Idea Transcript
Copyright Q 1993 by the Genetics Society of America
A Mutational Analysis of Killer Toxin Resistance in Saccharomyces cereuisiae Identifies New Genes Involved in Cell Wall (1 + G)=@-Glucan Synthesis Jeffrey L. Brown, Zuzana Kossaczka,' Bo Jiang and Howard Bussey Biology Department, McGill University, Montreal, Quebec, Canada H3A IBI Manuscript received November 2, 1992 Accepted for publication December 14, 1992
ABSTRACT Recessive mutations leading to killer resistance identify the KRE9, KRElO and K R E l l genes. Mutations in both the KRE9 and K R E l I genes lead to reduced levels of ( 1 + 6)-@-glucanin the yeast cell wall. The K R E l l gene encodes a putative 63-kD cytoplasmic protein,and disruption of the K R E l I locus leads to a 50% reduced level of cell wall (1 + 6)-glucan. Structural analysis of the (1 + 6)-@-glucanremaining in a krell mutant indicates a polymer smaller in size thanwild type, but containing a similar proportion of (1 -+6)- and (1 -+ 3)-linkages. Genetic interactions among cells harboring mutations at the K R E I I , K R E 6 and K R E l loci indicate lethality of krell kre6 double mutants and that krel I is epistatic to krel, with both gene products required to produce the mature glucan polymer at wild-type levels. Analysis of these KRE genes should extend knowledge of the Bglucan biosynthetic pathway, and of cell wall synthesis in yeast.
T
HE cellwall constitutes a massive extracellular organelle in fungi and plants. A major class of cell wall extracellular matrix polysaccharides are the large glucose-derived polymers, termed P-glucans (FLEET and PHAFF 198 1; KATO 1981; WESSELSand SIETSMA1981). T h e yeast cell wall glucan contains a major (1 + 3)-P-glucan, and a minor (1 + 6)-P-linked polymer (MANNERS, MASSON and PATTERSON 1973a,b). The (1 + 3)-P-~-glucancontains up to 1500 glucose residues and has a small number of ( 1 + 6)branch points and (1+ 3)-linked branches. The (1 6)-P-~-glucanis smaller with 150-200 glucose residuesand is highly branched with (1 + 3)-branch points and (1+ 6)-linked side chains. These polymers are found extensively cross-linked to chitin and mannoprotein, and arealkali insoluble (VAN RINSUM,KLIS and VAN DEN ENDE 1991). A 0-glucan fraction is, however, alkali-soluble, and has classically been viewed as a distinct polymer class. Genetic results have simplified this view, as in a mutant defective in chitin synthase 3, all yeast P-glucan is alkali-soluble (RONCERO et al. 1988).Chitin synthase 3appearstobe responsible for synthesis of cell wall chitin not found in the primaryseptum (SHAWet al. 1991).These findings suggest that the alkali-insoluble and alkalisoluble fractions differ only in the extent to which they are cross-linked to chitin in the cell wall. Understanding the synthesis of P-glucan polymers and their orderly assembly into the cell wall during growth, addresses some intrinsically interesting cell --j
' Present address: Institute of Chemistry, Slovak Academy of Sciences, Dubravska Cesta 9, 842 38 Bratislava, Slovakia. Genetics 133: 837-849 (April, 1993)
biological questions. Glucan synthesis also has broad practical applications, ranging from polymer biotechnology to the search for specific antifungal agents. The membrane-associated vectorial synthesis of Pglucans has been difficult to analyze biochemically. A genetic analysis to identify components involved in Pglucan biosynthesis circumvents this biochemical complexity, and once genes have been found, analysis of theirfunction may be possible. Important insights have comefromagenetic analysisof components involved in making a variety of cell wall polysaccharides, includingcellulose (WONCet al. 1990) andchitin (BULAWA et al. 1986; SHAWet al. 1991; VALDIVISEO et al. 1991; BULAWA1992). A number of studies have examinedmutants in both 0-glucan synthesis and hydrolysis in fungi, but despitethe tractability of such systems forgenetic analysis, the P-glucan area remains relatively unexplored. In S. cerevisiae, mutantsaffecting sensitivity (SHIOTAet al. 1985) or resistance, (JAMAS, RHA and SINSKEY1986) to (1 + 3)-P-glucanases have been reported. Yeast genes whose products have P-glucanase activity have been definedand examined (KLEBLand TANNER 1989; VASQUEZ DE ALDANA et al. 1991), but their biological roleremainsunclear. While these genes may be involved in P-glucan degradation, recentwork in Candida albicanssuggests that they could also function biosynthetically as transglycosylases (HARTLAND, EMERSONand SULLIVAN 1991). In Schizosaccharomyces pombe, two genes have been identified that affect Pglucan synthesis (RIBASet al. 1991). Mutations in these
838
Brown
J. L.
genesconferredtemperature-sensitive lysis phenotypes that could be prevented in the presence of an osmotic stabilizer. Mutations in the C W G l gene gave conditional reduction in cell wall (1 + 3)-P-glucan in vivo, and in a membrane-associated (1 + 3)-P-glucan synthaseactivity in vitro. Defects in a second gene, CWG2, conferred a conditional pleiotropic phenotype with a reduced level of both a- a n d P-glucan in vivo, and a reduced activity of a soluble possible regulatory component in the in vitro P-glucan synthase assay. A number of genes involved in P-glucan synthesis have also been identified through resistance to the S. cerevisiae K1 killer toxin, a protein that kills sensitive yeasts by forming lethal cation channelsin the plasma membrane (MARTINACet al. 1990; BUSSEY1991). T h e toxin binds initially t o yeast cells through a cell wall P-glucan receptor (HUTCHINSa n d BUSSEY1983). Selection for mutants resistant to this toxin, has defined a series of genes required for P-glucan synthesis.T h e KREI and K R E 5 genesencodesecretorypathway proteins necessary for (1 + 6)-glucan synthesis (BOONEet al. 1990; MEADENet al. 1990). K R E 6 encodes a membrane protein,possibly a glucan synthase, necessary to make normal levels of glucan (ROEMER and BUSSEY1991). K R E 2 is a gene whose product is neededforprotein glycosylation, andhasrecently been shown to encode a mannosyl transferase (HILL et al. 1992; HAUSLERand ROBBINS1992). Mutations in K R E 2 may thus lead to defects in thecross-linking of glucans with mannoprotein in the cell wall. Previous selections for killer resistant mutants were not exhaustive, (AL-AIDROOSa n d BUSSEY 1978; BOONEet al. 1990), and the present study aimed to more fully exploit the system. MATERIALS AND METHODS Yeast strains and culture conditions: Yeast strains used for thisstudy are displayed in Table 1. Killer resistant mutants were generated in the SEY6210 background from S. D. EMIR,California Institute of Technology, as described below. Yeast cells were grown under standard conditions, using media (YEPD, YNB and HALVORSON'S) as previously described (BUSSEY et al. 1982). Yeast genetics used standard procedures (SHERMAN,FINKand HICKS 1982). Routine transformations were by the lithium acetate procedure of ITOet al. (1983). Sorbitol-containing regeneration medium (SM2), was used for spheroplast transformations, and consisted of Halvorson's medium plus 2% agar, 1.2 M sorbitol and appropriate nutrilites. Toxin preparation andseeded plate tests: K 1 killer toxin was prepared using strain T158C/S14a as previously described (BUSSEYet al. 1982), with the final product approaching 1000-fold concentration. Yeast strains were assayed for resistance to killer toxin by resuspending approximately 1 X lo6 cells in 100 pI of sterile water. The slurry was used to inoculate 10 mlof 1X Halvorson's buffered YEPD or minimal medium (pH 4.7, 1% agar at 45") that was quickly poured into a Petri dish and allowed to cool. Concentrated toxin (20 11) was spotted onto the surface of
et al.
