YEAST VOL. 12: 17-29 (1996) H A M l , The Gene Controlling 6-N-Hydroxylaminopurine Sensitivity and Mutagenesis in the Yeast Saccharomyces cerevisiae V. N. NOSKOV,'*2,3K. STAAK,' P. V. SHCHERBAKOVA,' S. G. KOZMIN,' K. NEGISHI,3 B.-C. ON0,2 H. HAYATSU~AND Y. I. PAVLOV~* 'Department of Genetics, Sankt-Petersburg University, Sankt-Petersburg, 199034, Russia 2Faculty of Pharmaceutical Sciences, Okayama University, Tsushima, Okayama 700, Japan 3Gene Research Center, Okayama University, Tsushima, Okayama 700, Japan Received 20 January 1995; accepted 20 July 1995 The haml mutant of yeast Saccharomyces cerevisiae is sensitive to the mutagenic and lethal effects of the base analog, 6-N-hydroxylaminopurine (HAP). We have isolated a clone from a centromere-plasmid-based genomic library complementing HAP sensitivity of the haml strain. After subcloning, a 3.4 kb functional fragment was sequenced. It contained three open reading frames (ORFs) corresponding to proteins 353, 197 and 184 amino acids long. LEU2' disruptions of the promoter and N-terminal part of the gene coding 197 amino acids long protein led to moderate and strong sensitivity to HAP, respectively, and were allelic to the original haml-1 mutation. Thus this ORF represents the HAMI gene. The deduced amino acid sequence of HAM1 protein was not similar to any protein sequence of the SwissProt database. The H A M l gene was localized on the right arm of chromosome X between cdc8 and cdcll. Spontaneous mutagenesis was not affected by the haml::LEU2 disruption mutation. KEY WORDS - Saccharomyces cerevisiae; 6-N-hydroxylaminopurine; hypermutability; molecular cloning and sequencing; purine metabolism INTRODUCTION Analogs of natural precursors of nucleic acids are valuable tools for studies relevant to DNA structure and DNA synthesis (Kornberg and Baker, 1992). Some of them can provoke replication mistakes due to ambivalent pairing capacity and are strong mutagens (Freese, 1959). Recently it was shown that some base analogs, such as 8-oxoguanine, may be formed in vivo, thus being a normal cellular component affecting replication fidelity (Michaels and Miller, 1992). A specialized multicomponent enzyme system has evolved to protect organisms from such dangerous intrinsic mutagens. In the bacterium Escherichia coli, this system consists of the products of at least three genes-mut Y, mutM and mutT. Mutations in these genes lead to a strong mutator phenotype (see Michaels and Miller, 1992). A mammalian *Corresponding author: Department of Genetics, SanktPetersburg University, Universitetskaya emb., 7/9, SanktPetersburg, 199034, Russia. CCC 0749-503X/96/010017-13 0 1996 by John Wiley & Sons Ltd counterpart of mutT has been already described (Mo et al., 1992). 6-N-hydroxylaminopurine (HAP) is one of the potent base analog mutagens effective both for eukaryotes and prokaryotes (see Pavlov et al., 1991 and references therein). This base analog might like 8-oxoguanine be responsible for spontaneous mutagenesis because, in vitro, it can be formed from adenine by P-450 monooxygenases (Clement and Kunze, 1990). In an attempt to reveal the genes controlling HAP mutagenesis in yeast, we isolated the haml-1 mutant which was hypersensitive to HAP-induced growth inhibition and mutagenesis (Pavlov, 1986). The observed sensitivity was HAP-specific (no sensitivity to UV or EMS and even the related base analog, 2-amino-6-N-hydroxylaminopurine). It was shown that the haml-1 mutation did not influence spontaneous mutation and recombination rates but strongly reduced the capability of adenine-requiring ade2 haploid strains to grow on minimal medium with HAP as the sole adenine 18 V. N. NOSKOV ET AL. Table 1. Saccharomyces cerevisiae strains used in this study. Source or reference Genotype Strain Al-D795 YAH559 4A-D776 4078-8-2a MATa ade2-1 CANI MATa ade2-I CAN1 haml-1 MATa leu2-3,112 gal7 tyrl ura4 h a d - I MATa ura3 ade2-1 h a d - 1 MATa leu2-3,112 lysX his7-1 ade2 trpl-289 ura3 haml-1 MATa leu2-3,IIZ his3-11,IS ura3-52 trpl-289 MA Ta his7-2 ura3-52 leu2-3,112 ade5-1 trpl-289 CANI Same, but ura3-A lys2-Tn5-13 MATaIMATa his7-21+ ura3-52Iura3-52 leu2-3,11211eu2-3,112 ade5-llade.5-1 trpl-289Itrpl-28Y +lhis3-11,14 H A Mllhaml:: LEU2 CANI ICANI Same, but haml::LEU2/ham I:: LEU2 MATa cdc6 leu2-1 arg3 his3 ura2 MATa ilv3 ade5 trpl his7 met5 MATa ura3-52 his7 leu2-3.112 cdc8-1 IS488-9B M A Ta arg4 aro7-1 his6 ilv3 leu2-1 lysll trp3 ura2 cdcl I 197l2d hml-197l2d 139-D326 4-D365 57-D550 DBU746 CG379 CG379A Al-D794 Cox and Parry (1968) Pavlov (1986) Noskov (1988) This work' This work' Dr D. Botstein Morrison et al. (1991) Shcherbakova et al. (1995) This work' This work' Okayama Univ. collection This work4 Dr Barbara Garvick, University of Washington Okayama Univ. collection 'Segregate of diploid D365 from cross hml-l97/2d x XS122-57D (MATa ura3 rad52, Dr R. Mortimer). 'Segregate of the diploid D550 from a cross of 35G-D508 (MATa ura3 lys2-24 haml-1) and 33G-D373 (strain with high competence to lithium transformation, Kushnirov et al., 1990). 