Unique Aspects of Gene Expression during Candida albicans Mating and Possible G₁ Dependency

Abstract

Taking advantage of the high frequency of conjugation tube formation in mating mixtures and α-pheromone-treated a/a cells derived from saturation phase cultures of opaque cells of Candida albicans, 56 up-regulated and 30 down-regulated genes were identified employing microarray and Northern analyses. Combining these results with previous profiling studies of pheromone-induced cells, a more comprehensive transcript profile was developed for comparison with Saccharomyces cerevisiae. This comparison revealed the following: (i) that while a majority of mating-associated genes are regulated similarly between the two species, a significant minority are regulated dissimilarly; (ii) that filamentation genes are uniquely up-regulated and opaque-specific genes uniquely down-regulated during C. albicans mating; and (iii) that a newly identified class of genes is selectively down-regulated in opaque, but not white, cells that have entered saturation phase in a growth culture and then are up-regulated by pheromone. The observations that opaque cells are uniquely mating competent, that saturation phase facilitates mating, and that a newly identified group of genes is down-regulated only in opaque cells that have entered saturation phase led us to hypothesize that entering saturation phase may be requisite for mating. A test of this hypothesis revealed, however, that cells, whether in the exponential or saturation phase, may simply have to be in G₁ of the cell cycle to respond to pheromone and that the response includes G₁ arrest. These results add to the lists of similarities and dissimilarities between the mating processes of C. albicans and S. cerevisiae and underscore the unique regulation of filamentation and switching genes in the C. albicans mating process.

Mating in Candida albicans and Saccharomyces cerevisiae shares several fundamental characteristics, including similar key regulators harbored by the mating-type locus (25), a similar dependency of fusion on mating type (24, 34, 36), similar pheromone-induced shmooing (5, 35, 43), activation of a number of common mating-associated genes (5, 35), similar repression of haploid-specific genes by homologous repressor complexes (Mtla1p-Mtlα2p and Mata1p-Matα2p, respectively) (56), a similar central role for the pheromone-activated mitogen-activated protein (MAP) kinase pathway (11, 37), and similar cytological steps in the fusion process (34). This impressive list of similarities suggests that the intensively studied mating process of S. cerevisiae should serve as a paradigm for understanding C. albicans mating. However, the list of fundamental differences is almost as impressive. First, while the alternative a or α mating genes are expressed at the MAT locus and conserved at two unexpressed loci (HML and HMR) in an elegant cassette system for mating-type switching in S. cerevisiae (20), they are harbored at a single heterozygous locus in diploid C. albicans, which must undergo homozygosis for mating-type expression (25). Second, while a-specific genes in S. cerevisiae are expressed by default, they are positively regulated by Mtla2p in C. albicans (56). Third, while MAT-haploid strains are immediately mating competent in S. cerevisiae, MTL-homozygous strains of C. albicans are not; they must first undergo a complex phenotypic switch from white to opaque (34, 39) that has no counterpart in S. cerevisiae mating and then find mating partners. Furthermore, skin facilitates C. albicans mating (27), which involves the formation of very long, hypha-like conjugation tubes, much longer than those that have been described for S. cerevisiae. Hence, to employ effectively the information collected for the mating process of S. cerevisiae to understand the mating process of C. albicans, we must first distinguish those processes that are similar from those that are dissimilar.

Here we have taken advantage of the high frequency of conjugation tube formation obtained in mating mixtures of opaque a/a and α/α cells or α-pheromone-activated a/a cells derived from saturation phase cultures (16, 34) to identify 56 up-regulated and 30 down-regulated genes, employing an oligonucleotide-based microarray and Northern analysis. Regulated genes were identified that are involved in mating, hypha formation, switching, and a number of other diverse functions. By combining these results with data from earlier transcript profiling studies of a cells treated with α-pheromone (5, 35), a more comprehensive profile is presented of mating-associated genes regulated similarly and dissimilarly in C. albicans and S. cerevisiae. In addition, a number of genes associated with filamentation and switching were found to be uniquely regulated during C. albicans mating. Finally, a unique group of genes was identified that is expressed similarly in white and opaque cells in the exponential phase of a growth culture, selectively down-regulated in opaque, but not white, cells in saturation phase, and then selectively up-regulated by pheromone. The observations that only opaque cells are mating competent, that this new class of genes is selectively down-regulated in saturation phase cultures of opaque but not white cells, and that saturation phase facilitates mating led us to hypothesize that entry into saturation phase may be requisite to pheromone induction and mating. Tests of this hypothesis, however, suggested that opaque cells, whether in the exponential or saturation phase, must simply be in G₁ of the cell cycle to respond to pheromone and that the response includes G₁ arrest.

MATERIALS AND METHODS

Strain maintenance.

The unrelated C. albicans MTL-homozygous strains WO-1 (α/α) (48) and P37005 (a/a) (33) were stored in 20% glycerol at −80°C. For experimental purposes, cells were plated on agar containing Lee's medium (30) modified as described by Bedell and Soll (3) (modified Lee's medium). This agar was supplemented with 5 μg/ml phloxine B to differentially stain opaque-phase colonies and sectors red (2). The S. cerevisiae strains MGD353-13D and MGD353-46D, generous gifts from Brian Rymond at the University of Kentucky, were similarly stored in 20% glycerol at −80°C.

Preparation of samples for microarray and Northern analyses.

Samples from C. albicans mating cultures were prepared according to methods previously described (34) but on a larger scale. In brief, opaque cells from phenotypically homogeneous 7-day colonies of both mating types were individually inoculated into 250 ml of modified Lee's medium and grown at 25°C for 48 h to saturation phase in a water bath rotary shaker at 250 rpm. Cells were then pelleted and resuspended in fresh modified Lee's medium to the original saturation phase cell concentration (3 × 10⁷ per ml). In the case of homogeneous cultures (i.e., either a/a or α/α), 10 ml of a saturation phase culture of either a/a or α/α cells was pelleted and the cells resuspended in 10 ml of fresh growth medium in 50-ml Falcon tubes. In the case of mating mixtures, 5 ml each of stationary-phase cultures of a/a and α/α cells was mixed, pelleted, and resuspended in 10 ml of fresh medium in 50-ml Falcon tubes. Cell cultures were incubated at 25°C in a water bath rotary shaker at 250 rpm. Samples were collected from homogeneous cultures at 3.5 h and from mating mixtures at 3.5 and 7.5 h of incubation. Before samples were collected for microarray analysis, they were examined microscopically for shmooing and fusion. Only mating mixtures at 3.5 h containing over 60% cells with unconstricted evaginations (shmoos) or conjugation tubes and only mating mixtures at 7.5 h containing over 85% shmooed cells, of which 25% had fused, were used for microarray analyses. Three independent preparations of each sample were used for microarray analyses. For preparing pheromone-treated samples, opaque P37005 (a/a) cells from 7-day colonies were inoculated into 250 ml of modified Lee's medium and grown at 25°C for 48 h to stationary phase in a water bath rotary shaker at 250 rpm. Before the cells were treated with α-pheromone, 10-ml portions of cells were pelleted and resuspended in 10 ml of fresh modified Lee's medium. Chemically synthesized α-pheromone was then added to a final concentration of 3 μM (35), and the cell culture was incubated at 25°C in a rotary water bath shaker at 250 rpm.

Microarray analysis.