the solidified medium, and placed at 18 overnight. Following incubation at 30" for 24 h, either a zone of killed cells would appear when a strain was sensitive to the toxin or a confluent lawn of killer-resistant cells would emerge. Selection for killer-resistant mutants: Some 1-2 x 10' cells from freshly grown colonies of strain SEY6210 were taken with a toothpick, mixed with 100 pl of YEPD pH 4.7 and 25 pl of sterile killer toxin and spread onto aYEPD 1.2 M sorbitol agar-containing Petri plate, and incubated at 18 Thirty plates were prepared in this way,with a different colony being used for each plate to obtain independent spontaneous mutations. After 4-21 days, from 5-50 colonies of varying size grew up on each plate as survivors of the toxin treatment. From each plate up to four colonies of different size were taken for furtheranalysis. These resistant mutants were crossed with strains bearing mutant alleles of the previously identified genes KREI, 2, 5 , 6, and diploids tested for sensitivity to killer toxin. Forty-two mutants fell into oneof the known complementation groups; these comprised krel, 10; kre2, 15; kre5, 5; and kre6, 12. Of the remaining 35 mutants, 25 had astrong killerresistance phenotype. These mutants, presumed to identify new complementation groups, were crossed with SEY62 11, and all diploids were found to be toxin sensitive.Diploidswere sporulated, and from 8 to 16 tetrads dissected and analyzed for killer resistance to establish if the resistance segregated as a single gene. Spore progeny harboring single gene mutations were subjected to complementation analysis among themselves.Pairs of mutations that gave resistant diploids were judged to be in the same complementation group. Plasmids and recombinant DNA techniques: Yeast genomic libraries constructed in either YCp50 (ROSEet al. 1987) or pRS3 16 vectors (from C. BOONE, Universityof Oregon, Eugene, Oregon), were used to identifyisolate DNA fragments able to complement the kre9, and krell mutations. Both banksare based on single copycentromeric shuttle vectors, and contain the URA3 and Apr selectable markers. Plasmids were isolatedfrom yeast for transformation into Escherichia coli strain MC 106 1according to the procedures WINSTON(1987). Plasmids were purified of HOFFMAN and et al. (1989). Restricfrom E. coli as described by SAMBROOK tion endonucleases, Klenow and T 4 DNA polymerases, and T 4 DNA ligase were purchased from either Bethesda Research Laboratories, Inc. (Gaithersburg, MD), Pharmacia LKB Biotechnology (Piscataway, NJ), or New England Biolabs (Beverly, MA), and were used according to the manufacturers' instructions. Subcloning and disruption of KREl I : The KREl I gene was originally isolated from a YCp50-based genomiclibrary on a 12.5-kb insert. A restriction map of this region was generated and several subclonings into vector pRS316 localized the complementing activity toa 4.2-kb HindIIIEcoRI fragment (MP53). Further subclonings of this fragment were incapable of complementing the k r e l l - 1 mutation. To create an insertional disruption of the KREll gene, the 4.2-kb HindIII-EcoRI insert from plasmidMP53 was cloned into YCp50, linearized at the unique XbaI site, and treated with the Klenow DNA polymerase to render the termini blunt. The complete HIS3 gene was excised from a PBSK:HIS3 plasmid asa 1.8-kb fragment by digestion with BamHI, also filled in with Klenow, and blunt ligated into the XbaI site to create MP195. A linear 4.8-kb PuuII-EcoRV fragment containing the HIS3 gene at this XbaI site was excised from MP195 and used to disrupt the KREl I locus (see below). Two deletion disruptions of the KREll locus were alsoperformed, using the TRPl andURA3 auxotrophic O
O .
839
Yeast KRE Genes TABLE 1 Yeast strains Strain
SEY62 10 SEY62 1 1 HAB251-15B HAB522 HAB535 HAB544 HAB556 HAB562 HAB580-8D HAB635 HAB637-1-6A HAB669-2A HAB669-6C-3 HAB670 HAB676 HAB677 HAB751 HAB754 HAB756 HAB757 HAB758 HAB768 HAB792 HAB795 HAB800 HAB802 HAB805 HAB806 HAB807 HAB808 HAB809 HAB8 10 Y355 Strain 7B 7B- 1 1 T 158C/S 14a 165
Source
Genotype
MATa leu2-3,112 ura3-52 his3-A200 lys2-801 trpl-A9Ol suc2-A9 MATa ade2-I01 leu2-3,112 ura3-52 his3-A200 trpl-A901 suc2-A9 6 2 10 autodiploid MATaIMATa leu2-3, 112/leu2-3,112 ura3-52/ura3-52 his3-A200/his3200 lys2-801/lys2-801 trpI-A901/trpI-A901 suc2-A9/suc2-A9 MATa kre9-1 ade2-I01 leu2-3,112 ura3-52 his3-A200 lys2-801 trpl-A901 suc2-A9 MATa krel0-1 ade2-I01 leu2-3,I 12 ura3-52his3-A200 lys2-801 trpI-A901 suc2-A9 MATa krell-1 ade2-I01leu2-3,I 12 ura3-52 his3-A200 lys2-801 trpl-A90l suc2-A9 MATa kre9-4 ade2-I01 leu2-3,112 ura3-52 his3-A200 lys2-801 trpl-A901 suc2-A9 MATa krel0-2 ade2-101 leu2-3,112 ura3-52 his3-A200 lys2-801 trpl-A9Ol suc2-A9 krell-1 leu2-3,112 ura3-52 his3-A200 trpl-ASOl suc2-A9 MATa krel::HIS3 ade2-I01 leu2-3,112 ura3-52 his3-A200 trpl-A901 suc2-A9 MAT@krel A::HIS3 ade2 MATa krelI::HIS3 lys2 krel I::HIS3 Diploid from HAB637-1-6A X HAB669-2A Mala krelA::HIS3 krelI::HIS3, suppressor from HAB670 krelk:HIS3 krel I::HIS3, suppressor from HAB670 Mata kre9-1 ura3-52 leu2-3,112 his3-A200 trpI-A901 suc2-9 MATa krell-I ADE2 MATa krell-1 lys2-801 ade2-I01 leu2-3,I 12 ura3-52 his?-A200 trpl-A90l suc2-A9 HAB75 1 with pRS306 KRE9 integrated HAB757 X SEY6210 HAB669-6C-3 X HAB580-8D MATa krell::HIS3 derivative of HAB670 Diploid from HAB792 X Y355 MATa yurlA::HIS3 leu2-3,112 ura3-52 his3-A200 lys2-801 trpl-A901 suc2-A9 Diploid from HAB757 X HAB800 MATa krel1k:TRPI leu2-3,112 ura3-52 his3-A200 lys2-801 trpl-A901 suc2-A9 MATa kreIlA::URA3 leu2-3,112 ura3-52 his3-A200 lys2-801 trpl-A901 suc2-A9 Diploid from HAB805 X HAB635 krelA::HIS3 krell A::TRPl,suppressor from HAB807 krelA::HIS3 krel I A::TRPI, suppressor from HAB807 krelA::HIS3 krel I A::TRPI, suppressor from HAB807 MATa rsrlA::URA3 ade3 ura3 leu2 trpl ura3 his3 glcl krel I::HIS3 ura3 glcl MATaIMATa his4C-864/HIS4 ADE2/ade2-5 MATa tif463A::LEU2 leu2 trpl-289 ura3-52ade6 his3A1
markers. T o create a kre1lA::TRPl allele, a 1.17-kb ClaISnaBI fragment of the KREll open reading frame was replaced with an 860-bp fragment of the TRPl gene. Similarly, a krel lA::URA3 disruption was created by replacing the XbaI-SnaBI portion of KREll (798 bp) with a 1.2-kb fragment containing the URA3 gene (see Figure 1). Both deletion constructswere digested with PuuII and Xhol prior to deletionof the KREllwild-type locus by single step gene replacement. DNA sequencing: Subclones ofK R E l l were made in the pRS316 vector (Figure 1) and double-stranded DNA was prepared from MC1061 and DH5aF’ by standard procedures. Sequencing was by the dideoxy method (SANGER, NICKLENand COULSON1977) using Bluescript primers and syntheticoligonucleotides complementaryto specific regions of K R E l l , using the Sequenase Kit (U.S. Biochemicals, as a substrate (Amersham Cleveland, OH) and [a-35S]dATP Canada Limited,Oakville, Ontario). T h e complete sequence of K R E l l shown in Figure 2 was determinedforboth strands. Chromosomal localization and mapping of KRE9 and K R E l l : Chromosomal blots of S. cerevisiae were purchased
S. D. EMR S. D. EMR ROEMERand BUSSEY (1991) This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work This work M . LUSSIER This work This work This work This work This work This work This work A . BENDER BOONEet al. ( 1 990) This work ATCC 26427 C. COYER
from Clontech (Palo Alto, CA). Specific DNA probes were labeled using the Oligolabelling Kit (Pharmacia) according to the manufacturer’s instructions using [ C Y - ~ ~ P ] ~asCaT P substrate (Amersham Canada Limited, Oakville, Ontario). Hybridization of a 1.3-kb BglII/XbaI fragment containing the open reading frame of KRE9 (Figure 1A) identified chromosome X . T h e K R E l l gene was mapped to chromosome VII using a 750-bp ClaI/XhoI fragment (Figure 1B). After further exposure of the blot probed with the KREll sequence, an additional band corresponding to chromosome V was observed. T h e KRE9 and K R E l l probes were also hybridized to the X phage library of mapped yeast genomic DNA inserts of (L. RILESand M. V.OLSON, personal communication), and the KRE9 and KREll loci finally genetically mapped to known markers on chromosomeX and VU, using theequation of PERKINS(1949)to calculatemap distances. (1 + B)-Glucananalysis:(1 ”* 6)-Glucan was isolated and quantified according toBOONEet al. (1 990), outlined below. Yeast cells were grown in 10 ml YEPD until stationary phase,whentheywereharvested and washed once with distilled water. After threesubsequent alkali extractions (0.5
