. 13: 1–8 (1997)

Yeast PIG Genes: PIG1 Encodes a Putative Type 1Phosphatase Subunit that Interacts with the YeastGlycogen Synthase Gsy2p*

CHRISTINE CHENG, DONGQING HUANG AND PETER J. ROACH*

Department of Biochemistry and Molecular Biology, Indiana University School of Medicine, Indianapolis,Indiana 46202-5122, U.S.A.

Received 3 June 1996; accepted 9 July 1996

The biosynthesis of glycogen involves multiple proteins that associate with each other and the glycogenmacromolecule. In efforts to understand the nature of these proteins, a two-hybrid screen was undertaken to detectproteins able to interact with Gsy2p, a major form of glycogen synthase in Saccharomyces cerevisiae. Two positivesexpressed proteins derived from genes designated PIG1 and PIG2, on chromosomes XIIR and IXL respectively.PIG1 codes for a protein with 38% identity over a 230 residue segment to Gac1p, a protein thought to be a type 1protein phosphatase targeting subunit whose loss impairs glycogen synthesis. Pig2p has 30% identity to the proteincorresponding to an open reading frame, YER054, on chromosome V. Deletion of PIG1 on its own had little effecton glycogen storage but, in combination with loss of GAC1, caused a more severe glycogen-deficient phenotype thanseen in gac1 mutants. This result is consistent with Pig1p being functionally related to Gac1p and we propose thatPig1p may be a type 1 phosphatase regulatory subunit. Delection of PIG2, YER054, or both genes together causedno detectable change in glycogen metabolism under the conditions tested. Gac1p, Pig1p, Pig2p and the YER054p arethe only four proteins coded by the yeast genome that share a conserved segment of 225 residues, designated theGVNK motif, that is identifiable also in RGI, the mammalian type 1 phosphatase targeting subunit.

— glycogen; phosphatase type 1; targeting; PIG1; PIG2

INTRODUCTION

Glycogen is a branched polymer of glucosebelieved to function as a metabolic reserve in manycells (Preiss and Walsh, 1981; François et al.,1995). Formation of the basic á-1,4-glycosidiclinkages of the polymer is mediated by glycogensynthase, for which two genes, GSY1 and GSY2,exist in Saccharomyces cerevisiae (Farkas et al.,1990, 1991). The á-1,6-glycosidic branchpointscharacteristic of glycogen are introduced by thebranching enzyme encoded by the GLC3/GHA1gene (Rowen et al., 1992; Thon et al., 1992).In addition, we recently identified two self-glucosylating proteins encoded by the GLG1and GLG2 genes that may be homologues ofmammalian glycogenin, with a specialized functionin the initiation phase of glycogen biosynthesis(Cheng et al., 1995).

Glycogen synthase is controlled by proteinphosphorylation and three potential COOH-terminal sites have been identified (Hardy andRoach, 1993). Several protein phosphatases havebeen implicated in the control of yeast glycogenmetabolism and mutations in phosphatase genescan affect glycogen storage (Peng et al., 1990;Posas et al., 1991; Feng et al., 1991; François et al.,1992; Hardy and Roach, 1993; Cannon et al.,1994; Clotet et al., 1995). The best candidate fordirect dephosphorylation of glycogen synthase isthe type 1 phosphatase composed of Glc7p as thecatalytic and Gac1p as the regulatory subunit(François et al., 1992; Stuart et al., 1994). GLC7 isan essential gene but certain non-lethal alleles,such as glc7-1 (Cannon et al., 1994) and glc7Y170

(Hisamoto et al., 1994), lead to impaired glycogenaccumulation. In the case of glc7-1, the mutationdoes not destroy the phosphatase activity per sebut rather its ability to interact with the Gac1pregulatory subunit (Stuart et al., 1994). Similarly,*Corresponding author.

