JOURNAL OF BACTERIOLOGY,0021-9193/98/$04.0010

Apr. 1998, p. 2079–2086 Vol. 180, No. 8

Copyright © 1998, American Society for Microbiology

Isolation and Characterization of EPD1, an Essential Gene forPseudohyphal Growth of a Dimorphic Yeast, Candida maltosa

TAKANOBU NAKAZAWA, HIROYUKI HORIUCHI, AKINORI OHTA, AND MASAMICHI TAKAGI*

Department of Biotechnology, The University of Tokyo, Yayoi, Bunkyo-ku, Tokyo 113, Japan

Received 21 August 1997/Accepted 17 February 1998

Additional copies of the centromeric DNA (CEN) region induce pseudohyphal growth in a dimorphic yeast,Candida maltosa (T. Nakazawa, T. Motoyama, H. Horiuchi, A. Ohta, and M. Takagi, J. Bacteriol. 179:5030–5036, 1997). To understand the mechanism of this transition, we screened the gene library of C. maltosa forsequences which could suppress this morphological change. As a result, we isolated the 5* end of a new gene,EPD1 (for essential for pseudohyphal development), and then cloned the entire gene. The predicted amino acidsequence of Epd1p was highly homologous to those of Ggp1/Gas1/Cwh52p, a glycosylphosphatidylinositol-anchored protein of Saccharomyces cerevisiae, and Phr1p and Phr2p of Candida albicans. The expression ofEPD1 was moderately regulated by environmental pH. A homozygous EPD1 null mutant showed some mor-phological defects and reduction in growth rate and reduced levels of both alkali-soluble and alkali-insolubleb-glucans. Moreover, the mutant could not undergo the transition from yeast form to pseudohyphal forminduced by additional copies of the CEN sequence at pH 4 or by n-hexadecane at pH 4 or pH 7, suggesting thatEPD1 is not essential for yeast form growth but is essential for transition to the pseudohyphal form. Over-expression of the amino-terminal part of Epd1p under the control of the GAL promoter suppressed thepseudohyphal development induced by additional copies of the CEN sequence, whereas overexpression of thefull-length EPD1 did not. This result and the initial isolation of the 5* end of EPD1 as a suppressor of thepseudohyphal growth induced by the CEN sequence suggest that the amino-terminal part of Epd1p may havea dominant-negative effect on the functions of Epd1p in the pseudohyphal growth induced by the CENsequence.

Dimorphic fungi exhibit either a yeast-like form or a(pseudo)hyphal form mainly in response to environmentalconditions, and the switch from a yeast-like form to a filamen-tous form often correlates with their pathogenicities. Candidaalbicans is a typical dimorphic fungus which can cause life-threatening infections in immunocompromised patients. In thisorganism, the dimorphic transition is thought to be critical forpathogenesis, as an elongated hyphal form facilitates tissuepenetration (47–49). Environmental conditions for a dimor-phic transition have also been studied in the related speciesCandida tropicalis (52). Saccharomyces cerevisiae also shows adimorphic transition (12, 20, 23, 43, 57), which is inducedmainly by nitrogen starvation (12). A number of genes havebeen shown to participate in the dimorphism of C. albicans andS. cerevisiae. In S. cerevisiae, it has been reported that a mito-gen-activated protein kinase (MAPK) is involved in pseudohy-phal development and that nitrogen starvation and activatedRas proteins stimulate pseudohyphal growth through bothMAPK-dependent and MAPK-independent pathways (7, 23,25, 29). In addition, several genes have been reported whoseregulation influences pseudohyphal growth (2, 3, 9, 11, 56).Furthermore, a variety of genes, such as PHR1, HYR1, CPH1(ACPR), CST20, HST7, EFG1, RBF1, and TUP1, have beenshown to affect dimorphism in C. albicans (1, 4, 15, 18, 21, 22,24, 26, 30, 45, 50). In spite of these investigations, the mecha-nisms of dimorphism are still not clear.

Candida maltosa is a dimorphic fungus with a diploid ge-nome (14, 19, 31), and it has been shown by phylogenetic

analysis to be closely related to C. albicans (34). For recombi-nant DNA technology, host-vector systems have been con-structed in our laboratory (13, 16, 35–37, 51). Recently, wefound that a part of the centromeric DNA (CEN) region, whenpresent on a plasmid, induces pseudohyphal growth in thisyeast. It is suggested that some trans-acting factors which in-teract with a sequence essential for centromeric activity, GGTAGCG, are involved in the regulation of the transition fromthe yeast form to the pseudohyphal form (31). To help under-stand the mechanism of this transition, we screened a genelibrary of C. maltosa for DNA sequences which could suppressthe pseudohyphal growth that is induced by additional copiesof the CEN sequence.

MATERIALS AND METHODS

Strains and media. C. maltosa IAM12247 (the wild-type strain) and its deriv-atives, CHA1 (his5 ade1) (16) and CHAU1 (his5 ade1 ura3) (37), were used. Themedia for C. maltosa were YPD (1% yeast extract, 2% Bacto Peptone, 2%glucose), SD (pH 4) [0.17% yeast nitogen base without amino acid or ammoniumsulfate (Difco), 0.5% (NH4)2SO4, 2% glucose, appropriate nutrients], SG (pH 4)(SD in which the glucose was replaced with 2% galactose), and hexadecanemedium (pH 4) (the same as SD and SG, except that n-hexadecane, which issupplied as vapor, was the sole carbon source). The SD (pH 7), SG (pH 7), andhexadecane (pH 7) media were buffered with 150 mM HEPES and adjusted topH 7.0. When necessary, agar was added to a concentration of 2%.

Escherichia coli MV1190 [D(srl-recA)306::Tn10(Tetr) D(lac-pro) thi supE (F9proAB lacIq lacZDM15 traD36)], HB101 [hsdS20(r2 m2) recA13 ara-14 proA2lacY1 galK2 rpsL20(Smr) xyl-5 mtl-1 l2 mcrA1 mcrB], and DH5 (supE44 hsdR1recA1 endA1 gyrA96 thi-1 relA1) were grown in Luria-Bertani broth and used ashosts for the propagation of plasmids or the construction of genomic and sub-genomic libraries of C. maltosa.

