Release from Quorum-Sensing Molecules Triggers Hyphal Formation during Candida albicans Resumption of Growth

Candida albicans is a pathogenic fungus able to change morphologyin response to variations in its growth environment. Simpleinoculation of stationary cells into fresh medium at 37°C,without any other manipulations, appears to be a powerful buttransient inducer of hyphal formation; this process also playsa significant role in classical serum induction of hyphal formation.The mechanism appears to involve the release of hyphal repressioncaused by quorum-sensing molecules in the growth medium of stationary-phasecells, and farnesol has a strong but incomplete role in thisprocess. We used DNA microarray technology to study both theresumption of growth of Candida albicans cells and molecularregulation involving farnesol. Maintaining farnesol in the culturemedium during the resumption of growth both delays and reducesthe induction of hypha-related genes yet triggers expressionof genes encoding drug efflux components. The persistence offarnesol also prevents the repression of histone genes duringhyphal growth and affects the expression of putative or demonstratedmorphogenesis-regulating cyclin genes, such as HGC1, CLN3, andPCL2. The results suggest a model explaining the triggeringof hyphae in the host based on quorum-sensing molecules.

INTRODUCTION

Top

Introduction

Candida albicans is an opportunistic fungal pathogen found inthe normal gastrointestinal flora of most healthy humans. Itcan cause disorders ranging from localized infections to death,the latter especially in immunocompromised patients. As is thecase with many fungal pathogens, this yeast is able to use differentmorphological forms during host invasion (14). Although thelink between virulence and morphogenesis is circumstantial,the changes in morphology are predicted to increase the efficiencyof dissemination in the host (14). The wide range of morphologiesof C. albicans is one of its distinguishing features: it canexist in three common forms (yeast cells, pseudohyphae, andtrue hyphal cells) as well as in some less common shapes, likechlamydospores or opaque cells (6, 12, 23, 26, 32, 33, 35).The yeast-to-hyphal transition is the best-documented morphologicaltransition. Hyphae appear as microscopic tubes with or withoutbranches; to generate hyphae, budding cells change their growthmode to a continuous apical extension followed by septation.Pseudohyphae are created by unipolar budding, with the cellsremaining attached to the mother cell. Both pseudohyphae andtrue hyphae produce chains of cells. The difference betweenthe two can appear minor; however, it is commonly accepted thattrue hyphae have no constrictions at their septa, and they haveparallel cell walls with branches forming perpendicular to thesewalls (34).

Wide ranges of both chemicals and environmental conditions havebeen identified based on their ability to induce the yeast-to-hyphalswitch (12, 26, 28). Addition of serum appears to be a powerfultrigger with a dependence on low cell density and elevated temperature.Despite the widespread use of this technique to induce hyphae,the inherent mechanism is still poorly understood. Hudson etal. (16) recently suggested that the active role of serum ismainly the result of its nutrient properties (nitrogen and carbon).Another poorly understood method of hyphal induction consistsof inoculating fresh medium with stationary cells at 37°C(4, 5, 22). Despite its simplicity, this technique is not commonlyused (9, 10), although the process may have influenced recentobservations (17).

Quorum-sensing molecules allow bacteria to monitor their growthand to control cell density-dependent phenomena. Similar regulatorysystems have been recently identified in the eukaryotic pathogenCandida albicans. Hornby et al. (15) identified farnesol asa quorum-sensing molecule responsible for inhibiting hyphalformation in stationary-phase cells. As well, Chen et al. (10)characterized tyrosol as a candidate quorum-sensing moleculeinvolved in abolishing the lag phase preceding the resumptionof growth of stationary-phase cells.

In this paper, we reinvestigate the capacity of C. albicanscells to form hyphae when initiating growth from stationaryphase and suggest that classical serum-based induction is significantlydependent on this process. The basis of this process appearsto be a release from the inhibition caused by quorum-sensingmolecules, including farnesol, present in the medium of stationarycells. We performed transcriptional profiling of C. albicansinitiating hyphal development due to this trigger both in thepresence and in the absence of farnesol in the fresh medium.The results provide information about the transcriptional changesinvolved in the resumption of growth, in the hyphal developmentprocess, and in the mechanism of response to farnesol.

