogenicity, is regulated by zinc (3). In the pathogenic yeast Candida albicans, three out of six known superoxide dismutases (CaSod1, CaSod4, and CaSod6) are copper-zinc dependent. Enzymes of this class detoxify reactive oxygen species and thus contribute to virulence (4–6). Therefore, it is of particular importance for both benign and pathogenic microbes to ensure a sufficient zinc supply, especially when faced with a micronutrient-poor environment.
Exploiting this dependency, mammalian hosts manipulate levels of accessible zinc and other metals to inhibit pathogen growth and dissemination. This targeted limitation of micronutrients is known as nutritional immunity and is one of the main strategies used to defend against pathogenic microorganisms (7). To oppose zinc deprivation, pathogenic bacteria and fungi evolved specialized uptake mechanisms to obtain zinc (8, 9). For example, a high-affinity zinc transporter system is required for virulence of Salmonella enterica in mice (10). The intracellular zinc homeostasis is generally strictly controlled, and in C. albicans, the response to zinc deficiency is mediated by the transcription factor Csr1 (Candida suppressor of ROK1) (11), the ortholog of Saccharomyces cerevisiae Zap1 (zinc-responsive activator protein). Within the Candida clade, CSR1 orthologs have been found in all sequenced species. However, to date, this transcriptional factor has been investigated only in C. albicans in more detail, while the function of Csr1 in other pathogenic yeasts like Candida glabrata or even in the closest relative of C. albicans, Candida dubliniensis, is unknown.
Both C. dubliniensis and C. albicans are harmless gastrointestinal colonizers, but they can cause diseases ranging from superficial mucosal infections to life-threatening candidemia, especially in immunocompromised individuals. Interestingly, C. dubliniensis is less frequently isolated from patients with nosocomial bloodstream infections than C. albicans (2 to 3% versus 10%, respectively) (12–14). The overall lower virulence of C. dubliniensis has also been confirmed in mice infection models (15) and was found to be associated with differences in species-specific pathogenicity properties, such as the ability to adhere and to form true hyphae, which allow tissue invasion (16, 17). Finding differences in the genetic setup and infection-relevant phenotypes of these two fungi is, therefore, a promising avenue to dissect virulence in pathogenic yeasts and may provide insights into the mechanisms of evolutionary rewiring of regulatory factors among related microbes.
In C. albicans, Csr1 is known to have dual functions: it plays the key role both in transcriptional regulation of zinc homeostasis and in biofilm formation. C. albicans mutants lacking CSR1 hence cannot proliferate under low-zinc conditions and show reduced filamentation in the presence of serum (11) accompanied by altered biofilm formation (18). Further analysis of genes regulated by Csr1 of C. albicans (CaCsr1) under biofilm-inducing conditions revealed 60 targets, including CaZRT1-3, CaPRA1, and CaCSR1 itself (18). It is noteworthy that in biofilm-producing communities, a C. albicanscsr1Δ/Δ mutant strain secretes smaller amounts of the quorum-sensing molecule farnesol, which contributes to an altered morphology (19).
The Zrt proteins belong to the ZIP (Zrt/Irt-like proteins) transporter family and facilitate zinc ion transfer across membranes into the cytosol or cellular organelles. Zrt1 of S. cerevisiae (ScZrt1) is a high-affinity transporter in S. cerevisiae that mediates zinc uptake under strong zinc depletion, but it is downregulated under low-zinc conditions. There, the low-affinity zinc transporter ScZrt2 ensures import of zinc (20, 21). These complementary uptake systems are under the control of ScZap1 (22). Tightly controlled zinc uptake mechanisms in response to extracellular zinc levels have been observed not only in S. cerevisiae but also in Schizosaccharomyces pombe and Aspergillus fumigatus (22–24). Finally, Pra1 is a zinc-binding protein which is part of a novel zinc uptake mechanism of C. albicans recently discovered by Citiulo et al. (25).
A C. albicanscsr1Δ/Δ mutant is known to be proliferation defective during murine infections (26) and to elicit a decreased immune response in mice (27). In addition to this observation, expression of CSR1 and some of its target genes was increased up to 10-fold during the early stage of infection with the corresponding C. albicans wild-type strain (27).
In the present work, we analyzed the role of the C. dubliniensis transcription factor Csr1 (CD36_44490)—a homolog of C. albicans Csr1—in zinc homeostasis, germ tube formation, and virulence traits.
MATERIALS AND METHODSStrains and culture conditions.
Candida strains were routinely propagated on YPD agar (20 g peptone, 10 g yeast extract, 20 g glucose, 15 g agar per liter) at 30°C and stored as frozen stocks in YPD medium with 15% (vol/vol) glycerol at −80°C. For zinc starvation experiments, low-zinc medium (LZM) was prepared as described previously (22). The medium was supplemented with ZnSO4 as indicated (LZM0 contains no zinc; LZM25 and LZM2000 contain 25 µM and 2,000 µM ZnSO4, respectively), and 25 μM FeSO4 was used as a source of iron. Candida strains used in this work are listed in Table 1.
Germ tube assays.
Strains were grown in YPD overnight (30°C and 180 rpm), washed with double-distilled water (ddH2O), and transferred into filament-inducing medium at an optical density at 600 nm (OD600) of 0.2. To stimulate filamentation, we used spider medium (1% mannitol, 1% nutrient broth, 0.2% K2HPO4 [pH 7.2]), liquid YPD, or H2O plus 10% (vol/vol) fetal calf serum. Cultures were shaken (180 rpm) for 4 h at 37°C, and morphology was microscopically analyzed (Axiovert, Zeiss, Germany).
Chlamydospore formation.
Chlamydospore production was induced on rice extract-Tween 80 agar (BD, Heidelberg, Germany) or Staib agar (28), both prepared as described before. The plates were incubated at 28°C for 2 to 4 days in the dark, and chlamydospore formation was monitored microscopically.
The deletion cassette for CSR1 was constructed as follows. An ApaI-XhoI fragment with CSR1 upstream sequences was cloned after amplification by PCR with the primers CSR1-1 and CSR1-2 (see Table S1 in the supplemental material) using genomic DNA from C. dubliniensis Wü284 as the template. A SacII-SacI fragment containing CSR1 downstream sequences was obtained with the primers CSR1-3 and CSR1-4. The CSR1 upstream and downstream fragments replaced SSU2 upstream and downstream fragments in plasmid pSSU2M2 (29) via the introduced restriction sites, to result in pCSR1M2, in which the SAT1 flipper is flanked by CSR1 sequences.
The whole CSR1 gene for the gene reconstitution was amplified using the primers CdCSR1-1 and CdCSR1-5, the ApaI/BglII-cut DNA fragment was integrated into pSAP2KS1 (30), and the CSR1 downstream DNA element was inserted as described above.
For the generation of the PRA1 overexpression cassette, a XhoI-BglII fragment was amplified via PCR with the primers PRA1-1 and PRA1-2. Genomic DNA from C. dubliniensis Wü284 was used as the template. This DNA fragment was introduced behind the cdADH1 promoter into the pcdADH1E2 vector (31), and the plasmid was named pcdPRA1E1. The plasmids pcdZRT1E1 and pcdZRT2E1were constructed in a similar way by amplifying a XhoI-BglII fragment with the primer pair ZRT1-1 and ZRT1-2 or ZRT2-1 and ZRT2-2. The primer ZRT2-1 carried a SalI restriction site that is compatible with the XhoI overhang of the parental plasmid pcdADH1E2.
C. dubliniensis transformant construction.
Linear DNA fragments were transformed by electroporation into chemically competent C. dubliniensis cells (32), and clones were selected on YPD plates containing nourseothricin (Werner Bioagents, Jena, Germany). The usage of the SAT1 flipper strategy allowed the recycling of the selection marker, as described here (33). The insertion locus of the DNA fragment was confirmed by Southern blot analyses.
Southern blotting.
A 10-μg portion of isolated genomic DNA was digested with an appropriate restriction enzyme. After DNA separation on an agarose gel (1%), DNA was stained with ethidium bromide and transferred onto a nylon membrane using a vacuum blot system. UV-linked DNA was hybridized with chemiluminescence-labeled probes and detected via the Amersham ECL direct nucleic acid labeling and detection kit (GE Healthcare, Braunschweig, Germany) according to the manufacturer's instructions (see Fig. S1 in the supplemental material).
Growth curve analyses.
Proliferation under zinc depletion was evaluated via growth curve assays. Strains were pregrown overnight in YPD at 30°C and after repeated washing, cells with an OD600 of 0.4 were inoculated in LZM0 without additional zinc. After starvation in LZM0 for 24 h at 30°C, cells were diluted to an OD600 of 0.01 in LZM supplemented with various concentrations of ZnSO4. Cultures were incubated at 30°C in a Magellan TECAN plate reader with shaking for 30 s, and the OD600 was determined every 15 min over 48 h. Changes of the OD600 were plotted against the incubation time.
To determine gene expression rates, cells were precultured in YPD overnight (30°C and 180 rpm) and washed with phosphate-buffered saline (PBS). A total of 5 × 106 cells/ml were inoculated into 200 ml LZM plus 2,000 μM ZnSO4, and the cells were grown for an additional 24 h (30°C and 180 rpm). To remove residual zinc, cultures were washed four times with ultrapure water, and all yeast cells were transferred into 200 ml LZM0 without zinc.
Cells from 20 ml of liquid culture were sampled and frozen in liquid nitrogen at 0 h, 0.5 h, 4 h, and 24 h. RNA was isolated using an RNeasy kit (Qiagen, Hilden, Germany) following the manufacturer's instructions. A Bioanalyzer instrument (Agilent, Santa Clara, CA) was used to measure RNA quality, and RNA concentration was determined via NanoDrop (Thermo Fisher Scientific, Waltham, MA). A 700-ng portion of RNA was treated with DNase and transcribed into cDNA (enzymes by Promega [Fitchburg, WI]). Finally, a total amount of 13.3 ng cDNA was used for each qRT-PCR that included EvaGreen as fluorescent dye and ROX as an internal reference (Biosell, Feucht, Germany). The experiments were performed in a thermal cycler (Bio-Rad, Hercules, CA) and run in biological duplicates and technical triplicates. The expression rates reported here are relative to the expression values of the housekeeping gene TEF3. All primers are listed in Table S1 in the supplemental material.
Sequence analyses.
The protein sequences of C. dubliniensis Cd36_44490 (CdCsr1), C. albicans orf19.3794 (CaCsr1), C. glabrata CAGL0J05060g (CgCsr1), S. cerevisiae YJL056C (ScZap1), Aspergillus fumigatus Afu1g10080 (ZafA), and Cryptococcus neoformants (CnZap1) were compared using NCBI PBLAST (http://blast.ncbi.nlm.nih.gov/Blast.cgi); they were aligned with the ClustalW2 multiple-sequence-alignment tool (http://simgene.com/ClustalW), and the phylogenetic tree was constructed at http://www.phylogeny.fr (34). This phylogenetic analysis includes MUSCLE (v3.7) alignment, removal of ambiguous regions with Gblocks (v0.91b), and the reconstruction of the phylogenetic tree used the maximum-likelihood method implemented in the PhyML program (v3.0 aLRT). The tree was plotted with TreeDyn (v198.3).
Pro Coffee (35) was used as a tool to align homologous promoter regions of ZRT2 from C. dubliniensis, C. albicans, and S. cerevisiae.
Chicken embryo infection model.
The embryonated chicken infection model was used to study virulence as described previously (36). Briefly, overnight cultures of yeasts were washed with PBS and adjusted to 108 cells/ml. An inoculum of 107 yeast cells/egg was applied to the chorioallantoic membrane at developmental day 10 via an artificial air chamber. In each experiment, the viability of 20 eggs per group (Candida or PBS control) was evaluated daily by candling for 7 days. Experiments were performed twice. Surviving embryos were humanely terminated by chilling on ice at the end of the experiment. All experiments were performed in compliance with the German animal protection law. According to this, no specific approval is needed for work performed on avian embryos before the time of hatching. Experiments were terminated at the latest on developmental day 18.
RESULTSThe transcription factors CdCsr1, CaCsr1, and ScZap1 are orthologous proteins.
Protein sequence comparisons using NCBI PBLAST revealed a high similarity of the C. dubliniensis protein Cd36_44490 with the C. albicans transcription factor and zinc acquisition regulator Csr1 (also known as Zap1; 86% identities and 91% positives), the S. cerevisiae zinc-responsive activator protein Zap1 (37%/52%), and the Zap1 ortholog CNAG_05392 in Cryptococcus neoformans serotype A (43%/51%). Multiple sequence alignments showed the highest similarities for the C-terminal part of the protein sequence with a high degree of conservation of this domain. Phylogenetic tree reconstruction using the protein sequences of Csr1 homologs from the yeasts C. dubliniensis (CdCsr1), C. albicans (CaCsr1), C. glabrata (CgCsr1), and S. cerevisiae (ScZap1), from the filamentous fungus A. fumigatus (ZafA), and from the basidiomycete C. neoformans (CnZap1) shows the relationship of the Csr1 proteins in fungi (Fig. 1A). The close relationship of C. dubliniensis and C. albicans is well reflected in this analysis.
Structural analysis of the homologous Csr1 proteins. (A) Phylogenetic analysis of homologous Csr1/Zap1 protein sequences from distinct fungi: Candida dubliniensis Cd36_44490 (CdCsr1), C. albicans 19.3794 (CaCsr1), C. glabrata CAGL0J05060g (CgCsr1), S. cerevisiae YJL056C (ScZap1), Cryptococcus neoformans serotype A CNAG_05392 (CnZap1), and Aspergillus fumigatus Afu1g10080 (ZafA). Construction of the phylogenetic tree used the maximum likelihood method. The scale bar indicates the genetic distance, which is proportional to the number of amino acid substitutions. (B) Comparison of the Zap1/Csr1 protein domains. The protein structure was determined using the SMART analysis service. Zap1 from S. cerevisiae contained two activation domains (AD1 and AD2; in gray) that are absent in both Candida species. Seven C2-H2-like zinc finger domains (black) were found in all analyzed species, and comparison of the C-terminal DNA binding domain (DBD) regions showed highest similarities.
The protein domain architecture of Csr1 in S. cerevisiae, C. albicans, and C. dubliniensis was then analyzed using the SMART program (34, 37). In all these homologs, the C-terminal region contains seven C2H2-like zinc finger domains (Fig. 1B). Both activation domains (ADs) present in ScZap1, AD1 and AD2 (38), were not detected in C. dubliniensis or in C. albicans.
The characteristics of the ZRT2 promoter region in S. cerevisiae allow both transcriptional activation and repression of ZRT2 via Zap1 in response to zinc levels (39). An alignment of the homologous ZRT2 promoter regions in S. cerevisiae, C. albicans, and C. dubliniensis using Pro Coffee (35) revealed a strong divergence of the zinc-responsive elements (ScZRE1 and ScZRE2) between the Candida species and S. cerevisiae (39). Of particular note, the repressive ScZRE3 region was entirely absent in Candida spp. (see Fig. S2 in the supplemental material). However, a direct comparison of C. albicans and C. dubliniensis promoter sequences revealed a high similarity between two species. This hints at a promoter type-specific, distinct regulation of the ZRT2 genes among the different yeasts.
CSR1 is essential for C. dubliniensis growth in low-zinc medium.
The aim of this study was to elucidate the functions of the transcriptional factor CdCSR1, called CSR1 here, in C. dubliniensis. To this end, we created both a csr1Δ/Δ knockout and a CSR1-complemented mutant. Additionally, the zinc responsive genes ZRT1, ZRT2, and PRA1 were expressed under the control of the constitutive ADH1 promoter both in the C. dubliniensis wild-type strain Wü248 and in the csr1Δ/Δ mutant strains. All mutants were constructed as independent duplicates and gene deletions were confirmed by Southern blot analyses (see Fig. S1 in the supplemental material).
To investigate the role of Csr1 for zinc acquisition in C. dubliniensis, growth of prestarved (24 h without zinc) wild-type and mutants strains was monitored for 2 days in defined medium (LZM) with no (0 μM), little (25 μM), or plentiful (2,000 μM) zinc. The prestarvation step was designed to largely deplete the internal zinc storage, so that fungal growth depended on the ability to acquire zinc from the surrounding medium.
The growth of wild-type and all mutant strains was nearly abolished when no zinc was added to the LZM (Fig. 2A). Under low-zinc conditions (25 μM ZnSO4), all C. dubliniensis strains harboring at least one intact CSR1 allele proliferated robustly and at a rate virtually identical to the wild-type strain Wü284, whereas most mutants lacking CSR1 (the csr1Δ/Δ, csr1Δ/Δ+ZRT1OE, and csr1Δ/Δ+PRA1OE strains) failed to adapt and grow in the low-zinc medium (Fig. 2B). Only artificial overexpression of ZRT2 in the csr1Δ/Δ mutant could largely phenocopy wild-type growth under low-zinc conditions (Fig. 2B).
Growth of wild-type and mutant strains depends on extracellular zinc levels. C. dubliniensis Wü284 and the csr1Δ/Δ, csr1Δ/Δ+CSR1, ZRT1OE, csr1Δ/Δ+ZRT1OE, ZRT2OE, csr1Δ/Δ+ZRT2OE, PRA1OE, and csr1Δ/Δ+PRA1OE strains were assayed for growth. Cells were prestarved in LZM0 for 24 h at 30°C, and afterwards strains were grown in LZM without zinc (A) and with addition of 25 μM (B) or 2000 μM (C) ZnSO4. At the starting point, the optical density at 600 nm was adjusted to 0.01, and changes were monitored every 15 min for 48 h.
Addition of 2,000 μM ZnSO4 to the LZM rescued the growth defect of all csr1Δ/Δ mutant strains (Fig. 2C). We concluded that Csr1 is a key regulator of C. dubliniensis for growth in environments with low zinc. While overexpression of neither ZRT1 nor PRA1 in the csr1Δ/Δ background improved growth under conditions of low zinc, the ZRT2 overexpression mutant displayed intermediate growth in LZM plus 25 μM ZnSO4, which indicates that this zinc transporter may play an important role under conditions of low zinc.
The upregulation of CSR1 and its target genes facilitates adaption to low zinc.
In S. cerevisiae, more than 40 putative target genes of Zap1 are known. All of these are regulated in response to zinc levels and contain zinc-responsive elements (ZREs) to which Zap1 binds (40). Additionally, C. albicans Csr1 is known to control not only zinc homeostasis but also the hypha-associated gene HWP1 under filament-inducing conditions (11) and during biofilm formation (18). To determine whether selected homologs of these target genes are also zinc responsive in C. dubliniensis, the transcription levels of genes encoding putative zinc transporters (ZRT1 to ZRT3), the zincophore gene PRA1, and the hypha-associated gene HWP1 were analyzed by quantitative real-time PCR (qRT-PCR). Cells were precultured for 24 h in LZM plus 2,000 μM ZnSO4 before these LZM-adapted cells were shifted into LZM without added zinc. This ensured that changes in gene expression were solely due to zinc deficiency and not the medium per se. At 2,000 μM zinc in the preculture, the csr1Δ/Δ mutant strains proliferated at wild-type levels (Fig. 2C). The relative gene expression was normalized to TEF3, an established C. dubliniensis reference gene used for Northern blot analyses (41).
In the wild-type strain Wü284, a 10-fold increase of CSR 1 mRNA levels was observed within the first 4 h of starvation, reflecting the transcriptional response to the absence of external zinc. The transcript levels remained highly elevated until the end of the experiment at 24 h (1,440 min) (Fig. 3A). This gives additional support to the presumptive key role for CSR1 in the upkeep of zinc homeostasis. All ZRT genes were highly (>20×) upregulated at 24 h. By 4 h, the expression of the putative low-affinity zinc transporter ZRT2 and the vacuolar zinc exporter ZRT3 was increased 37-fold and 7-fold, respectively. In contrast, ZRT1 (likely encoding a high-affinity zinc transporter) transcript levels slightly decreased within the first 4 h but reached a 125-fold upregulation after 24 h compared to the zero time point. The transcript level of PRA1 reached its measured maximum after 24 h, where this zincophore-encoding gene showed the highest transcript level of all genes investigated in the wild-type strain.
qRT-PCR gene expression analysis of putative Csr1 target genes in the wild-type strain Wü284 (A), the csr1Δ/Δ knockout strain (B), and the csr1Δ/CSR1 revertant strain (C). The analyzed genes (CSR1, ZRT1, ZRT2, ZRT3, PRA1, and HWP1) are putatively regulated by Csr1. The cells were grown in LZM0 medium, and RNA samples were taken directly after inoculation in LZM0 and 30 min, 240 min, and 1,440 min postinoculation. The bars represent the relative change in expression normalized to expression of the housekeeping gene CdTEF3, and the results are the means and standard deviations (SD) from two biological and three technical replicates. The change in expression was significant (one-way analysis of variance [ANOVA], P < 0.05) within one strain (*) or compared to the expression level in the csr1Δ/Δ knockout strain at the same sampling time point (#).
Morphologically, no hypha formation was observed under zinc limitation (data not shown), in agreement with a negligible mRNA level of HWP1 at all time points (compared to TEF3) in the wild type. As expected, no CSR1 gene expression was measured in the csr1Δ/Δ mutant, and in addition, transcript levels of the ZRT genes and of PRA1 were significantly decreased compared to those in the wild type (Fig. 3B). The absolute amount of ZRT1 transcripts was mostly below the detection limit, showing its dependency on Csr1 during zinc depletion. A slight increase was observed for ZRT2, ZRT3, and PRA1 mRNA levels after 24 h, suggesting that their expression is regulated by other factors in addition to Csr1. Reintroduction of one CSR1 allele into the csr1Δ/Δ mutant restored the overall expression pattern of CSR1 as well as of the other zinc-responsive genes, although the transcript amounts of ZRT3 and PRA1 did not fully achieve the level of the wild type (compared to TEF3) (Fig. 3C). Reintroduction of CSR1 hence largely restored the transcriptional response to zinc limitation.
Morphology of C. dubliniensis is not coupled to CSR1.
Previously, Kim et al. reported a filamentation defect for the C. albicanscsr1Δ/Δ mutant in serum-containing medium (11). To test the possible relevance of C. dubliniensis Csr1 for initiation of germ tubes, the wild type, the csr1Δ/Δ mutant, and the revertant were tested for germ tube induction in water with 10% serum or in liquid spider medium at 37°C. Invariably, all strains formed proper germ tubes under these filament-inducing conditions (Fig. 4). In addition, germ tube formation was tested for the overexpressing strains ZRT1OE, ZRT2OE, PRA1OE, csr1Δ/Δ+ZRT1OE, csr1Δ/Δ+ZRT2OE, and csr1Δ/Δ+PRA1OE. No difference relative to the wild-type phenotype was detected in any strain (data not shown). These results confirm the findings from our gene expression analysis of HWP1, and together they demonstrate that in C. dubliniensis, in contrast to C. albicans, hypha formation is not regulated by Csr1 under all our investigated conditions.
Filamentation of the wild-type strain Wü284, the csr1Δ/Δ mutant, and the csr1Δ/CSR1 revertant in 10% (vol/vol) serum and spider medium. Cells were grown overnight in YPD at 30°C and shifted to germ tube-inducing medium for 4 h at 37°C. Germ tube induction was tested for Wü284 (A), the csr1Δ/Δ mutant (B), and the csr1Δ/CSR1 strain (C) in water plus 10% serum and liquid spider medium. Cell morphology was documented via differential interference contrast microscopy. The bar represents 20 μm.
The simultaneous deletion of two zinc transporter genes TZN1 and TZN2 in Neurospora crassa caused a growth defect under zinc depletion conditions, and this double mutant strain failed to exhibit conidiation (42). In this context, we tested production of chlamydospores on Staib and rice agar under chlamydospore-inducing conditions (see Fig. S3 in the supplemental material). All strains analyzed in this study were able to produce these morphological structures in wild-type-like quality and quantity.
CSR1 is crucial for full virulence of C. dubliniensisin vivo.
To study the role of CSR1 during an infection with C. dubliniensis, we used the embryonated chicken egg model (36). We compared the virulence of the C. dubliniensis wild type, the C. dubliniensiscsr1Δ/Δ mutants, and the respective complemented strains. To allow a better estimate of C. dubliniensis' virulence, we analyzed the C. albicans wild-type strain SC5314 in parallel. C. albicans is known to generally have a higher virulence than C. dubliniensis (16), which was confirmed in our study. The average survival rate 7 days after C. albicans infections in ovo was 14%, whereas C. dubliniensis infections were survived by 44% of the embryonated eggs at the end of the experiment (Fig. 5). One of the independent C. dubliniensiscsr1Δ/Δ deletion mutants showed a significantly decreased mortality rate (33%) versus the C. dubliniensis wild type (56%) and both reconstituted strains (62% and 67%). The second csr1Δ/Δ mutant (csr1Δ/ΔB) similarly exhibited a clear, but not statistically significant, attenuated virulence with a mortality rate of 46%. The reintegration of CSR1 into the knockout strains restored the virulence pattern of the wild type. These observations indicate an important role for CSR1 during in vivo infections by C. dubliniensis.
Virulence of the wild-type strain Wü284, the csr1Δ/Δ mutant, and the csr1Δ/CSR1 revertant in infected chicken embryos. Survival after infection is depicted as Kaplan-Meyer plots. There were 20 chicken embryos per group per experiment, and the combined results of two independent experiments are shown. The mutant csr1Δ/ΔA exhibited significantly attenuated virulence (P < 0.01) compared with the wild type and the reconstituted mutant csr1Δ/CSR1A, as calculated by the log-rank (Mantel-Cox) test.
DISCUSSION
C. dubliniensis is an important emerging pathogen but is generally considered less virulent than C. albicans (43, 44). While the two fungi share many similarities, genetic, regulatory, and/or phenotypic differences must exist between them to explain this gap in virulence potential (16). The very close evolutionary relationship between C. dubliniensis and C. albicans can thus provide us with important tools to investigate the genetic basis of virulence in fungi.
One important aspect of host-pathogen interaction is the struggle for micronutrients like iron and zinc (45). In this study, we hence focused on the role of the transcriptional factor Csr1 and other putative zinc-responsive genes in zinc homeostasis of C. dubliniensis. Between C. dubliniensis and C. albicans, the transcriptional regulators CdCsr1 and CaCsr1 share a high sequence similarity. Both differ in the N-terminal zinc-responsive activation domains from their S. cerevisiae homolog, the zinc-dependent regulator Zap1 (46). Both Candida species lack AD1 and AD2 (11 and this study). In S. cerevisiae, AD1 binds multiple Zn(II) ions and is required for proper catalytic function (47). The absence of the ADs indicates differences in the structure of this zinc-responsive regulator between S. cerevisiae and the Candida species. In support of that, multiple zinc finger domains were predicted in the C-terminal region of both CdCsr1 and CaCsr1, which could allow zinc binding even in the absence of the ADs. Two out of seven C2H2 domains in both Csr1 proteins were predicted with low confidence, and other authors thus describe only five zinc finger domains in CaCsr1 (11, 48).
More than 270 genes are known to have lower transcription levels in a C. albicanscsr1Δ/Δ mutant compared to the wild type during biofilm formation (18). The largest differences in expression were found for the zinc homeostasis genes PRA1, CSR1, ZRT2, and ZRT1. Our data indicate that C. dubliniensis Csr1 shares these target genes with C. albicans, as all four genes were not upregulated during zinc limitation in the csr1Δ/Δ knockout strain.
In C. albicans, the csr1Δ/Δ mutant shows impaired growth under zinc limitation (11, 18). We observed a similar growth defect of the C. dubliniensiscsr1Δ/Δ mutant. However, in C. albicanscsr1Δ/Δ, the overexpression of the zinc transporter genes ZRT1 and ZRT2 is known to improve growth of the mutant during zinc depletion (18), while overexpression of ZRT1 or PRA1 in C. dubliniensiscsr1Δ/Δ did not lead to any phenotypic rescue. The artificial expression of this zinc transporter or the zinc scavenger protein is evidently not sufficient to allow efficient zinc uptake by C. dubliniensis. Possibly, CdZRT1 is generally less efficient in zinc uptake than CaZRT1. Alternatively, both partners of the zincophore uptake system are required in C. dubliniensis for zinc acquisition. In C. albicans, zinc uptake can occur via a zincophore system comprising both Pra1 to sequester extracellular zinc ions and Zrt1 to transport zinc into the fungal cell (25). Based on their close genetic relationship, we expect a similar mechanism to be present in C. dubliniensis. Possibly, C. albicans, but not C. dubliniensis, has sufficient remaining transcriptional activation of PRA1 and ZRT1 in the absence of CSR1 to compensate for growth defects during artificial expression of only one reaction partner under low-zinc conditions.
On the other hand, overexpression of ZRT2 in the C. dubliniensiscsr1Δ/Δ mutant allowed growth under low-zinc conditions, which hints at an important role for this transporter in such environments. In S. cerevisiae, Zrt1 is known to be the high-affinity extracellular zinc transporter (21). Comparisons of the ScZrt1 protein with the C. dubliniensis proteome revealed a higher similarity with CdZrt2 (43% identities and 59% positives) than with CdZrt1 (27%/48%), in agreement with a recent report showing a rather distant relationship between ScZrt1 and its homologous proteins in several Candida species (49). Furthermore, regulatory ZREs were not detected in the promoter sequence of the two Candida species, which points to differences in the transcriptional regulation between Candida spp. and S. cerevisiae. Therefore, we hypothesize Zrt2, rather than Zrt1, to be the high-affinity zinc transporter in C. dubliniensis.
The zinc transporters Zrt1 and Zrt2 of C. dubliniensis exhibit only 30% amino acid sequence identity, suggesting nonredundant functions. Eide showed that in S. cerevisiae, the regulation of ZRT1 and ZAP1 transcription differs from that of ZRT2, with the first two being downregulated at higher zinc concentrations (50). This indicates, for baker's yeast, the presence of both a high-affinity zinc uptake system, comprising the regulator Zap1 and the transporter Zrt1, and a low-affinity zinc uptake system mediated by Zrt2 (20). Here, we measured the expression levels of putative zinc-responsive genes in a C. dubliniensiscsr1Δ/Δ mutant and noticed a strong dependency of ZRT1 on CSR1.
Overall, data on ZRT2 gene expression in S. cerevisiae are contradictory. Bird et al. showed a peak in ZRT2 mRNA accumulation at 300 to 1,000 μM zinc (39). In a different study, a β-galactosidase activity assay demonstrated ZRT2 promoter activity under low-zinc conditions, which was reduced under conditions of increased zinc abundance (250 μM or more) (46). In our experiments, we observed a strong upregulation of C. dubliniensisZRT2 during zinc depletion. ZRT2 transcription is hence in agreement with a role for Zrt2 as a high-affinity zinc transporter in C. dubliniensis.
A detailed study on the structural basis of the transcriptional regulation of ZRT2 in S. cerevisiae revealed that one of three ZREs (ZRE3) is located inside the promoter region. Zinc deprivation results in repressional binding of Zap1 to ZRE3, which inhibits the initiation of ZRT2 transcription (39). As we observed a significant upregulation of ZRT2 in the absence of zinc, promoter regions of ZRT2 in both Candida species were aligned with sequences from S. cerevisiae. The lack of the repressing ZRE3 domain in Candida species supports our finding that ZRT2 was upregulated during zinc limitation. These differences in the promoter sequence seem to be clade specific, as Bird et al. reported a conserved ZRT2 promoter region for different Saccharomyces species (39).
Furthermore, Eide suggested an at least partial independency of ZRT2 transcription from Zap1 in S. cerevisiae (50). We observed the same phenomenon in C. dubliniensis with a delayed and reduced but measurable upregulation of ZRT2 even in the csr1Δ/Δ background. Interestingly, in addition to ZRT2, ZRT3 and PRA1 also remained responsive to zinc starvation in a csr1Δ/Δ deletion mutant. Hence, additional factors besides Csr1 likely contribute to expression of zinc-responsive genes in C. dubliniensis. Finally, it is known that ZRT1 and PRA1 share the same intergenic promoter region in C. albicans, which allows efficient zinc assimilation by their coregulation (25). The synteny of this PRA1-ZRT1 locus is conserved in C. dubliniensis, and we detected largely synchronous shifts in gene expression during zinc starvation as long as CSR1 was present.
S. cerevisiae stores zinc intracellularly under zinc-replete conditions via the vacuolar importer Zrc1. Under conditions of low extracellular zinc availability, this intracellular storage is accessed via the vacuolar zinc exporter Zrt3 (51). We observed a clear upregulation of ZRT3 in C. dubliniensis within 4 h of zinc starvation. Likely, the cells had filled their vacuolar storage during the adaption phase in 2,000 μM zinc, which was then used to maintain zinc homeostasis under starvation. We found ZRT3 upregulation to be dependent on Csr1, as ZRT3 expression never exceeded the initial levels in the csr1Δ/Δ mutant. This is in agreement with the Zap1-mediated upregulation of ZRT3 in S. cerevisiae (51).
A highly interesting aspect of Candida pathobiology is that human infections with C. dubliniensis occur much less frequently than those with C. albicans. C. dubliniensis is also far less able to disseminate into the kidney and liver in oral-intragastrically infected mice. Histological analyses of these organs revealed that C. dubliniensis remained as yeast cells in vivo, whereas C. albicans formed true hyphae and caused major tissue damage (17). Due to their potential role as a pathogenicity factor differentiating C. albicans and C. dubliniensis, we characterized the ability of a C. dubliniensiscsr1Δ/Δ deletion mutant to produce hyphae in vitro.
A filamentation defect has been observed for the C. albicanscsr1Δ/Δ deletion mutant in inducing medium, accompanied by impaired gene expression of the hypha-associated HWP1 gene (11). Similar hypha formation defects were observed in in vitro-grown biofilms and in vivo using a rat intravenous catheter model (18). In the same study, expression of hypha-associated genes like HYR1, HWP1, IHD1, and RBT1 were found to be positively regulated by Csr1, while the yeast-specific YWP1 was downregulated in a C. albicans wild-type biofilm (18).
Therefore, we investigated the capacity of C. dubliniensiscsr1Δ/Δ to induce germ tubes and found no differences relative to the wild-type strain. Hence, in contrast to C. albicans, hypha induction is not coupled to the zinc-responsive transcription factor Csr1 in C. dubliniensis. This constitutes a species-specific phenotype which may help to explain the different in vivo morphologies of the two fungi. In fact, one of the main differentiation criteria between C. dubliniensis and C. albicans is the differences in regulation of true hypha formation (52). Compared to the common ancestor, C. dubliniensis underwent reductive evolution and pseudogenization, which affected several virulence factors, including genes known to be hypha associated in C. albicans. This includes the disappearance of members of the SAP gene family, ALS3 and HYR1, and a strong divergence in the HWP1 gene, among others (53). Interestingly, the latter two are also targets of Csr1 in C. albicans (18), which might contribute to the filamentation defect in the absence of CSR1. This offers a possible explanation for the filamentation of C. dubliniensis even with a csr1Δ/Δ background. Interestingly, a C. albicanscsr1Δ/Δ mutant was also shown to produce less of the quorum-sensing molecule farnesol during biofilm formation (19). As farnesol is also able to block hypha formation in C. dubliniensis (54), our data hint at possible species-specific differences in the relation of CdCsr1 and CaCsr1 to farnesol production and/or detection.
