mutation spectrum of heat-induced abasic sites on a single

THE JOURNAL OF BIOLOGICAL CHEMISTRY 0 1992 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 267, No. 27, Issue of September 25, pp. 19718-19723.1992 Printed in U. S.A.

Mutation Spectrum of Heat-induced Abasic Sites on a Single-stranded Shuttle Vector Replicated in Mammalian Cells*

(Received for publication, February 4, 1992)

Januario B. Cabral Neto, Alain GentilS, Rosa E. Caseira Cabral, and Alain Sarasin From the Laboratory of Molecular Genetics, UPR 42. Znstitut de Recherches Scientifiques sur le Cancer, BP8-94801 Villejuii France

The mutational potency of apuriniclapyrimidinic (AP) sites induced by heat-treatment under acidic conditions has been studied in mammalian cells. Abasic sites were induced on a single-stranded DNA shuttle vector carrying the supF tRNA gene, eliminating, therefore, any ambiguity concerning the damaged strand. This vector was able to replicate both in mammalian cells and in bacteria where the mutations induced in animal cells on the supF tRNA gene were screened by the white/blue &galactosidase assay in the presence of isopropyl- l-thio-/3-D-galactopyranoside and 5-bromo-4-chloro-3-indoyl-~-~-galactoside. All white colonies contained plasmid with a mutation on the target gene which was directly sequenced. Our results show that one A P site was induced122 min of heating as measured by sensitivity of DNA to alkali denaturation or treatment with the AP-endonuclease activity of the FPG protein (Fapy-DNA glycosylase). Putative AP sites decrease survival of the plasmid with a lethal hit of one A P sitelsingle-stranded molecule. Mutation frequency was increased by a factor of ap- proximately six after 2 h at 70 “C. Most of the induced mutations were point mutations not distributed at random and clustered in the gene region which will give rise to the mature tRNA. Mutations were abolished by treatments that eliminated AP sites such as alkali treatment or incubation with the Fapy-DNA glycosylase protein. Under our experimental conditions, when only single mutations were taken into account, the order of base insertion opposite A P sites was G > A > T > C.

Abasic (AP)’ sites are formed when the bonds connecting the purine or pyrimidine bases to the deoxyribose sugars are cleaved, leaving the DNA phosphodiester backbone intact. This kind of DNA alteration appears in bacterial or mammalian genomes with a high frequency by either spontaneous hydrolysis of the N-glycosylic bond or following treatments with physical or chemical agents (1-5). These agents usually produce DNA adducts which can lead with high probability to loss of the damaged base. It has consequently been sug-

* This work was supported by grants from the Association pour la Recherche sur le Cancer (Villejuif, France), the Fondation pour la Recherche Medicale (Paris, France), Commission of the European Communities Contract B17-0034, Brussels, Belgium, and The Bra- zilian National Research Council (CNPq). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “aduertkement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ T o whom correspondence should be addressed. Tel.: 33-1- 47264658; Fax: 33-1-47268836.

The abbreviations used are: AP, abasic sites; FPG, Fapy-DNA glycosylase; X-Gal, 5-bromo-4-chloro-3-indoyl-~-~-galactoside.

gested that AP sites may be intermediates in chemical mutagenesis (6, 7).

Heating DNA under acidic conditions induces the formation of abasic sites. Under these conditions depurination is severalfold higher than depyrimidination or cytidine deamination (1, 5). Nevertheless, deamination of cytosine residues, cleavage of pyrimidine glycosyl bonds, or destruction of deoxyribose residues may correspond to minor but possible types of heat-induced DNA damages. Heat-induced abasic sites have been shown to be mutagenic in procaryotes and eucary- otes (8-11). Indeed, introduction of abasic sites into single- stranded @X174 DNA decreases survival and increases mutation frequency of the phage progeny in “SOS-induced” Escherichia coli. The same phenomenon was observed in non- SOS-induced mammalian cells (i.e. in non-UV pre-irradiated cells) using SV40 as a probe (11).

The analysis of spontaneous and induced mutations in mammalian cells has been highly facilitated by using shuttle vectors. These plasmids may be replicated and eventually repaired by mammalian cellular enzymes. Mutations produced in mammalian cells are screened and analyzed quickly in bacteria (12, 13). Mutation spectra arising during passage of damaged double-stranded shuttle vectors in mammalian cells have been recently reported (14-21). However, using a double- stranded shuttle DNA vector as a probe does not allow one to know precisely which base was really mutated in a single base pair, due to the ambiguity concerning the damaged strand. The use of a single-stranded shuttle vector provides a solution to this problem (22).