CCC 0749–503X/97/010001–08? 1997 by John Wiley & Sons Ltd

disruption of GAC1 leads to reduced glycogenstorage (François et al., 1992). Gac1p has sequencesimilarity, some 26% identity over a region of2230 residues, with a mammalian protein, RGI,which is thought to target type 1 phosphatase toglycogen (Stalfors et al., 1985) and membranes(Tang et al., 1991). Recently, the sequence of aputative liver form of RGI has also been deter-mined (Doherty et al., 1995). The concept isemerging generally, for both protein kinases andphosphatases, that regulatory subunits may notonly have direct influences on activity but can alsobe involved in functional targeting (Hubbard andCohen, 1993), as is exemplified by the study ofGac1p (Stuart et al., 1994).In the course of our studies of yeast glycogen

metabolism, we undertook a screen for proteinsthat interact with Gsy2p, the major nutritionallyregulated form of glycogen synthase in yeast, usingthe two-hybrid system. We identified two clonesderived from what we termed the PIG1 and PIG2genes (for Proteins Interacting with Gsy2p). Wehave proceeded to make gene disruptions and toexamine relevant phenotypes of pig1 and pig2mutants. We hypothesize that Pig1p may functionas a type 1 phosphatase regulatory subunit.

MATERIALS AND METHODS

Yeast strains and methodsEG328-1A, used as a wild-type strain, was

derived from JC302-26B (Cannon et al., 1986) andobtained from Dr Kelly Tatchell, Louisiana StateUniversity. Other related strains are shown inTable 1. For the two-hybrid screening, strain

YPB2 was used (Chien et al., 1991). Yeast genomicDNA was purified according to the method ofHoffman and Winston (1987). Routine transfor-mation of yeast by the lithium acetate method wasas described by Schiestl and Gietz (1989). Cellnumbers were determined by counting 100–200cells in a hemocytomer. Other standard methodsfor yeast genetic analysis were as described (Roseet al., 1990).

Yeast two-hybrid screenThis screen was carried out as described by

Cheng et al. (1995). Briefly, it utilized the system ofChien et al. (1991) to search for proteins capableof interacting with Gsy2p sequences, expressedfrom the plasmid pGBT9-GSY2, to screen ayeast library in the pGAD vector. From 3·6#105

initial Leu+Trp+ transformants, three were con-firmed as positives dependent on the presence ofboth the pGAD and pGBT9-GSY2 plasmids. Atthe later stages of the screening, quantitativeâ-galactosidase solution assays (Platt et al., 1972)were performed.

Disruption of genesDisruption of the PIG1, PIG2 and YER054 loci

in yeast was achieved by one-step homologousrecombination (Rothstein, 1983) utilizing DNAfragments generated by polymerase chain reaction(PCR), essentially as described by Cheng et al.(1995). PCR primers contained 45 bp of targetgene sequence followed by 21 bp that matchedpBluescript sequences flanking the chosen markergene in an appropriate pRS plasmid (Sikorski and

Table 1. Yeast strains.

Strain Genotype Source

EG328-1A MATá trp1 leu2 ura3-52 K. TatchellEG328-2D MATá trp1 leu2 ura3-52 gac1::LEU2 K. TatchellYPB2 MATa ura3-52 his3-200 ade2-101 S. Fields

lys2-801 trip1-901 leu2-3,-112 can1R

gal4-542 gal80-538 LYS2::GAL1-HIS3URA3(GAL 17-mers)-lacZ

CC10 MATá trp1 leu2 ura3-52 pig2::URA3 This studyCC11 MATá trp1 leu2 ura3-52 yer054::URA3 This studyCC12 MATá trp1 leu2 ura3-52 pig2::URA3 yer054::TRP1 This studyCC13 MATá trp1 leu2 ura3-52 pig1::URA3 gac1::LEU2 This studyCC14 MATá trp1 leu2 ura3-52 pig1::URA3 This study

2 . .