840
Brown
J. L.
mlof 3% NaOH at 75" for 1 hr), the residual cellwall glucans were neutralized by washing once with 1 ml of 100 r n M Tris-HCI, pH 7.5, and twice with 1 ml of 10 mM TrisHCI, pH 7.5. The wall glucans were then resuspended in 1 rnl of 10 mM Tris-HCI, pH 7.5 and digested with 1 mg of the (1 + 3)-&glucanase, ZyrnolyaselOOT (ICN Biochemicals, Cleveland, OH) at 37" for 16 hr. After removing the residual insoluble material by centrifugation, the supernatant was dialyzed against distilled water for 16 hr using Spectra/por tubing, with a 6,000-8,000-D pore size (Spectrum Medical Industries, Los Angeles, CA). The carbohydrate present prior to dialysis represents the total alkaliinsoluble glucan (1 + 3)-@-linked plus(1 6)-linked, while the post-dialysis fraction represents (1 + 6)-&glucan. Carbohydrate in each fraction was measured as hexose by the JACKSON and SCHUBERT borosulfuric acid method (BADIN, 1953). For large scale isolation for structural analysis, the (1 + 6)-B-glucan was purified from cultures of 2 or 5 liters of wild-type cellsor yeast strains harboring disruptions at either the krelA or krellA loci, using the method of BOONEet al. (1990). Gel filtration chromatography of glucans: A Sepharose CL-GB (Pharmacia) column with dimensions 109 X 1 cm was used to estimate the molecular weight distribution of glucan samples (-1 mg) by gel filtration chromatography. The eluant was 0.1 M NaOH, and 0.33-ml fractions were collected at aflow rate of 10 ml/hr. Molecular weightswere estimated by comparison with dextran standards (Pharmacia), with dextran blue used to determine the column void volume. The hexose content of each fraction was estimated by the borosulfuric acid method of BADIN,JACKSON and SCHUBERT (1953). "C N M R Glucan samples (50-100 mg) were dissolved in -5 ml of D20, placed in5-mm diameter tubes, and proton decoupled "C NMR spectra obtained using a Bruker WH 400 spectrometer (Bruker Instruments, Billerica, MA) operated in the Fourier-transform mode, at 100.615 MHz. Each spectrum represents approximately 10,000 scans with a sweepwidthof 12,500 Hz, and an acquisition time of 0.655 sec. The probe temperature was held at 18",and the reference peak was external dioxane at 67.4ppm. ''C NMR spectra were generated for glucansamplesisolated from two independent yeast strains, SEY6210 and 7B, differing slightly with respect to genetic background. Both wild-type and kreA disrupted representatives from each strain produced an equivalent set of spectra. Additional cell wall analyses: Alkali-soluble glucan plus mannan cell wall fractions of wild-type or mutant strains were estimated based on the procedure described by RONCERO et al. (1988). Briefly, cells were extracted twice with 3% NaOH for 90 min at 80", and the alkali-soluble glucan plus mannan co-precipitated from the alkali extract with absolute ethanol. This material was subsequently digested overnight with 1 mg/ml Zymolyase lOOT in 100 mM TrisHCI (pH 7.5) at 37", and one-half the volumedialyzed against distilledwaterusing 6,000-8,000-D Spectra/por tubing. The total alkali-solubleglucan plus mannan fractions were estimated prior todialysis as hexose by the borosulfuric acid method. The carbohydrate material remaining after dialysis was taken to represent a fraction of the cellwall mannan, and alkali-soluble (1 + 6)-glucan. The (1 + 3)$glucan synthase activity was assayed from cell free extracts of wild-type or isogenic strains harboring a krelA::HIS3 or krel I::HIS3disruption, according to the procedure of CABIB and KANC (1987). Cell wall chitin was observed by fluorescence with Calcofluor White, and glycogenlevelsincell colonies were qualitatively assessed usingiodine vapors.
-
et al.
RESULTS
Isolation of killer resistant mutants in SEY6210: To search for mutations in possible new genes requiredfor killersensitivity, we selected for toxin resistant mutants, including those that grewslowly; in addition we selected in the presenceof 1.2 M sorbitol, so that osmotically fragile mutants might survive. Survivors of killer toxin treatment from a sensitive population of 3-6 X 1Os yeast, were selected as described in MATERIALS AND METHODS. Approximately 100 surviving colonies were taken and tested for toxin resistance on YEPD sorbitol seeded plates; 77 of these had a resistant phenotype. All grew on YEPD in the absence of sorbitol, and thus none appeared to be osmotically fragile.
Assignment of three new complementation groups: T h e resistant mutants were tested for complementation of mutant alleles of previously identified KRE genes (see MATERIALSANDMETHODS). Twentyfive mutants fell into no knowncomplementation group and were testedfor single gene defects andfor newcomplementationgroups (see MATERIALS AND METHODS). Three new complementation groups emerged; kre9, krelO and k r e l l , with 6 , 5 and 2 members, respectively. Allelism was tested between the krel0-2 and krellI mutations. Strain HAB562krel0-2 was crossed with HAB756 krell-1. T h e toxin-resistant phenotypes for thesemutationssegregatedindependently:parental ditypes (0), nonparental ditypes (2), tetratypes (lo), indicating that they were not allelic. In addition, the resistant krelO krell double mutants showed a more extreme slow growth phenotype than either the krel02 or the krell-I mutants alone. Mutantswith a defect in kre9 grew very poorly, were found to be more sensitive t o Zymolyase digestion than the wild type, and were difficult to work with. Cloning of the KRE9 gene indicated that it was distinct from K R E l l , see Figure 1 and below. (1 + 6)-&Glucan levels in new mutants: T h e alkali-insoluble (1 + 6)-glucan levels from kre9, 10 and 1 I mutantsandthe isogenicparentalstrainwere compared to determineif the killer resistance phenotypes correlated with a reduction in cell wall glucan. T w o kre9 alleles were tested for effects (Table 2) and both strains showeda reduction in the (1"* 6)-glucan fraction of approximately 60% as compared to the wild-type. T h e krel0-2 mutation caused no significant change in glucan level,while the krell-1 allele reduced the level of this polymer to about one half that of the SEY62 10 isogenic parental strain. Cloning of KRE9: T h e KRE9 gene was cloned by functional complementation of thekre9-1 mutation in HAB751, as described by BOONEet al. (1990) using pRS316 and YCp50-based yeast genomic banks (BOONEet al. 1990; ROSEet al. 1987). Strains harbor-
Yeast KRE Genes A ) KRE 9
84 1
TABLE 2
COMPLEMENTATION
Glucan levels of new kre mutants
-
PvuII Bg111
Kpnl
I
I
I
Spel Xbal
I I
1
PvuII
Kpnl
+
I
+/+/ -
I
I
I
8)KRE I t
I
Strain
SEY62 10 HAB522 HAB556 HAB535 HAB544
KRE allele
Wild type kre9-1 kre9-4 krel0-2 krell-1
@(1 + B)-Glucan (pglmg dry wt)
36.3 f 3.7 13.7 k 2.6 14.7 f 0.6 40.7 f 5.4 18.6 k 0.6
Cell wall (1 + 6)-glucan levels from representatives of the three new complementation groups are compared with the isogenic parental strain, SEY6210. Total alkali-insolubleglucan was not significantly different in any of these strains, with an average of 203 f 16 pg/mg dry weight. Error represents 1 SD. -
I
1
I
-
I kb
FIGURE 1.-Restriction maps of KRE9 and K R E l 1 . Restriction maps of the (A) KRE9, and (B) K R E l l loci are shown, with black arrows indicating the size and positions of the open readingframes. Only selected restriction sites are shown. Subclones of each fragment were tested for their ability to complement the respective kre9-2 or k r e l l - 1 mutation based on sensitivity to toxin on seeded plate tests (+, full complementation; +/-, partial complementation; -, failure to complement).
