The use of exogeneous DNA probes, which replicate extrachromosomally, is proposed in order to study spontaneous and induced mutagenesis in mammalian cells. Simian virus 40 has already proved to be very useful, since it has provided much important information in this field. Recently, several shuttle vectors have been designed for this purpose; however, it seems that these molecules have high spontaneous mutation frequencies when replicating in mammalian cells. We have developed new alternative systems, such as Epstein-Barr virus-based shuttle vectors that can be episomally maintained in human cells. Furthermore, we have constructed packageable shuttle vectors, which appear to be stable in the host cell and thus suitable for analysis of mutagenesis.

Genetic instability in higher organisms seems to be correlated with several deleterious physiological phenomena such as carcinogenesis and aging. The processes involved in this instability, either gene rearrangements or point mutations, are not known. Although they may occur ‘spontaneously’, the presence of induced DNA lesions, as a result of either a metabolic product or an external agent, may play an important role in the origin of genome alterations (Smith & Sargentini, 1985). Mutations can activate some cellular oncogenes (Cooper, 1984) leading to dramatic effects on the regulation of cellular growth. Recent experimental data provide demonstrations that such mutations in oncogenes can be induced by DNA-damaging agents (Sukumar et al. 1983; Guerrero et al. 1984). The relationship between unrepaired genomic lesions, mutations and carcinogenesis is also evident in the classical human genetic disease xeroderma pigmentosum. The absence of excision repair of some classes of DNA lesions in cells from these patients is correlated with increased mutagenesis and with a cancer-prone phenotype (Cleaver & Bootsma, 1975).

The analysis of induced genomic mutations in mammalian cells has been possible due to the existence of a few cell lines in which the mutated phenotype can be selected with appropriate media (Thacker, 1985). However, these analyses present technical difficulties, are time-consuming and yield very limited information. Hence, the use of small and easily manipulatable DNA probes, which would depend on cellular enzymic machinery for replication and repair, is desirable. These requirements can be fulfilled by certain animal viruses, such as simian virus 40 (SV40), and shuttle vectors. Lesions can be induced in the DNA molecules in vitro by treatment with DNA-damaging agents. Subsequently, the damaged probes are introduced into the recipient cells, by infection (virus) or transfection (naked DNA), where they will be repaired and/or mutagenized. A suitable mutation screening permits isolation of selected mutants, which may then be genetically mapped and characterized at the molecular level. The mutation sites can be correlated with the presence of putative lesions and thus the mechanisms by which a DNA lesion is processed into a mutational event can be approached. An additional advantage of this system is that the DNA probe and the host cell genome can be independently damaged, which permits a search for inducible SOS responses in mammalian cells (Sarasin, 1985).

SV40 is in many ways ideally suited as a probe for studying mutagenesis in mammalian cells. SV40 is a papovavirus with a supercoiled double-stranded DNA genome (5243 base-pairs) entirely dependent on the host cell machinery for replication as well as for DNA repair. Few proteins are encoded by the virus. The early transcription genes give rise to two mRNAs, which are translated into two proteins, the small t-antigen and the large T-antigen. This latter binds near the origin of replication and is the only virus-encoded protein needed to initiate DNA replication. Late transcription gives rise to the messenger RNA of the three capsid proteins, VP1, VP2 and VP3, and a fourth, agnoprotein, also involved in encapsidation. No enzyme is encoded for DNA replication or DNA repair. Moreover, SV40 DNA exists in vivo as a minichromosome resembling the chromatin organization of mammalian cells. One of the advantages in using such a viral probe is the facility to treat in vitro SV40 either as a virion or as DNA (when the compounds to be tested may react with the capsid proteins), and to treat the host cell separately.

Our mutation assay is based upon the reversion of a temperature-sensitive growth phenotype at the restrictive temperature of 41 °C to a wild-type growth phenotype. Two temperature-sensitive mutants are used, either the tsA58 SV40 mutant, unable to initiate DNA replication at 41 °C, or the tsB 201 late mutant, unable to produce virions at the restrictive temperature (Bourre et al. 1983).

