Tobacco Documents Online

Dr.Harmon McAllister,Ph.D. Research Director The Council for Tobacco Research 900 Third Avenue New York,NY 10022 July. 12,1988 Dear Dr.McAllister: In compliance with the requirements for the first step in your application procedure~I am enclosigg the pre- liminary outline of a project entitled "Mechanism~ of Spontaneous and Mutagen-Induced Deletions", for which I seek support from The Council for Tobacco Research.I am including the following appendix material:l) a manuscript in press in Mutation Research, 2) page proofs of an article to appear in a Symposium volume, and 3) a table summarizing results which are being prepared for publication, and which I refer to as Appendix I. Thank you very much for sending the application material promptly, and for your consideration. Sincerely yours [lias Balb~nder,Ph.D. Professor (Adjunct) Dept. of BBGN 40000193

MECHANISMS OF SPONTANEOUS A~D MUTAGEN-It~DUCED DELETIONS I) Synopsis of project.p~sent q~als and status. Evidence is accumulating daily that mutations at specific chromosomal locations are at the base of practically every type of cance~hat has been investigated, and that these mutations are predominantly major rearrangements of DNA such as deletions, du- plications,translocations,etc, rather than small changes of single base pairs.Deletions are frequently found in cancers(l-3) and are the most extensively studied rearrangements,parti- cularly in prokaryotes.Most of the deletions described to date are explained by misalignement mutagenesis models (4,5). Thes~postulate that,during DNA replication, single stranded regions can undergo ~lippage followed by interstrand mispairing between direct repeats (4)or formation of hairpin stuctures between inverted repeats (5),and these unstable intermediates are somehow resolved as deletions.While these models account for the many reports showing that deletions occur preferentially at the types of sequences mentioned above,they have not been tested. Also, they do not address the problem of mechanisms responsible for resolving the transient structures as deletions, or how deletions are induced by mutagens.The long term goal of this rese arch is to find answers to these complex questions.Our strategy is based on the use of simple engineered E.coli plasmids which allow full control of the structural and genetic parameters involved in causing deletions.This is explained in detail in the two enclosed papers (in press).Briefly,the insertion of DNA fragments of known size and sequence within unique restriction sites-in a selectable plasmid marker,such as lacZ or a gene for drug resistance ,will inactivate the gene and deletion of the insert will restore gene activity giving a selectable phenotype. This allows precise quantitation of deletion rates by a simple reversion test and, since the reversion is caused only by a very specific deletion event, no extensive sequencing of revertants is necessary.The processes of mutagenesis are extremely complex and require the participation of an intricate network of enzymes whose mission is to remove mutagen damage from DNA and prevent cell death. Mutations can occur occasionally as mistakes in the process of removing this damage.At this time the mechanisms of mutagenesis are better understood in E.coli than in any other organism, and a large number of strains carrying mutations in genes for DNA metabolism enzymes are available.Plasmids can be easily introduced into these strains to test the effect of such mutations on deletion formation.They also can be used as versatile genetic screens to isolate new mutants which stimulate the occurrence of specific deletions (see enclosed article). This strategy permits us to employ genetics, physiology and molecular biology in a combined and powerful approach to the problem. In a relatively short time and with minimal staffing we have obtained some new insights into the mechanisms of spontaneous deletions .These have helped us to formulate the working hypothesis which we will present in the next section. 4O0OO194

Because of space considerations, I will limit the summary of the current status of the project to those areas which are of relevance to this application. I- The use of pre-desiqned plasmids to study deletions See enclosed manuscript for details. ~;~. DNA fragments cloned into the unique EcoR~ the chloramphenicol acetyl transferase (CAT) gene of plasmid pBR325 are deleted precisely between the two EcoRl sites generated as a result of the insertion giving chloramphenicol resistant (Cmr) revertants.We have introduced different fragments of~70 bp into this site, the sequences of which are shown in fig.l of the manuscript. Our results (table 1 of manuscript) show:(1) Deletion of all inserts is independent of recA+ but there is a small increase in the presence of this allele.(2)A palindromic sequence is deleted ten times as frequently than a non palindromic one of equivalent size (pOCE15 vs pRSl).(3) Surprisingly, the presence of an extra EcoRl site on the 3' side of the non-palindromic insert(pRS4) increased its deletion frequency 5-10 fold over that~f a palindrome (pOCElS).This observation is novel and significant since it suggests that deletion intermediates can be stabilized in more ways than through formation of hairpin structures by palindromic sequences. 2- Can deletions occur as untarqeted mutations? We know practically nothing at present about how mutagenic agents induce deletions. The possibility had been suggested that deletions are predominantly untargeted mutations brought about by misreplication in cells induced for a repair system like SOS (6- 9).Miller and Low (10)reported that induction of the SOS system did not increase deletions in the lacI gene, but we have just found that deletions on a pBR325-derived plasmid were clearly and consistently increased in strains derepressed for the SOS response(see Appendix I, and also table 2 in manuscript).These results support the idea that deletions can be untargeted mutations caused bySOS processing. IIOutline of plans and qoals of propose4research-Next steps~t~ be taken. The following working hypothesis to explain how spontaneous as well as mutagen-induceddeletions originate is based on all the data available to date,including our own latest results. The i ikel ihood that slippage and mispairing occur on replicating DNA with the formation of transient structures (such as hairpins) which can interfere with DNA replication ( 11- 13) ,suggests that one possible role for DNA repair systems could be to correct or eliminate these structures to prevent cell death. In this respect, then, such transient structures resemble adducts formed on DNA by mutagenic agents with the difference that the former are reversible and DNA can return to its normal configuration spontaneously/ while adducts form covalent bonds with bases on DNA and must be removed by repair mechanisms.Thus,from the point of view of DNA repair systems,a very stable deletion intermediate can be the equivalent of a replication-interrupting D~A lesion.This would be consistent with 40000195