ing kreY mutations failed to transform by the standard lithium acetate or electroporationtechniques; so a modified spheroplast transformation procedure was employed. Approximately20,000 Ura+transformants were grown in sorbitol-containing selective regeneration medum (SM2), and the regeneration agar medum plus yeast colonies homogenized in a Waring-type blender. This slurrywas re-plated onto YNB selective medumto isolate single colonies, and then replicaplated onto methyleneblue-containing YEPD agar seeded with a diploid killer strain to test for toxin sensitivity. Cells with plasmids containing a wild-type copy of KREY should restore killer toxin sensitivity and be killed. Sensitivity in this test results in a white colony edged with a blue-stained ring of killed cells. Cells of toxin resistant colonies decolorize the methylene blue dye, and remain white in this test. Of the 20,000 colonies screened, two different sized plasmids from the pRS3 16 bank, and three from the YCp50 bank were found to completely restore growth and toxin sensitivity to kreY-I strain HAB751, in a plasmiddependent fashion. The smallest complementing fragment was a 7-kb insert from the pRS3 16 bank, and a restriction map of this fragment was generated. Subcloning experiments suggested KRE9 activity was located in a 3.8-kb KpnI fragment (Figure 1A). Additional subclonings of this regionrevealedthat two fragments overlapping by 1.3-kb were both able to
partially complement the kreY-l mutation, suggesting the approximate location of this gene. The position of an open reading frame within this fragment has recently been determined (J. BROWN,unpublished results) and is shown. KRE9 is allelic to kre9-I: To test if the3.8-kb genomic KpnI complementing fragment was KRE9, thefragment was first cloned intotheintegrating vector pRS306. The resulting constructwas linearized at a unique BglII site within the insert (Figure IA), and used to transform cellswith a kreY-1 mutation. The Ura+ prototrophs selected should contain a tandemly integrated URA3 gene at the kreY locus, as well as a restored wild-type copy of KREY. All transformants tested were sensitive to K 1 toxin and grew at wild-type rates, suggesting the homologous recombination event was capable of rescuing the kre9-1 mutation. One such transformantHAB757 was backcrossed to the isogenic wild-type parent, SEY62 10. The resulting diploid HAB758 was sporulated and 24 tetrads analyzed. All 4 spores of each tetrad were toxin sensitive, with the Ura+ phenotype segregating 2:2. This result indicates thatthecomplementing fragment directed integration to the kreY-1 locus and that the KRE9 gene had been cloned. Mapping of KRE9: KRE9 was mapped to chromosome X using a 1.3-kb BglII-XbaI fragment of the gene (Figure 1A) to probe a chromosome blot,and further to X clone 6699 of the X phage library of mapped yeast genomic DNA inserts of L. RILESand M. V. OLSON(personalcommunication). Clone 6699 lies between the cdcb and tzj2 genes, some 75-92 kb from the end of the left arm of the physical map of chromosome X (L. RILES, personal communication). We tested for linkage of KREY with y u r l , which is tightly linked physically to TIF2 in this region of chromosome X (MULLER,TRACHSEL and LINDER 1989; FOREMAN, DAVISand SACHS199 1). Tetrad analysisof spore progeny from diploid HAB802 YURl/yurl::HIS3 KRE9::URA3 kre9-IIKRE9 indicated a distance of 19 CM for the yurl-KRE9 interval (parental ditypes: 42, nonparental ditypes: 0, tetratypes: 25). This genetic
842
J. L. Brown et al.
distance is consistent with the physical mapping of KREY, approximately 50 kb centromere distal to t i p on the left arm of chromosome X . Cloning of K R E I I : The wild-type K R E l 1 gene was cloned by functional complementation of the k r e l l - I mutation in HAB754 as described above for KREY, using a YCp50-based yeast genomic bank. Approximately 15,000 Ura+ colonies were screened for the killer-sensitive phenotype, and three different sized genomic inserts with a common fragment (Figure 1B) were foundto completely complement the k r e l l - 1 mutation in a plasmid-dependent fashion. The smallest complementing fragment contained a 12.5-kb insert, which was shown (see below) to contain the wildtype K R E l I locus. Nucleotidesequenceof K R E I I : The subcloning and disruption experiments indicated anXbaI restriction site in the functional region of K R E l 1 , and the DNA sequence in this region was obtained (Figure 2). This revealed a single, continuous open reading frame of 1680 nucleotides spanning the XbaI site (bold in Figure 2).The K R E l I sequence predicts thatthe gene encodes a protein, Krel lp,of 560 amino acids, with a predicted molecular massof 63.3 kD (Figure 2). Structurally, Krel l p appears to bea soluble cytoplasmic protein, and is devoid of any hydrophobic potential transmembrane or signal sequences. Comparison of both the K R E l l nucleotide and deduced amino acid sequences with those in the EMBL and GenBank databases has been uninformative. Krel l p does however, contain the sequenceRLGG, which conforms to the consensus (RXGG) found in the proposed UDP-glucose bindingdomains ofyeast and rabbit glycogen synthases (FARKAS et al. 1990; MAHRENHOLZ et al. 1988). The presence of several Asnrich tracts located between amino acids 96-1 15, 343381 and 438-476 is also noteworthy. Mapping of KREII: A 750-bp ClaI/XhoI fragment present in theopenreadingframe of K R E l l (see Figure 1B) was used to probe a yeast chromosomal blot (see MATERIALS AND METHODS, and Figure 2),and hybridized to chromosome VZZ (data not shown). The K R E l l probe also indicated hybridization to X clone 3286 of the L. RILESand M. V. OLSONX phage library of yeast genomic inserts (see MATERIALS AND METHODS), which is anorphanand unassigned, andto cosmid clone 9943 which contains the X 3286 DNA and which was originally placed in the hip1 region of the right arm of chromosome VZZ, but could not be unambiguously assigned. In addition the gene TZF4631 hybridized to X 3286 as well as 6627, a X insert localized to the rsrl region of chromosome VZZ (L. RILESand C. COYER, personal communication). T o resolve the location of K R E l I, we testedfor linkage with rsrl (BENDERand PRINGLE 1989), and TZF4631. Preliminary data indicated that tif4631 and
