The Yeast RAD50 Gene Encodes a Predicted 153-Kd Protein Containing a Purine Nucleotide-Binding Domain Andtwo Large Heptad-Repeat Regions
Total Page:16
File Type:pdf, Size:1020Kb
Copyright 0 1989 by the Genetics Society of America The Yeast RAD50 Gene Encodes a Predicted 153-kD Protein Containing a Purine Nucleotide-Binding Domain andTwo Large Heptad-Repeat Regions Eric Alani, S. Subbiah and Nancy Kleckner Department of Biochemistry and Molecular Biology, Harvard University, Cambridge,Massachusetts 02138 Manuscript received December 24, 1988 Accepted for publication February 8, 1989 ABSTRACT The RAD50 gene of Saccharomyces cerevisiae is required for chromosomesynapsis and recombination during meiosis and for repair of DNA damage during vegetative growth. The precise role of the RAD50 gene productin these processesis not known. Most rad50 mutant phenotypes can be explained by the proposal that the RAD50 gene product is involved in thesearch for homology between interacting DNA molecules or chromosomes, but there is no direct evidence for this model. We present here the nucleotide sequence of the RAD50 locus and an analysis of the predicted 153-kD RAD50 protein. The amino terminal region of the predicted protein contains residues suggestiveof a purine nucleotide binding domain,most likely foradenine. The remaining 1 170 amino acids consist of two 250 amino acid segments of heptad repeat sequence separatedby 320 amino acids,plus a short hydrophobic carboxy-terminaltail. Heptad repeats occurin proteins such as myosin and intermediate filaments that form a-helical coiled coils. One of the two heptad regions in RAD50 shows similarity to the S-2 domain of rabbit myosin beyond that expected for two random coiled coil proteins. HE RAD50 gene of Saccharomyces cerevisiae plays one percent of the spores from a rad50 meiosis are T an important role in both production of viable viable (GAMEet al. 1980). spores in meiosis and in the repair of DNA damage During vegetative growth, RAD50 function is re- during vegetative growth (HAYNESand KUNZ 1981; quired for repair of some types of DNA damage, GAMEet al. 1980). including double strand breaks induced by ionizing During meiosis, RAD50 function is required for two radiation or the radiomimetic alkylating agent rneth- interrelated events of meiotic prophase, chromosome ylmethanesulfonate (MMS) (GAMEand MORTIMER synapsis and recombination, both of which are re- 1974; STRIKE,cited in HAYNESand KUNZ 198 1; J. quired for proper segregation of chromosomes at the NITISS and M. RESNICK,personal communication). first meiotic division (BAKERet al. 1976). In rad50 rad50mutants also exhibita delay in the spindle mutants, the defect in chromosome synapsis is mani- elongation step of normal vegetative growth, a phe- fested cytologically by the absence of mature synap- notype which suggests that they are defective in repair tonemal complexes (SC) (BYERS,as cited in MALONE, of spontaneous DNA damage as well (FARNETet al. JORDANand WARDMAN1985). More specifically, axial 1988; MONTELONE,HOEKSTRA and MALONE1988). elements form between each pair of sister chromatids RAD50-dependent DNA repair is thought to occurby but never come together to form mature tripartite a recombinationalpathway (HAYNESand KUNZ 1981). SCs (asynapsis) (FARNETet al. 1988; R. PADMOREand Both chromosome synapsis (SC formation)and N. KLECKNER,unpublished data). meiotic recombination require that homologous chro- The rad50 defect in meiotic recombination is ab- mosomes or DNA molecules find one another. Some solute and involves a block at a very early step in the models propose that synapsis is required for recom- recombination process as revealed by both phenotypic bination; other models propose more complex rela- and epistasis analysis (MALONEand ESPOSITO198 1 ; tionships. An economical hypothesis would be that a MALONE1983; MONTELONE, HOEKSTRAand MALONE single homology search is used for both processes. We 1988). RAD50 appears to act at a very early time in have proposed that RAD50is requiredfor such a meiosis; it is required for the appearanceof a specific homology search andthat the RAD50 dependent meiotic signal which occurs just at the conclusion of meiotic signal is a manifestation of this search (FARNET premeiotic DNA replication, and which manifests it- et al. 1988; L. CAO, E. ALANIand N. KLECKNER, self as a site specific double strand break or pair of unpublished data). double strand ends (L. CAO,E. ALAN and N. KLECK- The relationship between meiotic and mitotic phe- NER, unpublished data). The rad50 defects in meiotic notypes of rad50 mutantsis also not established. How- recombination and synaptonemal complex formation ever, theinvolvement of RAD50 inDNA repair makes are sufficient to account for the fact that fewer than it seem more likely that it is directly involved in (knrtics 122: 47-57 (May, 1080) 48 E. Alani, S. Subbiah and N. Kleckner meiotic chromosome metabolism rather than playing formed using standard methods (ITOet al. 1983). an indirect and/or regulatoryrole. The role proposed Nucleic acid techniques: All restriction enzymes, T4 DNAligase, T4 DNA polymerase and BAL31 were pur- meiosis could also account its for RAD50 in for role chased from New England Biolabs and used according to in DNA repair, since a recombinational repair path- the manufacturer's specifications. Plasmid DNAwas isolated way would alsorequire recognition and interaction of by a cleared lysate protocol (MANIATIS,FRITSCH and SAM- homologous molecules. Arguments thatRAD50 is re- BROOK 1982). DNA manipulations including isolation of quired for recognition of homologousDNA segments single stranded DNA from "13 origin plasmids was de- scribed previously. Chromosomal yeast DNA isolation and havealso been putforward by J. NITISSand M. Southern blotting were described previously (HOLMet al. RESNICK(personal communication). 