Copyright 0 1994 by the Genetics Society of America

The SUP35 Omnipotent Suppressor Gene Is Involved in the Maintenance of the Non-Mendelian Determinant [psi'] in the

Michael D. Ter-Avanesyan, Adilya R. Dagkesamanskaya, Vitaly V. Kushnirov and Vladimir N. Smirnov

Institute of Experimental Cardiology, Cardiology Research Center, 121552 Moscow, Russia Manuscript received November 20, 1992 Accepted for publication October 12, 1993

ABSTRACT The SUP35 gene of yeast Saccharomyces cerevisiae encodes a 76.5-kD -associated (Sup35p), the Gterminal part of which exhibits a high degree of similarity to EF-la elongation factor, while its N-terminal region is unique. Mutations in or overexpression of the SUP35 gene can generate an omnipotent suppressor effect. Inthe present study the SUP35wild-typegene was replaced with deletion alleles generated in vitrothat encode Sup35p lackingall or a part of the unique N-terminal region. These 5'deletion alleles lead, in a haploid strain, simultaneously to anantisuppressor effect and to loss of the non-Mendelian determinant [psi']. The antisuppressor effect is dominant while the elimination of the [psi'] determinant is a recessive trait. A set of the plasmid-borne deletion alleles of the SUP35 gene was tested for the ability to maintain[psi']. It was shown that the first 114amino acids of Sup35p are sufficient to maintain the [psi'] determinant. We propose that the Sup35p serves as a trans-acting factorrequired " for the maintenance of [psi'].

HE phenomenon of informational suppression is cytoplasmic genome to [psi-] (YOUNGand COX1971; T successfully exploited as a tool for elucidating the MCCREADYet al. 1977). genetic control of translational ambiguity in both pro- Herein we provide evidence that the omnipotentsup- karyotic and eukaryotic cells. Nonsense suppressors, the pressor SUP35, also called SUP2, SUPP, SAL3, SUF12 best studied class of informational suppressors of the and GSTl (INGE-VECHTOMOVand ANDRIANOVA 1970; yeast Saccharomyces cerevisiae, can be divided into two HAWHORNE and LEUPOLD1974; GERLACH1975; CROUSET categories, codon-specific and codon-nonspecific (om- and TUITE1987; CULBERTSONet al. 1982; IKUCHIet al. nipotent). As a rule, the nonsensesuppressors with pre- 1988), participates in maintenance of [psi'] in yeast cise codon specificity arise by anticodon mutations of cells. It was shown earlier that the SUP35 gene encodes tRNA genes (PIPERet al. 1976; GOODMAN et al. 1977; a 76.5-kD ribosome-associated protein in which the BROACHet al. 1981). In contrast, omnipotent suppres- C-terminal part, beginningfrom methionine-254, exhib- sors are presumed not to be mutations in tRNA genes its a high degree of similarityto EF-la elongation factor because of their lack of codon specificity (HAWTHORNE (KUSHNIROVet al. 1988; WILSONand CULBERTSON1988; and LEUPOLD1974; LIEBMANand ALL-ROBYN1984; ONO DIDICHENKOet al. 1991). Mutations in or overexpression et al. 1984). Itwas shown that mutations in these genes of this gene decrease translational fidelity, resulting in may cause an increased level of translational ambiguity anomnipotent suppressor effect (SURGUCHOVet al. (SURGUCHOVet al. 1980; MASUREKAR et al. 1981). The 1984; CHERNOFFet al. 1992). Overexpression of SUP35 function of the gene products in protein synthesis of also drasticallyreduces the growth rate of [psi'] strains only a few of the omnipotent suppressors are known possibly indicating that thecombination of different fac- (~~A~UREKARet al. 1981; SURGUCHOVet al. 1984; EUSTICE tors increasing translational ambiguity, such as the et al. 1986;ALL-ROBYN et al. 1990). Another approachto [psi'] determinant andextra copies of the SUP35 gene, identifying protein components involved in the control leads to a level of inaccuracy incompatible with cell vi- of translational accuracy consistsin the identification of ability (DAGKESAMANSKAYA and TER-AVANESYAN 1). 199 trans-acting mutations that modify the expression of Our previous studies revealed at least two functional suppressors (for reviews, see SHERMAN1982; SURGUCHOV domains within Sup35p. Deletion analysis ofthe SUP35 et al. 1984; HINNEBUSCHand LIEBMAN1991). In addition gene shows the essential function of theprotein to to nuclear mutations, cytoplasmically inherited factors becontained within the evolutionarily conserved [psi'] and [eta'] modify the action of suppressors. In Gterminal region. The N-terminal region of the spite of extensive studies, the physical entities cor- protein, which varies evolutionarily both in length and respondingto these cytoplasmic determinantsare sequence in S. cerevisiae, Pichia pinusand man still unknown [for reviews, see Cox et al. (1988) and (KUSHNIROVet al. 1990; HOSHINOet al. 1989) is not es- HINNEBUSCHand LIEBMAN(1991)l. Mutations in nuclear sential. Overexpression of this N-terminal region or the PNM genes have been described that convert a [psi'] entire Sup35p leads toomnipotent suppression and