Using the SV40 as probe, Sarasin & Hanawalt (1978) showed that a recovery pathway was acting in monkey cells, able to enhance the survival of the viral progeny from ultraviolet (u.v.)-irradiated virus, when cells were either u.v.-irradiated or treated with a chemical carcinogen prior to infection. This new recovery pathway has been shown to be error-prone since the mutation assay revealed an increased mutagenesis in the viral progeny of u.v.-irradiated virus grown in u.v.-irradiated cells (Sarasin & Benoit, 1986). Similar results have been obtained for human cells (Gentil et al. 1985). u.v.-induced revertants of the tsA58 mutant were isolated and their DNAs were analysed. The mutation responsible for the phenotype reversion, which leads to an active T-antigen at the restrictive temperature, was mapped by the marker rescue technique. DNAs of the revertants were digested by restriction enzymes, and the fragments obtained were hybridized with single-stranded tsA58 DNA. The restriction fragments that complement the tsA58 sequence carry the mutation. The nucleotide sequencing of these fragments was then carried out (Bourre & Sarasin, 1983). Our results showed that in all cases the tsA58 sequence was still present and that the reversion was always the result of a base substitution, changing one amino acid to another. All of these mutations were ‘targeted’, that is to say opposite a putative u.v. lesion, pyrimidine dimer or pyrimidine-(6-4)-pyrimidone.

As described above, one of the advantages of the system is the possibility of treating the DNA in vitro and therefore of quantifying precisely the number and type of lesions induced before transfection into permissive cells. This method has been used successfully to study the mutagenicity of u.v. lesions, abasic sites, and of acetoxy-acetylaminofluorene treatment (Bourre et al. 1983; Gentil et al. 1982). For u.v. treatment, identical viral data were obtained using DNA and using SV40 virion, except for a slight protective effect of the capsid proteins. When DNA is heated at 70°C under acidic conditions, abasic sites, chiefly apurinic sites (AP sites), are produced in SV40. One AP site is formed per 15 min of heating per SV40 genome (Gentil et al. 1984). This number is determined easily since these AP sites are alkali-labile and since AP endonucleases are available. We have shown that AP sites are very mutagenic for monkey cells and that they strongly decrease survival of the viral progeny with a lethal hit of three AP sites per SV40 genome. The use of SV40 DNA enabled us to compare the lethality and mutation efficiency of different kinds of lesion. Indeed, we showed that a great difference exists in cell killing potency between u.v. lesions, AP sites and acetylaminofluorene adducts. Four AP sites are as efficient in cell killing as 50 pyrimidine dimers or 200 acetylaminofluorene adducts. At these lethality levels similar mutation frequencies were obtained (Bourre et al. 1983).

These values, however, are the final result of many poorly understood phenomena that occur in the cell, including processes other than DNA metabolism. On this point we note that one restriction in the method we have used is that transfection of mammalian cells, itself, may well induce unusual phenomena. This is of relevance in the context of inducible repair. Indeed, one interpretation of our data is that AP sites are mutagenic per se, but an alternative interpretation is that mutator functions acting on AP sites may be induced by the transfection protocol. We have no data that bear directly on this possibility but we make the comparison with Shaaper & Loeb (1981) who, using ϕX174 bacteriophage, found that AP sites were mutagenic in bacteria only after SOS-induction of the host cells by u.v. irradiation. The molecular analysis of the mutations induced by AP sites and the carcinogen acetoxy-acetylaminofluorene is under investigation using the same experimental protocol as described above.

In conclusion, while it is valid to comment that this protocol is concerned with how viral DNA rather than chromosomal DNA is processed, the use of SV40 as a probe gives information that would not be accessible using other methods, given the complexity of a mammalian cell.

As part of other projects in our laboratory we have been investigating the potential of shuttle vectors for improving the efficiency of introduction, expression and rescue of genetic loci that are of interest in mammalian cells, particularly those from DNA repair-deficient and cancer-prone genetic syndromes. In addition such a vector system may facilitate the analysis of mutagenesis similar to the viral system described above but with less effort, more rapidly and with greatly improved logistics (Calos et al. 1983; Razzaque et al. 1983). This convenience arises from the ability to manipulate the vectors in mammalian cells but subsequently rescue and analyse them in bacteria.