Hieter, 1989). Deletion strains were derivedfrom haploid wild-type (EG328-1A) or gac1(EG328-2D) cells.

Analysis of glycogen accumulationFor routine assessment of glycogen accumu-

lation, yeast were grown on YPD or SD plates andexposed to iodine vapor. For quantitative analysisof glycogen and glycogen synthase activity, yeastwere grown in 1 liter cultures of the indicatedmedium at 30)C from a saturated overnightculture. Aliquots were removed at the indicatedtimes, the cells were harvested by centrifugationand stored at "80)C until analysis. The frozencell pellets were resuspended in 0·3 ml of asolution of 50 m-Tris–HCl, 1 m-EDTA,3 m-dithiothreitol, 100 m-NaF, 1 m-phenylmethylsulfonylfluoride, 0·1 m-Ná-p-tosyl--lysine chloromethyl ketone, 5 m-benzamidine,0·25 ìg/ml leupeptin, 0·5 ìg/ml aprotinin, pH 7·5.Cells were broken with glass beads. Glycogen wasdetermined enzymatically, as described previously(Hardy et al., 1994).

Sequence analysesSequence alignments and manipulations were

made using the GeneWorks program (Intelligen-etics). Database searches of GenBank were carriedout via the National Center for BiotechnologyInformation home page on the World Wide Webusing the programs blastp or tblastn. Searches ofTIGR human cDNA database, level I, were alsomade via the World Wide Web.

RESULTS

Identification of the PIG1 and PIG2 genesFrom a two-hybrid screen for genes encoding

proteins able to interact with Gsy2p, three posi-tives were identified, as detailed in Materials andMethods. One of these, GLG2, codes for a self-glucosylating protein that is a yeast homologue ofmammalian glycogenin (Cheng et al., 1995). Theother two genes, PIG1 and PIG2, had no counter-parts in the databases at the time of their originalidentification except that the two-hybrid fragmentof the protein encoded by PIG1 contained a shortregion with similarity, 14 out of 17 residues iden-tical, to the Gac1p sequence. Subsequently, thecomplete PIG1 and PIG2 genes have appeared inthe databases.

The original PIG1 two-hybrid clone, pGAD-PIG1, contained the 316 COOH-terminal residuesof the protein encoded by open reading frame(ORF) 9328.2 on cosmid 9328 from chromosomeXII (YLR273c). Pig1p has a sequence of 648amino acids with 38% identity to Gac1p over aregion of 230 residues (Figure 1). Like Gac1p,Pig1p is an acidic protein (Table 2) with a highcontent of Ser plus Thr residues, 23·3% for Gac1pand 18·3% for Pig1p. Both of these properties areshared with the mammalian phosphatase targetingsubunit RGI, which has some sequence similarityto Pig1p and Pig2p. The initial PIG2 clone,pGAD-PIG2, contained the 338 COOH-terminalresidues corresponding to an ORF (YIL045w)from chromosome IX. Pig2p is 538 amino acidslong and has weak sequence similarity to Pig1pand Gac1p. Database searches revealed the pres-ence of another gene, corresponding to an ORF onchromosome V (YER054), encoding a protein of548 residues with 30% identity to Pig2p spreadover the whole sequence (Figure 1). Pig2p andYER054p are basic proteins, which do not haveunusually high Ser plus Thr contents. Like Gac1pand Pig1p, they both have a somewhat elevatedAsn content (Table 2). YER054 was also identifiedin a two-hybrid screen as encoding a proteininteracting with Glc7p and was designated GIP2on this basis (Tu and Carlson, unpublished data).