C. albicans Array-Ready Genome Oligo set version 1.1 (QIAGEN, Valencia, CA) was resuspended at a concentration of 20 μM in 150 mM sodium phosphate buffer (pH 8.5). Oligonucleotides were printed in duplicate blocks on Code Link slides (Amersham Biosciences) at Microarrays, Inc. (Nashville, TN). DNA coupling and postcoupling processing of printed microarray slides were done according to the manufacturer's instructions (Amersham Biosciences, Piscataway, NJ). The methods for labeling probes and hybridization to microarrays were adapted from protocols developed by The Institute for Genomic Research (TIGR, http://pga.tigr.org/protocols.html). Total RNA was prepared using RNeasy (QIAGEN) according to the manufacturer's suggested protocol. Each sample started with a 10-min, 70°C incubation of 10 μg total RNA, Arabidopsis thaliana Spot Report spiking controls (Stratagene, La Jolla, CA), and 2 μl random hexamer primers (3 mg/ml) (Invitrogen, Carlsbad, CA). RNA samples were reverse transcribed with 400 U of Superscript II reverse transcriptase (Invitrogen) in the presence of deoxynucleoside triphosphates (aminoallyl-dUTP and deoxynucleoside triphosphates [4:1]). Samples were purified according to the TIGR protocol (22) and coupled to Cy3 or Cy5 (Amersham Biosciences) in the presence of 0.05 M Na₂CO₃, pH 9.0. cDNA samples were purified postcoupling according to the TIGR protocol. Cy dye-labeled cDNA samples were resuspended in hybridization buffer (10× SSC [1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate] and 0.2% sodium dodecyl sulfate) and applied to microarray slides. Prehybridized slides were blocked in buffer containing 5× SSC, 0.1% sodium dodecyl sulfate, and 1% bovine serum albumin (Sigma, St. Louis, MO) for 45 min at 37°C. Hybridizations were performed for 20 h in a 37°C water bath in the dark. Three independent biological replicates were performed for each hybridization comparison. Slides were washed and dried by centrifugation and immediately scanned using a GenePix 4000B scanner (Axon Instruments, Inc., Union City, CA). Images were processed using the software package GenePix Pro 4.0 (Axon Instruments, Inc.) to gather the raw fluorescence intensities of the Cy3 and Cy5 values of each spot.

Microarray data analysis.

To determine an appropriate normalization method, microarray plots, as described by Dudoit et al. (18), were used initially to analyze the fluorescence (R,G) data. The results of this analysis indicated that a global normalization was appropriate and all raw fluorescence intensities were exponentially transformed prior to fitting the model. The normalization model for this study was y_ijkgr = μ + A_i + D_j + A_dij + e_ijkgr, where y_ijkgr is the signal from spot r for gene g on array i, dye j, and target k; μ is the average signal across all factors; A_i is the global effect of slide i; D_j is the global effect of dye level j; A_dij is the interaction between array and dye; and e_ijkgr is residual error. The residuals from this model, which are regarded as a crude indicator of relative expression level and are referred to as normalized expression levels, were then subjected to gene-specific models of the form: r_ijkr = μ + A_ir + D_j + T_k + e_ijkr, where A_ir is r spot effects, D_k is gene-specific dye k effect, T_k is expression of gene specifically attributable to target k, and e_ijkr is residual error.

The normalization and gene models were fit using PROC MIXED in SAS/STA software version 8 (SAS Institute, Inc., Cary, NC). SAS code for this analysis was based on code available in the Introductory Manual for Mixed Model Analysis of Microarray Data (MANMADA) (2002) (http://statgen.ncsu.edu/ggibson/Manual.htm). In the normalization model, all effects were treated as random except for the fixed effect of dye. To assess the magnitude and significance of the effects of the cell phases and their interaction on the normalized expression level of each sequence, least-square means using the LSMEANS option in SAS (SAS Institute, Inc., 2000) were calculated as described in the MANMADA (2002). The LSMEANS statement calculates the mean value for each target category, adjusted for the other terms in the model. For each pairwise comparison, the difference between the two normalized expression levels and their standard errors are calculated along with the results of t tests (adjusted with Bonferroni's correction) for the statistical significance of the difference between the means in each pair (31). As fluorescence intensities were initially exponentially (base 2) transformed, the resulting difference between expression levels that is observed is interpreted as change (n-fold).

Website containing full transcript datasets.

The full transcript data sets can be found in the Gene Expression Omnibus (GEO) (http://www.ncbi.nlm.nih.gov/geo/) under the GEO accession number GSE1969.

Northern analysis.

Northern analysis was performed as previously described (59). Total RNA from each sample was prepared with the RNeasy kit as recommended by the manufacturer (QIAGEN, Valencia, CA). Five-microgram portions of RNA were separated on a 1.2% agarose-formaldehyde gel, transferred to a Zetabind nylon membrane, and probed with PCR products obtained from PCRs with the primers listed in Table 1. Prehybridization and hybridization were performed by the methods of Church and Gilbert (14). Autoradiography was performed by exposing membrane to Kodak X-OMAT film (Eastman Kodak, Rochester, NY) at −80°C with an intensifying screen. All Northern analyses were performed twice with independent preparations.

TABLE 1.

Primers used in this study

Primer	Sequence
AKL1F	5′ ATGGGTCATCTTCCAAAA 3′
AKL1R	5′ TCTGCTTTCTCGTCTATT 3′
CDC55F	5′ ATAATGCCGTCTGATAAA 3′
CDC55R	5′ TGATTCTCAAATCATCAG 3′
CPH1F	5′ TAACTGCCTAATACACTC 3′
CPH1R	5′ TCTCTGTATCTATTCATC 3′
CZF1F	5′ AACAATGAGTTCAATACC 3′
CZF1R	5′ CAACACAGAGAAGCTTAT 3′
DDR48F	5′ ACAATGGTTTTCGGTTTC 3′
DDR48R	5′ TTAATAGTCAGAAGATCC 3′
ERK2SmaIF	5′ TCCCCCGGGAAGAAATGAAGAAATCTACT 3′
ERK2SmaIR	5′ TCCCCCGGGTTACGACATGACTATTTC 3′
FANEFG15′	5′ GCGTCGCGAATGTCAACGTATTCTATA 3′
FANEFG13′	5′ GCGCCGCGGCTTTTCTTCTTTGGCAACAGTCGT 3′
GCN4F	5′ GATAAAATGCCTGCTACT 3′
GCN4R	5′ AACGAGAGGTATATATAG 3′
HAC1F	5′ ATGGAGTTAACTGTTGAT 3′
HAC1R	5′ TTCTTCTTGGTGTTCATC 3′
NRG1F	5′ AACTACTATATGGGACCA 3′
NRG1R	5′ TGACCACATCTATACTAG 3′
OP4NDEIF	5′ GGCATATGAAGTTTTCACAAGCC 3′
OP4SPHIR	5′ GGTGGTTGCTCTTCCGCACTAATAAAGTTTTCTTTT 3′
ORF6.8100F	5′ AATCATTATGGGTGCTTG 3′
ORF6.8100R	5′ GAACGTTTTATTGTGATG 3′
RCE1F	5′ TCAGATATGAAAGTGATAC 3′
RCE1R	5′ CTACATGTAATACAACAAC 3′
RIM101F	5′ ATGCTATGGGTAACAGTC 3′
RIM101R	5′ GCTTAGAAGCAGTTATAG 3′
SAP1F	5′ TCAATCAATTTACTCTTCCATTTCTAACA 3′
SAP1R	5′ CCAGTAGCATTAACAGGAGTTTTAATGACA 3′
TEC1SmaIF	5′ TCCCCCGGGATTATGATGTCGCAAGCT 3′
TEC1R	5′ TTTGTTCAAACCACTCTT 3′
STE3F	5′ AATGCAGATGTACAAAGGAGCAGCT 3′
STE3R	5′ CCGCTCGAGTCAGTTATCATACGATTT 3′
STE20F	5′ GTCTAACGTATATACGCGTAC 3′
STE20R	5′ CCGCTAGCCGTCTCAGCATTA 3′
FAR1F	5′ ATGTTCGTTGTTGTCAAGA 3′
FAR1R	5′ TCCCGCATCAATAGGTGT 3′

Quantitative fluorescence microscopy of nuclear DNA.