Here we have determined the effect of heat treatment under acidic conditions leading to the formation of abasic sites on a single-stranded SV4O-based shuttle vector. Our mutation target was the E. coli tyrosine suppressor tRNA gene (sup F ) which allows us to easily screen for mutants using the p- galactosidase assay (20). Our results showed that AP sites were indeed mutagenic giving rise to point mutations and that no preferential insertion of adenine opposite the abasic site was found in our experimental conditions.

MATERIALS AND METHODS

Cells and Bacteria-Monkey COS7 cells were grown as monolayers in Dulbecco’s modified Eagle’s medium supplemented with 7% fetal calf serum and antibiotics. The indicator bacterial strain employed as plasmid host for mutagenesis assays was Escherichia coli MBM 7070 which genotype was: F-lacZ(Am)CA7020, lac Y1 hsdR hsdM A(araABC-leu)7679 galU galK rpsL thi. In the presence of isopropyl- p-D thiogalactoside, an inducer of the lac operon, and of 5-bromo-4- chloro-3-indoyl-~-~-galactoside (X-Gal), an artificial substrate for p- galactoside, MBM 7070 strain, carrying an active supF tRNA gene forms blue colonies. Mutated supF tRNA gene will render the colonies white or occasionally light blue. The JM105 E. coli strain permissive for filamentous phages was used for production of single-stranded DNA.

Vector Design and Single-stranded DNA Preparation-The genetic map of the pCF3A plasmid used in this work is represented in Fig. 1.

19718

Mutagenesis Spectrum of Abasic Sites 19719 - fi ori

pCF3A

5207 bp SV40 late genes

II ori

FIG. 1. Genetic map of the pCF3A vector. The vector contains SV40 and P replication origins (SV40 ori and P ori) allowing replication, respectively, in mammalian cells and in bacteria. SV40 late genes allow encapsidation of the plasmid as pseudovirus in COS7 cells. The CAT gene regulated by the tryptophan promoter (pTrp) confers chloramphenicol resistance. The origin of phage f l ( f l ori) permits the production of the vector in a single-stranded form in bacteria. The supF tRNA gene was used as target sequence.

I t is derived from the general structure of a shuttle vector previously described (22). It can be selected in bacteria by chloramphenicol resistance, due to the expression of the chloramphenicol acetyltrans- ferase (cat) gene. It carries the supF tRNA gene and the bacterial origin of replication from plasmid aAN13. The vector contains the origin of replication and the late genes of the SV40 virus. It can replicate in mammalian cells which constitutively produce the SV40 large T antigen such as the COS7 cells. It also has the origin of replication from fl phage, which allows the plasmid to enter the phage replication mode in permissive bacteria after infection with a helper phage. The single-stranded DNA was prepared in JM105 strain by the procedure described in Analects (Pharmacia), using M13K07 as helper phage.

DNA Treatment and A P Site Quantification-Abasic sites were introduced into the single-stranded vector by heating at 70 “C for various periods of time in a 25 mM citrate, 250 mM KC1 buffer at pH 4.8. They were quantified: (a) by alkaline treatment: control or heat- treated DNA was incubated for 16 h at room temperature in a denaturation buffer at pH 12.1 as described previously (23); ( b ) by AP-endonuclease assay: ethanol-precipitated control or heat-treated DNA was dissolved in reaction buffer (24) and incubated with the FPG protein (a generous gift from Dr. J. Laval, UA 147, Villejuif, France) for 1 h at 37 ‘C.

After treatment with either alkaline buffer or FPG protein, DNA was run in a 1% agarose gel. Relative proportions of single-stranded linear (nicked) and circular (unnicked) DNA were determined by scanning photographs of the ethidium bromide gels with a Chromos- can 3 densitometer (Joyce-Loebl). The ratio between the two bands allows one to calculate the number of abasic sites/molecule using the Poisson distribution.

