Transcriptional Rewiring of Fungal Galactose-Metabolism Circuitry

Current Biology 17, 1007–1013, June 19, 2007 ª2007 Elsevier Ltd All rights reserved

DOI 10.1016/j.cub.2007.05.017

Article Transcriptional Rewiring of Fungal Galactose-Metabolism Circuitry Mikhail Martchenko,1 Anastasia Levitin,2 Herve Hogues,2 Andre Nantel,2 and Malcolm Whiteway1,2,* 1 Department of Biology McGill University Montreal, Quebec H3A 1B1 Canada 2 Genetics Group Biotechnology Research Institute National Research Council Montreal, Quebec H4P 2R2 Canada

Summary Background: The Leloir-pathway genes encode the enzymatic machinery involved in the metabolism of galactose. Results: In the distantly related fungi Saccharomyces cerevisiae and Candida albicans, the genes encoding these enzymes are syntenically arranged, but the upstream regulatory regions are highly divergent. In S. cerevisiae, the Leloir-pathway genes are positively regulated by Gal4p acting through the UASG sequence CGG(N11)CCG. However, in C. albicans, the Gal4p and UASG combination is found to regulate genes unrelated to galactose metabolism. We identified a palindromic sequence that acts to control GAL10 expression in C. albicans in the presence of galactose. This palindrome is found upstream of other Leloir-pathway genes in C. albicans, and in the absence of other regulatory sequences, activation of expression through this sequence in the presence of galactose requires Cph1p, the homolog of the Ste12p transcription factor of S. cerevisiae. Conclusions: Although the cellular process of galactose induction of the Leloir pathway is conserved between the two organisms, the regulatory circuits achieving the cellular process are completely distinct. Introduction The diploid fungal pathogen, Candida albicans, which causes the majority of human fungal infections, diverged from the bakers or brewers yeast S. cerevisiae approximately 250 million years ago [1]. Because many of the fundamental biological processes of C. albicans are similar to those of S. cerevisiae, budding yeast is an important guide for studying the pathogen. However, unlike S. cerevisiae, C. albicans can form true hyphae, can escape mammalian immune cells, and can proliferate in mammalian hosts and thus has the tools to be a pathogen. Despite a strong conservation of gene content, yeast species exhibit major phenotypic differences

*Correspondence: [email protected]

because of the different environmental selections imposed on them; these pressures may in turn drive changes in gene regulation. For example, the regulation of the ribosomal-protein genes in C. albicans is different from that in S. cerevisiae [2]. We are interested in whether transcriptional rewiring occurs for conserved processes such as sugar and amino acid metabolism. Galactose is utilized by almost all organisms through its conversion to glucose-6phosphate in a reaction catalyzed by the enzymes of the Leloir pathway [3]. This genetic regulatory circuit acting on the S. cerevisiae GAL genes is among the most highly studied and best understood of eukaryotic metabolic-control pathways. As such, it has acted as a paradigm for general principles of metabolic control [4, 5] and has served as well as a template for the application of a systems-biology approach to metabolic circuitry [6]. In S. cerevisiae, these enzymes are encoded by GAL1 (galactokinase), GAL7 (galactose-1-phosphate uridyl transferase), GAL10 (UDP-glucose-4-epimerase), and GAL5 (phosphoglucomutase) [7, 8]. Galactose enters S. cerevisiae cells through a specific permease, encoded by GAL2 [9]. Because eukaryotic genes are not organized in operons, it is intriguing that the GAL1, GAL7, and GAL10 genes are clustered near the centromere of chromosome 2 [10]. The expression of these GAL genes is controlled by a network that involves activating and repressing activities and that is encoded by the specific regulatory genes GAL3, GAL4, and GAL80. In S. cerevisiae, GAL4 serves as the transcriptional activator of galactose catabolism [5, 11], whereas Gal80p is involved in the repression of GAL genes by binding the transcriptional activation domain of Gal4p [12]. Gal3p functions by forming a complex with Gal80p to relieve inhibition of Gal4p; in the presence of galactose and ATP, Gal3p sequesters Gal80p in the cytoplasm, preventing inhibition of Gal4p and thus activating GAL gene expression [13, 14]. ScGal4p contains a DNA-binding domain that interacts with a specific upstream activating sequence (UASG; CGG(N11)CCG), located in the promoter regions of GAL1, GAL2, GAL3, GAL7, GAL10, and GAL80 [15]; the GAL5 gene is unregulated, being expressed under all conditions, and lacks the UASG [16]. In this investigation, we studied the regulation of galactose metabolism in the pathogenic fungus Candida albicans. This work helps define the C. albicans network that regulates GAL genes and shows that C. albicans and S. cerevisiae control the same process through different regulatory circuits. Results Elements of the C. albicans GAL Regulon The sequencing of the C. albicans genome [17] provides a framework for the postgenomic studies of this important fungal pathogen. We have recently completed the assembly of the Candida albicans genome [18] and