Plasmid construction and transformation. The construction of pUAH2A andpUAHHH1 was described previously (31). Ligation of DNA ends that were notcohesive with each other was done after blunt ending them with T4 DNApolymerase. The DraI-EcoT22I fragment of URA3 of C. maltosa, the AatII-ClaIfragment containing ARS (13, 51), and the 208-bp HincII-HindIII fragment ofthe CEN sequence were ligated into the NaeI site, the NspI site, and the EcoRIsite, respectively, of pBluescript II KS(1) to yield the plasmid pBLU1. The

* Corresponding author. Mailing address: Department of Biotech-nology, The University of Tokyo, Yayoi, Bunkyo-ku, Tokyo 113, Japan.Phone: 81-3-3812-2111, ext. 5169. Fax: 81-3-3812-9246. E-mail:amtakag@hongo.ecc.u-tokyo.ac.jp.

2079

on February 9, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

AatII-ClaI fragment of ARS and the DraIII-SspI fragment of ADE1 (16) of C.maltosa were ligated into the EcoO109I-SspI site and the SalI site, respectively,of pUC119 to yield the plasmid pUAA10. The XbaI-XbaI fragment containingEPD1 (see Fig. 4) was ligated into the XbaI site of pUC119, and then theDraIII-SspI fragment of ADE1 (16) was inserted into the PvuII site of EPD1 toyield the plasmid pEPD1::ADE1. The same XbaI-XbaI fragment containingEPD1 was ligated into the XbaI site of pUC119, and then the SalI fragmentcontaining HIS5 (13) was inserted into the PvuII site of EPD1 to yield theplasmid pEPD1::HIS5. The EcoRI-BstPI fragment containing the full-lengthEPD1 (see Fig. 4) and the XbaI-XbaI fragment of the GAL promoter (39) wereligated into the SmaI site and the XbaI site of pUAA10, respectively, to yield theplasmid pUAAEB5. The EcoRI-BglII fragment of the amino-terminal part ofEpd1p (see Fig. 4) and the XbaI-XbaI fragment of the GAL promoter (39) wereligated into the SmaI site of pUAA10 to yield the plasmid pUAAEBG3. TheEcoRI-SalI, XbaI-EcoRV, and SalI-XbaI fragments of pPHS1 (see Fig. 2) wereligated into the SmaI site of pUAA10 to construct plasmids pPHS1ES,pPHS1XE, and pPHS1SX, respectively.

Transformation of C. maltosa was performed by electroporation as describedpreviously (27). DNA manipulations and E. coli transformation were done asdescribed previously (44). DNA enzymes were purchased from Takara Shuzo Co.(Ohtu, Japan) and used according to the manufacturer’s instructions.

Construction of genomic and subgenomic libraries of C. maltosa. Total DNAwas prepared from C. maltosa IAM12247 as described previously (36). Afterpartial digestion with Sau3AI, the total DNA was size fractionated by sucrosedensity gradient centrifugation. The DNA fragments of 6 to 8 kb were ligated toBamHI-digested and phosphatase-treated pUAA10 and transformed into E. coliDH5. A subgenomic library on pUC119 was constructed as follows. Total DNAof C. maltosa IAM12247 was digested with EcoRI and electrophoresed througha 0.6% agarose gel. A piece of the gel containing EcoRI fragments of around 4.5kb was cut out, and the DNA contained in it was eluted, purified and ligated withEcoRI-digested and phosphatase-treated pUC119. E. coli MV1190 was trans-formed with the ligated DNA, and approximately 5,000 ampicillin-resistant col-onies were selected and used for screening by colony hybridization.

Deletion of EPD1. pEPD1::HIS5 was digested with ApaLI and used to trans-form C. maltosa CHAU1. Stable His1 transformants were selected to obtain thestrain UEP11. pEPD1::ADE1 was digested with ApaLI and used to transformstrain UEP11. Stable His1 and Ade1 transformants were selected to obtain thestrain UEP21 (epd1/epd1). Disruption of the EPD1 gene in these strains wasconfirmed by Southern hybridization and PCR.

Glucan analysis. The alkali-insoluble and -soluble 1,3- and 1,6-b-D-glucanswere isolated and quantified as described by Popolo et al. (40).

Southern blot analysis. Total DNA of C. maltosa was isolated from a 10-mlculture grown for 15 h in selective medium as described previously (36). South-ern blot analysis was performed by using the Amersham ECL direct nucleic acidlabelling and detection system according to the instructions of the supplier. A548-bp EcoRV-PvuII fragment of EPD1 was used as a probe. Colony hybridiza-tion was carried out with a Hybond-N1 membrane (Amersham) as follows. AnEcoRI-XbaI fragment from the insert of pPHS1SX was labelled with [a-32P]dCTP(Amersham) by use of a random primer labelling kit (Takara) and used as aprobe. Hybridization and membrane washing were carried out by the methodsuggested by the membrane supplier.

Northern blot analysis. Total yeast RNA from various culture conditions wasextracted by the method of Schmitt et al. (46), separated by agarose gel electro-phoresis, blotted on a Hybond-N membrane (Amersham), and analyzed. North-ern hybridization was carried out as described previously (38).

DNA sequence analysis of EPD1. Appropriate restriction fragments of theEPD1 gene were subcloned into the pUC series plasmids, and DNA sequencingwas carried out with an automated DNA sequencer (LI-COR model 4000L).

Nucleotide sequence accession number. The nucleotide sequence data re-ported in this paper have been submitted to the DDBJ, EMBL, and GenBanknucleotide sequence databases under accession no. AB005130.