A supply of micronutrients like zinc is essential for a microbial pathogen to survive and disseminate during an infection. Previous studies have shown that orthologs of Csr1 are essential for pathogenicity of different fungal pathogens: A murine infection with zap1 and zafA knockout strains resulted in milder forms of cryptococcosis and aspergillosis, respectively (48, 55). Very recently, the effect of a C. albicans csr1Δ/Δ deletion on virulence in mice and the associated transcriptome changes were assayed (27). In the present work, we used the embryonated egg infection model (36) for the first time to examine the virulence of wild type and mutant C. dubliniensis. This alternative infection model reflected the species-specific differences in virulence observed in human infections with C. albicans and C. dubliniensis. The survival rate of chicken embryos infected with C. dubliniensis Wü284 (44%) was more than three times higher than after infection with C. albicans SC5314 (14%) and paralleled previously published data on murine infections (68% versus 19% survival) (56).
The attenuated virulence of the C. dubliniensiscsr1Δ/Δ strains is of special interest, as hypha formation was still intact in vitro, and these results thus hint at an important role for zinc homeostasis during C. dubliniensis infections. This is also in agreement with data for CSR1 in C. albicans obtained by infection experiments in mice, where csr1Δ/Δ cells were strongly depleted in infected kidneys (26, 27). However, in C. albicans, a lack of filamentation by the CSR1 mutation may have played an additional or even dominant role besides the defect in zinc supply, although the C. albicanscsr1Δ/Δ mutant showed no reduction in hypha-associated gene expression during kidney invasion (27). Likely, important differences exist in hypha-related gene regulation by CdCsr1 and CaCsr1 (Zap1) in vitro and in vivo. Thus, our data provide an important hint at an independent contribution of the zinc supply to the success of fungal infection. Interestingly, the few virulence-associated genes verified in C. dubliniensis are generally associated with hypha formation, e.g., via calcineurin signaling (57) or telomere-associated open reading frames (ORFs) (58). Csr1, in contrast, seems be a virulence factor that is not mandatorily linked to a global filamentation defect.
In conclusion, we found that zinc homeostasis regulation by Csr1 seems to be generally conserved among C. dubliniensis, C. albicans, and S. cerevisiae, although there are important differences, especially with regard to its role in hypha formation. Furthermore, we identified Csr1 as a virulence factor in C. dubliniensis, which underlines the general relevance of micronutrient supply during fungal infections.
Supplementary Material
Supplemental material
Supplemental material for this article may be found at http://dx.doi.org/10.1128/EC.00078-15.
ACKNOWLEDGMENTS
This study was supported by Deutsche Forschungsgemeinschaft (DFG) grant STA 1147/1-1 and by the Hans-Knöll-Institute. This work was supported by the German Federal Ministry of Education and Health (BMBF) Germany, FKZ, 01EO1002, Integrated Research and Treatment Center, Center for Sepsis Control and Care (CSCC).
We thank Volha Skrahina and Daniela Schulz for technical assistance and advice and Duncan Wilson for stimulating discussions and his intellectual input.
REFERENCESSharmaA, PatniB, ShankhdharD, ShankhdharSC2013Zinc—an indispensable micronutrient.
19:11–20. doi:10.1007/s12298-012-0139-1.24381434AndreiniC, BanciL, BertiniI, RosatoA2006Zinc through the three domains of life.
5:3173–3178. doi:10.1021/pr0603699.17081069AroraG, SajidA, ArulanandhMD, MisraR, SinghalA, KumarS, SinghLK, MattooAR, RajR, MaitiS, Basu-ModakS, SinghY2013Zinc regulates the activity of kinase-phosphatase pair (BasPrkC/BasPrpC) in Bacillus anthracis.
26:715–730. doi:10.1007/s10534-013-9646-y.23793375FrohnerIE, BourgeoisC, YatsykK, MajerO, KuchlerK2009Candida albicans cell surface superoxide dismutases degrade host-derived reactive oxygen species to escape innate immune surveillance.
71:240–252. doi:10.1111/j.1365-2958.2008.06528.x.19019164MartchenkoM, AlarcoAM, HarcusD, WhitewayM2004Superoxide dismutases in Candida albicans: transcriptional regulation and functional characterization of the hyphal-induced SOD5 gene.
15:456–467. doi:10.1091/mbc.E03-03-0179.14617819FradinC, De GrootP, MacCallumD, SchallerM, KlisF, OddsFC, HubeB2005Granulocytes govern the transcriptional response, morphology and proliferation of Candida albicans in human blood.
56:397–415. doi:10.1111/j.1365-2958.2005.04557.x.15813733Kehl-FieTE, SkaarEP2010Nutritional immunity beyond iron: a role for manganese and zinc.
14:218–224. doi:10.1016/j.cbpa.2009.11.008.20015678WilsonD, CitiuloF, HubeB2012Zinc exploitation by pathogenic fungi.
8:e1003034. doi:10.1371/journal.ppat.1003034.23308062PorcheronG, GarenauxA, ProulxJ, SabriM, DozoisCM2013Iron, copper, zinc, and manganese transport and regulation in pathogenic Enterobacteria: correlations between strains, site of infection and the relative importance of the different metal transport systems for virulence.
3:90. doi:10.3389/fcimb.2013.00090.24367764AmmendolaS, PasqualiP, PistoiaC, PetrucciP, PetrarcaP, RotilioG, BattistoniA2007High-affinity Zn2+ uptake system ZnuABC is required for bacterial zinc homeostasis in intracellular environments and contributes to the virulence of Salmonella enterica.
75:5867–5876. doi:10.1128/IAI.00559-07.17923515KimMJ, KilM, JungJH, KimJ2008Roles of zinc-responsive transcription factor Csr1 in filamentous growth of the pathogenic yeast Candida albicans.
18:242–247.18309267WisplinghoffH, BischoffT, TallentSM, SeifertH, WenzelRP, EdmondMB2004Nosocomial bloodstream infections in US hospitals: analysis of 24,179 cases from a prospective nationwide surveillance study.
39:309–317. doi:10.1086/421946.15306996OddsFC, HansonMF, DavidsonAD, JacobsenMD, WrightP, WhyteJA, GowNA, JonesBL2007One year prospective survey of Candida bloodstream infections in Scotland.
56:1066–1075. doi:10.1099/jmm.0.47239-0.17644714GiriS, KindoAJ2012A review of Candida species causing blood stream infection.
30:270–278. doi:10.4103/0255-0857.99484.22885191Koga-ItoCY, KomiyamaEY, MartinsCA, VasconcellosTC, JorgeAO, CarvalhoYR, do PradoRF, BalducciI2011Experimental systemic virulence of oral Candida dubliniensis isolates in comparison with Candida albicans, Candida tropicalis and Candida krusei.
54:e278–e285. doi:10.1111/j.1439-0507.2010.01899.x.20492535MoranGP, ColemanDC, SullivanDJ2012Candida albicans versus Candida dubliniensis: why is C. albicans more pathogenic?2012:205921. doi:10.1155/2012/205921.21904553StokesC, MoranGP, SpieringMJ, ColeGT, ColemanDC, SullivanDJ2007Lower filamentation rates of Candida dubliniensis contribute to its lower virulence in comparison with Candida albicans.
44:920–931. doi:10.1016/j.fgb.2006.11.014.17251042NobileCJ, NettJE, HerndayAD, HomannOR, DeneaultJS, NantelA, AndesDR, JohnsonAD, MitchellAP2009Biofilm matrix regulation by Candida albicans Zap1.
7:e1000133. doi:10.1371/journal.pbio.1000133.19529758GangulyS, BishopAC, XuW, GhoshS, NickersonKW, LanniF, Patton-VogtJ, MitchellAP2011Zap1 control of cell-cell signaling in Candida albicans biofilms.
10:1448–1454. doi:10.1128/EC.05196-11.21890817ZhaoH, EideD1996The ZRT2 gene encodes the low affinity zinc transporter in Saccharomyces cerevisiae.
271:23203–23210. doi:10.1074/jbc.271.38.23203.8798516ZhaoH, EideD1996The yeast ZRT1 gene encodes the zinc transporter protein of a high-affinity uptake system induced by zinc limitation.
93:2454–2458. doi:10.1073/pnas.93.6.2454.8637895ZhaoH, EideDJ1997Zap1p, a metalloregulatory protein involved in zinc-responsive transcriptional regulation in Saccharomyces cerevisiae.
17:5044–5052.9271382VicentefranqueiraR, MorenoMA, LealF, CaleraJA2005The zrfA and zrfB genes of Aspergillus fumigatus encode the zinc transporter proteins of a zinc uptake system induced in an acid, zinc-depleted environment.
4:837–848. doi:10.1128/EC.4.5.837-848.2005.15879518BochA, TrampczynskaA, SimmC, TaudteN, KramerU, ClemensS2008Loss of Zhf and the tightly regulated zinc-uptake system SpZrt1 in Schizosaccharomyces pombe reveals the delicacy of cellular zinc balance.
8:883–896. doi:10.1111/j.1567-1364.2008.00414.x.18637840CitiuloF, JacobsenID, MiramonP, SchildL, BrunkeS, ZipfelP, BrockM, HubeB, WilsonD2012Candida albicans scavenges host zinc via Pra1 during endothelial invasion.
8:e1002777. doi:10.1371/journal.ppat.1002777.22761575NobleSM, FrenchS, KohnLA, ChenV, JohnsonAD2010Systematic screens of a Candida albicans homozygous deletion library decouple morphogenetic switching and pathogenicity.
42:590–598. doi:10.1038/ng.605.20543849XuW, SolisNV, EhrlichRL, WoolfordCA, FillerSG, MitchellAP2015Activation and alliance of regulatory pathways in C. albicans during mammalian infection.
13:e1002076. doi:10.1371/journal.pbio.1002076.25693184StaibF, ArastehK2001Chlamydospore formation on Staib agar. Observations made before Candida dubliniensis was described.
44:23–27.11398637HennickeF, GrumbtM, LermannU, UeberschaarN, PaligeK, BöttcherB, JacobsenID, StaibC, MorschhäuserJ, MonodM, HubeB, HertweckC, StaibP2013Factors supporting cysteine tolerance and sulfite production in Candida albicans.
12:604–613. doi:10.1128/EC.00336-12.23417561StaibP, LermannU, Blass-WarmuthJ, DegelB, WurznerR, MonodM, SchirmeisterT, MorschhäuserJ2008Tetracycline-inducible expression of individual secreted aspartic proteases in Candida albicans allows isoenzyme-specific inhibitor screening.
52:146–156. doi:10.1128/AAC.01072-07.17954688SvobodovaE, StaibP, LosseJ, HennickeF, BarzD, JozsiM2012Differential interaction of the two related fungal species Candida albicans and Candida dubliniensis with human neutrophils.
189:2502–2511. doi:10.4049/jimmunol.1200185.22851712KohlerGA, WhiteTC, AgabianN1997Overexpression of a cloned IMP dehydrogenase gene of Candida albicans confers resistance to the specific inhibitor mycophenolic acid.
179:2331–2338.9079920ReussO, VikA, KolterR, MorschhäuserJ2004The SAT1 flipper, an optimized tool for gene disruption in Candida albicans.
341:119–127. doi:10.1016/j.gene.2004.06.021.15474295SchultzJ, MilpetzF, BorkP, PontingCP1998SMART, a simple modular architecture research tool: identification of signaling domains.
95:5857–5864. doi:10.1073/pnas.95.11.5857.9600884ErbI, Gonzalez-VallinasJR, BussottiG, BlancoE, EyrasE, NotredameC2012Use of ChIP-Seq data for the design of a multiple promoter-alignment method.
40:e52. doi:10.1093/nar/gkr1292.22230796JacobsenID, GrosseK, BerndtA, HubeB2011Pathogenesis of Candida albicans infections in the alternative chorio-allantoic membrane chicken embryo model resembles systemic murine infections.
6:e19741. doi:10.1371/journal.pone.0019741.21603634LetunicI, DoerksT, BorkP2012SMART 7: recent updates to the protein domain annotation resource.
40:D302–D305. doi:10.1093/nar/gkr931.22053084BirdAJ, ZhaoH, LuoH, JensenLT, SrinivasanC, Evans-GaleaM, WingeDR, EideDJ2000A dual role for zinc fingers in both DNA binding and zinc sensing by the Zap1 transcriptional activator.
19:3704–3713. doi:10.1093/emboj/19.14.3704.10899124BirdAJ, BlankmanE, StillmanDJ, EideDJ, WingeDR2004The Zap1 transcriptional activator also acts as a repressor by binding downstream of the TATA box in ZRT2.
23:1123–1132. doi:10.1038/sj.emboj.7600122.14976557LyonsTJ, GaschAP, GaitherLA, BotsteinD, BrownPO, EideDJ2000Genome-wide characterization of the Zap1p zinc-responsive regulon in yeast.
97:7957–7962. doi:10.1073/pnas.97.14.7957.10884426MoranGP, SanglardD, DonnellySM, ShanleyDB, SullivanDJ, ColemanDC1998Identification and expression of multidrug transporters responsible for fluconazole resistance in Candida dubliniensis.
42:1819–1830.9661028KiranmayiP, TiwariA, SagarKP, HarithaA, Maruthi MohanP2009Functional characterization of tzn1 and tzn2-zinc transporter genes in Neurospora crassa.
22:411–420. doi:10.1007/s10534-008-9177-0.18989627SullivanDJ, MoranGP, PinjonE, Al-MosaidA, StokesC, VaughanC, ColemanDC2004Comparison of the epidemiology, drug resistance mechanisms, and virulence of Candida dubliniensis and Candida albicans.
4:369–376. doi:10.1016/S1567-1356(03)00240-X.14734017SullivanDJ, MoranGP, ColemanDC2005Candida dubliniensis: ten years on.
253:9–17. doi:10.1016/j.femsle.2005.09.015.16213674HoodMI, SkaarEP2012Nutritional immunity: transition metals at the pathogen-host interface.
10:525–537. doi:10.1038/nrmicro2836.22796883ZhaoH, ButlerE, RodgersJ, SpizzoT, DuesterhoeftS, EideD1998Regulation of zinc homeostasis in yeast by binding of the ZAP1 transcriptional activator to zinc-responsive promoter elements.
273:28713–28720. doi:10.1074/jbc.273.44.28713.9786867HerbigA, BirdAJ, SwierczekS, McCallK, MooneyM, WuCY, WingeDR, EideDJ2005Zap1 activation domain 1 and its role in controlling gene expression in response to cellular zinc status.
57:834–846. doi:10.1111/j.1365-2958.2005.04734.x.16045625Schneider RdeO, Fogaca NdeS, KmetzschL, SchrankA, VainsteinMH, StaatsCC2012Zap1 regulates zinc homeostasis and modulates virulence in Cryptococcus gattii.
7:e43773. doi:10.1371/journal.pone.0043773.22916306WilsonD522015An evolutionary perspective on zinc uptake by human fungal pathogens. doi:10.1039/c4mt00331d.EideDJ2003Multiple regulatory mechanisms maintain zinc homeostasis in Saccharomyces cerevisiae.
133:1532S-1535S.12730459MacDiarmidCW, GaitherLA, EideD2000Zinc transporters that regulate vacuolar zinc storage in Saccharomyces cerevisiae.
19:2845–2855. doi:10.1093/emboj/19.12.2845.10856230O'ConnorL, CapliceN, ColemanDC, SullivanDJ, MoranGP2010Differential filamentation of Candida albicans and Candida dubliniensis is governed by nutrient regulation of UME6 expression.
9:1383–1397. doi:10.1128/EC.00042-10.20639413JacksonAP, GambleJA, YeomansT, MoranGP, SaundersD, HarrisD, AslettM, BarrellJF, ButlerG, CitiuloF, ColemanDC, de GrootPW, GoodwinTJ, QuailMA, McQuillanJ, MunroCA, PainA, PoulterRT, RajandreamMA, RenauldH, SpieringMJ, TiveyA, GowNA, BarrellB, SullivanDJ, BerrimanM2009Comparative genomics of the fungal pathogens Candida dubliniensis and Candida albicans.
19:2231–2244. doi:10.1101/gr.097501.109.19745113HenriquesM, MartinsM, AzeredoJ, OliveiraR2007Effect of farnesol on Candida dubliniensis morphogenesis.
44:199–205. doi:10.1111/j.1472-765X.2006.02044.x.17257261MorenoMA, Ibrahim-GranetO, VicentefranqueiraR, AmichJ, AveP, LealF, LatgeJP, CaleraJA2007The regulation of zinc homeostasis by the ZafA transcriptional activator is essential for Aspergillus fumigatus virulence.
64:1182–1197. doi:10.1111/j.1365-2958.2007.05726.x.17542914VilelaMM, KameiK, SanoA, TanakaR, UnoJ, TakahashiI, ItoJ, YaritaK, MiyajiM2002Pathogenicity and virulence of Candida dubliniensis: comparison with C. albicans.
40:249–257. doi:10.1080/mmy.40.3.249.257.12146754ChenYL, BrandA, MorrisonEL, SilaoFG, BigolUG, MalbasFFJr, NettJE, AndesDR, SolisNV, FillerSG, AveretteA, HeitmanJ2011Calcineurin controls drug tolerance, hyphal growth, and virulence in Candida dubliniensis.
10:803–819. doi:10.1128/EC.00310-10.21531874HaranJ, BoyleH, HokampK, YeomansT, LiuZ, ChurchM, FlemingAB, AndersonMZ, BermanJ, MyersLC, SullivanDJ, MoranGP2014Telomeric ORFs (TLOs) in Candida spp. encode mediator subunits that regulate distinct virulence traits.
10:e1004658. doi:10.1371/journal.pgen.1004658.25356803GillumAM, TsayEY, KirschDR1984Isolation of the Candida albicans gene for orotidine-5′-phosphate decarboxylase by complementation of S. cerevisiae ura3 and E. colipyrF mutations.
198:179–182. doi:10.1007/BF00328721.6394964MorschhäuserJ, RuhnkeM, MichelS, HackerJ1999Identification of CARE-2-negative Candida albicans isolates as Candida dubliniensis.
42:29–32. doi:10.1046/j.1439-0507.1999.00259.x.10394844oai:pubmedcentral.nih.gov:44866762015-07-29eukcellpmc-openEukaryot CellEukaryotic CelleukcelleukcellEUKCELLEukaryotic Cell1535-97781535-9786American Society for Microbiology1752 N St., N.W., Washington, DCPMC4486676PMC44866764486676260027192600271900028-1510.1128/EC.00028-15ArticlesThe Candida albicans Exocyst Subunit Sec6 Contributes to Cell Wall Integrity and Is a Determinant of Hyphal BranchingSec6 in Cell Wall Integrity and FilamentationChavez-Dozal et al.Chavez-DozalAlba A.abBernardoStella M.abRaneHallie S.aHerreraGloriabKulkarnyVibhatiaWagenerJeanettecCunninghamIaincBrandAlexandra C.cGowNeil A. R.cLeeSamuel A.abSection of Infectious Diseases, New Mexico VA Healthcare System, Albuquerque, New Mexico, USADivision of Infectious Diseases, University of New Mexico Health Science Center, Albuquerque, New Mexico, USASchool of Medical Sciences, Institute of Medical Sciences, University of Aberdeen, Aberdeen, United KingdomAddress correspondence to Samuel A. Lee, SamALee@salud.unm.edu.
A.A.C.-D. and S.M.B. contributed equally to this work.
Citation Chavez-Dozal AA, Bernardo SM, Rane HS, Herrera G, Kulkarny V, Wagener J, Cunningham I, Brand AC, Gow NAR, Lee SA. 2015. The Candida albicans exocyst subunit Sec6 contributes to cell wall integrity and is a determinant of hyphal branching. Eukaryot Cell 14:684–697. doi:10.1128/EC.00028-15.
The yeast exocyst is a multiprotein complex comprised of eight subunits (Sec3, Sec5, Sec6, Sec8, Sec10, Sec15, Exo70, and Exo84) which orchestrates trafficking of exocytic vesicles to specific docking sites on the plasma membrane during polarized secretion. To study SEC6 function in Candida albicans, we generated a conditional mutant strain in which SEC6 was placed under the control of a tetracycline-regulated promoter. In the repressed state, the tetR-SEC6 mutant strain (denoted tSEC6) was viable for up to 27 h; thus, all phenotypic analyses were performed at 24 h or earlier. Strain tSEC6 under repressing conditions had readily apparent defects in cytokinesis and endocytosis and accumulated both post-Golgi apparatus secretory vesicles and structures suggestive of late endosomes. Strain tSEC6 was markedly defective in secretion of aspartyl proteases and lipases as well as filamentation under repressing conditions. Lack of SEC6 expression resulted in markedly reduced lateral hyphal branching, which requires the establishment of a new axis of polarized secretion. Aberrant localization of chitin at the septum and increased resistance to zymolyase activity were observed, suggesting that C. albicans Sec6 plays an important role in mediating trafficking and delivery of cell wall components. The tSEC6 mutant was also markedly defective in macrophage killing, indicating a role of SEC6 in C. albicans virulence. Taken together, these studies indicate that the late secretory protein Sec6 is required for polarized secretion, hyphal morphogenesis, and the pathogenesis of C. albicans.
INTRODUCTION
Exocytosis is an important process that plays a fundamental role in polarized growth in fungi. Exocytosis occurs when cargo-filled vesicles fuse with specific domains of the plasma membrane to provide material needed for cell wall synthesis and expansion. The final stages of polarized growth have been extensively studied in the model yeast Saccharomyces cerevisiae, and they require a highly conserved structure known as the exocyst complex (1, 2). This complex is composed of eight subunits, Sec3, Sec5, Sec6, Sec8, Sec10, Sec15, Exo70, and Exo84 (2), which function in providing spatiotemporal information for the recruitment and tethering of secretory vesicles. The process culminates when vesicles are delivered and tethered to the target membrane via SNARE-mediated membrane fusion, thereby completing the final steps of exocytosis. We have undertaken a detailed pathogenesis study in the role of the late stages of secretion; our work involves the detailed study of the components of the exocyst and SNARE complexes. We previously showed that the Candida albicans t-SNARE proteins Sso2 and Sec9 are required for hyphal growth and secretion (3). Here, we present our findings of the role of the exocyst subunit Sec6 in C. albicans secretion and filamentation.
In S. cerevisiae, SEC6 was originally discovered as a temperature-sensitive secretion mutation (4–6). SEC6 is essential for viability in S. cerevisiae, and a number of conditional mutant phenotypes associated with defective Sec6 function have been described. For example, accumulation of membrane-enclosed 80- to 100-nm vesicles and abnormal endomembrane morphology occur when a SEC6 conditional mutant strain (S. cerevisiaesec6-4) is grown at the restrictive temperature (5). Further, when grown at the permissive temperature, both wild-type Sec6 and Sec6-4 proteins localize to buds and septa; however, when these strains are grown at restrictive temperature, Sec6-4, but not wild-type Sec6, is mislocalized (6). This temperature-sensitive strain is also defective in invertase secretion (4) and polarized growth (7). The exocyst complex forms in the sec6-4 mutant at the restrictive temperature, but vesicle accumulation is still observed in the cytoplasm (8). In addition, S. cerevisiae Sec6 interacts with the plasma membrane t-SNARE Sec9, suggesting that the Sec6-Sec9 interaction is a critical intermediate in the assembly of SNARE complexes (8). Several studies have also suggested that the protein Sec1 interacts with Sec6 to regulate SNARE complex assembly (9, 10). It is thought that Sec6 interacts with the exocyst after Sec6 releases Sec9, and Sec1 is recruited simultaneously for coordinated SNARE complex formation and membrane fusion (10).
SEC6 function has been studied in multiple model systems, including Drosophila melanogaster (11), Caenorhabditis elegans (12), and S. cerevisiae (4–10), but there are no reports of its role in trafficking and polarized secretion in the pathogenic yeast Candida albicans, which has a more complex life cycle than S. cerevisiae, as it requires polarized secretion to be differentially regulated during growth in the yeast or hyphal stage. C. albicans is a polymorphic fungus of significant medical importance (13, 14) and has been used as a model for studying the molecular mechanisms of fungal pathogenesis, including polarity, secretion, and filamentation (15–17). Previous studies of the late secretory pathway in C. albicans (for example, the study of Sec3, Sec2, and the t-SNARE proteins Sso2 and Sec9) provided evidence for a key role of the exocyst and SNARE proteins in vesicle-mediated secretion and polarized hyphal growth of C. albicans (3, 18, 19). Therefore, we generated a C. albicans tetracycline-regulated SEC6 mutant strain to further investigate the role of the exocyst in polarized secretion and filamentation. We found that C. albicans Sec6 plays multiple roles in vegetative growth, cell wall biosynthesis, and virulence of this fungus.
MATERIALS AND METHODSStrains and media.
All strains used in this study are listed in Table 1. The URA3-complemented wild-type strain, THE1-CIp10 (20), was used as a control for all the experiments performed. Strains were grown at 30°C in YPD (1% [wt/vol] yeast extract, 2% [wt/vol] peptone, 2% [wt/vol] glucose) or CSM (complete synthetic medium; 0.67% [wt/vol] yeast nitrogen base without amino acids, 2% [wt/vol] glucose, 0.079% [wt/vol] complete synthetic mixture), supplemented with uridine (80 μg/ml) where required. Uracil auxotrophs were selected on 5-fluoroorotic acid (FOA) medium (CSM supplemented with 0.1 mg/ml uridine and 0.7 mg/ml FOA). Doxycycline (DOX) was added to a final concentration of 20 μg/ml to repress expression of SEC6. Solid media were prepared by adding 2% (wt/vol) agar.
Plasmids were maintained in Escherichia coli DH5α cells (Invitrogen, Carlsbad, CA) grown in LB medium (1% [wt/vol] tryptone, 0.5% [wt/vol] glucose, and 1% [wt/vol] NaCl) with ampicillin (100 μg/ml) at 37°C. Plasmid DNA was prepared from E. coli strains by using the PureYield plasmid miniprep system (Promega, Madison, WI). Genomic DNA was extracted from yeast cells by using the MasterPure yeast DNA purification kit (Epicentre Biotechnologies, Madison, WI) according to the manufacturer's instructions, with the addition of a 1-h incubation step on ice after the addition of the protein precipitation reagent.
Construction of a tetracycline-regulated C. albicansSEC6 mutant strain.
Table 2 lists the primers used in this study. Strain construction was performed as follows. First, we deleted one allele of SEC6 in the THE1 background to generate the strain sec6Δ/+ via a PCR-based gene disruption strategy previously described by Wilson et al. (21), using primers SEC6-5DR and SEC6-3DR and plasmid pDDB57 as a template. Colonies were verified by allele-specific PCR using primers SEC6-5Det and SEC6-3Det, which anneal to regions up- and downstream of the SEC6 open reading frame, respectively. Colonies that contained the correct integration of the URA3 disruption cassette (dpl200::URA3::dpl200) were designated sec6Δ/+. To regenerate uracil auxotrophy, the sec6Δ/+ strains were plated on agar medium containing 5-FOA to induce loss of URA3 by cis-recombination between the flanking dpl200 repeats. The resultant 5-FOA-resistant colonies were screened via PCR for the SEC6/sec6Δ::dpl200 genotype by using primers SEC6-5Det and SEC6-3Det. Next, the tetO promoter from plasmid p99CAU1 (22) was inserted upstream of the remaining SEC6 allele in the sec6Δ/+ FOA strains by using the PCR-based strategy previously described by Bates et al. (23) and primers tetSEC6-5DR and tetSEC6-3DR. Transformants were screened using primers tetSEC6-5Det and tetSEC6-3Det; the final constructed strain was named tetR-SEC6 (denoted as tSEC6 in the manuscript). Strain construction was verified by Southern blotting. In brief, genomic DNA digested with HindIII and EcoRV was separated on a 0.8% (wt/vol) agarose gel. DNA fragments were transferred to a positively charged nylon membrane (Roche Applied Science, Indianapolis, IN). A digoxigenin (DIG)-labeled probe (nucleotide [nt] −500 to nt 500 of orf19.5463) was prepared from genomic DNA isolated from strain THE1 with primers SEC6-5Sblt and SEC6-3Sblt (Table 2) and reagents supplied in the PCR DIG probe synthesis kit (Roche Applied Science, Indianapolis, IN). Detection of HindIII and EcoRV DNA fragments of the expected sizes for the wild-type allele (SEC6; 2.4 kb), sec6Δ/+ allele (sec6Δ::dpl200-URA3-dpl200; 2.1 kb), sec6Δ/+ FOA allele (sec6Δ::dpl200; 0.8 kb), and tSEC6 allele (URA3-tetR-SEC6; 2.5 kb and 0.5 kb) confirmed the genotype of each construct (data not shown).
Expression of SEC6 in the THE1-CIp10 control strain and tSEC6 after 2 and 4 h of growth in medium (with or without DOX) was assayed using reverse transcriptase PCR (RT-PCR) (see Fig. S1 in the supplemental material). Cells from an overnight culture were resuspended in fresh YPD with or without DOX and grown for 2 or 4 h. RNA was isolated using the RiboPure yeast RNA isolation kit (Life Technologies, Grand Island, NY) according to the manufacturer's instructions. RT-PCR was performed using the Access RT-PCR system (Promega, Madison, WI) according to the manufacturer's protocol, using primers RT-SEC6-5Det and RT-SEC6-3Det (Table 2) and 1 μg total RNA as the template. The absence of contaminating DNA was tested in parallel PCR-based analyses.
Analysis of in vitro growth.
In vitro growth was assessed using liquid and solid media. Growth on solid medium was carried out by spotting serial dilutions of cells on CSM with or without DOX, with cells prepared as described previously (24). Plates were incubated at 30°C for 24 h. Growth in liquid medium was assessed at two different temperatures (30 and 37°C) by measuring the optical density at 600 nm (OD600) at fixed intervals via an Ultraspec 2100 Pro spectrophotometer (GE Healthcare Life Sciences, Piscataway, NJ), after cells from overnight cultures were washed and transferred to fresh CSM with or without DOX and diluted to a starting OD600 of 0.1. Optical densities were recorded, and growth curves were generated using GraphPad Prism 6 (GraphPad Software, Inc., La Jolla, CA). At each time point, cells were counted using a hemocytometer, and approximately 300 cells were plated onto YPD agar to determine viability by counting CFU. CFU determinations were performed in triplicate independently.
Assessment of cellular and vacuolar morphologies.
Strains THE1-CIp10 and tSEC6 were grown overnight in 5 ml YPD, resuspended to an OD600 of 0.1, and incubated for 24 h in the presence or absence of DOX. Cells were visualized via differential interference contrast (DIC) microscopy (Carl Zeiss AG, Jena, Germany). To simultaneously stain vacuoles with FM4-64 [N-(3-triethylammoniumpropyl)-4-(6-(4-(diethylamino) phenyl) hexatrienyl) pyridinium dibromide] and CMAC (7-amino-4-chloromethyl coumarin) (both from Life Technologies, Grand Island, NY), cells were grown for 20 h in the presence or absence of DOX. Cells were inoculated into fresh medium and grown to early log phase (total incubation time, 24 h). For FM4-64 staining, cells were resuspended to an OD600 of 2 to 4 in YPD with 40 μM FM4-64 and incubated for 15 min at 30°C, resuspended in fresh YPD, and incubated for 45 min at 30°C. After incubation, the cells were resuspended to an OD600 of 0.1 in 10 mM HEPES, 5% (wt/vol) glucose, pH 7.4. CMAC was then added to a concentration of 100 μM, and cells were incubated at room temperature for 15 min prior to imaging via epifluorescence microscopy using Texas Red (FM4-64) and DAPI (CMAC) filters (25). Images were acquired and processed using AxioVision 4.7 (Carl Zeiss AG, Jena, Germany). Samples for thin section electron microscopy were prepared as described previously (3) from tSEC6 grown in the presence and absence of DOX for 24 h at 30°C, and micrographs were acquired using a Hitachi H7500 transmission electron microscope (Hitachi High-Technologies Corp., Japan).
Morphological characterization and counting of bud scars.
Strains THE1-CIp10 and tSEC6 were grown for 24 h in the presence or absence of DOX. Bud scars were visualized by chitin staining in growth medium containing 2 μg/ml calcofluor white (CW; Sigma-Aldrich, St. Louis, MO) for 15 to 30 min. Bud scar patterns were scored as follows: cells with three to five bud scars were considered “axial” if all bud scars were in a single chain, “bipolar” if at least two bud scars were at opposite ends of a cell, and “random” if the pattern was neither bipolar nor axial (26). Cells were visualized using DIC and fluorescence microscopy using a DAPI filter, and image assembly was completed using AxioVision 4.7 (Carl Zeiss AG, Jena, Germany).
Sensitivity to cell wall-perturbing agents.
The ability of strain tSEC6 to grow on medium containing cell wall stressors was tested on agar plates in the presence or absence of DOX and compared to that of controls. Plates included CSM with (i) 0.025 μg/ml caspofungin (CAS), (ii) 200 μg/ml Congo red (CR), (iii) 50 μg/ml CW, or (iv) 0.2% (wt/vol) SDS. Overnight cultures of strains tSEC6 and THE1-CIp10 were diluted for this experiment; a total of five 5-fold dilutions (starting at 108 cells/ml) were prepared in 96-well plates, and cells were stamped onto agar plates by using a multiblot replicator (VP 408H; VP Scientific Inc., San Diego, CA). Plates were incubated at 30°C for 24 h.
To test for defects in cell wall composition, cells were grown overnight with or without DOX and treated with Zymolyase 100T (β-1,3-glucanase; Sunrise Products Inc., Waterville, MN) as described previously (27). In brief, exponentially growing cells were adjusted to OD600 of 0.5 in 10 mM Tris-HCl (pH 7.5) containing 25 μg/ml of Zymolyase 100T; the decrease in optical density was monitored over 140 min, and light microscopy images were obtained every 20 min (27). Additionally, the number of cells that were unaffected by the zymolyase treatment was determined by counting the number of intact cells over 10 fields per slide per treatment at the time points indicated.
Determination of sensitivity to chitinase was analyzed as described previously (28). Chitinase from Streptomyces griseus (Sigma-Aldrich, St. Louis, MO) was dissolved in 200 mM potassium phosphate buffer (pH 6.0) at 25°C with 2 mM calcium chloride to a final concentration of 1 U/ml of chitinase. Exponentially growing cells were adjusted to an OD600 of 0.5 in phosphate buffer containing chitinase and incubated for 2 h. The decrease in optical density was then recorded every 15 min for a total of 2 h.