Interactions of Pig1p and Pig2p with glycogensynthase

Qualitative filter â-galactosidase assays wereemployed to identify interactions of proteinexpressed from pGAD-PIG1 and pGAD-PIG2with Gsy2p expressed from pGBT9-GSY2 duringthe initial screening. After characterization of thepGAD-PIG plasmids, quantitative â-galactosidaseassays were also performed (Figure 2). Interactionof full-length Gsy2p with the Pig1p fragment wassignificantly stronger than with the Pig2p protein.In addition, other forms of Gsy2p were tested fortheir interactions. The plasmid pGBT-GSY2ÄCexpresses Gsy2p lacking the 81 COOH-terminalresidues, whose loss did not impair interactionwith either Pig1p or Pig2p. Plasmid pGBT-GST2ÄN expresses only the 81 COOH-terminalresidues of Gsy2p and in this case there wasno evidence for interaction with either Pig1por Pig2p. Both the full-length Gsy2p and itsCOOH-terminus were expressed as judged byWestern analysis using antibodies specific to the

3

COOH-terminus of Gsy2p (data not shown).Therefore, the negative results with Pig1p andPig2p could not be explained by lack of expressionof the COOH-terminal fragment of Gsy2p. Weconclude that binding of neither Pig1p nor Pig2pinvolves the extreme COOH-terminus of Gsy2p,the region involved in phosphorylation controls(Hardy and Roach, 1993).

Analysis of PIG2 and YER045p deletion strainsIn order to assess whether PIG2 might be

involved in glycogen metabolism, we first made a

Figure 1. Alignment of Pig1p and Pig2p with other relevant proteins. All of the proteinsshown contain two highly conserved sequence motifs (indicated by the filled rectangles)which are separated by 28–48 intervening residues. Gac1p and Pig1p also share a thirdregion of high similarity, 14 identities in a stretch of 17 residues, as indicated by the openrectangles. Pig2p and YER054p share 230% identity over their entire sequences, as isrepresented by the thicker lines. The numbers of the right indicate the number of aminoacid residues in the polypeptide.

Table 2. Some properties of Gac1p, mammalian RGIand the proteins identified in the present study.

pI* Mr*No. ofresidues

Asncontent

Ser+Thrcontent

Gac1p 5·26 88,522 794 10·0 23·3Pig1p 4·86 74,064 648 8·6 18·3Muscle RGI 4·66 124,033 1109 5·4 18·2Liver RGI 5·26 32,589 284 4·2 14·1Pig2p 8·86 61,869 538 11·7 12·3YER054p 9·17 62,249 548 10·2 13·1

*Predicted from the amino acid sequences.

Figure 2. Interactions of Pig1p and Pig2p sequences withGsy2p as judged by two-hybrid analysis. Plasmids expressingfull-length Gsy2p (pGBT9-GSY2), a COOH-terminallytruncated form (pGBT9-GSY2ÄC) or the COOH-terminusalone (pGBT9-GSY2ÄN) were tested in combination withplasmids expressing portions of either Pig1p or Pig2pby two-hybrid assay using quantitative â-galactosidasemeasurements (see Material and Methods). We also performedcontrols with single plasmids, from left to right, pGBT9-GSY2,pGBT9-GSY2ÄC, pGBT9-GSY2ÄN, pGAD-PIG1 andpGAD-PIG2.

4 . .

targeted deletion of PIG2 and found no effect onglycogen accumulation as judged by iodine stain-ing after growth on rich (YPD) or defined (SD)media (data not shown). Because Pig2p sharedsignificant sequence similarity with YER045p, wereasoned that the proteins might have relatedfunctions, making it necessary to delete both genesin order to detect certain phenotypes. Thus, wealso disrupted PIG2 and the YER045 ORF.Even in these double mutants, glycogen storageappeared normal under all conditions tested (datanot shown). Deletion of these genes also had noobvious effects on growth rates on YPD or SDplates.