Current Biology 1008

established a detailed annotation of its genes [19]. Intriguingly, although overall the synteny between C. albicans and S. cerevisiae is very low, the genomic organization of the region involving the galactose metabolism cluster in S. cerevisiae is remarkably similar to that found in C. albicans. The galactokinase, galactose-1-phosphate uridyl transferase, and UDPglucose-4-epimerase genes show conservation of sequence (63%, 79%, and 71% similarity, respectively), and these genes are clustered together on chromosome 1 of C. albicans in the same relative arrangement as was found in S. cerevisiae. In the fungal pathogen, this cluster also includes hexose transporter (orf19.3668) with 48% similarity to ScGAL2. The analysis of GAL region across Ascomycota shows that GAL genes are arranged in a cluster that includes two currently uncharacterized open reading frames, between GAL10 and GAL7, that are conserved in C. albicans, C. dubliniensis, C. parapsilosis, and Debaryomyces hansenii. Schizosaccharomyces pombe contains only one of these open reading frames (Figure 1A). Although the galactose-metabolism structural genes are well conserved between S. cerevisiae and C. albicans, the regulatory components show less similarity. We were not able to detect even weak candidates for Gal3p in the C. albicans genome. A candidate for Gal80p, encoded by ORF19.6899, shares only 40% sequence similarity with ScGal80p. ORF19.5338, annotated as CaGal4p, shares strong sequence similarity (86%) with its S. cerevisiae homolog only in the DNA-binding domain (Figure 1B), with the six cysteine residues, the linker region, and the dimerization region all well conserved. This C. albicans gene encodes a much smaller protein of 261 amino acids compared to its S. cerevisiae homolog of 881 amino acids. The activation domains of those two proteins share no similarity, and the negatively charged region that serves as the ScGal80p interaction domain is missing in CaGal4p [20]. Galactose Regulates Expression of GAL1, GAL7, and GAL10 To investigate whether galactose regulates expression of the GAL1, GAL7, and GAL10 genes in C. albicans, we performed a transcriptional-microarray analysis of cells grown on galactose as the sole carbon source, as compared to cells grown on glucose. Just as in S. cerevisiae, galactose strongly induced the transcription of genes for the Leloir-pathway enzymes; the GAL1, GAL7, GAL10, and the hexose transporter ORF19.3668 genes were among the most highly expressed of all the induced genes. In addition, the expression of 78 genes involved in gluconeogenesis, glycogen degradation, formation of the cell wall, transport, fatty-acid metabolism, and a variety of unknown functions was greater in galactose-grown cells compared to glucose-grown cells (Figure S1A and Table S1 in the Supplemental Data available online), whereas the expression of a further 54 C. albicans genes, including genes for cell-wall components, transporters, transcription factors, ribosome-synthesis elements, as well as for unknown functions, was greater in glucose than in galactose (Figure S1B and Table S2). Surprisingly, putative glucose transporters are equally expressed in both glucose- and galactose-grown cells with the

exception of HGT1, HGT12, and HGT17. The expression of high-affinity glucose transporters HGT12 and HGT17 was previously shown to be repressed by 2% glucose, possibly by the action of the Mig1repressor [21–23]. Our microarray results show that the expression of GAL10 is 5.5 times greater in YPGal than in YPD. In order to confirm the set of genes whose expression was elevated in YPGal compared to YPD, we cloned the promoter of GAL10 upstream of lacZ. We studied the expression of GAL10-driven lacZ expression levels in cells grown on either SC dextrose or SC galactose. We show that the difference in the expression of GAL10driven lacZ is 5.1 times greater in SC galactose than in SC dextrose; this finding compares very well to our microarray results. In order to confirm the microarray results, we cloned the promoter of GDH3 gene, whose expression levels were higher in YPD compared to YPGal, upstream of a lacZ reporter and transformed it into the wild-type C. albicans strain. The microarray data showed that the levels of GDH3 transcription were 2.4 times higher in YPD than in YPGal. The expression of GDH3-driven lacZ expression was 3.3 times greater in SC dextrose compared to SC-galactosegrown cells, and such a finding relates well to the microarray data. The C. albicans GAL10 Promoter Has a Conserved and Functionally Important Element In order to study the regulation of the galactosemetabolizing genes, we used the promoter of GAL10 cloned upstream of the lacZ reporter [24] and transformed this reporter construct into a wild-type strain. Cells were grown in liquid medium with raffinose as the carbon source and shifted to medium containing either dextrose or galactose. After 90 min growth, the galactose-treated cells expressed 5-fold greater bgalactosidase activity than the dextrose-treated cells. The raffinose-grown cells showed intermediate levels of lacZ expression (Table 1); therefore, as in S. cerevisiae, GAL10 is apparently regulated by both activation and repression mechanisms. Because the presence of galactose coordinately upregulates the expression of the GAL genes, we searched for an upstream activating sequence (UASG). We compared the DNA sequences of the GAL1, GAL7, and GAL10 promoters and found a common palindromic stretch 50 -TGTAACGTTACA-30 that occurs only four times in the intergenic regions of the C. albicans genome, with all occurrences in the promoters of the GAL1, GAL7, and GAL10 genes, approximately 200 base pairs upstream of the protein start sites. As well, the promoter of the hexose transporter gene, ORF19.3668, has the sequence GTTACGT TAC, which is a GTTAC repeat rather than GTTAC palindrome. The palindromic sequence is also conserved in the promoters of the Leloir-pathway genes of C. dubliniensis, C. parapsilopsis, and D. hansenii. To test the importance of this GAL-palindrome sequence in the regulation of GAL10 expression in the presence of galactose, we deleted it from the GAL10 promoter. When cells were grown overnight in either galactose or dextrose medium, the GAL palindrome appears critical for the galactose-mediated induction of GAL10 expression. The palindrome-less full-length GAL10 promoter showed a slight 1.3-fold reduction of