Analysis of cell wall composition.
Cell walls from strains THE1-CIp10 and tSEC6 (grown in the presence of absence of DOX) were extracted as previously described (29). In brief, overnight strains were subcultured and grown (with or without DOX, for 24 h at 28°C, with shaking at 250 rpm). Cells were collected (3,000 × g for 5 min) and resuspended in deionized water. Cells were fractured in a FastPrep machine (Qbiogene, Carlsbad, CA) using beads for cell disruption. Cell debris was washed five times with 1 M NaCl and resuspended in buffer (500 mM Tris-HCl [pH 7.5], 2% [wt/vol] SDS, 0.3 M β-mercaptoethanol, 1 mM EDTA), boiled for 10 min, and then freeze-dried. For quantification of cell wall components (glucan, mannan, and chitin), 2 M trifluoroacetic acid was used to acid hydrolyze cell walls. Acid was evaporated at 65°C, and samples were washed and resuspended in deionized water. The hydrolyzed cell wall composition was analyzed by high-performance anion-exchange chromatography with pulsed amperometric detection (HPAEC-PAD) in a carbohydrate analyzer system from Dionex (Surrey, United Kingdom). The total concentration of each component was determined by calibration from the standard curves of glucosamine, glucose, and mannose monomers and is reported as a percentage relative to results for the total cell wall.
Assay for secreted aspartyl proteases.
Liquid assays for secreted aspartyl proteases were performed as follows: cells from 5-ml overnight cultures were resuspended in 30 ml bovine serum albumin (BSA) medium (0.34% [wt/vol] YNB without amino acids and without ammonium sulfate, 2% [wt/vol] glucose, 0.1% [wt/vol] BSA) to induce production of Sap2 proteases, and the mixtures were incubated for 24 h. Spent medium was removed, and the cells were resuspended to an OD600 of 30 in fresh BSA medium containing only 0.01% (wt/vol) BSA. The cell suspension was divided equally, and DOX was added to one sample. Cultures were incubated at 30°C for 24 h with shaking at 250 rpm. Cell-free culture supernatant was used for SDS-PAGE analyses as described previously (20). The triple deletion mutant sapΔ(1-3) was included as a control (30). Bands of intact BSA indicate reduced secretion of Saps.
Assay for lipase secretion.
The assay performed for secreted lipases was a modification of the turbidimetric esterase assay developed by von Tigerstrom and Stelmaschuk (31). Secretion of lipase was induced by growing cells from a standard overnight culture (diluted 1:100) in Tween 80 medium (0.54% [wt/vol] YNB without amino acids, 2.5% [vol/vol] Tween 80). Following incubation at 37°C for 24 h with shaking at 250 rpm, cells were washed twice in 1× phosphate-buffered saline (PBS), concentrated to an OD600 of 10 in Tween 80 medium, and incubated for 24 h with or without DOX at 37°C with shaking at 250 rpm. Supernatants from the cultures were tested for secreted lipases in a kinetic assay: 500 μl of cell-free supernatant was added to 5 ml of Tween 20 substrate (2% [vol/vol] Tween 20 in 20 mM Tris-HCl [pH 8.0] and 120 mM CaCl2) and incubated at 37°C with shaking at 250 rpm; OD500 readings were taken every 15 min for up to 75 min. Tween 20 substrate without lipase was similarly treated and used as a control to correct for background absorbance of the substrate at each time point. The kinetic profiles of the secreted lipases under each growth condition were visualized and compared by plotting the OD500 readings over time (31). The turbidimetric assay measures the precipitation of calcium ions by degradation products resulting from lipolytic activity of enzymes secreted into the culture supernatant. The rate of increase in optical density (OD500) is proportional to the concentration of lipases present in the supernatant.
Filamentation assays.
Filamentation in liquid medium was assayed at 37°C in RPMI 1640 supplemented with l-glutamine (Cellgro, Manassas, VA) and buffered with 165 mM 3-morpholinopropanesulfonic acid (Sigma, St. Louis, MO) to pH 7.0 (“buffered RPMI 1640”). In brief, buffered RPMI 1640 with or without DOX was inoculated with cells from overnight cultures at a final density of 5 × 106 cells/ml, followed by incubation at 37°C with shaking at 200 rpm. Cells were visualized by DIC microscopy after 24 h by using a Zeiss EC Plan-NeoFluar 63×/1.25× oil objective (Carl Zeiss AG, Jena, Germany).
Filamentation was assayed on solid YPD with 10% (vol/vol) fetal calf serum (FCS), Medium 199 supplemented with l-glutamine (M199), and Spider medium, as described previously (24). YPD agar was used for embedded colony observation as described previously (32). All plates were prepared with and without 20 μg/ml doxycycline. Aliquots of 3 μl of cells from overnight cultures were spotted onto agar plates and incubated at 37°C for 24 h. For embedded colony observation, molten YPD agar was cooled to approximately 45°C and cells from overnight cultures were added to a concentration of 20 cells/ml, poured into individual petri dishes, and incubated for 24 h at 30°C. Filamentation was observed with an inverted microscope (Fisher Scientific, Waltham, MA) at 100× and 400× total magnification.
Determination of hyphal branching patterns.
In order to characterize hyphal branching patterns, strains were grown in YPD with and without DOX for 15 h, and then filamentation was induced at 37°C in liquid YPD plus 20% (vol/vol) FCS (with and without DOX) for 8 h. The numbers of branched and unbranched hyphae were counted and averaged from 100 cells. Bud-shaped or pseudohypha-shaped filaments were discounted. The degree of branching was determined by counting the number of branches emanating from the primary hyphae (33).
Macrophage killing assays.
The in vitro model of macrophage infection with C. albicans was performed as previously published (34). The J774A.1 murine macrophage cell line was purchased from ATCC (American Type Culture Collection, Manassas, VA). Macrophages were grown in high-glucose Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% (vol/vol) FCS at 37°C, 5% CO2 for 72 h. Fresh DMEM plus 10% (vol/vol) FCS was inoculated with 2 × 105 macrophages in a final volume of 0.75 ml to seed Lab-Tek chambered slides (Nalge-Nunc, Rochester, NY). Slides were then incubated at 37°C with 5% CO2 overnight. Spent medium was removed, and adherent macrophage cells were washed twice with PBS. Overnight cultures of C. albicans strains THE1-CIp10 and tSEC6 were washed three times in PBS and added to DMEM with 10% (vol/vol) FCS, with or without DOX, and 0.75 ml of this cell suspension was used for coincubation with the macrophage cells at a multiplicity of infection of 2. C. albicans cells were coincubated with the adherent macrophage cells overnight at 37°C with 5% CO2. Cells were then washed twice with PBS, and macrophage viability was assessed using the LIVE/DEAD viability/cytotoxicity kit (Invitrogen, Carlsbad, CA) following the manufacturer's instructions. Live macrophages from 12 separate fields of each chamber were counted. The experiment was performed independently three times.
Visualization of C. albicans septin localization.
To tag with green fluorescent protein (GFP) the septin ring as has been previously described (35), GFP was inserted in frame at the C terminus of CDC10 by using a nourseothricin selection method (36). Briefly, the GFP-NAT1 cassette was amplified from plasmid pGFP-NAT1 using primers CDC10-GFP-5DR and CDC10-GFP-3DR; transformation of the cassette into THE1-CIp10 and tSEC6 strains was completed using the lithium acetate method, with a 4-h growth step in YPD added after the heat shock step to allow integration and translation of the NAT1 gene before exposing the cells to nourseothricin, as described previously (36). Transformants were selected on Difco Sabouraud-dextrose agar (BD, Franklin Lakes, NJ) containing 200 μg/ml nourseothricin (Gold Biotechnology, St. Louis, MO). Primers flanking the CDC10 open reading frame (Table 2) were used to screen for transformants carrying the CDC10-GFP allele. DIC and GFP fluorescence images were acquired after induction of filamentation in buffered RPMI for 6 and 24 h in the presence or absence of DOX and in yeast cells after 24 h in YPD with or without DOX. DIC and GFP fluorescence images were acquired using AxioVision 4.7 software (Carl Zeiss AG, Jena, Germany) and a Zeiss EC Plan-NeoFluar 63×/1.25× oil objective (Carl Zeiss AG, Jena, Germany).
Statistical analyses.
Results were analyzed using a one-way analysis of variance and Tukey's multiple comparison test (GraphPad Software, Inc., La Jolla, CA). Results were considered statistically significant if P was <0.05 compared to all other treatments.
RESULTSSEC6 is required for long-term viability in C. albicans.
Since the late secretory gene SEC6 is essential for viability in S. cerevisiae, we generated a tetracycline-repressible mutant of SEC6 in C. albicans (denoted strain tSEC6). Expression of SEC6 in tSEC6 and the control strain, THE1-CIp10, in the absence or presence of DOX was analyzed by RT-PCR (see Fig. S1 in the supplemental material). SEC6 expression was detected at 2 h in strain tSEC6 in the presence of DOX, but it was not detected after 4 h. Next, we assessed the contribution of SEC6 to cell growth and viability. No visible difference in growth was observed on solid CSM with or without DOX incubated for 24 h (data not shown). There were no significant differences in OD600 values and the rate of growth between the wild-type and mutant strains when grown at either 30°C (Fig. 1A) or 37°C (Fig. 1B), under repressing or derepressing conditions. We also measured cell viability by enumerating CFU at different time points of growth at these temperatures. In the repressed state, tSEC6 cells remained viable for up to 27 and 30 h at 30 and 37°C, respectively (Fig. 1C and D), after which cell death occurred. Based on these observations, all phenotypic assessments were performed at time points that did not exceed a total of 24 h, to ensure that any differences that were observed in tSEC6 with DOX compared to wild-type controls was not a direct result of cell death.
In vitro growth of the C. albicans tSEC6 mutant strain in the presence or absence of doxycycline. (A) Growth was assessed by measuring OD600 values of microcultures at a starting OD600 of 0.05 in CSM with or without DOX and incubated with shaking at 30°C. (B) A similar assessment of growth was performed at 37°C. There was no significant difference (P > 0.05) in optical density readings after incubation for 34 h. (C) CFU analysis was performed at the indicated time points at 30°C by plating approximately 300 cells from cultures grown in liquid YPD medium and counting CFU (n = 3; bars indicate standard deviations [SD]). There was a significant difference (P < 0.05) in CFU counts between the tSEC6 mutant strain in the presence of DOX and its controls after 27 h of incubation. (D) Similarly, CFU analysis was performed at 37°C (n = 3; bars indicate SD). There was a significant difference (P < 0.05) in CFU counts between the tSEC6 mutant strain in the presence of DOX and its controls after 30 h of incubation.
Repression of SEC6 results in altered cell morphology and increased axial bud scar localization.
Wild-type strains have a typical oval-shaped yeast cell morphology (4 to 6 μm in size). In contrast, light and DIC microscopy revealed altered gross cellular morphologies in the tSEC6 strain grown under restrictive conditions. tSEC6 cells were larger than the control strain and formed chain-like structures composed of three to five cells (Fig. 2A). Chains of cells became evident after 8 h under restrictive conditions, when approximately 10% of the cells were enlarged (data not shown); after 24 h, nearly all of the tSEC6 plus DOX cells had this abnormal phenotype (Fig. 2A).
Cell morphology of tSEC6 and THE1-CIp10 strains under restrictive and nonrestrictive conditions. (A) Yeast structures were observed after 24 h of incubation in YPD with or without DOX. tSEC6 cells were larger than wild-type yeast cells and formed chains of three to five cells in length. Bar, 10 μm. (B) THE1-CIp10 and tSEC6 were incubated in YPD with or without DOX for 24 h and then stained with calcofluor white, and bud scars were observed under fluorescence microscopy using a DAPI filter. Bud scar position was evaluated by counting 100 cells and classifying the budding pattern. There was a significant increase of axial scar pattern in the tSEC6 with DOX strain and a decrease in bipolar scars (indicated by the asterisks [P < 0.001]). Three biological replicates were analyzed. Bars indicate the standard deviations.
We observed an alteration in bud scar distribution in the tSEC6 mutant plus DOX. Bud scar localization can be classified as axial or unipolar (the bud located in the one-third portion closest to the birth scar), bipolar (bud opposite to the birth scar), or random (bud on the middle third of the cell) (26). Wild-type THE1-CIp10 cells displayed a predominantly bipolar bud scar pattern, whereas tSEC6 plus DOX bud scars were predominantly axial (Fig. 2B).
Repression of SEC6 results in defective cytokinesis and abnormal septal structure.
The formation of chains of cells suggested that the tSEC6 mutant is defective in cytokinesis. Cytokinesis is the process by which a cell divides its cytoplasm to produce two daughter cells. Cytokinesis is a common process in filamentous fungi and yeast cells, and it is mediated by the formation of septa (37). To determine whether the tSEC6 strain formed normal septa under restrictive conditions, CW was used to stain the chitin of the primary septal plate. Distinct septa were observed in strain THE1-CIp10 with or without DOX and tSEC6 without DOX (Fig. 3A). However, for the tSEC6 strain with DOX, CW staining revealed an increased amount of chitin in the region of the septum (Fig. 3A) similar to that seen in S. cerevisiae strains with mutations in the septins or proteins involved in chitin deposition that localize to the mother-daughter bud neck (37–39). An abnormally thickened septum within a wide bud neck was clearly evident when the tSEC6 mutant was analyzed using thin section electron microscopy (Fig. 3B). The abnormal septal structure and the presence of multiple cellular chains indicated that the absence of SEC6 produces a defect in cytokinesis.
Bud neck structure and septin localization in strain tSEC6. (A) Strains were incubated for 24 h in YPD with or without DOX, stained with calcofluor white, and observed under a fluorescence microscope using a DAPI filter. Microscopy of cells stained with calcofluor white revealed an increased concentration of chitin at the bud neck in C. albicans tSEC6 strains under restrictive conditions. Bar, 5 μm. (B) Cells were grown in YPD with or without DOX for 24 h and fixed for analysis using thin section electron microscopy. The electron micrographs revealed the presence of a thickened septum in the bud neck of the tSEC6 mutant cultured in the presence of DOX (black arrows). Bar, 2 μm. (C) Strains carrying CDC10-GFP alleles were grown in YPD with or without DOX for 24 h. Fluorescence microscopy was used to visualize localization of Cdc10-GFP to sites of polarized growth and septin ring assembly. Strain tSEC6 cultured in the presence of DOX showed septin rings stretched laterally, compared to those found in wild-type control strains. Bar, 5 μm.
The septin Cdc10 localizes to the septin ring that is formed at the bud neck between the mother and daughter cell, marking sites where septum formation occurs. Because the morphology of the septum was abnormal in tSEC6, we asked whether septin localization was altered. Cdc10-GFP was present in the control strains at the mother-daughter bud neck and sometimes appeared as a doublet when cell division had progressed further (Fig. 3C). In contrast, an abnormally elongated septin ring structure was observed in the tSEC6 mutant strain grown under repressing conditions (i.e., with DOX) (Fig. 3C).
Repression of C. albicans SEC6 expression results in defective endocytosis and accumulation of late endosomes and post-Golgi apparatus secretory vesicles.
The fungal vacuole plays a key role in C. albicans biology and pathogenesis (34). Since defects in vacuolar function result in a number of abnormal pathogenesis-related phenotypes, we visualized the vacuolar integrity by using the fluorescent dyes FM4-64 and CMAC. FM4-64 is a lipophilic dye that binds to cellular membranes and is actively endocytosed to the vacuole, whereas CMAC is a dye that passively accumulates in the vacuolar lumen and inside late endosomes. In addition to its utility as a marker for vacuolar morphology, FM4-64 is also a marker for endocytic trafficking from the plasma membrane to the vacuolar membrane (25). Fluorescence microscopy showed that the control strains (THE1-CIp10 with or without DOX and tSEC6 without DOX) had normal spherical vacuoles, and after 45 min of incubation, FM4-64 was mostly present at the vacuolar membrane (Fig. 4A). In contrast, in the tSEC6 mutant plus DOX, FM4-64 remained primarily at the plasma membrane, indicating an impairment in endocytosis. CMAC staining of the tSEC6 strain plus DOX revealed accumulation of ovoid structures resembling late endosomes, which have been described to be readily visualized with CMAC in Aspergillus nidulans (40). We further investigated the nature of these structures by visualization of tSEC6 under derepressing and repressing conditions using thin section electron microscopy. When grown under derepressing conditions, there was no difference in ultrastructural features of tSEC6 compared to THE1-CIp10 (20). Microscopy images revealed the accumulation of post-Golgi apparatus secretory vesicles (approximate size, 40 to 90 nm) and larger ovoid structures consistent with late endosomes (approximate size, 250 nm) when strain tSEC6 was grown under repressing conditions (Fig. 4B; see also Fig. S2 in the supplemental material).
Accumulation of late endosomes and post-Golgi apparatus secretory vesicles in strain tSEC6. (A) FM4-64 (red membrane stain) and CMAC (blue vacuolar lumen stain) double fluorescent staining was performed to visualize vacuole-related organelles in control and tSEC6 strains grown with or without DOX. Strain tSEC6 growth with DOX accumulated multiple structures which were stained by CMAC. FM4-64 remained at the plasma membrane, in contrast to the controls, where it was actively endocytosed into the vacuolar membrane. Bar, 10 μm. (B) Cells were grown in YPD with or without DOX for 24 h and fixed for analysis using thin section electron microscopy. Representative micrographs are shown: (1) tSEC6 without DOX; (2 to 4) tSEC6 with DOX. The micrographs revealed accumulation of structures resembling post-Golgi apparatus secretory vesicles (black arrows) and late endosomes (white arrows) in the tSEC6 strain grown under repressing conditions that were not present under derepressing conditions. Bar, 500 nm.
SEC6 contributes to cell wall integrity.
Our results indicated that chitin content was abnormally increased at the septin ring, suggesting that the chitin/glucan ratio might be altered. To further examine the role of SEC6 in maintenance of cell wall integrity, we performed a chitinase assay which revealed that strain tSEC6 with DOX had increased resistance to chitinase compared to controls (Fig. 5A). A zymolyase assay to indirectly assess glucan content showed that tSEC6 plus DOX was markedly resistant to zymolyase degradation compared to the control strains (Fig. 5B), suggesting that SEC6 repression increased the glucan content of the cell wall. Additionally, after zymolyase treatment, we visualized the formation of Candida spheroplasts via light microscopy (Fig. 5C). The number of unaffected cells (after zymolyase treatment) was markedly increased in strain tSEC6 with DOX compared to controls (Fig. 5D). These results are congruent with the turbidity endpoint data observed in the zymolyase kinetic assay.
Qualitative determination of cell wall components via enzymatic assays. (A) Strains were incubated for 24 h with or without DOX, in the presence of chitinase. Chitinase activity was followed spectrophotometrically at a wavelength of 600 nm, at which a decrease in optical density occurs as cell wall material is degraded. Optical density readings were significantly higher for the tSEC6 mutant in the presence of DOX compared to the controls after 2 h of incubation with chitinase, indicating resistance to cell lysis. The asterisk indicates a statistically significant difference (P < 0.005) at the time point indicated and later time points. (B) Strains were incubated for 24 h with or without DOX, in the presence of Zymolyase 100T. Zymolyase activity was followed spectrophotometrically at a wavelength of 600 nm. Optical density readings were significantly higher for tSEC6 in the presence of DOX after 45 min of incubation with zymolyase, indicating resistance to cell lysis by zymolyase. The asterisk indicates statistically significant difference (P < 0.005) at the time point indicated and later time points. (C) Following incubation with zymolyase, formation of spheroplasts was visualized using light microscopy. Spheroplast formation and cell lysis were reduced in strain tSEC6 grown under restrictive conditions. Bar, 10 μm. (D) The number of cells that were unaffected by zymolyase was determined over time. Three independent experiments were performed in which the number of intact cells over 10 fields per slide per treatment were counted. There was a significant increase in the number of cells that were not affected by zymolyase in strain tSEC6 grown in the presence of DOX as early as at 30 min of incubation with zymolyase compared to the controls. The asterisk indicates a statistically significant difference between wild-type controls and tSEC6 grown with DOX (P < 0.005).
These findings suggested a possible alteration in the content of chitin and glucan in the tSEC6 conditional mutant; therefore, we next assessed growth in the presence of cell wall stressors. When grown on plates with CW (20 μg/ml), SDS (0.005% [wt/vol]), CAS (0.025 μg/ml), or CR (50 μg/ml), tSEC6 plus DOX grew poorly on YNB agar supplemented with CW (Fig. 6). This was consistent with previous studies where strains with elevated chitin content were also found to be hypersensitive to CW (41). The tSEC6 strain plus DOX also grew poorly on CSM agar supplemented with SDS, which perturbs membrane integrity. Additionally, the tSEC6 strain plus DOX grew as well as control strains on agar plates with either CR or CAS (Fig. 6).
Growth of the tSEC6 strain under conditions of cell wall stress. Replicates of serial dilutions of cells were prepared on plates containing cell wall stressors (the four types of cells included in the experiment are listed at the top and correspond to the four rows for each stressor). The triangle at the bottom indicates decreasing cell densities (1.0 × 108, 2.0 × 107, 4.0 × 106, 8.0 × 105, 1.6 × 105, and 3.2 × 104 cells/ml, from left to right). There was no difference in growth between the strains grown on complete synthetic medium. Cell wall stressors were CAS, CR, CFW, and SDS. Under repressing conditions (+DOX), C. albicans tSEC6 grew poorly in the presence of CW and SDS.
To address the discrepancy in the results of these enzymatic assays, which indirectly assessed beta-glucan content, we next directly assayed cell wall composition in the tSEC6 and THE1-CIp10 strains by using HPAEC-PAD. These analyses revealed that glucan content was similar between the controls and tSEC6, but an increase in chitin content with a corresponding decrease in mannan content was observed in the mutant (Table 3). Although these differences were not statistically significant (P = 0.30), the overall trend was consistent with the sensitivity of the conditional tSEC6 mutant to CW (Fig. 6) and the increase of chitin staining by CW at the sites of cell division (Fig. 3A).
Cell wall component analysis of Candida albicans strains
Strain (presence of DOX in medium)
Component % of cell walla
Chitin
Glucan
Mannan
THE1-CIp10
3.42 ± 1.23
59.90 ± 3.37
36.67 ± 4.51
THE1-CIp10 (+DOX)
3.42 ± 1.23
60.19 ± 2.96
36.38 ± 4.06
tSEC6
4.33 ± 1.40
59.76 ± 2.68
37.63 ± 4.76
tSEC6 (+DOX)
8.22 ± 3.22
60.11 ± 3.22
31.68 ± 6.25
Results represent the averages and standard deviations of five biological replicates.
Repression of SEC6 results in defects in hypha formation.
One of the virulence traits implicated in C. albicans pathogenesis is the ability to switch between hyphal and yeast forms of growth (16). Thus, we examined the tSEC6 mutant under conditions that are optimal for hyphal growth on solid and in liquid media for up to 24 h. When incubated on agar medium, there was a clear impairment of filamentation in the tSEC6 mutant with DOX compared to the controls (Fig. 7).
Filamentation on solid hypha-inducing medium. Filamentation was analyzed using four different types of hypha-inducing medium: cells spotted on YPD with 10% (vol/vol) FCS, colonies embedded in YPD, and cells spotted on M199 or Spider medium. Images (10× magnification) are shown for YPD-embedded colonies and cells spotted on Spider and M199 media. A severe defect in filamentation was observed in the tSEC6 strain grown under restrictive conditions. Images were acquired after 24 h at 37°C; embedded cells in YPD agar were incubated at 30°C for 24 h.
We also tested in vitro filamentation in liquid, buffered RPMI. After inducing filamentation for 24 h, cultures of control cells were composed of approximately 20% yeast cells and 80% filamentous cells, whereas the proportion of tSEC6 plus DOX cells was 55% yeast cells and 45% filamentous cells (data not shown). Compared to control strains, a marked reduction in the number of branches in the hyphal cells in tSEC6 with DOX was observed (Fig. 8A). Quantification of the percentage of branched hyphae in those cells that were able to produce hyphae in all strains showed that the number of hyphal branches was significantly decreased in tSEC6 plus DOX (P < 0.05) (Fig. 8B). To ensure that these observations were not a direct result of arrest in growth of the tSEC6 strain under filamentation-inducing conditions, we assessed cell viability by enumerating CFU from each culture. The number of CFU in the tSEC6 plus DOX culture was similar to the CFU of the THE1-CIp10 plus DOX culture and derepressed THE1-CIp10 and tSEC6 cultures (P > 0.05) (data not shown), indicating that the defect in filamentation of the tSEC6 plus DOX cultures was not a result of decreased cell viability.
Assessment of the hyphal branching pattern of the tSEC6 strain. The number of lateral branches was quantified after 15 h of incubation in YPD (with or without DOX) and subsequent induction of filamentation in YPD with 20% (vol/vol) FCS for 8 h at 37°C with or without DOX. Experiments were performed in triplicate with 100 cells counted per treatment per experiment. (A) Light microscopy of the tSEC6 strain grown with or without DOX showed differences in branching patterns. Arrows indicate branches counted for each filament (4 total branches for the tSEC6 strain without DOX and 0 branches for the tSEC6 strain with DOX). Bar, 10 μm. (B) The proportion of hyphae with the indicated number of branches of each strain grown in the presence of absence of DOX. Strain tSEC6 grown with DOX had a decrease in the number of filaments with more than one branch. The asterisk indicates a statistically significant difference from the controls (P < 0.001).
SEC6 expression is required for secretion of aspartyl proteases and lipases.
We next tested for defects in secretion of virulence-associated degradative enzymes, including Saps and lipases. In THE1-CIp10, complete proteolysis of extracellular BSA occurred (Fig. 9A), but BSA was not degraded by a mutant strain [sapΔ(1-3)] that does not express secreted aspartyl proteases Sap1, Sap2, and Sap3 (30). Strain tSEC6 with DOX was similarly unable to degrade BSA, indicating a defect in secretion of Saps. The tSEC6 mutant plus DOX also exhibited a substantial reduction in secreted lipase activity (Fig. 9B). These results indicated that SEC6 plays an important role in the secretion of Saps, extracellular lipases, and presumably other secreted proteins.
Secretion of extracellular degradative enzymes by tSEC6 strain. (A) Cells from overnight cultures were transferred to BSA medium and incubated for 24 h to induce secretion of aspartyl proteases. The cells were then washed and concentrated into fresh BSA medium with or without DOX and incubated for 24 h with shaking. Supernatants collected from these cultures were analyzed by SDS-PAGE and subsequently stained with Coomassie blue to visualize the extent of BSA degradation that occurred as a result of proteolytic cleavage by Saps released into the medium. Liquid BSA medium alone was used as a control. The triple deletion mutant sapΔ(1-3) was also included as a negative control. Bands of intact BSA are indicative of reduced secretion of Saps. BSA degradation was markedly reduced in the tSEC6 strain grown with DOX compared to wild-type controls. (B) Secretion of degradative lipases was also assessed. Cells were grown overnight in liquid YPD and transferred to Tween 80 medium for 24 h to induce production and secretion of lipases. The cells were then washed and concentrated in fresh Tween 80 medium with or without DOX and incubated for 24 h at 37°C. The supernatant was collected and used in a turbidimetric kinetic assay that measured increases in optical density (500 nm) as a result of the precipitation of calcium salts formed when degradative products are combined with CaCl2. The rate of increase in the OD500 is proportional to the concentration of lipase present in the supernatant. Three biological replicates are represented and show a significant difference between the tSEC6 strain grown with DOX and the controls (P < 0.05). Error bars indicate standard deviations.
The SEC6 mutant is defective in macrophage killing.
Phagocytes (macrophages) are one of the host's first lines of defense against Candida infection (42). Therefore, to assess virulence potential in vitro, we used a macrophage killing assay to analyze the effect of repression of SEC6. After coincubation of macrophages and C. albicans for 24 h, the control strains efficiently killed macrophages, as expected (Fig. 10A and B). In contrast, macrophage killing was reduced 3-fold when the cells were incubated with tSEC6 plus DOX, with a significant impairment compared to control strains (P < 0.05) (Fig. 10A and B).
tSEC6 shows impaired macrophage killing in vitro. (A) Following coincubation with C. albicans, macrophages were stained with calcein AM (live cells) and ethidium bromide homodimer (dead cells) and visualized by fluorescence microscopy. Representative images from three independent experiments, incubated for 24 h, are shown. The right lower panel shows an increase in live macrophages, which is indicative of defective macrophage killing by the tSEC6 mutant strain. (B) The average numbers of live macrophage cells from 12 separate fields after 24 h of coincubation with C. albicans strains are shown. Error bars indicate standard deviations. The asterisk indicates a statistically significant difference (P < 0.05) between tSEC6 in the presence of DOX versus control strains.
DISCUSSION
The exocyst complex has been extensively studied in S. cerevisiae (43, 44) and is required for polarized exocytosis in multiple organisms, including C. albicans, where secretion is vital for growth of yeast and hyphal cells, and the delivery of secreted enzymes influences pathogenesis and the host-pathogen interaction. To expand our understanding of the function of the C. albicans exocyst, we performed a detailed investigation of each component of the complex and have presented here our findings for C. albicans SEC6. In S. cerevisiae, SEC6 encodes an 85-kDa protein composed of 733 amino acids; it is predicted to be hydrophilic, and it is found in the soluble fraction of the yeast lysate (45). Using the National Center for Biotechnology Information (NCBI) protein Basic Local Alignment Search tool (BLASTp [46]) and the Candida Genome Database (CGD [47]), we identified a homologous predicted protein in C. albicans (orf19.5463) that has 24% identity and 45% similarity with S. cerevisiae Sec6. Because SEC6 is essential in S. cerevisiae, we constructed a conditional mutant to analyze gene function in C. albicans. Growth assays indicated that SEC6 is also essential for viability in C. albicans, with cell death occurring 27 h after SEC6 expression was repressed. Additionally, the C. albicans tSEC6 conditional mutant exhibited striking morphological defects in both the yeast and hyphal forms. The majority of tSEC6 yeast cells were round instead of exhibiting a typical ellipsoidal shape, and their size was enlarged. We also found marked accumulation of secretory vesicles in the conditional mutant, a phenotype that has been observed in S. cerevisiae temperature-sensitive mutants (e.g., S. cerevisiae sec6-4). Intracellular accumulation of vesicles has been associated with defects in secretion (3, 5), and it is likely that these accumulated vesicles are associated with decreased secretion of extracellular proteases and lipases in the C. albicans tSEC6 conditional mutant.
Septins have a well-established role in cytokinesis and septum formation. Septins are filamentous GTP-binding proteins that have been implicated in fungi in multiple processes, including cell separation, conjugation, sporulation, and recruitment of proteins to the bud neck (48–50). In both C. albicans and S. cerevisiae, four septins (Cdc3, Cdc10, Cdc11, and Cdc12) have been described to form a ring on the inner surface of the plasma membrane at the bud neck during cell division (50). The process of cell division can be classified into two steps in terms of septin ring formation: (i) formation of a septin hourglass structure that is observed between the anaphase and telophase stages of the cell division cycle, and (ii) formation of a two-ring structure before the onset of cytokinesis, resulting from splitting of the major ring. It has been proposed that the exocyst mainly functions in the second step (septin double rings), where vesicle trafficking increases due to a rise in demand for cell wall remodeling components, including cell wall hydrolases and glucan synthases (50, 51). We found an elongated septin ring at the bud neck in the conditional mutant compared to the control strains. Moreover, the defect in cell division in the tSEC6 mutant strain was accompanied by a localized increase in chitin content at the site of the septum. Based on previous observations in S. cerevisiae (48), it is likely that the supply of proteins necessary for mother and daughter cells to separate is interrupted in the absence of Sec6. Our results suggest that septum morphology is altered in the conditional mutant; interestingly, a previous study showed that septin mutants of C. albicans (particularly the cdc10Δ null mutant) displayed defects in cytokinesis, diffuse mislocalization of chitin, and increased axial bud scars (35). The conditional tSEC6 strain also displayed defects in cytokinesis and a high degree of axial budding, but in contrast to the cdc10Δ null mutant, we observed a focal increase in chitin content at the bud neck.
Repression of SEC6 also resulted in hypersensitivity to the cell wall stressors CW and SDS. Zymolyase, CR, and CAS affect cell wall composition by inhibiting the synthesis of β-1,3-glucan (52, 53), whereas chitinase degrades chitin in the cell wall. CW binds to nascent chitin and represses its deposition in chains, resulting in inhibition of growth (54). In agreement with previous studies where it has been shown that a decrease in content of one of the main C. albicans cell wall components (glucan, chitin, and mannoproteins) results in a compensatory effect in order to maintain cell wall integrity (52, 55, 56), we observed a trend toward decreased mannose and increased chitin content in the tSEC6 mutant strain cultured with DOX, although these differences were not statistically significant. The observed increased resistance to chitinase degradation was consistent with the raised levels of chitin in the cell wall of the tSEC6 strain grown under repressing conditions. However, it was interesting to also see increased resistance to zymolyase degradation despite the unaltered total glucan content. The resistance to zymolyase could be a result of structural changes in the glucan network in the tSEC6 mutant strain. One possibility is that repression of SEC6 results in trafficking defects which impair delivery of cell wall maintenance and remodeling enzymes. Alternatively, there may be alterations in cell wall structure in the conditional mutant that cause architectural changes in the cell wall, making it less accessible to degradation by zymolyase and chitinase.
Under repressing conditions, the tSEC6 mutant had a striking impairment in filamentation, producing hyphae that lacked the normal branching pattern. During hyphal development of wild-type strains, true hyphae grow with parallel cell walls (57). In the conditional mutant, hyphal cells were less parallel in the distal segments and, notably, exhibited a marked decrease in formation of lateral branches. We also studied the contribution of Sec6 to pathogenesis by performing an in vitro model of macrophage infection (34). Macrophage killing was significantly attenuated in the conditional mutant grown under repressing conditions compared to the controls. Wild-type C. albicans strains induce macrophage death by forming filaments in response to the phagosome environment (34, 58). Thus, it is likely that the severe defect in filamentation was responsible for the attenuated virulence in macrophages. Taken together, we have demonstrated that C. albicansSEC6 is required for polarized growth, which is vital for tissue penetration, lateral branching of hyphae, cytokinesis, and the secretion of virulence-associated enzymes.
