MolGenGenet (1989) 218:64-71 © Springer-Verlag 1989

Molecular cloning of SNM1, a yeast gene responsible for a specific step in the repair of cross-linked DNA Eekard Haase, Dorothea Riehl, Michael Mack, and Martin Brendel Institut ftir Mikrobiologie der Johann Wolfgang Goethe-Universit/it, D-6000 Frankfurt/Main, Federal Republic of Germany Summary. We have isolated yeast gene SNM1 via complementation of sensitivity towards bi- and tri-functional alkylating agents in haploid and diploid yeast DNA repairdeficient snml-1 mutants. Four independent clones of plasmid DNA containing the S N M 1 locus were isolated after transformation with a YEp24-based yeast gene bank. Subcloned SNMl-containing DNA showed (i) complementation of the repair-deficiency phenotype caused by either one of the two different mutant alleles snml-1 and snml-2ts; (ii) complementation in haploid and diploid yeast snml-1 mutants by either single or multiple copies of the SNM1 locus; and (iii) that the SNM1 gene is at most 2.4 kb in size. Expression of SNM1 on the smallest subclone, however, was under the control of the G A L l promotor. Gene size and direction of transcription was further verified by mutagenesis of SNM1 by Tnl0-LUK transposon insertion. Five plasmids containing T n l 0 - L U K insertions at different sites of the SNMl-containing DNA were able to disrupt the function of genomic SNM1 after gene transplacement. Correct integration of the disrupted S N M I : : T n l O - L U K at the genomic site of S N M 1 was verified via tetrad analysis of the sporulated diploid obtained after mating of the S N M I : : T n l O - L U K transformant to a haploid strain containing the URA3 S N M 1 wild-type alleles. The size of the poly(A) + RNA transcript of the SNM1 gene is 1.1 kb as determined by Northern analysis. Key words: Saccharomyees cerevisiae- DNA repair- Crosslink - Transposon mapping - Nitrogen mustard

Introduction The unicellular eucaryote Saceharomyces cerevisiae contains at least 50 genes that have a function in the repair of damaged DNA (Friedberg 1988; Haynes and Kunz 1981). Some of these genes seem to control highly specific steps in the repair of certain DNA lesions, i.e., yeast strains containing their mutant Mleles yield phenotypes with sensitivity towards mutagens with common functionality. As an example, defects in gene SNM1 leading to mutant alleles snml-1 and snml-2 ts, result in extreme sensitivity towards cross-linking agents such as nitrogen mustard (Ruhland Abbreviations: Tnl0-LUK, TnlO-laeZ-URA3-kanR Offprint requests to: M. Brendel

et al. 1981 a; Siede and Brendel 1981), triaziquone (Ruhland etal. 1981b), and cis-diamminedichloroplatinum (II) (cisDDP; Brendel and Ruhland 1984) and to only moderate or wild-type-like sensitivity towards monofunctionally alkylating agents, ultraviolet radiation, and ionizing radiation (Ruhland et al. 1981 b). Genetic analysis of double mutants in which snml was paired with mutant alleles of other repair genes, mostly with different rad alleles, revealed epistatic interaction (cf. Brendel and Haynes 1973) with mutant alleles of the excision repair pathway (Siede and Brendel 1982); thus SNM1 has been assigned to the RAD3 repair pathway in yeast (Friedberg 1988). However, SNM1 differs from the RAD1 to RAD4 genes in that it apparently does not control the incision step in either UV- or nitrogen mustard (HN2)-damaged DNA (Magafia-Schwenke et al. 1982; Wilborn 1989) but seems to lack a later function in the excision repair of DNA cross-links. Recently, SNM1 has been shown to be allelic to P S 0 2 (Cassier-Chauvat and Moustacchi 1988) whose defective allele pso2-1 was isolated as sensitive towards photoactivated bifunctional psoralens (Henriques and Moustacchi 1980). The genotoxicity of such a treatment in yeast (Averbeck and Moustacchi 1975) is correlated with the formation of thymidyl mono- and di-adducts (interstrand cross-links) in yeast DNA (Bankmann and Brendel 1989). The above-mentioned biological and genetical data suggest an important role of S N M I in the repair of DNA cross-links, most probably those of the interstrand type [8-MOP+UVA does not produce any intrastrand crosslinks like the bifunctional mustards or triaziquone (see Bankmann and Brendel 1989; cf. Brendel and Ruhland 1984)]. We therefore think it of great importance to molecularly clone this gene in order to elucidate its structure and to obtain some knowledge of its function. Materials and methods Yeast strains, bacteria, bacteriophage, and plasmids. The genotypes of the haploid and diploid strains of S. cerevisiae used in this study are shown in Table 1. Diploid EH3851 was constructed by mating of the two haploid uracil prototrophic strains MKPo snml-O:: Tnl0-LUK with MB11145D/A SNM1 (Ruhland et al. 1981 a). The yeast-Escherichia coli shuttle vectors YEp24 and pBM272 carrying the yeast URA3 gene, were kindly provided by Dr. R. Fleer, Stanford University, and by Dr. M. Johnston, Washington University. A yeast genomic library obtained by ligation of a limited

65 Table 1. Relevant genotypes and sources of yeast strains Strain WS8069-115 EH3709-2A

Genotype

MATa ade2-1 ura3-53 snml-1 MATa ade2-1 ura3-52 snml-1 his5-2 lysl-1 EH3714-2B MATa ade2-1 ura3-52 SNM1 his5-2 lysl-1 leu2-3,112 MBltl4-5D/A MATa his5-2 lysl-1 SNM1 EH3798-1D MA T~ ade2-1 ura3-52 snml-2 ts leu2-3,112 his5-2 his4 trplA-901 canl-lO0 MKPo MAT~ ade2-1 ura3-52 SNM1 his3A-200 lys2-1 leu2-3,112 trplA-901 canl-lO0 MA Ta/MA Tc~lysl-I/L YS1 EH3636 ade2-1/ADE2 ura3-52/ura3-52 leu2-3,112/LEU2 snml-1/snml-1

within the SNM1 gene was physically mapped by restriction analysis of yeast plasmids amplified in E. coli strain 490.