1986; SOUTHERN1975; MELTONet al. 1984). To further probe the nature of the RAD50 gene Plasmid constructions: All plasmids were derived from product, we have carried out DNA sequence analysis the following vectors (relevant genotype and/or descrip- of a previously cloned RAD50 gene. This analysis, tion): YEP24 (URA3,2p; ROSE, GRISAFIand BOTSTEIN1984; New England Biolabs Catalog 1986), YCP50 (URA3, ARSI, describedbelow, has identified a 1312 aminoacid CEN4; STINCHCOMB,MANN and DAVIS1982) pSG205 (7-kb open reading frame which corresponds to the geneti- RAD50 HindIII fragmentinserted into YCP50, from S. cally defined RAD50 gene. The predicted 153-kD GOTTLIEBand R.ESPOSITO), pOL13 (pSG205 derivative protein has three interesting features:(1) The amino with RAD50 in opposite orientation), and "13 sequencing terminal region of the predicted protein contains res- vectors pGC-1, pGC-2, POL10 (pGC-I with the HindIIIsite destroyed) and pNKY47 (POL10 with a HindIII linker at idues diagnostic of a purine nucleotide binding do- BamHI site) (MYERSand MANIATIS1985). main (ROSSMAN,MORAS and OLSEN1974), probably pNKYlOO is a vector that was used to clone chromosomal for adenine(FRY, KUBY and MILDVAN1986). (2)Two sequences upstream of the 5' HindIII site in RAD50. A 4.0- regions, one at amino acids 177-421 and another at kb HindIII-Sal1 subclone from pSG205 was inserted into amino acids 743-995, are segments of heptad repeat the HindIII-Sal1 backbone of pBR322 to make pNKY 19. A HindIII URA3 subclone isolated from YEP24 was inserted sequencecharacteristic of alphahelical coiled coil into the HindIII site of pNKY19 to form pNKY100. In proteins such as myosin and intermediate filaments pNKYlOO URA3 transcription isin the same direction as (MCLACHLANand KARN 1982; STEINERT,STEVENS RAD50 and EcoRI cuts only once in the plasmid, between and Roop 1985). (3) Heptad 743-995 also shows a amp' and URA3. significant similarity to S2the domain of rabbit myosin pNKYlOl is a vector bearing the RAD50 gene and its upstream sequences. pNKYlO0 was linearized within the (CAPONYand ELZINGA198 1). RAD50 gene by an NsiI restriction digest and then inte- grated into the RAD50 locus of NKY501 (a,his#-7 13, lys2- MATERIALS AND METHODS 20, sufl0, ura3, Ura- derivative of RE82 1, kindly supplied by R. ESPOSITO)by lithium acetate transformation (blot not Strains: Escherichiacoli stain "294 (F-, endA, hsdR, shown). Chromosomal DNA from the integrant was isolated, supE44, thiA) was used for all plasmid manipulations (GUAR- digested with EcoRI, ligated and transformed into "294. ENTE et al. 1980). JMlOl (Alac, thi, supE/F, traD36, proA, An amp' plasmid bearing the HindIII-Sal1 RAD50 fragment proB, laciq, lacZA, M15) was used as a host to preparesingle plus 2.8 kb of additional upstream RAD50 sequence was stranded DNA for dideoxy sequencing (MESSING 1983). isolated and designated pNKY 10 1. Saccharomycescerevisiae strain NKY551 is a diploid that pNKY74 is a pGC derived vector (MYERSand MANIATIS contains a complete coding region deletion of RAD50 1985) that contains an 8.8-kb RAD50 clone. RAD50 DNA (HindIII-Sal1 deletion, Figure 1). NKY551 (a/a, rad50A:: upstream of the coding region was isolated on a 2 kb BglII- hisG/rad50A::hisG, ura3lura3, lys2/lys2, ho::LYSB/ho::LysP) HindIII fragmentfrom pNKYlOland inserted into the was constructed by transforming NKY278 (a/a, ura3/ura3, BglIl and HindIII backbone of pNKY47 to form pNKY73. lys2/lys2, ho::LYS2/ho::LYS2 to UraC with BglII and EcoRI The RAD50 locus, containing the coding region plus both digested pNKY83 (described below). The integration of this 5' and 3' flanking sequence, was reconstructed by inserting fragment resulted in a complete coding region deletion of the 7-kb HindIII RAD50 fragmentfrom pSG205 into RAD50 (blot not shown). One such transformant, NKY492, pNKY73 to form the integrating vector pNKY74. was plated onto 5-fluoro-orotic (5-FOA) acid plates and a pNKY1070 is an ARSI, CEN4 vector that contains 6.6 kb Ura- strain, NKY546, was selected (ALANI,CAO and KLECK- of RAD50 sequence. pNKY74 was digested with EcoRI and NER 1987). Tetrad dissection of NKY546 yielded spores a 6.6-kb RAD50 fragment was inserted into the EcoRI site which segregated 2:2, MMS', wild-type growth: MMS", slow of YCP50. growth (FARNETet al. 1988). Two rad50A haploid segre- pNKY83 was used to delete the Hind-Sal1 RAD50 se- gants from NKY546 were mated to form the homozygous quences from the chromosome in order to createcomplete rad50A strain NKY551. This strain sporulated at high fre- chromosomal coding region deletions of RAD50.