Genetics 137: 671-676 Uuly, 1994) 672 M. D. Ter-Avanesyan et al. reduced growth of strains. The expression of [psi'] Aa the Cterminal portion of Sup35p, on single or multi- G copy plasmids, causes an antisuppressor effect (TER- SUP35 AVANESYANet al. 1993). ATG ATG ATG TAA In this study we demonstrate that deletionswithin the 1I sup35-AN3 N-terminal region of Sup35p cause elimination of the [psi'] determinant. We suggest that Sup35p serves as a ~~p35-AN2 tramacting factor required for the maintenance of [psi']. SUp35-ANE MATERIALS AND METHODS Strains and plasmids: The following segregants of the dip loid strain H19 described in (DAGKESAMANSKAYAand TER- B 1234 AVANESYAN1991) were used: 1A-H19 (MATa ade2-1lysl-1 his3-11, 15leu2-3, 112SUQ5[psi']);5V-H19 (MATaade2-I - 2.4kb canl-100 leu2-3, 112 ura3-52 SUQ5 [psi']). A strain 10B- - 2,3kb H49 of genotype MATa ade2-1 SUQ5 leu2-3,112 lysl his3- -2,lkb 11,15 karl-1 [rho'] [psi'] was derived from a cross between -1,8kb the strain CKOl (MATa leu1 karl-1 canR [psi'] [rho']) that was obtained from P. M. LUND(Oxford, Great Britain) and 1A-H19 [psi']. The [psi-] derivatives of all of these strains were also used. FIGURE1.Ponstruction of deletion alleles of the chromo- All deletion alleles used in this study,with the exception of somal SUP35 gene. (A) 5"Deletion alleles used for replace- sup35-AN2 and sup35-AN3, were obtained by the deletion of ment of the wild-type SUP35 gene. The solid bar represents internal restriction fragments of SUP35 gene as shown in Fig- the coding sequence of the SUP35 gene, lines (-) non- ures 1 and 3. In the case of of 3'-terminal SUP35 deletion coding regions. The restriction map of the SUP35 gene is constructs, terminated in adjacent plasmid DNA shown at the top of this picture. The first, second and third thus adding to the proteinencoded extensions of 7-17 amino in-frame ATG codons and terminator TAA codon are shown acids ofnon-functional sequence. The sup35-AN2 and sup35- under the restriction map. The XhoI-XbaI fragments carrying AN3 deletion alleles were generated by joining the SUP35 the sup35-A N3, sup35-AN2or sup35-ANB alleles were inte- promoter to the EcoRV or HpaI restriction sites of the SUP35 grated into a through recombination at free gene so that translation can start from ATG codons corre- DNA ends (see text for details). (B) Southern blot analysis. sponding to Met-124 or Met-254, respectively. Alldeletion con- DNAfrom the sup35dNdeletion mutanrs 1-5V-H19 (lane l), structs, with the exception of ABst-Hind, were able to direct 2-5V-H19 (lane 2), 3-5V-H19 (lane 3) and original strain 5V- the synthesis ofcorresponding truncated versions of Sup35p, H19 (lane 4) was isolated, digested with BcnI and hybridized as was shown by Western blot analysis (TER-AVANESYANet al. to the M13 single stranded probe, containing the BcnI-XbaI 1993). The S'deletion alleles sup35-AN3, sup35-AN2 and fragment of the SUP35 gene. A 2.4-kb fragment is observedfor sup35-ANB (Figure 1) used ingene-replacement experiments the strain 5V-H19 corresponding to the wild type SUP35 gene were cloned in the plasmid pFL44 (F.LACROUTE, Gif sur Yvette, and a 2.3-kb band is present in strain 3-5V-Hl9, a 2.1-kb band France). Other deletion constructs of the SUP35 gene were in strain 2-5V-H19 and a 1.8-kb band in strain 1-5V-H19 as cloned in the plasmids pEMBLyex4 (CESAREMand MURRAY expected for the sup35-ANB,sup35-AN2 and sup35-AN3 1987) and pRG415 (R. CABER, Evanston, Illinois). For details, alleles, respectively. see TER-AVANESYANet al. (1993). The YEp13 plasmid havebeen described elsewhere (BROACHet al. 1979). yielded only rare complementing diploids in the mixed cul- Genetic methods: Standard yeast genetic procedures of ture and were easily distinguishable fromthe KARl segregants crossing, sporulation and tetrad analysis were usedto construct