MATERIALS AND METHODS

Top

Materials and Methods

Growth medium. Cultures of C. albicans strain SC5314 (13) were grown in bufferedyeast extract-peptone-dextrose (YPD) (2% glucose, 2% Bacto peptone,1% yeast extract)-based medium; the required pH is obtainedby adding the correct proportion of an Na₂HPO₄-NaH₂PO₄ mix toa 100 mM final concentration. The buffering ensures pH stabilitythroughout the experiments. Fetal bovine serum was purchasedfrom Gibco (Grand Island, N.Y.); trans,trans-farnesol was purchasedfrom Sigma-Aldrich (Oakville, Ontario, Canada).

Culture conditions. Cells were cultured in a shaking incubator (180 rpm) at 30°Cor 37°C. Absorbance was read at 600 nm. Serial dilutionswere achieved by lightly inoculating a 4-ml YPD tube from aYPD plate culture no older than 2 weeks. One milliliter of thistube was transferred to a second 4-ml YPD tube; the latter isvortexed and 1 ml taken to inoculate a third tube and so onto generate 10 cultures. After overnight growth, this approachgenerates a range of growth states from early-exponential- tostationary-phase cells.

The cells were counted under a microscope (40x lens), and about200 cells were categorized as budded (less than three joinedcells), true hyphae (no constriction at the septa, walls remainingparallel throughout the hyphae, and branching perpendicularto the cell walls), or pseudohyphal. Several positions per slidewere examined to ensure a representative selection of cellswas characterized. Hyphal lengths were measured with the softwareOpenlab (Improvision).

RNA extraction, mRNA purification, cDNA labeling, and microarray analysis. An overnight culture (optical density at 600 nm [OD₆₀₀] >18) was used to initiate the resumption of growth in a volumeof 300 ml of buffered YPD at an OD₆₀₀ of 0.5. Aliquots (50 ml)were removed and centrifuged before being quick-frozen and storedat –80°C at 0, 10, 30, 60, and 180 min following theinoculation. The experiments were done at 30°C or at 37°C;the experiments at 37°C with farnesol were the same as thoseat 37°C except for the addition of 30 µM farnesolin the medium prior to inoculation. We used slight variationsof the methods and microarrays described by Enjalbert et al.(11); results presented consist of the average of four completelyindependent experiments (two Cy3/Cy5 and two Cy5/Cy3 labelings)except for the 37°C plus farnesol condition (two Cy3/Cy5and one Cy5/Cy3 labeling). For each experiment and replicate,the control corresponds to the time zero of the same replicate.Detailed protocols and full results can be obtained from thesupplementary material (http://candida.bri.nrc.ca/papers/Enjalbert2005/index.cfm).

RESULTS

Top

Results

Dilution of stationary cells is a potent and transient trigger of hyphal induction. We reinvestigated previous results that demonstrated the potentialof inducing hyphae in Candida albicans during release from stationaryphase (4, 5, 21, 22, 31). We confirmed that dilution of cellsgrown to stationary phase at 37°C into fresh YPD triggersinduction of hyphae, generating as much as 90% synchronizedhyphal cells after 3 h (Fig. 1). However, this induction istransient: the hyphae start to produce budded cells after 3h, so that buds rapidly become the major growth form (Fig. 1).The initial hyphal structures remain stable through the 15-hexperimental period. We observed no hyphal formation at 37°Cother than those arising from the initial dilution of the stationary-phasecells.

View larger version (23K):
[in this window]
[in a new window]
FIG. 1. Hyphal triggering by dilution of stationary cells. Cells were grown overnight at 37°C in buffered YPD, pH 7.5, in order to obtain stationary cells (OD₆₀₀ $≥$ 20). This culture was diluted to an OD₆₀₀ of 0.5 in the same medium, same temperature, and same pH. At each hour following the initiation, the cell forms were counted ( ${blacklozenge}$ , hyphae; ${blacksquare}$ , pseudohyphae; •, yeast cells), and the OD₆₀₀ was measured ( ${blacktriangleup}$ ). The figure represents the average of two independent experiments, and the standard deviations of the hyphal points only are shown for clarity.