Source W. Siede This study This study A. Ruhland This study B. Kunz This study

Sau3AI digestion of yeast D N A with the BamHI-restricted D N A of the shuttle vector YEp24 (Carlson and Botstein 1982) was kindly provided by Dr. D. Botstein, MIT. Bacteria, plasmids, and bacteriophage 2NK1224-containing T n l 0 - L U K were kindly furnished by Dr. N. Kleckner, Harvard University. Preparation of DNA and transformation procedures. Plasmid D N A was isolated from E. coil by alkaline lysis (Birnboim and Doly 1979). Purification, restriction, ligation, and analysis on agarose gels of plasmid D N A were performed as described by Maniatis et al. (1982). Electro-eluted D N A was further purified on Elutip columns (Schleicher and Schiill). E. coli was transformed according to Dagert and Ehrlich (1979), and yeast according to the method of Ito et al. (1983) with modifications as described by Rodriguez and Tait (1983). Yeast plasmid D N A was isolated according to methods described by Rodriguez and Tait (1983) or by Holm et al. (1986). Tn 10 transposon mutagenesis and mapping of SNM 1. E. coli strain NK5830, containing the transposase plasmid pNK629 was transformed with a monomer of the multicopy shuttle vector pEH1612 containing the SNM1 gene (cf. Fig. 2). Transposition of TnlO-lacZ-URA3-kanR (Huisman et al. 1987) - from now on abbreviated T n l 0 - L U K into the SNMl-containing shuttle vector was carried out by the " l a m b d a - h o p " procedure as described by Way et al. (1984). Transformants of E. coli resistant to kanamycin and ampicillin were selected and D N A of the pooled transformants was isolated by alkaline lysis (Maniatis et al. 1982). After retransformation of the lambda-resistant E, eoli strain NK8017, plasmid D N A was prepared from some 600 transformants; the pooled D N A was then used to transform either haploid yeast strain EH3709-2A snml-1 or diploid yeast strain EH3636 which is homozygous for the snml-1 mutant allele. Transformants were selected on Synco medium lacking uracil. In the next step transformants prototrophic for uracil were screened for complementation of the snml-1 mutant allele, i.e., they were tested for HN2 resistance/sensitivity. Those not complementing the snml-1 mutant allele were scored as putative T n l 0 - L U K integrants. The location of T n l 0 - L U K insertions within pEH1612 and

Gene transplacement. Insertion mutants sensitive to HN2 were constructed according to the one-step gene transplacement method of Rothstein (1983). Plasmid pEHI612 with the integrated T n l 0 - L U K at positions assumed within the SNM1 locus (cf. Fig. 4) was restricted with XbaI, in order to eliminate a 4.5 kb D N A segment containing the URA3 gene and part of the adjacent original 2 gm plasmid D N A of the YEp24 vector. The remaining large D N A fragment was ligated and amplified in E. coli strain 490. A mini-prep of this D N A was linearized by SalI and XbaI restriction and used directly for transformation of the haploid yeast strains M K P o and EH3714-2B (both SNM1 ura3-52), with subsequent screening for HN2 sensitivity amongst the uracil prototrophic transformants. Yeast strain EH3709-2A snml ura3-52 was transformed in parallel with the same linearized D N A mixture and screened for uracil prototrophic transformants able to complement the HN2-sensitive phenotype. Subcloning of SNMl-containing DNA fragments. D N A of pEH3616 and pEH111, originally isolated via complemenration of HN2 sensitivity of diploid and haploid snmi-1 mutants was partially digested with Sau3AI and the resulting fragments separated by preparative agarose gel electrophoresis (Maniatis et al. 1982), Fragments sized between 2 and 6 kb were isolated by electro-elution and ligated into the vectors YEp24 and pBM272, respectively, both of which had been cut by BamHI and treated with alkaline phosphatase. The mixture was transformed into E. coli 490 and transformants that were sensitive to tetracycline and resistant to ampicillin were selected. Some 300 clones were washed off the agar plates using 10 ml MgCI2 (20 mM) and grown for 3 h in liquid culture (40 ml LB medium + ampicillin). Isolated plasmids were used to transform diploid yeast strain EH3636, homozygous for snml-1 and some 100 transformants were tested for their wild-type-like (WT) resistance to HN2. Plasmid D N A was prepared from transformants that demonstrated the appropriate WT phenotypes (cf. Fig. 1) and after retransformation and amplification in E. coli 490 were used for further restriction analysis (Fig. 4). Complementation of HN2 sensitivity by individual yeast insert DNAs in plasmid pBM272 was analyzed in growth media containing either glucose or galactose as energy source, since the SNM1 gene in some of the constructs may be under the control of the GALl promotor.