Our recent work on C. albicans t-SNAREs Sso2 and Sec9 demonstrated their role in hyphal growth and secretion; however, the conditional mutants exhibited clear phenotypic differences despite their predicted interactions in the SNARE complex (3). Repression of SSO2, but not SEC9, resulted in hyphal growth arrest and a return to isotropic growth with a corresponding loss of the C. albicans Spitzenkörper at the hyphal tip. Interestingly, when grown under hypha-inducing conditions as described previously (3), the Spitzenkörper is also still present at the hyphal tip of the tSEC6 strain when grown under repressing conditions (A. A. Chavez-Dozal et al., unpublished results). This similarity can be explained in part by the suggested physical interaction of Sec6 and Sec9 in S. cerevisiae (8), which brings two proteins from different complexes together, spatially and temporally, in a key step in exocytosis. It is further believed that release of Sec9 from this interaction allows formation of the SNARE complex (8) and is concomitant with binding of Sec6 with other components of the exocyst complex (10), leading to fusion between the vesicles and target membranes. As we continue our detailed examination of the role of the individual subunits of the exocyst in C. albicans, we are finding differences in their involvement in secretion and polarized growth that are distinct from those reported in previously published work regarding S. cerevisiae and that represent a novel finding in this opportunistic pathogen. It would be of further interest to see if these findings extend to filamentous fungi and/or are limited to pathogenic fungi.
Supplementary Material
Supplemental material
Supplemental material for this article may be found at http://dx.doi.org/10.1128/EC.00028-15.
ACKNOWLEDGMENTS
We thank Hironobu Nakayama (Suzuka University of Medical Science, Japan) for providing strain THE1 and plasmid p99CAU1; Aaron P. Mitchell (Carnegie Mellon University) for providing plasmid pDDB57; and Steven Bates (University of Exeter) for providing plasmid pGFP-NAT1. We thank Stephen Jett (UNM Health Sciences Center Electron Microscopy Facility) for assistance with thin section electron microscopy. Sequence data for C. albicans was obtained from the Stanford DNA Sequencing and Technology Center website (http://www-sequence.stanford.edu/group/candida). Sequencing of C. albicans was accomplished with the support of the NIDCR, NIH, and the Burroughs Wellcome Fund.
This work was supported by funding from the Department of Veterans' Affairs (Merit Award 5I01BX001130 to S.A.L.), the Biomedical Research Institute of New Mexico (S.A.L.), National Institutes of Health grants T32AI007538 (S.M.B.) and K12GM088021 (A.A.C.-D.). N.A.R.G. and A.C.B. thank the Wellcome Trust and Royal Society, respectively, for funding (097377, 101873, and UF080611).
REFERENCESHeiderMR, MunsonM2012Exorcising the exocyst complex.
13:898–907. doi:10.1111/j.1600-0854.2012.01353.x.22420621TerBushDR, MauriceT, RothD, NovickP1996The exocyst is a multiprotein complex required for exocytosis in Saccharomyces cerevisiae.
15:6483–6494.8978675BernardoSM, RaneHS, Chavez-DozalA, LeeSA2014Secretion and filamentation are mediated by the Candida albicans t-SNAREs Sso2p and Sec9p.
14:762–775. doi:10.1111/1567-1364.12165.24911595NovickP, FieldC, SchekmanR1980Identification of 23 complementation groups required for post-translational events in the yeast secretory pathway.
21:205–215. doi:10.1016/0092-8674(80)90128-2.6996832NovickP, FerroS, SchekmanR1981Order of events in the yeast secretory pathway.
25:461–469. doi:10.1016/0092-8674(81)90064-7.7026045LampingE, TanabeK, NiimiM, UeharaY, MonkBC, CannonRD2005Characterization of the Saccharomyces cerevisiaesec6-4 mutation and tools to create S. cerevisiae strains containing the sec6-4 allele.
361:57–66. doi:10.1016/j.gene.2005.07.014.16185821SongerJA, MunsonM2009Sec6p anchors the assembled exocyst complex at sites of secretion.
20:973–982. doi:10.1091/mbc.E08-09-0968.19073882SivaramMV, SaporitaJA, FurgasonML, BoettcherAJ, MunsonM2005Dimerization of the exocyst protein Sec6p and its interaction with the t-SNARE Sec9p.
44:6302–6311. doi:10.1021/bi048008z.15835919HongW, LevS2014Tethering the assembly of SNARE complexes.
24:35–43. doi:10.1016/j.tcb.2013.09.006.24119662MorgeraF, SallahMR, DubukeML, GandhiP, BrewerDN, CarrCM, MunsonM2012Regulation of exocytosis by the exocyst subunit Sec6 and the SM protein Sec1.
23:337–346. doi:10.1091/mbc.E11-08-0670.22114349BeronjaS, LapriseP, PapoulasO, PellikkaM, SissonJ, TepassU2005Essential function of Drosophila Sec6 in apical exocytosis of epithelial photoreceptor cells.
169:635–646. doi:10.1083/jcb.200410081.15897260MabonME, MaoX, JiaoY, ScottBA, CrowderCM2009Systematic identification of gene activities promoting hypoxic death.
181:483–496. doi:10.1534/genetics.108.097188.19047414MorganJ, MeltzerMI, PlikaytisBD, SofairAN, Huie-WhiteS, WilcoxS, HarrisonLH, SeabergEC, HajjehRA, TeutschSM2005Excess mortality, hospital stay, and cost due to candidemia: a case-control study using data from population-based candidemia surveillance.
26:540–547. doi:10.1086/502581.16018429WisplinghoffH, BischoffT, TallentSM, SeifertH, WenzelRP, EdmondMB2004Nosocomial bloodstream infections in US hospitals: analysis of 24,179 cases from a prospective nationwide surveillance study.
39:309–317. doi:10.1086/421946.15306996MoleroG, Diez-OrejasR, Navarro-GarciaF, MonteolivaL, PlaJ, GilC, Sanchez-PerezM, NombelaC1998Candida albicans: genetics, dimorphism and pathogenicity.
1:95–106.10943347CalderoneRA, FonziWA2001Virulence factors of Candida albicans.
9:327–335. doi:10.1016/S0966-842X(01)02094-7.11435107KabirMA, HussainMA, AhmadZ2012Candida albicans: a model organism for studying fungal pathogens.
2012:538694. doi:10.5402/2012/538694.23762753LiCR, LeeRT, WangYM, ZhengXD, WangY2007Candida albicans hyphal morphogenesis occurs in Sec3p-independent and Sec3p-dependent phases separated by septin ring formation.
120:1898–1907. doi:10.1242/jcs.002931.17504812BishopA, LaneR, BenistonR, Chapa-y-LazoB, SmytheC, SudberyP2010Hyphal growth in Candida albicans requires the phosphorylation of Sec2 by the Cdc28-Ccn1/Hgc1 kinase.
29:2930–2942. doi:10.1038/emboj.2010.158.20639857BernardoSM, KhaliqueZ, KotJ, JonesJK, LeeSA2008Candida albicans VPS1 contributes to protease secretion, filamentation, and biofilm formation.
45:861–877. doi:10.1016/j.fgb.2008.01.001.18296085WilsonRB, DavisD, EnloeBM, MitchellAP2000A recyclable Candida albicans URA3 cassette for PCR product-directed gene disruptions.
16:65–70. doi:10.1002/(SICI)1097-0061(20000115)16:1<65::AID-YEA508>3.0.CO;2-M.10620776NakayamaH, MioT, NagahashiS, KokadoM, ArisawaM, AokiY2000Tetracycline-regulatable system to tightly control gene expression in the pathogenic fungus Candida albicans.
68:6712–6719. doi:10.1128/IAI.68.12.6712-6719.2000.11083786BatesS, HughesHB, MunroCA, ThomasWP, MacCallumDM, BertramG, AtrihA, FergusonMA, BrownAJ, OddsFC, GowNA2006Outer chain N-glycans are required for cell wall integrity and virulence of Candida albicans.
281:90–98. doi:10.1074/jbc.M510360200.16263704RaneHS, BernardoSM, RainesSM, BinderJL, ParraKJ, LeeSA2013Candida albicansVMA3 is necessary for V-ATPase assembly and function and contributes to secretion and filamentation.
12:1369–1382. doi:10.1128/EC.00118-13.23913543ConibearE, StevensTH2002Studying yeast vacuoles.
351:408–432.12073360NiL, SnyderM2001A genomic study of the bipolar bud site selection pattern in Saccharomyces cerevisiae.
12:2147–2170. doi:10.1091/mbc.12.7.2147.11452010CasanovaM, ChaffinWL1991Cell wall glycoproteins of Candida albicans as released by different methods.
137:1045–1051. doi:10.1099/00221287-137-5-1045.1865179MonrealJ, ReeseET1969The chitinase of Serratia marcescens.
15:689–696. doi:10.1139/m69-122.4894282LeeKK, MaccallumDM, JacobsenMD, WalkerLA, OddsFC, GowNA, MunroCA2012Elevated cell wall chitin in Candida albicans confers echinocandin resistance in vivo.
56:208–217. doi:10.1128/AAC.00683-11.21986821HubeB, SanglardD, OddsFC, HessD, MonodM, SchaferW, BrownAJ, GowNA1997Disruption of each of the secreted aspartyl proteinase genes SAP1, SAP2, and SAP3 of Candida albicans attenuates virulence.
65:3529–3538.9284116von TigerstromRG, StelmaschukS1989The use of Tween 20 in a sensitive turbidimetric assay of lipolytic enzymes.
35:511–514. doi:10.1139/m89-079.2501015ClearyIA, SavilleSP2010An analysis of the impact of NRG1 overexpression on the Candida albicans response to specific environmental stimuli.
170:1–10. doi:10.1007/s11046-010-9297-2.20232156BarelleCJ, RichardML, GaillardinC, GowNA, BrownAJ2006Candida albicansVAC8 is required for vacuolar inheritance and normal hyphal branching.
5:359–367. doi:10.1128/EC.5.2.359-367.2006.16467476PalmerGE, KellyMN, SturtevantJE2005The Candida albicans vacuole is required for differentiation and efficient macrophage killing.
4:1677–1686. doi:10.1128/EC.4.10.1677-1686.2005.16215175WarendaAJ, KonopkaJB2002Septin function in Candida albicans morphogenesis.
13:2732–2746. doi:10.1091/mbc.E02-01-0013.12181342MilneSW, CheethamJ, LloydD, AvesS, BatesS2011Cassettes for PCR-mediated gene tagging in Candida albicans utilizing nourseothricin resistance.
28:833–841. doi:10.1002/yea.1910.22072586HowellAS, LewDJ2012Morphogenesis and the cell cycle.
190:51–77. doi:10.1534/genetics.111.128314.22219508SanzM, CastrejonF, DuranA, RonceroC2004Saccharomyces cerevisiae Bni4p directs the formation of the chitin ring and also participates in the correct assembly of the septum structure.
150:3229–3241. doi:10.1099/mic.0.27352-0.15470103SchmidtM, VarmaA, DrgonT, BowersB, CabibE2003Septins, under Cla4p regulation, and the chitin ring are required for neck integrity in budding yeast.
14:2128–2141. doi:10.1091/mbc.E02-08-0547.12802080AbenzaJF, GalindoA, PinarM, PantazopoulouA, de los RiosV, PenalvaMA2012Endosomal maturation by Rab conversion in Aspergillus nidulans is coupled to dynein-mediated basipetal movement.
23:1889–1901. doi:10.1091/mbc.E11-11-0925.22456509WalkerLA, GowNA, MunroCA2013Elevated chitin content reduces the susceptibility of Candida species to caspofungin.
57:146–154. doi:10.1128/AAC.01486-12.23089748NeteaMG, BrownGD, KullbergBJ, GowNA2008An integrated model of the recognition of Candida albicans by the innate immune system.
6:67–78. doi:10.1038/nrmicro1815.18079743LipschutzJH, MostovKE2002Exocytosis: the many masters of the exocyst.
12:R212–R214. doi:10.1016/S0960-9822(02)00753-4.11909549BoydC, HughesT, PypaertM, NovickP2004Vesicles carry most exocyst subunits to exocytic sites marked by the remaining two subunits, Sec3p and Exo70p.
167:889–901. doi:10.1083/jcb.200408124.15583031PotenzaM, BowserR, MullerH, NovickP1992SEC6 encodes an 85 kDa soluble protein required for exocytosis in yeast.
8:549–558. doi:10.1002/yea.320080706.1523887AltschulSF, GishW, MillerW, MyersEW, LipmanDJ1990Basic local alignment search tool.
215:403–410. doi:10.1016/S0022-2836(05)80360-2.2231712InglisDO, ArnaudMB, BinkleyJ, ShahP, SkrzypekMS, WymoreF, BinkleyG, MiyasatoSR, SimisonM, SherlockG2012The Candida genome database incorporates multiple Candida species: multispecies search and analysis tools with curated gene and protein information for Candida albicans and Candida glabrata.
40:D667–D674. doi:10.1093/nar/gkr945.22064862KozubowskiL, LarsonJR, TatchellK2005Role of the septin ring in the asymmetric localization of proteins at the mother-bud neck in Saccharomyces cerevisiae.
16:3455–3466. doi:10.1091/mbc.E04-09-0764.15901837BerepikiA, ReadND2013Septins are important for cell polarity, septation and asexual spore formation in Neurospora crassa and show different patterns of localisation at germ tube tips.
8:e63843. doi:10.1371/journal.pone.0063843.23691103BertinA, McMurrayMA, GrobP, ParkSS, GarciaG3rd, PatanwalaI, NgHL, AlberT, ThornerJ, NogalesE2008Saccharomyces cerevisiae septins: supramolecular organization of heterooligomers and the mechanism of filament assembly.
105:8274–8279. doi:10.1073/pnas.0803330105.18550837SudberyPE2001The germ tubes of Candida albicans hyphae and pseudohyphae show different patterns of septin ring localization.
41:19–31. doi:10.1046/j.1365-2958.2001.02459.x.11454197LesageG, SdicuAM, MenardP, ShapiroJ, HusseinS, BusseyH2004Analysis of beta-1,3-glucan assembly in Saccharomyces cerevisiae using a synthetic interaction network and altered sensitivity to caspofungin.
167:35–49. doi:10.1534/genetics.167.1.35.15166135RonceroC, DuranA1985Effect of Calcofluor white and Congo red on fungal cell wall morphogenesis: in vivo activation of chitin polymerization.
163:1180–1185.3897187ElorzaMV, RicoH, SentandreuR1983Calcofluor white alters the assembly of chitin fibrils in Saccharomyces cerevisiae and Candida albicans cells.
129:1577–1582.6352860KaeberleinM, GuarenteL2002Saccharomyces cerevisiaeMPT5 and SSD1 function in parallel pathways to promote cell wall integrity.
160:83–95.11805047KapteynJC, RamAF, GroosEM, KollarR, MontijnRC, Van Den EndeH, LlobellA, CabibE, KlisFM1997Altered extent of cross-linking of beta-1,6-glucosylated mannoproteins to chitin in Saccharomyces cerevisiae mutants with reduced cell wall beta-1,3-glucan content.
179:6279–6284.9335273SudberyP, GowN, BermanJ2004The distinct morphogenic states of Candida albicans.
12:317–324. doi:10.1016/j.tim.2004.05.008.15223059LorenzMC, BenderJA, FinkGR2004Transcriptional response of Candida albicans upon internalization by macrophages.
3:1076–1087. doi:10.1128/EC.3.5.1076-1087.2004.15470236oai:pubmedcentral.nih.gov:45197522015-08-27eukcellpmc-openEukaryot CellEukaryotic CelleukcelleukcellEUKCELLEukaryotic Cell1535-97781535-9786American Society for Microbiology1752 N St., N.W., Washington, DCPMC4519752PMC45197524519752260710342607103400064-1510.1128/EC.00064-15ArticlesSpotlightA Minimal Set of Glycolytic Genes Reveals Strong Redundancies in Saccharomyces cerevisiae Central MetabolismYeast Minimal GlycolysisSolis-Escalante et al.Solis-EscalanteDanielKuijpersNiels G. A.Barrajon-SimancasNuriavan den BroekMarcelPronkJack T.http://orcid.org/0000-0003-3136-8193DaranJean-Marchttp://orcid.org/0000-0002-4097-7831Daran-LapujadePascaleDepartment of Biotechnology, Delft University of Technology, Delft, The NetherlandsAddress correspondence to Pascale Daran-Lapujade, p.a.s.daran-lapujade@tudelft.nl.
D.S.-E. and N.G.A.K. contributed equally to this article.
Citation Solis-Escalante D, Kuijpers NGA, Barrajon-Simancas N, van den Broek M, Pronk JT, Daran J-M, Daran-Lapujade P. 2015. A minimal set of glycolytic genes reveals strong redundancies in Saccharomyces cerevisiae central metabolism. Eukaryot Cell 14:804–816. doi:10.1128/EC.00064-15.
As a result of ancestral whole-genome and small-scale duplication events, the genomes of Saccharomyces cerevisiae and many eukaryotes still contain a substantial fraction of duplicated genes. In all investigated organisms, metabolic pathways, and more particularly glycolysis, are specifically enriched for functionally redundant paralogs. In ancestors of the Saccharomyces lineage, the duplication of glycolytic genes is purported to have played an important role leading to S. cerevisiae's current lifestyle favoring fermentative metabolism even in the presence of oxygen and characterized by a high glycolytic capacity. In modern S. cerevisiae strains, the 12 glycolytic reactions leading to the biochemical conversion from glucose to ethanol are encoded by 27 paralogs. In order to experimentally explore the physiological role of this genetic redundancy, a yeast strain with a minimal set of 14 paralogs was constructed (the “minimal glycolysis” [MG] strain). Remarkably, a combination of a quantitative systems approach and semiquantitative analysis in a wide array of growth environments revealed the absence of a phenotypic response to the cumulative deletion of 13 glycolytic paralogs. This observation indicates that duplication of glycolytic genes is not a prerequisite for achieving the high glycolytic fluxes and fermentative capacities that are characteristic of S. cerevisiae and essential for many of its industrial applications and argues against gene dosage effects as a means of fixing minor glycolytic paralogs in the yeast genome. The MG strain was carefully designed and constructed to provide a robust prototrophic platform for quantitative studies and has been made available to the scientific community.
INTRODUCTION
Gene duplication plays a key role in evolution by providing DNA templates for evolutionary innovation, while preventing interference with the cellular function of the original genes (1–3). Immediately after gene duplication, the resulting paralog pairs are usually identical and therefore functionally redundant. Unless duplication confers a selective advantage, either via gene dosage effects or via mutational acquisition of modified or new functions, duplicated genes will eventually be pseudogenized and/or lost from the genome (2, 3).
Gene duplication played a key role in the evolutionary history of the yeast Saccharomyces cerevisiae, an intensively investigated model in eukaryotic evolutionary biology. A whole-genome duplication event (WGD) in an ancestor of S. cerevisiae, ca. 100 million years ago, was followed by loss of ca. 90% of the resulting gene duplications (4, 5). Despite the long time interval that separates current S. cerevisiae strains from the WGD, many surviving paralog pairs still exhibit a substantial degree of functional redundancy, as indicated by neutral or weak phenotypes of single-paralog deletion mutants (6–9). However, the selective pressures that caused the long-term retention of these functionally redundant paralogs remain elusive (1, 10).
The Embden-Meyerhof-Parnas (EMP) pathway, the main route for oxidation of glucose to pyruvate in all eukaryotes and in many other organisms, is among the most slowly evolving metabolic pathways (11, 12). In all taxa, paralog families are found for the structural genes that encode the EMP enzymes (12–14). Under conditions of oxygen limitation and/or sugar excess, S. cerevisiae couples the EMP pathway to the fermentative production of ethanol via pyruvate decarboxylase and NAD+-dependent alcohol dehydrogenase (15, 16). In this article, we use the term “glycolysis” to indicate the set of 12 enzyme-catalyzed reactions in yeast that convert intracellular glucose to ethanol.
In S. cerevisiae, no fewer than 8 of the 12 enzyme reactions in glycolysis are represented by multiple paralogous genes (Fig. 1). This incidence represents a significant overrepresentation (P = 1.9 × 10−10, based on hypergeometric distribution analysis) relative to the ca. 26% of the yeast genome that consists of parologous combinations (17). The WGD event and resulting duplication of genes involved in central metabolism have been implicated in the appearance of several key physiological characteristics of S. cerevisiae. In particular, the duplication of glycolytic genes has been proposed to have contributed to the strong tendency of S. cerevisiae to produce ethanol under aerobic conditions (Crabtree effect) and its high glycolytic capacity (18–20). However, the impact of reducing the number of glycolytic paralogs on these and other physiological characteristics of S. cerevisiae has not been systematically explored. In all paralogous gene sets in yeast glycolysis, with the notable exception of the phosphofructokinase gene, gene expression and gene deletion studies support the definition of a single major paralog and one to four minor paralogs (Fig. 1). Except for the pseudogenes GPM2 and GPM3, all paralogs have retained their original catalytic function, although their context-dependent expression profiles differ (Fig. 1). Deletion of minor paralogs for individual glycolytic enzymes has minor effects on enzyme activities in cell extracts and on the specific growth rate under standard laboratory conditions (usually shake flask cultivation on yeast extract, peptone, and glucose) (21–25) (Fig. 1). Several hypotheses have been forwarded to explain the neutral effect of paralog deletion (6, 26–31). These include experimental limitations, such as the poor sensitivity of fitness screens (32) and the narrow range of cultivation condition tested (generally complex medium) (33). Additionally, analysis of deletion mutants in which paralogs that encode a single glycolytic enzyme are inactivated cannot reveal synergistic effects of the minor paralogs of different glycolytic enzymes. Such synergistic effects might, for example, arise from the well-documented phenomenon of distribution of metabolic control over multiple enzymes in metabolic pathways (34, 35) or from other regulatory or catalytic interactions.
Separation of major and minor glycolytic paralogs in Saccharomyces cerevisiae. The following eight enzymes in yeast glycolysis are encoded by parologous genes: hexokinase/glucokinase (HXK), phosphofructokinase (PFK), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), phosphoglycerate mutase (GPM), enolase (ENO), pyruvate kinase (PYK), pyruvate decarboxylase (PDC), and alcohol dehydrogenase (ADH). Percentages of gene and protein similarity (GeSy and PrSy, respectively) between paralogs, the type of duplication event (DuEv), whole-genome duplication (WGD) or other small scale duplication (SSD), and the proposed fate (pseudogenization [P], subfunctionalization [S], or neofunctionalization [N]) of the different paralogs are indicated in the tables in each panel. Separation between WGD and small-scale duplications pre-WGD (SSD*) and post-WGD (SSD) was based on the information in the Yeast Gene Order Browser (YGOB; http://ygob.ucd.ie/) (119). Scatter plots show the comparisons between expression levels of the different glycolytic paralogs in S. cerevisiae measured under 170 different conditions (75) (see Table S6 in the supplemental material). Except for PFK, where both paralogs PFK1 and PFK2 were considered paralogs with equivalent contributions, the expression levels of the major glycolytic paralogs (labeled in red) are indicated by a red line. Bar graphs display in vitro enzyme activities of glycolytic enzymes in mutants carrying individual and multiple glycolytic gene knockouts. Values are presented as percentages of in vitro enzyme activities measured in reference strains (RS [denoted by an orange dotted line]) as reported in the literature (21–25).
Insight into the importance, under laboratory conditions, of the minor glycolytic paralogs is essential for understanding an apparent genetic redundancy in a key ubiquitous metabolic pathway in an important model organism and industrial platform. In addition, analysis of the extent of genetic redundancy in central metabolism is highly relevant for the complete redesign and construction of entirely synthetic yeast genomes, as pursued in the Synthetic Yeast 2.0 initiative (36). Moreover, if complexity in yeast glycolysis can be significantly reduced by elimination of “redundant” isoenzymes, this could eliminate uncertainties and thereby facilitate the formulation and validation of mathematical models that describe the kinetics of this key metabolic pathway (37, 38).
The goals of the present study are to experimentally explore genetic redundancy in yeast glycolysis by cumulative deletion of minor paralogs and to provide a new experimental platform for fundamental yeast research by constructing a yeast strain with a functional “minimal glycolysis” (MG). To this end, we deleted 13 minor paralogs, leaving only the 14 major paralogs for the S. cerevisiae glycolytic pathway. The cumulative impact of deletion of all minor paralogs was investigated by two complementary approaches. A first, quantitative analysis focused on the impact on glycolytic flux under a number of controlled cultivation conditions that in wild-type strains result in different glycolytic fluxes. These quantitative growth studies were combined with transcriptome, enzyme activity, and intracellular metabolite assays to capture potential small phenotypic effects. A second, semiquantitative characterization explored the phenotype of the “minimal glycolysis” (MG) strain under a wide array of experimental conditions to identify potential context-dependent phenotypes.
MATERIALS AND METHODSStrains and strain construction.
Plasmid propagation and isolation were performed with chemically competent Escherichia coli DH5α (Z-competent transformation kit; Zymo Research, Orange, CA) cultivated in lysogeny broth (LB) medium (39, 40) supplemented with 100 mg liter−1 ampicillin (LBAmp) when required. All yeast strains are derived from the CEN.PK family (41–43) and are listed in Table S3 in the supplemental material. All strains were stored at −80°C in 1-ml aliquots of 30% glycerol and the appropriate medium. CEN.PK102-12A was selected as the parental strain for the minimal glycolysis (MG) strain. The order of gene deletions that led to the final MG strain IMX372 was GLK1, HXK1, TDH1, TDH2, GPM2, GPM3, ENO1, PYK2, PDC5, PDC6, ADH2, ADH5, and ADH4. The genes GLK1, HXK1, TDH1, and TDH2 were deleted using the auxotrophic and dominant markers Sphis5 (44, 45), KlLEU2 (45, 46), KanMX (47, 48), and hphNT1 (49, 50), respectively. These markers remained in the genome during the whole strain construction process. GPM2 to ADH4 deletions were performed using a strategy of selection and counterselection with the KlURA3/5-fluoroorotic acid (5-FOA) system (51) for the recovery of the marker module. KlURA3 was removed seamlessly as previously described (52, 53). The dominant marker modules KanMX and hphNT1 were removed using deletion cassettes containing KlURA3 and amdSYM, respectively. These markers were sequentially removed as reported previously (54). To obtain a prototrophic strain, a cassette containing the marker module KlURA3 was integrated in the TDH1 locus; this generated the MG strain. All genetic modifications were confirmed by PCR and later by whole-genome sequencing.
Molecular biology techniques.
All integrative cassettes were constructed using Phusion Hot Start II high-fidelity polymerase (Thermo Scientific, Landsmeer, The Netherlands) following the manufacturer's recommendations and with the primer pairs and plasmids listed in Tables S4 and S5 in the supplemental material, respectively, as the templates. Correct integrations, deletions, and marker excision were confirmed by PCR using Dreamtaq polymerase (Thermo Scientific) and following the manufacturer's recommendations. The primers used for confirmation are listed in Table S4. Genomic DNA that served as the template for PCR was obtained by extraction with 0.05 N NaOH directly from single colonies or by purification using the YeaStar genomic DNA kit (Zymo Research, Orange, CA) following the manufacturer's recommendations. All PCR products were loaded on gels containing 1% (wt/vol) agarose (Thermo Scientific) and 1× Tris-acetate-EDTA (TAE) buffer (Thermo Scientific). Integrative cassettes were gel purified using the Zymoclean gel DNA recovery kit (Zymo Research). Yeast strain transformations were performed with the lithium acetate protocol as previously described (55). Plasmids were isolated from E. coli with the GenElute plasmid miniprep kit (Sigma-Aldrich, St. Louis, MO).
Media.
Complex and nonselective media for growth rate determination and propagation contained 10 g liter−1 yeast extract, 20 g liter−1 peptone, and 20 g liter−1 glucose (YPD). When selection was required, YPD was supplemented with 200 mg liter−1 G418 (YPD+G418) or 200 mg liter−1 hygromycin (YPD+Hyg). Synthetic medium (SM) containing 3 g liter−1 KH2PO4, 0.5 g liter−1 MgSO4·7H2O, 5 g liter−1 (NH4)2SO4, 1 ml liter−1 of a trace element solution, and 1 ml liter−1 of a vitamin solution as previously described (56). SM was supplemented with 20 g liter−1 glucose (SMG) for propagation, growth rate determination, and batch cultures. When auxotrophic strains were cultivated, SMG was supplemented with 150 mg liter−1 uracil, 125 mg liter−1 histidine, and/or 500 mg liter−1 leucine, according to strain auxotrophy (57). Selection of strains lacking the KlURA3 marker was done in SMG supplemented with 150 mg liter−1 uracil and 1 g liter−1 5-fluoroorotic acid (SMG–5-FOA). Solid media were obtained by addition of 20 g liter−1 agar or agarose. Unless otherwise stated, liquid cultures were performed in 500-ml shake flasks with a working volume of 100 ml and incubated at 30°C and 250 rpm, the targeted starting optical density at 660 nm (OD660), measured using a Libra S11 spectrophotometer (Biocrom, Cambridge, United Kingdom), being 0.5. Unless otherwise stated, cultures on solid media were incubated for 3 days at 30°C.
Semiquantitative phenotypic characterization on plates.
Growth of the MG strain and of the prototrophic control strain CEN.PK113-7D was performed on agar plates containing YPD or SM plus 2% glucose (SMG), SM plus 2% galactose (SMGal), SM plus 2% maltose (SMMal), SM plus 2% sucrose (SMSuc), SM plus 3% (vol/vol) ethanol (SMEtOH), SM plus 6% glycerol (SMGlyc), SM plus 5% glucose (SMHG), SM plus 0.5% glucose (SMLG), SMG plus 2.5, 5, or 10 mM LiCl (SMG-Lit), SMG plus 200 or 500 mM NaCl (SMGSa), SMG plus 50, 100, or 200 μM CdCl2 (SMGCad), SMG plus 1, 1.5, or 2 M sorbitol (SMGSor), SMG at pH 4, 6, or 7.5 (SMGpH), or SMG plus 1 mM H2O2 (SMGOx). Spot plates were inoculated with 101 to 105 cells obtained by serial dilutions of liquid cultures grown to the exponential phase on SMG. All plates were incubated at 30°C for 3 days, except for plates containing SMGlyc, which were incubated for 6 days.
Quantitative characterization of the MG strain's physiology in shake flasks.
Growth rate determination was performed by OD660 measurement of cultures grown in 500-ml shake flasks containing 100 ml of YPD, SMG, or SMEtOH. Cultures for growth rate measurements were inoculated with precultures grown until the late exponential phase under identical conditions (i.e., the same medium and temperature).
Glucose/ethanol/glucose switches were started by inoculation of SMG shake flasks with a preculture grown overnight on the same medium. At the mid-exponential phase (OD660 of ca. 4) samples were taken, washed twice with sterile demineralized water, and used to inoculate shake flasks containing 100 ml of SMEtOH. When reaching an OD660 of 3, a sample from the culture was washed twice with sterile demineralized water and used to inoculate a 500-ml shake flask containing 100 ml SMG. Growth was then monitored until the late exponential phase.
To evaluate tolerance of MG to oxidative stress, cells grown to the exponential phase in shake flasks containing SMG were transferred to a shake flask containing SMG supplemented with 4.5 mM H2O2. Samples were then taken every 20 min for 1 h, diluted, and plated on SMG plates to reach approximately 100 cells per plate. CFU were used to determine the percentage of surviving cells.
Quantitative characterization of the MG strain's physiology in the bioreactor.
Aerobic and anaerobic batch cultures were performed in 2-liter laboratory fermenters (Applikon, Schiedam, The Netherlands) with a 1-liter working volume. SM was used and set at pH 5 before autoclaving at 121°C. Antifoam emulsion C (Sigma) was added to a concentration of 0.2 g liter−1 from a 20% (vol/vol) solution autoclaved separately (121°C). Prior to inoculation, glucose was added to a final concentration of 20 g liter−1 from a sterile (110°C) 50% glucose solution and 1 ml of a vitamin solution (56). Anaerobic cultures were supplemented with 0.01 g liter−1 ergosterol and 0.42 g liter−1 Tween 80 dissolved in ethanol as previously described (56). The pH was maintained at 5 by the automatic addition of 2 M KOH. Compressed air or gaseous nitrogen (quality, 4.5; >99.995% vol N2, <10 volumes per million [vpm] O2 pollution) (Linde Gas, Schiedam, The Netherlands) for aerobic and anaerobic cultures, respectively, was sparged to the bioreactor at a rate of 500 ml min−1 via an Ion Science Saga digital flow meter (Cambridge, United Kingdom). The temperature in the fermenters was kept at 30°C. Complete mixing of the medium was ensured by stirring at 800 rpm. Four independent batch cultures were performed for each condition and each strain. Precultures were started by inoculation of SMG shake flasks with 1 stock vial of the appropriate strain. After ca. 16 h of incubation, these cultures were used to inoculate new SMG shake flasks. The starting OD660 was tightly controlled to reach an OD660 of 1 after 8 h of incubation. These exponentially growing cultures were washed twice with demineralized water (spinning at 5,000 rpm for 4 min) and resuspended in 100 ml demineralized water. This was the cell suspension that was used to inoculate fermenters with a starting OD660 of 0.1.
The OD660 of diluted cell suspensions was measured as described above. Biomass dry weight was determined by filtration as previously described (56). The concentration of extracellular metabolite in culture supernatants was measured by high-performance liquid chromatography (HPLC) analysis using a Aminex HPX-87H ion-exchange column operated at 60°C with 5 mM H2SO4 as the mobile phase at a flow rate of 0.6 ml min−1.
Chromosome separation using CHEF electrophoresis.
Agarose plugs for the different strains were prepared using the contour-clamped homogenous electric field (CHEF) yeast genomic DNA plugs kit (Bio-Rad, Richmond, CA), following the manufacturer's recommendations, and used for CHEF electrophoresis. CHEF electrophoresis was performed as previously described (54).
Vacuole staining.
MG and CEN.PK113-7D vacuoles were stained with the red fluorescent dye FM4-64 (excitation/emission, ∼515/640 nm) (Thermo Fisher Scientific) following the manufacturer's recommendations. Yeast cells and vacuoles were visualized with an Imager-Z1 microscope equipped with an AxioCam MR camera, an EC Plan-Neofluar 100×/1.3 oil Ph3 M27 objective, and the filter set BP 535/25, FT 580, and LP 590 (Carl-Zeiss, Oberkochen, Germany).
Enzyme activities.
Samples equivalent to 62.5 mg of dry weight biomass taken at the mid-exponential phase (ca. 10 h after inoculation) of aerobic batch cultures were used to obtain cell extracts as previously described (58). Measurement of the activity of glycolytic enzymes was carried out as previously described (59), except for phosphofructokinase, the activity of which was determined as described in reference 60. Enzyme activities are expressed as micromoles of substrate per minute per milligram of protein or units per milligram of protein. Protein concentrations in cell extracts were determined according to reference 61 with bovine serum albumin as a standard.
Intracellular metabolite determination.