RESULTS

Isolation of a multicopy suppressor of pseudohyphal growthinduced by additional copies of the CEN sequence. Previously,we found that the CEN region on a plasmid could inducepseudohyphal growth when it was introduced into C. maltosa(Fig. 1A) (31). To analyze the mechanism of induction, wecarried out screening for genes that can suppress pseudohyphalgrowth under these conditions. First, we confirmed that thehost strain does not show pseudohyphal growth on SG agarplates at both pH 4 and pH 7 but that the host strain containingpUAHHH1 carrying the truncated CEN sequence does show iton SG agar plates at both pHs, although the pseudohyphalmorphology is more evident at pH 4. Then the latter strain wastransformed with a C. maltosa genomic library constructed inthe high-copy-number vector pUAA10. After 3 days, about

FIG. 1. Colony morphology of C. maltosa CHA1 containing plasmidspUAHHH1 and pUAA10 (vector alone) (A), pUAHHH1 and pPHS1 (suppres-sor) (B), and pUAHHH1 and pPHS1SX (subclone of the insert of pPHS1) (C).The cells were grown on SG agar plates (pH 4) for 3 days.

FIG. 2. Delimitation of the inserted C. maltosa DNA of pPHS1. Colony morphology was observed after 3 days on SG agar plates (pH 4). 1, suppression of thepseudohyphal growth induced by additional copies of the CEN sequence; 2, no suppression of the pseudohyphal growth.

2080 NAKAZAWA ET AL. J. BACTERIOL.

on February 9, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

35,000 transformants were obtained on SG (pH 4) agar platesand examined microscopically. Most colonies showed pseudo-hyphal growth, but four showed reduced pseudohyphal growth.Plasmids were isolated from these four colonies and reintro-duced into C. maltosa CHA1 containing pUAHHH1, and thesuppression of pseudohyphal growth by these plasmids wasconfirmed. One of the plasmids was designated pPHS1 (Fig.1B). Restriction enzyme analyses and subcloning of the insertof pPHS1 were performed as shown in Fig. 2. The plasmidpPHS1SX, which carried the smallest fragment, suppressedpseudohyphal growth at the same level as pPHS1 (Fig. 1C).Only a part of one open reading frame (ORF) and its possiblepromoter region were found in the inserted DNA ofpPHS1SX. The deduced amino acid sequence of the ORF ishomologous to the N-terminal parts of the S. cerevisiae proteinGgp1/Gas1/Cwh52p and of the C. albicans proteins Phr1p andPhr2p. In the other three isolated plasmids, the HIS5 gene wascloned. The transformants with these YRp-type plasmids car-rying ADE1 and HIS5 had lost pUAHHH1 carrying HIS5 andthe truncated CEN sequence, and hence, they did not showpseudohyphal growth. We next investigated whether this newlyisolated plasmid (pPHS1) could suppress the pseudohyphaltransition induced by n-hexadecane, which we found to be astrong inducer of the transition (see Fig. 6 and 7). We did notobserve suppression by the plasmid under these conditions(data not shown).

Isolation of the entire EPD1 gene. To clone the full-lengthORF, the EcoRI-XbaI fragment from the insert of pPHS1SXwas used as a hybridization probe. Southern blot analysis onEcoRI digests of C. maltosa IAM12247 genomic DNA wascarried out under conditions of high stringency, and one pos-itive band at 4.5 kb was detected (data not shown). Then, theC. maltosa subgenomic DNA library that contained a DNAfragment of about 4.5 kb was constructed in pUC119 andscreened by colony hybridization with the same probe. Onepositive clone was obtained, and the isolated plasmid was des-ignated pPCO10. Subcloning and sequencing of the insertedDNA in this plasmid revealed a single, uninterrupted ORF of1,650 nucleotides, which could encode a 549-residue protein.We designated this gene EPD1. The plasmid pPHS1SX con-tained the promoter region and the region encoding aminoacids Met-1 to Asn-126 of Epd1p. The deduced amino acidsequence of the entire Epd1p is shown in Fig. 3.

Three proteins with amino acid sequences similar to that ofthe full-length Epd1p were identified from the databases by theFASTA program. They were products of S. cerevisiae GGP1/GAS1/CWH52 and C. albicans PHR1 and PHR2. The Epd1protein was 58.4, 56.0, and 74.0% identical to the proteinsGgp1/Gas1/Cwh52p, Phr1p, and Phr2p, respectively. GGP1/GAS1/CWH52 codes for a GPI-anchored plasma membraneglycoprotein (6, 32, 33, 40–42, 53–55), and PHR1 encodes aputative GPI-anchored cell surface glycoprotein, production ofwhich is differentially regulated in response to the pH of thegrowth medium (45). PHR2 encodes a functional homolog ofPHR1, and its expression is also regulated by pH, but in a wayexactly the inverse of that of PHR1 (30). Mutants of thesegenes showed defects in general morphogenesis (30, 40, 41,45). The four homologous proteins showed similarity alongtheir entire sequences. Several regions which seemed function-ally significant were well conserved among them (Fig. 3).Epd1p has a hydrophobic amino terminus, characteristic ofsecretory signal sequences, and a hydrophobic carboxy termi-nus, characteristic of GPI-linked proteins (8, 28). The Ser-523,Ala-524, and Ala-525 residues of Epd1p correspond to the v,v11, and v12 sites of the consensus GPI attachment site,

respectively (10, 32). The serine-rich region located near thecarboxy terminus is also conserved.

Disruption of EPD1. An epd1 null mutant was constructed toanalyze the role of EPD1 in morphogenesis. Two EPD1 allelesin C. maltosa CHAU1 were inactivated by successive genereplacements with the insertion alleles epd1::ADE1 andepd1::HIS5 (Fig. 4A). Single and double EPD1 mutants weregenerated as described in Materials and Methods and desig-nated UEP11 (EPD1/epd1) and UEP21 (epd1/epd1), respec-tively. Replacement of the EPD1 gene by homologous recom-bination was confirmed by hybridization to a genomicSouthern blot (Fig. 4B) and by PCR (data not shown). The nullmutants were viable, indicating that EPD1 is not essential for

FIG. 3. Comparison of predicted amino acid sequences of Epd1p, Ggp1/Gas1/Cwh52p, Phr1p, and Phr2p. Identical and similar residues are indicated byasterisks and dots, respectively. The hydrophobic amino and carboxy termini areboxed. The proposed GPI attachment sites of Epd1p and Ggp1/Gas1/Cwh52pare indicated by bold letters, and the serine-rich regions are underlined. Amino-acid N-126, from which the EPD1 gene on plasmid pPHS1SX is truncated (seeResults), is boxed inversely.