3Diploids D794 and D795 were obtained from cross of tl-CG379 (ham1::LEUZ) to DBU746 or t2-DBU746 (haml::LEU2), respectively. Mitotic adenine-requiring segregants (add-llade.5-I) of these diploids were obtained after UV-light irradiation (35 J / M ~ ) . 4Meiotic segregate of XS195-23B (Dr R.K. Mortimer) and hum1 disruption transformant of CG379. Has been used for meiotic mapping (chromosome X right arm markers ilv3 and m e d ) . source. On the basis of these observations, it was proposed that this mutation interferes with HAP metabolism in the yeast cell (Pavlov, 1986). In this study we have cloned, mapped and sequenced the HAMl gene. This enabled us to obtain haml disruption mutants in various strains. We have shown that haml/haml diploid strains can grow on HAP as the adenine source, thus this mutation does not interfere with HAP conversion to adenine precursors. We have shown that spontaneous mutation frequency is unaffected by the haml disruption mutation, thus endogenous HAP does not contribute to spontaneous mutation under normal growth conditions. The possible role of the HAMl gene product in HAP mutagenesis in yeast is discussed. haml-1. The pol2-1 mutants of the strain CG379A and its haml::LEU2 derivative were constructed by gene disruption as described (Shcherbakova et al., 1995). Escherichia coli strains HB101, DH5a, MW1184 and JM103 (Sambrook et al., 1989) were used for plasmid propagation. MATERIALS AND METHODS Yeast genomic DNA library and basic plasmids Yeast genomic DNA library constructed on a centromere plasmid p366 was kindly provided by Dr P. Hieter. p366 is a derivative of YCp.50 with the LEU2 selective marker and is 8.2 kb long (P. Hieter, personal communication). pFL44 is a multicopy yeast-E. coli shuttle vector (Bonneaud et al., 1991). Single-stranded mp19/mp18 vectors or phagemids pUCl18/119(KS+/KS - ) were used for subcloning for DNA sequence determination (see Sambrook et al., 1989). Yeast and bacterial strains Basic yeast strains used in this study and their genotypes are listed in Table 1. We named the previously isolated haml allele (Pavlov, 1986) as Media and supplies Standard yeast YPD and minimal (SD) media were used (see Sherman et al., 1986). SD medium with appropriate supplements and L-canavanine 19 HAMl GENE (40 mg/l) was used for selection of Can‘ mutants. The medium for selection of Adp’ mutants, capable of growing on a-aminoadipic acid as the sole nitrogen source, was as described by Chatoo et al. (1979). Escherichia coli minimal medium, LB, and 2 x YT media were prepared as described (Miller, 1972). 6-N-hydroxylaminopurine was synthesized from 6-chloropurine according to the method of Giner-Sorolla and Bendich (1958). 2-amino-6-Nhydroxylaminopurine was a gift of Dr F. J. deSerres (Research Triangle Institute, USA). Both were dissolved in dimethylsulfoxide (DMSO). Restriction endonucleases and other DNA metabolism enzymes were either from ‘Fermentas’ (Lithuania) or ‘Takara’ (Japan). DNA size markers (1-kb ladder) were from BRL (USA). Genetic methods Yeast transformation was performed according to Ito et al. (1983). Other standard methods were as described by Sherman et al. (1986). All incubations were at 30°C. HAP-mutability of individual clones was analysed as described (Pavlov, 1986). Briefly, strains to be tested were streaked on a YEPD master plate (usually no more than ten strains per plate). For strains containing unstable plasmids, SD medium was used to select for the plasmid marker. After 2-3 days of incubation, cells were replica-plated onto three plates with the same medium. Then, a long strip of filter paper soaked with DMSO or HAP solution (20pg and 200pg per strip) was placed perpendicular to the streaks on the corresponding plate. The plates were incubated for 2 days. HAP mutagenesis occurred during this incubation. Then cells were replica-plated onto medium selective for Canr or Adp’ mutants. The density of resistant colonies appearing in the vicinity of the HAP-treated area was an indicator of mutagenesis. Wild-type strains are immutable at the low HAP dose and moderately mutable at the high dose (phenotype Hamf). The haml mutants are mutable at low doses and hypermutable at high doses. For the growth test on SD medium with HAP as a sole adenine source, a diluted yeast suspension (an equal cell concentration for all strains) was spread onto adenineless SD medium. Then a paper strip soaked with adenine or HAP solution was applied. One version of this test is to plate cells of several strains on one plate in a radial fashion and apply a paper square in the center of the plate (Figure 1). The spontaneous frequency of Canr mutants was measured as follows. Ten independent colonies of each strain to be tested were inoculated into 1 ml of liquid YEPD medium and grown to stationary phase. 