-

Determination of W T phenotype. Mutagen resistance was determined by replica-plating of transformants on to icecold Synco-ura medium (Ruhland and Brendel 1979) supplemented with HN2, and verification of the HN2-resistant phenotype was rechecked by the agar diffusion test of Ruhland et al. (1981a). Survival curves were calculated as described by Ruhland and Brendel (1979). The cells were pregrown in liquid Synco-ura medium at 30 ° C on a gyrotary shaker to exponential growth phase, washed three times with phosphate buffer (0.067 tool/l, pH 7.0) and treated with HN2 as described by Ruhland and Brendel (1979). Media and general growth conditions. Standard growth medium for yeast was YEPD as described by Fleer and Brendel (1979). Prototrophic yeast transformants were screened on the appropriately supplemented synthetic media (Synco),

66 as described in Ruhland and Brendel (1979). lacZ gene fusions were detected as blue-colored colonies on buffered M63 medium prepared according to Clifton et al. (1978) with 40 lag/ml X-gal (5-bromo-4-chloro-3-indolyl-/~-D-galactopyranoside) added. E. eoli was grown in LB medium as described by Miller (1972), with antibiotics supplemented according to procedures given in Huisman et al. (1987). lO 4

Isolation of yeast RNA, and Northern blotting and hybridization. Total yeast nucleic acid was prepared by glass bead disruption of intact cells in the presence of phenol and sodium dodecylsulfate (Nicolet et al. 1985). Poly(A) + R N A was selected by passage over oligo(dT)-cellulose with subsequent elution by TE buffer (10 mM TRIS, p H 8.0; 1 m M EDTA). Poly(A) + R N A was precipitated with 2.5 volumes of ethanol, washed, dissolved in sterile water, and stored at - 7 0 ° C. Aliquots of 15 lag of the poly(A) + R N A were precipitated with ethanol, dried, and dissolved in formamide buffer (250 ~tl formamide, 83 lal formaldehyde, 50 lal 10x concentrated MOPS, pH 8.0, bromophenol blue). After incubation at 55 ° C for 15 min, RNAs were electrophoresed through an agarose gel (1%) containing formaldehyde, as described by Maniatis et al. (1982). Southern and Northern blottings and hybridizations were performed essentially as described by Maniatis et al. (1982) with the following modifications: filters were processed, pre-hybridized (4 h, 68 ° C), and hybridized (18 h, 68 ° C). The filters were then washed in 2 x SSC-0.5% SDS (two changes of 100 ml for 5 rain each at room temperature), in 2 x SSC-0.1% SDS (two changes of 100 ml for 5 min each at room temperature), and finally in 0.1 x SSC-0.5% SDS (two or three changes of 100ml for 30 rain each at 68 ° C; 1 x SSC is 0.15 M NaC1, 15 mM sodium citrate). Autoradiography was performed at - 7 0 ° C with an intensifying screen. All hybridization probes were prepared by random primed D N A labeling (Boehringer).

5

> "7 i.._

10-2

10-3 -i n o

B

31 10-5 HN2 (mot/L)

10-5

Fig. I A and B. Sensitivity to HN2 of haploid yeast containing different alleles of the SNM1 gene. Survival after HN2 treatment of exponentially growing cells of haploid strains of Saceharomyces cerevisiae. Am ~ , EH3709-2A snml-1 (YEp24);---, EH3709-2A snml-l(pEH3616); ~ a , EH3709-2A snmlI(pEHll 1);o o, EH3714-2B SNMl(YEp24);-_ ;, EH3714a, EH3714-2B SNMI(pEHlll). B 2B SNMl(pEH3616);a Complementation studies with the snml-2 t~ allele of haploid yeast strain EH3798-1D; incubation at 36° and 23° C is denoted by open and closedsymbols, respectively.,, A,-" -~ (YEp24);v v, • ---,, (pEH3616);a ~, (pEHlll at both temperatures)

Results

a strong indication that we did not obtain complementation by suppression of a specific snml mutant allele.

Isolation of SNM1 by complementation

Subcloning and physical mapping of SNM1

We followed two strategies when screening for yeast insert D N A containing the SNM1 locus: (a) complementation of HN2/triaziquone sensitivity in haploid snml-1 mutant WS8069-115, and (b) of HN2 sensitivity alone in diploid yeast strain EH3636 homozygous for the snml-1 mutation via insert D N A contained in a YEp24-based yeast genomic library (Carlson and Botstein 1982). Both approaches yielded uracil prototrophic transformants with wild-type resistance to the cross-linking agents; most of the haploid transformants, however, were false positive, i.e., the HN2/ triaziquone resistance was not due to the co-transformation of SNM1- and URA3-containing yeast insert DNA. Further analysis revealed that screening for resistance with H N 2 as sole cross-linking mutagen in the URA3 transformants led to very high numbers of transformants with wild-typelike resistance to this alkylating agent. Figure 1 shows successful complementation of HN2 sensitivity by two yeast insert DNAs, with a size of 12 kb and 15 kb, respectively, (from a total of 4 true SNM1 transformants), one each derived from the haploid and diploid selection strategy. The fact that the two different mutant alleles snml-1 and snml2 ts could be complemented by both yeast insert DNAs is

Most of the originally cloned yeast insert D N A is not needed for expression of the SNM1 phenotype, since limited Sau3AI digestion yielded substantially smaller subclones which still fully complement the snml-1 and snml-2 ts haploid mutants (Fig. 2). Two of the subclones, pEHI612 with 5 kb (based on episomal vector YEp24) and pEH1102 with 7 kb (based on centromer vector pBM272) carry their own SNM1 promotor, while a third construct, subclone pEH1104 with a size of 2.4 kb (contained in centromer vector pBM272) is under the control of the GALl promotor; snml-1 mutants transformed with this plasmid reveal complementation of HN2 sensitivity only after induction in galactose-containing growth medium (Fig. 3 B). Complementation with plasmids p E H l l 0 2 and p E H l l 0 4 is achieved with a single copy of subcloned yeast insert D N A containing the SNM1 wild-type allele (Fig. 3A, B). Transcription of SNM1 is from right to left (Fig. 2). Further characterization of SNM1 by TnlO mutagenesis

The monomeric target plasmid pEHI612(YEp24) was subjected to TnlO-LUK mutagenesis in E. coli harboring the

67 pEH 1612 U I

pEH 1104 G X ,I I ARS1 CEN4

E I

B/S Nc L~E/4 11"