Effects of PIG1 and/or GAC1 disruption onglycogen accumulationWe adopted a similar approach to test whether

PIG1 had any influence on glycogen metabolism(Figure 3). Glycogen accumulates at late log andstationary phase in cultures and cells grown ondefined SD medium accumulated significantlymore polysaccharide than cells grown on rich YPDmedium. Disruption of PIG1 on its own had littleeffect on glycogen accumulation. However, asnoted above, Pig1p has significant sequencesimilarity to Gac1p and so we also tested PIG1mutations in the presence of a GAC1 deletion. Asexpected, mutation of GAC1 resulted in a signifi-cant reduction in glycogen stores after growth oneither medium. Interestingly, pig1 gac1 double

mutants had a more extreme glycogen defect thangac1 mutants, as is especially evident after longertimes of culture. The result was true for cells grownon either rich (YPD) or defined (SD) media.Similar results were obtained when glycogen accu-mulation was monitored by iodine staining (datanot shown). There was no large difference in thetotal glycogen synthase activity in gac1, pig1 orpig1 gac1 mutants (data not shown). The differ-ences in glycogen accumulation noted abovewere not due to any alterations in growth sincemutations in GAC1 and PIG1, alone or in combi-nation, had no effect on growth in either rich ordefined media (data not shown). Mutation of PIG1thus accentuates the glycogen storage defect ofgac1 mutants, and suggests that Pig1p may havesome similar functions to Gac1p.

DISCUSSION

The first protein identified by a two-hybrid screenfor proteins that interact with Gsy2p was theself-glycosylating protein, Glg2p, which is almostcertainly involved in glycogen synthesis (Chenget al., 1995). From studies of null mutants of thePIG2 gene, we thus far have no evidence for itsinvolvement in glycogen metabolism. Loss of thePIG2 gene alone did not affect glycogen depositionand even when we additionally deleted theYER054 ORF, which encodes a protein with sig-nificant sequence similarity to Pig2p, there was no

Figure 3. Glycogen accumulation in pig1, gac1 and pig1 gac1 mutants. Wild-type ormutant cells, as indicated, were grown for 18 or 42 h and glycogen was determinedenzymatically as described in Materials and Methods. The amount of glycogen isnormalized to the amount of protein measured in the cell extracts. Cells were growneither on SD (A) or YPD (B) media.

5

effect on glycogen accumulation. It should benoted that the interaction of Gsy2p with Pig2p wassignificantly weaker than with either Pig1p (Figure2) or Glg2p (Cheng et al., 1995). However, wecannot yet formally exclude that Pig2p (orYER054p) has a functional relationship withglycogen synthase under some as-yet untestedconditions.Pig1p readily identifies Gac1p in database

searches and has regions of significant sequencesimilarity, 38% over a core region of 230 residues;the COOH-terminal segment of this region hassequences sharing 14/17 identities (Figure 1). Inaddition, both have some sequence similaritywith the mammalian RGI. Furthermore, thoughdeletion of PIG1 alone caused little reduction inglycogen storage, its loss clearly accentuated thedefective glycogen accumulation in gac1 mutants.Thus, Pig1p may not ordinarily have a major rolein glycogen metabolism but, in the absence ofGac1p, may sustain some Gac1p functions asregards glycogen synthesis. When both Pig1p andGac1p are not present, an even stronger defect inglycogen is observed. Taking together these resultson glycogen storage phenotype with the sequencesimilarity between Pig1p and Gac1p leads us topropose that Pig1p and Gac1p may have relatedfunctions and that Pig1p may be a phosphatasesubunit. Interestingly, K. Thatchell (personal com-munication) has observed an interaction betweenGsy2p and Gac1p, which would be consistent withour results. Targeting of phosphatase by bothGac1p and Pig1p may involve direct interactionwith substrates. This property could represent adifference from the mammalian RGI-containingphosphatase in which the targeting would beindirect via binding to glycogen (however, see nextparagraph).One intriguing observation from this work is

that, although Pig2p and YER054p are overallonly weakly similar in sequence to Gac1p andPig1p, all four proteins contain two regions of verystrong homology (Figure 1). One is a region of 15residues, motif II, that is also found in liver andmuscle RGI type 1 phosphatase subunits (Figure4). Notable are two Asn and one Tyr residue thatare completely conserved. Database searches withthis motif identify a rice glucoamylase (Ashikariet al., 1986). The other motif is of about 25residues, dubbed the GVNK domain, that againcan be identified in both liver and muscle RGI type1 phosphatase subunits (Figure 5). The yeastgenome encodes only these four examples of