Transcriptional Rewiring of Galactose Metabolism 1009

Figure 1. Analysis of the GAL Circuit across Ascomycota (A) The structural members of the GAL regulon are conserved in Ascomycota and are also organized in a cluster. Species in which whole-genome duplication occurred are identified by the star and by the color. The regions of D. hansenii, C. parapsylopsis, C. dubliniensis, and C. albicans also contain two uncharacterized open reading frames X and Y, marked in black and white color; S. pombe contains only one of these elements. The primary sequences of the GAL promoters were analyzed for the presence of Gal4-, Cph1-, and Mig1-binding sites by the motif-finding program MEME (http://weblogo.berkley.edu/). The phylogenetic relationships among the fungi are shown as defined in [45]. (B) ScGal4p acts as the activator of the S. cerevisiae GAL regulon. The putative CaGal4p protein, encoded by C. albicans ORF19.5338, shares strong sequence similarity (86%) with S. cerevisiae Gal4p only within the DNA-binding domain; the DNA-binding domain has the six cysteine residues, the linker region, and the dimerization region all well conserved.

expression in the presence of dextrose and a strong 5-fold drop of expression in the presence of galactose, compared to the intact GAL10 promoter (Table 1). A truncated version of the promoter was also assayed. In these constructs, the lacZ reporter was linked to a minimal promoter containing or lacking the palindromic sequence. After overnight growth, the palindrome directed 10-fold greater expression in the presence of galactose compared to dextrose, and this result suggests that the GAL palindrome acts as a galactosedextrose-dependent regulator of the GAL10 promoter under these growth conditions (Figure 2). Definition of the C. albicans Gal4p Regulon In S. cerevisiae, Gal4p activates the expression of GAL genes to metabolize the galactose present in the environment, and the deletion of ScGal4 leads to the inability of cells to grow on galactose [5]. To investigate which C. albicans genes are regulated by the Gal4p homolog, we created a C. albicans gal4 knockout strain. The C. albicans gal4 strain grew identically to the wildtype strain when utilizing galactose, dextrose, ethanol, or glycerol as a sole carbon source (data not shown). By using microarrays, we then compared the transcriptional behavior of the wild-type and gal4 strains when both were grown on galactose-containing media (YPGal). The expression of GAL genes remained unchanged (Tables S3 and S4). In addition, deletion of

the GAL4 gene had no effect on the regulation of GAL10-driven lacZ expression, suggesting that CaGal4p does not transactivate the expression of GAL10. Further transcriptional analysis of the C. albicans gal4 strain compared to the control grown on galactose showed that loss of gal4 reduces the expression of 87 genes and increases the expression of 37 genes. The Gal4p-influenced gene set represents 2% of the genome and includes genes whose products are involved in a diverse range of cellular processes. The 87 genes that are reduced in the absence of Gal4p encode transporters, cell-wall proteins, nucleic-acid-binding proteins, and hypothetical proteins (Figure S2A and Table S3). Interestingly, one third of these downregulated genes are members of a newly discovered subtelomeric gene family TLO (for TeLOmere associated) [18], and 14

Table 1. b-galactosidase Activities in Miller units of C. albicans Promoters in C. albicans Strains Construct

Dextrose

Raffinose

Galactose

pGAL10-lacZa pGAL10-lacZD176-188b

98 6 7 75 6 5

196 6 15 82 6 5

501 6 30 96 6 6

a

pGAL10-lacZ contains full-length GAL10 promoter. pGAL10-lacZD176-188 contains full-length GAL10 promoter, in which the GAL-specific palindrome located 188 base pairs upstream of the start site was deleted. b