Nuclear DNA content was quantitated using a modification of a protocol by J. A. Huberman and colleagues (http://saturn.roswellpark.org/huberman/Quant_Flu_Microscopy/Quant_Flu_Micro.html).

Cells (2.5 × 10⁷) were washed three times with water prior to fixation in 1 ml of 70% ethanol. After 5-min fixation, the ethanol was removed by centrifugation, and the preparation was rinsed three times in 50 mM sodium citrate (SC) and resuspended in 1 ml SC. RNase A (100 μg/ml) (QIAGEN, Valencia, CA) was added and incubated overnight at 37°C. The cells were harvested by centrifugation, rinsed three times in SC, and resuspended in 33 μl SC. The cells were mixed in a 1:2 ratio with solution containing 37.5 μM Sytox Green (Molecular Probes, Eugene, OR), 1.5 mM p-phenylenediamine, and 75% glycerol. The cells were distributed in the solution by gentle pipetting, and 2 μl of stained cells was placed on a glass slide and covered with an 18-mm square cover glass to ensure a uniform single-cell layer. The edges of the cover glass were sealed with fingernail polish and the slides incubated overnight at room temperature. Quantitation of Sytox Green-labeled DNA fluorescence was performed using a Bio-Rad Radiance 2000MP confocal microscope system on a Nikon TE2000U microscope, with 60× plan apochromat water immersion objective (numerical aperture of 1.2) and 3× digital zoom. LaserSharp 2000 (release 5.2) software was used for image acquisition and fluorescence quantitation. Sytox Green was excited using the 488-nm argon laser line at 3% power, and simultaneous transmitted light differential interference contrast images were generated using the 476-nm argon laser line at 10% power. Acquisition parameters, including laser power, pinhole size, and photomultiplier detector gain and black levels, were set by imaging with unreplicated DNA (see Results and Discussion) from a yeast cell in exponential-phase culture. A nucleus in a mother cell near the bud neck, with replicated DNA (see Results and Discussion) was used to optimize the gray-scale dynamic using the false color setcol.lut lookup table to eliminate pixel saturation and to ensure linear pixel intensity and the full dynamic range from 0 to 255 of the 8-bit photomultiplier detector. Once the parameters were set, the acquisition software remained unchanged for all subsequent image acquisition. A 4-μm z series was collected at 0.2-μm increments at a scan rate of 50 lines per second without frame averaging. The z series was projected as a single stacked image with a maximum voxel intensity value. Using the projected image, a line profile was drawn across the nucleus at its greatest fluorescence intensity to measure pixel intensity in the x-y plane. The line profile data were exported to Microsoft Excel software for analysis. Confocal projection images and differential interference contrast single optical plane images were exported to Adobe Photoshop for presentation.

Cytological analysis of shmooing.

Saturation phase cultures of S. cerevisiae a cells (MGD353-13D) or C. albicans a/a cells (P37005) were resuspended at high density (2 × 10⁷ per ml) in fresh medium containing 10⁻⁶ M synthetic α-pheromone, the 13-mer in the case of S. cerevisiae (Zymo Research Corp., Orange, CA) and the 14-mer in the case of C. albicans (35), and rotated in a water bath shaker at 25°C. Shmooing and conjugation tube growth were monitored microscopically for 17 h.

RESULTS AND DISCUSSION

Microarray analysis of gene expression in mating mixtures.

For microarray analyses, natural opaque a/a cells of strain P37005 and opaque α/α cells of strain WO-1 were individually grown to saturation phase in liquid medium, in which over 90% of both strains accumulated as unbudded cells (Fig. 1A and B). Although opaque P37005 cells (Fig. 1B) were rounder and larger than opaque WO-1 cells (Fig. 1A), both switched to white at the same frequency, regulated the same repertoire of opaque-specific and white-specific genes, and exhibited the same unique pimples on the cell surface and the same large intracellular vacuole in the opaque phase (33, 35) (data not shown). For microarray analyses, cells were incubated under the following conditions: (i) opaque a/a cells alone, (ii) opaque α/α cells alone, and (iii) a mixture of a/a and α/α cells at equal concentrations. After 3.5 h of incubation, the majority of cells in both a/a and in α/α cultures remained unbudded (Fig. 1C and D). In marked contrast, after 3.5 h the majority (>60%) of cells in mixed a/a and α/α cultures had formed shmoos (Fig. 1E). After 7.5 h of incubation, over 80% had shmooed, and approximately 25% of shmooed cells (approximately 20% of the entire cell population) had fused (Fig. 1F). For analysis, cultures of either opaque a/a cells or opaque α/α cells alone were sampled at 3.5 h. Mating mixtures were sampled at 3.5 and 7.5 h. The microarray comparisons included (i) 3.5-h a/a versus 3.5-h a/a-α/α mating mixture, (ii) 3.5-h α/α versus 3.5-h a/a-α/α mating mixture, (iii) 3.5-h a/a versus 7.5-h a/a-α/α mating mixture, and (iv) 3.5-h α/α versus 7.5 h a/a-α/α mating mixture. Each comparison was repeated three times with independent preparations.

FIG. 1.

Description of the preparations and cell phenotypes used in transcript profiling studies of opaque-phase a/a and α/α mating mixtures. The original 0-h preparations were from saturation phase cultures. Opaque (Op) cells of strain P37005 (a/a) were consistently larger than those of strain WO-1 (α/α), but both switched back to white at the same frequency, possessed cell wall pimples, and expressed the same phase-specific genes (data not shown). Small white arrows in panels E and F point to shmoos and fusions, respectively. Cells were resuspended in fresh medium at time zero at the original saturation phase concentration (3 × 10⁷ per ml). Bars in panels A and D, 5 μm (all panels at the same magnification).

A gene was considered up-regulated if one or more of the four comparisons demonstrated an increase of ≥2.0-fold, the remaining comparisons demonstrated increases between 1.1- and 1.9-fold, and the P values adjusted with Bonferroni's correction were significant. Thus, it included genes that peaked at 3.5 h or exhibited maxima at 7.5 h. Fifty genes were identified by these criteria as up-regulated. On the basis of known or deduced functions of the C. albicans genes or the S. cerevisiae orthologs, up-regulated genes included the following genes: 7 encoding proteins involved in mating and/or the cell cycle, 7 encoding membrane proteins, 5 encoding cell wall-related proteins, 9 encoding transcription regulators, 3 encoding cytoskeleton-related proteins, 7 encoding proteins involved in energy metabolism, and 12 encoding proteins with other or unknown functions (Table 2). Seven of the 50 genes had been identified by Bennett et al. (5) as up-regulated in α-pheromone-treated opaque a cells derived from log-phase cultures employing a cDNA microarray (Table 2). Orthologs of 13 of the 50 C. albicans genes identified as up-regulated in mating mixtures had also been demonstrated to be up-regulated in S. cerevisiae a cells treated with α-pheromone (Table 2) (46).

TABLE 2.