10 ml of each resulting culture were inoculated in 10 ml of liquid YEPD medium and incubated with good aeration for 15-17 h. Then cells were plated onto canavanine-containing medium and YEPD medium (after 105-fold dilution). Mutant frequency was calculated for each of ten cultures according to the formula: F= M1N.k where M = a number of mutants, N=number of colonies on YPD, k=dilution factor. Frequencies for ten cultures of each strain were compared using Krusker-Wallis non-parametric criteria. Other methods The molecular genetic methods such as subcloning, restriction analysis, blot hybridization and DNA preparation for sequencing were performed according to Sambrook et al. (1989). A deletion set for DNA sequence determination was constructed using the kit from ‘Takara’ (Japan) according to the published protocol. DNA sequence analysis was performed using an ABI Model 373A automatic sequencer according to the protocol of the manufacturer. Initial mapping of the cloned gene by hybridization with pulse-electrophoresisseparated yeast chromosomes was performed on the blot commercially available from Clonetech Labs (USA). RESULTS Cloning of the HAMl gene by functional complementation of haml-1 mutation Saccharomyces cerevisiae strain 139-D326 (hamf- I , phenotype Ham -) was transformed with the yeast genomic library to Leu’. Transformants were individually tested for their Ham phenotype. From 3185 transformants tested, 137 were HAPimmutable at low dose (presumably Ham+). Seventy-seven of them were unable to grow on non-fermentable carbon source, i.e. were respiratory-deficient. Since low HAP-mutability of these transformants is at least partially attributable to their very slow growth, they were discarded. The remaining 60 transformants were tested for Ham phenotype after plasmid loss. It was found that 58 20 V. N. NOSKOV ET AL. A. CG379A ham I:: LE U2 pol2-1 haml::LEU2 pol2-I B. CG379A ham 1::LE U2 ~ 0 1 2 - Ihaml::LEU2 po12-I Figure 1. Growth on minimal medium with adenine or HAP as the sole adenine source of the strain CG379-A and its disruption mutants, haml::LEUZ, po12-I and the double mutant, haml::LEUpo12-1.30 p1 of cell suspension of each strain was applied as a small spots forming a rectangle. (A) Paper strip soaked with adenine solution (2 mg total); (B) paper strip soaked with HAP solution (2 mg total). clones were still immutable, so their low mutability was associated with changes in chromosomal genes or in strain ploidy. For example, they might be diploids, which are at least one order of magnitude less mutable after HAP exposure than haploids (Pavlov et al., 1991). Thus, we obtained two clones possessing the plasmid-associated Hamf phenotype. From the two transformants, plasmid DNA was rescued after E. coli retransformation. Restriction analysis revealed that the plasmids were p366 derivatives with different inserts. Strain 139-D326 containing one of them was H a m- on a-aminoadipic acid medium but was completely HAP-immutable on canavanine medium. This plasmid likely contained a cloned CAN1 gene as it conferred canavanine HAMI GENE x H x RIS H G2 G2 p366HAM1 AB-B = = 21 HAP hypermutability - AB-G2 AH-H AS-S + + 1 kbp Figure 2. Restriction map of original insert containing the H A M l gene and its deletion derivatives. Insert of the yeast DNA in the plasmid p366HAM1 is represented by a solid bar. Dashed bar is the vector sequence. Restriction sites are above the insert map. Note that the left Hind111 site is in the vector sequence. The boundaries of deletion variants are below the insert map. Phenotype of the strain 57-D550 transformed with plasmid variants is shown in the right part of the Figure. H, HindIII; S, Sufi; X, XhoI; B, BamHI; RI, EcoRI; G2, BglII. sensitivity to the yeast strain DC5 which carries the can1 mutation. Transformants with the second plasmid had typical Ham+ phenotype in both forward mutation systems. Thus, only one plasmid carried the gene complementing haml-1. The plasmid was named p366HAM1. The restriction map of insert in p366HAM1 plasmid is shown in Figure 2. Several deletion variants of this plasmid were constructed: deletion of the central BamHI fragment (AB-B), deletion of the left part of the insert flanked with HindIII sites (AH-H), deletion of the BamHI-BglII region (AB-G2), deletion of the SalI region (AS-S). Yeast strain 57-D550 (haml-I) was transformed with these plasmid variants and also the original plasmid p366HAM 1 and transformants were examined for the level of HAP mutability. The gene conferring Ham+ phenotype was located at the left part of the cloned region as only variants AB-B and AB-G2 complemented the haml-1 mutation. We cloned the 3.4 kb HindIII-BamHI DNA fragment into the autonomously replicating plasmid pFL44 and demonstrated that it was capable of restoring the Ham- phenotype in haml-1 strains, 57-D550 and 4-D365. To verify that we have cloned the gene itself and not its suppressor, we constructed an integrative plasmid p366HI with selectable URA3 marker. For this purpose, the parent plasmid p366HAM1 was cleaved with BglII, and the resulting HAMI-pBR322 fragment was ligated with the BgnI fragment of the URA3 gene isolated from pFL44. The resulting plasmid was linearized with BamHI (see Figure 2 for the map of the insert) and