GVNK sequences. However, searches of othersequence databases using the GVNK consensusregion as a query identified three other sequences(Figure 5). Two of these are human EST sequencesfor which little sequence is available outside of theconserved region. The third is again a sequencefrom rice glucoamylase (Ashikari et al., 1986); thisis the most divergent sequence. Nonetheless, fourresidues, G, V, N and K, are completely invariantand, in several other positions, only conservativedifferences are detected.The structure and function corresponding to the

GVNK motif is still unknown but some interestingpossibilities can be raised. For example, this isthe very region of RGI that was postulated byHubbard and Cohen (1993) to be a candidatefor interacting with glycogen. Their argumentwas based on sequence similarity to a region ofglycogen phosphorylase known from structuralwork to interact with glycogen (Newgard et al.,1989). As seen in Figure 5, the sequence similaritywith phosphorylase is quite poor and it remains tobe seen if the GVNK region of RGI is related toglycogen binding. However, support for a rolein carbohydrate binding comes from the riceglucoamylase. In fact, both of the conserved motifsare present in a 90 residue NH2-terminal region ofglucoamylase that has been implicated in bindingto raw starch (Ashikari et al., 1986). Furtherexperimentation is suggested by these analyses.Even if the GVNK domain is not involved in

interacting with glycogen, it is likely that it under-lies some conserved structural unit. RGI andGac1p are well established as subunits of type 1phosphatase, and we are proposing, from thisstudy, a similar role for Pig1p.YER054p/Gip2 may interact with Glc7p (Tu

and Carlson, unpublished data) and it is conceiv-able that Pig2p might have similar properties.

Figure 4. Conserved motif II. Motif II is a region of 15residues that is highly conserved in the yeast proteins Pig1p,Pig2p, Gac1p and YER054p, the mammalian type 1 phos-phatase targeting subunit RGI from rabbit muscle (M) or ratliver (L), and rice glucoamylase (D00049). Residues identical inmore than three sequences are boxed. The lower bullets indicatepositions where residues are highly conserved.

6 . .

Thus, a common feature of many of the proteinscontaining the GVNK sequence is interaction witha phosphatase and this motif might somehow beimplicated in this function, mediating perhapsbinding to the type 1 catalytic subunit. Note thatthe original PIG1 two-hybrid clone did not containthe GVNK motif, suggesting it has no role inGsy2p interaction. Even though we have not so farimplicated Pig2p in any function related to glyco-gen metabolism, it is intriguing that it is one ofonly four yeast proteins containing the GVNKmotif, two of which were identified by a two-hybrid screen as interacting with glycogen syn-thase. It seems improbable that the presence ofthis conserved sequence motif and the ability tointeract with Gsy2p are linked purely by chance.

ACKNOWLEDGEMENTS

This research was supported by grant DK42576from the National Institute of Diabetes, Digestiveand Kidney Diseases. We thank Drs AnnaDePaoli-Roach, Mark Goebl and Ron Wek forhelpful discussions in the course of this investi-gation. We are also grateful to Ron Wek for acritical review of the manuscript.

REFERENCES

Ashikari, T., Nakamura, N., Tanaka, Y., et al. (1986).Rhizopus raw-starch degrading glucoamylase: Itscloning and expression in yeast. Agric. Biol. Chem. 50,957–964.

Cannon, J. F., Gibbs, J. B. and Tatchell, K. (1986).Suppressions of the ras2 mutation of Saccharomycescerevisiae. Genetics 113, 247–264.