Samples (1.2 ml each) were taken from aerobic batch cultures using a rapid sampling setup (62) and placed directly into vials containing 5 ml 80% methanol precooled at −40°C. The samples were washed with precooled 80% methanol, and the extraction was performed with boiling ethanol as previously described (63). 13C-labeled cell extract was used as an internal standard for metabolite quantification (64). Intracellular glucose (Gluc), glucose-6-phosphate (Gluc-6-P), fructose-6-phosphate (Fruc-6-P), dihydroxyacetone phosphate (DHAP), glyceraldehyde-3-phosphate (GAP), 3-phosphoglycerate (3PG), 2-phosphoglycerate (2PG), and pyruvate (Pyr) were measured by gas chromatography-mass spectrometry (GC-MS) according to reference 65. Intracellular fructose-1,6-bisphosphate (Fruc-1,6-bP) and phosphoenolpyruvate (PEP) were measured by liquid chromatography-mass spectrometry (LC-MS) according to reference 66. Acetaldehyde determination was performed as previously described (67).
RNA-seq.
Samples for transcriptome sequencing (RNA-seq) were obtained from aerobic batch cultures at the mid-exponential phase of growth on glucose (ca. 10 h after inoculation). Sampling, rapid quenching in liquid nitrogen, and RNA extraction were performed as previously described (68). Sequencing was performed with the Illumina HiSeq 2500 and carried out by BaseClear (Leiden, The Netherlands). A data set of 51-bp single reads of at least 1 Gb was generated. The genome sequence of CEN.PK113-7D (42) was used for all analyses. The data were aligned to the reference using the Burrow-Wheeler alignment tool BWA (69, 70). Gene expression levels were estimated using fragments per kilobase per million (FPKM) values by the Cufflinks software (71). To identify differential gene expression between strains CEN.PK113-7D and IMX372, RNA-seq data comparison was performed and statistically assessed using Cuffdiff (71).
Whole-genome sequencing.
High-quality genomic DNA of the strains CEN.PK102-12A and IMX372 (MG) was obtained using the Qiagen 100/G kit (Qiagen, Hilden, Germany) following the manufacturer's recommendations. Libraries of 300-bp inserts were constructed and paired end sequenced (100-bp reads) using an illlumina HiSeq 2500 sequencer (Baseclear BV, Leiden, The Netherlands). A minimum data quantity of 950 Mb was generated, representing a minimum 80-fold coverage. The sequence reads were mapped onto CEN.PK113-7D genome (42) using the Burrows-Wheeler Alignment tool BWA and further processed using SAMtools (69, 70). Single-nucleotide variations were extracted from the mapping using SAMtools' varFilter. Default settings were used, except that the maximum read depth was set to 400× (−D400). To minimize false-positive mutation calls, custom Perl scripts were used for further mutation filtering as follows: (i) mutation calls containing ambiguous bases in mapping consensus were filtered out, (ii) only the single-nucleotide variations with a quality of at least 20 were kept (with variant quality defined as the Phred-scaled probability that the mutation call is incorrect [72, 73]), (iii) mutations with a depth of coverage below 10× were discarded, and (iv) the mutations found in CEN.PK102-12A were subtracted from the list sequence. Eventually, the single-nucleotide variations were physically positioned and functionally annotated according to the CEN.PK113-7D sequence annotation.
Microarray data accession number.
The RNA-seq data generated in this study have been submitted to the Genome Expression Omnibus database and assigned accession no. GSE63884. The sequence data generated in this study are searchable at NCBI–Entrez (http://www.ncbi.nlm.nih.gov/) under Bioproject PRJNA269221.
RESULTSFrom 27 to 14 glycolytic genes: minimal yeast glycolysis.
To enable construction of a “minimal glycolysis” (MG) yeast strain, the first goal was to identify the major paralogs that should be retained in the final strain design. This assessment was based on information from the literature on phenotypes of relevant deletion mutants and on nonglycolytic roles (“moonlighting functions”) (74) of glycolytic isoenzymes. Furthermore, a compendium of S. cerevisiae transcriptome data obtained under a wide range of controlled cultivation conditions (75) was used to compare expression profiles of glycolytic paralogs. Major paralogs were elected based on the following, nonexclusive criteria (Fig. 1; see Fig. S1 in the supplemental material): (i) the highest transcript level over a range of growth conditions, (ii) the most extensive loss of enzyme activity in cell extracts upon deletion, (iii) moonlighting functions with a strong impact on specific growth rate or robustness (HXK2, ENO1, and ENO2) (76, 77), and (iv) the strongest decrease in specific growth rate upon deletion.
Based on these criteria, 11 of the 27 glycolytic genes in S. cerevisiae were assessed to be major paralogs: HXK2, PGI, FBA1, TPI1, TDH3, GPM1, ENO2, PGI1, PYK1 (CDC19), PDC1, and ADH1, while 13 minor paralogs (i.e., HXK1, GLK1, TDH1, TDH2, GPM2, GPM3, ENO1, PYK2, PDC5, PDC6, ADH2, ADH4, and ADH5) were selected for gene deletion. In S. cerevisiae, the PKF1 and PFK2 paralogs have evolved into subunits of a hetero-octameric phosphofructokinase (78). Since deletion of either gene substantially decreases fitness (79–81) (see Fig. S1 in the supplemental material), both were retained in the minimal glycolysis design. ADH3 encodes a mitochondrial alcohol dehydrogenase involved in an anaerobic redox shuttle across the mitochondrial inner membrane (82). To prevent reduced growth rates under anaerobic conditions, ADH3 was therefore also retained in the minimal glycolysis design.
Sequential deletion of the 13 minor paralogs in a haploid S. cerevisiae strain belonging to the CEN.PK family (41–43) yielded the prototrophic strain IMX372, which we will refer to as the “MG” (minimal glycolysis) strain. Resequencing of the genome of the MG strain confirmed the correct deletion of all 13 minor glycolytic genes. Although transformation of S. cerevisiae can be mutagenic (83, 84), only 20 single-nucleotide differences were detected in the MG strain relative to its ancestor strain, CEN.PK102-12A. Eleven of these differences occurred within open reading frames (ORFs), of which 10 resulted in amino acid changes (see Table S1 in the supplemental material). None of these mutations affected glycolytic genes or genes related to central carbon metabolism. Whole-genome sequencing and karyotyping indicated the duplication of two short sections of chromosome III (16.5 kbp, from YCR019W to YCR027C) and chromosome V (19.1 kbp, from YER093C-A to YER104W) (see Fig. S2 and Table S1 in the supplemental material). These regions do not carry genes involved in central carbon metabolism, and no interchromosomal rearrangements were observed.
Elimination of all minor glycolytic paralogs has minimal impacts on growth kinetics, intracellular metabolite concentrations, and gene expression.
To explore the physiological impact of the simultaneous deletion of all 13 minor glycolytic paralogs, we quantitatively compared specific growth rates, product formation, and gene expression in the MG strain and in a congenic, prototrophic reference strain with a full complement of glycolytic genes. In these studies, a synthetic, chemically defined medium was used, since complex media do not enable cells to express their full genetic potential (33) and complicate quantitative physiological analysis.
In aerobic, glucose-grown bioreactor batch cultures, specific rates of growth, substrate consumption, and product formation were not significantly affected by deletion of all 13 minor glycolytic paralogs (Fig. 2A). The only exception to this observation concerned acetate production, which was slightly higher in the MG strain. Also after the diauxic shift, where the glycolytic flux changes direction as the aerobic yeast cultures consumed the ethanol and acetate produced during the initial growth phase on glucose, the biomass formation and substrate consumption rates of the two strains were virtually identical. In anaerobic yeast cultures, the absence of oxidative phosphorylation makes glycolysis the only pathway for energy conservation. Also under these more demanding conditions, the specific growth rate and metabolic fluxes of the MG strain did not differ significantly from those of the congenic reference strain (Fig. 2B).
Biomass production and extracellular metabolite profiles from aerobic and anaerobic batch cultures in bioreactors of the minimal glycolysis (MG) strain and a congenic reference strain (indicated by “C”). Shown are biomass and extracellular metabolite profiles from aerobic (A) and anaerobic (B) controlled batch cultures of the minimal glycolysis strain (MG) (open circles) and the prototrophic reference strain CEN.PK113-7D (closed circles). The data shown in the graphs are the average and average deviation of the mean from two independent cultures for each strain. The specific growth rate (μ [per hour]) and biomass-specific rates of glucose consumption (qs), ethanol production (qEtOH), glycerol production (qGlyc), and acetate production (qAce) (all expressed in millimoles per gram dry weight per hour) represent the average and standard deviation of data from four independent cultures for each strain. *, statistically significant (by two-tailed t test assuming the same variance in the populations) differences between the two tested strains (P = 0.03 and P = 0.02 for μ and qAce, respectively, in aerobic cultures).
To further compare growth of the MG strain in aerobic bioreactor batch cultures, its glycolytic enzyme activities in cell extracts, intracellular metabolite concentrations, and transcriptome were compared with those of a congenic reference strain. Remarkably, no significant differences were observed in the transcript levels of any of the major glycolytic paralogs or in vitro glycolytic enzyme activities (Fig. 3). With the exception of slightly higher intracellular concentrations of acetaldehyde, concentrations of glycolytic intermediates in the MG strain did not significantly differ from those in the reference strain. To explore potential impacts of the deletion of 13 minor glycolytic paralogs outside glycolysis, genome-wide transcript levels of the MG strain and the reference strain were compared. As few as 17 genes showed a significantly different transcript level (see Table S2 in the supplemental material), 12 of which were located on the duplicated regions on chromosomes III and V. An in-depth, comprehensive analysis of the MG strain in glucose-grown bioreactor batch cultures therefore failed to identify substantial impacts of the minor glycolytic paralogs on fluxes, intracellular metabolite concentrations, or gene expression.
In vitro enzyme activities and intracellular metabolite concentrations in aerobic batch cultures in bioreactors of the minimal glycolysis (MG) strain and a congenic reference strain. Thirteen minor glycolytic paralogs were deleted in the MG strain. (A) Average values of four independent culture replicates of the in vitro activities for the glycolytic enzymes in the MG strain (white bars) and the prototrophic reference strain CEN.PK113-7D (black bars). HXK, hexokinase/glucokinase; PGI, phosphoglucose isomerase; PFK, phosphofructokinase; FbPA, fructose-bisphosphate aldolase; TPI, triose-phosphate isomerase; GAPdh, glyceraldehyde-3-phosphate dehydrogenase; PGK, phosphoglycerate kinase; GPM, phosphoglycerate mutase; ENO, enolase; PYK, pyruvate kinase; PDC, pyruvate decarboxylase; ADH, alcohol dehydrogenase. The denoted error bars represent standard deviations. (B) Intracellular glycolytic metabolite profiles of the MG strain (open circles) and of CEN.PK113-7D (close circles) from aerobic batch cultures. Average values from two independent culture replicates are shown, and the average deviations of the mean are indicated by error bars. The vertical orange dotted line indicates the time at which glucose was depleted.
Minimal impact of minor glycolytic paralogs under a wide range of conditions.
If during evolution of S. cerevisiae, its glycolytic paralogs have evolved different roles through sub- or neofunctionalization, their deletion may only cause an observable phenotype under specific growth conditions. The growth rates of the MG and reference strains were therefore compared under a wide range of selected growth conditions. During fast growth in shake flasks on complex (yeast extract-peptone-glucose [YPD]) medium (Fig. 4) and during growth at low temperature (12°C) (Fig. 4), the glycolytic pathway operates at rates closer to its maximum capacity than during growth at 30°C on synthetic medium (60, 85). However, no difference in specific growth rates between the MG and reference strains was observed under these conditions. The MG strain also showed the same growth rate as the reference strain at high temperature (37°C) (Fig. 4).
Growth of the minimal glycolysis (MG) strain and a congenic reference strain under different growth conditions in shake flasks. (A) Specific growth rates of S. cerevisiae strains MG (white bars) and CEN.PK113-7D (reference strain, black bars). SMG, synthetic medium with glucose; SMEtOH, synthetic medium with ethanol as the carbon source; YPD, complex medium with glucose as the carbon source. The labels “37°C” and “12°C” indicate growth on SMG at 37° and 12°C, respectively. Average specific growth rates (per hour) are denoted above each bar. (B) Growth, measured as change in the culture's optical density at 660 nm (OD660), of the MG strain (open circles) and reference strain (closed circles) during carbon source switches. Strains were successively grown in glucose, ethanol, and glucose. All data represent the average and average deviation of the mean from two independent culture replicates.
Absence of the minor glycolytic paralogs did not affect growth of the MG strain on ethanol (Fig. 2A and 4A). These results supported the conclusion from the aerobic bioreactor batch cultures that efficient gluconeogenesis does not require any of the minor paralogs (as has previously been proposed for Adh2 [86]). While minor glycolytic paralogs apparently do not contribute to maximum specific growth rates on individual carbon sources, they might be involved in substrate transitions. For instance, Pyk2, the glucose-repressed and fructose-1,6-bisphosphate-insensitive paralog of Pyk1 (21), has been proposed to prevent futile cycling resulting from simultaneous operation of glycolysis and gluconeogenesis during transitions between fermentable and nonfermentable carbon sources (87, 88). However, no significant impact of the deletion of minor glycolytic paralogs was observed during transitions from glucose to ethanol and back to glucose (Fig. 4B).
To investigate possible phenotypes of the MG strain under a wider range of environmental conditions, growth on solid medium was investigated under 24 experimental conditions. Some of these were chosen without a specific focus on individual glycolytic paralogs. For example, growth was analyzed at high osmotic pressure (Fig. 5a to c), on different carbon sources (Fig. 5p to w), and at high concentrations of lithium and sodium ions, to which strains from the CEN.PK lineage are hypersensitive (89) (Fig. 5h to l). Only growth with glycerol as a carbon source indicated a slightly reduced growth rate of the MG strain. However, this phenotype was not observed during growth on liquid medium with glycerol (see Fig. S3 in the supplemental material) and may therefore have been caused by a contaminating substrate in the agar used for the plate experiments.
Growth of the minimal glycolysis (MG) strain and a congenic reference strain on different solid media. Serial dilutions of cell suspensions of the MG strain and of the reference strain CEN.PK113-7D were plated on agar media with 24 different compositions. The conditions included osmotic stress (a to c), different pHs (d to f), oxidative stress (g), three different salts (h to o), six different carbon sources (p to w), and complex medium (x). With the exception of panel x (YPD), all plates contained synthetic medium (SM). All plates contained 2% glucose, with the exception of plates with different carbon sources, which contained galactose (2%), maltose (2%), sucrose (2%), ethanol (3%, vol/vol), or glycerol (6%) and plates with low (0.03 M [5%]) and high (0.3 M [50%]) concentrations of glucose. Colonies of reference and MG strains are denoted by black bands and orange bands, respectively, at the bottom of each panel.
Additional growth conditions tested on solid media were chosen based on information from the literature (74, 90) on sub- or neofunctionalization of specific deleted glycolytic paralogs. In addition to their enolase activity, Eno1 and Eno2 have both been implicated in vacuolar assembly, but in the absence of Eno1, Eno2 is able to support both functions (76). In the vacuole and the ATPase associated with it, disruption causes a deleterious phenotype in S. cerevisiae when cultivated under alkaline conditions (91). As expected, the MG strain grew normally at pH 7.5, thereby indicating the absence of major vacuolar malfunction (Fig. 5f; see Fig. S4 in the supplemental material). Tdh3 and Tdh2, the two minor isoenzymes of glyceraldehyde-3-phosphate dehydrogenase, have been proposed to protect S. cerevisiae against oxidative stress because of their different levels of thiolation (92). However, exposure to oxidative stress by growth on hydrogen peroxide-containing plates (Fig. 5g) showed that the absence of Tdh2 did not affect the oxidative stress resistance of the MG strain. Sub- and neofunctionalization have resulted in a different glucose-dependent regulation of GLK1/HXK1 and PYK2 and in different regulatory properties of the proteins that they encode from those encoded by HXK2 and PYK1, respectively (21, 93). However, growth of the MG strain at low, intermediate, and high glucose concentrations was not impaired relative to that of the reference strain (Fig. 5p to r). Finally, PDC6, which encodes a pyruvate decarboxylase isoenzyme with a low cysteine and methionine content, is specifically induced under sulfur limitation (94). Cultivation in the presence of cadmium increases abundance of Pdc6 (94–96); since cells require a high level of glutathione production for detoxification, sulfur amino acids are then used for this process and are less available for protein synthesis. Despite the absence of PDC6, the MG strain grew as well as the reference strain in the presence of cadmium.
DISCUSSIONAbsence of phenotypes upon deletion of 13 glycolytic genes.
Deletion of 13 of the 27 glycolytic paralogs in S. cerevisiae yielded a “minimal glycolysis” (MG) yeast strain whose most spectacular characteristic was the absence of any pronounced phenotype under a wide range of laboratory growth conditions. One of the hypotheses that has been proposed to explain retention of functionally redundant paralogs during evolution is a contribution to gene dosage and, thereby, to the capacity of the pathway or process in which they operate (97, 98). The high glycolytic rates in anaerobic cultures of the MG strain (18 mmol glucose per g of dry biomass per h) did not significantly differ from those of wild-type strains (99) (Fig. 2). This result argues against gene dosage effects as a means for fixing minor glycolytic paralogs in the yeast genome. Instead, our observations indicate that duplication of glycolytic genes is not a prerequisite for achieving the high glycolytic fluxes and fermentative capacities that are characteristic of S. cerevisiae and essential for many of its industrial applications (100, 101). It might be argued that the gene dosage hypothesis does apply to phosphofructokinase, as deletion of either PFK1 or PFK2 substantially reduces enzyme activity and fitness (79–81) (see Fig. S1 in the supplemental material). However, the presence of Pfk1 and Pfk2 in a hetero-octameric complex can also be seen as a case of neofunctionalization, in which two paralogs have been fixed by an acquired mutual dependency. The near-wild-type growth kinetics of the MG strain in aerobic and anaerobic cultures is difficult to reconcile with the hypothesis that duplication of glycolytic genes during the WGD event played a major role in increasing its glycolytic capacity or in causing the phenomenon of aerobic fermentation (the Crabtree effect) (20).
Our inability to identify a phenotype after deletion of all minor paralogs of glycolytic genes in S. cerevisiae does not imply that such a phenotype does not exist. The range of conditions tested represents an infinitesimal fraction of the environmental conditions to which S. cerevisiae may have been exposed to in its evolutionary history. The absence of a clear phenotype under standard laboratory conditions makes the MG strain an even more interesting platform for future high-throughput studies to investigate its phenotype under more conditions. Other characteristic features of S. cerevisiae like tolerance to high ethanol concentrations or growth at high gravity are interesting conditions to test, as well as the robustness of the MG strain under nonstandard conditions, such as sporulation, starvation, severe calorie restriction, and dynamic nutrient supply regimens. However, guessing the environmental factors that have conferred an evolutionary advantage to strains carrying gene duplications presents a formidable challenge, more particularly because the real niche for S. cerevisiae is still a matter of debate, to the extent that S. cerevisiae may be considered a generalist that does not necessarily favor a specific type of environment (102). For further functional analysis studies of the minor glycolytic paralogs, the set of intermediate strains used for construction of the MG strains (see Table S3 in the supplemental material) can be used to rapidly identify which minor paralog or paralogs contribute to any newly identified phenotypes. Moreover, the sensitivity of fitness analyses can be improved—for example, by competitive cultivation of wild-type and MG strains in mixed cultures.
In S. cerevisiae, overrepresentation of paralogous gene sets is not unique for glycolysis (23 of 27 glycolytic genes are paralogs, corresponding to 85%) but also occurs for metabolic genes in general. Of all metabolic genes, 44% are paralogs (98), while this percentage is only 5% for the entire yeast genome (5, 17). Currently, a large research effort is under way to enable design and assembly of entirely synthetic yeast genomes (36). For the design of compact, functional synthetic genomes, it will be highly interesting to investigate whether our results on yeast glycolysis can be extrapolated to minor paralogs of genes encoding key enzymes in other central metabolic pathways. At present, predictions of the outcome of such experiments can only be highly speculative. The clear impact of “minor paralogs” of transaldolase and transketolase on pentose fermentation kinetics by engineered S. cerevisiae (103) is a clear example that, at least under some conditions, minor paralogs can make a clear contribution to metabolic flux.
The deletion of all minor glycolytic paralogs had a surprisingly small effect on the yeast transcriptome in aerobic, glucose-grown batch cultures. This result is in marked contrast with several studies in which deletion of individual major glycolytic paralogs led to transcriptional upregulation of its minor paralogs. An example of such a “compensatory” upregulation of minor paralogs is the upregulation of PDC5 upon deletion of PDC1 (104). Such an asymmetric cross-regulation is consistent with the backup theory for fixation of duplicated genes (17), which postulates that (minor) paralogs provide a buffer against deleterious mutations. Although upregulation of minor paralogs and even repair of a null mutation in a major glycolytic paralog by recombination with a minor paralog have been demonstrated (pdc1/PDC5) (104), the general evolutionary significance of the backup theory is still a matter of debate (98, 105). The selective advantages responsible for the overrepresentation of paralogous genes among the structural genes of yeast glycolysis therefore remain an intriguing conundrum.
Minimal glycolysis yeast: a versatile new platform for quantitative research.
As the MG strain was constructed before the advent of CRISPR-Cas9 technology (106), its construction involved 16 consecutive rounds of transformation and 11 marker recycling steps. A previous study (107) in which all 20 hexose transporter (HXT) genes in S. cerevisiae were deleted by repeated rounds of transformation and marker recycling resulted in massive genomic rearrangements, which were recently explained from repeated use of the LoxP system for marker recovery (120). Repeated use of the LoxP system enables recombination across LoxP sites that are left in the genome after marker removal (108). The “HXT-null” strain is mainly used as an excellent platform for functional analysis of individual transporter strains (107, 109–117), in which role its chromosomal rearrangements are hardly relevant. While the MG strain similarly offers an interesting platform for functional analysis of (heterologous) glycolytic genes by one-step gene replacement, we designed and constructed it with the specific aim of building a robust platform for quantitative studies in systems biology. By using URA3 and amdSYM (53) as counterselectable, recyclable marker genes, major chromosomal rearrangements could be avoided. Furthermore, interference of auxotrophies in the interpretation of quantitative growth studies (57) was avoided by making the MG strain prototrophic.
The availability of a well-defined yeast platform with a minimal complement of glycolytic enzymes should provide clear advantages for quantitative modeling of the kinetics and regulation of glycolysis, as it eliminates the intrinsic uncertainties caused by the simultaneous, context-dependent expression of different isoenzymes. In view of the important role of glucose transport in the kinetics of yeast glycolysis, we are currently constructing MG variants with a single hexose transporter. In addition to providing a relevant test bed for mathematical modeling of glycolysis, the MG strain provides an interesting, simplified starting point for laboratory studies on yeast glycolysis. Previous studies have indicated that long-term laboratory evolution can have a major impact on glycolytic genes and their expression (59, 118). Comparison of evolution of the MG strain with that of strains that carry a full complement of glycolytic genes provides an interesting starting point for studies on the impact of gene duplication on evolutionary flexibility of a ubiquitous central metabolic pathway.
Supplementary Material
Supplemental material
Supplemental material for this article may be found at http://dx.doi.org/10.1128/EC.00064-15.
ACKNOWLEDGMENTS
This work was supported by the Technology Foundation STW (Vidi grant 10776).
We thank Walter van Gulik for expert support for metabolome analysis and Lizanne Bosman for experimental contribution to the construction of the MG strain.
REFERENCESInnanH, KondrashovF2010The evolution of gene duplications: classifying and distinguishing between models.
11:97–108. doi:10.1038/nrg2689.20051986OhnoS1970
. Springer-Verlag, New York, NY.ZhangJZ2003Evolution by gene duplication: an update.
18:292–298. doi:10.1016/S0169-5347(03)00033-8.WolfeKH, ShieldsDC1997Molecular evidence for an ancient duplication of the entire yeast genome.
387:708–713. doi:10.1038/42711.9192896KellisM, BirrenBW, LanderES2004Proof and evolutionary analysis of ancient genome duplication in the yeast Saccharomyces cerevisiae.
428:617–624. doi:10.1038/nature02424.15004568DeanEJ, DavisJC, DavisRW, PetrovDA2008Pervasive and persistent redundancy among duplicated genes in yeast.
4:e1000113. doi:10.1371/journal.pgen.1000113.18604285DeLunaA, VetsigianK, ShoreshN, HegrenessM, Colon-GonzalezM, ChaoS, KishonyR2008Exposing the fitness contribution of duplicated genes.
40:676–681. doi:10.1038/ng.123.18408719GuX, ZhangZQ, HuangW2005Rapid evolution of expression and regulatory divergences after yeast gene duplication.
102:707–712. doi:10.1073/pnas.0409186102.15647348MussoG, CostanzoM, HuangfuMQ, SmithAM, PawJ, LuisBJS, BooneC, GiaeverG, NislowC, EmiliA, ZhangZL2008The extensive and condition-dependent nature of epistasis among whole-genome duplicates in yeast.
18:1092–1099. doi:10.1101/gr.076174.108.18463300ConantGC, WolfeKH2008Turning a hobby into a job: how duplicated genes find new functions.
9:938–950. doi:10.1038/nrg2482.19015656RomanoAH, and ConwayT1996Evolution of carbohydrate metabolic pathways.
147:448–455. doi:10.1016/0923-2508(96)83998-2.9084754Fothergill-GilmoreLA, MichelsPAM1993Evolution of glycolysis.
59:105–235. doi:10.1016/0079-6107(93)90001-Z.8426905DandekarT, SchusterS, SnelB, HuynenM, BorkP1999Pathway alignment: application to the comparative analysis of glycolytic enzymes.
343:115–124. doi:10.1042/0264-6021:3430115.10493919SteinkeD, HoeggS, BrinkmannH, MeyerA2006Three rounds (1R/2R/3R) of genome duplications and the evolution of the glycolytic pathway in vertebrates.
4:16. doi:10.1186/1741-7007-4-16.16756667DashkoS, ZhouN, CompagnoC, PiskurJ2014Why, when, and how did yeast evolve alcoholic fermentation?14:826–832. doi:10.1111/1567-1364.12161.24824836van DijkenJP, WeusthuisRA, PronkJT1993Kinetics of growth and sugar consumption in yeasts.
63:343–352. doi:10.1007/BF00871229.8279829GuZL, SteinmetzLM, GuX, ScharfeC, DavisRW, LiWH2003Role of duplicate genes in genetic robustness against null mutations.
421:63–66. doi:10.1038/nature01198.12511954GordonJL, ByrneKP, WolfeKH2009Additions, losses, and rearrangements on the evolutionary route from a reconstructed ancestor to the modern Saccharomyces cerevisiae genome.
5:e1000485. doi:10.1371/journal.pgen.1000485.19436716MericoA, SuloP, PiskurJ, CompagnoC2007Fermentative lifestyle in yeasts belonging to the Saccharomyces complex.
274:976–989. doi:10.1111/j.1742-4658.2007.05645.x.17239085ConantGC, WolfeKH2007Increased glycolytic flux as an outcome of whole-genome duplication in yeast.
3:129. doi:10.1038/msb4100170.17667951BolesE, SchulteF, MiosgaT, FreidelK, SchluterE, ZimmermannFK, HollenbergCP, HeinischJJ1997Characterization of a glucose-repressed pyruvate kinase (Pyk2p) in Saccharomyces cerevisiae that is catalytically insensitive to fructose-1,6-bisphosphate.
179:2987–2993.9139918HeinischJJ, MullerS, SchluterE, JacobyJ, RodicioR1998Investigation of two yeast genes encoding putative isoenzymes of phosphoglycerate mutase.
14:203–213.9544241HohmannS1991Characterization of PDC6, a third structural gene for pyruvate decarboxylase in Saccharomyces cerevisiae.
173:7963–7969.1744053McAlisterL, HollandMJ1985Differential expression of the three yeast glyceraldehyde-3-phosphate dehydrogenase genes.
260:15019–15027.3905788WalshRB, CliftonD, HorakJ, FraenkelDG1991Saccharomyces cerevisiae null mutants in glucose phosphorylation: metabolism and invertase expression.
128:521–527.1874414ConantGC, WagnerA2004Duplicate genes and robustness to transient gene knock-downs in Caenorhabditis elegans.
271:89–96. doi:10.1098/rspb.2003.2560.15002776De SmetR, Van de PeerY2012Redundancy and rewiring of genetic networks following genome-wide duplication events.
15:168–176. doi:10.1016/j.pbi.2012.01.003.22305522HsiaoTL, VitkupD2008Role of duplicate genes in robustness against deleterious human mutations.
4:e1000014. doi:10.1371/journal.pgen.1000014.18369440KafriR, SpringerM, PilpelY2009Genetic redundancy: new tricks for old genes.
136:389–392. doi:10.1016/j.cell.2009.01.027.19203571KrakauerDC, PlotkinJB2002Redundancy, antiredundancy, and the robustness of genomes.
99:1405–1409. doi:10.1073/pnas.032668599.11818563NowakMA, BoerlijstMC, CookeJ, SmithJM1997Evolution of genetic redundancy.
388:167–171. doi:10.1038/40618.9217155ThatcherJW, ShawJM, DickinsonWJ1998Marginal fitness contributions of nonessential genes in yeast.
95:253–257. doi:10.1073/pnas.95.1.253.9419362DeLunaA, SpringerM, KirschnerMW, KishonyR2010Need-based up-regulation of protein levels in response to deletion of their duplicate genes.
8:e1000347. doi:10.1371/journal.pbio.1000347.20361019FellDA, ThomasS1995Physiological control of metabolic flux: the requirement for multisite modulation.
311:35–39.7575476KacserH, BurnsJA, FellDA1995The control of flux.
23:341–366.7672373DymondJS, RichardsonSM, CoombesCE, BabatzT, MullerH, AnnaluruN, BlakeWJ, SchwerzmannJW, DaiJB, LindstromDL, BoekeAC, GottschlingDE, ChandrasegaranS, BaderJS, BoekeJD2011Synthetic chromosome arms function in yeast and generate phenotypic diversity by design.
477:471–476. doi:10.1038/nature10403.21918511TeusinkB, PassargeJ, ReijengaCA, EsgalhadoE, van der WeijdenCC, SchepperM, WalshMC, BakkerBM, van DamK, WesterhoffHV, SnoepJL2000Can yeast glycolysis be understood in terms of in vitro kinetics of the constituent enzymes? Testing biochemistry.
267:5313–5329. doi:10.1046/j.1432-1327.2000.01527.x.10951190SmallboneK, MessihaHL, CarrollKM, WinderCL, MalysN, DunnWB, MurabitoE, SwainstonN, DadaJO, KhanF, PirP, SimeonidisE, SpasicI, WishartJ, WeichartD, HayesNW, JamesonD, BroomheadDS, OliverSG, GaskellSJ, McCarthyJEG, PatonNW, WesterhoffHV, KellDB, MendesP2013A model of yeast glycolysis based on a consistent kinetic characterisation of all its enzymes.
587:2832–2841. doi:10.1016/j.febslet.2013.06.043.23831062BertaniG1951Studies on lysogenesis. 1.
62:293–300.14888646BertaniG2004Lysogeny at mid-twentieth century: P1, P2, and other experimental systems.
186:595–600. doi:10.1128/JB.186.3.595-600.2004.14729683EntianKD, KötterP200725 yeast genetic strain and plasmid collections, p 629–666. InStansfieldI, StarkJ (ed), . Academic Press, Amsterdam, The Netherlands.NijkampJF, van den BroekM, DatemaE, de KokS, BosmanL, LuttikMA, Daran-LapujadeP, VongsangnakW, NielsenJ, HeijneWHM, KlaassenP, PaddonCJ, PlattD, KotterP, van HamRC, ReindersMJT, PronkJT, de RidderD, DaranJM2012De novo sequencing, assembly and analysis of the genome of the laboratory strain Saccharomyces cerevisiae CEN.PK113-7D, a model for modern industrial biotechnology.
11:36. doi:10.1186/1475-2859-11-36.22448915van DijkenJP, BauerJ, BrambillaL, DubocP, FrancoisJM, GancedoC, GiuseppinMLF, HeijnenJJ, HoareM, LangeHC, MaddenEA, NiederbergerP, NielsenJ, ParrouJL, PetitT, PorroD, ReussM, van RielN, RizziM, SteensmaHY, VerripsCT, VindelovJ, PronkJT2000An interlaboratory comparison of physiological and genetic properties of four Saccharomyces cerevisiae strains.
26:706–714. doi:10.1016/S0141-0229(00)00162-9.10862876WachA, BrachatA, Alberti-SeguiC, RebischungC, PhilippsenP1997Heterologous HIS3 marker and GFP reporter modules for PCR-targeting in Saccharomyces cerevisiae.
13:1065–1075.9290211GueldenerU, HeinischJ, KoehlerGJ, VossD, HegemannJH2002A second set of LoxP marker cassettes for Cre-mediated multiple gene knockouts in budding yeast.
30:e23. doi:10.1093/nar/30.6.e23.11884642ShusterJR, MoyerD, IrvineB1987Sequence of the Kluyveromyces lactisURA3 gene.
15:8573–8573. doi:10.1093/nar/15.20.8573.3671096WachA, BrachatA, PohlmannR, PhilippsenP1994New heterologous modules for classical or PCR-based gene disruptions in Saccharomyces cerevisiae.
10:1793–1808. doi:10.1002/yea.320101310.7747518GuldenerU, HeckS, FiedlerT, BeinhauerJ, HegemannJH1996A new efficient gene disruption cassette for repeated use in budding yeast.
24:2519–2524. doi:10.1093/nar/24.13.2519.8692690de KokS, YilmazD, SuirE, PronkJT, DaranJM, van MarisAJA2011Increasing free-energy (ATP) conservation in maltose-grown Saccharomyces cerevisiae by expression of a heterologous maltose phosphorylase.
13:518–526. doi:10.1016/j.ymben.2011.06.001.21684346GritzL, and DaviesJ1983Plasmid encoded hygromycin-B resistance: the sequence of hygromycin-B phosphotransferase gene and its expression in Escherichia coli and Saccharomyces cerevisiae.
25:179–188. doi:10.1016/0378-1119(83)90223-8.6319235AlaniE, CaoL, KlecknerN1987A method for gene disruption that allows repeated use of URA3 selection in the construction of multiply disrupted yeast strains.
116:541–545. doi:10.1534/genetics.112.541.test.3305158AkadaR, KitagawaT, KanekoS, ToyonagaD, ItoS, KakiharaY, HoshidaH, MorimuraS, KondoA, KidaK2006PCR-mediated seamless gene deletion and marker recycling in Saccharomyces cerevisiae.
23:399–405. doi:10.1002/yea.1365.16598691Solis-EscalanteD, KuijpersNG, BongaertsN, BolatI, BosmanL, PronkJT, DaranJM, Daran-LapujadeP2013amdSYM, a new dominant recyclable marker cassette for Saccharomyces cerevisiae.