used for transformation of the yeast strain 4-D365 to Ura+. All ten transformants tested were Ham+. The Ura+ phenotype was stable, indicating that the plasmid was really integrated. We crossed one of these transformants with the strain DBU746 (HAM1 ura3-52, see Table 1). Resulting diploids were sporulated and subjected to tetrad analysis. Of 17 tetrads analysed, 16 showed 2:2 Ura+:Uraand 4:O Ham+:Ham- segregation, which is expected if integration occurred into the HAMl locus. One tetrad showed 3:l segregation for both markers and probably resulted from plasmid excision. This analysis confirmed that we really had cloned the gene of interest. Open reading frames (ORFs) found in the 3.4 kb DNA fragment carrying the HAMl gene We have determined the nucleotide sequence of the 3.4 kb HindIII-BamHI DNA fragment after cloning its sub-fragments into phagemid vectors and creation of a nested deletion set. Three ORFs V. N. NOSKOV ET AL. 22 1800 60 C W L T C C R C A C G T T G C T G T G C T G T G T G A A C T G G C C CTATTAGRATGTCRTGAGGTACCACACCAGCTAATTCTTCCACTT~~AGAAGTTATGT ^Ball "BamHI 120 CCATGGAGGTGCCAAGGATATTGCAACAATCCCTATACTCGAGlATTGACCC ^Xho i 180 ~CCCACCTCTACCATTAGAAAAGGATGCTACCATCCCAGAACTACAGGCCTTATTGRA 1920 TTCTC~GTCACCTGCCGAARTGTCCAAGATCCT~TCC~CTCCATCATCACATTTRA 240 TGATCCTAAGCMCCTTTGTTCCAIIAGATAAACC~~~CCATGTTCAGACTGA~~GATATCffi 1980 300 'LACPGRTGAAGCAA? CCT(jGCCTTUjCCA~TGLTTTCAGTtiC~TCT'rCCCTTT1 CAA CATlTTCTTGlTCGCTGATAAACCTCAAA~TCRATC~ATlACTTGCGTCT~CTT 'Ball 360 2040 TGRATCTAAATTTGCGCT~C~TCRATRATlCTTGTCACATAATTAC GCATGAAATCGiCTACGTCTTCU;TC~TAGGTAGTCCGGCTGCTGTC~C-TTTGAT 420 "NheI 2100 RAATCRAACAGRATCTTGTTACTCCTGRATAGGTTTCCATTGTTCTACGT~SGCACTCT TGrlP-GTTTTGGGCAGAAAGGARGAAGCTCCAATG5TTAGGCATW.AGCTGCTGAAGCCTT 480 GGGTGCCATTGCTTCICCAGAAGlTGTCGACGTCTlGAAATCTTACCTCAACGATGAAGT "SalG1 2160 540 GTGCATCGGCAGTCATTGAATCGGCCTCATC~GRAlAATGATCTTATACGGTGGGCACG (CGTTG~~A~it~~~~~GA~CTG'(Tjr'lAjlC 600 CGAACTAGAAT~)TGCTCCAACTGCTAATlA~;RA~~TCAG=AT~TCAATAATGATAGTA 2220 GATAGTTCTCCAhGTCATGllT~CTACCCTTC~ATACACTCAATCTTC~TTTTTTA ""Y ACAh~AATCATTGTATAATCTTGlATTTTATACTTTTCETTCCGTTATA 720 2280 CCTTTTCTCTTACRATAGAGATA~CA~~TTCG~CAGL%CGTTCRRCrCCRATA~~CTTG TATACNMTCTATCAAATACCAAAARAATTATTTATTTTCGNMCCGTTCTCTCGAGCTT . "Xhu 1 "SSP1 780 2340 ATTTCATCARGTCGGGTCCATATRATTCTTTTCTTIICCGCT~TGGTAGA~GTTTTAC CAGCTTTCRACACTGGTCTTCAIIAGCAlAGGRAACACTCACATAGAGATATClAAWdCG 840 M S N N E I V F V T G N A N K RACGATAGCAGAATGliCCA-~T~;TATTTGTCACTGGCRATGCCRAC-TTA L 1 6 2400 CAGTTCCTWjGWjACCATRCAACATAlGTGGTAGATTAGCTGACTllAAGGTTTTCT ^NdeI 800 K E V Q S I L T Q E V D N N N K T I H L 3 6 AAGGRAGTGCAGTCGATTTTGACTCRAGAWTCGACAACRACAhCAAMCCATCCACTTA "SSJGI ~~~ ~ 960 I N E A L O L E E L Q D T D L N A I A L 5 6 ATCAhTGAGGCCTTAGACCTGGAAG2*RTTGCAGGACACGGATCTGRATGCCATlGCGllA 2460 TCIllVUlCAGTAACGGCATGATCTTGAGCTSlCACTTCATCTAGGTTTTTGGGCCTGTATT 2520 TCTCRACCCRGGGTTGTTGTGCCAATGATTGCTCTGCGGCGCT A K G K Q A V A A L G K G K P V F V E D 7 6 GCTAAUiGARAGCAAGCTGlTGCGXCTTGGGTRAGGGTAAGCCAGTGTTTGTGGRAGAC 1080 T A L R F D E F N G L P G A Y I K W F L 9 ACTGCGTTGAGATTTGACGAATTTAACGGTCTACCAGGTGCTTATATAAMTGGTTTTTA 2580 TTTTATTTGGACLRARCCCTTCAAACATCTCRAGTGATTGCCAATACTTTCCCC~TAT 6 2640 I140 S M G L E K I V K M L E P F E N K N A 1 1 6 AAGAGCATGGGAlTGTGTGAPAATGTTGGRACCCTTTGAAMTAAWdCGCT TTCTCGCTGTCTTTCTATTTTACTTCTCTTATAGCATTGACTTGTTCGCTTGAGTTACTT K 1200 G I36 E A V T T I C F A D S R G E Y H F F Q GAAGCAGTTACLACCATTTGTT 1TGCTtiAT1'CSC(*%j(iTljAA',ATCA~,'T~STTC~T 1260 I T R ti K 1 V P S R G P T T F O W D S I 1 5 6 ATTACRAGAGTCRAGATCGTTCCL%CCGGGGACCCACTACATTCKXTGGGATTC?ATT 2700 CCTGATTGCTTTGRAATACGCGAAGC~lGAAAAATAACGATGAGAAAGSTTTTT 2760 GATAGAAGA~~AACAAAAAXXAACAGTGCTACRATAGTATATA@.GXTGTCG?ATACTTG 2820 GGRCGATGTATG(~GCATCAGACAGTGATGTCGAAACTGAACGATCTCCAGATTTAGT~ 2880 1320 B P F D S H G L 1 Y A E M S K D A K N176 TTTCAACCTTTTGAT~=TCAT~A=TGACTl'ACCCAGACATG~C~ACG~GAAT GTTAAGA~A-TCAlAG~GAMjGGGTACTTUjACGGAATAGTGAGllCT~G~A B 2940 RAAGCTGCAAGRA~TTTCRATUjTGG1TTCCCRAC~GCTRAATTA~AAAC-T 1380 A I S X K ti K A F A Q F K 6 Y L Y 0 N 0 1 3 6 G C T A T T T C T C A T C ~ T G G T - G C ~ T ~ C G C T C R S T T T A A R 3000 TGGCATAATTATGGGGATTTTACTT~~~CCTACGAACACGTTTTGGTGATGRAGATGA~ 1440 F 197 TTTTAGGTGATGTGCTGACRAGATAAGA~TTATffiATAGATTTTlTTTTTTTATTCT CCTCAGlRAMjCATATATTTGATGCTCAAAAAGAGTTACGCATTRACRAAGTACTlAGTAA 3060 3120 1500 TACTTTATGTCGCCTCTA~ATT~AATTATArNMGATGCCTGTAlATTAGAGTT~ATA GTCCATATTTGATCCACAGGACGGGTGTGGTCGCCATGATCGCGTAGTCGATAGTGSCTC 3180 CAAGTAGCGRAGCGAGCAGSACTGGGCGGCGGCCAAAGCGSTCGGACAGTGCTCCGAW 1560 TTTTAAIIICCAATARGTTTAATAATTGGA~~~~CTCGTTCGTTCCGTTGTT~AATCTGGAAT ~~- . ~~~~~~~~~ ~ . ~ ~~ 1620 CCGTACTAIIACAGCAhCCRAGATATTTGATTTTTG~TTCGTATC~TTATCGTTGG 3240 CGGGTGCGCATRGAMTTGCRTCAACGCATATAGCGCTAGCAGCACGCCATAGTGACTGG "Nhel 3300 CGATGCTGTCGGAATGGZCGATATCCCGCAAGAGGCCCGGCAGTACCGSCATRACCRAGC 3360 CTATGCCTACAGCATCCAGGGTGACGGTGCCGAffiATGACGATGAG~GCATTGTTAGAT~ "SSPPl ^HpaI 1740 ALCGTA~TGACCTATTTCTT~T~T=A~~AAMT~CCAC~TTTGACTTTTT~CAhTTT 3421 TCATACACffiTGCCTGACTGCGTTAGCRATTTAkCTGTGATAIIACTACCGCATT~T "Hind111 Figure 3 . Nucleotide sequence of the 3.4 kb DNA fragment containing the HAMl gene. Only a few restriction sites, used for plasmid construction, are marked. Amino acid sequence deduced from the HAMl reading frame is shown above the nucleotide sequence. ORFl corresponding to the RFC2 gene is underlined. 