Current Biology 1010

Figure 2. Cph1p Regulates the Expression of GAL10 The promoter of GAL10, a member of the C. albicans Leloir pathway, was fused to a downstream lacZ reporter. The promoter either contained or lacked a potential GAL-specific palindrome (TGTAACGT TACA). The resulting b-galactosidase activities were analyzed in wild-type and cph1 strains grown either in galactose or dextrose. The values of the liquid b-galactosidase assays are expressed in Miller units.

genes encode products involved in glycolysis (e.g., PGI12, LSC2, and LAT1). The expression of a further 37 genes, which encode transporters, cell-wall proteins, nucleic-acid-binding proteins, iron and copper uptake regulators, and hypothetical proteins, is elevated in the absence of CaGal4p (Figure S2B and Table S4). The strongest Gal4p-transactivated genes are members of the C. albicans-specific subelomeric TLO gene family of unknown function, previously known as CTA2. Almost all of the TLO genes share an identical promoter region, which contains the sequence CGG(N11)CCG that represents the binding site for ScGal4p. When cloned upstream of lacZ, a 126-basepairs-long TLO promoter that includes the Gal4-binding site was 5.5 times less active in a gal4 mutant strain, confirming that CaGal4p is required for proper expression of this gene through the UASG site (Figure 3B). The absence of the Gal4-binding site alone leads to a 3-fold drop in reporter expression; the combined absence of both Gal4p and its binding site leads to a 5-fold drop of TLO-driven reporter expression. This suggests that Gal4p transactivates the expression of

TLO through a classical UASG. Because the absence of the Gal4p-binding site in the wild-type allows greater b-galactosidase activity than does the absence of Gal4p, it is possible that there is additional Gal4pdependent regulation of TLO expression. The promoters of a number of the C. albicans Gal4pupregulated glycolysis genes (e.g., LAT1 and LSC2) contain a potential regulatory site CGG(N11)CCA that is conserved in the promoters of the same genes in C. dubliniensis and C. tropicalis and that closely resembles a UASG site (Figure 3A). The removal of this site from the 270-base-pair-long LAT1 promoter, the loss of Gal4p, or the combined loss of both Gal4p and the regulatory site all lower expression by about 2.5 fold, suggesting Gal4p works through this site to regulate LAT1 expression (Figure 3). This implies that the proper expression of TCA cycle genes is dependent on both the presence of a functional Gal4p and the Gal4p-binding sites within their promoters. To test whether Gal4p is responsible for the differential expression of the genes whose transcription changes in response to a carbon source of either dextrose or galactose (Figure S1, see also Tables S1 and S2), we used microarrays to compare the transcriptional behavior of the gal4 strain to that of its wild-type equivalent grown on dextrose-containing media (YPD). We observed that Gal4p is needed for the expression of the same genes in the presence of either dextrose or galactose. The Regulators of the Expression of GAL10 Extended incubation of C. albicans with macrophage cells leads to a differential expression of GAL genes through a process that may involve the MAP kinase pathway acting through Cph1p [25]. The half-palindrome sequence TGTAAC is similar to the Ste12p/Cph1pbinding site TGAAAC, so we analyzed the expression of the GAL10 promoter in a cph1 knockout strain [26]. In order to exclude the influence of the upstream GAL10-promoter regulatory sequence on the actions of the palindrome, we examined the expression patterns of the 188-base-pair-long GAL10 promoter containing the palindrome. When cloned upstream of lacZ,

Figure 3. Gal4 Regulates Expression through Gal4-DNA-Binding Site (A) The promoters of C. albicans, C. dubliniensis, and C. tropicalis glycolysis genes (e.g., LAT1 and LSC2) contain conserved Gal4p-binding sites. Deduced nucleotide sequences were aligned with the CLUSTALW program; identical nucleotides are boxed in black. Numbers at the ends of the sequences represent the distances from the ATG start site. (B) The promoter of TLO, a member of the C. albicans telomeric-gene family, and the promoter of LAT1, a member of the TCA cycle, were fused to drive the expression of the downstream lacZ reporter.The promoters either contained or lacked a potential Gal4p-binding site (BS) ‘‘CGG(N11)CCG/A,’’ and their resulting b-galactosidase activities were analyzed in wild-type and gal4 strains. The values of the liquid b-galactosidase assays are expressed in Miller units.