that gave a massive complementation reaction. Nonsuppres the appropriate strains and analyze gene segregation. Nutri- sive petites in transformants of the strain 1-5V-H19 were ob- tional markers were scored by growth on synthetic (SC) me- tained by ethidium bromide treatment (COLDBRING et al. dium lacking specificamino acids or nucleic acidbases. Sporu- 1970). For performing "cytoduction" experiments, strains of lation and YF'D media were also used (SHERMANet al. 1986). opposite matingtype, one of which desthe kurl-1 mutation Yeast strains werecured of the [psi'] determinant by growth that blocks karyogamy, were mated by mixing them togetheron on YF'D medium supplemented with 5 mM guanidine hydro- the surface of a WD plate and incubated for1 day at 30". This chloride (GuHCl). The [psi-] colonies of ade2-1 SUQ5 car- mixture was then spread on appropriate mediumand cytoduc- rying strains were chosen by pink color and adenine require- tants were selected as described in detail in RFSULTS. ment because the serine-insertingdominant suppressor SUQ5 Yeast transformation: DNA transformation of lithium (also called SUPl6) cannot suppress the add-1 ochre muta- acetate-treated yeast was done as described by ITO et al. (1983). tion inthe absence of the [psi'] determinant (Cox 1965; LIEE Manipulations with DNA and RNA Isolation of plasmid MAN et al. 1975; ONOet al. 1979). Strains to bescored for [psi] and yeast chromosomal DNA and total cellular RNA was per- were crossedto SUQ5 [psi-] tester and the efficiency ofSUQ5 formed as given elsewhere(MANIATIS et al. 1982; HOFFMANand was examined in the meiotic progeny. If SUQ5 was efficient WINSTON1987; SHERMANet al. 1986). Restrictionendonuclease enough to suppress ade2-1 in the progeny of such a cross, the cleavage, ligation and other enzymatic procedures as well as tested strain was scored as [psi']. The meiotic segregation of Southern and Northern blot analyses were done as described the karl phenotype was determined by the strong-weak by MANIATIS et al. (1982) and SHERMANet al. (1986). The DNA complementation testof CONDEand FINK (1976). All seg- and RNA blots were hybridized to M13 single-stranded probe regants of the diploid heterozygous for the karl-1 mutation canying the SUP35 fragment, labeled by primer extension were crossed with MATa or MATa ade8 testers. karl strains according to MESSING(1983). Antisuppression and [psi'] Loss 673

RESULTS TABLE 1 Construction of deletions in the chromosomalSUP35 data from crosses of the sup35-AN mutauts with the [psi'] and [psi-] variants of gene: The sequence of Sup35p can be divided into strain 1A-H19 three regions (KUSHNIROVet al., 1988). The Gterminal Segregation in tetrads (Ade+:Ade-) region (amino acids 254-685) shows similarity to elon- Cross gation factor EF-la, while the N-terminal (amino acids Strain with1:3 2:2 0:4 1-123) and middle (amino acids 124-253) regions are 1-5V-H19 [psi'] 77 14 1 unique and differ significantly in their amino acid con- 1-5V-H19 [psi-] 0 0 30 tent. The beginning of the middle and Gterminal re- 2-5V-H19 [psi'] 6 2 0 2-5V-H19 [psi-] 0 0 13 gions are defined by the second (Met-124) and third 3-5V-H19 [psi'] 10 3 0 (Met-254) methionines in the protein. Transformation 3-5V-H19 [psi-] 0 0 15 of the 5V-H19 [psi'] strain with multicopy or centro- Strains 1-5V-H19,2-5V-H19and 3-5V-H19 carry sup35-AN3, sup35 meric plasmids carrying either one of the 5'deletion AN2 and sup35-ANB chromosomal deletion alleles, respectively.The deviation from the 2 Ade+:2Ade- segregation pattern observed in the alleles of the SUP35 gene (termed sup35-AN3, sup35- progeny of diploids formed in crosses with the [psi'] tester resembles A N2, or sup35-ANB or in common sup35-AN, Figure the excess of [psi-] spore cultures usually yielded by diploids het- l), causes a non-suppressed phenotype, resulting in erozygous for the recessive pnm mutations (COXet al., 1980). pink colonies and adenine auxotrophy (TER-AVANESYAN et al., 1993). pressed:2 nonsuppressed spores in every tetrad, and all To further examine the propertiesof the deletedver- segregants were expected to be [psi'l. In fact, the dip sions of the SUP35 gene, the chromosomal wild-type loids formed by crosses withthe [psi'] tester yielded, as SUP35 gene was replaced with the sup35-AN3, sup35- a rule, tetrads that segregated 2 Ade':2 Ade- rather than AN2 or sup35-ANB allele. This was performed by co- 4 Ade':O Ade-. That the adenine-requiring segregants transformation of the [psi'] haploid strain 5V-H19 with from these tetrads carried sup??