23 HAMl GENE HAP hypermutability X Ba H N Nd SP N Hp 'p S X S B a t i , B pFL44HAMI - AN-N - ORF2 ORFl r - ',,ORF3 ABa-Ba AX-X AS-S +/- + 1kbp ' Figure 4. Deletions in 3.4 kb DNA fragment containing the H A M l gene: map position and phenotypes of appropriate mutants. Yeast strain 57-D550 was transformed with plasmid variants to Ura' and their Ham phenotypes analysed as described in Materials and Methods. See text for other explanations. Nd, NdeI; Ba, Ban; X, XhoI; SaA; N, NheI; Ss, SspI. (ORF1, ORF2 and ORF3) have been found, which might encode proteins 353, 197 and 184 amino acids long, respectively (Figures 3 and 4). ORFl appeared to be a gene with a deduced amino acid sequence similar to several subunits of human replication factor C, bacterial z subunit of replicative DNA polymerase I11 and gene 44 product of bacteriophage T4, and was called RFC2 gene (Noskov et al., 1994). Amino acid sequences deduced from ORF2 and ORF3 had no similarity to any of the protein sequences of the SwissProt database. ORF3 had an ATG codon but not a promoter sequence, thus it is likely a truncated gene. Efects of the deletionlinsertion mutations in ORF2 Several deletion mutations were constructed using the 3.4 kb DNA fragment cloned in pFL44 (Figure 4). Strain 57-D550 (ade2 ura3 haml-I) was transformed with the pFL44-HAM1 deletion derivatives, and phenotypes of transformants were scored. Deletions AX-X and AS-S failed to complement the ham1 -1 mutation, suggesting that ORF2 is the H A M l gene; note that the deletion AX-X which leads to the intermediate Ham+'phenotype (Figure 4) does not cover ORF2 itself but the majority of the promoter region of the gene, while deletion AS-S removes part of the coding region. For further characterization of these mutant alleles, we have constructed disruption mutations analogous to AX-X and AS-S. Fragments flanked by XhoI and SalI sites, respectively, were replaced by the LEU2 gene. Plasmid LHAM was constructed on the basis of p366HAM1 (see section on HAMl cloning) by replacing the small XhoI-XhoI fragment with the 2.2 kb LEU2 gene SalI fragment. For construction of HLAM plasmid we have used pUCll9 carrying the 3.4 kb H A M l fragment. We cut it with SalI and ligated the LEU2 gene Sun fragment to a larger fragment. Linear DNA fragments with disruption were released from the plasmids HLAM and LHAM by cleavage with Hind111 and BamHI. The yeast strain CG379 was transformed with these DNA fragments to Leu+. DNA isolated from five transformants of each variant was analysed by Southern blot hybridization to verify an expected disruption (results not shown). Then the transformants were tested for HAP mutability and growth on HAP as the sole adenine source (note that adenine biosynthesis is blocked by the ade5 mutation in strain CG379). All transformants with AX-X::LEU2 were Hamf'-, and all transformants with AS-S::LEU2 were Ham- by these two criteria (see Figure 1 for example of typical result of the growth test). Complementation testing showed that insertion mutations were allelic to 24 V . N. NOSKOV ET AL. Table 2. Spontaneous Can‘ mutant frequencies in the CG379 strain and its haml::LEU2 derivatives. Culture number 1 2 3 4 5 6 I 8 9 10 Canavanine-resistant mutants frequencies ( x 10 ‘) in strains: CG379 LHAM-CG379 HLAM-CG379 0.8 1 0.80 0.75 2.30 6.10 0.85 0.74 1.70 0.80 3.15 2.87 0.96 0.29 0.29 0.29 0.40 0.40 7.96 0.38 1.03 0.26 0.40 6.95 0.51 0.30 1.23 1.oo 0.74 0.36 0.30 haml-1 (cross to 4-D365). The data prove that ORF2 represents the H A M l gene. Features of HAM 1 gene and its predicted product The coding region of the H A M l gene is 591 nucleotides long. Its 5’ and 3’ flanking regions have sequences common to many yeast genes (see Breathnach and Chambon, 1981): the promoter region of the gene contains at least three putative TATA boxes (starting from positions - 178, - 163 and - 135, Figure 3); positions - 3 and - 1 are occupied by ‘A’. The promoter lacks ACGCGT (MZu1)-like sequences which are necessary for coordinated expression during S-phase. Proposed termination and polyadenylation consensus T A G . . . TATGT . . . TTT is found at base 37 after the termination codon. The H A M l gene ORF encodes a 197 amino acids long protein, the predict isoelectric point of which is 5.27. Searches for amino acid sequence similarity with known sequences using FASTA and DNASUN programs showed only weak homologies with short stretches of unrelated genes of different origin. Additional H A M phenotypes Availability of the H A M l gene disruption mutants enabled us to study the Ham- phenotype in comparison with isogenic wild-type strains. First, we addressed the question of whether the inability of ade5 adenine-requiring strains with hum1 mutations to grow on HAP as the sole adenine source is due to accumulation of recessive