Transcriptional Rewiring of Galactose Metabolism 1011

a GAL10 promoter that includes the palindrome was five times less active in a cph1 mutant strain grown on galactose, confirming that Cph1p is required for proper expression of this gene through the palindrome (Figure 2). The absence of the palindrome alone leads to 3.3-fold drop in the reporter expression; the combined absence of both Cph1p and palindrome leads to a 5-fold drop of GAL10-driven reporter expression, suggesting that Cph1p transactivates the expression of GAL10 by interacting with the palindrome. Because the absence of the palindrome in the wild-type strain permits greater b-galactosidase activity than does the absence of Cph1p, it is possible that there is additional Cph1pdependent regulation of GAL10 expression. The removal of the Cph1p had a very mild effect on GAL10 expression in the absence of the palindrome. These results suggest that in the presence of galactose, Cph1p acts as an activator of the GAL10 expression through the palindrome sequence. Discussion In this study, we show the same biochemical process, the induction of the Leloir-pathway genes by galactose, is regulated by different transcriptional circuits in S. cerevisiae and C. albicans. This transcriptional rewiring has resulted in the C. albicans protein that is the closest homolog of the S. cerevisiae transcriptional regulator Gal4p not controlling the expression of GAL genes in the pathogen. Because the presumptive C. albicans GAL4 has a different function from its S. cerevisiae homolog, we suggest the name CGF1, for Candida Gal Four homolog, for this transcription regulator, to reduce the confusion of the functional implications of the name GAL4. Our observations show that the mechanism of regulation of transcription of GAL genes is fundamentally different in S. cerevisiae and C. albicans. In S. cerevisiae, the promoters of GAL genes contain a UASG, which functions as a galactose-inducible enhancer. We observed that the regulation of the GAL10 promoter in C. albicans is controlled by at least two regulatory sites: an enhancer and a galactose-dextrose-responsive element. Our study shows that the latter is regulated by Cph1p, a homolog of Ste12p of S. cerevisiae. The difference in the regulation of the Leloir-pathway genes may be due to the fact that galactose plays important roles in C. albicans adhesion and biofilm formation, processes that contribute to the virulence of this pathogen and which are absent in S. cerevisiae [27]. To investigate when the changes of the cis-regulatory element in the GAL promoters occurred during the evolution of the yeast species, we used available genomic data of the Ascomycota that have GAL genes conserved in a cluster in their genome: Schizosaccharomyces pombe, S. cerevisiae, S. paradoxus, S. mikatae, S. bayanus, S. castellii, Kluyveromyces lactis, Debaryomyces hansenii, C. parapsilosis, C. dubliniensis, and C. albicans (Figure 1). S. cerevisiae and its close relatives contain classic tandem Gal4p-binding sites and lack binding motifs that are similar to Cph1p-binding motifs, whereas C. albicans and its close relatives contain binding sites that are similar to Cph1p-binding sites and lack any motifs that are similar to Gal4p-binding motifs in

their GAL promoters. Interestingly, the intermediate species S. castellii and K. lactis contain both Gal4plike and Cph1p-like motifs. S. pombe, which lies outside of the Hemiascomycetes lineage, lacks a strong Gal4p homolog and lacks Gal4p-binding motifs. However, it contains Cph1p-like motifs upstream of its GAL genes, and such a finding suggests a Gal4p role in galactolysis may be specific to S. cerevisiae and its relatives. The loss of these open reading frames of unknown function occurs in the Saccharomyces lineages concurrent with the appearance of Gal4 UASGs and the disappearance of Cph1 UASGs in the GAL regulatory regions. Recent studies show that the evolution of the regulatory circuits that control cell-type regulation coincides with a whole-genome-duplication event in yeast [28]. We observed that evolution of metabolic GAL regulation appears to have a somewhat different pattern; the Saccharomyces-lineage control circuits are separate from other fungi, with K. lactis and S. castellii perhaps representing intermediates having both regulatory motifs (Figure 1A). Saccharomyces species primarily degrade hexoses to pyruvate and ethanol by fermentation, even in the presence of oxygen, whereas C. albicans degrades hexoses by respiration [29]. The emergence of this capacity is linked to the apparent whole-genome-duplication event [30]. This phenomenon relies on a ‘‘glucoserepression’’ circuit that represses the respiratory part in the presence of glucose [31]. The transcriptional repressor Mig1p controls the glucose repression in S. cerevisiae. The analysis of GAL promoters in Ascomycota shows that Mig1 recruitment to the GAL region precedes the whole-genome-duplication event in Saccharomyces, and it is present in K. lactis [32] (Figures 1A and 4). Finally, GAL3 is thought to have arisen from GAL1 as a result of the S. cerevisiae whole-genome duplication [33]. These observations suggest that the rewiring of transcriptional control could occur through an intermediate, in which both old and new regulators control gene expression (Figure 4). Because Gal4p is present only in fungi, it would be interesting to know what regulates GAL genes in humans, where the absence of those enzymes results in galactosemia [34]. It is interesting to note that the core of the MAPK-controlled AP1-binding site, tgACGTca [35], shares similarity to the GAL palindrome, tgtaACGTtaca, and human GAL genes contain potential AP1-binding sites. Experimental Procedures C. albicans Strains and Plasmid Constructions The C. albicans strains used in this study are listed in Table S5. BWP17 [36] was used for generating GAL4/gal4, gal4, and its wildtype prototrophic equivalent BWP17-HA. CAI4 [37] strain was used for mapping pGAL10 regulatory motifs and for studying the influence of hexoses on GDH3 promoter. JKC19 strain (cph1) [26] was used for identification of the transcriptional regulator of the GAL10 promoter. Strains gal4 and its wild-type equivalent, CMM1, were used to show that CaGal4p regulates the expression from the TLO and LAT1 promoters through the Gal4p-DNA-binding site. JKC19 (cph1) was used to show that Cph1p regulates expression from the GAL10 promoter through a GAL-specific palindrome. Plasmids and oligonucleotides are shown in Tables S6 and S7. We created the GAL4 deletion cassettes, pHIS1-GAL4 and pARG4GAL4, by cloning the 500 bp GAL4 flanking regions into pFA-HIS1