krel I were closely linked (parental ditypes: 14, nonparental ditypes: 0, tetratypes: 0). In a further cross, k r e l l linkage to rsrl and ade3 was examined, see Table 3. Close linkage was found of k r e l l to ade3, a marker close to hip1 on the right armof chromosome V U , and based on linkage to rsrl, k r e l l is centromere proximal to ade3. Disruption of K R E I I : A 1.8-kb fragment containing the HIS3 gene was inserted at the XbaI site in the K R E l I open reading frame, and a linear 4.8-kb PvuIIl EcoRV fragment used to disrupt the wild-type K R E l I locus in an isogenic diploid strain, HAB 251-15B,(see below for proof thatthe cloned gene is K R E I 1 ). Killer resistance co-segregated 2:2 with the His+ phenotype in 30 tetrads dissected. Resistant spore progeny from 5 tetrads had a 50% reduction in the cellwall (1 + 6)-glucan level, a representative example is shown in Table 4. Two large deletionsof the K R E l 1 locus were also created using the T R P l and U R A 3 selectable markers (see Materials and Methods). Deletion disruptions made in both haploid SEY6210 and in the isogenic diploid strain, HAB 25 1-15B, gave growth and killer resistance phenotypes indistinguishable from the kreII::HIS3 insertionabove, see Table 4. Correct integrations at the K R E l I locus were confirmed by genomic Southern blots (data not shown). A mutant strain harboring the larger of these deletions, HAB805 krel IA::TRPI, was assayed for cell wall (1 + 6)-glucan levels. Table 4 shows the kreI1A::TRPl deletion has a similar effect on cell wall glucan, reducing the in uiuo levels of the (1 + 6) polymer approximately 50%. The reduction in (1 + 6)-glucan levels found with the krel1A::TRPI deletion was comparable to those found in the kreII::HIS3 insertion and the k r e l l - I allele, suggesting all three mutations generate null alleles of the KREI I gene. KREII is allelic to k r e l l - I : Since the insertion at the K R E l I locus did not severely impair cell growth it allowed an allelism test to be performed.Strain HAB 669-6C-3,krel I::HZS? harboring a disruption at the locus of the cloned gene, was crossed with HAB 580-8D, k r e l l - I , andthe resulting toxin resistant diploid, HAB 768, sporulated and 27 meiotic tetrads analyzed. All tetrads segregated 4:O for killer resistance and 2:2 for His+; the HIS? insertion was, therefore, closely linked to the krell-1 mutation. Analysis of glucan from a k r e l l mutantby gel filtrationand "C NMR spectroscopy: The killer resistance phenotype and the reduction in the level of the (1 + 6)-glucan fraction, suggested a possible role for K R E l 1 in (1 += 6)-glucan biosynthesis. The residual (1 + 6)-glucan from each of two strains, disrupted at the K R E I I locus, was purifiedfor analysis. The (1 += 6)-glucan was isolated from genetic backgrounds SEY6210 and 7B, size fractionated by gel filtration, and analyzed by "C NMR spectroscopy, to examine
Yeast KRE Genes
843
ATAAGAATAC -535 -525 G A A G G T T A T A T A T G G T ~ G C T n : T ~ G G l Y X % ? i G A T C A C A A T A T ~ ~ A C T A A C G ~ G G T ~ G -450 n ; G G 3 T A C C G T T A C G G T A G C D G 9 A A T A ~ T A A C A A G A A A G A C A G A A ~ ~ C T T T A A T A A A C ~ ~ -375 TGGTATATATATAACCACTTTTTWCTTATGAGTGCTTl'CTTGTATATGCGTATATGTCGTCATATT~GCCA -300 TTATTACTTGTACATAAAATATTCTACATAAATTTATATATCCCAlYX%?iGAGCCATGGTCACGlVZGGCACATT -225 T C T n ; 4 A T A C T C G T T T G C A A G G C T T A T T l l ' ~ T C ! l ' T T T C A C A ! l ' C T T T C T A T T T T A G T G ~ G A A A ~ ~ C -150 CTATGAAAAATTGGTAACAAAAATAATTAAMGTAAATGCTATAAAATTGlWXTAAATAAAAGTGGMiCAAGAG - 75 C n ; q G T T C A C A A C T C T A C A C T C T A T A ~ C A A T A T T ~ T T A A ~ C A A ~ T A A A ~ C A A C T C A T ~ C C 1 ATG GAA TGT TTTGTA CCG CTG CGT TGC GAC TTA GAT GGA AGTAAT A T A GAA CAG T T A E C F V P L R C D L D G S N I E Q L 1 M 58 CGT CAA TCA CAC TTA AGCCGT AAA TTTATTATATTT GAT GAG CAA CTG AATCTT TGG 20 R Q S H L S R K F I I F D E Q L N L W 115 CTG Tcx: TTT CAA GGT AAT TCG CAA GAG AAC AAG AGA TTTGTACTA CAG AATATGATA W F Q G N S Q E N K R F V L Q N M I 39 L 1 72 ATA TTA ATA AAT GAA GCG CAA GTT ACC A 0 9 ACA AGC ACTATC GAT GAT TATTTT ACC 58 I L I N E A Q V T R T S T I D D Y F T 229 GU GTT GAG AACAAT GAA AATCTA TGG AGG TTG AAA AAC GACTGC TGT TCG AAG ATT 77 Q V E N N E N L W R L K N D C C S K I 286 CTT TTC AAA TCAAATGTTGTTATGAATAAT GGT TATAATAAT CAG ATC AAA TTT GTC 96 L F K S N V V M N N G Y N N Q I K F V CAG GAT TCATTA CAA GAT CCA 343 TTT GAA TAT AAA TCT GTG GAT GCA AATTTCAATAAC 115 F E Y K S V D A N F N N Q D S L Q D P 400 CAGGCA AAGTAT ACA TTA GAT AAGTAC TCT AGCGAGGAG ATTTTG CCA AGT TTT GAG 134 Q A K Y T L D K Y S S E E I L P S F E 457 CCA GTTTAT TCC TGG TCTTCT GCA GCC ACC AAA TCATCC AAA AATACTAATAATCAT 153 P V Y S W S S A A T K S S K N T N N H 51 4 CTG GAG AAA AATAAC AGG GCG ACT CAT CGA GTT AGT TCT AAA AAT AGC GAA GTC CAC FIGURE2.-Thenucleotideand 172 L E K N N R A T H R V S S K N S E V H predicted amino acid sequence of 5 71 GAA GCG GAC GTT !KT Mu AAT CCG AATACATTTACTCTT AAG CTC CAA TAC C c 4 A T A 191 E A D V S R N P N T F T L K L Q Y P I KREZI. (GenBank accession number 62 TTC 8 TCACTTTTGAATATG AGA Tl'G AGA AACATC TCC TTG AAA TCT GAG CAT TGC A T A L10667)A,1680-bpopen reading 21 0 F S L L N M R L R N I S L K S E H C I frame is shown along with the pre685 TTA TCA TCG TTA GAC TTT CAA ACT TCT AAA GCA TCC GAA CAA CTG ACC AAA AAA TTT dicted amino acid sequence of 229 L S S L D F Q T S K A S E Q L T K K F Krel l p (560 amino acids). The XbaI 742 ATTTAT CCG CAA GAA CAC AAT TCT TTT CTC AAA CTG AACTTT CAA GAA A T A TCG TAT restriction site used for subcloning 248 I Y P Q E H N S F L K L N F Q E I S Y and the insertional disruption of 799 AAA CTAATC GAC Go4 ACC TCT CAA ATT GAA CTA GAT CCA ATC TGT CCT TTG AAA GTG KREll isin bold. A possible UDP267 K L I D G T S Q I E L D P I C P L K V glucose binding consensus sequence 856 CCA CTCACT GCA TTTTCATAC GAT AGC ATT AGCGCT ACTTTT AAA CTGGTT CTG TTA is underlined. Several Asn-rich re286 P L T A F S Y D S I S A T F K L V L L gions of Krel l p can be seen between 913 CCC AAA TCAACT CAA CCA CAT CGT GTG AAA ATC ACC TTA GCG TAC GAA CTC GAG CTG amino acids 96-115,343-381and 3 05 P K S T Q P H R V K I T L A Y E L E L 438-476. 970 CAT CCC AATCTGAAG TGG GAA ACA GAA GTC ACATTGAAG TTA CCT GTG AGAACATCA 324 H P N L K L P V R T S W E T E V T L K 1027 CGT TCTATG CCA ATT TCC TCG ACATCTTCT CAA TAC TCG AGT AACAACAATAAT ACC 343 R S M P I S S T S S Q Y S S N N N N T 1084 AAT CAT AGCGCT TCTTTTAAT QZTGCGGCC AACAACGTTAATTCT GGTGGT TTG GCC 3 62 N H S A S F N G A A N N V N S G G L A 1141 AATCTAAGA TCG ACT TTA GGT GGG GTT TCC TCC T o 9 AGA TTT AGT CTT GGA GCTGCT 381 N L B L G G V S S S R F S L G A A S T 1198 ACA TCATTG GTG AAT AGC AAA TTA AGC AATGTA AAA TTCAAGTTTATTAAT AGC AAT 400 T S L V N S K L S N V K F K F I N S N 1255 A T A AAA GTTATTAAG GGC GAA AAG TTTACTATG AGG CTT CAG ATCATTAAC TCG TCA 419 I K V I K G E K F T M R L Q I I N S S 1312 TCA TCC CCC CTG GAT CTTGTT GTG TATTATAACAATACAATAAAC CCA ATC CCT TCA 438 S S P L D L V V Y Y N N T I N P I P S 1369 GCT AATAACGTA CGT AAC AGC AAT GGT ATAAACAAC TGT GGC ATGAATAATACT 457 A N N V R N S N G I N N C G M N N G T 1426 ATC CCC AAT TCG CCC TTGACA CIY: GAA AAT CAG TAC GU CTG CATAAT AAA TATAGA 4 76 I P N S P L T L E N Q Y Q L H N K Y R 1483 TCC AAC GAT TAC AAA ATT CCA GTTGTA CCT A m GCGGAG GGG ATTATACTATTA 495 K I A E G I I L L S N D Y K I P V V P 1540 CCG A m GAA ACATACTTC GCG GAT TTA CGA TTTATT GGT ATTATG TCC GGA TATTAT 514 P R E T Y F A D L R F I G I M S G Y Y 1597 GGcACTCTCTCC Go4 CTTAAGGTA Tn: GAT TTAAATACAAAT GAA CTTATAGTT 533 G T L S G L K V L D L N T N E L I E V 1654 GGA AAT GGC GCA TCT GTG TTAATC CAG TAA GCGGUCAAGTATACATTTTATTAACATACTAC G552 N G A SL V I QSTOP 1711 A T A C T A A T C T G T A A C A T T T C T A T A A A A A A G T T A T T A T C ~ T
J. L. Brown et al.
844 TABLE 3 Mapping of K R E l l Tetrad type Interval krell-ade3 krell-rsrl rsrl-ade3 4
PD
91 3539 28 52
Map distance
NPD
TT
0 3
11 49 72
0.6
(CM)
5
"
0.5
-0 3 --04
Diploid HAB795, isolated from crossing strains Y355 and HAB792, was sporulated and tetrad analysis was performed. Map distances were calculated using the equation of PERKINS(1949), and a graphical correction for longer map distances (MORTIMERand SCHILD (1980).