13:126–139. doi:10.1111/1567-1364.12024.23253382Solis-EscalanteD, KuijpersNG, van der LindenFH, PronkJT, DaranJM, Daran-LapujadeP2014Efficient simultaneous excision of multiple selectable marker cassettes using I-SceI-induced double-strand DNA breaks in Saccharomyces cerevisiae.
14:741–754. doi:10.1111/1567-1364.12162.24833416GietzRD, WoodsRA2002Transformation of yeast by lithium acetate/single-stranded carrier DNA/polyethylene glycol method.
350:87–96.12073338VerduynC, PostmaE, ScheffersWA, VandijkenJP1992Effect of benzoic acid on metabolic fluxes in yeasts: a continuous culture study on the regulation of respiration and alcoholic fermentation.
8:501–517. doi:10.1002/yea.320080703.1523884PronkJT2002Auxotrophic yeast strains in fundamental and applied research.
68:2095–2100. doi:10.1128/AEM.68.5.2095-2100.2002.11976076PostmaE, VerduynC, ScheffersWA, VandijkenJP1989Enzymic analysis of the Crabtree effect in glucose-limited chemostat cultures of Saccharomyces cerevisiae.
55:468–477.2566299JansenMLA, DiderichJA, MashegoM, HassaneA, de WindeJH, Daran-LapujadeP, PronkJT2005Prolonged selection in aerobic, glucose-limited chemostat cultures of Saccharomyces cerevisiae causes a partial loss of glycolytic capacity.
151:1657–1669. doi:10.1099/mic.0.27577-0.15870473CruzALB, HeblyM, DuongGH, WahlSA, PronkJT, HeijnenJJ, Daran-LapujadeP, van GulikWM2012Similar temperature dependencies of glycolytic enzymes: an evolutionary adaptation to temperature dynamics?6:151. doi:10.1186/1752-0509-6-151.23216813LowryOH, RosebroughNJ, FarrAL, RandallRJ1951Protein measurement with the Folin phenol reagent.
193:265–275.14907713LangeHC, HeijnenJJ2001Statistical reconciliation of the elemental and molecular biomass composition of Saccharomyces cerevisiae.
75:334–344. doi:10.1002/bit.10054.11590606CanelasAB, RasC, ten PierickA, van DamJC, HeijnenJJ, van GulikWM2008Leakage-free rapid quenching technique for yeast metabolomics.
4:226–239. doi:10.1007/s11306-008-0116-4.WuL, MashegoMR, van DamJC, ProellAM, VinkeJL, RasC, van WindenWA, van GulikWM, HeijnenJJ2005Quantitative analysis of the microbial metabolome by isotope dilution mass spectrometry using uniformly 13C-labeled cell extracts as internal standards.
336:164–171. doi:10.1016/j.ab.2004.09.001.15620880CipollinaC, ten PierickA, CanelasAB, SeifarRM, van MarisAJA, van DamJC, HeijnenJJ2009A comprehensive method for the quantification of the non-oxidative pentose phosphate pathway intermediates in Saccharomyces cerevisiae by GC-IDMS.
877:3231–3236. doi:10.1016/j.jchromb.2009.07.019.van DamJC, EmanMR, FrankJ, LangeHC, van DedemGWK, HeijnenSJ2002Analysis of glycolytic intermediates in Saccharomyces cerevisiae using anion exchange chromatography and electrospray ionization with tandem mass spectrometric detection.
460:209–218. doi:10.1016/S0003-2670(02)00240-4.KozakBU, van RossumHM, BenjaminKR, WuL, DaranJMG, PronkJT, van MarisAIJA2014Replacement of the Saccharomyces cerevisiae acetyl-CoA synthetases by alternative pathways for cytosolic acetyl-CoA synthesis.
21:46–59. doi:10.1016/j.ymben.2013.11.005.24269999PiperMDW, Daran-LapujadeP, BroC, RegenbergB, KnudsenS, NielsenJ, PronkJT2002Reproducibility of oligonucleotide microarray transcriptome analyses: an interlaboratory comparison using chemostat cultures of Saccharomyces cerevisiae.
277:37001–37008. doi:10.1074/jbc.M204490200.12121991LiH, DurbinR2009Fast and accurate short read alignment with Burrows-Wheeler transform.
25:1754–1760. doi:10.1093/bioinformatics/btp324.19451168LiH, DurbinR2010Fast and accurate long-read alignment with Burrows-Wheeler transform.
26:589–595. doi:10.1093/bioinformatics/btp698.20080505TrapnellC, WilliamsBA, PerteaG, MortazaviA, KwanG, van BarenMJ, SalzbergSL, WoldBJ, PachterL2010Transcript assembly and quantification by RNA-Seq reveals unannotated transcripts and isoform switching during cell differentiation.
28:511–515. doi:10.1038/nbt.1621.20436464EwingB, HillierL, WendlMC, GreenP1998Base-calling of automated sequencer traces using phred. I. Accuracy assessment.
8:175–185.9521921EwingB, GreenP1998Base-calling of automated sequencer traces using phred. II. Error probabilities.
8:186–194.9521922GancedoC, FloresCL2008Moonlighting proteins in yeasts.
72:197–210. doi:10.1128/MMBR.00036-07.18322039KnijnenburgTA, DaranJMG, van den BroekMA, Daran-LapujadePAS, de WindeJH, PronkJT, ReindersMJT, WesselsLFA2009Combinatorial effects of environmental parameters on transcriptional regulation in Saccharomyces cerevisiae: a quantitative analysis of a compendium of chemostat-based transcriptome data.
10:53. doi:10.1186/1471-2164-10-53.19173729DeckerBL, WicknerWT2006Enolase activates homotypic vacuole fusion and protein transport to the vacuole in yeast.
281:14523–14528. doi:10.1074/jbc.M600911200.16565073PelaezR, HerreroP, MorenoF2010Functional domains of yeast hexokinase 2.
432:181–190. doi:10.1042/BJ20100663.20815814KopperschlagerG, BarJ, NisslerK, HofmannE1977Physicochemical parameters and subunit composition of yeast phosphofructokinase.
81:317–325. doi:10.1111/j.1432-1033.1977.tb11954.x.145942Breitenbach-SchmittI, HeinischJ, SchmittHD, ZimmermannFK1984Yeast mutants without phosphofructokinase activity can still perform glycolysis and alcoholic fermentation.
195:530–535. doi:10.1007/BF00341458.CliftonD, FraenkelDG1982Mutant studies of yeast phosphofructokinase.
21:1935–1942. doi:10.1021/bi00537a037.6211191LoboZ, MaitraPK1983Phosphofructokinase mutants of yeast: biochemistry and genetics.
258:1444–1449.6218165BakkerBM, BroC, KotterP, LuttikMAH, van DijkenJP, PronkJT2000The mitochondrial alcohol dehydrogenase adh3p is involved in a redox shuttle in Saccharomyces cerevisiae.
182:4730–4737. doi:10.1128/JB.182.17.4730-4737.2000.10940011CakarZP, Turanli-YildizB, AlkimC, YilmazU2012Evolutionary engineering of Saccharomyces cerevisiae for improved industrially important properties.
12:171–182. doi:10.1111/j.1567-1364.2011.00775.x.22136139TengX, Dayhoff-BranniganM, ChengWC, GilbertCE, SingCN, DinyNL, WheelanSJ, DunhamMJ, BoekeJD, PinedaFJ, HardwickJM2013Genome-wide consequences of deleting any single gene.
52:485–494. doi:10.1016/j.molcel.2013.09.026.24211263TaiSL, Daran-LapujadeP, LuttikMAH, WalshMC, DiderichJA, KrijgerGC, van GulikWM, PronkJT, DaranJM2007Control of the glycolytic flux in Saccharomyces cerevisiae grown at low temperature: a multi-level analysis in anaerobic chemostat cultures.
282:10243–10251. doi:10.1074/jbc.M610845200.17251183de SmidtO, du PreezJC, AlbertynJ2012Molecular and physiological aspects of alcohol dehydrogenases in the ethanol metabolism of Saccharomyces cerevisiae.
12:33–47. doi:10.1111/j.1567-1364.2011.00760.x.22094012GruningNM, RinnerthalerM, BluemleinK, MullederM, WamelinkMMC, LehrachH, JakobsC, BreitenbachM, RalserM2011Pyruvate kinase triggers a metabolic feedback loop that controls redox metabolism in respiring cells.
14:415–427. doi:10.1016/j.cmet.2011.06.017.21907146RingJ, SommerC, Carrnona-GutierrezD, RuckenstuhlC, EisenbergT, MadeoF2012The metabolism beyond programmed cell death in yeast.
318:1193–1200. doi:10.1016/j.yexcr.2012.03.019.22480867Daran-LapujadeP, DaranJM, LuttikMAH, AlmeringMJH, PronkJT, KötterP2009An atypical PMR2 locus is responsible for hypersensitivity to sodium and lithium cations in the laboratory strain Saccharomyces cerevisiae CEN.PK113-7D.
9:789–792. doi:10.1111/j.1567-1364.2009.00530.x.19519766KimJW, DangCV2005Multifaceted roles of glycolytic enzymes.
30:142–150. doi:10.1016/j.tibs.2005.01.005.15752986KanePA2006The where, when, and how of organelle acidification by the yeast vacuolar H+-ATPase.
70:177–191. doi:10.1128/MMBR.70.1.177-191.2006.16524922GrantCM, QuinnKA, DawesIW1999Differential protein S-thiolation of glyceraldehyde-3-phosphate dehydrogenase isoenzymes influences sensitivity to oxidative stress.
19:2650–2656.10082531RodriguezA, de la CeraT, HerreroP, MorenoF2001The hexokinase 2 protein regulates the expression of the GLK1, HXK1 and HXK2 genes of Saccharomyces cerevisiae.
355:625–631.11311123BoerVM, de WindeJH, PronkJT, PiperMDW2003The genome-wide transcriptional responses of Saccharomyces cerevisiae grown on glucose in aerobic chemostat cultures limited for carbon, nitrogen, phosphorus, or sulfur.
278:3265–3274. doi:10.1074/jbc.M209759200.12414795CormierL, BarbeyR, KurasL2010Transcriptional plasticity through differential assembly of a multiprotein activation complex.
38:4998–5014. doi:10.1093/nar/gkq257.20392822FauchonM, LagnielG, AudeJC, LombardiaL, SoularueP, PetatC, MarguerieG, SentenacA, WernerM, LabarreJ2002Sulfur sparing in the yeast proteome in response to sulfur demand.
9:713–723. doi:10.1016/S1097-2765(02)00500-2.11983164QianWF, LiaoBY, ChangAYF, ZhangJZ2010Maintenance of duplicate genes and their functional redundancy by reduced expression.
26:425–430. doi:10.1016/j.tig.2010.07.002.20708291KuepferL, SauerU, BlankLM2005Metabolic functions of duplicate genes in Saccharomyces cerevisiae.
15:1421–1430. doi:10.1101/gr.3992505.16204195VerduynC, PostmaE, ScheffersWA, VandijkenJP1990Physiology of Saccharomyces cerevisiae in anaerobic glucose-limited chemostat cultures.
136:395–403. doi:10.1099/00221287-136-3-395.1975265HongK, NielsenJ2012Metabolic engineering of Saccharomyces cerevisiae: a key cell factory platform for future biorefineries.
69:2671–2690. doi:10.1007/s00018-012-0945-1.22388689NevoigtE2008Progress in metabolic engineering of Saccharomyces cerevisiae.
72:379–412. doi:10.1128/MMBR.00025-07.18772282GoddardM, GreigD2015Saccharomyces cerevisiae: a nomadic yeast with no niche?15:fov009. doi:10.1093/femsyr/fov009.25725024WisselinkHW, CipollinaC, OudB, CrimiB, HeijnenJJ, PronkJT, van MarisAJA2010Metabolome, transcriptome and metabolic flux analysis of arabinose fermentation by engineered Saccharomyces cerevisiae.
12:537–551. doi:10.1016/j.ymben.2010.08.003.20816840HohmannS, CederbergH1990Autoregulation may control the expression of yeast pyruvate decarboxylase structural genes PDC1 and PDC5.
188:615–621. doi:10.1111/j.1432-1033.1990.tb15442.x.2185016ZhangJZ2012
, p 279–300. Springer, New York, NY.DiCarloJE, NorvilleJE, MaliP, RiosX, AachJ, ChurchGM2013Genome engineering in Saccharomyces cerevisiae using CRISPR-Cas systems.
41:4336–4343. doi:10.1093/nar/gkt135.23460208WieczorkeR, KrampeS, WeierstallT, FreidelK, HollenbergCP, BolesE1999Concurrent knock-out of at least 20 transporter genes is required to block uptake of hexoses in Saccharomyces cerevisiae.
464:123–128. doi:10.1016/S0014-5793(99)01698-1.10618490DelneriD, TomlinGC, WixonJL, HutterA, SeftonM, LouisEJ, OliverSG2000Exploring redundancy in the yeast genome: an improved strategy for use of the Cre-LoxP system.
252:127–135. doi:10.1016/S0378-1119(00)00217-1.10903444CoelhoMA, GoncalvesC, SampaioJP, GoncalvesP2013Extensive intra-kingdom horizontal gene transfer converging on a fungal fructose transporter gene.
9:e1003587. doi:10.1371/journal.pgen.1003587.23818872Dae-HeeL, Soo-JungK, Jin-HoS2014Molecular cloning and characterization of two novel fructose-specific transporters from the osmotolerant and fructophilic yeast Candida magnoliae JH110.
98:3569–3578. doi:10.1007/s00253-013-5225-y.24048639PriceDRG, TibblesK, ShigenobuS, SmertenkoA, RussellCW, DouglasAE, FitchesE, GatehouseAMR, GatehouseJA2010Sugar transporters of the major facilitator superfamily in aphids; from gene prediction to functional characterization.
19:97–112. doi:10.1111/j.1365-2583.2009.00918.x.20482643SchneidereitA, Scholz-StarkeJ, ButtnerM2003Functional characterization and expression analyses of the glucose-specific AtSTP9 monosaccharide transporter in pollen of Arabidopsis.
133:182–190. doi:10.1104/pp.103.026674.12970485SchusslerA, MartinH, CohenD, FitzM, WipfD2006Characterization of a carbohydrate transporter from symbiotic glomeromycotan fungi.
444:933–936. doi:10.1038/nature05364.17167486StasykOG, MaidanMM, StasykOV, Van DijckP, TheveleinJM, SibirnyAA2008Identification of hexose transporter-like sensor HXS1 and functional hexose transporter HXT1 in the methylotrophic yeast Hansenula polymorpha.
7:735–746. doi:10.1128/EC.00028-08.18310355VignaultC, VachaudM, CakirB, GlissantD, DedaldechampF, ButtnerM, AtanassovaR, Fleurat-LessardP, LemoineR, DelrotS2005VvHT1 encodes a monosaccharide transporter expressed in the conducting complex of the grape berry phloem.
56:1409–1418. doi:10.1093/jxb/eri142.15809282XuanYH, HuYB, ChenLQ, SossoD, DucatDC, HouBH, FrommerWB2013Functional role of oligomerization for bacterial and plant SWEET sugar transporter family.
110:E3685–E3694. doi:10.1073/pnas.1311244110.24027245YoungE, PoucherA, ComerA, BaileyA, AlperH2011Functional survey for heterologous sugar transport proteins, using Saccharomyces cerevisiae as a host.
77:3311–3319. doi:10.1128/AEM.02651-10.21421781MashegoMR, JansenMLA, VinkeJL, van GulikWM, HeijnenJJ2005Changes in the metabolome of Saccharomyces cerevisiae associated with evolution in aerobic glucose-limited chemostats.
5:419–430. doi:10.1016/j.femsyr.2004.11.008.15691747ByrneKP, WolfeKH2005The Yeast Gene Order Browser: combining curated homology and syntenic context reveals gene fate in polyploid species.
15:1456–1461. doi:10.1101/gr.3672305.16169922Solis-EscalanteD, van den BroekM, KuijpersNG, PronkJT, BolesE, DaranJM, Daran-LapujadeP2015The genome sequence of the popular hexose-transport-deficient Saccharomyces cerevisiae strain EBY.VW4000 reveals LoxP/Cre-induced translocations and gene loss.
15:fou004. doi:10.1093/femsyr/fou004.oai:pubmedcentral.nih.gov:46213102015-11-05eukcellpmc-openEukaryot CellEukaryotic CelleukcelleukcellEUKCELLEukaryotic Cell1535-97781535-9786American Society for Microbiology1752 N St., N.W., Washington, DCPMC4621310PMC46213104621310263858922638589200130-1510.1128/EC.00130-15ArticlesEvidence that the Entamoeba histolytica Mitochondrial Carrier Family Links Mitosomal and Cytosolic Pathways through Exchange of 3′-Phosphoadenosine 5′-Phosphosulfate and ATPImportance of MCF in Sulfate Metabolism in EntamoebaMi-ichi et al.Mi-ichiFumikaaNozawaAkirabYoshidaHirokiaTozawaYuzurucNozakiTomoyoshideDivision of Molecular and Cellular Immunoscience, Department of Biomolecular Sciences, Faculty of Medicine, Saga University, Saga, JapanProteo-Science Center, Ehime University, Matsuyama, Ehime, JapanGraduate School of Science and Engineering, Saitama University, Saitama, JapanDepartment of Parasitology, National Institute of Infectious Diseases, Shinjuku-ku, Tokyo, JapanGraduate School of Life and Environmental Sciences, University of Tsukuba, Tsukuba, Ibaraki, JapanAddress correspondence to Yuzuru Tozawa, tozawa@mail.saitama-u.ac.jp, or Tomoyoshi Nozaki, nozaki@nih.go.jp.
F.M. and A.N. contributed equally to this article.
Citation Mi-ichi F, Nozawa A, Yoshida H, Tozawa Y, Nozaki T. 2015. Evidence that the Entamoeba histolytica mitochondrial carrier family links mitosomal and cytosolic pathways through exchange of 3′-phosphoadenosine 5′-phosphosulfate and ATP. Eukaryot Cell 14:1144–1150. doi:10.1128/EC.00130-15.
Entamoeba histolytica, a microaerophilic protozoan parasite, possesses mitosomes. Mitosomes are mitochondrion-related organelles that have largely lost typical mitochondrial functions, such as those involved in the tricarboxylic acid cycle and oxidative phosphorylation. The biological roles of Entamoeba mitosomes have been a long-standing enigma. We previously demonstrated that sulfate activation, which is not generally compartmentalized to mitochondria, is a major function of E. histolytica mitosomes. Sulfate activation cooperates with cytosolic enzymes, i.e., sulfotransferases (SULTs), for the synthesis of sulfolipids, one of which is cholesteryl sulfate. Notably, cholesteryl sulfate plays an important role in encystation, an essential process in the Entamoeba life cycle. These findings identified a biological role for Entamoeba mitosomes; however, they simultaneously raised a new issue concerning how the reactions of the pathway, separated by the mitosomal membranes, cooperate. Here, we demonstrated that the E. histolytica mitochondrial carrier family (EhMCF) has the capacity to exchange 3′-phosphoadenosine 5′-phosphosulfate (PAPS) with ATP. We also confirmed the cytosolic localization of all the E. histolytica SULTs, suggesting that in Entamoeba, PAPS, which is produced through mitosomal sulfate activation, is translocated to the cytosol and becomes a substrate for SULTs. In contrast, ATP, which is produced through cytosolic pathways, is translocated into the mitosomes and is a necessary substrate for sulfate activation. Taking our findings collectively, we suggest that EhMCF functions as a PAPS/ATP antiporter and plays a crucial role in linking the mitosomal sulfate activation pathway to cytosolic SULTs for the production of sulfolipids.
INTRODUCTION
Entamoeba histolytica is a microaerophilic protozoan parasite causing intestinal and extraintestinal amebiasis in humans. These infectious diseases have spread worldwide and have become a serious public health problem (1). E. histolytica possesses mitosomes, a type of mitochondrion-related organelle (MRO) (2–5). This organelle was originally called crypton when its discovery was reported, and such organelles are currently known as mitosomes, the widely accepted name (6, 7). MROs are derived from canonical mitochondria and are ubiquitously found in anaerobic/microaerophilic eukaryotes (2, 4). MROs display a variety of unique characteristics which are conferred by essentially reduced and/or modified mitochondrial functions and that occasionally result from the addition of new functions acquired through lateral gene transfer (5, 8). It has been postulated that the uniqueness of MROs helps organisms to adapt to various niches for their survival (2–5, 8). Entamoeba mitosomes have largely lost typical mitochondrial functions, such as those involved in the tricarboxylic acid cycle, electron transport, oxidative phosphorylation, and β-oxidation of fatty acids (4, 5). This raises two fundamental questions. Why does Entamoeba maintain mitosomes? What are their biological roles in this organism?
Despite being recognized, these important issues have not been satisfactorily addressed. We previously showed that sulfate activation, which is not generally compartmentalized to mitochondria, is a major function in E. histolytica mitosomes (3, 5, 9). Furthermore, we demonstrated that 3′-phosphoadenosine 5′-phosphosulfate (PAPS), which is synthesized through mitosomal metabolism, acts as an activated sulfur donor mainly to produce sulfolipids by the catalytic actions of putative cytosolic sulfotransferases (SULTs) (5). Cholesteryl sulfate (CS) is one such sulfolipid that plays an important role in encystation, a cell differentiation process necessary for maintaining the Entamoeba life cycle (9). These findings provide not only an explanation for the biological role of Entamoeba mitosomes but also evidence to support the uniqueness of MROs.
Importantly, our findings indicate that in Entamoeba, PAPS, a metabolite synthesized through the mitosomal pathway, needs to be translocated to the cytosol for the production of sulfolipids. Therefore, a molecule that can translocate PAPS from mitosomes to the cytosol must be present. Three candidates which could be responsible for this important function are encoded in the E. histolytica genome. One is a PAPS transporter, and the others are mitochondrial carrier (MC) proteins (EHI_068590, EHI_095150, and EHI_153760, respectively) (AmoebaDB; http://amoebadb.org/amoeba/). However, the PAPS transporter can be ruled out because its nonmitosomal localization has already been demonstrated (5). Hence, we focused on the MC proteins as the most likely candidates.
MC proteins belong to a large family of mitochondrial inner membrane carriers that transport various metabolites between the cytosol and mitochondrial matrix (10, 11). Most MC proteins are localized in mitochondria, but atypical localizations in chloroplasts and peroxisomes have recently been reported (11). The structural features conserved in MC proteins include a size of 30 to 35 kDa, three tandemly repeated homologous domains, each of which has ∼100 amino acid residues, and six transmembrane helices forming an aqueous cavity. Substances transported by MC proteins include nucleotides, amino acids, keto acids, and inorganic phosphate (Pi) (10, 11).
In this study, to address the issue of how the mitosomal sulfate activation pathway and putative cytosolic SULTs cooperate in E. histolytica, we performed biochemical and cell biological characterizations of an E. histolytica MC protein, E. histolytica mitochondrial carrier family (EhMCF), and related sulfate metabolism enzymes, the E. histolytica SULTs (EhSULTs) and E. histolytica 3′(2′),5′-bisphosphate nucleotidases (EhPAPases).
MATERIALS AND METHODSMaterials.
[14C]leucine (stock radioactivity, 100 μCi/ml), [32P]ATP (stock radioactivity, 10 mCi/ml), and [35S]PAPS (stock radioactivity, 1 mCi/ml) were purchased from PerkinElmer Japan (Yokohama, Japan). Asolectin was from Fluka (Buchs, Switzerland).
Culture of E. histolytica.
The E. histolytica HM-1:IMSS cl6 strain was routinely cultured in vitro in Diamond's BI-S-33 medium at 37°C as described previously (3, 5). Transformants were also obtained in Diamond's BI-S-33 medium as described below.
Indirect immunofluorescence analyses of E. histolytica transformants expressing HA-tagged EhSULTs or EhPAPases.
Expression plasmids for hemagglutinin (HA)-tagged EhSULTs and EhPAPases were constructed by PCR amplification of the corresponding open reading frames (ORFs) using appropriate primers sets (Table 1). Amplicons, except for the one containing EhSULT7, were then digested with BglII and inserted into the corresponding sites of the pEhEx-HA plasmid (5). The PCR-amplified EhSULT7 fragment was directly cloned into BglII-digested plasmid pEhEx-HA using an In-Fusion HD cloning kit from TaKaRa Bio (Otsu, Japan) according to the manufacturer's instructions. Lipofection transfection of E. histolytica trophozoites with the constructed plasmids, drug selection, maintenance of selected transformants, and indirect immunofluorescence analysis of independent established transformants were performed as described previously (5).
The recombinant EhMCF (rEhMCF) protein that is encoded in the E. histolytica genome (EHI_095150 in AmoebaDB; http://amoebadb.org/amoeba/) was produced using a wheat germ cell-free translation system in the presence of asolectin liposomes, which were freshly prepared just prior to use, as described previously (12). Two plasmids, pYT08-EhMCF and pYT08-codon-optimized EhMCF, were constructed as the templates for in vitro mRNA synthesis. For pYT08-EhMCF, a DNA fragment encoding the EhMCF ORF was amplified from E. histolytica cDNA by PCR with an appropriate primer set (Table 1), digested with SpeI and SalI, and inserted into the corresponding sites of the pYT08 vector (12). For the pYT08-codon-optimized EhMCF, a synthetic DNA encoding an EhMCF ORF in which the codon usage is optimized to that of the wheat germ translation system (GenBank accession number LC036596) was custom synthesized by GenScript Japan (Tokyo, Japan) and cloned into SpeI and SalI sites of the pYT08 vector (12). The mRNAs synthesized from the constructed plasmids were then used in the cell-free translation system in the presence of asolectin liposomes and, when needed, with [14C]leucine (final radioactivity, 2 μCi/ml) as described previously (12). Subsequently, the reaction mixtures were centrifuged at 20,000 × g for 20 min at 4°C, and the supernatant and the precipitate were separately collected. Finally, the precipitate was suspended in 150 μl 10 mM PIPES-NaOH (pH 6.5) and subjected to ultrasonication (Digital Sonifier model 250 D; Branson, Danbury, CT, USA) for 18 s (50% duty cycle). The suspension obtained was used for vesicle preparations for the transporter activity assay as described in the following section. As a blank suspension, the reaction mixture operated in the absence of mRNA was treated in the same way as the rEhMCF protein suspension. To verify the production of rEhMCF and its purity, 1/100 volumes from each fraction were analyzed by SDS-PAGE, followed by visualization through either autoradiography or Coomassie brilliant blue staining.
Transport activity assay of rEhMCF in vesicles.
Two types of the vesicles, namely, substrate-loaded vesicles and empty vesicles, were prepared and used for the assays. The substrate-loaded vesicles were prepared essentially as described previously (12). In detail, the rEhMCF protein suspension (see the section above) was mixed with an equal volume of solution containing liposomes into which one of the countersubstrates examined was preloaded. To prepare the liposome solution, 200 mM PIPES–NaOH (pH 6.5) containing 40 mM potassium gluconate and 50 or 20 mM countersubstrate was added on acetone-washed asolectin (80 mg lipid/ml), followed by ultrasonication (Digital Sonifier model 250 D; Branson, Danbury, CT, USA) for 5 min on ice. Subsequently, the obtained mixture was frozen in liquid nitrogen, thawed at room temperature, and subjected to ultrasonication (Digital Sonifier model 250 D; Branson, Danbury, CT, USA) for 18 s (50% duty cycle). Finally, the countersubstrate that was not loaded into vesicles was removed by gel filtration with a Dowex AG-1X8 column (Bio-Rad, Tokyo, Japan) using 10 mM PIPES–NaOH (pH 6.5) containing 40 mM potassium gluconate and 100 mM sodium gluconate. The empty vesicles were prepared in a same way as described for the substrate-loaded vesicles except that no countersubstrates were preloaded. Blank controls for each type of vesicle were likewise prepared using the blank (see the section above) suspension in place of the rEhMCF protein suspension.
The transport activity of the rEhMCF in vesicles was measured as described previously (12). Briefly, a quantitative evaluation of the uptake of either [32P]ATP or [35S]PAPS into substrate-loaded or empty vesicles was performed. The final concentration of preloaded countersubstrate in the transport assay of the uptake of [32P]ATP or [35S]PAPS was 25 mM or 10 mM, respectively, and that of added radiolabeled substrate (either [32P]ATP or [35S]PAPS) was 0.5 mM. The reaction was initiated at 25°C upon addition of 5 μl radiolabeled substrate into 100 μl 10 mM PIPES–NaOH (pH 6.5) containing either type of vesicle, [32P]ATP (final radioactivity, 2.7 μCi/ml) or [35S]PAPS (final radioactivity, 1.5 μCi/ml). Reactions were terminated by the addition of 15 μl stop solution (360 mM pyridoxal 5′-phosphate, 64 mM mersalyl acid). After removal of extravesicle radiolabeled substrate by gel filtration performed with a Dowex AG-1X8 column (Bio-Rad, Tokyo, Japan) using 200 mM sodium acetate, the radioactivity associated with each type of vesicle was measured as units of disintegrations per minute using an LSC-6100 liquid-scintillation counter from Aloka (Tokyo, Japan). In parallel, a blank control for each type of vesicle was assayed. Finally, the amount of substrate transported into the vesicles was calculated after subtracting the radioactivity of a blank control from that of the corresponding sample. The amount of protein which was estimated prior to vesicle preparation was used to determine the specific transporter activity. Its estimation was accomplished by measuring the intensity of the corresponding bands in SDS-PAGE gels stained by Coomassie brilliant blue. Bovine serum albumin was used as a standard.
For a time course analysis, a quantitative evaluation of the uptake of [32P]ATP into substrate-loaded vesicles into which nonradiolabeled ATP was preloaded or into empty vesicles was performed at 0, 5, 10, or 20 min after starting the incubation. For a countersubstrate specificity analysis, a quantitative evaluation of the uptake of either [32P]ATP or [35S]PAPS into the substrate-loaded vesicles into which one of the various substances tested was preloaded or into empty vesicles was performed after 10 min of incubation. The obtained activities were statistically analyzed with Student's t test.
Nucleotide sequence accession numbers.
The proteins described in this study (GenBank accession numbers, E. histolytica genome identification numbers) and the accession number of the EhMCF synthetic gene are as follows: MCF (XP_649800, genome identification number EHI_095150), Pi transporter (XP_656350, EHI_153760), PAPS transporter (XP_654175, EHI_068590), SULT1 (XP_654200, EHI_140740), SULT2 (XP_654101, EHI_166030), SULT3 (XP_651675, EHI_197340), SULT4 (XP_655605, EHI_092490), SULT5 (XP_650013, EHI_090430), SULT6 (XP_649714, EHI_146990), SULT7 (XP_651768, EHI_114190), SULT8 (XP_652989, EHI_181190), SULT9 (XP_653539, EHI_031640), SULT10 (XP_650544, EHI_062680), PAPase1 (XP_651950, EHI_193350), PAPase2 (XP_655585, EHI_179820), PAPase3 (XP_650613, EHI_175410), and the synthetic EhMCF cDNA (LC036596).
RESULTS AND DISCUSSIONAll EhSULTs are localized in the cytosol.
The E. histolytica genome contains 10 genes that encode putative SULTs, which can catalyze the production of sulfated molecules, e.g., sulfolipids, from PAPS and sulfur acceptors (AmoebaDB; http://amoebadb.org/amoeba/ [5, 9]). We previously demonstrated the cytosolic localization of EhSULT1 and -2 (5). To fully determine the localization of all EhSULTs, we established independent transformants expressing HA-tagged EhSULT3 to -10 in E. histolytica and analyzed the transformants obtained. The fluorescent signals detected in these transformants were distributed throughout cells, except in organelles such as the nucleus, vacuoles, and small vesicles (Fig. 1). This result, combined with the data corresponding to the cytosolic localization of EhSULT1 and -2 (5), indicates that all EhSULTs are localized in the cytosol (Fig. 1). Importantly, it also confirms the cytosolic localization of EhSULT6, which catalyzes the production of CS, a sulfolipid important for Entamoeba encystation (9). These findings are consistent with the requirement for PAPS, a mitosomal sulfate-activated metabolite, to be transported into the cytosol for it to be a substrate for EhSULTs in E. histolytica.
Localization of EhSULT3 to -10 in E. histolytica. Indirect immunofluorescence images of transformants expressing HA-tagged EhSULT3 to -10 are shown. Bars, 10 μm.
Production, purification, and characterization of rEhMCF protein.
Two E. histolytica MC proteins (EHI_095150 and EHI_153760, respectively) have been characterized as an ADP/ATP carrier (AAC) and a Pi carrier (AmoebaDB; http://amoebadb.org/amoeba/ [6, 9, 13]). In addition, phylogenetic analysis showed that only the E. histolytica AAC (EhAAC) and not the Pi carrier is a member of a subfamily cluster of MC proteins, which includes carriers transporting adenine nucleotides and coenzyme A (CoA) (13). Recently, a carrier that was originally characterized as an AAC of the thylakoid membrane (14) was shown to possess a capacity for countertransporting PAPS; therefore, it is now known as a PAPS transporter (15). These findings narrow down the mitosomal membrane PAPS transporter candidates and are consistent with our hypothesis that EhAAC acts as an antiporter that can transport PAPS across mitosomal membranes in E. histolytica. EhAAC is more commonly termed E. histolytica mitochondrial carrier family (EhMCF) (3, 5, 6, 16); therefore, unless otherwise stated, we use the name EhMCF here.
Obtaining functional rEhMCF protein is necessary to address how the mitosomal sulfate activation pathway cooperates with cytosolic SULTs in E. histolytica. To achieve this, we exploited a wheat germ cell-free translation system in the presence of asolectin liposomes because this system circumvents several problems encountered in expressing recombinant membrane proteins in surrogate organisms or cells (17, 18). In this system, we can trace only the target recombinant protein as a nascent protein using a radiolabeled leucine. The molecular size of the synthesized radiolabeled protein was ∼30 kDa, which is close to the deduced molecular mass of EhMCF (30,443 Da), and it was predominantly recovered from the fraction precipitated at 20,000 × g (Fig. 2A). The yield of rEhMCF could be improved by using a synthetic EhMCF cDNA in which the codon usage is optimized to that of the wheat germ translation system (Fig. 2A; GenScript Japan [Tokyo, Japan]). Most importantly, rEhMCF was highly enriched and became a major protein in the precipitated fraction, indicating that centrifugal fractionation is sufficient to purify rEhMCF nearly to homogeneity (Fig. 2B).