Mutational Assay-Subconfluent COS7 cells were transfected with plasmid DNA (1 pg/90-mm Petri dish) by the DEAE-dextran method (25). After 3 days, cells were harvested for extrachromosomal DNA extraction using an adaptation of the procedure described by Stary and Sarasin (26). DNA preparation was then incubated with DpnI endonuclease restriction enzyme at 37 “C for 1 h in order to eliminate any unreplicated DNA (input DNA) and shuttled into MBM7070 E. coli strain by electroporation (Eurogentec SA) as described by Miller et al. (27). Transformants were plated on LB medium containing 34 Fg/ml of chloramphenicol, 0.08 mg/ml of X-Gal, and 0.1 mM of isopropyl-P-D-thio-galactoside, to determine the total yield of trans- formed colonies and identify colonies carrying mutated plasmids. The supF tRNA gene present in the vector is used as target gene for screening the mutants. Colonies bearing plasmids with mutations inactivating the suppressor tRNA (white or light blue colonies) were restreaked three times for confirmation of the phenotype, and plasmid DNA was prepared for further analysis by a simplified alkaline lysis “mini prep” method. The tRNA region of the vector from mutant colonies was sequenced by the chain elongation terminator method (28).

DNA Synthesis Termination Site Assay-Location and quantification of sites of heat-induced lesions on the supF gene were made using a modified in vitro DNA polymerase-stop assay (29). 1 pg of ethanol-precipitated control or depurinated DNA was hybridized with the sequencing primer and incubated 5 min at room temperature with 10 mM dithiothreitol, 0.5 p M [(u-~’S]~ATP, 0.75 p~ dGTP, dCTP, dTTP, and 3 units of T4 DNA polymerase. 10 mM of the four deoxynucleotide triphosphates were added to the labeling reaction described above, and the incubation was continued at 37 “C for 5 min. Reactions were stopped by adding 4 pl of a stop solution (95% formamide, 20 mM EDTA, 0.05% bromphenol blue, and 0.05% xylene cyanol). Samples were heated to 80 “C immediately before loading on a 6% polyacrylamide gel. The supF DNA sequence was carried out as described above and run as sequencing marker in the same gel.

RESULTS

Abasic sites were generated in the single-stranded form of the shuttle vector pCF3A by heating at 70 “C under pH 4.8 conditions. Their number was measured either by alkaline denaturation at pH 12.1 or by treatment with the AP-endonuclease activity of the E. coli FPG protein (24). Fig. 2 shows the number of breaks/genome after treatment with either the denaturation buffer or the FPG protein as a function of the time of heat treatment. These two treatments gave the same result. According to the linear regression (correlation coeffi- cient, 0.98) and the Poisson distribution, one abasic site was formed in 22 min of heating. As observed in Fig. 2, almost no breaks were directly introduced by the heating treatment alone.

Plasmid survival decreases as a function of heat treatment duration as shown in Fig. 3. From the curve one can calculate that one lethal hit (37% survival) is produced every 22 min of heating, and this is correlated with the formation of one abasic site/DNA molecule. Treatment with alkali, which destroys abasic sites, or digestion with the FPG protein containing an associated AP-endonuclease activity does not affect the survival. This is expected since a break of the phosphodiester

A,

0 20 4 0 6 0 MIN. AT 70°C

FIG. 2. Quantification of abasic sites. Abasic sites were quantified by the number of breaks/genome introduced on heated DNA following treatment with either alkali-denaturation buffer (0) or with the FPG protein (A). Linear regression used to draw the straight line was calculated from experimental values we obtained with both treatments. The number of breaks induced by heating alone was measured at the same time (0).

0 30 6 0 00 120 MIN AT 70%

FIG. 3. Effect of either the FPG protein or alkali treatment on the survival of heated pCF3A DNA in COS cells. Heated- DNA was either left untreated (0) or treated with the FPG protein (A) or with a denaturation buffer at pH 12.1 (0) before transfection into COS cells. Survival was measured in bacteria as described under “Material and Methods.”

19720 Mutagenesis Spectrum of Abasic Sites

backbone at an AP site in a single-stranded circular molecule gives rise to a linear one which is not able to be replicated.

According to the length of the treatment, between lo4 and 80 X lo4 bacterial colonies were screened from 10 independent cell transfections for each time of treatment to calculate mutation frequency. As shown in Fig. 4, mutation frequency increases with the number of abasic sites introduced into the supF tRNA gene: the mutation frequency is proportional to the number of introduced abasic sites. This enhancement of the mutation yield was abolished when heat-treated DNA was further incubated with alkali buffer or digested with the FPG protein before cell transfection. Spontaneous mutation frequency varied from 2.4 to 4 x in these experiments.