Current Biology 1012

The levels of sugars used in this study (2%) do not represent physiological levels found in the normal environment of the pathogen. The growth of the gal4 strain was assayed as described [44]. Blast Searches The search for the members of C. albicans GAL regulon was performed with the standard blastp setting at CGD (http://www. candidagenome.org/cgi-bin/nph-blast), at NRC-BRI (http:// candida.bri.nrc.ca/candida/index.cfm?page=blast), and at SGD (http://seq.yeastgenome.org/cgi-bin/blast-fungal.pl). Supplemental Data Two figures and seven tables are available at http://www. current-biology.com/cgi/content/full/17/12/1007/DC1/. Acknowledgments We would like to thank Daniel Dignard for bioinformatics assistance and two anonymous reviewers for insightful comments and corrections. This work was supported by Canadian Institutes of Health Research grant MOP-42516 (to M.W.). M.M. gratefully acknowledges a FRSQ-FCAR-Sante Scholarship and National Research Council Graduate Student Scholarship Supplement. This is National Research Council publication 47554.

Figure 4. Model for the Evolutionary Change in GAL Gene Regulation A possible scheme through which the Cph1-like regulator of the GAL genes in an ancestor of C. albicans and S. cerevisiae could have been replaced, in S. cerevisiae, by the Gal4 factor. First, Gal4p and Mig1p were recruited to coregulate GAL genes along with Cph1p prior to whole-genome duplication. Eventually, Cph1p was lost, and GAL expression became Gal4p and Mig1p dependent. and pFA-ARG4 [38]. The TLO, LAT1, and GAL10 promoters as well as their shorter versions were cloned into the plac-poly backbone [24], and the GAL palindromic sequence was cloned into pCRlacZ [39]. All of the constructs created in this study were integrated into the genome of C. albicans with digestion with StuI to target them to the RPS1 locus [40]. Microarray Analysis Transcription profiling and analysis were performed with long oligonucleotide microarrays as described [41]. Total RNA was extracted by the hot-phenol extraction method [42]. GeneSpring software (Agilent Technologies) was used for normalization and for the identification of significantly regulated transcripts. We used a t test p value cutoff of 0.05 to be statistically significant. For analysis of the transcriptional response under each experimental condition, at least seven individual biological replicates were used. C. albicans strain gal4 and its wild-type equivalent (CMM3 and CMM1) were inoculated overnight in media containing 1% yeast extract, 2% peptone, with either 2% dextrose (YPD) or 2% galactose (YPGal) supplemented with uridine (25 mg/ml) for the growth of C. albicans Ura auxotrophic strains. These overnight cultures were diluted to an OD600 of 0.1 in their respective media and grown to an OD600 of 1.0 (approximately three generations). The wild-type strain SC5314 [37] was either grown in YPD or in YPGal media. The overnight cultures were inoculated from a fresh colony and were grown in either YPD or in YPGal at 30 C. These overnight cultures were diluted to OD600 of 0.1 in their respective media: overnight YPD and YPGal-grown cells were diluted into a fresh YPD and YPGal media and grown to an OD600 of 1.0. b-galactosidase and Growth Assays b-galactosidase assays were performed as described in [43]. All of the assays were performed on standard synthetic yeast media (synthetic complete: 0.67% YNB, 0.2% amino acid mix) that contained 2% dextrose, 2% galactose, or 2% raffinose as carbon source.