Connection with serum induction. Classical serum induction consists of diluting stationary cellsat 30°C into fresh medium at 37°C containing serum (24).As this method includes the dilution of stationary-phase cellsand has the same pH and temperature requirements as the precedingprotocol, we investigated the role of stationary-phase conditioningduring serum induction. We measured the capacity of cells toproduce serum-induced hyphae after growth for various timesin fresh YPD at 30°C before the switch to 37°C plusserum. Figure 2 demonstrates that the cells lose most of theircapacity to produce hyphae with increasing lengths of time at30°C. Ultimately, transfer of cells growing exponentiallyfor 24 h to 37°C plus serum triggered hyphal developmentin less than 20% of the cells. No hyphae were observed in theabsence of serum with these long-term exponential-phase cells(data not shown). Thus, the classical serum induction appearslinked to outgrowth from stationary-phase cells even if theserum possesses genuine inducer properties.

View larger version (20K):
[in this window]
[in a new window]
FIG. 2. Serum induction of stationary cells with increased time of growth resumption at 30°C. Cells were grown overnight at 30°C in YPD in order to obtain stationary cells (OD₆₀₀ $≥$ 20). These cultures were diluted to an OD₆₀₀ of 0.01 in the same medium, same temperature, and same pH. At each hour from 0 to 6 at 30°C, 10% of final serum was added to a sample, and the tube was transferred at 37°C for 3 h. The 24-h time point corresponds to an overnight exponentially growing culture at 30°C with a final OD₆₀₀ of 0.01 treated the same way. Then, the cell forms were counted ( ${blacklozenge}$ , hyphae; ${blacksquare}$ , pseudohyphae; •, yeast cells). These results are from three independent experiments.

Acquisition of hyphal competence increases with entry into stationary phase. We investigated at which stage of growth the cells become competentto produce hyphae (Fig. 3). Diluting different samples at differentphases of growth of a culture results in a range of morphologicalbehaviors. As seen in Fig. 1, only cells that have exited theexponential phase of growth can produce hyphae after the reinoculation.The closer the cells are to stationary phase, the higher isthe percentage of hyphae 3 h after the dilution. Finally, completelystationary-phase samples are the only ones that generate anextensive production of true hyphae.

View larger version (21K):
[in this window]
[in a new window]
FIG. 3. Influence of the growth phase on hyphal formation after dilution in fresh medium. Cells were grown overnight in serial dilution at 37°C in buffered YPD, pH 7.5, in order to obtain a range of concentrations. These cultures were diluted to an OD₆₀₀ of 0.01 in fresh medium at the same temperature and same pH. At 3 h following the initiation, the cell forms were counted. The data are the results from three independent experiments, and the standard deviations of the hyphae percentages only are shown for the clarity of the graph. ${blacklozenge}$ , hyphae; ${blacksquare}$ , pseudohyphae; •, yeast cells.

Identification of the inherent mechanism. There could be several explanations for a requirement of stationary-phasecells to permit the induction of hyphae: (i) the release ofthe frequent cellular contacts found in an environment withhigh cell concentration, (ii) the sudden availability of nutrients,(iii) the exit from stationary phase, or (iv) the release frominhibitory molecules. To distinguish among these hypotheses,we transferred stationary-phase cells at high concentration(OD₆₀₀ = 20) into fresh medium (inoculation to OD₆₀₀ = 1 to20) or into water, always at 37°C (Fig. 4a). All of theseconditions triggered germ tube formation, and the germ tubelength was proportional to the cell concentration (Fig. 4b).The initiation of hyphal development after transfer of the cellsto the fresh medium without dilution (thus maintaining the sameOD) suggests that cellular contacts are not preventing hyphalformation. Moreover, the induction of hyphae after the transferto distilled water makes it unlikely that hyphal induction requiresthe nutrients available in the new medium. This result correlateswith and supports the observation of Bell and Chaffin (4) thatestablished that 54% of stationary cells diluted in a minimalmedium without a carbon source were able to form germ tubes.Therefore, a plausible explanation for the induction of hyphaeduring the resumption of growth of stationary-phase cells isa release from the repression generated by inhibitory moleculespresent in the conditioned medium.