VOL. 180, 1998 EPD1 ESSENTIAL FOR PSEUDOHYPHAL GROWTH OF C. MALTOSA 2081

on February 9, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

cell viability. The duplication times of strains CHAU1 andUEP21 during the exponential growth phase in SD (pH 4)were 80 and 120 min, respectively.

The yeast form cells of these strains grown in SD (pH 4)were examined microscopically. UEP11 (EPD1/epd1) cellswere somewhat larger than the wild-type cells and slightlyround. At the early time of incubation, UEP21 (epd1/epd1)cells were larger than the wild-type cells and round, and theyfrequently had two buds instead of one. This phenotype issimilar to that observed in the ggp1 mutant of S. cerevisiae andthe C. albicans phr1/phr1 and phr2/phr2 mutants (30, 40, 45).After prolonged incubation, these morphological characteris-tics of the mutants became more evident (Fig. 5). These results

indicate that EPD1 is involved in maintaining normal yeastform growth in SD at pH 4. The morphology of the yeast formcells of UEP21 (epd1/epd1) was not much different from that ofthe wild-type cells in SD at pH 7 (data not shown).

Next, we examined the effect of EPD1 disruption onpseudohyphal growth. The wild-type strain CHAU1 (Fig. 6A)and UEP11 (data not shown), when they contained pBLU1(containing a part of the CEN sequence), showed pseudohy-

FIG. 4. Disruption of EPD1. (A) Construction of the disrupted genes epd1::ADE1 and epd1::HIS5. The detailed methods for the disruption are described inMaterials and Methods. (B) Southern blot analysis of the EPD1 mutants. Genomic DNAs from the indicated strains were digested with XbaI and hybridized with thelabelled EPD1 DNA region indicated in panel A as a probe. The electrophoretic positions of DNA size markers are indicated on the left.

FIG. 5. Effect of disruption of EPD1 on yeast form growth of C. maltosaCHA1 (EPD1/EPD1) (A), UEP11 (EPD1/epd1) (B), and UEP21 (epd1/epd1)(C). Cells were grown in SD (pH 4) until the early stationary phase.

FIG. 6. Effect of disruption of EPD1 on colony morphology at pH 4. (A andB) CHAU1 bearing pBLU1 and UEP21 containing pBLU1, respectively, weregrown on SG agar plates (pH 4) for 1 day. (C and D) CHAU1 and UEP21,respectively, grown on plates containing n-hexadecane as a sole carbon source(pH 4) for 3 days.

2082 NAKAZAWA ET AL. J. BACTERIOL.

on February 9, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

phal growth on SG agar plates (pH 4) after 1 day of incubation,whereas the homozygous mutant, UEP21, containing pBLU1did not (Fig. 6B). In addition, strain CHAU1 growing on n-hexadecane as a sole carbon source (pH 4) showed pro-nounced pseudohyphal growth in 3 days (Fig. 6C), while strainUEP21 could not undergo a transition from the yeast form tothe pseudohyphal form (Fig. 6D). Strain UEP11 showed onlya slight transition to the pseudohyphal form on hexadecaneplates (pH 4) (data not shown). In contrast to the C. albicansphr2 null mutant, which failed to grow at pH 4 (30), the epd1null mutant could grow in SG (pH 4) and hexadecane medium(pH 4) as well as in SD (pH 4), where the duplication time ofUEP21 (120 min) was 1.5 times that of the wild-type strain (80min). Thus, the incapability of strain UEP21 to demonstratepseudohyphal growth cannot be attributed to its growth defect.We also examined morphological characteristics of thesestrains on SG agar plates (pH 7). Strain CHAU1 containingpBLU1 showed pseudohyphal growth (Fig. 7A), although thefrequency of the transition was not as high as that seen at pH4. To our surprise, strain UEP21 containing pBLU1 showedpseudohyphal growth equal to that of the parent strain (Fig.7B). On the other hand, UEP21 could not undergo a transitionfrom yeast form to pseudohyphal form on hexadecane plates(pH 7) (Fig. 7D), in contrast to the pseudohyphal growth of theparent strain (Fig. 7C).

In the ggp1 mutant of S. cerevisiae, the content of alkali-insoluble 1,6-b-D-glucan was shown to be about 50% of that ofthe wild-type strain (40). Therefore, we also examined theamount of glucan in the epd1 mutants at their entry into sta-

tionary phase in SD (pH 4) (Table 1). In strain UEP21, thetotal amount of alkali-insoluble fraction decreased and theamount of alkali-insoluble 1,6-b-D-glucan was about 50% ofthat in the wild-type strain. The carbohydrate content of thealkali-insoluble material was obtained after Zymolyase diges-tion (Table 1). This fraction was increased in strain UEP21 incomparison with that of the wild-type strain. The carbohydratecontents of all fractions of strain UEP11 were similar to thoseof the wild-type strain. The carbohydrate content of the alkali-soluble fraction of strain UEP21 was lower than that of thewild-type strain; this was different from the case of the S.cerevisiae ggp1 mutant, where it was about 150% of that of thewild-type strain (40).

Transcriptional regulation of EPD1. We also examined thetranscriptional regulation of EPD1 by Northern blot analysis.Total RNA was isolated from CHA1 containing the plasmidpUAH2A, which does not induce pseudohyphal growth, or thecentromere plasmid pUAHHH1, which induces pseudohyphalgrowth at various growth phases (early log phase and mid-logphase) in SG (pH 4). The transcriptional level was not influ-enced by the growth phase or the pseudohyphal morphogen-esis, and the transcript was not detected in stationary-phasecells (data not shown). We also examined the effect of pH onthe transcriptional regulation of EPD1. Total RNA was iso-lated at early log phase (optical densities at 600 nm [OD600],0.5 and 2) in SD (pH 4) and SD (pH 7). The transcripts weredetected in all samples, and the level of EPD1 mRNA wasreduced in the cells grown at pH 7 (Fig. 8). This result is similarto that found with PHR2 (30). However, the expression ofEPD1 is not so strictly regulated by the environmental pHsignal as is that of PHR2.