Microarray analysis identified 50 genes up-regulated^a during the mating process of C. albicans

ORFb	Genec	Pheromone response ind:		Function(s)e
		S. cerevisiae	C. albicans
Mating and cell cycle related
orf6.4306	MFα1	Induced	Induced	α-Pheromone
orf6.3131	FUS1	Induced	Induced	Surface protein, cell fusion
orf6.2854	ERK2	Induced	Induced	MAP kinase
orf6.6562	PRM1	Induced	Induced	Pheromone-regulated membrane protein
orf6.5090	CDC55	No	No	Cell cycle protein kinase
orf6.7520	RCE1	No	No	a-Pheromone maturation
orf6.8018	MPT5	Induced	No	Cell cycle reentry
Membrane proteins
orf6.7231	RSN1	No	No	Membrane protein
orf6.4763	DSE4	No	No	Daughter cell-specific glucanase
orf6.1470	-	NA	No	Membrane-spanning permease
orf6.1684	MDN1	No	No	Similar to membrane protein YLR106c
orf6.6464	FRP6	Induced	No	Membrane protein, ammonia production
orf6.5714	YPT7	No	No	GTP binding protein, endosome-vacuole fusion
orf6.1292	SNX4	No	No	Nexin-like protein, sorting
Cell wall-related proteins
orf6.4883	HWP1	NA	Induced	Hyphal wall protein
orf6.4889	RBT1	NA	Induced	Repressed by TUP1, similar to MUC1
orf6.1267	EMC9	No	No	Cell wall organization and biogenesis
orf6.5738	MNT2	No	No	N-myristoyltransferase for wall biosynthesis
orf6.1074	FMP45	Induced	No	Cell wall organization and biogenesis
Transcriptional regulator
orf6.5290	GCN4	No	No	Amino acid biosynthetic genes
orf6.7973	RTF1	No	No	DNA binding properties of TBPf
orf6.9111	ARG11	Induced	No	Mitochondrial amino acid biosynthesis
orf6.3349	RMS1	No	No	Transcription regulator
orf6.7972	MET28	Induced	No	Sulfur metabolism
orf6.1395	HAC1	Induced	No	Phospholipid metabolism
orf6.7210	CTA9	NA	No	Transcriptional regulator
orf6.5049	TYE7	Induced	No	Glycolytic gene expression
orf6.7119	SPT5	No	No	Transcription initiation protein
Cytoskeleton related
orf6.8346	AKL1	No	No	Serine/threonine protein kinase, actin cytoskeleton
orf6.6411	ENT3	Induced	No	Recruitment of clathrin, Golgi traffic
orf6.7150	CDC55	No	No	Protein phosphatase, actin organization, pseudohypha growth
Energy and metabolism
AF28526	COX3	NA	No	Cytochrome oxidase subunit 3
orf6.8673	RHR2	No	Induced	dl-Glycerol phosphatase, glycerol biosynthesis, stress response
orf6.6577	IDI1	No	No	Isopentenyl-diphosphate delta-isomerase, ergosterol biosynthesis
orf6.4518	PRO3	No	No	Δ1-Pyrroline-5-carboxylate reductase
orf6.9158	AYR1	No	No	1-Acyl dihydroxyacetone phosphate reductase
orf6.5186	DPS1	No	No	Aspartyl-tRNA synthetase
orf6.6334	OLE1	No	No	Δ-9-Fatty acid desaturase
Other functions
orf6.555	-	NA	No	Similar to RP108 of Rickettsia prowazekii
orf6.1938	CTF5	No	No	Minichromosome maintenance
orf6.6854	DDR48	Induced	No	Stress response
orf6.9095	-	NA	No	C3HC4-type zinc finger protein, translation termination
orf6.1039	-	NA	No	Lipase, peroxisome biogenesis
orf6.2192	CDC91	No	No	GPIg transamidase subunit
Unknown functions
orf6.6035	-	NA	No	Unknown
orf6.8658	-	NA	No	Unknown
orf6.7430	-	NA	No	Unknown
orf6.147	-	NA	No	Unknown
orf6.8877	-	NA	No	Unknown
orf6.7824	-	NA	No	Unknown

^aSee “Microarray analysis of gene expression in mating mixtures” in Results and Discussion for criteria for considering a gene to be up-regulated.

^bORF, open reading frame.

^cNames of C. albicans genes were obtained from the C. albicans genome database (http://www.genolist.pasteur.fr/CandidaDB). In case the genes were not annotated, the names of their S. cerevisiae orthologs were listed. If an unambiguous S. cerevisiae ortholog was not identified from the S. cerevisiae genome database (http://www.yeastgenome.org), the name of the closest ortholog was listed. -, unnamed sequence.

^dPheromone response data of S. cerevisiae were obtained from Roberts et al. (46). Pheromone response data of C. albicans were obtained from Bennett et al. (5). Induced, demonstrated induction; No, demonstrated no induction; NA, not available (not reported or tested).

^eKnown or deduced from the S. cerevisiae ortholog.

^fTBP, TATA binding protein.

^gGPI, glycosylphosphatidylinositol.

A gene was considered to be down-regulated if one or more of the four comparisons demonstrated a decrease of ≥2.0-fold, the remaining comparisons demonstrated decreases between 1.1- and 1.9-fold, and the Bonferroni-adjusted P values were significant. Thus, this definition included genes that exhibited minima at 3.5 or 7.5 h. Thirty genes were identified by these criteria as down-regulated. On the basis of known or deduced function of the C. albicans genes or the S. cerevisiae orthologs, down-regulated genes included 3 genes encoding secreted proteinases and lipases, 3 encoding cell surface proteins, 13 encoding proteins involved in energy and metabolism, and 11 encoding proteins involved in other or unknown functions (Table 3). Bennett et al. (5) reported down-regulation of four genes (MCM6, MCM7, PRI, and POL5a) in opaque a cells treated with α-pheromone, none of which overlapped the 30 down-regulated genes identified in the present study (Table 3). Orthologs of six of the 30 genes identified here as down-regulated (ALG8, SLC1, COX11, PFK2, HEM13, and TRM9) had been demonstrated to be down-regulated in S. cerevisiae a cells treated with α-pheromone; however, homologs of five (FLO9, WSC2, LKH1, TH14, and HSP31) had been demonstrated to be up-regulated by α-pheromone (46) (Table 3). The OP4, SAP1, and SAP3 genes, which are all opaque specific, had been previously demonstrated by Northern analysis to be down-regulated in C. albicans opaque a/a cells treated with α-pheromone (35).

TABLE 3.

Microarray analysis identified 30 genes down-regulated^a during the mating process of C. albicans

ORFb	Genec	Pheromone response in S. cerevisiaed	Function(s)
SAPs and LIPse
orf6.9036	SAP3	NA	Secreted aspartyl proteinase 3
orf6.4644	SAP1	NA	Secreted aspartyl proteinase 1
orf6.5210	LIP2	NA	Secretory lipase
Cell surface proteins
orf6.8640	FLO9	Induced	Similar to hypharegulated protein HYR1
orf6.1042	WSC2	Induced	Cell wall integrity, stress response protein
orf6.3288	YWP1	NA	Putative cell wall protein
Metabolism
orf6.3413	APE2	No	Aminopeptidase
orf6.4694	SUL1	No	High-affinity sulfate transporter
orf6.2190	SAM2	No	S-adenosylmethionine synthetase 2
orf6.5806	LKH1	Induced	Probable leukotriene A hydrolase
orf6.8469	ALG8	Repressed	Glycosyl transferase
orf6.6671	PUT4	R/I	Proline permease
orf6.9051	THI4	Induced	Thiazole biosynthetic enzyme precursor
orf6.1992	SLC1	Repressed	Fatty acyltransferase
orf6.6385	COX11	Repressed	Cytochrome c oxidase assembly protein
orf6.4023	PFK26	No	6-Phosphofructose kinase
orf6.4161	LEU42	No	2-Isopropylmalalate synthase
orf6.5274	PFK2	Repressed	Phosphofructokinase beta-subunit
orf6.1494	HEM13	Repressed	Oproporphyrinogen III oxidase
Other functions
orf6.6056	OP4	NA	Opaque-phase-specific protein
orf6.8863	CHS7	No	Probable membrane protein
orf6.5830	TRM9	Repressed	tRNA methyltransferase
orf6.4726	CTF8	No	Putative kinetochore protein
orf6.6713		NA	Probable heat shock protease
orf6.6475	HSP31	Induced	Heat shock protein
orf6.6564	PSP1	No	High-copy suppressor of TS mutant of DNA Polf
orf6.6535	PTC4	No	Ser/Thr protein phosphatase PP2C
orf6.8075	HLJ1	No	Tail-anchored ERg membrane protein, unknown function
Unknown functions
orf6.3229		NA
orf6.6060		NA

^aSee “Microarray analysis of gene expression in mating mixtures” in Results and Discussion.