View larger version (51K):
[in this window]
[in a new window]
FIG. 4. Dilution of stationary-phase cells at different final concentrations. Cells were grown overnight at 37°C in buffered YPD, pH 7.5, in order to obtain stationary cells (OD₆₀₀ = 20). The cells were centrifuged and resuspended in fresh medium at the same temperature and same pH at an OD₆₀₀ of 1 (1), an OD₆₀₀ of 5 (2), an OD₆₀₀ of 10 (3), an OD₆₀₀ of 15 (4), or an OD₆₀₀ of 20 (5) or in water at an OD₆₀₀ of 0.5 (6). After 3 h, a sample of cells was photographed (a). Hyphal lengths were measured, and the averages and standard deviations are reported in the graph (b).

Release from quorum-sensing molecules triggers hyphal formation. We studied the impact of the initial medium on hyphal formationusing serial dilutions: we diluted stationary-phase cells fromthe first tube into the conditioned media of the other cultures.As shown in Fig. 5a, there is a clear step between an OD₆₀₀of 5 and an OD₆₀₀ of 10 where the medium becomes effective atinhibiting hyphal development. This observation suggests thepresence of inhibitory quorum-sensing molecules released inthe spent medium by stationary cells. Of the two quorum-sensingmolecules described so far for C. albicans, only farnesol inhibitshyphal formation (10, 15). A 30 µM concentration of thismolecule has been reported to inhibit, by 50%, the hyphal formationtriggered by serum induction (15). We also found that additionof 30 µM farnesol to the new medium prior to inoculationinhibited hyphal formation by 61% ± 18% (Fig. 5b). Unlikethe spent medium, increasing concentrations of farnesol werenot able to completely suppress hyphal formation. In addition,while the spent medium blocked cells in the yeast form, farnesoltreatment generated a mix of yeast, pseudohyphal, and hyphalcells. Finally, we found that conditioning exponential-phasecells with farnesol followed by farnesol removal was not sufficientto trigger hyphal formation (data not shown). The inabilityof farnesol addition to completely mimic the properties of spentmedium suggests the existence of other inhibitor molecules.

View larger version (25K):
[in this window]
[in a new window]
FIG. 5. Release from quorum-sensing molecules triggers hyphal formation. Cells were grown overnight at 37°C in buffered YPD, pH 7.5, in serial dilution in order to obtain a range of cultures from early-exponential- to stationary-phase cells (OD₆₀₀ $≥$ 20). All of the samples but the more grown were centrifuged, and their medium was sterilized by filtration. Then, the more grown culture was used to inoculate the range of spent media to an OD₆₀₀ of 0.5 at the same temperature (a) or diluted to an OD₆₀₀ of 0.5 in fresh medium at the same temperature plus a range of farnesol concentrations (b). After 3 h, the cell forms were counted. The data are the results from at least three independent experiments, and the standard deviations of the hyphal percentages only are shown for the clarity of the graph. ${blacklozenge}$ , hyphae; ${blacksquare}$ , pseudohyphae; •, yeast cells.