lethal mutations or due to an intrinsic defect of HAP salvage as an adenine source. In the latter case, adenine-requiring strains, blocked in de novo adenine synthesis reactions prior to IMP formation, and carrying huml mutation, would be unable to grow on HAP as the sole adenine source. We constructed two sets of strains: diploids Al-D794 and Al-D795 (Table 1) which should tolerate recessive lethal mutations (see Pavlov et al., 1991) and derivatives of strain CG379A with haml::LEU2, pol2-1 mutations (the latter is the HAP-antimutator mutation; Shcherbakova et al., 1995) and a double huml::LEU2 po12-1 mutation. Both haml::LEU2lhaml::LEU2 diploid (results now shown) and haml::LEU2 po12-1 double mutant haploids were able to grow on HAP as the sole adenine source, albeit at a reduced rate as compared to wild-type strains (Figure 1A and B). Thus, inability of the ham1 adenine-requiring haploid strains to grow on HAP as the sole adenine source is not a result of the defect of HAP utilization as adenine. We propose that HAP is much more mutagenic on SD medium than on YPD due to the absence of competing adenine in the former, and hypermutable haml strains stop growth due to accumulation of lethal mutations. Spontaneous mutation frequency in isogenic wild-type and haml strains To address the question of whether possible endogenous HAP contributes to spontaneous mutation, we have analysed spontaneous Can‘ mutant frequencies in the strain CG379 and its huml derivatives obtained by one-step gene disruption (see previous section). The results of one of the two 25 HAM1 GENE experiments performed are presented in Table 2. Statistical analysis indicated that differences between all variants are not significant (p> 0.05). Thus, haml strains (which are at least two orders of magnitude more sensitive to HAP-induced mutagenesis) are not distinguishable from the isogenic wild-type by spontaneous mutability. The result suggests that either yeast cytochrome P450 is unable to convert adenine to HAP or its expression under conditions used is insufficient to produce HAP. It was noted that HAP production from adenine by mammalian microsomes is inefficient, which questions the biological relevance of this reaction (Clement and Kunze, 1991). Chromosome mapping of the HAMl gene The 3.4 kb DNA fragment containing the H A M l gene was oligolabeled to high specific activity and hybridized to blotted separated yeast chromosomes. A strong positive signal was detected with chromosome X (result not shown). To map the H A M l gene in the chromosome, strain hml- 197/2d was crossed to several strains carrying chromosome X markers and standard genetic mapping was performed; haml segregation was followed by its HAP-mutability phenotype. It was shown that haml is linked to ilv3 and met5 markers of the right chromosomal arm, but not to arg3 and cdc6 on the left arm. For precise mapping, a strain carrying ham1::LEUZ was crossed to cdc8 and cdcl I marker strains and segregation of the ham1 mutation was monitored by segregation of Leuf phenotype. Thus haml was located between cdc8 and cdcll (Figure 5A and B). DISCUSSION We cloned and sequenced a gene controlling HAP mutagenesis in the yeast S. cerevisiae. The analysis of predicted amino-acid sequence coded by this gene gave no direct clues as to its function. Availability of the cloned gene enabled us to study the effects of mutations in this gene in a strictly isogenic background. Examination of the phenotypes of disruption mutants permits consideration of several hypothesis on the nature of the product of this gene. It is important to consider pathways of dATP synthesis to understand the possible mechanism of HAP mutagenesis (Figure 6). Both de novo synthesis and salvage of adenine have the same intermediate-IMP. It is known that yeast strains A. PD NPD haml-met5 haml-ilv3 haml-cdc8 haml-cdcll 43 28 67 67 T 4 56 3 69 23 9 0 0 Map distance (cM)* 38.8 43.3 12.8 5.9 chromosome X~ight Figure 5 Mapping of the HAM1 gene (A) Tetrad analysis data *Map distance was calculated according to the formula Xp=(100/2)x [(T+6NPD)/CPD+NPD+T)] (Barratt et u l , 1954) (B) Position of hum1 on genetic map of chromosome X with defects in purine biosynthesis de novo, prior to the IMP+AMP conversion step (most adeninerequiring mutants, Figure 6), can utilize various 6-substituted purine analogs as the adenine source, while mutants with a block of the IMP+AMP step cannot (Abbondandolo et al., 1971). This observation is explained by the broad specificity of aminohydrolase, an enzyme converting adenine and its 6-substituted analogs into hypoxanthine. Hypoxanthine is converted further to AMP, unless synthesis and conversion of adenylosuccinate is blocked (mutants adel2 and adel3, Figure 6). This seems to be the case with HAP, too. Both ade2 and ade5 mutants are able to grow on HAP as the sole adenine source (Pavlov, 1986; this study, Figure 1) but, according to preliminary unpublished data of S. G. Kozmin, A. M. Zekhnov and D. V. Domkin, yeast adel2 mutants are not able to grow on HAP. Thus, HAP cannot substitute adenine per se, via phosphoribosyl transferase reaction. It should be converted to hypoxanthine. The main feature of the haml-1 point mutation and the haml::LEU2 null allele is recessive HAP-hypermutability (Pavlov, 1986; this study). We have shown, in the present work, that inability of ade2 ham1 or ade5 ham1 mutants to 26 V. N. NOSKOV ET AL. exogeneous adenine Adenine biosynthesis de novo 4 1 1 4 i ude4 ade5 ude8 ade6 ade7 ade2 adel adel3 minohydrolase ,/ hypoxanthine I HGPRT / .