The amino-terminal part of Epd1p has a dominant-negativeeffect on the pseudohyphal growth induced by the CEN se-quence. The 59 end of EPD1 was initially isolated as a sup-pressor of the pseudohyphal growth induced by additionalcopies of the CEN sequence. To confirm that this phenotypewas due to the truncated EPD1 gene, we did the followingexperiments. Plasmids pUAAEB5 and pUAAEBG3, which ex-press full-length and truncated EPD1, respectively, under thecontrol of the GAL promoter (39), were constructed and trans-formed into the wild-type strain CHA1 containing pUAHHH1.After 3 days of incubation on SG (pH 4), transformants wereexamined under the microscope. CHA1 containing bothpUAHHH1 and pUAAEB5 showed pseudohyphal growth,whereas CHA1 containing both pUAHHH1 and pUAAEBG3did not (Fig. 9). This result and the initial isolation of the 59end of EPD1 suggest that the overexpression of only the ami-no-terminal region of Epd1p in a multicopy plasmid has adominant-negative effect on pseudohyphal growth induced bythe extra CEN sequence. On the other hand, overexpression ofthe full-length EPD1 using the GAL promoter in CHA1 did

FIG. 7. Effect of disruption of EPD1 on colony morphology at pH 7. CHAU1carrying pBLU1 (A) and UEP21 bearing pBLU1 (B) were grown on SG (pH 7)plates for 1 day. CHAU1 (C) and UEP21 (D) were grown on hexadecanemedium (pH 7) for 2 and 3 days, respectively.

TABLE 1. b-Glucan contents of epd1D mutant cellsa

Strain

Alkali insoluble Alkalisoluble

1,3- 11,6-b-D-glucan

1,6-b-D-glucan ZUPb

1,3- 11,6-b-D-glucan

CHAU1 (EPD1/EPD1) 94 72 28 100UEP11 (EPD1/epd1) 96 75 30 107UEP21 (epd1/epd1) 61 43 62 87

a All fractions are expressed as micrograms per milligram (dry weight) of cells.b ZUP, Zymolyase-undigestible pellet.

VOL. 180, 1998 EPD1 ESSENTIAL FOR PSEUDOHYPHAL GROWTH OF C. MALTOSA 2083

on February 9, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

not induce pseudohyphal growth on SG agar plates (pH 4)(data not shown).

DISCUSSION

Cells of yeasts and fungi are surrounded by walls that areresponsible for the mechanical strength of the cell (17) and areessential for maintenance of the cell shape (58). Fungal cellwalls are composed primarily of three classes of polysaccha-rides: mannoproteins, b-glucans, and chitins (5). The b-glucansconstitute about 60% of the cell wall carbohydrate and arebelieved to be the most important elements in determining cellmorphology in S. cerevisiae (58) and probably in C. albicansand C. maltosa. The deduced protein product of EPD1 isolatedfrom C. maltosa in the present work contains several conservedstructural features of the GPI-anchored proteins, such as ahydrophobic signal sequence at the N terminus, a C-terminalhydrophobic domain, and a GPI attachment site (8, 10, 28, 32).These and other structural similarities suggest that Epd1p ismodified by the addition of GPI and that it is a homolog of theS. cerevisiae GGP1/GAS1/CWH52 gene product and the C.albicans PHR1 and PHR2 gene products. Although the primaryfunction of this family of genes is currently unknown (30, 40,45), it may be essential for the assembly of b-glucan, since thelack of EPD1 in C. maltosa and of GGP1 in S. cerevisiae (40)reduces the level of b-glucans (Table 1).

It was reported that C. albicans phr1/phr1 mutants displaydefects in the formation of germ tubes and apical growth ofhyphal forms at the restrictive pH. Prolonged incubation ofwild-type C. albicans at higher pH resulted in the mixture ofyeast form and pseudohyphal-form cells, but incubation of thephr1/phr1 null mutant did not result in pseudohyphal-formcells (45). The effect of EPD1 disruption is similar to the resultsobtained by Saporito-Irwin et al. (45), and the function ofEPD1 in the transition from yeast form to pseudohyphal formmay be similar to that of PHR1 in the morphological transitionof C. albicans. But, in the case of the induction of pseudohy-phae by n-hexadecane, the disruption of EPD1 affectedpseudohyphal morphology at both pH 4 and 7. Although thegrowth rate of the homozygous null mutant, UEP21, was lowerthan that of the wild-type strain in SD at pH 4 in the yeast

form, it was almost comparable to that of the pseudohypha-forming wild-type strain in hexadecane medium at pH 4, incontrast to the failure of growth of the disruptant of PHR2, ahomolog of EPD1. The incapability of strain UEP21 to showpseudohyphal growth, therefore, cannot be attributed to itsgrowth defect. UEP21 cells were round and larger than wild-type cells in SD (pH 4), as was also observed for ggp1, phr1, andphr2 mutants (30, 40, 45), which could be due to a lack of cellwall integrity caused by defects in b-glucan assembly. All theseresults suggest that the function of Epd1p may not be identical,but is quite similar, to those of the GGP1/GAS1/CWH52,PHR1, and PHR2 products.