-A N deletion alleles was the LEU2-canying YEp13 plasmid and DNA fragments verified in several cases by Southern blot analysis (data obtained after XhoI and XbaI digestion of the plasmids not shown). Surprisingly, the diploids from crosses with pFL443ATG, pFL442ATG and pFL44ABst carrying the [psi-] tester gave no suppressed segregants (Table alleles sup35-AN?, sup35-AN2 and sup35-AB, respec- 1).Nonsuppressed segregants derived from the crosses tively (Figure 1). Approximately 5-10%of leucine- to the [psi'] tester appeared to be [psi-], rather than independent colonies in each transformation experi- the expected [psi']. This was shown as follows. The ment were pink and adenine-requiring. Three of them sup35-AN-carrying adenine-requiring segregants from (1-5V-H19,2-5V-H19and 3-5V-H19 presumably carrying several tetrads showing monogenic segregation for ad- sup35-AN?, sup35-AN2 and sup35-ANB, respectively) enine auxotrophy were crossed again with the [psi-] were studied in detail. Plasmid-less colonies of these tester strain 1A-H19. The diploids formed by these transformants were selected after streaking on non- crosses did not segregate spore cultures with the sup- selective WD medium. The color of the colonies did not pressed phenotype. The control crosses with the [psi'] depend onthe presence of the YEpl3 plasmid. Because tester resulted in [psi'] diploids, since they produced as these transformants were pink and adenine-requiring, a rule tetrads with 2 Ade':2 Ade- segregation (data not we expected them tocarry the antisuppressor sup35-AN shown). alleles instead of the wild-type SUP35 gene. This was Taken together, these observations suggest that the proved by Southern analysis (Figure 1). sup35-AN deletion alleles cause elimination of the The sup35-AN mutations cause elimination of the [psi'] determinant in haploid strains. The elimination [psi+] determinant: All independentlyobtained of the [psi'] factor is a recessive trait, since crosses be- sup35-AN mutants were crossed with the [psi'] and tween the sup35dNbearing strains and the [psi'] tester [psi-] variants of the tester strain 1A-H19.Hybrids produced diploids that could segregate whiteand adenine- formed by these crosses were pink and unable to grow independent spore cultures and therefore carried the on adenine omission medium. Therefore the antisup- [psi'] determinant. This property of deletion mutations in pressor effect of allthree sup35-ANalleles is a dominant the SUP35 gene makes them similar to the recessive pnm trait. Diploids formed between strains carrying mutations already described (Cox et aL 1980). sup35dNalleles and either[psi'] or [psi-] versions of Rescue of [psi'] by the expression of truncated ver- the strain 1A-H19 wereexpected tohave the same geno- sions of Sup35p: The results presented above suggest type:[psi'] ade2-l/ade2-l SUQ5/SUQ5 sup35-AN/ that the N-terminal part of Sup35p is necessary for SUP35, since all gene replacement experiments were [psi'] maintenance. The identification of the minimal performed in the [psi+]-carrying strain 5V-H19. Since region of Sup35p sufficient to maintain the [psi'] de- all of these diploids were heterozygous for one of the terminant was performed as follows(Figure 2).First, the antisuppressor sup35-AN alleles and homozygous for [rho'] and [psi'] determinants were transferred by cy- the SUQ5 allele, the diploids should segregate 2 sup- toduction from the [rho'][psi'] karl-1 strain, 10B- 674 M. D. Ter-Avanesyan et al.

10B-H49 [psi'][rho'] Rescue of (ade2-1 SUQS karl-1) the @si+] > I I/ \r II I I SUP35 Transfer of [psi'] and [rho'] + by cytoduction aBst - + aHpa + Transformants of 1-5V-H19 [psi-] [rho-] ABal + (adel-1 SUQS sup35-/ N) ASal + I aBcl + Selection of [rho'] cytoductants ~Bal2 + \I/ aBst-Hind - Transformants of 1-5V-H19 [psi?] [rho'] - (ade2-1 SUQ5 sup35-/ N) 2ATG-&Sal - 1 2ATG-aBCI Transfer of [psi'] AECO + + FIGURE3.-The ability of the sup35 deletion constructs to 10B-H49 [psi-] (ade2-1 SUQS karl-1) rescue the [psi'] determinant. Coding regions of the con- structs are represented by solid bars. "+ ," ability to rescue the \1 [psi'] determinant; "-," inability to rescue the [psi']. For Selection of [psi'] cytoductants other details, see the legend to Figure 1.