Received: February 22, 2007 Revised: May 7, 2007 Accepted: May 8, 2007 Published online: May 31, 2007 References 1. Heckman, D.S., Geiser, D.M., Eidell, B.R., Stauffer, R.L., Kardos, N.L., and Hedges, S.B. (2001). Molecular evidence for the early colonization of land by fungi and plants. Science 293, 1129– 1133. 2. Tanay, A., Regev, A., and Shamir, R. (2005). Conservation and evolvability in regulatory networks: The evolution of ribosomal regulation in yeast. Proc. Natl. Acad. Sci. USA 102, 7203–7208. 3. Leloir, L.F. (1951). The enzymatic transformation of uridine diphosphate glucose into a galactose derivative. Arch Biochem. 33, 186–190. 4. Johnston, M. (1987). A model fungal gene regulatory mechanism: The GAL genes of Saccharomyces cerevisiae. Microbiol. Rev. 51, 458–476. 5. Lohr, D., Venkov, P., and Zlatanova, J. (1995). Transcriptional regulation in the yeast GAL gene family: A complex genetic network. FASEB J. 9, 777–787. 6. Hwang, D., Smith, J.J., Leslie, D.M., Weston, A.D., Rust, A.G., Ramsey, S., de Atauri, P., Siegel, A.F., Bolouri, H., Aitchison, J.D., et al. (2005). A data integration methodology for systems biology: Experimental verification. Proc. Natl. Acad. Sci. USA 102, 17302–17307. 7. Douglas, H.C. (1961). A mutation in saccharomyces that affects phosphoglucomutase activity and galactose utilization. Biochim. Biophys. Acta 52, 209–211. 8. Douglas, H.C., and Hawthorne, D.C. (1966). Regulation of genes controlling synthesis of the galactose pathway enzymes in yeast. Genetics 54, 911–916. 9. Tschopp, J.F., Emr, S.D., Field, C., and Schekman, R. (1986). GAL2 codes for a membrane-bound subunit of the galactose permease in Saccharomyces cerevisiae. J. Bacteriol. 166, 313– 318. 10. Bassel, J., and Mortimer, R. (1971). Genetic order of the galactose structural genes in Saccharomyces cerevisiae. J. Bacteriol. 108, 179–183. 11. Laughon, A., and Gesteland, R.F. (1982). Isolation and preliminary characterization of the GAL4 gene, a positive regulator of transcription in yeast. Proc. Natl. Acad. Sci. USA 79, 6827–6831. 12. Timson, D.J., Ross, H.C., and Reece, R.J. (2002). Gal3p and Gal1p interact with the transcriptional repressor Gal80p to form a complex of 1:1 stoichiometry. Biochem. J. 363, 515–520.

Transcriptional Rewiring of Galactose Metabolism 1013

13. Peng, G., and Hopper, J.E. (2002). Gene activation by interaction of an inhibitor with a cytoplasmic signaling protein. Proc. Natl. Acad. Sci. USA 99, 8548–8553. 14. Platt, A., and Reece, R.J. (1998). The yeast galactose genetic switch is mediated by the formation of a Gal4p-Gal80p-Gal3p complex. EMBO J. 17, 4086–4091. 15. West, R.W., Jr., Yocum, R.R., and Ptashne, M. (1984). Saccharomyces cerevisiae GAL1–GAL10 divergent promoter region: Location and function of the upstream activating sequence UASG. Mol. Cell. Biol. 4, 2467–2478. 16. Bevan, P., and Douglas, H.C. (1969). Genetic control of phosphoglucomutase variants in Saccharomyces cerevisiae. J. Bacteriol. 98, 532–535. 17. Jones, T., Federspiel, N.A., Chibana, H., Dungan, J., Kalman, S., Magee, B.B., Newport, G., Thorstenson, Y.R., Agabian, N., Magee, P.T., et al. (2004). The diploid genome sequence of Candida albicans. Proc. Natl. Acad. Sci. USA 101, 7329–7334. 18. van Het Hoog, M., Rast, T.J., Martchenko, M., Grindle, S., Dignard, D., Hogues, H., Cuomo, C., Berriman, M., Scherer, S., Magee, B.B., et al. (2007). Assembly of the Candida albicans genome into sixteen supercontigs aligned on the eight chromosomes. Genome Biol. 8, R52. 19. Braun, B.R., van Het Hoog, M., d’Enfert, C., Martchenko, M., Dungan, J., Kuo, A., Inglis, D.O., Uhl, M.A., Hogues, H., Berriman, M., et al. (2005). A human-curated annotation of the Candida albicans genome. PLoS Genet. 1, 36–57. 20. Martchenko, M., Levitin, A., and Whiteway, M. (2007). Transcriptional activation domains of the Candida albicans Gcn4p and Gal4p homologs. Eukaryot. Cell 6, 291–301. 21. Fan, J., Chaturvedi, V., and Shen, S.H. (2002). Identification and phylogenetic analysis of a glucose transporter gene family from the human pathogenic yeast Candida albicans. J. Mol. Evol. 55, 336–346. 22. Maidan, M.M., Thevelein, J.M., and Van Dijck, P. (2005). Carbon source induced yeast-to-hypha transition in Candida albicans is dependent on the presence of amino acids and on the G-protein-coupled receptor Gpr1. Biochem. Soc. Trans. 33, 291–293. 23. Brown, V., Sexton, J.A., and Johnston, M. (2006). A glucose sensor in Candida albicans. Eukaryot. Cell 5, 1726–1737. 24. Munro, C.A., Selvaggini, S., de Bruijn, I., Walker, L., Lenardon, M.D., Gerssen, B., Milne, S., Brown, A.J., and Gow, N.A. (2007). The PKC, HOG and Ca2+ signalling pathways co-ordinately regulate chitin synthesis in Candida albicans. Mol. Microbiol. 63, 1399–1413. 25. Singh, V., Sinha, I., and Sadhale, P.P. (2005). Global analysis of altered gene expression during morphogenesis of Candida albicans in vitro. Biochem. Biophys. Res. Commun. 334, 1149–1158. 26. Liu, H., Kohler, J., and Fink, G.R. (1994). Suppression of hyphal formation in Candida albicans by mutation of a STE12 homolog. Science 266, 1723–1726. 27. Jin, Y., Samaranayake, L.P., Samaranayake, Y., and Yip, H.K. (2004). Biofilm formation of Candida albicans is variably affected by saliva and dietary sugars. Arch. Oral Biol. 49, 789–798. 28. Tsong, A.E., Tuch, B.B., Li, H., and Johnson, A.D. (2006). Evolution of alternative transcriptional circuits with identical logic. Nature 443, 415–420. 29. Ihmels, J., Bergmann, S., Berman, J., and Barkai, N. (2005). Comparative gene expression analysis by differential clustering approach: Application to the Candida albicans transcription program. PLoS Genet. 1, e39. 30. Piskur, J., and Langkjaer, R.B. (2004). Yeast genome sequencing: The power of comparative genomics. Mol. Microbiol. 53, 381–389. 31. Johnston, M. (1999). Feasting, fasting and fermenting. Glucose sensing in yeast and other cells. Trends Genet. 15, 29–33. 32. Dong, J., and Dickson, R.C. (1997). Glucose represses the lactose-galactose regulon in Kluyveromyces lactis through a SNF1 and MIG1- dependent pathway that modulates galactokinase (GAL1) gene expression. Nucleic Acids Res. 25, 3657– 3664. 33. Platt, A., Ross, H.C., Hankin, S., and Reece, R.J. (2000). The insertion of two amino acids into a transcriptional inducer