Xb G B/S X.E ~,E,~ ,G! NeE Xb E , . . . . . . ' II I I URA3 2~m

IBIS B/5 L~ E X'b~E . I'. '. ', '.'.~ ' +-GALl URA3 Nc

I

I

,

I

I I

pEH1102 E I

G X Nc 'I I ~ ARSI CEN4 URA3

B/S~ L~:K III

,G' XBt G II I

L B/S Net ~ E II I I ---GALl lkb

t

I

Fig. 2. Restriction mapping of three vectors whose yeast insert DNA will complement the snml-1 mutant allele. From top to bottom: pEH1612 is based on multi-copy vector YEp24 and contains yeast insert DNA of 5 kb; pEHl104 and pEHl102 are based on the single-copy vector pBM272 and contain yeast insert DNAs of 2.5 and 7.8 kb, respectively. The SNM1 gene in pEH1612 and pEH1102 is under the control of its own promoter while the GALl promotor controls the SNM1 gene in pEHll04. The closed bar represents yeast insert DNA, the open bar represents yeast vector DNA, and the thin line represents DNA from bacterial plasmid pBR322. Abbreviations: B, BamHI; E, EcoRI; G, BglII; L, SalI; Nc, NcoI; S, Sau3AI; U, PvuII; X, Xhol; Xb, XbaI

pNK629 plasmid that contains the gene for transposase (Way et al. 1984). Tnl0-LUK derivatives (integrants) of the target plasmid were isolated via selection for kanamycin/ ampicillin resistance in E. coli NK5830 according to the "lambda-hop" protocol. Some 600 Tnl0-LUK insertion mutants in pEH1612 were obtained with this method (approx. 2% of all TnlO insertions). This plasmid D N A was extracted, pooled, and used for transformation of the haploid and diploid yeast strains EH3709-2A snml-1 ura3-52 and EH3636 homozygous for both mutant alleles, respectively. Screening for uracil prototrophy was followed by screening for complementation of HN2 sensitivity. About half (187 of 354) of the uracil prototrophic transformants were found resistant to HN2, i.e., could complement the snmI-1 mutant allele. Those 167 Tnl0-LUK integrants not complementing the HN2 sensitivity should lie within the S N M 1 locus. Figure 4 depicts the location found by restriction mapping of 12 Tnl0-LUK integrants (transpositions) within and of 28 transpositions outside of the S N M 1 gene. A 13th Tnl0-LUK integrant, located to the left of the XhoI site (open square in Fig. 4) also was non-complementing. However, this construct was found to have a deletion of approx. 1 kb in the D N A flanked by the XhoI and XbaI restriction sites (data not shown), so that non-complementation of snml-1 can be expected by this fact alone. Also, all Tnl0-LUK integrants were screened for functionality of the laeZ gene. Of 354 integrants 100 had the insertion in frame, i.e., lacZ was induced by the promotor of either the S N M 1 gene or of promotors of other loci on the vector. Gene disruption yields viable snml-0: : Tnl O-LUK mutants

]0-I

ul

t

10-2

lo 4

A o

10-s 0 HN2 (mot/I)

10-5

Fig. 3A and B. Complementation of HN2 sensitivity in haploid yeast snml-1 mutants by subclones of yeast DNA present on singleor multi-copy vectors (described in Fig. 2). A o o, EH3709-2A snm1-1(pBM272) ; H , EH3709-2A snml-l(pEH1612); w-----~, EH3709-2A snm1-1(pEHl102); o o, EH3714-2B SNM1(pBM272) ;A A, EH3714-2B snml-1 : : Tnl0-LUK. B Cells were pre-grown in medium in which glucose had been replaced by galactose.[] [], EH3709-2A snm1-l(pBM272);= ", EH3709-2A snm1-l(pEH1104);o o, EH3714-2B SNMl(pBM272)

Five individual gene disruption integrants were isolated after transforming haploid yeast with the genotype ura3 S N M 1 using the linearized vector EH1612 (see the Materials and methods). All five transformants were prototrophic for uracil (URA3) and showed identical or slightly higher sensitivity to HN2 than mutant strains containing the s n m l 1 allele (Fig. 3A), i.e., contained a disrupted snml-O: : Tnl0LUK locus. The correct integration at the S N M 1 locus was verified for one integrant through tetrad analysis of diploids obtained by crossing MPKo snml-O::TnlOLUK(92/01) (cf. Fig. 4) with the wild-type strain MB11145D/A S N M 1 URA3. The segregation of both markers is given in Table 2; we find 2: 2 segregation for the HN2 resistance/sensitivity phenotype and all HN2-sensitive clones exhibit uracil prototrophy. The uracil marker shows aberrant segregation in the tetrads, i.e., 4: 0, 3:1, and 2:2 for prototrophy/auxotrophy as would be expected for an integration of URA3 at the site of the S N M 1 gene. Also, none of the HN2-sensitive s n m l - O : : T n l O - L U K integrants could, after mating to an HN2-sensitive haploid snml-1 mutant strain, complement its HN2-sensitive phenotype. Insertion of Tnl0-LUK in pDR20/01 approx. 1 kb downstream of the XhoI site (cf. Fig. 4) does not disrupt the S N M 1 gene since this plasmid, when linearized by XbaI/SalI restriction and used for gene transplacement, can restore HN2 resistance in an snml-1 mutant (data not shown). Transcript size o f the SNM1 gene

The size of the S N M 1 gene transcript was determined by preparation of poly(A) + R N A and Northern transfer, as shown in Fig. 5 A. Lanes 1 and 2 each contain 15 pg po-