Mutation frequency calculated for each class of template bases of the supF tRNA is shown in Fig. 5, as a function of the heating time. Spontaneous mutations were only observed at cytosine sites. In 6% of the cases, spontaneous deletions ranging from 2 to 80 base pairs were also observed, and no insertion was found (not shown). Heat treatment did not increase mutation frequency at C sites while it did a t adenine and chiefly at guanine residues. Thymine sites were never implicated in mutation. This representation confirms, therefore, the induction of mainly apurinic sites by heating the DNA under acidic conditions.

The specificity of induced base substitutions and the heating time to produce them are presented in Table I. Only independent mutants were taken into account (identical mutants isolated from the same cell transfection were excluded). 71 out of 105 mutations were located a t purine sites in the template strand. Of the 105 mutations, 73 (69%) were transversions and 32 (31%) were transitions. Six multiple mutations (two or more mutation events found within a unique mutated plasmid) were found (Fig. 6).

K

MIN. AT 70°C

FIG. 4. Mutation enhancement factor in the progeny of heat-treated pCF3A DNA recovered from COS cells. The enhancement factor was calculated from the mutation frequencies of heat-treated DNA (0), and of heat-treated DNA further incubated with either an alkali buffer at pH 12.1 (0) or the FPG protein (A), divided by the mutation frequencies of untreated DNA (2.4 X or control DNA treated by alkali or the FPG protein (4 X

. z 10 x > Y s w 7

U U

z

$ 6

P 4 G 2 2

0 0 15 30 60 120

MIN. AT 70°C

FIG. 5. Frequency and distribution of the mutations induced by heat treatment for each type of base in the treated template. Mutation frequency of base substitutions was calculated as a function of heating time on each template base: cytosine (O), adenine (a), and guanine (0) along the supF tRNA gene. No mutation was found at thymine sites.

Fig. 6 presents the treated single-stranded DNA sequence of the supF gene. This figure illustrates all base substitutions recovered and shows the non-random distribution of base changes among the available sites within the target gene for both induced and spontaneous spectrum. Most of point mutations were clustered within the region which will give rise to the mature tRNA. The distribution of mutations/site observed within the supF tRNA gene sequence is quite different of that expected from the Poisson distribution. Given the 80 potential sites of detected mutations scored from published sequence changes within the supF tRNA gene (14-17,20,21, 30-32) seven or more mutations at a single site are considered to represent a hot spot ( p < 0.005) (33-34). Five hot spots, at positions 51, 52, 65, 75, and 111 were therefore observed with heat-treated DNA and only one with control DNA at position 65. After treatment by either NaOH or FPG protein, 66 and 72% of the mutations for heat-treated and control DNA, respectively, were found at position 65. These mutations should not be due to AP sites.

Fig. 7 shows the sites of T4 DNA polymerase terminations in both untreated and heat-treated DNA, beside a sequencing ladder produced with untreated supF DNA. Abasic template was hybridized with our usual sequencing primer, elongated with the T4 DNA polymerase, and run on a 6% polyacrylamide gel as described under "Material and Methods." Dis- crete halts due to a block of the polymerization may be seen in the heat-treated DNA (lane 2 ) which do not exist for untreated DNA (lane 1) . These discrete bands show that termination sites are targeted at DNA lesions which have been created along the supF sequences. Intensity of these bands is quite uniform indicating therefore that no hot spots of lesions were created by heating the template. With the sequencing ladder, termination sites were located precisely. Since it has been shown that DNA polymerases usually stop one nucleotide before abasic sites (35), stops clearly occur on the treated template only at adenine and guanine sites as indicated in Fig. 7. Similar results were obtained by using the Klenow fragment of the E. coli DNA polymerase I instead of the T4 DNA polymerase (not shown).