Cannon, J. F., Pringle, J. R., Fiechter, A. and Khalil, M.(1994). Characterization of glycogen-deficient glcmutants of Saccharomyces cerevisiae. Genetics 136,485–503.

Cheng, C., Mu, J., Farkas, I., Huang, D., Goebl, M. G.and Roach, P. J. (1995). Requirement of self-glucosylating initiator proteins, Glg1p and Glg2p, forglycogen accumulation in Saccharomyces cerevisiae.Mol. Cell. Biol. 15, 6632–6640.

Chien, C.-T., Bartel, P. L., Sternglanz, R. and Fields, S.(1991). The two-hybrid system: a method to identifyand clone genes for proteins that interact with aprotein of interest. Proc. Natl. Acad. Sci. USA 88,9578–9582.

Clotet, J., Posas, F., Hu, G.-Z., Ronne, H. and Arino, J.(1995). Role of protein phosphatase 2A in the controlof glycogen metabolism in yeast. Eur. J. Biochem. 229,207–214.

Doherty, M. J., Moorhead, G., Morrice, N., Cohen, P.and Cohen, P. T. W. (1995). Amino acid sequenceand expression of the hepatic glycogen-binding (GL)-subunit of protein phosphatase-1. FEBS Lett. 375,294–298.

Farkas, I., Hardy, T. A., Goebl, M. G. and Roach, P. J.(1991). Two glycogen synthase isoforms in Saccharo-myces cerevisiae are coded by distinct genes that aredifferentially controlled. J. Biol. Chem. 266, 15602–15607.

Farkas, I., Hardy, T. A., DePaoli-Roach, A. A. andRoach, P. J. (1990). Isolation of the GSY1 geneencoding yeast glycogen synthase and evidence for theexistence of a second gene. J. Biol. Chem. 265, 20879–20886.

Figure 5. Conserved motif I (GVNK). This highly conserved segment of Pig1p, Pig2p,Gac1p and YER054p is shown together with the corresponding sequence from fourmammalian proteins, rabbit muscle (M) RGI, rat liver (L) RGI, and two human ESTs(Z42059 and R10183), as well as a sequence from rice glucoamylase (D00049). Residuesidentical in more than five sequences are boxed. The lower bullets indicate positions atwhich the residues are highly conserved. Above is shown a region from rabbit muscleglycogen phosphorylase (GP) implicated in glycogen binding and suggested by Hubbardand Cohen (1993) to align with RGI. Asterisks indicate residues in the phosphorylasesequence that show some similarity to, or are not incompatible with, the consensus ofthe GVNK motif.

7

Feng, Z., Wilson, S. F., Peng, Z.-Y., Schlender, K.,Riemann, E. M. and Trumbly, R. J. (1991). The yeastGLC7 gene required for glycogen accumulationencodes a type I protein phosphatase. J. Biol. Chem.266, 23796–23801.

François, J. M., Thompson-Jaeger, S., Skroch, J.,Zellenka, U., Spevak, W. and Tatchell, K. (1992).GAC1 may encode a regulatory subunit for proteinphosphatase type 1 in Saccharomyces cerevisiae.EMBO J. 11, 87–96.

François, J., Blazquez, M. A., Arino, J. and Gancedo,C. (1995). Storage carbohydrates in the yeast Sac-charomyces cerevisiae. In Zimmermann, F. K. (Ed.),Yeast Sugar Metabolism. Technomics Pub. Co. Inc.,Lancaster, PA, in press.

Hardy, T. A., Huang, D. and Roach, P. J. (1994).Interactions between cAMP-dependent and SNF1protein kinases in the control of glycogen accumu-lation in Saccharomyces cerevisiae. J. Biol. Chem. 269,27907–27913.

Hardy, T. A. and Roach, P. J. (1993)., Control of yeastglycogen synthase-2 by COOH-terminal phosphoryl-ation. J. Biol. Chem. 268, 23799–23805.