0.2
--
01 0
TABLE 4
,
:
.'-
'.
i
m' '8
,.a'
06
Glucan levels in krel and krel I mutants
0.5
Strain
KRE allele(s)
SEY 62 10 HAB 637 HAB 669 HAB 676
Wild type krelA::HIS3 krel l:HIS3 krelA::HIS3 krell:HIS3 krel&:HIS3 krel l:HIS3 krell A::TRPl krel&:HIS3 k r e l l A::TRPI krelk:HIS3 krell A::TRPl krelk:HlS3 krel1A::TRPl
HAB 677 HAB 805 HAB 808 HAB 809 HAB 810
B(1
-P
(Pi+%
6)-Glucan
0.4
dry wt)
03
32.8 f 2.5 14.3 f 0.8 15.4 f 2.7 8.4 f 1.5 7.0 f 1.2 15.0 rfr 0.6 7.7 2.1
*
7.5 f 1.0 7.4 f 0.3
02 0. I
0 0 P v) D
0
0.5 0.4
0.3 0.2 0.1
kre I ::HIS3
-'
-a
-'
-I
--
The levels of (1 + G>glucan were analyzed in strains harboring disruptions at the h e 1 locus, the k r e l l locus, or independently arising suppressors to krelkrell double mutants. Total alkaliinsoluble glucan was an average of 145 f 23 ag/mg dry weight in all these strains. Error represents 1 SD.
the structure of the polymer. Gel filtration chromatography of (1 + 6)-glucan isolated from krell::HZS3 or wild-type strains (Figure 3), suggested the mutant glucan was somewhat smaller than that from thewildtype strain. The apparent weight average molecular massof the wild-type glucan was approximately 40 kD (containing about 200 glucose residues), while the krell::HZS3 mutant glucan eluted with a profile corresponding approximately to a 30-kD polymer. 13C NMR was performed on the k r e l l mutant glucan to assess the relative proportions of (1 ".* 3)-/3- and ( 1 + 6)-/3-linked residues in the polymer. The "C NMR spectrum of krel 1 (1 ".* 6)-glucan was similar to that of (1 + 6)-glucan purified from the isogenic wildtype strains SEY62 10 and 7B (Figure 4).These results suggest that disruption of the K R E l l gene leads to a (1 + 6)-glucan a little smaller in size than the wildtype polymer, but containing a similar proportion of (1 + 3)- and (1 + 6)-linkages. The k r e l l mutations affect (1+ 6)-glucan synthesis in a different manner than those in the K R E l gene (BOONEet al. 1990). Krelp is involved in the attach-
100
120
140
160
180
200
220
240
260
Fractlon Number
FIGURE3.--Gel filtration chromotagraphy of purified ( 1 + 6)glucan isolated from kre mutant strains. A Sepharose CL-GB column was used to determine the average molecular weights of (1 + 6)glucan isolated from strains: (A) SEY6210,(B) krel l::HlS3, (C) krelA::HlS3, (D) double mutant (krel&:HlS3 krell::HIS3), as described in the MATERIALS AND METHODS. Each glucan sample was purified from two independent strainsand runtwo or three times. The elution volumes of dextran standards are indicated. The column void volume was 30.5 ml, or fraction 94.
ment of (1 + 6)-/3-glucan side chains to a (1 + 3)branched, (1 + 6)-glucan backbone. The (1 + 6)-glucan from a krelA mutant is smaller in size than wild type, but differs fromthe krel1 mutant glucan above, in having a reduced proportion of (1 + 6)-linked residues in the mutant polymer. T o explore a possible
Yeast KRE Genes A)
c-3
SEV8210
c-2
I
c-1
I / I'
6 )k r e l l
845
I'
C-B(linked)
HIS3
I
c) krel
FIGURE4.-% N M R spectra of purified ( 1 * 6)-glucan isolated from kre mutant strains. Purified (1 + 6)-glucan samples were subject to I3C N M R spectroscopy to determinethe relative proportions of ( 1 + 3)-/3 and (1 + 6)-/3 linkages present in the polymers. The predominant signals (A) C-1, C-3 (linked), C-3, C-5, C2, C-4, C-6 (linked) and C-6) characteristic of (1 + 6)-/3-linked glucan in S. cereuisiae are indicated as previously described by WNE et al. (1990). (A) SEY6210, (B) hrell::HIS3, (C) k r e l k : H I S 3 , (D) double mutant (krelA::HIS3 krel l::HIS3).
I
HIS3
1
I
D) k r e l l : . H I S 3 krel HIS3
I
100
90
80
70
80
PPn
846
J. L. Brown et al.
interaction of the K R E l and K R E l l gene products, we examined the phenotype of k r e l A k r e l l double mutants. Phenotype of k r e l l Krel double mutants:A krel A krell double disruption was created by mating strain HAB637- 1-6A krelA::HIS3, to the isogenic strain HAB669-2A krel I::HIS3, and the resulting diploid HAB670sporulated,andtetrads dissected. Spore progeny harboring both the krelA krel I mutations showed a slow growth phenotype thatwas more severe than the modest growth reduction seen with either mutation alone. Indeed the double mutants gave microcolonies that were so severely compromisedfor growth that they could not be tested directly. Spontaneous second site suppressors of the slow growing doublemutantphenotypearoseatafrequency of approximately 1 X 1O"j. These suppressors partially restored growth but did not restore toxin sensitivity. The krelA krelI double mutants bearing the suppressor could be further characterized. The retention of disruptions at both the K R E l and K R E l I loci in these suppressed strains was confirmed by genomic Southern blots. Previously, partial suppressors of the severe growthdefect in a kre5A null mutant allowed the determination that the mutant lacked (1 + 6)-glucan (MEADENet al. 1990), suggesting that suppression may occur through an alternative pathway. Glucan analysis on two independent suppressed variants of the k r e l A k r e l l double mutant, HAB 676 and HAB 677, revealed a 75-80% reduction in the(1 + 6)-glucan fraction,areduction greaterthanthat caused by either a k r e l A or k r e l l mutation alone, see Table 4. Similar results were also obtained in HAB808, HAB809 and HAB810, three independently derived suppressors of k r e l A k r e l I A double mutants generated from crossing HAB637-1-6A krelA::HIS3, with the k r e l l A deletion mutant, HAB805 krel1A::TRPl (Table 4). Characterization of HAB677 k r e l A k r e l l glucan by gel filtration indicated an average molecular massof 15 kD, similar tothat of the kreIA::HIS3 glucan (Figure3). The ''C NMR spectrum of the glucan from double mutant cells showed a structure with linkage ratios similar to those of a k r e l A null mutant(Figure4). These results suggest thatboth Kre l p and Kre 1l p are required to produce the mature glucan polymer at wild-type levels. KREII interacts with KRE6: As with krel I , mutants with defects in the KRE6 gene show a reduced level of a wild-type (1 + 6)-glucan polymer (ROEMER and BUSSEY1991). We tested whether mutants harboring thek r e l l - 1 allele interacted with a kre6A::HZS3 containing mutant, by attempting to construct strains containing mutations in both genes. Diploid HAB6 13 heterozygous for k r e l l - 1 and kre6A::HIS3 was made, sporulated, and13tetrads dissected and analyzed. Three tetrads were parental ditypes for k r e l l - 1 and
kre6A::HIS3, with 4 killer-resistant spore progeny, 2 being His+. Nine tetrads were tetratypes and gave 3 viable spores, 1 sensitive His-, 2 killer resistant with one of these His+, and the fourth inviable spore divided 2-6 times before ceasing to grow. One tetrad was a nonparental ditype, and gave 2 viable, sensitive His- spore progeny. This segregation pattern is consistent with the k r e l l - I kre6A::HIS3 double mutation leading to lethality. Additional cell wall analyses: The major cell wall components were examined in krel and krel I mutants to explore the possibility that defects in these genes led to pleiotropic effects incell wall synthesis. The alkali-soluble glucan plus mannan fractions were determined from a wild-type strain, and isogenic kreIA::HIS3, krel1A::TRPI or krelA::HIS3 krel1A::TRPI double disruptant strains. The results indicated that total alkali-soluble glucan plus mannan fractions in the wild-type or mutant strains were not significantly different, with an average value of 167.1 f 3 &mg dry weight. After digestion with Zymolyase and dialysis to remove alkali-soluble (1 + 3)-glucan, no differences in the remaining fractions were seen, with average values of 72.8 f 4 pg/mg dry weight. Alkali-insoluble (1 + 3)-glucan levels were also measured, and appeared normal in these strains (legend, Table4), while the alkali-insoluble (1 + 6)-glucan fraction showed areductionfrom the wild type in either single mutant, anda cumulative reduction (7580%) in the kreIA::HIS3 krel1A::TRP double mutants (Table4). Cell wall chitin levelsin wild-type and mutant strains were examined by Calcofluor White fluorescence. While no observable differences in fluorescence were seen between the parental strain and either a krelA::HIS3 or kre1lA::TRPl single mutant, the krelA::HIS3krel1A::TRP doublemutantsappeared to stain more intensely. Whether this apparent increase incell wall chitin levels is the basis the suppression seen in the krel A::HIS3 krel I A::TRP double mutants has not presently been determined. A kreIA::HIS3, and a krelI::HIS3 mutant,along with the isogenic parentalstrain 7B, were assayed for (1 + 3)-/3-glucan synthase activity in vitro according to CABIBand KANC (1987). No significant difference inspecific activity was seen between these mutants and the wild-type strain, withvalues for strain 7B; 97.4 f 13.6 nmol glucose incorporated/ mg protein/hr, krelA::HIS3; 109 25.5 nmol glucose incorporated/mg protein/hr, and krel l::HIS3; 133.1 2 49.2 nmol glucose incorporated/mgprotein/hr. Finally, glycogen levels were tested by staining with iodine vapors, and appeared indistinguishable in wild type, kreIA::HIS3,krelIA::TRP, and kreIA::HIS3 krel1A::TRP doublemutant strains. Cumulatively, these results suggest the cell wall defect observed in mutants harboring disruptions of K R E l or K R E l I is