FIGURE2.-Scheme of cytoduction experiments performed minant could grow on the adenine omission medium. to identify the plasmids with the sup35 deletion mutations capable of rescuing the [psi'] determinant. Genotypes of Transformants of strain 1-5V-H19 and diploid cells can- strains are presented in MATERIALAND METHODS. Cytoductants not grow on this medium because they carry the domi- were selected as described in the text. [psi?],the [psi] status nant antisuppressor mutation sup35-AN3. Strain 1OB- is to be determined. HI9 is Ade- because SUQ5 cannot suppress the ade2-1 ochre mutation in the absence of The selectivity H49, to the sup35-AN3-bearing strain, 1-5V-H19, that [psi']. of this system allowed usto develop a qualitative test for had beentransformed with plasmidsthat carry different [psi'] transfer by cytoduction (Figure 4). The adenine- truncated versions of the SUP35 gene (Figure 3). [rho'] independent clones selected in these experiments pos- and [psi'] show high coincidence of transfer (Cox et al. sess the same nuclear genotype as the tester strain 10B 1988). Therefore, cytoductants and diploids were se- H49 and can be easily converted to theadenine- lected from the mating mixtures of cells by transfer to dependent phenotype after curing of the histidine omission medium containing glycerol asa sole [psi'] determinant by plating on GuHCl-containing medium. carbon source. Respiratory competent colonies pheno- Results of the study of the ability of different truncated typically indistinguishable in other respects from the versions of Sup35p to rescue the determinant in transformants of strain 1-5V-H19, were scored as cyto- [psi'] the sup35-AN3-carrying strain are summarized in Fig- ductants. Surprisingly, none of the cytoductants ob ure 3. It is noteworthy that in spite of the fact that most tained showed the noticeable decrease in growth rate of experiments described above were performed with expected for [psi'] strains carrying SUP35containing multicopy plasmids, the results did not depend on the multicopy plasmids (DAGKESAMANSKAYAand TEK- plasmid copy number, since the centromeric plasmid AVANESYAN1991; TEK-AVANESYANet al. 1993), although carrying the "ASal" construct could efficiently rescue the some of them possessed this non-Mendelian determi- determinant in the 1-5V-H19 strain (datanot nant as will be demonstrated below. This means that [psi'] shown). chromosomal antisuppressor sup35-AN3 allele amelio- rated thedeleterious effect of overexpression of Sup35p DISCUSSION or its C-terminally truncated versions in [psi'] strains, possibly due to reduction of the high translational am- Our previous studies have shown that single or mul- biguity caused by the combined action of multicopy ticopy plasmids carrying the sup35-AN alleles generate SUP35 plasmids and [psi']. The [psi] status of the cy- in haploid strains an antisupressor effect but do not toductants was revealed in the second round of cyto- cause [psi'] elimination (TEK-AVANESYANet al. 1993). In duction experiments (Figure 2). Only those [rho+] cy- this report we have shownthat replacementof the chro- toductants of transformants that could transfer [psi'] by mosomal wild-type SUP35 gene with the sup35-AN de- cytoduction to the [psi-] tester strain, 1ORH49, were letion alleles inthe ade2-l SUQS [psi'] strain causes, in considered to possess the [psi'] determinant. Mating of addition to a previously observed non-suppressed phe- the transformants with strain 1OBH49 allowed us to iso- notype, a loss of the [psi'] determinant. These effects late cytoductants selectively, since only those cells ofthe are expressed differently in heterozygous diploids: the recipient strain 1OB-H49 that received the [psi'] deter- antisuppressor effect is dominant while the elimination Antisuppression and [psi'] Loss 675

A B AB AB causing amino acid substitution at position 58 of the encodedprotein (S. M. DOEL, C. R. NIERRAS,S. J. MCCRFADYand B. S. COX, personal communication). This confirms our conclusion about the critical role of 7 the N-terminal domain of Sup35p in the maintenanceof 8 the [psi'] determinant. It is noteworthy that in contrast 9 to the sup35-ANdeletion mutations, the loss 10 [psi'] caused by the PNM2 pointmutation is a dominanttrait 11 and a heterozygous diploid gradually loses over succes- 12 sive generations the abilityto produce [psi'] spores (MCCREADYet al. 1977).Another feature that distin- guishes the sup35-AN deletion alleles from the PNM2 point mutation is that, unlike deletion mutations, point mutation in the N-terminal part of the protein does not FIGURE4.