In the first instance, we have inserted the replication origins or entire early-gene regions of monkey and human papovaviruses into a small cosmid (see Fig. 1A). The mutation locus we have chosen to insert in the vectors is lacZ′, which may be scored easily in appropriate bacterial hosts. These vectors replicate transiently in monkey COS-7 cells (Gluzman, 1981) and in human cell lines, including a xeroderma pigmentosum strain XP4PA-SVwt (group C), transformed with an origin-defective SV40 recombinant, because these cells provide T-antigen in trans (Daya-Grosjean et al. 1987). Two to three days following transfection with the vectors, plasmid that has replicated may be rescued in bacteria. In agreement with the data from other laboratories, the spontaneous mutation frequency of shuttled vectors is high, greater than 10−22, in all cell types with the exception of human 293 cells (Lebkowski et al. 1984). Agarose gel electrophoresis of rescued plasmids shows many insertions and deletions. The inclusion in these vectors of the hybrid neomycin resistance gene permits selection in mammalian cells using the drug G418. Selection of XP4PA-SVwt cells transfected with these vectors results in rapid establishment of cell clones that show episomally maintained plasmid, a phenomenon that appears much less toxic to the human cells than to monkey COS cells. A high copy-number of the vector was found in the low molecular weight DNA fraction of these XP cells, persisting for more than 2 months. When this persisting vector was rescued into bacteria, however, an even higher spontaneous mutation frequency was observed, and electrophoresis revealed a concomitant increase in gross alterations in the plasmid.

Fig. 1.

General structure of SV40-based (A) and EBV-based (B) shuttle vectors. Divisions within the circles indicate functional components including bacterial-plasmid (ColEl, pBR) and SV40 replication origins, EBV sequences necessary for plasmid replication and maintenance in mammalian cells (EBNA-1 and oriP), mutation locus (lacZ′), selectable genes (kanr/G418r, ampr and hygromycinr). Inner arrows show major transcripts. Restriction enzyme sites indicated in brackets are deleted. Constructions will be described in detail elsewhere.

Fig. 1.

General structure of SV40-based (A) and EBV-based (B) shuttle vectors. Divisions within the circles indicate functional components including bacterial-plasmid (ColEl, pBR) and SV40 replication origins, EBV sequences necessary for plasmid replication and maintenance in mammalian cells (EBNA-1 and oriP), mutation locus (lacZ′), selectable genes (kanr/G418r, ampr and hygromycinr). Inner arrows show major transcripts. Restriction enzyme sites indicated in brackets are deleted. Constructions will be described in detail elsewhere.

More recently we have inserted lacZ′ into Epstein-Barr virus (EBV)-based shuttle vectors to investigate the claim by Yates et al. (1985) that these vectors may be maintained, without gross alteration, in an exclusively episomal manner in human cells. Indeed, preliminary results of several cell lines containing between 30 and 100 copies of the vector p205-Z (Fig. IB) have indicated a far lower spontaneous mutation frequency of lacZ′ (10−4 to 10−5). These results suggest that this system may be suitable for examining induced mutagenesis and repair phenomena in a more systematic and efficient manner than has been possible previously.

Another possibility for studying mutagenesis is to combine the advantages of both virus and shuttle vectors by means of a packageable shuttle vector. Such a shuttle-virus would enable DNA to be introduced into mammalian cells without the need for transfection, which has certain disadvantages including variable and low efficiency and variable toxicity to many human cell lines. This method may also be one cause of the high spontaneous frequency of mutagenesis observed with shuttle vectors. It seemed feasible to construct a shuttle vector, similar to those described above, which was packageable as a mammalian virus by use of a helper system. We have accomplished this goal by the construction of a series of SV40-based shuttle vectors containing the entire SV40 late region, which codes for capsid proteins, and its intact replicon origin. These vectors, one of which is presented in Fig. 2, have part of the SV40 early genes, which normally transcribe for the small t- and large T-antigen, substituted by the miniplasmid πΔlac (Little et al. 1983). This plasmid has the minimal essential DNA sequences for replication in Escherichia coli, the supF suppressor tRNA gene, which provides selection in the appropriate host, and the lacO sequence. Although unable to produce the large T-antigen these plasmids can replicate efficiently in monkey COS-7 cells, which produce this protein constitutively, thus avoiding the use of a helper virus. Since all the other genetic information necessary for virus production is present in these vectors, they also produce infectious viral particles (Menck, James & Sarasin, unpublished data).