*
Yeast KRE Genes
A) Krel lp KreSp Kre Kre6p
1p Core Glucan
UDPG
p( 1 -+6)
+
p( 1 +6)
backbone addltlon chaln synthesls and p( 1-3) branchlng
Mature
p( 1 + 6 )
Glucan
side
Kre9p K r e 1 Op
Kre UDPG
UDPG
Krel 1p KreSp Kre6p
1p
Core Glucan I
K r e 1 1p-
Core Glucan 1 1
M a t u r e pC 1 + 6 ) Glucan Type I
~
~
Krel p
M a t u r e PC 1 -+6) Glucan Type I I
Homolog? KreSp Kre6p
Kre9p K r e 1 Op
FIGURE5.-Possible models for the functional role of K R E l I in the assembly of (1 -+ 6)-glucan. (A) (1 + 6)-Glucan is synthesized in a pathway from UDP-glucose, with Krelip, Kre6p and Krel Ip involved in making a glucan with (1 + 3)-branch points (Core Glucan), which is a substrate for Krelp-dependent side-chain addition to elaboratethe mature glucan polymer. (B) Depicts two independent pathways for (1 + 6)-glucan synthesis. Krel Ip would be involved only in the upper pathway, in making Core Glucan I, whose matureproduct is largerthan type I1 glucan. A second pathway (lower part of B), uses a postulated functional homolog of Krel Ip to make core glucan 11, which is matured to the smaller type I1 glucan found in a K R E I l disrupted strain. The role of Kre9p and KrelOp in these processes is not known.
confined to a reduction in the (1 + 6)-P-glucan polymer. DISCUSSION
This study extends genetic analysis of the yeast cell surface using killer toxin resistant mutants. The repeated isolation of resistant alleles in seven genes indicates that further isolations using this method are unlikely to identify new genes. We have identified three genes implicated in cell wallsynthesis. Mutations in K R E 9 and K R E I I show reduced levels of cell wall (1 + 6)-glucan. KREIO genetically interacts with
847
K R E l l , as cellswith mutations in both genes grow more slowly than cells with a mutation in either gene alone. Disruption of the K R E l I gene and analysis of the resultant phenotype indicates that cell walls contain a reduced amount of a smaller glucan polymer. T h e exact structure of the (1 + 6)-glucan polymer remains unknown, but it is composed of a (1 -6)linked core, with (1 3)-branch points and (1 + 6)linked side chains. Failure to make normal amounts of the wild-type (1 + 6)-glucan structure likely leads to an impaired toxin receptor and killer resistance. Functions and genetic interactions of the K R E l , K R E l l and KRE6 genes involved in 8-glucan synthesis: The K R E l gene product is involved in (1 + 6)-side chain synthesis; the defect in strains with krel mutations is consistent with a failure to extend (1 -+ 6)-glucan sidechains to(1 -+ 3)-branchpoints on a (1 + 6)-glucan core (BOONEet al. 1990). The P-glucan phenotype in a k r e l l mutation appears to preserve the linkage arrangement of the wild-type polysaccharide in a polymer that is somewhat smaller insize. T h e cumulative reduction of the amount of the (1 + 6)-glucan polymer in the cellwall, andthe slow growth phenotype of the k r e l A k r e l double IA mutant strain argues that the defect in the polymer is more extreme than that caused by eithera krel A or a krel I A mutation alone. The glucan structures and the mutantphenotypes imply thatKrel l p precedes Krelp in a biosynthetic pathway, and that the glucan made in a krel I A mutant remains as a substrate for Kre 1p. The finding that double mutations in K R E l 1 and KREG are lethal is consistent with these two gene products being required for (1 -+ 6)-glucan synthesis. Single mutants in these genes lead to reduced synthesisof a wild-type polymer. Thus both KREG and K R E l l appeartoprecede K R E l andtogetherare essential for growthof this strain. In seeking to explain these results in an ordered glucan assembly pathway, there are a number of possible functions for Krel lp, outlined in Figure 5. In a simple explanation (Figure 5A),Krel l p may be required in some way for efficient glucan synthesis, perhaps as a cytoplasmic component, with Kre6p as a membrane-associated synthase complex. In a second model, Krel l p may be an essential component of one of several (1 + 6)-P-D-glucan synthases (Figure 5B)and involved in synthesis of a type I glucan. When Krel l p is absent, a redundant second synthase, possibly involving a Kre 1 1 p homolog, makes a lesser amount of a somewhat smaller glucan, called type I1 in Figure 5B. This model proposes two independent pathways for synthesis of the (1 -+ 6)-glucan (Figure 5B). The wild-type glucan would then be composed of a mixed population of the two types of mature (1 -+ 6)-glucans, which differ slightly in size. A formal, andmorecomplex, possibility is that
-
J. L. Brown et al.
848
Krel l p is not involved in glucan synthesis but in preventing glucan degradation, and thatin its absence wild-type (1 + 6)-glucan is made normally, but then partially degraded. Pathway of (1 -+ 6)-glucan synthesis: The genes identified here augment the previous genetic model of (1 + 6)-glucan assembly in S. cerevisiae (MEADENet al. 1990). While the roles of the Kre9p and KrelOp are unknown, analysis of glucan from krel l A disrupted cells suggests that the polymer is synthesized in a sequential manner involving several gene products, some of which are in the yeast secretory pathway, and Krell p which is likely cytoplasmic. Identification of these genes should allow access to their products, and eventually, through biochemistry and cell biology, to their functions. The search for further genes interacting with those described could usefully be continued, as our presentunderstandingremains incomplete. We thank TERRY ROEMERand ARTHURPERLINfor helpful discussions, ANNE-MARIE SDICUfor technical assistance, DIANEOKI for manuscript preparation, SILVIBILODEAU at the NMR Laboratory, Universiti de MontrGaI, for NMR spectroscopy, SHAWN DELANEY for oligonucleotides, ALANBENDER and CHARLE~ ~ Y E for R strains, and LINDARILESfor help with mapping. Supported by a Biotechnology Strategic Grantfrom The Natural Sciences and Engineering Research Council of Canada.
LITERATURECITED AL-AIDROOS,K.,and H. BUSSEY,1978 Chromosomal mutants of Saccharomyces cereuisiae affecting the cell wall binding site for killer factor. Can. J. Microbiol. 2 4 228-237. BADIN, J., C. JACKSON and M. SCHUBERT, 1953 Improved method for determination of plasma polysaccharides with tryptophan. Proc. SOC.Exp. Biol. Med. 8 4 228-291. BENDER, A., AND J. R. PRINGLE,1989 Multicopy suppression of the cdc24 budding defect in yeast by CDC42 and three newly identified genes including the ras-related RSRl. Proc. Natl. Acad. Sci. USA 8 6 9976-9980. BOONE, C., S. S. SOMMER, A. HENSEL and H. BUSSEY, 1990 Yeast KRE genes provide evidence for apathway of cell wall 8-glucan assembly.J. Cell Biol. 1 1 0 1833-1843. BULAWA, C. E. , 1992 CSD2,CSD3, and CSD4 genes required for chitin synthesis in Saccharomyces cerevisiae: the CSD2 gene product is related to chitin synthases and to developmentally regulated proteins in Rhizobium species and Xenopus laevis. Mol. Cell. Biol. 12: 1764-1776. BULAWA, C. E.,M. L. SLATER, E. CABIB, J. Au-YOUNG, A. SBURLATI, W.L. ADAIRand P.W. ROBBINS,1986 The S. cerevisiae structural gene for chitin synthase is not required for chitin synthesis invivo. Cell 4 6 2 13-225. BUSSEY, H., 1991 K l killer toxin,a pore-forming protein from yeast. Mol. Microbiol. 5: 2339-2343. BUSSEY, H., W. SACKS,D. GALLEY and D. SAVILLE,1982 Yeast killer plasmid mutations affecting toxin secretion and activity and toxin immunity function. Mol. Cell. Biol. 2: 346-354. CABIB,E., and M. S. KANG,1987 Fungal 1,3-@-glucansynthase. Methods Enzymol. 1 3 8 637-642. FARKAS, I., T . A. HARDY, A. A. DEPAOLI-ROACH and P. J. ROACH, 1990 Isolation of the GSYl geneencoding yeastglycogen synthase and evidence for the existence of a second gene. J. Biol. Chem. 265: 20879-20886.