-An illustration of the qualitative tests to deter- generate an antisuppressor phenotype. It is possible to mine the capacity of theplasmids to support maintenance of explain the dominant phenotype of the PNM2 point [psi'] determinant instrain 1-5V-H19. The vertical streaks: mutation by the assumption that the mutant proteinacts (A) 1A-H19 (MATa nde2-1 SUQ5 [psi-]); (B)10BH19 by titration of its wild-typecounterpart or some factor(s) (MATa ade2-1 SUQ5 karl-1 [psi-]). The strains streaked essential for the maintenanceof It should be also horizontally are cytoductants of transformants of 1-5V-H19 [psi']. (MATa ade2-1 SUQ5 sup35dN3) with multicopy plasmids noted that expression of Sup35p is necessary but not suf- carrying wild-typeSUP35 gene and SUP35 deletion constructs ficient for the [psi'] phenotype since strains carryingthe being tested for ability to transfer [psi']: 1, SUP35 gene; 2, wild-type SUP35 gene may either possess or not the [psi'] "ABst"; 3, "AHpa"; 4, "ABal"; 5, "ABa12";6, "ASal"; 7, "ABcl"; determinant. Moreover, the expression of the SW35gene 8, "ABst-Hind";9, "2ATGASal"; 10, "2ATGABcl";11, "AEco"; does not depend on the[pa] status of the cell, as was de- 12 is the transformant carrying the control plasmid pEM- BLyex. The complete genotypes of strains mentioned above termined by RNA blot analysis (not shown). are presented in MATERIAL AND METHODS. The appearance of Simultaneous but separateexpression of both the N- [psi'] cytoductants depends upon the presence of the karl-1 and Gterminaldomains of Sup35p is not thesame as the mutation in the tester strain. expression of the entire protein,since segregants com- biningthe determinant, chromosomal sup35- of the [psi'] determinant is a recessive trait, i.e., heterozy- [psi'] gous diploids possess an antisuppressor phenotype but AN3 allele and plasmid-borne alleles of the SUP35 gene cany the [psi'] determinant. Dominant and recessive mu- encoding a Gterminally truncated (e.g., the tationscausing either an antisuppressor phenotype or "AEco," "ABal," "ABcl" or "ASal" constructs) possess an antisuppressor phenotype. Therefore,for the Sup35 [psi'] loss have been described in S. CertTVisiae (MCCREADY protein to possess the wild-type activity the integrity of and COX1973; MCCRFADYet uL 1977; Cox et aL 1980). The sup35-ANalleles differ from thesemutations by their ability these domains is required. In spite of this, the to cause both effects simultaneously. N-terminal domain of the Sup35 protein expressed It seems likely that the dominant antisuppressoreffect alone allows maintenance of the [psi'] determinant. is not the directconsequence of mutations that simply In conclusion we would like to emphasize that the delete or reduce some of the SUP35 functions. In con- Gterminal domainof Sup35p is essential for cell viability trast, the recessive pnm-like effect suggests the loss of while the N-terminal domain of this protein is essential some function that is essential for [psi'] maintenance. for [psi'] "viability." We suggest that the Sup35 protein serves as a truns- We wish to thank B. S. Cox (Oxford, Great Britain) for providing acting factor essential for the [psi'] maintaining ma- us with the sequence of thePNMP mutant allele prior to publication chinery. Deletion analysis performed in this study re- and for critical reading of ourmanuscript. The authors arealso grate- vealed that thefirst 114 amino acids of Sup35p (e.g.,the ful to P. M. LUND (Oxford,Great Britain) for the strain CKOI. This work wassupported in part by grant SRG 1 ROS TWO01 29-01 from the "AEco" construct) are sufficient for maintenance of the Fogarty International Center. [psi'] determinant. This sequence also represents the smallest region of Sup35p whose overexpression gener- LITERATURECITED ates an omnipotent suppressor effect. Thus, this study ALL-Ronw, J. A., N. BROWN,E. OTAKAand S. W. LIERYAN,1990 has revealed a specific role of this domain of Sup35p in Sequence and functional similarity between a yeast ribosomal maintaining the [psi'] determinant. Theseresults are in proteinand the E. coli S5 ram protein. Mol. Cell.Biol. 10: 6544-6553. agreement with the data obtained recently in the labo- BROACH,J. R., J. N. STRATHBKNand J. B. HICKS,1979 Transformation ratory of B. S. Cox (Oxford, Great Britain). These au- in yeast: Development ofa hyhrid cloning vector andisolation of thors have cloned and sequenced the PNM2gene, the CAN1 gene. Gene 8: 121-127. BROACH,.].R., L.R. FRIEDMANand F. SHERMAN,1981 Correspondence which was found to be identical to the SUP35 omnipo- of yeat UAA suppressors to cloned tRNA,,,,, genes. J. Mol. Biol. tent suppressor. A mutation in this gene was identified 150 375-387. 676 M. D. Ter-Avanesyan et al.