DISCUSSION

We have shown that heating a single-stranded DNA under acidic conditions induces easily detectable alterations which can be correlated with an increased mutation frequency after replication in mammalian cells. This treatment has been reported to induce mainly apurinic sites although other chemical modifications may occur a t much lower frequency, chiefly cytosine deamination. It has been reported (36, 37) under similar conditions the occurrence of deamination of free cytosine and dCMP. However, according to Drake and Baltz (38) only one cytosine deamination is produced for lo3 depu- rinations under our experimental conditions. In our experiments a significant participation of cytosine in the mutation spectrum obtained is observed. However, these point mutations at cytosine level do not seem to be induced by the treatment since, first, they do not increase in frequency with increasing times of heating (Fig. 5 ) and, second, they do not disappear after treatment with either NaOH or the AP endonuclease activity of the FPG protein. Moreover, single base substitutions at cytosine are of the same nature in all conditions: heated DNA, control, or heated DNA treated with NaOH or digested with the FPG protein (Fig. 6). We have previously described, using a single-stranded vector replicated in mammalian cells, that spontaneous mutations were only located at cytosine sites (30). Consequently, mutations located at cytosine sites are not due to induced abasic sites and can be considered as arising from spontaneous origin.

Mutagenesis Spectrum of Abasic Sites TABLE I

Location and specificity of base substitution mutations Insertion

?Elt Colonies No. of point No. of single Mutations Opposite No. of mutations at a given site mutation mutation sites site in the SupF gene"

Base N"

A C 1 1 (3126

A 3 1 (101); 1 (65)T'; 1 (66IT

T 8 4 (65); 1 (9215; 1 (92); 1 (98); 1 (101)

A 1 1 (6)'

T 8 2 (50)T; 2 (51)T; 2 (52IT; 2 (85)

15 50834 28 16 C C 4 1 (57); 1 (98); 1 (102); 1 (106)

G G 3 1 (51)5; 2 (75)

30 44034 32

60 22967 25

120 12301 20

24

23

16

A A 4 1 (76); 1 (103); 1 (104); l (113) C 1 1 (76)

A 4 1 (46)4; 1 (65)4; 2 (98) C C 3 1 (55)'; 1 (98); 1 (106)4

T 4 2 (65); 1 (98); 1 (101)

A 6 3 (75); 1 (84)4; 2 (110)

T 3 1 (4)'; 2 (52)

A 1 1 (104) A C 1 1 (103)

G 2 2 (76)

A 2 2 (65)

T 2 1 (65); 1 (106)

A 6 1 (75); 1 (81); 2 (97); 2 (110) G G 7 3 (51); 1 (85); 3 (111)

T 3 1 (17)3; 1 (50); 1 (52)

A A 1 1 (113)T

G G 7 1 (51)T; 1 (52)T; 1 (97); 1 (110); 3 (111)

C C 1 1 (102)~

C 2 1 (80); 1 (103)

C A 1 1 (65) T 2 1 (65); 1 (102)

A 5 l (51 ) ; l (75 ) ; 1 (110)'j; 1 (120)6; l (120) G G 6 3 (97); 2 (111); 1 (114)T

T 3 2 (52); 1 (91)

A 6 (46%) A C 5 (39%)

G 2 (15%)

A 10 (29%) Total 130184 105 79 C C 8 (24%)

T 16 (47%)

A G G

18 (31%) 23 (40%)

T 17 (29%) Numbers in parenthesis correspond to nucleotide numbers in Fig. 6. Small numbers in superscript refer to multiple mutation events.

' T in superscript refers to tandem or triplet mutation events.

The mutation frequency increases with the heating time at adenine and chiefly guanine sites (Fig. 5) and is reverted to spontaneous level by alkali or FPG treatments (Fig. 4). Mu- tagenesis observed at these sites is therefore due to apurinic sites, and it seems unlikely that other lesions may contribute significantly to the mutation spectrum.

The mutations observed after heat treatment of the single- stranded DNA of the vector were essentially single base substitutions with a great participation of purines, mainly guanine as target sites (Fig. 5). Thymine was never implicated as a premutagenic site. As described by Kunkel (IO), a major-

ity of transversions have been observed, but no preferential incorporation of adenine was found in our experiments. It has been reported that a preferential insertion of adenine opposite abasic sites replicated either in bacteria or using purified polymerase in vitro, including the avian myeloblastosis virus DNA polymerase (10, 35, 39-42). In our experiments such a phenomenon did not occur, even if we consider that mutations occurring at cytosine sites were not due to abasic sites induced by the treatment as suggested above. Numerous hypothesis have been proposed to explain the preferential adenine insertion opposite abasic sites during replication of these non-