68

B/S URA 3

B/S

/~J

/-"2

pm

j

/

I

E

I

U

I I

L ~. B/S

I

Nc /

T

1,//

~Xb

II

~2~%

E

I

lkb

I

20101 1711145101 10811 92101 20711

Fig. 4. Tnl0-LUK mapping and mutagenesis of the SNM1 locus. The position within pEH1612 of 41 Tnl0-LUK integrants that complement the ura3-52 mutant is shown in the upper panel. Of these Tnl0-LUK integrants, 13 (marked by arrows) destroy the SNM1 gene. Their position is within the right-hand end of the yeast insert DNA and the Xhol cleavage site is approx. 2.2 kb to the left. Transpositions to the left of this site do not destroy the function of the SNM1 gcne. The two orientations of the inserted elements are depicted by circles (~-) and squares (~), respectively. Filled circles represent insertions that yield blue color (laeZ function) when transformed yeast is fed with X-gal; open circles and open squares represent insertions without lacZ expression. The lower part of the figure depicts constructs in which a 4.3 kb fragment was deleted from the pEH1612 Tnl0-LUK insertion plasmid. The constructs were used for gene disruption experiments at the genomic SNM1 locus. For the other symbols and abbreviations of restriction sites cf. Fig. 2 Table 2. Segregation pattern for six markers in 9 tetrads of sporulated diploid EH3851

Genetic marker

SNM1 ADE2 URA3 LEU2 TRP1 CAN1

Segregation of alleles 4:0

3:1

2:2

0 0 1 0 0 0

0 0 6 0 0 0

9 9 2 9 9 9

ly(A) + R N A isolated from either yeast strain A H 2 2 S N M 1 (lane 1) or EH3255-6A S N M 1 (lane 2). Lane 1 shows two discrete bands o f hybridization, 2.4 kb and approx. 1.1 kb in size. Lane 2 shows hybridization with the lower b a n d o f approx. 1.1 kb in size and very weak hybridization with the 2.4 kb band. Lane 3 contains 40 gg of p o l y ( A ) - R N A as a control and yields no visible specific hybridization. W h e n probing these two poly(A) + R N A s with two smaller D N A fragments obtained from pEH1612, one flanked by the two BglII sites, the other by the two EcoRI sites (cf. Fig. 2) we could only observe hybridization at the 1.1 kb b a n d (data not shown). Southern analysis o f chromosomal D N A

Integration o f t r a n s p o s o n T n l 0 - L U K into S N M 1 was targeted by using the plasmids pRH45/01 and pRH108/1, respectively (cf. Fig. 4), linearized by XbaI/SalI restriction for t r a n s f o r m a t i o n o f yeast strain EH3714-2B ura3-52 S N M 1 to the uracil p r o t o t r o p h i c and HN2-sensitive phenotype. Southern analysis o f c h r o m o s o m a l D N A p r e p a r e d from the wild type and two uracil p r o t o t r o p h i c integrants revealed that T n l 0 - L U K (on pRH45/01 and p R H I 0 8 / 1 ) had integrated at the S N M 1 locus (Fig. 5 B). The transposon T n l 0 - L U K has a size o f 6.1 kb and contains one XhoI site approx. 500 bp from one end. XhoI restriction o f chrom o s o m a l D N A o f the wild type, EH3714-2B, gave one fragment o f 14.5 kb in size (Fig. 5B, lane 2), while restriction

Fig. 5. A Northern blot analysis of the SNM1 transcript. Lanes 1 and 2 each contain 15 gg of poly(A) + RNA, lane 3 contains 40 gg of poly(A)- RNA. Formamide-denatured RNAs were fractionated on a 1% formaldehyde-agarose gel prior to transfer to nitrocellulose. The filter was probed with the 1.7 kb XhoI/BglII fragment of the SNM1 gene (cf. Fig. 2, upper construct) that had been labeled with 32p (2 x 10s cpm/gg) by random primed DNA labeling. Size standards at the right are Escherichia coli rRNAs. B Southern analysis of chromosomal DNA isolated from the wildtype strain EH3714-2B and from two transformants carrying the Tnl0-LUK insertion that disrupted SNM1 (pRH45/01). Lane 1 contains HindIII-digested lambda DNA as size standards, lane 2 contains DNA from the wild-type strain EH3714-2B digested with XhoI, and lanes 3 and 4 contain DNA from transformants carrying the Tnl0-LUK insertions. DNAs were fractionated on a 0.7% agarose gel, transferred to nitrocellulose, and hybridized with the 1.7 kb XbaI/BglII fragment of the SNM1 gene exactly as described for Fig. 5A. All sizes are given in kb o f the D N A from the two transformants with integrated T n l 0 - L U K yielded two fragments each, approx. 20 kb and 1.1 kb in size. The large D N A fragment contains most o f the T n l 0 - L U K t r a n s p o s o n (minus 500 bp) while the small fragment consists o f 500 bp from T n l 0 - L U K plus approx.

69 600 bp from the yeast insert D N A (located to the right of the XhoI site in yeast insert DNA; cf. Fig. 4). Discussion