We conclude that the increased expression of EPD1 may notbe required for pseudohyphal growth, not only because over-expression of the full-length EPD1 did not affect the transitionbut also because Northern blot analysis showed that the ex-pression level of EPD1 was not affected by the pseudohyphalmorphogenesis. EPD1 expression seems constitutive duringthe early and mid-log phases, in contrast to the expression ofHYR1, a C. albicans gene encoding a GPI-anchored surfaceprotein which is activated in response to hyphal development(1). The EPD1 gene product may have a more general functionin the transition from yeast form to pseudohyphal form. Wealso examined the effect of pH on EPD1 expression and found

FIG. 8. Northern blot analysis of the EPD1 transcript. C. maltosa CHA1 wasincubated in SD (pH 4) and SD (pH 7) media. RNA was isolated, and Northernblot hybridization was performed as described in Materials and Methods. Lane1, OD600 of 0.5 in SD (pH 4); lane 2, OD600 of 0.5 in SD (pH 7); lane 3, OD600of 2 in SD (pH 4); lane 4, OD600 of 2 in SD (pH 7). The upper panel displays theresults of hybridization with EPD1, and the lower panel displays the results ofethidium bromide (EtBr) staining of rRNA. The positions of 25S and 18S rRNAare shown on the right.

FIG. 9. Effect of GAL promoter-induced expression of EPD1 on colonymorphology. CHA1 containing pUAHHH1 and pUAAEB5 with full-lengthEPD1 (A) and pUAHHH1 and pUAAEBG3 with the amino-terminal part ofEPD1 (B) were grown on SG agar plates (pH 4) for 3 days.

2084 NAKAZAWA ET AL. J. BACTERIOL.

on February 9, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

that the expression of EPD1 is regulated by pH, similar to butnot as strictly as PHR2 (30). In C. maltosa, pseudohyphalgrowth is not as pronounced at pH 7 as it is at pH 4. This mightbe explained by the lower level of expression of EPD1 at pH 7.These results suggest that EPD1 works mainly at pH 4, andpartially at pH 7. In addition, on hexadecane plates, UEP11(EPD1/epd1) showed reduced pseudohyphal growth andUEP21 (epd1/epd1) did not show pseudohyphal growth even atpH 7. Pseudohyphal growth on hexadecane may requiregreater amounts of Epd1p than it does on SG agar plates.

The most interesting finding in this work was that the amino-terminal part of Epd1p has a dominant-negative effect on thepseudohyphal growth induced by the CEN sequence. It mayinhibit the functions of Epd1p in the pseudohyphal growthinduced by the CEN sequence in the cells. As the total glucancontent of the wild-type strain CHA1 that contains pUAAEBG3bearing the truncated EPD1 downstream of the GAL promoterwas identical to that of strain CHA1 containing vector pUAA10(data not shown), the expression of the amino-terminal part ofEpd1p might cause an abnormal architecture, but not an ab-normal composition, of the cell wall. However, the expressionof the amino terminus of Epd1p by its own promoter had nosuppressive effect on the pseudohyphal growth induced by n-hexadecane as a sole carbon source. These differential pheno-types might be due to the differences between the cell wallarchitectures of the pseudohypha-forming cells induced by n-hexadecane and those induced by the CEN sequence.

There are several examples suggesting that the dose of agene affects the filamentous growth of yeast. For example,hyphal formation is sensitive to the dose of the MAPK cascadecomponents in C. albicans (18). Our previous data indicatedthat three copies of YCp-type plasmid induce more markedpseudohyphal growth than the single copy in C. maltosa (31),and the present results indicate that a heterozygous disruptantof EPD1 shows only a slight transition from a yeast form to apseudohyphal form induced by n-hexadecane. These resultstaken together suggest that the copy numbers of some cellularcomponents are important for the induction of pseudohyphalgrowth.

EPD1 is not essential for growth, but it is essential for thepseudohyphal transition. Defining the precise role of Epd1pwill be helpful for understanding the mechanism of the dimor-phism of fungi.

ACKNOWLEDGMENTS

Part of this work was performed using facilities of the BiotechnologyResearch Center of The University of Tokyo. T.N. thanks the membersof the Takagi Laboratory for their helpful discussions and support.

REFERENCES

1. Bailey, D. A., P. J. F. Feldmann, M. Bovey, N. A. R. Gow, and A. J. P. Brown.1996. The Candida albicans HYR1 gene, which is activated in response tohyphal development, belongs to a gene family encoding yeast cell wall pro-teins. J. Bacteriol. 178:5353–5360.

2. Blacketer, M. J., C. M. Koehler, S. G. Coats, A. M. Myers, and P. Madaule.1993. Regulation of dimorphism in Saccharomyces cerevisiae: involvement ofthe novel protein kinase homolog Elm1p and protein phosphatase 2A. Mol.Cell. Biol. 13:5567–5581.

3. Blacketer, M. J., P. Madaule, and A. M. Myers. 1995. Mutational analysis ofmorphologic differentiation in Saccharomyces cerevisiae. Genetics 140:1259–1275.

4. Braun, B. R., and A. D. Johnson. 1997. Control of filament formation inCandida albicans by transcriptional repressor TUP1. Science 277:105–109.

5. Cabib, E., R. Roberts, A. Sburlati, and J. Silverman. 1982. Synthesis of yeastcell wall and its regulation. Annu. Rev. Biochem. 51:763–793.

6. Conzelmann, A., H. Riezman, C. Desponds, and C. Bron. 1988. A major125-kd membrane glycoprotein of Saccharomyces cerevisiae is attached to thelipid bilayer through an inositol-containing phospholipid. EMBO J. 7:2233–2240.

7. Cook, J. G., L. Bardwell, and J. Thorner. 1997. Inhibitory and activatingfunctions for MAPK Kss1 in the S. cerevisiae filamentous-growth signallingpathway. Nature 390:85–88.

8. Ferguson, M. A. J. 1988. Cell-surface anchoring of proteins via glycosylphos-phatidylinositol structures. Annu. Rev. Biochem. 57:285–320.

9. Gavrias, V., A. Andriannopoulos, C. J. Gimeno, and W. E. Timberlake. 1996.Saccharomyces cerevisiae TEC1 is required for pseudohyphal growth. Mol.Microbiol. 19:1255–1263.

10. Gerber, L. D., K. Kodukula, and S. Udenfriend. 1992. Phosphatidylinositolglycan (PI-G) anchored membrane proteins. J. Biol. Chem. 267:12168–12173.

11. Gimeno, C. J., and G. R. Fink. 1994. Induction of pseudohyphal growth byoverexpression of PHD1, a Saccharomyces cerevisiae gene related to tran-scriptional regulators of fungal development. Mol. Cell. Biol. 14:2100–2112.