^bORF, open reading frame.

^cNames of C. albicans genes were obtained from the C. albicans genome database (http://www.genolist.pasteur.fr/CandidaDB). In case the genes were not annotated, the names of their S. cerevisiae orthologs were listed. If an unambiguous S. cerevisiae ortholog was not identified from the S. cerevisiae genome database (http://www.yeastgenome.org), the name of the closest ortholog was listed. -, unnamed sequence.

^dPheromone response data of S. cerevisiae were obtained from Roberts et al. (46). Induced, demonstrated induction; No, demonstrated no induction; NA, not available (not reported or tested); Repressed, demonstrated repression; R/I, repressed/induced.

^eSAPs, secreted aspartyl proteinases; LIPs, lipases.

^fTS, mutant of DNA Pol, temperature-sensitive mutant of DNA polymerase.

^gER, endoplasmic reticulum.

Eight of the genes identified as up-regulated (AKL1, CDC55, CAP1, DDR48, GCN4, ERK2, RCE1, and HAC1) and three genes identified as down-regulated (OP4, SAP1, and SAP3) by microarray analysis (Table 2) proved to be similarly regulated when analyzed by Northern blot hybridization (Fig. 2). However, while 7 of 12 genes identified as unchanged by microarray analysis proved to be similarly unchanged by Northern analysis (WH11, EFG1, TEC1, OF6.8100, SLN1, CST20, and FAR1), five proved to be up-regulated (RIM101, CZF1, NRG1, CPH1, and STE3) (Fig. 2; also data not shown). Hence, while the microarray analysis performed here was generally effective in identifying genes that changed transcript levels during mating, they apparently did not identify all genes that were so regulated. The same limitation was probably true for the microarray analysis performed by Bennett et al. (5), which may explain why the genes identified by that study and the present study as up- or down-regulated only partially overlap. However, some differences may be due to the cell preparations, since Bennett et al. (5) analyzed α-pheromone-treated a/− cells, while the present study analyzed a/a-α/α mating mixtures.

FIG. 2.

Northern analyses of gene expression in opaque a/a and α/α mating mixtures and α-pheromone-treated a/a cells. (A) Opaque-specific genes. (B) White-specific genes. (C) New class of genes, which are selectively down-regulated in opaque, but not white, cells in saturation phase cultures and then up-regulated in mating mixtures of opaque cells or opaque a/a cells treated with α-pheromone. (D) Genes expressed exclusively during mating. (E) Genes involved in the mating process that are not regulated (i.e., expressed constitutively) by the mating process. WO-1 α/α white (wh) cells, WO-1 α/α opaque (op) cells, P37005 a/a white cells, and P37005 a/a opaque cells were analyzed after 3.5 h of incubation. The mating mixture, consisting of opaque WO-1 α/α cells and opaque P37005 a/a cells, were analyzed after 3.5 h and 7.5 h. Opaque P37005 a/a cells treated with 3 × 10⁻⁶ M α-pheromone (35) were analyzed after 2.5 h. All original cell preparations were derived from saturation phase cultures that consisted of over 90% unbudded cells.

Regulation of mating-associated genes.

Bennett et al. (5) demonstrated that α-factor induced a number of mating-associated genes in a cells. These included genes encoding proteins involved in receptor-mediated signal transduction (STE2, CEK1, and CEK2), transcription activation (CPH1), pheromone maturation and transport (RAM1 and AXL1), fusion and membrane specialization (PRM1, FUS1, HWP1, FIG1, and RBT1), pheromone adaptation (SST2), and karyogamy (KAR4). Lockhart et al. (35) demonstrated by Northern analysis that α-pheromone induced the STE2, STE4, CAG1, FIG1, HWP1, and KAR4 genes in opaque a/a cells. Here, using both microarray and Northern analyses, we also identified a number of mating-associated genes up-regulated in mating mixtures, including several of the genes previously identified by Bennett et al. (5) and Lockhart et al. (35). Genes previously demonstrated to be up-regulated included CEK2, CPH1, HWP1, FUS1, PRM1, and RBT1 (Table 2 and Fig. 2D). New genes identified as up-regulated included the following: STE3, which encodes the receptor for a-pheromone (21); MFα1, which encodes the α-pheromone (37); RCE1, an ortholog of the S. cerevisiae gene that encodes a CaaX prenyl proteinase involved in a-pheromone maturation (7); and MPT5, an ortholog of the S. cerevisiae gene that encodes a protein involved in reentry into the mitotic cell cycle after pheromone arrest (12) (Table 2 and Fig. 2D). Northern analysis also revealed that neither CST20, the homolog of S. cerevisiae STE20, which encodes a key kinase that activates the MAP kinase pathway (58), nor the homolog of S. cerevisiae FAR1, which encodes a key kinase inhibitor that blocks cell cycle progression during mating (44), was regulated in C. albicans mating mixtures or a/a cells treated with α-pheromone (Fig. 2E). Both genes were equally expressed in untreated and pheromone-treated opaque α/α cells and in mating mixtures.

The results obtained in this study have been combined with those of Bennett et al. (5) and Lockhart et al. (35) to provide a more comprehensive picture of mating-associated gene regulation during the mating process of C. albicans (Fig. 3). Twenty mating-associated genes have been demonstrated to be up-regulated in α-pheromone-treated opaque a cells and/or in mating mixtures of opaque cells (Fig. 3A). Orthologs of all but four of these genes (STE4, CEK1, RCE1, and RAM1) are also up-regulated during S. cerevisiae mating (Fig. 3B). CST20 was constitutively expressed, as was its ortholog in S. cerevisiae STE20 (58), while FAR1, which is pheromone induced in S. cerevisiae (46), was constitutively expressed in C. albicans (Fig. 3B). Hence, while a majority (20 of 25) of mating-associated genes were similarly regulated in C. albicans and S. cerevisiae mating, a minority (5 of 25) were dissimilarly regulated. The latter five genes were dispersed throughout the pheromone-activated program (Fig. 3A) and, hence, provided no clue as to why their regulation differed between the two species.

FIG. 3.

A profile of mating-associated gene regulation during Candida albicans mating (A) and comparison with Saccharomyces cerevisiae mating (B). The data used to develop the model in panel A were from this study and studies by Bennett et al. in 2003 (5) and Lockhart et al. in 2003 (35). (A) Up-regulation of a gene is noted by a vertical arrow, color coded according to the source of data. No definitive information existsfor HST11 and HST7 in the MAP kinase cascade. Boxed genes indicate differences in regulation between C. albicans and S. cerevisiae. Note that the scaffold gene STE5, a necessary component of the MAP kinase cascade in S. cerevisiae mating (13), has not been identified in C. albicans. (B) Induction is compared between C. albicans and S. cerevisiae. The latter data are from the study of Roberts et al. in 2000 (46). Induction or no induction is noted by a plus or minus symbol, respectively. n.a., data not available. The dashed line means no homolog. Genes with differences in regulation between C. albicans and S. cerevisiae are shown on light blue background.

Regulation of filamentation-associated genes.