CESARENI,G., and A. H. MURRAY,1987 Plasmid vectors carrying KUSHNIROVV. V., M. D. TER-AVANESYAN,M. V. TELCKOV, A. P. SURGUCHOV, the replicationorigin of filamentoussingle-stranded phages, V.N. SMIRNOVand S. G. INGE-VECHTOMOV,1988 Nucleotide se- pp. 135-154 in GeneticEngineering: Principles and Methods, quence of the supZ(sup35) gene of Saccharomycescereuisiae. Vol. 4, edited by J. K. SETLOW.Plenum Press, New York. Gene 66 45-54. CHERNOFF,Y. O., S. G.INGE-VECHTOMOV, I. L. DERKACH, M. V. PTWSHKINA, KUSHNIROV,V.V.,M. D. TER-AVANEW,S. A. DIDICHENKO,~.N. SMIRNOV, 0. V. TARUNINA,A. R. DAGKESAMANSKAYAand M. D. TER-AVANESYAN, Yu. 0. CHERNOFF,I. L. DERKACH,0. N. NOVIKOVA,S. G. 1992 Dosagedependent translational suppression in yeast Sac- INGE-VECHTOMOV,M. A. NEISTATand I. I. TOLSTORUKOV,1990 Di- charomyces cereuisiae. Yeast 8: 489-499. vergence and conservation of SUPZ( SUP35) gene ofyeasts Pichia CONDE,J., and G. R. FINK,1976 Amutant of Saccharomyces cereuisiae pinus and Saccharomyces cereuisiae. Yeast 6: 461-472. defective fornuclear fusion.Proc. Natl. Acad. Sci. USA 73: LIEBMAN,S. W., J. W. STEWARTand F. SHERMAN,1975 Serine substitu- 3651-3655. tions caused by an ochre suppressor in yeast. J. Mol. Biol. 94: COX,B. S., 1965 psi, a cytoplasmic suppressor of super-suppressor in 595-610. yeast. Heredity 20: 505-521. LIEBMAN,S. W., and J. A. ALL-ROBYN,1984 A non-Mendelian factor, COX,B. S., M. F. TUITEand C. J. MUNDY,1980 Reversion from sup- [eta'], caused lethality of yeast omnipotent-suppressor strains. pression to nonsuppression in SUQ5 [psi'] strains of yeast: the Curr. Genet. 8: 567-573. 95: classification of mutations. Genetics 589-609. MANIATIS,T., E. F. FRITSCHand J. SAMBROOK,1982 Molecular Cloning: COX,B. S., M. F. TUITEand C. S. MCLAUGHUN,1988 The psi factor of A Laborato-ry Manual. Cold Spring Harbor Laboratory, Cold yeast: a problem of inheritance. Yeast 4: 159-178. Spring Harbor, N.Y. CROUZET,M., and M. F. TUITE,1987 Genetic control of translational MASUREKAR, M., E. PALMER,B.4. ONO,J. M. WILHELMand F. SHERMAN, fidelity in yeast: molecular cloning and analysis of the allosup- 1981 Misreading of the ribosomal suppressor SUP46 due to an pressor gene sa13. Mol. Gen. Genet. 210 581-583. altered 40s subunit in yeast. J. Mol. Biol. 147: 381-390. CULBERTSON,M. R., R. F. GABER and C. C. CUMMINS,1982 Frameshift MCCREADY,S. J., and B. S. Cox, 1973 Antisuppressors in yeast. Mol. suppression in Saccharomyces cereuisiae. V. Isolation and genetic Gen. Genet. 124 305-320. properties of nongroup-specific suppressors. Genetics 102 MCCRFADY,S. J., B. S. Cox and C. S. MC~UGHLIN,1977 The extra- 363-378. chromosomal control of nonsense suppression in yeast: an analy- DAGKESAMANSKAYA,A. R., and M. D. TER-AVANESYAN,1991 Interaction sis of the elimination of [psi'] in the presence of a nuclear gene of the yeast omnipotent suppressors SUPI(SUP45) and SUPZ(SUP35) with non-Mendelianfactors. Genetics 128: PNM-. Mol. Gen. Genet. 150 265-270. 513-520. MESSING,L., 1983 New M13 vectors for cloning. Methods Enzymol. DIDICHENKO,S. A., M. D. TER-AVANESYANand V. N. SMIRNOV,1991 EF- 101: 20-78. 1-like ribosome bound protein of yeast Saccharomyces cereuisiae. ONO,B.-I., J. STEWARTand F. SHERMAN,1979 Yeast UAA suppressors Eur. J. Biochem. 198: 705-711. effective in psi' strains. Serine inserting suppressors.J. Mol. Biol. EUSTICE,D. C., L. P. WAKEM,J. M. WILHELMand F. SHERMAN,1986 Al- 128: 81-100. tered 40s ribosomal subunits inmisreading suppressors of yeast. ONO, B.-I.,N. MORIGA, K. ISHIHARA,J. ISHIGUROand S. SHINODA, J. Mol. Biol. 188: 207-214. 