mutagen-induced deletions being predominantly untargeted and their frequency increased by SOS processing, since the misalignement itself could be considered as the target of this repair system. This hypothesis is also consistent with the participation of other repair systems in producing deletions, and the possibility that the same deletions can be caused by more than one pathway. The goals of this application are to test specific questions arising from this hypothesis. The specific objectives are.(1)Obtain a better understanding of how the asymmetric duplication of a direct repeat can increase deletion frequency.(2)Are mUtagen-induced deletions ~ntargeted?.(3)To what extent does RecA-mediated recombination contribute to deletion frequency increase?.(4)Attempt to isolate deletion intermediates. Qbjective l-The observation that asymmetric duplication of a d~rect repeat has a large stimulatory effect on deletions was totally unexpected and is important in that it indicates that (a)deletion intermediates can be stabilized in more ways than by hairpin formation at palindromic sequences, (b) sequences bracketed by direct repeats which are normally stable,can become deletion-prone by the tandem duplication of one repeat.The model shown in fig.7 of the manuscript fits our results, but we need more information to understand how asymmetric duplications of terminal repeats facilitate the occurrence of deletions..We have constructed new plasmids showing deletion frequencies fom one to four orders of magnitude higher than pRS4 (see manuscript,fig. 1 and table I) by adding more EcoRl linkers (monomers,dimers,etc.) to the~70bp EcoR~ insert of pRS1 (manuscript,fig.1).With these plasmids we propose to c~nduct studies along the same lines as those described in the enclosed manuscript,namely sequence the termini of the inserts atthe EcoRl site and correlatethese with deletion rates.Using this direct approach we expect (a) to confirm and extend our original observation that tandem duplication of a terminal repeat stimulates deletions,(b) determine whether increasing asymmetry increases deletion frequency,(c)expand our data base about this particular group of deletions to develop a better and testable model. Objective 2-The results shown in Appendix I lend plausibility to the idea that mutagen-induced deletions can be untargeted and caused by an induced SOS response.The next step is to test this idea using the directed mutagenesis method of Livneh (14) .With pBR325-derived plasmids we can mutagenize the insert' prior to cloning in into the plasmid,or the plasmid prior to the insertion of the fragment at the EcoR1 site: targeted mutations should result only from mutagenizing the insert.Since deletion rates can be readily determined,quantitative data bearing on how different parameters(type and dose of mutagen,extent of SOS induction,etc.) affect deletion frequency are easy to obtain.The use of these plasmids together with mutants for various genes in the SOS response allows great flexibility in designing experiments which should yield clear cut answers.Strains are available in which SOS mutations are in the same background with fusions of Mu d(Ap lac) in SOS controlled genes, so the extent of SOS induction can be determined from the level of ~-galactosidase produced (15,16)).A 40000198

typical experiment would consist of mutagenizing one of the plasmid components as described above,reconstituting the p~3m~mid by ligation, introducing it by transformation into desired strains and determining the frequency of Cmr revertants. Objective 3-Some results in Appendix I suggest that RecA- mediated recombination can contribute to the enhancement of deletion frequency in SOS-constitutive strains.We propose to conduct experiments comparing the effects of various m~tant alleles of recA on deletions,taking advantage of the large collection of protease constitutive (Prtc) recA mutants iso3~ted by Tessman and Peterson (15,16).These mutants have been extensively characterized for various functions of RecA an~many of them are recombinase positive (Prtc Rec+) while othe~ are recombinase negative (Prtc Rec-).These should be extremely useful to settle the question of whether RecA recombinase can stim~/ate deletions on plasmids. If this is the case,we should observe a large stimulation of deletion frequency in the presence of the Rec+ and no stimulation in the presence of Rec- alleles,while if RecA recombinase plays~no role in this process there will be no difference between Rec+ and Rec-in this respect. Objective 4-A major tenet of misalignement mutagenes~els for deletions' is that these occur through the resolution of unstable misaligned intermediates, yet the presence of such intermediates has not been demonstrated.Since deletions are rare events occurring at frequencies between 10-8-10-9 ,it is possible that deletion intermediates are much too rare to be recuwered from cells, as reported by Sinden et al (17).These anthors attempted to trap these structures through the formation of intrastrand T-T crosslinks by psoralen plus UV light (I~/VA) treatment, but ~sed 9 plasmid with a low deletion frequency.We have constructed.new plasmids which show deletion frequencie~ of the order of 10-3, and we would like to attempt the same experiments of Sinden et al (17) using these plasmids. The demonstration of misaligned deletion intermediates would go a long way in establishing the validity of misalignement mutagenesis models for deletions. These experiments are important and should be tried. 4 40000197