Possible function of the SNM1 gene product in DNA repair The unicellular lower eucaryote Saccharomyces cerevisiae is rapidly becoming an indispensable model when studying genome and chromosome structure, gene control elements, and even the function of individual genes of a living cell. One of the most important genetically controlled functions of a cell is its D N A repair machinery since correct genetic information is a prerequisite for cellular growth and propagation. At least 50 genes are presently known to control steps in D N A repair in yeast (Haynes and Kunz 1981); by epistatic interaction they have been ordered into three pathways of D N A repair (Game and Cox 1972; Brendel and Haynes 1973). Of those genes about a dozen have been cloned and enzymatic activity can be attributed to an even smaller number (cf. the reviews by Moustacchi 1987; Friedberg 1988). Nevertheless, exact phenotypic description of D N A repair mutants coupled with results of molecular dosimetry of D N A lesions has produced plausible suggestions concerning the role of certain D N A repair genes. Presently 15 genes, amongst them SNM1, are allocated to the RAD3 epistasis group which is involved in nucleotide excision repair (Henriques and Moustacchi 1981; Friedberg 1988). Functional alleles of the genes RAD1, RAD2, RAD3, RAD4, and RADIO are absolutely necessary for incision next to pyrimidine dimers (Reynolds and Friedberg 1981; Wilcox and Prakash 1981) and other D N A lesions, such as those caused by mono- and bifunctional photoactivated psoralens (Moustacchi et al. 1983; Miller et al. 1982), i.e., code for an early function in nucleotide excision repair. The function of SNM1 must be lie in control of a different and a later step of this repair pathway, since snml (pso2) mutants show very little sensitivity to UV254 am, X-rays, and monofunctional radiomimetic chemicals (Ruhland etal. 1981b; Moustacchi et al. 1983) and are thought to excise and repair these D N A lesions more or less normally. Also, mutant snml (pso2) shows normal incision in D N A cross-linked by nitrogen mustard (Wilborn 1989) or by photoactivated psoralen (Magafia-Schwenke et al. 1982), indicating that sensitivity to such treatments must be due to a defect in a post-incision step. This information and the fact that mutant snml (pso2) is sensitive to all agents able to form interstrand cross-links, be it photoactivated psoralens (Henriques and Moustacchi 1980), nitrogen mustard, and related bifunctional alkylating agents (Ruhland et al. 1981b) or bifunctional platinum coordination compounds (Brendel and Ruhland 1984; Wilborn and Brendel 1989), points to a possible role of the SNM1 gene product in a repair step specific for damage simultaneously involving both D N A strands. At present it is believed that the P S 0 2 (SNM1) gene product is needed for the repair of D N A double-strand breaks that result during excision of ICL (Moustacchi et al. 1983). The successful cloning of this gene with its suggested reading frame of approximately 1 kb (Fig. 5), just described in this communication, will enable us to determine its structure and hence the nature of the protein it encodes, which in turn may shed some light on its function.

Specific genes for the repair of cross-links in eucaryotes ? There is a phenotypic resemblance between yeast snml mutants and cells from patients suffering from Fanconi's anemia (FA; either complementation group A or B); both are sensitive to cross-linking agents, i.e., mitomycin C and photoactivated 8-methoxypsoralen but not to UV (254 nm) (Ruhland et al. 1981b; Duckworth-Rysiecki et al. 1985; Moustacchi et al. 1987), indicating a possible similarity in the function of both genes. Complementation of the sensitivity of FA cells by the SNM1 gene would give a strong indication that genetic information for this type of D N A repair would be ubiquitous for eucaryotes. Indeed mutant mus308 of Drosophila exhibits a similar mutagen-sensitive phenotype (Boyd et al. 1981). That there is a considerable structural and functional relationship for excision repair enzymes in yeast and mammalian cells was demonstrated by the partial complementation of UV sensitivity of the excision repair-defective CHO cell line by introduction of the RADIO yeast gene (Thompson et al. 1987). Also, substantial structural conservation between proteins of the human Erccl and yeast Radl0 proteins has been demonstrated (van Duin et al. 1986).

Cloning of the SNM1 gene Strong evidence for the cloning of the SNM1 gene comes from the fact that single- and multi-copy plasmids containing at least a 2.4 kb fragment of the originally cloned yeast insert D N A could complement the two individual mutant alleles snml-1 and snml-2 ts (Figs. 1 and 3). The subcloning of SNMl-containing yeast insert D N A in multi- and singlecopy vectors as depicted in Fig. 2 gave a good indication of the maximal size of this repair gene. Since SNM1 is under the control of the GALl promotor in p E H l l 0 4 we have delimited the 5' region of the gene near the XbaI site. The 3' end of SNM1 must lie within the yeast D N A insert of pEH1102 which ends 400 bp downstream from the XhoI site, since this single-copy plasmid still fully complements the HN2-sensitive phenotype of snml-1. Further evidence concerning the maximal size of SNM1 comes from TnlOL U K transposon mutagenesis (Fig. 4): of 41 T n l 0 - L U K insertions into the yeast insert D N A of pEH1612 only the 5 located on the fragment flanked by the XbaI and XhoI sites (Fig. 4) led to genomic disruption of SNM1 after gene transplacement. One other T n l 0 - L U K insertion which occurred downstream from the XhoI site led to non-complementation when the yeast insert D N A was introduced into an snml-1 mutant on a multi-copy plasmid. Upon restriction analysis this isolate (open square in upper construct of Fig. 4) was shown to contain a deletion of approx. 1 kb in the yeast insert D N A (data not shown), which could explain its non-complementing character. Northern analysis showed two bands of hybridization, 2.4 and 1.1 kb in size (Fig. 5A). Because of the following observations we may assume that the 1,1 kb poly(A) + R N A is the transcript of gene SNM1 : (i) Northern blotting of two independent R N A preparations only yielded specific hybridization with the 1.1 kb poly(A) + R N A when probing with two smaller D N A fragments derived from yeast insert D N A of pEH1612, one flanked by the two BglII sites, the other flanked by the two EeoRI sites with sizes of 0.8 and 1.3 kb, respectively (Fig. 2); (ii) comparison of subcloned SNMl-containing

70 yeast insert D N A suggests a gene size of less than 2 kb (Fig. 2); and (iii) the transplacement studies with TnlOL U K disrupted S N M 1 also point to a gene size smaller than 2 kb (Fig. 4). The 2.4 kb poly(A) + R N A found with the larger D N A probe apparently represents the transcript from the next gene, downstream of S N M I , which by TnlOL U K transposition revealed the same direction of transcription (Fig. 4).

for the "lambda-hop" protocol and Dr. D. Botstein for providing the YEp24-based yeast genomic library. We are thankful to Ms. M. Niesen for excellent technical assistance, Ms M. Marcovic for drawing of the figures and Mr. W. Barth for construction of scientific instruments. Most of the reported data come from the Doctoral Thesis of E.H. This research was supported by the Herrmann Willkomm-Stiftung. M.M. was supported by a doctoral fellowship of the Fonds der Chemischen Industrie.