Bennett et al. (5) found that several genes involved in the regulation of filamentation (CPH1, CEK1, and FGR23) and several regulated by filamentation (RBT1, HWP1, SAP4, RBT4, SAP6, SAP5, and ECE1) were up-regulated in a cells treated with α-pheromone. Our microarray analysis also identified three of these genes, CPH1, HWP1 and RBT1, as up-regulated, as well as an additional gene, DDR48 (29) (Table 2 and Fig. 2C and D). Our results further revealed that YWP1, previously demonstrated to be down-regulated during hypha formation (49), was similarly down-regulated in mating mixtures (Table 3). To investigate further the regulation of filamentation-associated genes during mating, we analyzed by Northern blot hybridization five additional genes, NRG1 (8), EFG1 (55), TEC1 (47), CZF1 (10), and RIM101 (17). Three of these genes, NRG1, CZF1, and RIM101, were up-regulated in mating mixtures and in a/a cells treated with pheromone (Fig. 2C). Together, these results demonstrate that a number of genes associated with filamentation are uniquely up-regulated during C. albicans mating (Table 4).

TABLE 4.

Filamentation-associated genes are uniquely regulated by C. albicans in response to α-pheromone

C. albicans gene	S. cerevisiae ortholog	Response to pheromone in:		Hypha regulation or function in C. albicans
		C. albicansa	S. cerevisiae
CPH1	STE12	Up-regulated (B, TS)	Up-regulated	Hyphal growth (Liu et al. [32])
CEK1	KSS1	Up-regulated (B)	Up-regulated	Hyphal growth (Csank et al. [15])
FGR23	-b	Up-regulated (B)	-	Hypha regulated (Bennett et al. [5])
RBT1	-	Up-regulated (B, TS)	-	Wall growth (Braun et al. [9])
HWP1	-	Up-regulated (TS)	-	Hyphal wall protein (Staab et al. [54])
SAP4	-	Up-regulated (B)	-	Up-regulated in hypha (Hube et al. [23])
RBT4	-	Up-regulated (B)	-	Secretion signal (Braun et al. [9])
SAP6	-	Up-regulated (B)	-	Up-regulated in hyphae (Hube et al. [23])
SAP5	-	Up-regulated (B)	-	Up-regulated in hyphae (Hube et al. [25])
ECE1	-	Up-regulated (B)	-	Hypha-specific expression (Birse et al. [6])
DDR48	DDR48	Up-regulated (TS)	Down-regulated	Up-regulated in hyphae (Lane et al. [29])
CZF1	-	Up-regulated (TS)	-	Hypha regulator (Brown et al. [10])
RIM101	RIM101	Up-regulated (TS)	No change	pH-induced filamentation (Ramon et al. [45])
YWP1	-	Down-regulated (TS)	-	Bud-specific (Sohn et al. [49])

^aB, Bennett et al. (5); TS, this study.

^b-, no homolog.

The filamentation-associated genes regulated by pheromone in opaque C. albicans cells have not been reported to be similarly regulated in S. cerevisiae, primarily because for the majority, there are no S. cerevisiae orthologs (Table 4). Comparison of the cell biology of pheromone-induced shmooing in the two species provided a clue to the reason for the difference. While a cells of S. cerevisiae form relatively short projections in response to pheromone (Fig. 4A through D), C. albicans forms long, hypha-like conjugation tubes (Fig. 4E through H) (34). Since meiosis in S. cerevisiae results in both a and α offspring, fusions can and probably do occur between a and α cells in close proximity. Hence, the mating process of S. cerevisiae may not require long conjugation tubes. In marked contrast, meiosis has not been demonstrated in C. albicans. Because mating types are expressed through MTL homozygosis rather than cassette switching, a and α offspring do not arise in proximity in a single cell lineage as they do in S. cerevisiae. C. albicans conjugation tubes may, therefore, have to travel longer distances to fuse, as has been observed in cell clumps formed in suspension cultures (34) and on skin (27).

FIG. 4.

Candida albicans forms long conjugation tubes that superficially resemble hyphae, while Saccharomyces cerevisiae forms comparatively short evaginations upon pheromone induction. Cells (S. cerevisiae MGD353-13d [a] and C. albicans P37005 [a/a]) were treated in a similar manner with chemically synthesized α-pheromone for 17 h. While S. cerevisiae shmooed and remained morphologically fixed, C. albicans formed conjugation tubes that continued to grow until they reverted apically (panel H, rev). Note that only the tube in panel H has septated, in this case at the junction with the apical bud.

Regulation of white and opaque-phase-specific genes.

It was previously demonstrated by Northern analysis that the opaque-specific genes OP4, SAP1, and SAP3 were down-regulated in cultures of a/a cells treated with α-pheromone (35). Microarray and Northern analyses performed in the present study revealed that these genes were down-regulated in mating mixtures as well (Table 3 and Fig. 2A). Because the Northern blot hybridization signals of these genes were high in a/a and α/α cells prior to mating and negligible in mating mixtures (Fig. 2B), these results further demonstrate that a-pheromone must down-regulate these opaque-specific genes in opaque α/α cells, just as α-pheromone has been shown to down-regulate them in opaque a/a cells (35).

Down-regulation of opaque-specific genes suggested that opaque cells may switch to white in response to pheromone. If this were the case, then one would expect coordinate up-regulation of white-specific genes. Our microarray analysis did not reveal up-regulation of the white-specific genes WH11 (52), EFG1 (53), and TEC1 (R. Zhao, S. R. Lockhart, and D. R. Soll, unpublished observations). Northern analysis further confirmed that these three white-specific genes were not up-regulated during mating or in a/a cells treated with α-pheromone (Fig. 2B). Hence, pheromone does not appear to induce an immediate switch from opaque to white. Rather, opaque-specific genes are independently down-regulated by switching and by pheromones in the mating process. The absence of white-specific gene up-regulation in mating mixtures that contained between 20 to 25% fusants also indicated that a/a-α/α fusion did not immediately cause a switch from opaque to white.

New class of genes regulated by switching, mating, and growth phase.

Northern analyses performed to verify mating-induced up-regulation of a number of genes (AKL1, CDC55, DDR48, HCA1, and GCN4) revealed that while these genes were expressed at low levels in opaque cells and up-regulated in mating mixtures upon pheromone treatment, they were expressed at high levels in white cells (Table 2). None of these genes was identified as up-regulated in a cells treated with α-pheromone by Bennett et al. (5). Northern analyses further revealed that CZF1, RIM101, and NRG1, which encode trans-acting factors involved in filamentation (10, 17, 42), and CAP1, which encodes a trans-acting factor involved in oxidative stress (1), were also expressed at high levels in white cells, low to negligible levels in opaque cells, and high levels in opaque mating mixtures or in α-pheromone-treated a/a cells (Fig. 2C). It was puzzling that none of these nine genes had previously been identified as white specific in a number of diverse screens for white-specific genes (28, 50, 52).

An explanation can be offered on the basis of the growth history of the cells. In the protocol for both the microarray and Northern analyses of mating mixtures in the present study, opaque a/a and opaque α/α cells were derived from liquid growth cultures that had entered saturation phase, since we had qualitatively observed that these cells appeared to respond to pheromone and to fuse at higher levels than cells derived from exponential-phase cultures. In previous screens for opaque-specific genes, cells in the exponential phase of growth had consistently been employed (28, 40, 41, 52). In addition, Bennett et al. (5) employed log-phase cells in their microarray analysis of α-pheromone-induced genes. Therefore, the possibility that differential expression of this newly discovered class of apparently white-specific genes might be limited to saturation phase was entertained. In other words, these genes may be expressed in both white and opaque cells in exponential phase but down-regulated in opaque, but not white, cells that had entered saturation phase. Furthermore, selective down-regulation of these genes could be related to the mating competency of opaque, but not white, cells: i.e., these genes may have to be down-regulated to achieve mating competency. To test these possibilities, we examined expression of two of these genes, AKL1 and CDC55, in white cells and in opaque cells in the exponential and saturation phases of a growth culture (Fig. 5A and B, respectively). The transcript levels of the two genes were similarly high in exponential-phase cultures of both white and opaque cells and remained high in saturation phase cultures of white cells (Fig. 5C). However, the transcript levels were low in saturation phase cultures of opaque cells (Fig. 5C). These results demonstrate that down-regulation of this new class of genes occurs only in opaque cells that have entered saturation phase in a growth culture.