REFERENCES l-T.P.Dryja,J.M.Rapaport,J.M.Joyce and R.A.Petersen (1986).Proc.Natl.Acad.Sci.USA 83.:7391-7394. 2-S.H.Orkin,D.S.Goldman and S.E.Sallan(1984) Nature 30--9:172-1743 3- C.Theillet,R.Lidereau,C.Escot,P.Hutzell,M.Brunet,J.Gest,. J.Schlom and R.Callahan (1986). Cancer Research 4--6:4776-4781. 4-A.M.Albertini,M.Hofer,M.P.Calos and J.H.Miller (1982).Celi 29:319-328. 5-B.W.Glickman and L.S.Ripley (1984)Proc.Natl.A~ad. Sci.US.81:512- 516. 6-J.W.Little and DoW.Mount (1982) Cell 29:11-22. 7-J.Cairns (1981) Nature 28__~9:353-357o 8-H.Echois (1981) Cell 2__5:1-2. 9-E.Balbinder,D.Kerry and C.I.Reich (1983) Mut.Res.l12:147-168. 10-J.H.Miller and K.B.Low (1984) Cell 3_/7:675-682. ll-D.T.Weaver and M.L.DePamphilis (1984) J.Mol.Biol.180:961-986. 12-R.J.LaDuca,P.J.Fay,C.Chuang,C.S.McHenryand R.A.Bambara (1983) Biochemistry 2__2:5177-5187. 13-I.Baumel,T.F.Meyer and K.Geider (1984) Eur.J.Biochem. 13_~8:247- 251. 14-Z.Livneh (1986) Proc..Natl.Acad~Sc~.USA 83:4599-4603. 15-E.S.Tessman and P.K.Peterson (1985) J.Bact. 16--3:677-687. 16-E.S.Tessman and P.K.Peterson (1985) J.Bact. 163:688-695. 17-R.R.Sinden,S.S.Broyles and D.E.Petijohn (1983) Acad.Sci.USA 80:1797-1801 18-S.E.Luria and M.Delbr~ck (1943) Genetics 2--0:491-511. 19-D.E.Lea and C.A.Coulson (1949) J.of Genetics 49:264-285. 20-E.M.Witkin,J.O.McCall,R.M.Volkert and I.Wermundsen (1982) Mol.Gen.Genet.185:43-50. 21-J.H.Krueger,S.J.Elledge and G.C.Walker (1983) J.Bact.15__2:1368- 1378. 22-N.J.Sargentini and K.C.Smith (1984) Mut.Res.128:l-9. 40000198

Appendix I-Results being prepared for publication. Deletion rates were calculated by the method of Luria and Delbruck(18) as modified by Lea and Coulson (19).The values represent the frequency of deletion per cell per generation and were obtained from the expression R=m/N where m is the mean number of mutations per sample and N is the final viable cell count perl Iculture.T~e value of m was calculated from the median of the -distribution of the number of Cmr revertants per culture, in a total sample of 90 cultures for each experiment.N was always obtained from direct viable counts.Relative deletion rates are 9xpressed in relation to EB312 1 (#7) taked as a standard (column A), or in relation to the lexA+umuC+ strain in each group of three strains with the same ;ecA allele (column B).Deletion frequencies represent the number of Cmr revertants per number of cells plated on selective agar, and were calculated from the average number of revertants in the entire sample of 90 cultures (column c) or from overnight broth cultures (column D). For columns C and D, relative frequencies are shown in parentheses.Point mutations in each strain were monitored through reversion of t~E+ to ~ on enriched minimal medium in the absence of adenine (SEM) or its presence (SEMA). Adenine stimulates reversion frequency in strains constitutive for the SOS response (20). The different alleles of recA, lexA, and umuC have the following properties:recA730 results in constitutive expression of the SOS response (high protease activity) as well as enhanced recombination (high recombinase activity) (20);.lexA71::Tn5 (Def) eliminates LexA activity and results in constitutive expression of the SOS response (21); umuC36 is a leaky mutation in umuC which decreases the mutagenic response (22). The results summarized in the table show (a) that derepression of SOS increases deletion frequency ( compare #1&#2, #4 & #5, and #7 & #8), (b) umuC36 lowers it (compare #i &#3,and #4 & #6),(c)RecA-mediated recombination may contribute to cause these deletions (compare #2 &#5). 40000199

OOZO000~ 000 -F ,, -F -}- -, -I.-