Production, purification, and characterization of rEhMCF. Production (A) and purification (B) of rEhMCF and the assay system for measurement of its transporter activity (C). Autoradiograph (A) and Coomassie brilliant blue-stained (B) SDS-PAGE gels are shown. The loaded samples, which were prepared by centrifugal fractionation of a reaction mixture from the cell-free translation system, were as follows: I, initial material; S, supernatant; P, precipitate. The different DNA fragments carrying an EhMCF ORF in the pYT-08 vector are as follows: native, a PCR-amplified DNA from E. histolytica cDNA; codon-optimized, a synthetic DNA in which the codon usage of the EhMCF ORF is optimized to that of the wheat germ translation system. (C) Time course of the uptake of ATP into vesicles. The uptake of external [32P]ATP into the substrate-loaded vesicles preloaded with nonradiolabeled ATP (open) or into empty vesicles (filled) was measured. The data are shown as means ± standard deviations (SDs) (indicated with error bars) of the results from more than three independent experiments. Asterisks indicate significant differences from empty vesicles (*, P < 0.05; **P < 0.01).
We then examined whether the rEhMCF purified as the precipitated fraction was functional by measuring its ability to transport ATP, a standard substrate for MC proteins. Uptake of 32P-labeled ATP into substrate-loaded vesicles proceeded in a time-dependent manner, while uptake of 32P-labeled ATP into empty vesicles could not be detected (Fig. 2C). These results validate the biochemical methodology used in this study. Furthermore, they indicate that rEhMCF is properly reconstituted into the lipid bilayer of asolectin liposomes, indicating that rEhMCF can function as an antiporter.
The EhMCF functions as a PAPS/adenosine 3′,5′-bisphosphate (PAP) and PAPS/PAPS antiporter in vitro.
Our primary question—does EhMCF, a mitosomal protein, participate in the translocation of PAPS across mitosomal membranes in E. histolytica?—can be now addressed. To address this question, we measured the countertransport activity of the reconstituted rEhMCF for ATP with PAPS, ATP, ADP, or AMP. The uptake of 32P-labeled ATP into the substrate-loaded vesicles was significantly higher when PAPS was used as a countersubstrate (11.7 ± 0.4 nmol/mg protein) than when adenosine mono-, di-, or triphosphates were used (3.2 ± 0.2, 0.0 ± 0.5, or 1.8 ± 0.7 nmol/mg protein, respectively). The uptake of 32P-labeled ATP into empty vesicles could not be detected (Fig. 3A). These data are partly inconsistent with those from a previous study owing to differences in the ATP/ADP exchange activity (6). In this study, no uptake of 32P-labeled ATP into the ADP-loaded vesicles or the empty vesicles could be detected, while in the previous study, a significant uptake of 14C-labeled ATP into ADP-loaded vesicles was observed. Assessment of this difference is hampered by the lack of demonstration of uptake of 14C-labeled ATP into the empty vesicles in the former work. They also showed a significant uptake of 14C-labeled ADP into ATP-loaded vesicles, a result which was well supported by showing competitive inhibition by excess cold ADP as well as the absence of uptake into the empty vesicles. This inconsistency in the reported ATP/ADP exchange activity may be due to differences in proteoliposome preparations, in components in the reactions (e.g., the use of radiolabeled materials as a substrate or of preloaded molecules as a countersubstrate), and/or in assay conditions (e.g., incubation time or concentrations of substrate added and of countersubstrate preloaded). Another explanation is possible differences in the lipid compositions because of different membrane sources having been used for protein reconstitutions (this study and reference 6). More importantly, the data (Fig. 3A) clearly indicate that EhMCF indeed has the capacity for countertransport of ATP using PAPS as a preferred countersubstrate.
Countersubstrate specificities of rEhMCF reconstituted in vesicles. The substrate-loaded vesicles (preloaded substances are indicated by names) and the empty vesicles (indicated by a minus sign) were assayed for the uptake of either [32P]ATP (A) or [35S]PAPS (B). The specific transport activity (A) and the transport activity relative to that of PAPS as the control (set as 100%) (B) are shown. The data are presented as the means ± SDs calculated from the results of three independent experiments. The raw data used to calculate the relative activity levels are as follows: PAPS, 32.8 ± 7.2; PAP, 39.1 ± 13.9; ATP, 14.1 ± 5.0; ADP, 13.6 ± 3.2; AMP, 2.4 ± 0.7; APS, 7.3 ± 1.6; CoA, 4.2 ± 0.7; sulfate, 0.1 ± 0.1; GTP, 0.1 ± 0.2; −, 0.0 ± 0.1 (nmol/mg protein/min). Asterisks indicate significant differences between PAPS and other countersubstrates (*, P < 0.05; **P < 0.01).
To investigate further the countersubstrate specificity, we measured the activity of PAPS countertransport with various substances. Among potential countersubstrates examined, adenosine 3′,5′-bisphosphate (PAP) gave the highest activity for the uptake of 35S-labeled PAPS into the substrate-loaded vesicles (119.2 ± 42.4% relative to that of PAPS as 100% control). Adenosine mono-, di-, and triphosphates all gave moderate activities, but a preference for ATP and ADP over AMP (43.0 ± 15.2% and 41.5 ± 9.6% over 7.3 ± 2.2%, respectively) was observed. Similarly, adenosine 5′-phosphosulfate (APS) and coenzyme A (CoA) also gave moderate activities (22.3 ± 4.8% and 12.8 ± 2.2%, respectively). Sulfate and GTP gave nearly untraceable activities (0.3 ± 0.2% and 0.3 ± 0.5%, respectively). Consistent with the assays using 32P-labeled ATP, uptake of 35S-labeled PAPS into empty vesicles could not be detected at all (Fig. 3B). These data (Fig. 3) indicate that EhMCF is different from archetypal AACs regarding countersubstrate specificity and are in agreement with the previous finding that carboxyatractyloside and bongkrekic acid, specific inhibitors of most AACs, are not effective against EhAAC (6). More importantly, these data indicate that EhMCF functions mainly as a PAPS/PAP and PAPS/PAPS antiporter in vitro.
Evidence that EhMCF functions mainly as a PAPS/ATP antiporter in vivo.
Generally, PAP is produced together with a sulfated molecule through the catalytic action of SULT with respect to PAPS and a sulfur acceptor. PAP is then degraded to adenosine 5′-phosphate and Pi by 3′(2′),5′-bisphosphate nucleotidase [19, 20; 3′(2′),5′-bisphosphate nucleotidase is described here as PAPase for ease of reading]. In E. histolytica, all the EhSULTs are localized in the cytosol (Fig. 1) (5), and EhMCF has a high capacity for countertransport of PAPS with PAP or PAPS in vitro (Fig. 3). Therefore, we inferred that knowing the localization of EhPAPase is important for predicting the availability of the substrates for EhMCF in vivo; this will help unravel its precise role.
In the E. histolytica genome, three genes encoding putative PAPase1 to -3 (EHI_193350, EHI_179820, and EHI_175410, respectively) (AmoebaDB; http://amoebadb.org/amoeba/) are present. To determine the localization of all the PAPases in E. histolytica, we established independent transformants expressing each HA-tagged EhPAPase. In all the transformants analyzed, except in organelles such as the nucleus, vacuoles, and small vesicles, the fluorescent signals were evenly distributed throughout cells, indicating that EhPAPase1 to -3 are localized in the cytosol (Fig. 4). Consistent with these observations, the cytosolic localizations of EhPAPase1 and -2 have been recently reported (21, 22). Moreover, EhPAPase1 has been biochemically characterized (21) and active transcription of EhPAPase1 to -3 has been reported (23). Collectively, these findings indicate that all the EhPAPases are functional in the E. histolytica cytosol.
Localization of EhPAPase1 to -3 in E. histolytica. Indirect immunofluorescence images of transformants expressing HA-tagged EhPAPase1 to -3 are shown. Bars, 10 μm.
This notion, together with the cytosolic localization of all the EhSULTs (Fig. 1) (5), suggests that PAP produced by the catalytic action of EhSULTs is sequentially degraded by EhPAPases; therefore, the PAP concentration in the cytosol is maintained at a low level in E. histolytica. This is consistent with the general idea that PAP is toxic to cells (24) but contradicts another interpretation of the results (Fig. 3), i.e., that EhMCF functions mainly as a PAPS/PAP and PAPS/PAPS antiporter in vitro. However, the results (Fig. 3) also showed that the reconstituted rEhMCF had high activity for the exchange of ATP with PAPS. ATP, which is mainly synthesized through cytosolic pathways, is a crucial molecule for the sulfate activation pathway to produce PAPS in E. histolytica mitosomes, whereas PAPS is a necessary substrate for EhSULTs to produce sulfolipids in the cytosol. Hence, maintaining a molecule such as EhMCF that has a high capacity for exchanging ATP with PAPS across the mitosomal membrane could be beneficial for E. histolytica.
In agreement with this interpretation, we previously demonstrated that MCFgs, an E. histolytica G3 strain in which EhMCF was knocked down by gene silencing, showed a significant reduction in sulfate activation activity and a marked growth defect (3). This finding can be now explained by that the shortage of ATP in the mitosomes as well by as a shortage of PAPS in the cytosol. This would impair the synthesis of sulfolipids, which plays important roles in E. histolytica growth. However, a study to confirm the importance of sulfolipids in E. histolytica is needed.
Mechanistically, the EhMCF knockdown causes an evident reduction in the exchange activity of PAPS with ATP across the mitosomal membrane in E. histolytica; this reduction lowers the flow of cytosolic ATP into mitosomes, leading to a mitosomal ATP shortage. This shortage severely impairs mitosomal sulfate activation activity; therefore, PAPS production significantly decreases. In addition, even the PAPS produced cannot be efficiently translocated into the cytosol because of the defect in EhMCF activity. Consequently, PAPS levels in the cytosol become remarkably low, resulting in the halting of almost all EhSULT sulfolipid synthesis.
In conclusion, we suggest that EhMCF functions mainly as a PAPS/ATP antiporter and links the mitosomal sulfate activation pathway to the cytosolic chain reaction that is composed of EhSULTs and EhPAPases in E. histolytica (Fig. 5).
A scheme for sulfate metabolism in E. histolytica. The flow of metabolites and the enzymes involved are depicted, based on evidence from previous studies (3, 5, 13, 21, 22) as well as from the present study. APS, adenosine 5′-phosphosulfate; APSK, APS kinase; AS, ATP sulfurylase; IPP, inorganic pyrophosphatase; MCF, mitochondrial carrier family; NaS, sodium/sulfate symporter; PAP, adenosine 3′,5′-bisphosphate; PAPase, 3′(2′),5′-bisphosphate nucleotidase; PAPS, 3′-phosphoadenosine 5′-phosphosulfate; Pi, inorganic phosphate; PiC, Pi carrier; PPi, pyrophosphate; SULT, sulfotransferase.
ACKNOWLEDGMENTS
This work was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan to F.M. (22890136, 24117517, and 26117719), to A.N. (24580094, 24589511, and 15K07006), to H.Y. (25460594), to Y.T. (24117516 and 26117717), and to T.N. (23117001, 23117005, and 26293093) and from the Research Program on Emerging and Re-emerging Infectious Diseases from Japan Agency for Medical Research and Development (AMED) to T.N. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
We thank Shouko Takao, Ritsuko Yoshida, and Shizuko Furukawa for technical assistance.
F.M. and A.N. designed and performed the experiments; F.M., A.N., H.Y., Y.T., and T.N. analyzed the data and wrote the paper.
We declare that we have no conflict of interest.
REFERENCESRalstonKS, PetriWAJr2011Tissue destruction and invasion by Entamoeba histolytica.
27:254–263. doi:10.1016/j.pt.2011.02.006.21440507MakiuchiT, NozakiT2014Highly divergent mitochondrion-related organelles in anaerobic parasitic protozoa.
100:3–17. doi:10.1016/j.biochi.2013.11.018.24316280Mi-ichiF, MakiuchiT, FurukawaA, SatoD, NozakiT2011Sulfate activation in mitosomes plays an important role in the proliferation of Entamoeba histolytica.
5:e1263. doi:10.1371/journal.pntd.0001263.21829746van der GiezenM2009Hydrogenosomes and mitosomes: conservation and evolution of functions.
56:221–231. doi:10.1111/j.1550-7408.2009.00407.x.19527349Mi-ichiF, Abu Yousuf Nakada-TsukuiM, NozakiK, T2009Mitosomes in Entamoeba histolytica contain a sulfate activation pathway.
106:21731–21736. doi:10.1073/pnas.0907106106.19995967ChanKW, SlotboomDJ, CoxS, EmbleyTM, FabreO, van der GiezenM, HardingM, HornerDS, KunjiER, Leon-AvilaG, TovarJ2005A novel ADP/ATP transporter in the mitosome of the microaerophilic human parasite Entamoeba histolytica.
15:737–742. doi:10.1016/j.cub.2005.02.068.15854906MaiZ, GhoshS, FrisardiM, RosenthalB, RogersR, SamuelsonJ1999Hsp60 is targeted to a cryptic mitochondrion-derived organelle (“crypton”) in the microaerophilic protozoan parasite Entamoeba histolytica.
19:2198–2205.10022906MüllerM, MentelM, van HellemondJJ, HenzeK, WoehleC, GouldSB, YuRY, van der GiezenM, TielensAG, MartinWF2012Biochemistry and evolution of anaerobic energy metabolism in eukaryotes.
76:444–495. doi:10.1128/MMBR.05024-11.22688819Mi-ichiF, MiyamotoT, TakaoS, JeelaniG, HashimotoT, HaraH, NozakiT, YoshidaH2015Entamoeba mitosomes play an important role in encystation by association with cholesteryl sulfate synthesis.
112:E2884–E2890. doi:10.1073/pnas.1423718112.25986376MonnéM, PalmieriF2014Antiporters of the mitochondrial carrier family.
73:289–320. doi:10.1016/B978-0-12-800223-0.00008-6.24745987PalmieriF, PierriCL, De GrassiA, Nunes-NesiA, FernieAR2011Evolution, structure and function of mitochondrial carriers: a review with new insights.
66:161–181. doi:10.1111/j.1365-313X.2011.04516.x.21443630NozawaA, FujimotoR, MatsuokaH, TsuboiT, TozawaY2011Cell-free synthesis, reconstitution, and characterization of a mitochondrial dicarboxylate-tricarboxylate carrier of Plasmodium falciparum.
414:612–617. doi:10.1016/j.bbrc.2011.09.130.21986531DolezalP, DagleyMJ, KonoM, WolynecP, LikicVA, FooJH, SedinovaM, TachezyJ, BachmannA, BruchhausI, LithgowT2010The essentials of protein import in the degenerate mitochondrion of Entamoeba histolytica.
6:e1000812. doi:10.1371/journal.ppat.1000812.20333239ThuswaldnerS, LagerstedtJO, Rojas-StutzM, BouhidelK, DerC, Leborgne-CastelN, MishraA, MartyF, SchoefsB, AdamskaI, PerssonBL, SpeteaC2007Identification, expression, and functional analyses of a thylakoid ATP/ADP carrier from Arabidopsis.
282:8848–8859. doi:10.1074/jbc.M609130200.17261580GigolashviliT, GeierM, AshykhminaN, FrerigmannH, WulfertS, KruegerS, MugfordSG, KoprivaS, HaferkampI, FluggeUI2012The Arabidopsis thylakoid ADP/ATP carrier TAAC has an additional role in supplying plastidic phosphoadenosine 5′-phosphosulfate to the cytosol.
24:4187–4204. doi:10.1105/tpc.112.101964.23085732HjortK, GoldbergAV, TsaousisAD, HirtRP, EmbleyTM2010Diversity and reductive evolution of mitochondria among microbial eukaryotes.
365:713–727. doi:10.1098/rstb.2009.0224.20124340NozawaA, TozawaY2014Modifications of wheat germ cell-free system for functional proteomics of plant membrane proteins.
1072:259–272. doi:10.1007/978-1-62703-631-3_19.24136528BernhardF, TozawaY2013Cell-free expression—making a mark.
23:374–380. doi:10.1016/j.sbi.2013.03.012.23628286MurguíaJR, BellesJM, SerranoR1995A salt-sensitive 3′(2′),5′-bisphosphate nucleotidase involved in sulfate activation.
267:232–234. doi:10.1126/science.7809627.7809627RamaswamySG, JakobyWB1987(2′)3′,5′-Bisphosphate nucleotidase.
262:10044–10047.3038862Faisal TariqueK, Arif Abdul RehmanS, GourinathS2014Structural elucidation of a dual-activity PAP phosphatase-1 from Entamoeba histolytica capable of hydrolysing both 3′-phosphoadenosine 5′-phosphate and inositol 1,4-bisphosphate.
70:2019–2031. doi:10.1107/S1399004714010268.25004978Faisal TariqueK, Arif Abdul RehmanS, BetzelC, GourinathS2014Structure-based identification of inositol polyphosphate 1-phosphatase from Entamoeba histolytica.
70:3023–3033. doi:10.1107/S1399004714021245.25372691HusainA, JeelaniG, SatoD, NozakiT2011Global analysis of gene expression in response to L-Cysteine deprivation in the anaerobic protozoan parasite Entamoeba histolytica.
12:275. doi:10.1186/1471-2164-12-275.21627801ToledanoE, OgryzkoV, DanchinA, LadantD, MecholdU20123′-5′ phosphoadenosine phosphate is an inhibitor of PARP-1 and a potential mediator of the lithium-dependent inhibition of PARP-1 in vivo.
443:485–490. doi:10.1042/BJ20111057.22240080oai:pubmedcentral.nih.gov:46648792015-12-10eukcellpmc-openEukaryot CellEukaryotic CelleukcelleukcellEUKCELLEukaryotic Cell1535-97781535-9786American Society for Microbiology1752 N St., N.W., Washington, DCPMC4664879PMC46648794664879264536502645365000142-1510.1128/EC.00142-15MinireviewsAdaptations of the Secretome of Candida albicans in Response to Host-Related Environmental ConditionsMinireviewKlisFrans M.BrulStanleyDepartment of Molecular Biology and Microbial Food Safety, Swammerdam Institute for Life Sciences, University of Amsterdam, Amsterdam, The NetherlandsAddress correspondence to Frans M. Klis, F.M.Klis@uva.nl.
Citation Klis FM, Brul S. 2015. Adaptations of the secretome of Candida albicans in response to host-related environmental conditions. Eukaryot Cell 14:1165–1172. doi:10.1128/EC.00142-15.
The wall proteome and the secretome of the fungal pathogen Candida albicans help it to thrive in multiple niches of the human body. Mass spectrometry has allowed researchers to study the dynamics of both subproteomes. Here, we discuss some major responses of the secretome to host-related environmental conditions. Three β-1,3-glucan-modifying enzymes, Mp65, Sun41, and Tos1, are consistently found in large amounts in culture supernatants, suggesting that they are needed for construction and expansion of the cell wall β-1,3-glucan layer and thus correlate with growth and might serve as diagnostic biomarkers. The genes ENG1, CHT3, and SCW11, which encode an endoglucanase, the major chitinase, and a β-1,3-glucan-modifying enzyme, respectively, are periodically expressed and peak in M/G1. The corresponding protein abundances in the medium correlate with the degree of cell separation during single-yeast-cell, pseudohyphal, and hyphal growth. We also discuss the observation that cells treated with fluconazole, or other agents causing cell surface stress, form pseudohyphal aggregates. Fluconazole-treated cells secrete abundant amounts of the transglucosylase Phr1, which is involved in the accumulation of β-1,3-glucan in biofilms, raising the question whether this is a general response to cell surface stress. Other abundant secretome proteins also contribute to biofilm formation, emphasizing the important role of secretome proteins in this mode of growth. Finally, we discuss the relevance of these observations to therapeutic intervention. Together, these data illustrate that C. albicans actively adapts its secretome to environmental conditions, thus promoting its survival in widely divergent niches of the human body.
INTRODUCTION
The fungal pathogen Candida albicans is a highly specialized inhabitant of warm-blooded animals (mammals and birds). It preferentially colonizes mucosal surfaces and the skin but can also invade deeper-lying tissues and cause systemic infections that are difficult to treat and frequently lethal (1). To survive under the challenging and divergent conditions associated with the various mucosal surfaces in the human body, C. albicans disposes of a wide arsenal of virulence traits that help it to cope with antimicrobial peptides, the complement system, engulfment by macrophages, antibodies, hypoxic conditions, iron restriction, etc. A fascinating trait is its ability to switch reversibly between various growth forms, including among others the single-cell yeast form, which is especially suitable for dispersion of the fungus; the hyphal form, which facilitates adhesion to host tissues and promotes invasive growth and escape from engulfment by immune cells; and an intermediate, pseudohyphal growth form. C. albicans also forms biofilms (surface-associated microbial communities), which clinically speaking represent a highly relevant mode of growth and in which yeast, pseudohyphal, and hyphal cells cooccur and become encapsulated by substantial amounts of extracellular, macromolecular material. Biofilm formation on abiotic surfaces of medical devices and prostheses and on teeth has therefore been extensively studied (2, 3). The first contacts between C. albicans and host cells occur predominantly at the cell surface, and this presumably explains why the external protein coat of C. albicans cell walls consists of a wide variety of glycoproteins with specialized functions, many of which are under tight control, thus promoting survival under diverse stress conditions (4, 5). Equally important, C. albicans secretes a variety of glycoproteins that help to forage for nutrients by degrading host proteins, lipids, and glycogen, while others acquire iron and zinc ions and provide protection against antimicrobial peptides. Other glycoproteins help to form and strengthen biofilms and to accumulate extracellular matrix material. Together, we designate these secreted proteins as the secretome sensu stricto (see below). The introduction of mass spectrometry in protein research has made it possible to study the protein assortment of entire cells or tissues and also well-defined subsets of proteins (subproteomes such as the cell wall proteome and the secretome), not only qualitatively but also quantitatively. This review discusses recent mass spectrometric explorations of the dynamics of the secretome of C. albicans depending, for example, on growth form and pH or in response to cell surface stress. For complementary reviews, the reader is referred to reference 4, which includes an extensive section about secretome proteins, and to a more recent review (6). In addition, the Candida albicans PeptideAtlas and the Candida Genome Database are recommended for detailed information about mass spectrometrically identified peptides (7, 8).
CLASSIFICATION OF MAJOR SECRETOME PROTEINS (SENSU STRICTO) OF C. ALBICANS
In this review, we will restrict ourselves to the secretome in the narrow sense of the word (secretome sensu stricto), that is, we will discuss only those proteins that possess an N-terminal signal peptide for entering the classical secretory pathway and lack internal transmembrane sequences. The major advantages of analyzing this protein category separately are that it is physiologically well defined, is limited in size, and is not affected by accidentally released proteins. This facilitates statistical analysis of the data (9), simplifies discussion of the results, and leads to physiologically relevant conclusions and testable hypotheses. For more information about the secretome in the wide sense of the word (secretome sensu lato), which includes a number of known cytosolic proteins, we refer to references 6 and 10.
The experimentally identified secretome proteins (sensu stricto) (currently, about 70 [9–15]) can be classified into two major groups: nonglycosylphosphatidylinositol (non-GPI) proteins, which lack a C-terminal signal sequence for the addition of a GPI anchor in the endoplasmic reticulum, and GPI proteins (Fig. 1). To avoid an excessively long list, this review focuses on the more abundant secretome proteins that are observed under more than one growth condition and/or have a known or predicted function.
Major features of the secretome (sensu stricto) of C. albicans. The wall proteins are represented as short line segments perpendicular to the cell surface. M, mother cell; D, daughter cell; GPI-WP, GPI-modified wall protein. The role of Csa2 in heme binding is speculative (6, 40, 41). Note that Als3 and Phr1 are possibly directly released from the cell wall by Sap9/10 activity (47).
NON-GPI PROTEINSModification of cell wall polysaccharides.
Eleven enzymes (distributed over six families of glycosylhydrolases [GHs]) are involved in glycan chain elongation and branching and in glycan degradation (for example, during cell separation and emergence of a new bud or hyphal branch). Throughout this paper, we will follow the GH family classification according to the CAZy (for carbohydrate-active enzymes) database (16).
(i) β-1,3-Glucan-modifying enzymes.
β-1,3-Glucan-modifying enzymes include Xog1 (GH5); Tos1 (GH16); Bgl2, Mp65, and Scw11 (GH17); Eng1 (GH81); Sim1 and Sun41 (GH132); and Dag7 (Barwin-like endoglucanase domain, PF03330) and GH (unspecified) (8).
(ii) Chitin-degrading enzymes.
Chitin-degrading enzymes include Cht1 and Cht3 (GH18); they carry out degradation of the primary septum between mother and daughter cells, thus initiating cell separation. Note that the GH18 family member Cht2 is a GPI protein.
Three of the 11 corresponding genes (CHT3, which encodes the major chitinase; ENG1, which encodes an endo-β-1,3-glucanase; and SCW11, which encodes a β-1,3-glucan-modifying enzyme) are periodic genes with maximal expression in the M/G1 period of the cell cycle and are target genes of the transcription factor Ace2 (17–20), consistent with a role in cell separation. Agar-grown colonies, mucosal biofilms, and biofilms formed on abiotic surfaces produce abundant amounts of extracellular matrix material (3, 21–23). The transglucosylase Bgl2 and the exoglucanase Xog1 (together with the GPI-modified, putative β-1,3-glucan-elongating enzyme Phr1) seem to be directly involved in formation and modification of extracellular matrix material in biofilms (24).
Nutrient acquisition. (i) Glucoamylases.
Glucoamylases include Gca1 and Gca2 (GH31); the predicted substrates are glycogen and starch (4, 16, 25, 26). Interestingly, maltose, which is a degradation product of both polysaccharides, is a carbon source known to promote hyphal growth (27). It has been suggested previously that Gca1 and Gca2 are directly involved in promoting matrix production in biofilms by enzymatic release of soluble β-1,3-glucan fragments from insoluble β-1,3-glucan chains (28). However, Gca1 and Gca2 possess the hallmarks of an α-glucosidase/glucoamylase, cleaving alpha-1,4-glucosidic linkages (8), and hence, it seems unlikely that they might cleave beta-1,3-glucosidic linkages, which have a spatial structure highly different from that of alpha-1,4-glucosidic linkages.
(ii) N-Acetylhexosaminidase.
HEX1 is a glucose-repressed gene (29). Hex1 (GH20) is found in the periplasm and in the culture supernatant (29). It is possibly involved in the release of GlcNAc residues from host tissues for use as a carbon or nitrogen source (29). Note that GlcNAc also acts as a signaling molecule and induces and maintains hyphal growth in glucose-derepressed cells (30–33).
(iii) Lipid degradation by Lips and Plbs.
Lipids are degraded by lipases (Lips) including Lip1 to Lip6 and Lip8 to Lip10 and by phospholipases (Plbs) including Plb1 and Plb2, reviewed in references 4 and 34.
(iv) Protein degradation by aspartyl proteases.
Protein degradation by aspartyl proteases includes, for example, degradation of mucins and host immune proteins (4, 34–37), by aspartyl proteases such as Sap1 to Sap8. Note that Sap9 and Sap10 are cell surface-associated GPI proteins (38).
(v) Metal ion acquisition.
Metal ion acquisition includes acquisition of zinc by Pra1 (39) and, probably, acquisition of heme (iron) by Csa2 (6, 12, 40, 41). Intriguingly, Pra1 is also involved in immune evasion (42, 43).
Pry family.
The Pry family (11) consists of five members: Pry1, Rbe1, Rbt4, and two uncharacterized open reading frames (ORFs) (19.6200 and 19.2336). Rbe1 and Rbt4 have been often identified in culture supernatants (10–14). Interestingly, whereas Rbe1 is much more abundant in yeast culture supernatants, Rbt4 is much more abundant in hyphal culture supernatants (11–13) (see also Table 2). The virulence of an Δrbe1 Δrbt4 double deletion strain in the mouse model for systemic infection is strongly diminished (11), but the precise function of the Pry family proteins is still unknown.
Coi1.
Coi1 is a small protein of 191 amino acids that is relatively abundant and is consistently found in the medium under all growth conditions tested (10–14). Homologous proteins are found in only a limited number of Candida spp.—Candida dubliniensis, Candida orthopsilosis, Candida parapsilosis, and Candida tropicalis—and in Lodderomyces elongisporus (8). Its function is unknown.
Signaling protein Msb2.
The signaling protein Msb2 is located in the plasma membrane and involved in signaling through activation of the Cek1 mitogen-activated protein (MAP) kinase (44). Although Msb2 has an internal transmembrane sequence, it is included in this classification and in the secretome sensu stricto, because it has a large extracellular, highly O-glycosylated domain that is shed into the medium and consistently identified in culture supernatants (10–14, 45). Interestingly, it also serves as a broad-range protectant against antimicrobial peptides (45, 46).
GPI PROTEINS
GPI proteins are targeted to the plasma membrane or become covalently linked to the β-glucan layer of the cell wall through their GPI anchor or are found at both locations. However, most GPI proteins are also identified in culture supernatants (10–14). There are several (possible) explanations for their presence.
Target proteins of Sap9 and Sap10.
Some wall-bound GPI proteins are released by the surface-bound aspartyl proteases Sap9 and Sap10, such as Cht2 (47). Other candidates for active and controlled release, such as the adhesion protein Als3 and the transglucosylase Phr1, will be discussed below.
GPI proteins from the neck region.
In single-cell yeast cultures, wall-bound GPI proteins are released from the neck region during cell separation. The wall between mother cell and growing bud is continuous, and complete cell separation therefore requires not only degradation of the primary septum by chitinase activity but also degradation of the lateral wall in the neck region.
Accidental release.
Especially in shaken cultures, wall-bound GPI proteins might be released during emergence of a new bud or hyphal branch, which requires localized cell wall softening, or during periods of isotropic growth, which requires insertion of new cell wall polysaccharides and wall proteins into the existing wall.
Wall protein precursors may be washed out into the medium before they become covalently linked to the glucan-chitin network, especially in shaken cultures.
DYNAMICS OF THE SECRETOME OF C. ALBICANS
In this section, the term “apparent abundance” is introduced. It is defined as the number of spectral counts per protein divided by the total number of spectral counts of all secretome proteins (sensu stricto) and expressed as a percentage (48). This is a semiquantitative measure that allows comparison of the individual contributions of the secretome proteins and, importantly, allows estimating and comparing the fold changes of individual secretome proteins upon changes in environmental conditions, including conditions that induce growth as single yeast, pseudohyphal, or hyphal cells. We prefer the use of apparent abundances to that of normalized spectral abundance factors (NSAFs [9, 10, 49]). In the latter approach, the number of spectral counts per protein is normalized for protein length, which for nonglycosylated proteins results in more accurate estimates of protein abundance. However, secretome proteins sensu stricto frequently contain long, heavily O-glycosylated sequences, which rarely result in detectable tryptic peptides and thus lead to serious underestimations of protein abundances (for example, about 10-fold in the case of Msb2 [7, 8]) and decreased accuracy. A rough estimate of the number of secretome proteins per cell present in yeast culture supernatants, based on the data in references 12 and 50 and assuming an average protein mass of 40 kDa, is about 4 × 105 to 5 × 105.
Three prominent secretome proteins.
The three β-glucan-modifying enzymes Mp65 (GH17), Sun41 (GH132), and Tos1 (GH16) belong to the most prominent (detectable) secretome proteins, both in single-cell yeast culture supernatants and hyphal culture supernatants and during pseudohypha-like growth induced by fluconazole (Table 1). Similar values for their apparent abundances have been obtained under diverse growth conditions (6, 9, 10, 12–14). Conceivably, they are involved in various ways in the construction and remodeling of the β-1,3-glucan layer in the cell wall during growth (8, 16, 51–55).Their combined apparent abundance accounts for one-fourth to one-third of all secretome proteins. Consistent with this, the gene sequences of Mp65, Sun41, and Tos1 have a relatively high codon bias index (MP65, 0.71; SUN41, 0.64; TOS1, 0.59 [8]), suggesting that these genes are strongly expressed. Surprisingly, although Mp65 lacks a C-terminal GPI anchor to connect it covalently to the β-glucan network, it is usually also found in (hot-SDS-extracted) cell walls (10, 13, 14, 30, 37, 40). Scw11 and Sim1 belong to the same families as Mp65 and Sun41, respectively, and both show high apparent abundances in single-cell yeast cultures (Tables 2 to 4). However, in hyphal cultures and to a lesser extent also in pseudohyphal-growth cultures, their apparent abundances are considerably lower, suggesting that Scw11 and Sim1 might play a direct role in cell separation.
Three prominent secretome (sensu stricto) proteins of C. albicans grown under various conditions
Protein
Apparent abundance (%)a
Yeast
Hyphal
FCZb
Mp65
12.4
14.3
8.5
Sun41
8.0
8.6
7.4
Tos1
9.9
12.4
8.7
Based on data from references 12 to 14. Yeast cultures were grown at 30°C and pH 7.4; hyphal cultures were grown at 37°C and pH 7.4.
FCZ, fluconazole-treated cells grown at 37°C and pH 7.4. They form pseudohyphal aggregates (14).
Yeast state- and hyphal state-enriched secretome proteins
Change and protein
Apparent abundance (%)a
Fold change
Yeast
Hyphal
Yeast to hyphal change
Cht3b
8.6
1.9
4.5
Scw11b
8.6
1.9
4.5
Xog1
8.4
1.2
7.1
Sim1
7.8
2.6
3.0
Eng1b
7.3
NDd
>28
Bgl2
3.6
0.5
7.5
Rbe1
3.0
ND
>12
Cht1
2.5
ND
>9
Dag7
2.3
ND
>9
Hyphal to yeast change
Sap6
ND
13.1
>62
Rbt4
1.8
9.7
5.7
Als3c
ND
9.3
>44
Sap4
ND
8.8
>42
Based on data from reference 12. Yeast cultures were grown at 30°C and pH 7.4; GlcNAc-induced hyphal cultures were grown at 37°C and pH 7.4 (12).
Periodically expressed genes with maximal expression in the M/G1 phase of the cell cycle.
The GPI protein is in bold.
ND, not detected.
Main features of the secretomes of fluconazole-supplemented cultures
Change and protein
Apparent abundance (%)a
Fold change
FCZ
Yeast
FCZ to yeast
Phr1b
10.4
ND
>49
Yeast to FCZ
Cht3c
4.3
8.6
2.0
Scw11c
5.6
8.6
1.5
Sim1
4.2
7.8
1.9
Eng1c
0.3
6.7
22
Cht1
0.3
2.5
7.5
Based on data from references 12 and 14. Yeast cultures were grown at 30°C and pH 7.4 (12); fluconazole (FCZ)-treated cultures were grown at 37°C and pH 7.4. ND, not detected.
The GPI protein is in bold.
Periodically expressed genes with maximal expression in the M/G1 phase of the cell cycle.
Main features of secretomes of low-pH-grown cultures
Change and protein
Apparent abundance (%) at pHa:
Fold change
4
7.4
pH 4 to pH 7.4 change
Utr2b
8.0
3.2
2.5
Bgl2
7.4
2.8
2.6
Plb4.5b
6.0
ND
>32
Pir1
3.6
1.1
3.2
Data based on reference 14. ND, not detected.