Hisamoto, N., Suigimoto, K. and Matsumoto, K.(1994). The Glc7 type 1 protein phosphatase ofSaccharomyces cerevisiae is required for cell cycleprogression in G2/M. Mol. Cell. Biol. 14, 3158–3165.

Hoffman, C. S. and Winston, F. (1987). A ten-minuteDNA preparation from yeast efficiently releasesautonomous plasmids for transformation ofEscherichia coli. Gene (Amst) 57, 267–272.

Hubbard, M. J. and Cohen, P. (1993). On targetwith new mechanism for the regulation of proteinphosphorylation. TIBS 18, 72–177.

Newgard, C. B., Hwang, P. K. and Fletterick, R. J.(1989). The family of glycogen phosphorylases: struc-ture and function. Crit. Rev. Biochem. Mol. Biol. 24,69–99.

Peng, Z.-Y., Trumbly, R. J. and Reimann, E. M. (1990).Purification and characterization of glycogen synthasefrom a glycogen deficient strain of Saccharomycescerevisiae. J. Biol. Chem. 265, 13871–13877.

Platt, T., Müller-Hill, B. and Miller, J. H. (1972). InMiller, J. H. (Ed.), Experiments in Yeast Genetics.Cold Spring Harbor Laboratory, Cold SpringHarbor, NY, pp. 352–355.

Posas, F., Clotet, J. and Arino, J. (1991). Saccharomycescerevisiae gene SIT4 is involved in the control ofglycogen metabolism. FEBS Lett. 279, 341–345.

Preiss, J. and Walsh, D. A. (1981). The comparativebiochemistry of glycogen and starch. In Ginsburg, V.and Robbins, P. (Eds), Biology of Carbohydrates,Vol. 1. Wiley, New York, pp. 199–314.

Rose, M. D., Winston, F. and Hieter, P. (1990).Methodsin Yeast Genetics, A Laboratory Course Manual. ColdSpring Harbor Laboratory Press, Cold SpringHarbor, N.Y.

Rothstein, R. (1983). One-step gene disruption in yeast.Methods Enzymol. 101C, 202–210.

Rowen, D. W., Meinke, M. and LaPorte, D. C. (1992).GLC3 and GHA1 of Saccharomyces cerevisiae areallelic and encode the glycogen branching enzyme.Mol. Cell. Biol. 12, 22–29.

Schiestl, R. H. and Gietz, R. D. (1989). High efficiencytransformation of intact yeast cells using singlestranded nucleic acids as a carrier. Curr. Genet. 16,339–346.

Sikorski, R. S. and Hieter, P. (1989). A system of shuttlevectors and yeast host strains designed for efficientmanipulation of DNA in Saccharomyces cerevisae.Genetics 122, 19–27.

Stralfors, P., Hiraga, A. and Cohen, P. (1985). Theprotein phosphatases involved in cellular regulation;purification and characterization of the glycogen-bound form of protein phosphatase-1 from rabbitskeletal muscle. Eur. J. Biochem. 149, 295–303.

Stuart, J. S., Frederick, D. L., Varner, C. M. andTatchell, K. (1994). The mutant type I protein phos-phatase encoded by glc7-1 from Saccharomycescerevisiae fails to interact productively with the GAC1encoded regulatory subunit. Mol. Cell. Biol. 14,896–905.

Tang, P. M., Bondor, J. A., Swiderek, K. M. andDePoaoli-Roach, A. A. (1991). Molecular cloning andexpression of the regulatory (RGL) subunit of theglycogen-associated protein phosphatase. J. Biol.Chem. 266, 15782–15789.

Thon, V. J., Vigneron-Lesens, C., Marianne-Pepin, T.,et al. (1992). Coordinate regulation of glycogenmetabolism in the yeast Saccharomyces cerevisiae.Induction of glycogen branching enzyme. J. Biol.Chem. 267, 15224–15228.

8 . .