FIG. 5.

The genes in the newly identified group of genes selectively down-regulated in opaque cells and then pheromone induced are expressed at high levels in both log-phase white cells and log-phase opaque cells. (A) Growth kinetics of white-phase WO-1 cells. (B) Growth kinetics of opaque-phase WO-1 cells. In growth kinetics curves of both cell types, the time at which samples were taken from log-phase (LP) and saturation phase (SP) cultures are boxed. (C) Northern analysis of the expression of two genes, AKL1 and CDC55, in white (W) and opaque (O) cells derived from log-phase (LP) and saturation phase (SP) cell cultures.

Saturation phase facilitates mating.

The observations that opaque, but not white, is the mating-competent phenotype (33, 39), that a group of pheromone-inducible genes are selectively down-regulated in opaque but not white cells when they enter saturation phase, and that cells at saturation phase mate more efficiently than cells in the exponential phase of growth led us to hypothesize that entrance into saturation phase may be requisite for mating. To test this hypothesis, we first compared the rate of shmooing in low-density mating mixtures derived from either exponential- or saturation phase growth cultures. Cell multiplication continued in the low-density mating mixtures derived from exponential-phase growth cultures (Fig. 6A) and resumed in low-density mating mixtures derived from saturation phase growth cultures (Fig. 6B). In the mixture derived from exponential-phase cells, the proportion of unbudded cells was 58% at 0 h and 45% at 8 h. Ten percent of the cell population had shmooed after 8 h (Fig. 6A). In the mating mixture derived from saturation phase cells, the proportions of unbudded cells were 92% at 0 h and 35% at 8 h. Forty percent had shmooed after 8 h (Fig. 6B). In both cases, only unbudded cells shmooed. These results were consistent with the hypothesis that opaque cells may have to enter saturation phase to be pheromone responsive. The far smaller proportion of cells that shmooed in mixtures derived from exponential-phase cultures could be explained as that proportion of the cell population that had exited the cell division cycle.

FIG. 6.

Shmoo formation in mating mixtures of opaque α/α (WO-1) and opaque a/a (P37005) cells in suspension is facilitated by deriving cells from saturation phase rather than log-phase growth cultures and resuspending them at a high cell density that inhibits further cell multiplication. (A, B) Opaque a/a and α/α cells derived from log-phase or saturation phase growth cultures, respectively, were mixed in fresh medium at low cell density (2 × 10⁶ per ml). (C, D) Opaque a/a and α/α cells derived from saturation phase growth cultures, respectively, were mixed in fresh medium at high cell density (3 × 10⁷ per ml). The cell concentration and the proportions of unbudded singlets, budding cells, and “shmooing” cells were monitored in each case as a function of time.

In high-density mating mixtures derived from exponential-phase cultures, budding cells, which initially made up approximately 40% of the cell population, divided as the cell population entered saturation phase (Fig. 6C). By 8 h, 40% of the population were unbudded and 50% had shmooed (Fig. 6C). Only unbudded cells had shmooed. In high-density mating mixtures derived from saturation phase cultures, the proportion of unbudded cells at 0 h was 92%, and no further cell multiplication was evident (Fig. 6D). At 8 h, the great majority of unbudded cells had shmooed (Fig. 6D). These results were consistent with the hypothesis that mixtures of opaque a/a and α/α cells in the exponential phase became pheromone responsive as they entered the saturation phase and that the majority of cells in a mating mixture that were in saturation phase were pheromone responsive.

Saturation phase per se, however, is not requisite for pheromone induction.

However, because mating mixtures rely upon the endogenous production of pheromone, we considered the alternative possibility that the low proportion of shmooing in exponential-phase mating mixtures may simply be due to low levels of pheromone, rather than to a difference in pheromone responsiveness. If this were the case, the addition of chemically synthesized α-pheromone to low-density a/a cells derived from exponential-phase cultures should increase the proportion of shmooing cells over that observed in mating mixture to which no exogenous pheromone was added. This was observed (Table 5). In exponential-phase cultures of a/a cells to which chemically synthesized α-pheromone was added, 35% of the cell population shmooed after 3 h, a frequency nine times higher than that for a comparable low-density mating mixture that relied on endogenously generated pheromone (Table 5). Shmooing in α-pheromone-treated cultures was restricted to unbudded cells and included 50% of that portion of the population, or roughly a quarter of the entire cell population (Table 5). The addition of α-pheromone to low-density a/a cells derived from saturation phase cultures induced 73% of the cell population to shmoo after 3 h, three times the level of comparable mating mixtures that relied on endogenously generated pheromone (Table 5). These results demonstrate that pheromone was limiting in mating mixtures, but more importantly, they show that approximately half of the unbudded cells in an exponential-phase population are responsive to α-pheromone and that shmooing is, therefore, not an exclusive characteristic of saturation phase cells.

TABLE 5.

Shmooing in mating cultures and α-pheromone-treated cells derived from log- versus stationary-phase cultures^a

Condition	Cell origin	Proportion of shmoos at 3 h (%)	Fold increase in cell number at 3 h	Proportion of singlets (%) at:		% Change after 3 h	Proportion of budding cells (%) at:		% Change after 3 h
				0 h	3 h		0 h	3 h
a/a-α/α mating mixture	Log phase	4	2.0	58	50	−8	42	45	+3
	Stat.b phase	25	2.4	92	60	−30	8	15	+7
α-Pheromone-treated a/a cells	Log phase	35	2.3	58	44	−14	42	24	−20
	Stat. phase	73	1.7	92	19	−73	8	12	+4

^aLog- and stationary-phase cells were diluted into fresh medium containing 3 × 10⁻⁶ M chemically synthesized α-pheromone (Lockhart et al. [35]) at a final cell density of 2 × 10⁶ per ml.

^bStat., stationary.

G₁ may be a requisite for shmooing.