19722 Mutagenesis Spectrum of Abasic Sites

TA 5CA

CA CA CA

$A1 T 2 A 3 T Bash 4 M T . . 3 ' G T C G C C G C G C A G T A A A C T A T A C T A C G C G G G G C G A A G G G G C T C G

-30 -20 - 1 0 0 1 0 2 0 3 0 4 0 5 0

C C C

TTTTA T g C TC

GGAAAAA T CT A TC TT G A TC

AAAAAAAAA T CA

AAA TC

TTT cc A TG T G ~ A TC T CT 4 A A CA AGGT G 4 CC T 6T

G TT TG GT TC AA CG AATT G TC x T C C G G T T T C C C T C G T C T G A G A T T T A G A C G G C A G T A G C T G A A G T T T T C A G G C . . 5 '

TA A T GA G AAA TG A

TAGGA TT A AAAAAGG T A TTTGT T

60 70 80 90 1 0 0 1 1 0 1 2 0 1 3 0 1 4 0

""""""""""""-"~--"-- -b

FIG. 6. Spectrum of base substitutions induced on the supF tRNA gene by heat treatment. Heat-induced mutations are shown above the wild type sequence of the treated supF tRNA strand, and spontaneous mutations are displayed below. Tandem or triplet mutations are underlined. Mutations found in the same mutant (multiple mutations) are marked by identical numbers. Base substitutions in bold are from heat-treated or control DNA incubated with either alkaline denaturation buffer or the FPG protein prior to transfection of COS cells.

MATURE tRNA REGION FLANKING REGION

coding lesions, such as preferential dNTP binding on the polymerase, stabilization of dNTP by stacking interactions with the previous base inserted, editing bias of the proofread- ing activity, dNTP pool bias, dATP being considerably increased when DNA is damaged, nucleotide hydrophobicity, DNA structure at the AP site, etc. (for review, see Ref. 5 ) . From our results of transfection of mammalian cells with single-stranded DNA containing abasic sites, all these factors do not seem to play a major role, since whatever the template base, the three possible bases have been inserted opposite the abasic site when crude data are considered in their totality.

This result is somewhat surprising since we were expecting a preferential insertion of adenine opposite AP sites as has been reported in bacteria. The absence of preferential insertion of A opposite AP site in our experiments does not seem be due to a bias in the selection of mutants since the supF target has been shown to allow the recovery of all types of mutants. This target has been used extensively by several authors to study mutagenesis. Among the 140 base pairs of the mature tRNA sequence, 80 sites have been scored for which a single base mutation gives rise to a detectable phenotype (14-17, 20, 21, 30-32). Among these 80 sites, it is possible to phenotypically detect misincorporation of 50 A, 46 T, 31 C, and 25 G. If we consider the bases only inserted opposite purines, these numbers become A (29), G (19), T (18), and C (8). This target which detects the incorporation of all bases exhibits, however, the highest probability to recover misincorporation of A.

Since the probability of having more than one lesion on the supF target leading to more than one mutational event/

molecule is almost nil, tandem or multiple mutations could not be due to several AP sites on the same target. Indeed, an error prone replication during conversion of single-stranded DNA into double-stranded DNA in mammalian cells could be responsible for such multiple mutations as previously reported for single-stranded (30) or double-strandedvectors (43). Thus, it could be interesting to look only at the single changes at purines, eliminating therefore all tandem, triplet, and multiple mutations which could be due to something else than AP site. In fact, among the 53 single mutations observed a t purines, the misinsertion order is G (21), A (19), T (9), and C (4). This distribution is not a t random and not very different from what has been reported for bacteria. On the other hand, using an oligonucleotide containing a unique AP site chemically synthesized, inserted opposite A, C, G, or T in a double- stranded shuttle vector, we did not find a preferential adenine insertion (44, 45). However, separate examination of the hot spots reveals a different repartition of bases inserted in rela- tion to the hot spot considered. I t is therefore possible that a sequence effect or the secondary structure of the molecule may intervene, leading to the insertion of a preferential base, which may be any one of the four bases at a definite site.