SNM1 is not an essential gene

References

The correct integration at the genomic site of S N M 1 of one of the five T n l 0 - L U K insertions (construct pRH92/01) that lead to disruption o f S N M 1 and hence inability to c o m p l e m e n t the H N 2 sensitivity of snml-1 was verified by tetrad analysis (Table 2) and Southern analysis (Fig. 5 B). With the exception of the U R A 3 and including the S N M 1 marker all tetrads showed the expected 2:2 segregation. The presence of ascospores auxotrophic for uracil from a cross that involved two uracil prototrophic parents is evidence that S N M 1 was disrupted by the URA3-containing transposon and that the linearized construct transplaced the genomic S N M 1 locus, thus generating a haploid transformant carrying both the URA3 and ura3-52 alleles. Genomic D N A o f an S N M 1 wild-type strain, restricted with XhoI, yielded one 14.5 kb band when probed with D N A containing the S N M 1 gene (cf. the Materials and methods) while the two HN2-sensitive strains, containing T n l 0 - L U K disrupted S N M 1 , gave two bands, approx. 20 kb and 1.1 kb in size (Fig. 5 B). The sum of both D N A fragments is about 6 kb larger and thus corresponds to a D N A size expected after integration of the 6.1 kb T n l 0 - L U K transposon (Huisman et al. 1987) into the S N M 1 locus. Since disruption of the S N M 1 gene is not lethal it is not an essential yeast gene. The subcloned S N M 1 gene in p E H l l 0 4 (Fig. 2), under the control o f the G A L l promotor, cannot fully complement the snml-1 phenotype (Fig. 3 B) although it should be fully induced by growth in galactose-containing medium. This could be attributed most probably to the fact that the G A L l p r o m o t o r is not correctly placed in front of the S N M 1 locus; either, part of the S N M 1 p r o m o t o r might still be present or part of the gene is missing in p E H l l 0 4 . Both scenarios could lead to the observed effect. Partial complementation of the radl phenotype can still be achieved although up to 11 codons of the R A D 1 reading frame are missing (Higgins et al. 1983). It is unlikely that an overexpressed correct S N M 1 gene product in p E H I 1 0 4 cannot fully complement the s n m l mutant phenotype, since S N M 1 contained in multi-copy plasmid pEH1612 (apparently overexpressed) does restore the wild-type phenotype to yeast strains carrying s n m l mutant alleles (Fig. 3A). Expression of S N M 1 controlled by the G A L l p r o m o t o r shows that transcription is from right to left. This is verified by the 4 T n l 0 - L U K integrants that allow expression of the lacZ gene which is now under the control of the S N M 1 p r o m o t o r (Fig. 4). Insertion of T n l 0 - L U K in the inverse orientation as in pRH92/01 does not lead to expression of laeZ. Other genes located on the yeast insert D N A to the left o f S N M 1 must have the same direction of transcription since we find lacZ expression only when T n l 0 - L U K is inserted in likewise orientation (Fig. 4, upper part). Acknowledgements. We wish to thank Dr. N. Kleckner for providing us with bacterial strains, plasmids, and bacteriophage needed

Averbeck D, Moustacchi E (1975) 8-Methoxypsoralen plus 365 nm light effects and repair in yeast. Biochim Biophys Acta 395 : 393-404 Bankmann M, Brendel M (/989) UVA-induced 8-MOP binding to cells of Saccharomyces cerevisiae: Comparison of DNA-adducts produced in vitro and in vivo. J Photochem Photobiol, in press Birnboim HG, Doly J (1979) A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res 7:1513-1523 Boyd JB, Golino MD, Shaw KES, Osgood CJ, Green MM (1981) Third chromosome mutagen-sensitive mutants of Drosophila melanogaster. Genetics 97: 607-623 Brendel M, Haynes RH (1973) Interaction among genes controlling sensitivity to radiation and alkylation in yeast. Mol Gen Genet 125:197-216 Brendel M, Ruhland A (1984) Relationships between functionality and genetic toxicology of selected DNA-damaging agents. Mutat Res 133:51-85 Carlson M, Botstein D (1982) Two differentially regulated mRNAs with different 5' ends encode secreted and intracellular forms of yeast invertase. Cell 28:145-154 Cassier-Chauvat C, Moustacchi E (1988) Allelism between psol-1 and rev3-1 mutants and between pso2-1 and snml mutants in Saeeharomyees cerevisiae. Curt Genet 13:37-40 Clifton D, Weinstock SB, Fraenkel DG (1978) Glycolysis mutants in Saccharomyces cerevisiae. Genetics 88:1-11 Dagert M, Ehrlich SD (1979) Prolonged incubation in calcium chloride improves the competence of E. eoli cells. Gene 6: 23-28 Duckworth-Rysiecki G, Cornish K, Clarke CA, Buchwald M (1985) Identification of two complementation groups in Fanconi anemia. Somatic Mol Cell Genet 11:35-41 Fleer R, Brendel M (1979) Formation and fate of cross-links induced by polyfunctional anticancer drugs in yeast. Mol Gen Genet 176:41-52 Friedberg EC (1988) Deoxyribonucleic acid repair in the yeast Saceharomyees eerevisiae. Microbiol Rev 52:70~102 Game JC, Cox BS (1972) Epistatic interactions between four rad loci in yeast. Murat Res 16:353-362 Haynes RH, Kunz BA (1981) DNA repair and mutagenesis in yeast. In: Strathe{n J, Jones EW, Broach JR (eds) The molecular biology of the yeast Saccharomyees. Life cycle and inheritance. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY, pp 371-414 Henriques JAP, Moustacchi E (1980) Isolation and characterization ofpso mutants sensitive to photo-addition of psoralen derivatives in Saceharomyces cerevisiae. Genetics 95:273-288 Henriques JAP, Moustacchi E (1981) Interactions between mutations for sensitivity to psoralen photoaddition (pso) and to radiation (rad) in Saeeharomyces cerevisiae. J Bacteriol 148: 248-256 Higgins DR, Prakash S, Reynolds R, Prakash L (1983) Molecular cloning and characterization of the RAD1 gene of Saccharomyces cerevisiae. Gene 26:119-126 Holm C, Meeks-Wagner DW, Fangman WL, Bostein D (1986) A rapid, efficient method for isolating DNA from yeast. Gene 42:169-173 Huisman O, Raymond W, Froehlich KU, Errada P, Kleckner N, Botstein D, Hoyt MA (1987) A TnlO-lacZ-kanR-URA3 gene