1984 Omnipotent suppressors effective in bsi' strains of Sac- GERLACH,W. L., 1975 Mutational properties of supP amberachre charomycescereuisiae: recessiveness anddominance. Genetics supersuppressors in Saccharomyces cerevisiae. Mol. Gen. Genet. 107: 219-230. 144: 213-215. PIPER,P. W., M. WASSERSTEIN,F. ENGBAEK,K. KALTOFT,J. E. CELIS,J. GOLDBRING,E. S., L. J. GROSSMAN,D. KRUPNICK,D.R. CRYER and J. ZEUTHEN,S. LIEBMANand F. SHERMAN,1976 Nonsense suppres- ~~~ARMuR,1970 The petite mutation in yeast. Loss of mitochon- sors of Saccharomyces ceruisiae can be generated by mutation of drialdeoxyribonucleic acid duringinduction of petites with the tyrosine tRNA anticodon. Nature 262 757-761. ethidium bromide. J. Mol. Biol. 52 323-335. SHERMAN,F., 1982 Suppression in the yeast Saccharomyces cereuisiae, GOODMAN,H. M., M.V. OLSON and B.D. HALL, 1977 Nucleotide pp. 463-486 in The Molecular Biology of the Yeast Saccharomyces: sequence of a mutant eukaryotic gene: the yeast tyrosine-insert- Metabolism and Gene Expression,edited by J. N. STRATHERN,E. W. ing ochre suppressor SUP4-0. Proc. Natl. Acad. Sci. USA 74: JONESand J. R. BROACH.Cold Spring Harbor Laboratory, Cold 5453-5457. Spring Harbor, N.Y. HAWTHORNE,D. C. and U. LEUPOLD,1974 Suppressor mutations in SHERMAN, F., G. R.FINK and J. B. HICKS,1986 Methds in Yeast Genetics. yeast. Curr. Top. Microbiol. Immunol. 64 1-47. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. HINNEBUSCH,A. G. and S. W. LIEBMAN,1991 Protein synthesis and SURGUCHOV,A. P., Yu. V. BERESTETSKAYA,E. S. FOMINYCH, E. M. POSPELOVA, translational control in Saccharomyces cerevisiae, pp. 627-735 in V. N. SMIRNOV,M. D. TER-AVANESYANand S. G. INGE-VECHTOMOV, The Molecular and Cellular Biology of the Yeast Saccharomyces, 1980 Recessive suppression in yeast Saccharomyces cerevisiae is Vol. 1, edited by J. R. BROACH,J. R. PRINCEand E. W. JONES. Cold mediated by a ribosomal mutation. FEBS Lett. 111: 175-178. Spring Harbor Laboratory, Cold Spring Harbor, NY. SURGUCHOV,A. P., V. N. SMIRNOV,M. D. TER-AVANESYANand S. G. INGE- HOFFMAN, C.S., and F. WINSTON,1987 A ten-minute DNApreparation VECHTOMOV,1984 Ribosomal suppression in . Physi- from yeast efficiently releases autonomous plasmids for transfor- cochem. Biol. Rev. 4: 147-205. mation of Escherichia coli. Gene 57: 267-272. TER-AVANESYAN,M. D., V.V. KUSHNIROV,A. R. DAGKFSAMANSKAYA,S. A. HOSHINO,%-I., H. MIYAZAWA,T. ENOMOTO,F. HANAOKA,Y. KIKUCHI, A. DIDICHENKO,Yu. 0. CHERNOFF,S. G. INGE-VECHTOMOVand V. N. KIKUCHI and M. UI, 1989 A human homologueof the yeast GSTI SMIRNOV,1993 Deletion analysis of the SW2(SW3J) gene of yeast gene codes for a GTP-binding protein and is expressed in a Saccharomyces cer&m reveals two non-overlapping functional re- proliferation-dependent manner inmammalian cells. EMBO J. 8: gions in the encoded protein. Mol. Microbiol. 7: 683-692. 3807-3814. WILSON,P. G., andM. R. CULBERTSON,1988 SUFIZsuppressor protein INGE-VECHTOMOV,S. G., and V. M. ANDRIANOVA,1970 Recessive super- of yeast: a fusion protein related to the EF-1 family of elongation suppressors in yeast (in Russian). Genetika 6: 103-115. factors. J. Mol. Biol. 199: 559-573. ITO,H., Y. FUKUDA,K. MURATAand A. KIMURA, 1983 Transformation YOUNG,C. S. H., and B. S. Cox, 1971 Extrachromosomal elements in of intact yeast cells treated with alkali cations. J. Bacteriol. 153 a super-suppressionsystem of yeast. I. A nuclear genecontrolling 163-168. the inheritance of the extrachromosomal elements. Heredity 26: KIKUCHI, Y., H. SHIMATAKEand A. KIKUCHI,1988 A yeast gene required 413-422. for G1 to S transition encodes a protein containing anA-kinase target site and GTPase domain. EMBO J. 7: 1175-1182. Communicating editor: E. W. JONES