In bacteria AP sites have been shown to be mutagenic only after "SOS" induction of the cells by UV light prior to trans- formation (8, 9). In mammalian cells mutagenesis occurs without such preconditioning treatment. These results are in agreement with the data we reported previously using either SV40 DNA (11) or a double-stranded shuttle vector carrying a unique AP site chemically synthesized (44, 45). However, we cannot exclude, as we previously suggested, the induction

Mutagenesis Spectrum of Abasic Sites 19723

1 2 T G C A

C C

T C

T 10 +A

T C

T +A +A

A G T

+G T T A

-c 3'

FIG. 7. Sites of termination of in vitro DNA synthesis in heat-treated pCF3A DNA. Polyacrylamide gel showing the sites of termination in the supF tRNA gene of the T4 DNA polymerase. Single-stranded DNA was either untreated (lane I ) or heated a t 70 "C under acidic conditions for 2 h (lane 2). T, G, C, and A refer to the wild type supF tRNA sequence run in the gel as sequencing ladder. Arrows indicate stops of synthesis due to heat-induced abasic sites not present in control DNA, taking into account that halts of the T4 polymerase occur one nucleotide before the abasic sites.

of some mutator functions in mammalian cells by the transfection of a damaged DNA (11) which could then explain the multiple mutations observed.

The mutation spectrum was not random as shown in Fig. 6. Some hot spots were found and, as most of the mutations detected, they were clustered in the mature tRNA region. This location agrees with the data reported by Kraemer and Seid- man (20) who reviewed mutagenesis induced by different chemicals and physical agents also using the supF tRNA gene as a target in shuttle vector plasmids. We can effectively notice that four out of five hotspots are located in the double- stranded regions of the hypothetical secondary structure of the supF tRNA gene and one at the anticodon position. These locations are not due to a bias in the detection of the mutants since mutations could be found in almost all parts of the gene. However, since lesions are induced at random along the supF tRNA gene, it is surprising that mutations are not induced at random. It is possible that the primary or the secondary structure of the supF could be involved in this phenomenon. It has to be recalled that AP sites should constitute a complete

block to DNA replication since one AP site corresponds to a lethal hit. A small percentage of abasic sites could, however, be bypassed by the polymerase due to specific sequences or structures. Such events could lead to the mutagenesis observed. The structural facilitation for the polymerizing com- plex to go through an A P site could explain the existence of mutational hot spots and the absence of correlation with DNA lesion spectrum.

Acknowledgments-We thank C. Madzak for help in the production of single-stranded DNA when starting these experiments and S. Boiteux and J. Laval (UA 147, Villejuif) for the generous gift of the FPG protein. Thanks are also due to Drs. T. Kunkel, L. Loeb, and P. C. Hanawalt for helpful criticisms in the preparation of the manuscript.

REFERENCES 1. Lindahl, T., and Nyberg, B. (1972) Biochemistry 11.3610-3618 2. Lindahl, T., and Annika, A. (1972) Biochemistry 11,3618-3623 3. Marelson. G. P.. and OConnor. P. J. (1973) Btochem. Btophys. Acta 331,

343-359 ' . .

4. Strauss, B., Scudiero, D., and Henderson, E. (1975) in Molecular Mecha- nisms for Repair of DNA (Hanawalt, P. C., and Setlow, R. B., eds), pp.

5. Loeb, L. A., and Preston, B. D. (1986) Annu. Reu. Genet. 2 0 , 201-230 13-24;Plenum, New York

6. Loeb, L. A. (1985) Cell 40,483-484 7. Shaauer. R. M.. Glickman. B. W.. and Loeb. L. A. (1982) Cancer Res. 42,

3480-3485 '

1773-1777

1-9

8. Shaaper, R. M., and Loeb, L. A. (1981) Proc. Natl. Acad. Sci U. S. A. 78 ,

9. Shaaper, R. M., Glickman, B. M., and Loeb, L. A. (1982) Mutat. Res. 106.

10. Kunkel, A. T. (1984) Proc. Natl. Acad. U. S. A. 81,1494-1498 11. Gent11 A. Margot, A., and Sarasm, A. (1984) Mutatton Res. 129, 141-147 12. DuBriide: R. B., and Calos, M. P. (1988) Mutagenesis 3 , l - 9 13. Sarasin, A. (1989) J. Photochem. Photobiol. 3, 143-155 14. Hauser, J., Seidman. M. M.. Sidur. K.. and Dixon, K. (1986) Mol. Cell. Biol.

fi.377-3A5 15. Bredberg, A., Kraemer, K. H., and Seidman, M. M. (1986) Proc. Natl. Acad.