GPI proteins are in bold.
Yeast state-enriched secretome proteins: yeast versus hyphal cultures.
In single-cell yeast cultures, the daughter cells become separated from the mother cell, a process that requires chitinase activity to degrade the primary septum formed during cytokinesis and β-1,3-endoglucanase activity to degrade the β-1,3-glucan layer in the lateral wall of the neck region. Conceivably, also transglucosylase activity is needed for repair activity. This results in much higher apparent abundances of the cell separation enzymes Cht3 and Eng1 and the potential repair enzyme Scw11 in the culture solution of unicellular budding yeast than in hyphal cells (Table 2). Although the apparent abundances of Bgl2, Cht1, Dag1, Rbe1, Sim1, and Xog1in the culture solution of hyphal cultures are also strongly reduced, the corresponding genes do not seem to be periodically expressed (8, 17).
Hyphal state-enriched secretome proteins: hyphal versus yeast cultures.
Many host-related chemical and physical conditions, such as a neutral pH, a temperature of 37°C, low oxygen levels and high CO2 concentrations, the presence of GlcNAc (derivatives) or serum, iron restriction, and low glucose concentrations, promote hyphal growth (30, 56–58). The natural inducer N-acetylglucosamine (GlcNAc) offers several advantages. Under glucose-derepressing conditions, which occur in many niches in the human body, GlcNAc and also GlcNAc derivatives strongly initiate and sustain hyphal growth (31, 32, 59). GlcNAc is, for example, found in the polysaccharide hyaluronan, which consists of alternating residues of glucuronic acid and GlcNAc and is an abundant component of the extracellular matrix of epithelial and connective tissues (60) and is a known inducer of hyphal growth (59). An additional advantage of using GlcNAc to induce hyphal growth is that it, in contrast to bovine serum, allows (mass spectrometric) analysis of the hyphal secretome (12). GlcNAc-induced hyphal cultures contain four additional proteins in the culture medium that become about as equally prominent as Mp65, Sun41, and Tos1 (Tables 1 and 2), namely, the aspartyl proteases Sap4 and Sap6; the Pry family protein Rbt4, whose function is unknown; and the GPI-modified adhesion protein Als3 (Tables 1 and 2). This is consistent with the strongly increased expression levels of the corresponding genes in GlcNAc-induced hyphal cultures (31; see also reference 61) for serum-induced hyphal cultures. Of interest, Als3 is a hypha-specific adhesive GPI protein that is targeted to the cell wall and required for biofilm formation (62); it is also known to promote the formation of mixed fungal-bacterial biofilms through its N-terminal adhesion domain (63, 64). This raises the question of whether biofilm matrix material might become enriched with released Als3 and whether it might promote the cohesion of the extracellular matrix material. It is unknown whether Als3 is passively released from the cell wall or is released through a regulated, enzymatic process, possibly involving the GPI proteins Sap9 and/or Sap10 (38, 47). The secreted aspartyl proteases Sap4 and Sap6 are strongly associated with the yeast-to-hypha transition (65). Note that the presence of an ammonium salt as a rich nitrogen source in the culture medium at neutral pH does not seem to interfere with the accumulation of Sap4 and Sap6 at high levels in the medium (Table 2).
Major changes in apparent abundance in the secretome of fluconazole-treated cells.
Pseudohyphal (or pseudohypha-like) growth can be induced in various ways, for example, by treating single-cell yeast cultures with sublethal concentrations of the antifungal compound fluconazole (14). Fluconazole blocks an Erg11-mediated demethylation step of the planar ring structure during ergosterol synthesis, and this results in the formation of structurally suboptimal sterols and increased fluidity of the plasma membrane (66). The cells respond by suppression of cell separation and by increased phosphorylation of the MAP kinase Mkc1 (14), which mediates the cell wall integrity pathway. Pseudohypha-like cell aggregates are formed, which separate into single cells when treated with exogenous chitinase (14). These effects are not specific for fluconazole but also occur when the cells are treated with the detergent SDS, grown at the harmful temperature of 42°C, or treated with the cell wall construction-perturbing compound calcofluor white or Congo red (14, 51) (Fig. 2). This suggests that many forms of plasma membrane stress or cell wall stress, together here referred to as cell surface stress, will cause pseudohypha-like growth and constitutively activate the cell wall integrity pathway. Indeed, cell separation suppression is a common phenotype among all genetic mutants that have to cope with a weakened cell wall resulting from defective N- or O-glycosylation of secretory glycoproteins (67–70) or lack proteins involved in cell wall formation, such as Bgl2 (71); Ecm33, which is a GPI protein required for cell wall integrity (72); the secretome enzyme Sun41 (51, 52); and the GPI-modified and cell surface-located aspartyl proteases Sap9 and Sap10 (38), as well as Pir1, a putative cross-linker of β-glucan chains (73), and this list could be easily extended. Interestingly, CHT3 transcription is markedly downregulated in micafungin-treated cells (74), suggesting that this inhibitor of β-1,3-glucan synthesis might also induce cell separation suppression. Similar observations have been described by O'Meara and coworkers, who treated wild-type cultures with serum in combination with various antifungal drugs at sublethal concentrations and found that under these conditions hyphal growth was suppressed and, instead, pseudohyphal cell clusters were formed (see Fig. 2c in reference 75). A relevant implication of these observations is that “cell separation suppression” as a phenotype does not necessarily prove that the mutated gene is directly involved in cell separation.
Cell surface stress induces pseudohyphal growth. For further discussion, see the text and references 14 and 75.
Consistent with the partial suppression of cell separation in fluconazole-treated cells, resulting in pseudohypha-like growth, the major chitinase Cht3 (GH18), the endo-β-1,3-glucanase Eng1 (GH81), and the β-1,3-glucan-modifying enzyme Scw11 are all decreased in the supernatants of fluconazole-treated cultures, although less so than in hyphal cultures (Tables 2 and 3). Other chitin- and β-1,3-glucan-modifying enzymes, namely, Cht1 (GH18) and Sim1 (GH132), are also decreased in fluconazole-treated and hyphal cells. How exactly suppression of cell separation is regulated in response to cell surface stress is currently unknown. As CHT3, ENG1, and SCW11 are known target genes of the transcription factor Ace2 (18–20), a regulator of cell separation that controls the expression of M/G1-specific genes, it seems likely that fluconazole treatment and many other forms of cell surface stress lead to suppression of Ace2 activity in cell separation. Possibly, cell separation is suppressed through phosphorylation of the transcription factor Efg1, which in turn represses Ace2 target genes and promotes pseudohypha-like growth (76).
Another striking observation is the high apparent abundance of the GPI protein Phr1 in the medium of fluconazole-treated cells, whereas it is not detectable in the medium of single-cell yeast cultures and hyphal cultures. This observation is consistent with the finding that the antifungal compounds ketoconazole and caspofungin stimulate the expression of PHR1 (77). Phr1 (GH72) is a pH-responsive transglucosylase and probably involved in expansion of the β-1,3-glucan layer (78, 79). It also plays an important role in the accumulation of β-1,3-glucan in the extracellular matrix of biofilms, possibly by extending the β-1,3-glucan chains there and thus increasing the cohesiveness of the biofilm (24). This raises the interesting question of whether biofilm-associated Phr1in fluconazole-treated cells might contribute to protection against fluconazole. It is currently unknown how Phr1 is released from the cell surface. However, Phr1 contains two consecutive lysine residues (K451 and K452), suggesting that it might be a substrate of Sap9/10 (47).
Finally, the culture medium of fluconazole-treated yeast cultures contains several low-abundance secretory proteins that so far have not been observed under other growth conditions (13). This does not necessarily prove that this secretome response is specific for fluconazole-treated cells, because, as we have seen above, many other conditions lead to cell surface stress, and this might be in itself enough to trigger the release of members of this subset of proteins.
Secretome proteins increased at low pH.
Acidic conditions are found on the skin (80); in the vagina, due to formation of lactic acid by vaginal epithelial cells and by lactobacilli (81); and in dental plaque, where they are caused by the formation of lactic acid by Streptococcus mutans (82). Two GPI-modified wall proteins (Plb4.5 and Utr2) and the β-1,3-glucan-linked wall protein Pir1 become much more abundant in the culture medium upon lowering the environmental pH, either because at acidic pH their incorporation into the cell wall becomes less efficient or because they are actively released from the wall (Table 4). The apparent abundance of the non-GPI, β-1,3-glucan-modifying protein Bgl2 (GH17) also shows a considerable increase. As this protein contributes to β-1,3-glucan accumulation in biofilms (24), this might mean that at acidic pHs biofilms produce more extracellular matrix.
Lactate-induced secretome proteins.
Although at many infection sites glucose levels are low and, consequently, C. albicans cells are in a glucose-derepressed state, depending instead on alternative carbon sources for growth, glucose-grown cells are often used for research. The results obtained in this way can therefore not always be directly extrapolated to glucose-derepressed cells (83). For example, the levels of three secretome proteins, glucoamylase(s) Gca1/2 (GH31), the aspartyl protease Sap7, and the exo-β-1,3-glucanase Xog1 (GH15), strongly increase when lactate, a nonfermentable carbon source and a much poorer carbon source than glucose, is used to support growth (9, 84). On the other end of the spectrum, there are also a number of secretome proteins whose apparent abundances strongly decrease in lactate-grown cultures (9). The induction of hyphal growth by GlcNAc as discussed above is already blocked at a glucose concentration of 20 mM (33), and this represents another striking example of the importance of selecting the appropriate carbon source and/or glucose concentration in Candida research.
CONCLUDING REMARKS
Similarly to the wall proteome, the secretome of C. albicans operates at the fungus-host interface, and pronounced changes in abundance of individual proteins occur in distinct host niches. For example, the switch from single-cell yeast growth to hyphal growth results in the abundant secretion of the aspartyl proteases Sap4 and Sap6, which significantly contribute to virulence (85). As expected, the pH of the host niche also strongly affects the composition of the secretome. The ability to form biofilms, in which yeast, pseudohyphal, and hyphal cells are found, with the concurrent accumulation of extracellular matrix material and the often strongly lowered susceptibility to antifungal agents, is another important virulence trait. Several secretome proteins, such as Als3, Bgl2, Mp65, Phr1, Xog1, and Sun41, are believed to be involved in biofilm formation (8). In view of their high apparent abundances under widely diverging growth conditions, the secretome proteins Mp65, Sun41, and Tos1 seem attractive candidates for diagnostic purposes. In addition, as Candida infections are usually associated with invasive, hyphal growth, and as Als3, Rbt4, Sap4, and Sap6 become abundant in the supernatants of hyphal cultures, these proteins also are potential diagnostic candidates. Two secretome proteins, Als3 (actually, a recombinant protein that covers the N-terminal immunoglobulin-like domain of Als3) and Sap2, are currently being actively pursued for vaccine development (86–88). Als3 is an abundant, multifunctional GPI-modified cell wall protein (86, 89, 90) and as such is a highly attractive vaccine candidate. The finding that Als3 also becomes an abundant secretome protein in hyphal cultures (12) with a proposed role in biofilm formation further increases its attractiveness. It has been argued that a multivalent vaccine might be more effective to combat the various types of Candida infections (91). Conceivably, a recombinant protein consisting of two or three of the most immunogenic epitopes from various cell wall proteins (92) and from secretome proteins such as the aspartyl proteases Sap4 and Sap6 (12, 85); the three secretome proteins involved in the formation of biofilm matrix beta-glucan (Bgl2, Phr1, and Xog1) (24); Sun41 and Sim1, which form a two-member family and are synthetically lethal (52); and the Pry proteins Rbe1 and Rbt4 (11), each separated from the other by a nonimmunogenic linker sequence, might be an option. Similar approaches might be considered for other pathogenic fungi. Finally, many regulatory pathways have been shown to be involved in controlling the abundance of individual secretome proteins, but more-systematic studies of how the secretome as a whole is regulated are still scarce.
ACKNOWLEDGMENTS
We thank Alice Sorgo and Clemens Heilmann for their stimulating involvement in the research presented here.
REFERENCESMoranG, ColemanD, SullivanD2012An introduction to the medically important Candida species, p 11–25. InCalderoneRA, ClancyCJ (ed), , 2nd edASM Press, Washington, DC.BonhommeJ, d'EnfertC2013Candida albicans biofilms: building a heterogeneous, drug-tolerant environment.
16:398–403. doi:10.1016/j.mib.2013.03.007.23566895FinkelJS, MitchellAP2012Biofilm formation in Candida albicans, p 299–315. InCalderoneRA, ClancyCJ (ed), , 2nd edASM Press, Washington, DC.ChaffinWL2008Candida albicans cell wall proteins.
72:495–544. doi:10.1128/MMBR.00032-07.18772287KlisFM, SosinskaGJ, de GrootPW, BrulS2009Covalently linked cell wall proteins of Candida albicans and their role in fitness and virulence.
9:1013–1028. doi:10.1111/j.1567-1364.2009.00541.x.19624749SorgoAG, HeilmannCJ, BrulS, de KosterCG, KlisFM2013Beyond the wall: Candida albicans secret(e)s to survive.
338:10–17. doi:10.1111/1574-6968.12049.23170918VialasV, SunZ, Loureiro y PenhaCV, CarrascalM, AbiánJ, MonteolivaL, DeutschEW, AebersoldR, MoritzRL, GilC2014A Candida albicans PeptideAtlas.
97:62–68. doi:10.1016/j.jprot.2013.06.020.23811049InglisDO, ArnaudMB, BinkleyJ, ShahP, SkrzypekMS, WymoreF, BinkleyG, MiyasatoSR, SimisonM, SherlockG2012The Candida genome database incorporates multiple Candida species: multispecies search and analysis tools with curated gene and protein information for Candida albicans and Candida glabrata.
40:D667–D674. doi:10.1093/nar/gkr945.22064862EneIV, HeilmannCJ, SorgoAG, WalkerLA, de KosterCG, MunroCA, KlisFM, BrownAJ2012Carbon source-induced reprogramming of the cell wall proteome and secretome modulates the adherence and drug resistance of the fungal pathogen Candida albicans.
12:3164–3179. doi:10.1002/pmic.201200228.22997008Gil-BonaA, Llama-PalaciosA, ParraCM, VivancoF, NombelaC, MonteolivaL, GilC2015Proteomics unravels extracellular vesicles as carriers of classical cytoplasmic proteins in Candida albicans.
14:142–153. doi:10.1021/pr5007944.25367658RöhmM, LindemannE, HillerE, ErmertD, LemuthK, TrkuljaD, SogukpinarO, BrunnerH, RuppS, UrbanCF, SohnK2013A family of secreted pathogenesis-related proteins in Candida albicans.
87:132–1351. doi:10.1111/mmi.12087.23136884SorgoAG, HeilmannCJ, DekkerHL, BrulS, de KosterCG, KlisFM2010Mass spectrometric analysis of the secretome of Candida albicans.
27:661–672. doi:10.1002/yea.1775.20641015SorgoAG, HeilmannCJ, DekkerHL, BekkerM, BrulS, de KosterCG, de KoningLJ, KlisFM2011Effects of fluconazole on the secretome, the wall proteome, and wall integrity of the clinical fungus Candida albicans.
10:1071–1081. doi:10.1128/EC.05011-11.21622905HeilmannCJ, SorgoAG, MohammadiS, SosinskaGJ, de KosterCG, BrulS, de KoningLJ, KlisFM2013Surface stress induces a conserved cell wall stress response in the pathogenic fungus Candida albicans.
12:254–264. doi:10.1128/EC.00278-12.23243062MaddiA, BowmanSM, FreeSJ2009Trifluoromethanesulfonic acid-based proteomic analysis of cell wall and secreted proteins of the ascomycetous fungi Neurospora crassa and Candida albicans.
46:768–781. doi:10.1016/j.fgb.2009.06.005.19555771LombardV, Golaconda RamuluH, DrulaE, CoutinhoPM, HenrissatB2014The carbohydrate-active enzymes database (CAZy) in 2013.
42:D490–D495. doi:10.1093/nar/gkt1178.24270786CôteP, HoguesH, WhitewayM2009Transcriptional analysis of the Candida albicans cell cycle.
20:3363–3373. doi:10.1091/mbc.E09-03-0210.19477921MulhernSM, LogueME, ButlerG2006Candida albicans transcription factor Ace2 regulates metabolism and is required for filamentation in hypoxic conditions.
5:2001–2013. doi:10.1128/EC.00155-06.16998073KellyMT, MacCallumDM, ClancySD, OddsFC, BrownAJ, ButlerG2004The Candida albicansCaACE2 gene affects morphogenesis, adherence and virulence.
53:969–983. doi:10.1111/j.1365-2958.2004.04185.x.15255906EstebanPF, RíosI, GarcíaR, DueñasE, PláJ, SánchezM, de AldanaCR, Del ReyF2005Characterization of the CaENG1 gene encoding an endo-1,3-beta-glucanase involved in cell separation in Candida albicans.
51:385–392. doi:10.1007/s00284-005-0066-2.16328626JoshiKR, WheelerEE, GavinJB1973Scanning electron microscopy of colonies of six species of Candida.
115:341–348.4197906DouglasLJ2003Candida biofilms and their role in infection.
11:30–36. doi:10.1016/S0966-842X(02)00002-1.12526852HarriottMM, LillyEA, RodriguezTE, FidelPLJr, NoverrMC2010Candida albicans forms biofilms on the vaginal mucosa.
156:3635–3644. doi:10.1099/mic.0.039354-0.20705667TaffHT, NettJE, ZarnowskiR, RossKM, SanchezH, CainMT, HamakerJ, MitchellAP, AndesDR2012A Candida biofilm-induced pathway for matrix glucan delivery: implications for drug resistance.
8:e1002848. doi:10.1371/journal.ppat.1002848.22876186SturtevantJ, DixonF, WadsworthE, LatgeJP, ZhaoXJ, CalderoneR1999Identification and cloning of GCA1, a gene that encodes a cell surface glucoamylase from Candida albicans.
37:357–366. doi:10.1046/j.1365-280X.1999.00244.x.10520161DennersteinGJ, EllisDH2001Oestrogen, glycogen and vaginal candidiasis.
41:326–328. doi:10.1111/j.1479-828X.2001.tb01238.x.11592551ShepherdMG, SullivanPA1976The production and growth characteristics of yeast and mycelial forms of Candida albicans in continuous culture.
93:361–370. doi:10.1099/00221287-93-2-361.6622NobileCJ, NettJE, HerndayAD, HomannOR, DeneaultJS, NantelA, AndesDR, JohnsonAD, MitchellAP2009Biofilm matrix regulation by Candida albicans Zap1.
7:e1000133. doi:10.1371/journal.pbio.1000133.19529758NiimiK, NiimiM, ShepherdMG, CannonRD1997Regulation of N-acetylglucosaminidase production in Candida albicans.
168:464–472. doi:10.1007/s002030050523.9385137HeilmannCJ, SorgoAG, SiliakusAR, DekkerHL, BrulS, de KosterCG, de KoningLJ, KlisFM2011Hyphal induction in the human fungal pathogen Candida albicans reveals a characteristic wall protein profile.
157:2297–2307. doi:10.1099/mic.0.049395-0.21602216NaseemS, ArayaE, KonopkaJB2015Hyphal growth in Candida albicans does not require induction of hyphal-specific gene expression.
26:1174–1187. doi:10.1091/mbc.E14-08-1312.25609092NaseemS, KonopkaJB2015N-Acetylglucosamine regulates virulence properties in microbial pathogens.
11:e1004947. doi:10.1371/journal.ppat.1004947.26226264ShepherdMG, PoulterRT, SullivanPA1985Candida albicans: biology, genetics, and pathogenicity.
39:579–614. doi:10.1146/annurev.mi.39.100185.003051.3904613SchallerM, BorelliC, KortingHC, HubeB2005Hydrolytic enzymes as virulence factors of Candida albicans.
48:365–377. doi:10.1111/j.1439-0507.2005.01165.x.16262871SchallerM, BeinM, KortingHC, BaurS, HammG, MonodM, BeinhauerS, HubeB2003The secreted aspartyl proteinases Sap1 and Sap2 cause tissue damage in an in vitro model of vaginal candidiasis based on reconstituted human vaginal epithelium.
71:3227–3234. doi:10.1128/IAI.71.6.3227-3234.2003.12761103GroppK, SchildL, SchindlerS, HubeB, ZipfelPF, SkerkaC2009The yeast Candida albicans evades human complement attack by secretion of aspartic proteases.
47:465–475. doi:10.1016/j.molimm.2009.08.019.19880183SosinskaGJ, de KoningLJ, de GrootPW, MandersEM, DekkerHL, HellingwerfKJ, de KosterCG, KlisFM2011Mass spectrometric quantification of the adaptations in the wall proteome of Candida albicans in response to ambient pH.
157:136–146. doi:10.1099/mic.0.044206-0.20864472AlbrechtA, FelkA, PichovaI, NaglikJR, SchallerM, de GrootP, MaccallumD, OddsFC, SchäferW, KlisF, MonodM, HubeB2006Glycosylphosphatidylinositol-anchored proteases of Candida albicans target proteins necessary for both cellular processes and host-pathogen interactions.
281:688–694. doi:10.1074/jbc.M509297200.16269404CitiuloF, JacobsenID, MiramónP, SchildL, BrunkeS, ZipfelP, BrockM, HubeB, WilsonD2012Candida albicans scavenges host zinc via Pra1 during endothelial invasion.
8:e1002777. doi:10.1371/journal.ppat.1002777.22761575SorgoAG, BrulS, de KosterCG, de KoningLJ, KlisFM2013Iron restriction-induced adaptations in the wall proteome of Candida albicans.
159:1673–1682. doi:10.1099/mic.0.065599-0.23728625KuznetsG, VigonskyE, WeissmanZ, LalliD, GildorT, KauffmanSJ, TuranoP, BeckerJ, LewinsonO, KornitzerD2014A relay network of extracellular heme-binding proteins drives C. albicans iron acquisition from hemoglobin.
10:e1004407. doi:10.1371/journal.ppat.1004407.25275454LuoS, BlomAM, RuppS, HiplerUC, HubeB, SkerkaC, ZipfelPF2011The pH-regulated antigen 1 of Candida albicans binds the human complement inhibitor C4b-binding protein and mediates fungal complement evasion.
286:8021–8029. doi:10.1074/jbc.M110.130138.21212281ZipfelPF, SkerkaC, KupkaD, LuoS2011Immune escape of the human facultative pathogenic yeast Candida albicans: the many faces of the Candida Pra1 protein.
301:423–430. doi:10.1016/j.ijmm.2011.04.010.21565550RománE, CottierF, ErnstJF, PlaJ2009Msb2 signaling mucin controls activation of Cek1 mitogen-activated protein kinase in Candida albicans.
8:1235–1249. doi:10.1128/EC.00081-09.19542310Szafranski-SchneiderE, SwidergallM, CottierF, TielkerD, RománE, PlaJ, ErnstJF2012Msb2 shedding protects Candida albicans against antimicrobial peptides.
8:e1002501. doi:10.1371/journal.ppat.1002501.22319443SwidergallM, ErnstAM, ErnstJF2013Candida albicans mucin Msb2 is a broad-range protectant against antimicrobial peptides.
57:3917–3922. doi:10.1128/AAC.00862-13.23733470SchildL, HeykenA, de GrootPW, HillerE, MockM, de KosterC, HornU, RuppS, HubeB2011Proteolytic cleavage of covalently linked cell wall proteins by Candida albicans Sap9 and Sap10.
10:98–109. doi:10.1128/EC.00210-10.21097664LiuH, SadygovRG, YatesJRIII2004A model for random sampling and estimation of relative protein abundance in shotgun proteomics.
76:4193–4201. doi:10.1021/ac0498563.15253663McIlwainS, MathewsM, BeremanMS, RubelEW, MacCossMJ, NobleWS2012Estimating relative abundances of proteins from shotgun proteomics data.
13:308. doi:10.1186/1471-2105-13-308.23164367KlisFM, de KosterCG, BrulS2014Cell wall-related bionumbers and bioestimates of Saccharomyces cerevisiae and Candida albicans.
13:2–9. doi:10.1128/EC.00250-13.24243791HillerE, HeineS, BrunnerH, RuppS2007Candida albicans Sun41p, a putative glycosidase, is involved in morphogenesis, cell wall biogenesis, and biofilm formation.
6:2056–2065. doi:10.1128/EC.00285-07.17905924FironA, AubertS, IraquiI, GuadagniniS, GoyardS, PrévostMC, JanbonG, d'EnfertC2007The SUN41 and SUN42 genes are essential for cell separation in Candida albicans.
66:1256–1275. doi:10.1111/j.1365-2958.2007.06011.x.18001349SandiniS, StringaroA, AranciaS, ColoneM, MondelloF, MurtasS, GirolamoA, MastrangeloN, De BernardisF2011The MP65 gene is required for cell wall integrity, adherence to epithelial cells and biofilm formation in Candida albicans.
11:106. doi:10.1186/1471-2180-11-106.21575184GasteboisA, AimaniandaV, Bachellier-BassiS, NesseirA, FironA, BeauvaisA, SchmittC, EnglandP, BeauR, PrévostMC, d'EnfertC, LatgéJP, MouynaI2013SUN proteins belong to a novel family of β-(1,3)-glucan-modifying enzymes involved in fungal morphogenesis.
288:13387–13396. doi:10.1074/jbc.M112.440172.23508952SteczkiewiczK, KnizewskiL, RychlewskiL, GinalskiK2010TOS1 is circularly permuted 1,3-beta-glucanase.
9:201–204. doi:10.4161/cc.9.1.10510.20016268SudberyP, GowN, BermanJ2004The distinct morphogenic states of Candida albicans.
12:317–324. doi:10.1016/j.tim.2004.05.008.15223059ThompsonDS, CarlislePL, KadoshD2011Coevolution of morphology and virulence in Candida species.
10:1173–1182. doi:10.1128/EC.05085-11.21764907HameedS, PrasadT, BanerjeeD, ChandraA, MukhopadhyayCK, GoswamiSK, LattifAA, ChandraJ, MukherjeePK, GhannoumMA, PrasadR2008Iron deprivation induces EFG1-mediated hyphal development in Candida albicans without affecting biofilm formation.
8:744–755. doi:10.1111/j.1567-1364.2008.00394.x.18547332ShepherdMG, SullivanPA1984The control of morphogenesis in Candida albicans.
63:435–440. doi:10.1177/00220345840630031501.6366000LaurentTC, FraserJR1992Hyaluronan.
6:2397–2404.1563592HubeB, MonodM, SchofieldDA, BrownAJ, GowNA1994Expression of seven members of the gene family encoding secretory aspartyl proteinases in Candida albicans.
14:87–99. doi:10.1111/j.1365-2958.1994.tb01269.x.7830564NobileCJ, SchneiderHA, NettJE, SheppardDC, FillerSG, AndesDR, MitchellAP2008Complementary adhesin function in C. albicans biofilm formation.
18:1017–1024. doi:10.1016/j.cub.2008.06.034.18635358BamfordCV, NobbsAH, BarbourME, LamontRJ, JenkinsonHF2015Functional regions of Candida albicans hyphal cell wall protein Als3 that determine interaction with the oral bacterium Streptococcus gordonii.
161:18–29. doi:10.1099/mic.0.083378-0.25332379SilvermanRJ, NobbsAH, VickermanMM, BarbourME, JenkinsonHF2010Interaction of Candida albicans cell wall Als3 protein with Streptococcus gordonii SspB adhesin promotes development of mixed-species communities.
78:4644–4652. doi:10.1128/IAI.00685-10.20805332NaglikJR, ChallacombeSJ, HubeB2003Candida albicans secreted aspartyl proteinases in virulence and pathogenesis.
67:400–428. doi:10.1128/MMBR.67.3.400-428.2003.12966142AbeF, UsuiK, HirakiT2009Fluconazole modulates membrane rigidity, heterogeneity, and water penetration into the plasma membrane in Saccharomyces cerevisiae.
48:8494–8504. doi:10.1021/bi900578y.19670905MunroCA, BatesS, BuurmanET, HughesHB, MaccallumDM, BertramG, AtrihA, FergusonMA, BainJM, BrandA, HamiltonS, WestwaterC, ThomsonLM, BrownAJ, OddsFC, GowNA2005Mnt1p and Mnt2p ofCandida albicans are partially redundant alpha-1,2-mannosyltransferases that participate in O-linked mannosylation and are required for adhesion and virulence.
280:1051–1060. doi:10.1074/jbc.M411413200.15519997BatesS, HughesHB, MunroCA, ThomasWP, MacCallumDM, BertramG, AtrihA, FergusonMA, BrownAJ, OddsFC, GowNA2006Outer chain N-glycans are required for cell wall integrity and virulence of Candida albicans.
281:90–98. doi:10.1074/jbc.M510360200.16263704SouthardSB, SpechtCA, MishraC, Chen-WeinerJ, RobbinsPW1999Molecular analysis of the Candida albicans homolog of Saccharomyces cerevisiaeMNN9, required for glycosylation of cell wall mannoproteins.
181:7439–7448.10601199TimpelC, Strahl-BolsingerS, ZiegelbauerK, ErnstJF1998Multiple functions of Pmt1p-mediated protein O-mannosylation in the fungal pathogen Candida albicans.
273:20837–20846. doi:10.1074/jbc.273.33.20837.9694829SarthyAV, McGonigalT, CoenM, FrostDJ, MeulbroekJA, GoldmanRC1997Phenotype in Candida albicans of a disruption of the BGL2 gene encoding a 1,3-beta-glucosyltransferase.
143:367–376. doi:10.1099/00221287-143-2-367.9043114Martinez-LopezR, MonteolivaL, Diez-OrejasR, NombelaC, GilC2004The GPI-anchored protein CaEcm33p is required for cell wall integrity, morphogenesis and virulence in Candida albicans.
150:3341–3354. doi:10.1099/mic.0.27320-0.15470113MartínezAI, CastilloL, GarceráA, ElorzaMV, ValentínE, SentandreuR2004Role of Pir1 in the construction of the Candida albicans cell wall.
150:3151–3161. doi:10.1099/mic.0.27220-0.15470096KanekoY, OhnoH, KohnoS, MiyazakiY2010Micafungin alters the expression of genes related to cell wall integrity in Candida albicans biofilms.
63:355–357.20859005O'MearaTR, VeriAO, KetelaT, JiangB, RoemerT, CowenLE2015Global analysis of fungal morphology exposes mechanisms of host cell escape.
6:6741. doi:10.1038/ncomms7741.25824284WangA, RanigaPP, LaneS, LuY, LiuH2009Hyphal chain formation in Candida albicans: Cdc28-Hgc1 phosphorylation of Efg1 represses cell separation genes.
29:4406–4416. doi:10.1128/MCB.01502-08.19528234LiuTT, LeeRE, BarkerKS, LeeRE, WeiL, HomayouniR, RogersPD2005Genome wide expression profiling of the response to azole, polyene, echinocandin, and pyrimidine antifungal agents in Candida albicans.
49:2226–2236. doi:10.1128/AAC.49.6.2226-2236.2005.15917516De BernardisF, MühlschlegelFA, CassoneA, FonziWA1998The pH of the host niche controls gene expression in and virulence of Candida albicans.
66:3317–3325.9632601MouynaI, FontaineT, VaiM, MonodM, FonziWA, DiaquinM, PopoloL, HartlandRP, LatgéJP2000Glycosylphosphatidylinositol-anchored glucanosyltransferases play an active role in the biosynthesis of the fungal cell wall.
275:14882–14889. doi:10.1074/jbc.275.20.14882.10809732ZlotogorskiA1987Distribution of skin surface pH on the forehead and cheek of adults.
279:398–401. doi:10.1007/BF00412626.3674963LinharesIM, SummersPR, LarsenB, GiraldoPC, WitkinSS2011Contemporary perspectives on vaginal pH and lactobacilli.
204:120.e1–5. doi:10.1016/j.ajog.2010.07.010.20832044MetwalliKH, KhanSA, KromBP, Jabra-RizkMA2013Streptococcus mutans, Candida albicans, and the human mouth: a sticky situation.
9:e1003616. doi:10.1371/journal.ppat.1003616.24146611EneIV, ChengSC, NeteaMG, BrownAJ2013Growth of Candida albicans cells on the physiologically relevant carbon source lactate affects their recognition and phagocytosis by immune cells.
81:238–248. doi:10.1128/IAI.01092-12.23115042EneIV, AdyaAK, WehmeierS, BrandAC, MacCallumDM, GowNA, BrownAJ2012Host carbon sources modulate cell wall architecture, drug resistance and virulence in a fungal pathogen.
14:1319–1335. doi:10.1111/j.1462-5822.2012.01813.x.22587014SanglardD, HubeB, MonodM, OddsFC, GowNA1997A triple deletion of the secreted aspartyl proteinase genes SAP4, SAP5, and SAP6 of Candida albicans causes attenuated virulence.
65:3539–3546.9284117LiuY, FillerSG2011Candida albicans Als3, a multifunctional adhesin and invasin.
10:168–173. doi:10.1128/EC.00279-10.21115738IbrahimAS, LuoG, GebremariamT, LeeH, SchmidtCS, HennesseyJPJr, FrenchSW, YeamanMR, FillerSG, EdwardsJEJr2013NDV-3 protects mice from vulvovaginal candidiasis through T- and B-cell immune response.
31:5549–5556. doi:10.1016/j.vaccine.2013.09.016.24063977SandiniS, LaValleR, DeaglioS, MalavasiF, CassoneA, DeBernardisF2011A highly immunogenic recombinant and truncated protein of the secreted aspartic proteases family (rSap2t) of Candida albicans as a mucosal anticandidal vaccine.
62:215–224. doi:10.1111/j.1574-695X.2011.00802.x.21535228ColemanDA, OhSH, ZhaoX, ZhaoH, HutchinsJT, VernachioJH, PattiJM, HoyerLL2009Monoclonal antibodies specific for Candida albicans Als3 that immunolabel fungal cells in vitro and in vivo and block adhesion to host surfaces.
78:71–78. doi:10.1016/j.mimet.2009.05.002.19427882AlmeidaRS, BrunkeS, AlbrechtA, ThewesS, LaueM, EdwardsJE, FillerSG, HubeB2008The hyphal-associated adhesin and invasin Als3 of Candida albicans mediates iron acquisition from host ferritin.
4:e1000217. doi:10.1371/journal.ppat.1000217.19023418CassoneA2013Development of vaccines for Candida albicans: fighting a skilled transformer.
11:884–891. doi:10.1038/nrmicro3156.24232568KlisFM, de KosterCG, BrulS2011A mass spectrometric view of the fungal wall proteome.
6:941–651. doi:10.2217/fmb.11.72.21861624