Microarray analysis of the resumption of growth. We monitored the transcriptional reprogramming caused by theresumption of growth of stationary cells transferred to freshmedium at 37°C in the presence of farnesol. We took advantageof the similarity of concentration between the maximum impactof farnesol on hyphal growth (30 µM) and the in vitromaximum accumulation of farnesol during stationary phase (10to 50 µM [15]). We hypothesized that applying 30 µMfarnesol in the new medium would likely result in a minimalchange of farnesol concentration from that of the spent medium.As the cells leave stationary phase and resume growth, transcriptionalprofiling highlights the induction of ribosomal protein genesand RNA production components as well as the repression of stationary-phasegenes, like the stress-related genes (supplementary web data).To bypass the complexity caused by the metabolic changes andto identify farnesol-dependent components, we compared the responsesof cells resuming growth at 37°C in the presence or absenceof farnesol at different times (Fig. 6, top panel). We examinedwhether the same patterns of expression were obtained when wecompared the hyphal form resumption of growth at 37°C tothe yeast form resumption of growth at 30°C. We hypothesizedthat the presence of genes in the two comparisons would be theconsequence of the partial hyphal deficiency created by farnesol.This pattern was observed for the functional groups relatedto biogenesis, to organization and properties of the cell wall,and to ergosterol metabolism as well as the groups describedas "chromatin assembly and disassembly" and "DNA replication."Beside the morphological consequences, the comparison exposedthe gene sets whose expression was modulated by farnesol independentlyof morphology. The quorum-sensing molecule decreases the expressionof some fatty acid oxidation genes after 30 min and createscell cycle changes (from G₂/M transition at 60 min to cell separationat 180 min). The continued presence of farnesol appears to generatea greater expression of genes implicated in the response tostress, as well as to drug and DNA damage stimuli. We observedalso an increase in the expression of genes involved in variousmetabolic processes, such as alcohol metabolism, iron ion transport,and lipid metabolism. Recently, Cao et al. (8) performed a genomicinvestigation of farnesol addition to Candida albicans biofilms.Despite the use of a dissimilar experimental model and a semicompletegenome with a different annotation, they identified equivalentfunctional categories (i.e., iron transport, cell wall, drugresistance, and cell cycle) but with very few overlaps for specificgenes (see supplementary web data). This enforces the idea thatthe specific presence of farnesol, rather than variations inconcentration, is directly related to the expression of thesegenes. Maintaining a similar concentration of farnesol throughoutthe experiment leads to the expected variations based on themorphology of the cells but also causes specific adaptations.

View larger version (55K):
[in this window]
[in a new window]
FIG. 6. Classification of the functional groups with divergent responses. For each time point ratio (time 10, 30, 60, or 180 min/time zero), the log₂ value at 37°C was subtracted from the log₂ at 37°C plus farnesol. Genes with a score greater than 0.8 or lower than –0.8 were selected (both with about 150 genes). The list was refined to keep only genes with an S. cerevisiae homologue in order to use the GO resources (http://www.yeastgenome.org/GOContents.shtml). The lists of function were simplified to remove redundant categories and weak probability groups with no clear significance in the C. albicans response. The upper part of the table indicates the functional categories overexpressed at 37°C without farnesol, while the lower part indicates the functional categories with an increased expression in the presence of farnesol. The finding of functional categories at a time point is represented by a square. The same analysis has been done with the 37°C experiment versus the 30°C study in order to distinguish between the consequences of hyphal growth and the consequences of continued farnesol treatment. Functional categories found in both analyses are represented by a gray square (), while functional categories specific to the presence of farnesol are represented by black squares ( ${blacksquare}$ ). The best probability for each group occurrence for the 37°C plus farnesol versus 37°C analysis is presented on the right. Illustrations of the predominant growth forms for each time point of the 37°C and 37°C plus farnesol experiment are displayed on top of the figure.