Our results (Table 5), therefore, indicated that approximately half of the unbudded cells in an exponential-phase growth culture, representing between 25% and 35% of the entire cell population, can be induced by pheromone to shmoo, just as the majority of cells in a saturation phase population can be induced. What characteristics do saturation phase cells and half of the unbudded cells in an exponential-phase culture have in common? First, they are both in the unbudded phase of the cell cycle. Second, they may both be in G₁ of the cell cycle. Prior observations indicated that in modified Lee's medium (3), cells entering saturation phase accumulated in G₁ of the cell cycle (4, 51). Hence, the requisite for mating competency may not be saturation phase, per se, but simply G₁ of the cell cycle. To test this hypothesis, we developed a method for measuring the DNA content of the nucleus of individual cells that could be staged microscopically in the budding cycle. Sytox Green was used to stain DNA, and scanning confocal microscopy was used to quantify staining. Measurements were made from a line scan through the nucleus (Fig. 7A through F, left panels of each pair of panels). The line scan intensities were then plotted as a function of pixel distance (Fig. 7A through F, right panels of each pair of panels). A representative intensity scan through the nucleus of an unbudded saturation phase cell is presented in Fig. 7A. The peak was at 125 pixel intensity units. Peak measurements of pixel intensity for the nuclei of 28 unbudded saturation phase cells, presented in a histogram of ascending values in Fig. 8A (blue), ranged from 100 to 125 pixel intensity units, with a mean ± standard deviation of 113 ± 7 pixel intensity units. Peak measurements of the pixel intensities of 28 nuclei of exponential-phase unbudded cells, presented in a histogram of ascending values in Fig. 8A (red), included two groups. Measurements of the first group, which included the nuclei of 13 cells (46%), ranged from 105 to 125 pixel intensity units, mimicking the distribution of measurements of the nuclei of saturation phase cells. The mean peak pixel intensity for this group was 113 ± 5, similar to that of saturation phase cells (113 ± 7). Measurements of the second group, which included the nuclei of 15 cells (54%), ranged between 145 and 245 pixel intensity units. The mean peak pixel intensity for the second group was 184 ± 37. The three highest values in the second group (242, 245, and 247 pixel intensity units) were approximately twice the average value of nuclei in the first group (Fig. 8A), suggesting that the range of values in the second group reflected different stages of DNA replication. The peak measurements of nuclei in mother cells with anucleate buds ranged between 240 and 255 pixel intensity units (n = 10). A value of 245 pixel intensity units was, therefore, deemed the replicated state, since it was approximately twice the value of stationary-phase nuclei or nuclei of unbudded exponential-phase cells in the first group (Fig. 8A). A scan of a nucleus of an exponential-phase cell in the first group, with a peak pixel intensity of 124 (unreplicated), is presented in Fig. 7B, and a scan of the nucleus of an exponential-phase cell that had formed an anucleate bud, with peak pixel intensity of 245 (replicated), is presented in Fig. 7C. Scans of an unreplicated nucleus in a mother cell and a replicated nucleus in a daughter cell that had in turn budded in a pseudohyphal sequence are presented in Fig. 7D.

FIG. 7.

Nuclear DNA was quantitated by staining with Sytox Green and quantitating fluorescence by obtaining line scans of pixel intensity using a laser-scanning confocal microscope. (A) Saturation phase unbudded, unreplicated DNA. (B) Log-phase unbudded cell, unreplicated DNA. (C) Log-phase budded cell, replicated DNA. (D) Log-phase cell undergoing pseudohyphal growth, with a nucleus in the mother cell containing unreplicated DNA (white) and a nucleus in the first daughter cell containing replicated DNA (red). (E, F) Shmoos derived from log-phase cells containing unreplicated DNA. Fluorescence intensity is noted as pixel intensity, plotted along the scan. Bar, 5 μm.

FIG. 8.

Histograms of peak fluorescent intensity (pixel intensity) demonstrate that the nuclei of all saturation phase unbudded cells contain unreplicated DNA, half the nuclei of log-phase unbudded cells contain unreplicated DNA and half DNA undergoing replication, and the nuclei of unbudded cells forming conjugation tubes (shmooing) contain unreplicated DNA. (A) Comparison of saturation phase unbudded cells and log-phase unbudded cells. (B) Comparison of log-phase unbudded cells and α-pheromone-induced log-phase shmoos. Cells were treated with α-pheromone for 3.5 to 4.0 h. In each histogram, cells are ordered in ascending DNA content.

If shmooing occurred exclusively in unbudded cells in G₁ in an α-pheromone-treated culture of dividing cells, then the distribution of nuclear DNA intensities of cells that had just shmooed should mimic the distribution of nuclei in saturation phase cells or untreated exponential-phase unbudded cells in the first group (Fig. 8A). That was exactly the result we obtained. The range for 28 α-pheromone-treated exponential-phase cells that had just shmooed was between 100 and 145 pixel intensity units (Fig. 8B). The mean was 120 ± 11 pixel intensity units, very close to the average value of a nucleus with unreplicated DNA (113 ± 5 pixel intensity units) and approximately half that of nuclei with replicated DNA. Examples of scans of the nuclei of representative cells that had shmooed in α-pheromone-treated exponential-phase cultures are presented in Fig. 7E and F. These results indicate that shmooing induced by α-pheromone in a low-density, multiplying culture of cells may occur only in unbudded cells in the G₁ phase of the cell cycle.

Pheromone blocks cells in G₁.

Measurements of the relative DNA content of the nuclei of cells that had shmooed in exponential-phase cultures treated for 4 h with pheromone revealed that in every case nuclear DNA was unreplicated (Fig. 8B). For the majority of these cells, the length of the evagination (conjugation tube) was between one-quarter and one cell diameter. These results suggested not only that cells had to be in G₁ to shmoo but also that pheromone blocked cells in G₁. To demonstrate this point, we searched for those cells in exponential-phase cultures treated with α-pheromone for 4 h that had formed conjugation tubes greater than two cell diameters and that had not reverted apically to the budding growth form (34). The relative DNA content of the nuclei in these cells, which remained localized in the parent cell body, ranged between 110 and 125 pixel intensity units (n = 10), the same range as the nuclei of the general population of cells that had shmooed. These results indicate that cells must not only be in G₁ to respond to α-pheromone but that pheromone then blocks cells in G₁ as the conjugation tube grows. In S. cerevisiae, it has been demonstrated that pheromone similarly blocks cells in G₁ (57). When the conjugation tubes of C. albicans reverted apically to the budding growth form (34) in 7-h cultures, the relative DNA content of the nucleus ranged between 150 and 240 pixel intensity units (n = 4), indicating that upon reversion, the block is lifted.

Conclusions.

The mating system of C. albicans is remarkable not only for the characteristics it shares with S. cerevisiae but also for its unique features. C. albicans, like S. cerevisiae, relies upon pheromones from opposite mating types to induce the formation of conjugation projections, which find each other through chemotropism (34). These tubes then fuse so that the nuclei of opposite mating types can undergo karyogamy and recombination. The pheromone response pathway is similar in the two organisms, as are the genes regulated by the pathway (5, 11, 25a, 37). In addition, we have shown here that C. albicans cells respond to pheromone in G₁ and are then blocked in G₁ by pheromone, characteristics similar to those of S. cerevisiae. However, there are differences, some of which are related to the incorporation of two developmental programs into the C. albicans mating process. While the majority of mating-related genes analyzed here and in previous studies (5, 35) were similarly regulated in C. albicans and S. cerevisiae, a minority were regulated quite differently. A few of these latter genes, such as STE4 and FAR1, are pivotal in the mating response pathway, so it seems quite unlikely that the differences are without purpose, although that purpose is not obvious. On the other hand, the unique up-regulation of hypha-associated genes correlates with the unique hypha-like conjugation tubes formed by C. albicans. On skin, which facilitates shmooing and fusion, the conjugation tubes have been shown to extend long distances over complex terrain to fuse (27). It seems likely that pheromone activation of filamentation-associated genes and induction of hypha-like conjugation tubes reflect integration of the filamentation program into the C. albicans mating process, which may have evolved through host-pathogen interactions (38). This may also be at least part of the explanation for why cells must switch from white to opaque to mate and why pheromone down-regulates opaque-phase-specific genes. Cells in the opaque phase selectively colonize skin (26), which facilitates mating (27). Cells in the opaque phase also do not release a white blood cell chemoattractant and are, hence, invisible to human polymorphonuclear leukocytes (19). Down-regulation of opaque-specific genes suggests that while the opaque phenotype may be required for the first steps in the mating process, expression of some opaque-specific genes may be incompatible with later steps in the mating process. Finally, it is not obvious why a set of genes identified for the first time in the present study are selectively down-regulated in opaque, but not white, cells entering saturation phase in liquid culture. Since exponential-phase cells can mate, the selective down-regulation of these genes in saturation phase may serve an opaque-related function other than mating. What is clear is that C. albicans, in contrast to S. cerevisiae, has incorporated two developmental programs, switching and hypha formation, into the mating process. These differences may be due to the fact that C. albicans, unlike S. cerevisiae, is a true pathogen that has developed a life cycle that depends upon its interactions with a host.