( AF’RT inosinate (IMP) ,,/ 4 inosinate (IMP) adel2 adenylosuccinate synthase udel2 adenylosuccinate synthase .c 1 adenylate lunase nucleoside diphosphate lanase grow efficiently on SD minimal medium containing HAP as the sole adenine source (Pavlov, 1986; present study) is a manifestation of their hypermutability, and not their HAP salvage defect. Strains bearing a hamZ mutation but resistant to HAP-induced mutagenesis (diploid and antimutator) were apparently proficient in HAP reutilization. To find a possible role for the HAMZ gene product, let us consider a hypothetical pathway leading to mutation induction by HAP (Figure 7). According to available data, HAP produces 27 HAM/ GENE HAP extracellular hypoxanthine IMP - AMP I aminohydrolase kinase i HAPDP I ribonucleotide diphosphate reductase I nucleotide diphosphate b a s e I base selection, proof-reading by DNA polymerases dHAPDP dHAPMP triphosphatase 4 dHAPTP HAP incorporated into DNA i base selection, proof-reading by DNA polymerases HAP-containing templates I repair mutation Figure 7. Hypothetical pathway of HAP mutagenesis. Explanations are in the text mutations during DNA replication, thus it should be converted into dHAPTP in the yeast cell (see discussion in Shcherbakova et al., 1995). The product of the HAMI gene should act during some stage of the HAP mutagenicity pathway (Figure 7) to prevent HAP incorporation into DNA and its further persistence in DNA templates. It is evident that haml cannot be a transport defect or a defect in phosphoribosyl transferases. It is well known that such defects lead to resistance to base analogs (Woods et al., 1983; Sahota et al., 1987) unless enzyme activity is increased (Lomax and Woods, 1969). The latter event should lead to a dominant effect and it is hard to expect this from the null allele. We may assume that, in wild-type strains, HAP is quickly metabolized to adenine by a novel reaction catalysed by xanthine oxidase (Clement and Kunze, 1992), or to hypoxanthine or IMP by corresponding aminohydrolases (Figure 7). If this pathway is blocked in haml mutants, more HAP would be available for further conversion to the deoxytriphosphate form. Both possibilities are unlikely. The first pathway does not contribute to HAP destruction significantly, as HAP utilization is mediated by the hypoxanthine pathway (see above, results with adel2 mutants). Moreover, Ham1 protein has no homologies to rat xanthine oxidase, the protein sequence of which is known. The second pathway is evidently not impaired in 28 ham1 mutants as they can utilize HAP as the adenine source. AMP deaminase, which has been cloned and sequenced in yeast (Meyer et al., 1989), has no similarity to Hamlp. We can propose that the HAM1 gene product is involved in some unidentified set of reactions protecting cells from HAP, either on the level of deoxynucleoside triphosphate or the DNA level (Figure 7). Such a system is known for the endogenous analog, 8-oxoguanine (see Michaels and Miller, 1992). This proposal raises the question of why such a system might be necessary, as we have shown that endogenous HAP does not contribute to the spontaneous mutation rate in yeast (at least under standard growth conditions in rich media) and thus is not a common cellular component. Further biochemical characterization of HAMI gene product is in progress to test the possibilities discussed. It may be of interest to mention that strains of Salmonella typhimurium and E. coli bearing a deletion encompassing the uwB-bio region share some common properties (HAP hypermutability and hypersensitivity) with ham1 mutant (Pavlov et al., 1991; Pavlov et al., in preparation). This defect is due to deletion of some unidentified gene. Thus genes analogous to HAMI might be widespread. ACKNOWLEDGEMENTS We are grateful to Dr M. G. Samsonova for help with computer searches of homology and Dr V. D. Domkin for interest in this work. We are grateful to referees for helpful comments. This work was supported in part with All-Russia Program ‘Frontiers in Genetics’, Russian Fund for Fundamental Research and Japanese Society of Promotion of Science. Part of this work was made possible in part by grant #JDH100 from the International Science Foundation and Russian Government to Y.I.P. REFERENCES Abbondandolo, A,, Weyer, A,, Heslot, H. and Lambert, M. (1971). Study of adenine aminohydrolase in the yeast, Schizosaccharomyces pombe. J. Bacteriol. 108, 959-963. Barratt, R. W., Newmeyer, D., Perkins, D. D. and Garnjobst, L. (1954). Map construction in Neurospora crassa. Adv. Genet. 