FLEET,G. H., andH. J. PHAFF,1981 Fungal glucans-structure and metabolism, pp. 416-440 in Encyclopedia of Plant Physiology,New Series, Vol. 13b, Plant Carbohydrates 11, edited by and F. A. LOEWUS. Springer-Verlag, Berlin. W. TANNER FOREMAN, P. K., R. W. DAVISand A. B. SACHS,1991 The Saccharomyces cerevisiae RBP4gene is tightly linked to the TIF2 gene. Nucleic Acids Res. 1 9 2781. HARTLAND, R. P., G.W. EMERSON and P. A. SULLIVAN, 1991A secreted @-glucan-branchingenzyme from Candida albicans. Proc. R. SOC.Lond. B 246: 155-160. HAUSLER, A., and P. W. ROBBINS,1992 Glycosylation inSaccharomycescerevisiae: cloning and characterization of an a-1,2mannosyltrans-ferase structural gene. Glycobiology 2: 77-84. HILL,K . , C.BOONE, M. GOEBL,R. PUCCIA,A. M. SDICUand H. BUSSEY, 1992 Yeast KREP defines a new gene family encoding probable secretory proteins, and is required for the correct N-glycosylation of proteins. Genetics 1 3 0 273-283. HOFFMAN, C. S. and F. WINSTON, 1987 A ten-minute DNA p r e p aration from yeast efficiently releases autonomous plasmids for transformation of Escherichia coli. Gene 57: 267-272. HUTCHINS, K., and H. BUSSEY, 1983 Cellwall receptor for yeast killer toxin: involvement of (1 + 6)-@-~-ghcan. J. Bacteriol. 154: 161-169. ITO, H., Y. FUKUDA,M. MURATA and A. KIMURA, 1983 Transformations of intact yeast cells with alkali cations. J. Bacteriol. 153: 163-168. JAMAS, S., C. K. RHA and A. J. SINSKEY, 1986 Morphology of yeast cell wall as affected by genetic manipulation of p(1-6)glycosidic linkage. Biotechnol. Bioeng. 28: 769-784. KATO, K . , 1981 Ultrastructure of the plant cellwall: biochemical viewpoint, pp. 29-46 in Encyclopedia of Plant Physiology, New Series, Vol. 13b, Plant Carbohydrates 11, edited by W. TANNER and F. A. LOEWUS. Springer-Verlag, Berlin. KLEBL,F., and W. TANNER, 4989Molecular cloning ofa wall exo1,3-@-glucanasefrom Saccharomyces cerevisiae.J. Bacteriol. 171: 6259-6264. MANNERS,D. J., A. J. MASSONand J. C. PATTERSON, 1973a The structure of a p-(1 + 3)-D-ghCan from yeastcellwalls. Biochem. J. 135: 19-30. MANNERS, D. J., A. J. MASSONand J. C. PATTERSON, 1973b The structure of a p-(1 + 6)-~-ghcan from yeastcellwalls. Biochem.J. 135: 3 1-36. MARTINAC,B., H. ZHU,A. KUBALSKI,X. ZHOU,M. CULBERTSON, H. BUSSEYand C. KUNG,1990 YeastK1 killer toxin forms ion channels in sensitive yeast spheroplasts and in artificial liposomes. Proc. Natl. Acad. Sci. USA 87: 6228-6232. MEADEN,P., K. HILL,J. WAGNER, D. SLIPETZ,S. S. SOMMER and H. BUSSEY, 1990 The yeastKRE5 gene encodes a probable endo-plasmic reticulum protein required for (1 + 6)@-~-glucan synthesis and normal cell growth. Mol. Cell. Biol. 1 0 301330 19. MORTIMER, R. K., and D. SCHILD,1980 The genetic map of Saccharomyces cerevisiae. Microbiol. Rev. 44: 5 19-57 1. MULLER,P. P., H. TRACHSEL and P. LINDER, 1989 Genetic localization of the Saccharomyces cerevisiae genes tiyland hf2. Curr. Genet. 16: 127-128. PERKINS, D., 1949 Detection of linkage in tetrad analysis. Genetics 3 8 187-197. RIBAS,J. C., M. DIAZ,A. DURANand P. PEREZ,1991 Isolation and characterization of Schizosaccharomyces pombe mutants defective incellwall(1-3)@-D-gluCan. J. Bacteriol. 173: 34563462. T.,and H. BUSSEY, 1991 Yeast @-glucan synthesis: KRE6 ROEMER, encodes a predicted type I1 membrane protein required for glucan synthesis in vivo and for glucan synthase activity an vitro. Proc. Natl. Acad. Sci. USA 88: 11295-1 1299. RONCERO, C., M. H. VALDIVIESO, J. C.RIBAS and A. DURAN, 1988 Isolation and characterization of Saccharomyces cereuisiae
Yeast KRE Genes mutants resistant to calcofluor white. J. Bacteriol. 170: 1950-
1954. ROSE,M. D., P. NOVICK, J. H. THOMAS,D. BOTSTEIN and G. R. FINK,1987 A Saccharomyces cerevisiae genomic plasmid bank based on a centromeric-containing shuttle vector. Gene 6 0
237-243. SAMBROOK, J., E.F. FRITSCHand T. MANIATIS,1989 Molecular Cloning:ALaboratoryManual, Ed. 2. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. SANGER, F., S. NICKLEN and A. R. COULSON, 1977 DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74: 5463-5467. SHAW,J. A., P.C. MOL, B. BOWERS,S. J. SILVERMAN, M. H. VALDIVIESO, A. DURAN and E. CABIB,1991 The function of chitin synthases 2 and 3 in the Saccharomyces cerevisiae cell cycle. J. Cell Biol. 114 1 1 1-123. SHERMAN, F., G. R. FINKand J. B. HICKS,1986 Methods in Yeast Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. SHIOTA,M., T. NAKAJIMA, A. SATOHAN, M. SHIDAand K. MATSUDA, 1985 Comparison of @-glucanstructures in a cell wall mutant of Saccharo-myces cerevisiae andthe wild type. J. Biochem. 9 8 1301- 1307.
a49
VALDIVIEW, M. H., P. C . MOL,J. A. SHAW,E. CABIBand A. DURAN, 1991 C A L I , a gene required for activity of chitin synthase 3 in Saccharomyces cereuisiae.J. Cell Biol. 114 101-109. VAN RINSUM, J., F. M. KLIS and H. VAN DEN ENDE,1991 Cell wall glucomannoproteins of Saccharomyces cereuisiae mnn9. Yeast 7:
717-726. VASQUEZ DE ALDANA, C. R., J. CORREA, P. S. SEGUNDO, A. BUENO, A. R. NEBREDA, E. MENDEZand F. DEL REY,1991 Nucleotide gene, E X G l , of sequence of the exo-l,3-B-glucanase-encoding the yeast Saccharomyces cereuisiae. Gene 97: 173-182. WESSELS,J. G. H., and J. H.SIETSMA,1981 Fungal cellwalls: a survey, pp 352-394 in Encyclopedia of Plant Physiology, New Series, Vol. 13b,Plant Carbohydrates 11, edited by W. TANNER and F. A. LOEWUS. Springer-Verlag, Berlin. G. H. EICHINGER, R. WONG,H. C., A.L. FEAR,R.D. CALHOON, MAYER,D. AMIKAM,M. BENZIMAN, D. H. GELFAND, J. H. MEAD,A.W. EMERICK, R. BRUNER,A.BEN-BASSATand R. TAL,1990 Genetic organization of the cellulose synthase operon in Acetobacter xylinum. Proc. Natl. Acad. Sci. USA 87:
8130-8
134. Communicating editor: E. W. JONES