16. Moraes, E. C., Keyse, S. M., and Tyrrell, R. M. (1990) Carcinogenesis 11,

-, - . . - - - Scl. U. S. A. 83,8273-8277

3Q9-909

17. Keyse, S. M., Amaudruz, F., and Tyrrel, R. M. (1988) Mol Cell Bid. 8, -\," -""

18. Stary, A., Menck, C. F. M., and Sarasin, A. (1992) Mutat. Res., in press

20. Kraemer, K. H., and Seidman, M. M. (1989) Mutat. Res. 220,61-72 19. Bourre, F., Renault, G., and Gentil, A. (1989) Mutat. Res. 220 , 107-113

21. Murlel, W. J., Lamb, J. R., and Lehmann, A. R. (1991) Mutat. Res 254 ,

5425-5431

22. Madzak, C., Menck, C. F. M., Armier, J., and Sarasin, A. (1989) J. Mol.

23. Shaper, N. L., and Grossman, L. (1980) Methods Enzymol. 65,216-224 24. Boiteux, S., O'Connor, T. R., Lederer, F., Gouyette, A., and Laval, J. (1990)

25. M ~ ~ ~ u t c h a m , J. H., and Pagano, J. S. (1968) J. Natl. Cancer Inst. 41,351-

119-123

Biol. 205,501-509

J . Biol. Chem. 265,3916-3922

26. 27.

28.

29. 30.

31.

32.

33.

34. 35. 36. 37. 38. 39. 40.

41.

42.

43.

44.

45.

Stary, A. and Sarasin, A. (1992) J. Gen. Virol. 7 3 , in press Miller, J. F., Dower, W. J., and Tompkins, L. S. (1988) Proc Natl. Acad.

Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Pm. Natl. Acad. Sci.

Moore, P., and Strauss, B. S. (1979) Nature 278,664-666 Madzak. C.. Cabral Neb . J. B.. Menck. C. F. M.. and Sarasin. A. (1992)

33 I

Sci. U. S. A. 85,856-860

U. S. A. 74,5463-5467

Mutat. R& 274,135-i45 '

Yagi, T., Sato, M., Tatsumi-Miyajima, J., and Takebe, H. (1992) Mutat.

Seetharam, S., Protic-Sabljic, M., Seidman, M. N., and Kraemer, K. H.

Gordon, A. J. E., Burns, P. A., and Glickman, B. W. (1990) Mutat. Res.

Res. 273,213-220

(1987) J. Clin. Inuest. 8 0 , 1613-1617

Sockett, R., S., and Hutchinson, F. (1991) Mol. Gen. Genet. 227,252-259 Sagher, D., and Strauss, B., (1983) Biochemistry 22,4518-4526 Shapiro, R., and Klein, R. S. (1966) Biochemistry 6,2358-2362 Lindahl, T., and Nyberg, B. (1974) Biochemistry 13,3405-3410 Drake, J. W., and Baltz, R. H. (1976) Annu. Reo. Biochem., 45 , l l -38 Boiteux, S., and Laval, J. (1982) Biochemistry 21 , 6746-6751 Kunkel, T. A,, Schaaper, R. M., and Loeb, L. A. (1983) Biochemistry 22 ,

Randall, S. K., Eritja, R., Kaplan, B. E., Petruska, J., and Goodman, M. F. 2378-2384

Shaaper, R. M., Kunkel, T. A,, and Laeb, L. A. (1983) Pm. Natl. Acad. (1987) J. Biol. Chem. 262,6864-6870

Seidman, M. M., Bredberg, A., Seetharam, S., and Kraemer, K. H. (1987) Sci. (1. S. A. 80,487-491

Gentil, A., Renault, G., Madzak, C., Margot, A,, Cabral Neto, J. B., Vasseur, Proc. Natl. Acad. Sei. U. S. A. 84,4944-4948

J. J., Rayner, B., Imbach, J. L., and Sarasin, A. (1990) B k h e m . Biophys. Res. Comm 173,704-710

Gentil, A., Cabral Neto, J. B., Mariage-Samson, R:, Margot, A., Imbach, J. L., Rayner, B., and Sarasin, A. (1992) J. MOL Bud., In press

223,95-103

mutation spectrum of heat-induced abasic sites on a single

Documents