Expression profiles of farnesol-responsive genes. In order to unravel the quorum-sensing-dependent mechanismsgoverning the choice of the growth form, we focused on the functionalcategories identified after 10 min of growth. We reasoned thattranscriptional variations observed after 10 min imply thatthe cells have already sensed the new conditions and selectedtheir next growth mode. Figure 7 presents the responses of geneswith farnesol-modulated early variation during the resumptionof growth at 37°C with farnesol, at 30°C, and at 37°C.Farnesol has a major and expected impact on the hypha-relatedgenes (Fig. 7A). We observed a delay in the induction of HWP1,ECE1, and RBT1 as well as a reduction in their level of induction.The mix of cellular forms in the presence of farnesol (Fig.4b) does not explain the reduction of the level of inductionas simply an intermediary result between the extremes of 30°C(yeast cells) and 37°C (hyphal cells), because farnesolresults in a total absence of induction for HWP1, ECE1, andRBT1 10 min after the inoculation, and PHR1 remains noninducedthrough the whole time course. One hypothesis could be a directimpact of farnesol on these genes, with a partial transcriptionalinhibition impairing hyphal growth. However, none of these hypha-relatedgenes have yet been demonstrated as essential for hyphal development(7). So far, the only hypha-specific gene needed for hyphalinduction is HGC1, which encodes a G₁ cyclin (36). The earlyinduction of HGC1 during hyphal growth, which is blocked bythe presence of farnesol, supports the importance of this genein the morphogenesis control (Fig. 7B). Other cyclin-relatedgenes present an early specific profile in the presence of farnesol:the G₁/S cyclin gene CLN3 is repressed, while the G₁/S cyclingene PCL2 no longer displays the strong repression observedduring hyphal growth. Similarly, the repression of the histonegenes during hyphal growth is blocked in the presence of farnesolin fresh medium or at 30°C (Fig. 7C). This tendency is amplifiedafter 60 min, with the induction of the histone genes dramaticallyopposed to their repression at 37°C without farnesol.

View larger version (61K):
[in this window]
[in a new window]
FIG. 7. A selection of genes with different profiles due to farnesol. Genes from some of the functional categories described in Fig. 6 at time 10 min were selected, and their profiles in response to the 37°C plus farnesol, 30°C, and 37°C experiments are presented. The data correspond to the ratio between the times 10, 30, 60, or 180 min after the dilution versus time zero. A, cell wall/hyphal genes; B, cell cycle genes; C, chromatin/histone genes; D, ß-oxidation/glyoxylate cycle genes; and E, drug response genes.

The continued presence of farnesol can trigger specific earlyresponses (Fig. 7D and E). The glyoxylate and ß-oxidationpathways are induced by starvation or during macrophage internalization(19). Following exit from stationary phase, we observed a reductionof expression of some of these genes (see supplementary webdata); this repression is enhanced by the presence of farnesolat the earlier time points (Fig. 7D). We also observed specificinduction of the drug response genes CDR1 and CDR2, as wellas the putative aryl-alcohol dehydrogenase YPL88 that is regulatedby the multidrug transcriptional factors Yrm1 and Yrr1 in Saccharomycescerevisiae (20).

DISCUSSION

Top

Discussion

This study investigates the mechanism of hyphal induction triggeredby the resumption of growth of Candida albicans cells held instationary phase. Changing the medium of stationary cells isa strong but transient way to generate hyphae by releasing theinhibition caused by the conditioned medium. This phenomenonrepresents a major component of the process of serum-inducedhyphal development (24). The quorum-sensing molecule farnesolis implicated in this process, but its inability to completelyinhibit hyphal development suggests the possible existence ofother unknown inhibitory molecules. We studied the transcriptionalchanges due to growth resumption in the presence of farnesolin fresh medium at 37°C to mimic a continuous farnesol concentrationthroughout the experiment. We identified the classes of geneswhose expression was influenced by the presence of farnesolthrough comparisons among experiments at 37°C, at 30°C,and at 37°C in presence of farnesol. This strategy allowedus to distinguish between the classes identified because ofthe morphological differences created by the farnesol (celladhesion, cell wall formation, chromatin, DNA replication, andcell cycle) and those due specifically to the continued presenceof farnesol (drug response and fatty acid oxidation).

Several genes of interest that were influenced by farnesol wereidentified. The hypha-specific genes RBT1, ECE1, and HWP1 showdelayed and reduced induction in the presence of farnesol. However,as these genes are not necessary for hyphal formation, theirinhibition likely reflects an effect of the block in hyphaldevelopment rather than a cause. Drug response genes are quicklyand significantly induced by the continued presence of farnesolin the fresh medium, a response similar to that generated byantimycotic drugs (25, 27, 30). This induction of the Candidadrug resistance genes was surprising because 30 µM farnesolshould be similar to the farnesol concentration in the conditionedmedium of stationary cells (15). The appearance of stress responsegenes in a specific response to farnesol may indicate that thepresence of the quorum-sensing molecule in the fresh mediumis creating damage in the cells. However, this farnesol concentrationdoes not reduce growth (15, 29; our unpublished data). Moreover,the glyoxylate and ß-oxidation pathways are not inducedas they are in other stress conditions (19; supplementary webdata) but in fact are repressed (Fig. 7D). We hypothesize thatthis response could reflect a misinterpretation of the situationby the cells, due to the artificial presence of the stationary-relatedquorum-sensing molecule in the new medium.