71 fusion transposon for insertion mutagenesis and fusion analysis of yeast and bacterial genes. Genetics 116:191-199 Ito H, Fukuda J, Murata K, Kimura A (1983) Transformation of intact yeast cells treated with alkali cations. J Bacteriol 153:163-168 Magafia-Schwenke N, Henriques JAP, Chanet R, Moustacchi E (1982) The fate of 8-methoxypsoralen photoinduced crosslinks in nuclear and mitochondrial yeast DNA: comparison of wildtype and repair-deficient strains. Proc Natl Acad Sci USA 79:1722-1726 Maniatis T, Fritsch EF, Sambrook J (1982) Molecular cloning. A laboratory manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York Miller JH (1972) Experiments in molecular genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY Miller RD, Prakash L, Prakash S (1982) Genetic control of excision of Saccharomyces cerevisiae interstrand DNA cross-links induced by psoralen plus near-UV light. Mol Cell Biol 2: 939-948 Moustacchi E (1987) DNA repair in yeast: genetic control and biological consequences. Adv Radiat Biol 13:1-30 Moustacchi E, Cassier C, Chanet R, Magafia-Schwenke N, Saeki T, Henriques JAP (1983) Biological role of photo-induced crosslinks and monoadduets in yeast DNA: genetic control and steps involved in their repair. In: Friedberg EC, Bridges BA (eds) Cellular responses to DNA damage. Liss, New York, pp 87-106 Moustacchi E, Papadopoulo D, Diatloff-Zito C, Buchwald M (1987) Two complementation groups of Fanconi's anemia differ in their phenotypic response to a DNA-crosslinking treatment. Hum Genet 75 : 45-47 Nicolet CM, Chenevert JM, Friedberg EC (1985) The RAD2 gene of Saccharomyces cerevisiae: nucleotide sequence and transcript mapping. Gene 36:225-234 Reynolds RJ, Friedberg EC (1981) Molecular mechanisms of pyrimidine dimer excision in Saccharomyces cerevisiae: incision of ultraviolet-irradiated deoxyribonucleic acid in vivo. J Bacteriol 146:692-704 Rodriguez RL, Tait RC (1983) In: Recombinant DNA techniques (an introduction). Addison-Wesley, London Rothstein RJ (1983) One-step gene disruption in yeast. Methods Enzymol 101:202-209 Ruhland AR, Brendel M (1979) Mutagenesis by cytostatic alkylat-

ing agents in yeast strains of differing repair capacities. Genetics 92: 83-97 Ruhland AR, ttaase E, Siede W, Brendel M (1981a) Isolation of yeast mutants sensitive to the bifunctional alkylating agent nitrogen mustard. Mol Gen Genet 181:346-351 Ruhland AR, Kircher M, Wilborn F, Brendel M (1981 b) A yeast mutant specifically sensitive to bifunctional alkylation. Mutat Res 91:457-462 Siede W, Brendel M (1981) Isolation and characterization of yeast mutants with thermoconditional sensitivity to the bifunctional alkylating agent nitrogen mustard. Curr Genet 4:145-149 Siede W, Brendel M (1982) Interactions among genes controlling sensitivity to radiation (RAD) and to alkylation by nitrogen mustard (SNM) in yeast. Curt Genet 5:33-38 Thompson LH, Salazar EP, Brookman KW, Collins CC, Stewart A, Busch DB, Weber CA (1987) Molecular biology of DNA repair. J Cell Sci 6 (Suppl):97-110 van Duin M, de Witt J, Odijk H, Westerveld A, Yasui A, Koken MHM, Hoeijmakers JHJ, Bootsma D (1986) Molecular characterization of the human excision repair gene ERCCI: cDNA cloning and amino acid homology with the yeast DNA repair gene RADIO. Cell 44: 913-923 Way JC, Davis MA, Morisato D, Roberts DE, Kleckner N (1984) New TnlO derivatives for transposon mutagenesis and for construction of lacZ operon fusions by transposition. Gene 32: 369-379 Wilborn F (1989) Genetik und Biochemie der DNA-Dunkelreparatur yon Crosslink-Sch/iden bei Saceharomyees eerevisiae. Dissertation, J.W. Goethe-Universit/it, Frankfurt/Main Wilborn F, Brendel M (1989) Formation and stability of interstrand cross-links induced by eis- and trans-diamminedichloroplatinum (II) in the DNA of Saecharomyces cerevisiae strains differing in repair capacity. Mutat Res, in press Wilcox DR, Prakash L (1981) Incision and post-incision steps of pyrimidine dimer removal in excision-defective mutants of Saccharomyces cerevisiae. J Bacteriol 148 : 618-623 C o m m u n i c a t e d by C.P. Hollenberg Received December 12, 1988