12. Gimeno, C. J., P. O. Ljungdahl, C. A. Styles, and G. R. Fink. 1992. Unipolarcell divisions in the yeast S. cerevisiae lead to filamentous growth: regulationby starvation and RAS. Cell 68:1077–1090.

13. Hikiji, T., M. Ohkuma, M. Takagi, and K. Yano. 1989. An improved host-vector system for Candida maltosa using a gene isolated from its genome thatcomplements the his5 mutation of Saccharomyces cerevisiae. Curr. Genet.16:261–266.

14. Hofmann, K. H., and E. Polnisch. 1990. Characterization of a mutant of theyeast Candida maltosa defective in catabolite inactivation of gluconeogenicenzymes. Arch. Microbiol. 154:514–517.

15. Ishii, N., M. Yamamoto, F. Yoshihara, M. Arisawa, and Y. Aoki. 1997.Biochemical and genetic characterization of Rbf1p, a putative transcriptionfactor of Candida albicans. Microbiology 143:429–435.

16. Kawai, S., T. Hikiji, S. Murao, M. Takagi, and K. Yano. 1991. Isolation andsequencing of a gene, C-ADE1, and its use for a host-vector system inCandida maltosa with two genetic markers. Agric. Biol. Chem. 55:59–65.

17. Klis, F. M. 1994. Cell wall assembly in yeast. Yeast 10:851–869.18. Kohler, R. J., and G. R. Fink. 1996. Candida albicans strains heterozygous

for mutations in mitogen-activated protein kinase signaling components havedefects in hyphal development. Proc. Natl. Acad. Sci. USA 93:13223–13228.

19. Komagata, K., T. Nakase, and N. Katsuya. 1964. Assimilation of hydrocar-bons by yeasts. II. Determination of hydrocarbon-assimilating yeast. J. Gen.Appl. Microbiol. 4:323–331.

20. Kron, S. J., C. A. Styles, and G. R. Fink. 1994. Symmetric cell division inpseudohyphae of the yeast Saccharomyces cerevisiae. Mol. Biol. Cell 5:1003–1022.

21. Lambrechts, M. G., F. F. Bauer, J. Marmur, and I. S. Pretorius. 1996. Muc1,a mucin-like protein that is regulated by Mss10, is critical for pseudohyphaldifferentiation in yeast. Proc. Natl. Acad. Sci. USA 93:8419–8424.

22. Leberer, E., D. Harcus, I. D. Broadbent, K. L. Clark, D. Dignard, K.Ziegelbauer, A. Schmidt, N. A. R. Gow, A. J. P. Brown, and D. Y. Thomas.1996. Signal transduction through homologs of the Ste20p protein kinase cantrigger hyphal formation in the pathogenic fungus Candida albicans. Proc.Natl. Acad. Sci. USA 93:13217–13222.

23. Liu, H., C. A. Styles, and G. R. Fink. 1993. Elements of the yeast pheromoneresponse pathway required for filamentous growth of diploid. Science 262:1741–1744.

24. Liu, H., J. Kohler, and G. R. Fink. 1994. Suppression of hyphal formation inCandida albicans by mutation of a STE12 homolog. Science 266:1723–1725.

25. Madhani, H. D., and G. R. Fink. 1997. Combinational control required forthe specificity of yeast MAPK signaling. Science 275:1314–1317.

26. Malathi, K., K. Ganesan, and A. Datta. 1994. Identification of a putativetranscription factor in Candida albicans that can complement the matingdefect of Saccharomyces cerevisiae ste12 mutants. J. Biol. Chem. 269:22945–22951.

27. Mauersberger, S., M. Ohkuma, W.-H. Schunck, and M. Takagi. 1995. Can-dida maltosa, p. 411–580. In K. Wolf (ed.), Nonconventional yeasts in bio-technology. Springer, Heidelberg, Germany.

28. McConville, M. J., and M. A. J. Ferguson. 1993. The structure, biosynthesisand function of glycosylated phosphatidylinositols in the parasitic protozoaand higher eukaryotes. Biochem. J. 294:305–324.

29. Mosch, H.-U., R. L. Roberts, and G. R. Fink. 1996. Ras2 signals via theCdc42/Ste20/mitogen-activated protein kinase module to induce filamentousgrowth in Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. USA 93:5352–5356.

30. Muhlschlegel, F. A., and W. A. Fonzi. 1997. PHR2 of Candida albicansencodes a functional homolog of the pH-regulated gene PHR1 with aninverted pattern of pH-dependent expression. Mol. Cell. Biol. 17:5960–5967.

31. Nakazawa, T., T. Motoyama, H. Horiuchi, A. Ohta, and M. Takagi. 1997.Evidence that part of a centromeric DNA region induces pseudohyphalgrowth in a dimorphic yeast, Candida maltosa. J. Bacteriol. 179:5030–5036.

32. Nuoffer, C., A. Horvath, and H. Riezman. 1993. Analysis of the sequencerequirements for glycosylphosphatidylinositol anchoring of Saccharomycescerevisiae Gas1 protein. J. Biol. Chem. 268:10558–10563.

33. Nuoffer, C., P. Jeno, A. Conzelmann, and H. Riezman. 1991. Determinantsfor glycophospholipid anchoring of the Saccharomyces cerevisiae GAS1 pro-tein to the plasma membrane. Mol. Cell. Biol. 11:27–37.

34. Ohkuma, M., C. W. Hwang, Y. Masuda, H. Nishida, J. Sugiyama, A. Ohta,

VOL. 180, 1998 EPD1 ESSENTIAL FOR PSEUDOHYPHAL GROWTH OF C. MALTOSA 2085

on February 9, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

and M. Takagi. 1993. Evolutionary position of n-alkane-assimilating yeastCandida maltosa shown by nucleotide sequence of small-subunit ribosomalRNA gene. Biosci. Biotechnol. Biochem. 57:1793–1794.