34. 35.

36.

37. 38.

39.

40.

41.

42.

43. 44.

45.

converts it into a galactokinase. Proc. Natl. Acad. Sci. USA 97, 3154–3159. Bosch, A.M. (2006). Classical galactosaemia revisited. J. Inherit. Metab. Dis. 29, 516–525. John, M., Leppik, R., Busch, S.J., Granger-Schnarr, M., and Schnarr, M. (1996). DNA binding of Jun and Fos bZip domains: Homodimers and heterodimers induce a DNA conformational change in solution. Nucleic Acids Res. 24, 4487–4494. Wilson, R.B., Davis, D., and Mitchell, A.P. (1999). Rapid hypothesis testing with Candida albicans through gene disruption with short homology regions. J. Bacteriol. 181, 1868–1874. Fonzi, W.A., and Irwin, M.Y. (1993). Isogenic strain construction and gene mapping in Candida albicans. Genetics 134, 717–728. Gola, S., Martin, R., Walther, A., Dunkler, A., and Wendland, J. (2003). New modules for PCR-based gene targeting in Candida albicans: Rapid and efficient gene targeting using 100 bp of flanking homology region. Yeast 20, 1339–1347. Russell, C.L., and Brown, A.J. (2005). Expression of one-hybrid fusions with Staphylococcus aureus lexA in Candida albicans confirms that Nrg1 is a transcriptional repressor and that Gcn4 is a transcriptional activator. Fungal Genet. Biol. 42, 676–683. Murad, A.M., Lee, P.R., Broadbent, I.D., Barelle, C.J., and Brown, A.J. (2000). CIp10, an efficient and convenient integrating vector for Candida albicans. Yeast 16, 325–327. Nantel, A., Rigby, T., Hogues, H., and Whiteway, M. (2006). Microarrays for studying pathology in Candida albicans. In Medical Mycology: Cellular and Molecular Techniques, K. Kavanaugh, ed. (Hoboken, NJ: Wiley Press). Carlson, M., and Botstein, D. (1982). Two differentially regulated mRNAs with different 50 ends encode secreted with intracellular forms of yeast invertase. Cell 28, 145–154. Rupp, S. (2002). LacZ assays in yeast. Methods Enzymol. 350, 112–131. Martchenko, M., Alarco, A.M., Harcus, D., and Whiteway, M. (2004). Superoxide dismutases in Candida albicans: Transcriptional regulation and functional characterization of the hyphalinduced SOD5 gene. Mol. Biol. Cell 15, 456–467. Diezmann, S., Cox, C.J., Schonian, G., Vilgalys, R.J., and Mitchell, T.G. (2004). Phylogeny and evolution of medical species of Candida and related taxa: A multigenic analysis. J. Clin. Microbiol. 42, 5624–5635.