6, 1-93. Bonneaud, N., Ozier-Kalogeropoulos, O., Li, G., Labouesse, M., Minvielle-Sebastia, L. and Lacroute, V. N. NOSKOV ET AL. F. (1991). A family of low and high copy replicative, integrative and single-stranded S. cerevisiaelE. coli shuttle vectors. Yeast 7, 609-615. Breathnach, R. and Chambon, P. (1981). Organization and expression of eukaryotic split genes coding €or proteins. Annu. Rev. Biochem. 50, 349-383. Chatoo, B. B., Sherman, F., Azubalis, D., Fjellstedt, T. A., Mehnert, D. and Ogur, M. (1979). Selection of lys2 mutants of the yeast Saccharomyces cerevisiae by the utilization of a-aminoadipate. Genetics 93, 5 1-65. Clement, B. and Kunze, T. (1990). Hepatic microsomal N-hydroxylation of adenine to 6-N-hydroylaminopurine. Biochem. Pharmacol. 39, 925 -933. Clement, B. and Kunze, T. (1992). The reduction of 6-N-hydroxylaminopurine to adenine by xanthine oxidase. Biochem. Pharmacol. 44, 1501-1 509. Cox, B. S. and Parry, J. M. (1968). The isolation, genetics, and survival characteristics of ultravioletsensitive mutants in yeast. Mutat. Res. 6, 3746. Freese, E. (1959). The difference between spontaneous and base-analogue induced mutations of phage T4. Proc. Natl. Acad. Sci. USA. 45, 622-633. Ginner-Sorolla, A. and Bendich, A. (1958). Synthesis and properties of some 6-substituted purines. J. Amer. Chem. Soc. 80, 3932-3937. Ito, H., Fukuda, Y., Murata, K. and Kimura, A. (1983). Transformation of intact yeast cells treated with alkali cations. J. Bacteriol. 153, 163-168. Jones, E. W. and Fink, G. R. (1982). Regulation of aminoacid and nucleotide biosynthesis in yeast. In Strathern, J. N., Jones, E. W. and Broach, J. R. (Eds), The Molecular Biology of the Yeast Saccharomyces. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Kornberg, A. and Baker, T. A. (1992). DNA Replication. W.H. Freeman and Company, New York. Kushnirov, V. V., Ter-Avanesyan, M. D., Didichenko, S. ’A. et al. (1990). Divergence and conservation of SUP2(SUP35) gene of yeast Pichia pinus and S. cerevisiae. Yeast 6,461472. Lomax, C. A. and Woods, R. A. (1969). Mutant of yeast sensitive to 2,6-diaminopurine. J. Bacteriol. 100, 817-822. Meyer, S., Kvalnes-Krick, K. L. and Schramm, V. L. (1989). Characterization of A M D , the AMP deaminase gene in yeast. Production of amd strain, cloning, nucleotide sequence, and properties of the protein. Biochemistry 28, 873&8743. Michaels, M. L. and Miller, J. H. (1992). The GO system protects organisms from the mutagenic effect of the spontaneous lesion 8-hydroxyguanine (7,8-dihydro8-oxoguanine). J. Bacteriol. 174, 6321-6325. Miller, J. H. (1972). Experiments in Molecular Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Mo, J.-Y., Maki, H. and Sekiguchi, M. (1992). Hydrolytic elimination of a mutagenic nucleotide, H A M I GENE 29 8-oxodGTP, by human 18-kilodalton protein: Sahota, A., Ranjekar, P. K., Alfonco, J., Lewin, A. S. sanitization of nucleotide pool. Proc. Nutl. Acud. Sci. and Taylor, M. S. (1987). Mutants of Succharomyces USA 89, 11021-11025. cerevisiue deficient in adenine phosphoribosyltransMorrison, A., Bell, J. B., Kunkel, T. A. and Sugino, A. ferase. Mutation. Res. 180, 81-87. (1991). Eukaryotic DNA polymerase aminoacid se- Sambrook, J., Fritsch, E. F. and Maniatis, J. (1989). quence required for 3l-5' exonuclease activity. Proc. Molecular Cloning: A Laboratory Manual. Cold Nutl. Acud. Sci. USA 88, 9473-9477. Spring Harbor Laboratory, Cold Spring Harbor, NY. Noskov, V. N., Maki, S., Kawasaki, Y., et ul. (1994). Shcherbakova, P. V., Noskov, V. N., Pshenichnov, The RFC2 gene encoding a subunit of replication M. R. and Pavlov, Y. I. (1995). Base analog 6-Nfactor C of Succhuromyces cerevisiue. Nucl. Acids. hydroxylaminopurine mutagenesis in yeast Succharomyces cerevisiue is controlled by replicative DNA Res. 22 (9), 1527-1535. Noskov, V. N. (1988). The studies of the mutapolymerases. Mutation Res., submitted. genic action of 6-N-hydroxylaminopurine and b- Sherman, F., Fink, G. R. and Hicks, J. B. (1986). propiolactone in diploid yeast Succhuromyces cereviLaboratory Course Manual for Methods in Yeast Genetics. Cold Spring Harbor Laboratory, Cold siae strains. PhD Thesis, University of Leningrad. Pavlov, Y. I. (1986). Sacchuromyces cerevisiue mutants Spring Harbor, NY. highly sensitive to the mutagenic action of 6-N- Woods, R. A,, Roberts, D. G., Friedman, T., Jolly, D. and Filpula, D. (1983). Hypoxanthine: guanine hydroxylaminopurine. Soviet Genetics 22, 1099-1 107. phosphoribosyltransferase mutants in Succharomyces Pavlov, Y. I., Noskov, V. N., Lange, E. K., Moiseeva, E. V., Pshenichnov, M. R. and Khromovcerevisiue. Mol. Gen. Genet. 191, 401412. Borisov, N. N. (1991). The genetic activity of N6- Woods, R. A,, Roberts, D. G., Stein, D. S. and Filpula, D. (1984). Adenine phosphoribosyltransferase muhydroxyadenine and 2-amino-N6-hydroxyadenine in Escherichiu coli, Sulmonellu typhimurium and tants in Succhuromyces cerevisiue. J. Gen. Microbiol. Succhuromyces cerevisiue. Mutut. Res. 253, 3346. 130, 2629-2637.