35. Ohkuma, M., K. Kobayashi, S. Kawai, C. W. Hwang, A. Ohta, and M.Takagi. 1995. Identification of a centromeric activity in the autonomouslyreplicating TRA region allows improvement of the host-vector system forCandida maltosa. Mol. Gen. Genet. 249:447–455.

36. Ohkuma, M., S. Muraoka, C. W. Hwang, A. Ohta, and M. Takagi. 1993.Cloning of the C-URA3 gene and construction of a triple auxotroph (his5,ade1, ura3) as a useful host for the genetic engineering of Candida maltosa.Curr. Genet. 23:205–210.

37. Ohkuma, M., S. Muraoka, T. Tanimoto, M. Fujii, A. Ohta, and M. Takagi.1995. CYP52 (cytochrome P450alk) multigene family in Candida maltosa:identification and characterization of eight members. DNA Cell Biol. 14:163–173.

38. Ohtomo, R., K. Kobayashi, S. Muraoka, M. Ohkuma, A. Ohta, and M.Takagi. 1996. Peroxisome proliferators activate cytochrome P450 genes in analkane-assimilating yeast, Candida maltosa. Biochem. Biophys. Res. Com-mun. 222:790–793.

39. Park, S. M., M. Ohkuma, Y. Masuda, A. Ohta, and M. Takagi. 1997. Ga-lactose-inducible expression system in Candida maltosa using promoters ofnewly-isolated GAL1 and GAL10 genes. Yeast 13:21–29.

40. Popolo, L., D. Gilardelli, P. Bonfante, and M. Vai. 1997. Increase in chitin asan essential response to defects in assembly of cell wall polymers in the ggp1Dmutant of Saccharomyces cerevisiae. J. Bacteriol. 179:463–469.

41. Popolo, L., M. Vai, E. Gatti, S. Porello, P. Bonfante, R. Balestrini, and L.Alberghina. 1993. Physiological analysis of mutants indicates involvement ofthe Saccharomyces cerevisiae GPI-anchored protein gp115 in morphogenesisand cell separation. J. Bacteriol. 175:1879–1885.

42. Ram, A. F. J., S. S. C. Brekelmans, L. J. W. M. Oehlen, and F. M. Klis. 1995.Identification of two cell cycle regulated genes affecting the b 1,3-glucancontent of cell walls in Saccharomyces cerevisiae. FEBS Lett. 358:165–170.

43. Roberts, R. L., and G. R. Fink. 1994. Elements of a signal MAP kinasecascade in Saccharomyces cerevisiae mediate two developmental programs inthe same cell type: mating and invasive growth. Genes Dev. 8:2974–2985.

44. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: alaboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, ColdSpring Harbor, N.Y.

45. Saporito-Irwin, S. M., C. E. Birse, P. S. Sypherd, and W. A. Fonzi. 1995.PHR1, a pH-regulated gene of Candida albicans, is required for morpho-genesis. Mol. Cell. Biol. 15:601–613.

46. Schmitt, M., T. A. Brown, and B. L. Trumpower. 1990. A rapid and simplemethod for preparation of RNA from Saccharomyces cerevisiae. NucleicAcids Res. 18:3091.

47. Soll, D. R. 1996. The emerging molecular biology of switching in Candidaalbicans. ASM News 8:415–420.

48. Soll, D. R., B. Morrow, and T. Srikantha. 1993. High-frequency phenotypicswitching in Candida albicans. Trends Genet. 9:61–65.

49. Stewart, S., and P. T. Magee. 1990. Genetics of Candida albicans. Microbiol.Rev. 54:226–241.

50. Stoldt, V. R., A. Sonneborn, C. E. Leuker, and J. F. Ernst. 1997. Efg1p, anessential regulator of morphogenesis of the human pathogen Candida albi-cans, is a member of a conserved class of bHLH proteins regulating mor-phogenetic processes in fungi. EMBO J. 16:1982–1991.

51. Takagi, M., S. Kawai, M. C. Chang, I. Shibuya, and K. Yano. 1986. Con-struction of a host-vector system in Candida maltosa by using an ARS siteisolated from its genome. J. Bacteriol. 167:551–555.

52. Tani, Y., Y. Yamada, and T. Kamihara. 1979. Morphological change inCandida tropicalis pk233 caused by ethanol and its prevention by myo-inositol. Biochem. Biophys. Res. Commun. 91:351.

53. Vai, M., E. Gatti, E. Lacana, L. Popolo, and L. Alberghina. 1991. Isolationand deduced amino acid sequence of the gene encoding gp115, a yeastglycophospholipid-anchored protein containing a serine-rich region. J. Biol.Chem. 266:12242–12248.

54. Vai, M., E. Lacana, E. Gatti, D. Breviario, L. Popolo, and L. Alberghina.1993. Evolutionary conservation of genomic sequences related to the GGP1gene encoding a yeast GPI-anchored glycoprotein. Curr. Genet. 23:19–21.

55. Vai, M., L. Popolo, R. Grandori, E. Lacana, and L. Alberghina. 1990. Thecell cycle modulated glycoprotein GP115 is one of the major yeast proteinscontaining glycosylphosphatidylinositol. Biochim. Biophys. Acta 1038:277–285.

56. Ward, M. P., C. J. Gimeno, G. R. Fink, and S. Garrett. 1995. SOK2 mayregulate cyclic AMP-dependent protein kinase-stimulated growth andpseudohyphal development by repressing transcription. Mol. Cell. Biol. 15:6854–6863.

57. Wright, R. M., T. Repine, and J. E. Repine. 1993. Reversible pseudohyphalgrowth in haploid Saccharomyces cerevisiae is an aerobic process. Curr.Genet. 23:388–391.

58. Zlotnik, H., M. P. Fernandez, B. Bowers, and E. Cabib. 1984. Saccharomycescerevisiae mannoproteins form an external cell wall layer that determineswall porosity. J. Bacteriol. 159:1018–1026.

2086 NAKAZAWA ET AL. J. BACTERIOL.

on February 9, 2020 by guest

http://jb.asm.org/

Dow

nloaded from