The presence of cell cycle and histone genes under the influenceof farnesol can be more easily related to the choice of thegrowth form. The set of histone genes is repressed under thehyphal growth conditions and up-regulated at 30°C or inthe presence of farnesol. It is likely that their inductionreflects the growth requirement of the cells. However, thisdoes not explain the need to reduce the amount of histones duringhyphal growth. This suggests other possible functions for histoneregulation, especially in morphogenesis. In mammalian cells,hydroxyurea creates a rapid decrease of the histone mRNA pool(3), while in Candida albicans, hydroxyurea is able to inducefilament formation (1). It is therefore possible that a decreasein the amount of histone message plays a role in hyphal development.The presence of farnesol in the fresh medium affects the expressionprofiles of the cell cycle genes PCL2, HGC1, and CLN3. Thesethree genes possess putative or demonstrated function in morphogenesis.In S. cerevisiae, pcl1/2 family mutants display an abnormalcell morphology upon starvation and random budding in diploids(18). The cyclin HGC1 is the only reported hypha-induced genenecessary for hyphal growth (36). Repression of the cyclin geneCLN3 causes yeast cells to develop into hyphae and pseudohyphae(2). The presence of farnesol prevents the PCL2 repression andthe HGC1 induction and represses CLN3. Therefore, these cyclinsare attractive as potential links between farnesol and morphogenesis.

The triggering of hyphal development through growth resumptionof stationary-phase cells appears to be a major component ofthe standard serum induction process: even if serum brings robustnessto this process, its role in initiation may be modest. Thiscould explain some of the difficulties encountered in understandingthe mechanism of serum induction (16). Dilution of stationarycells provides a clear bias in the interpretation of resultsrelated to the C. albicans morphology switch; true exponentialcells should always be used to measure the influence of moleculeson hyphal growth. This could help resolve discrepancies in theliterature about the effect of molecules that are describedas both inducers and repressors of hyphal formation (26).

There have been many investigations aiming to demonstrate whichof the yeast or hyphal states represents the virulent form ofC. albicans in the host. Odds (26) concluded that both formsare implicated in virulence. One clue is the presence of bothyeast cells and hyphae in the majority of established candidosiscases. The suggestion is that each form possesses its own function:hyphae would carry the yeast infection through the cell walls,and yeast would efficiently spread in the bloodstream. Basedon this work, we propose that the first step of infection byisolated cells would involve yeast form growth. Yeast cellssurrounded by others will be in a high concentration of farnesoland other inhibitors, while cells in contact with the host cellsor otherwise isolated will be relieved of this inhibition. Thus,the latter cells would be able to induce hyphae that would allowthem to adhere better to the host cell walls and to progressthrough in order to carry the infection elsewhere (26). Thishypothesis could explain the presence of the different typesof C. albicans morphologies and the mechanism of their creation.

ACKNOWLEDGMENTS

This work was supported by the Genomics and Health Initiativeof the National Research Council of Canada, by grant CRP004from the British Council/NRC Collaborative Research Program,and by CIHR grant MOP-42516 to M.W. B.E. received an NSERC visitingfellow award through the GHI.

We are grateful to Alistair J. P. Brown and Neil A. R. Gow forcomments, and we thank members of the Genetics group for theirsupport.

FOOTNOTES

* Corresponding author. Mailing address: Aberdeen Fungal Group, School of Medical Sciences, Institute of Medical Sciences, University of Aberdeen, Aberdeen AB25 2ZD, United Kingdom. Phone: (44)-1224-555882. Fax: (44)-1224-555844. E-mail: brice.enjalbert{at}abdn.ac.uk.

This is National Research Council publication number 47468.

REFERENCES

Top