Fig. 2.

Scheme for the construction of a shuttle-virus. Divisions within the circles indicate functional components including bacterial-plasmid and SV40 origins, selectable gene (supF) and mutation locus (lacO). Inner arrows show major transcripts. Restriction enzyme sites indicated in brackets are deleted. Details will be described elsewhere.

Fig. 2.

Scheme for the construction of a shuttle-virus. Divisions within the circles indicate functional components including bacterial-plasmid and SV40 origins, selectable gene (supF) and mutation locus (lacO). Inner arrows show major transcripts. Restriction enzyme sites indicated in brackets are deleted. Details will be described elsewhere.

Plasmid DNA from cells infected with these viruses can be transferred back into E. coli for analysis of its stability during replication in mammalian cells. Many DNA alterations, mostly deletions, are observed in these plasmids and appear to be due to limits in the genome size that could be packaged as virus. Plasmids with sizes smaller than SV40 are more stable. Moreover, the lac operator sequence present in these vectors permits the analysis of mutations induced in this DNA target, which can be monitored by a sensitive colour assay. The presence of multiple copies of a nonmutated plasmid in a lacZ-proficient bacterial host leads to binding of the Lac repressor to the plasmid and thus to the derepression of the lac operon. In this situation, the bacteria metabolize the synthetic substrate X-gal to produce a chromogenic (blue) metabolite. Mutations in the lacO sequence that inhibit Lac repressor binding lead to formation of white colonies. Preliminary studies indicate that at least one of the vectors is very stable (lacO spontaneous mutation frequency after passage as shuttle-virus is lower than 0·03%). Analysis of mutations induced by u.v. irradiation of the shuttle-virus is in progress.

We have outlined a variety of strategies that we are using to analyse, at the molecular level, mutagenesis in mammalian cells.

The EBV-based shuttle vectors show considerable promise as this system appears to function in all immortalized human cells tested to date. Such a stable episomal system offers unique advantages in experimental versatility; for example, time course studies and more complex treatment protocols. It remains to be seen if the low copy number of these vectors contributes to their stability and may more closely approximate the host chromosome replication and repair than the high copy number, and rapidly replicating, papovaviruses.

On the other hand the approaches using SV40 as described above retain considerable advantages, namely: (1) high efficiency, approaching 100%, of infection compared to 5-10% transfection of mammalian cells; (2) ideal delivery to cells via virus-particle transport to the cell nucleus; (3) efficient virus replication: at least 10- to 100-fold greater than shuttle vectors; (4) efficient recovery of very pure DNA in the form of virus particles. An attractive possibility with these viral systems is the ability to work with untransformed diploid fibroblasts as this would permit studies in all of the human DNA repair deficiency syndromes.

To obtain, for mammalian cells, mutation data at the detailed molecular level comparable to that which exists for bacteria remains a formidable task. Modern experimental tools in biology are, however, very powerful and we are confident that one of the variety of approaches being examined by ourselves and others will succeed eventually in this aim.

C. F. M. Menck holds a postdoctoral fellowship from CNPq, Brazil; M. R. James acknowledges the support of IARC (Lyon, France) during the earliest stages of these projects and is now supported by ARC (Villejuif, France). This work has been supported by grants from ARC (Villejuif, France) and from the Commission of the European Communities (no. B16-163 F, Brussels, Belgium).

Bourre
,
F.
,
Gentil
,
A.
&
Sarasin
,
A.
(
1983
). Molecular analysis of ultraviolet induced mutagenesis on simian virus 40 DNA. In
Genetics: Nevi Frontiers
(ed.
V. L.
Chopra
,
B. C.
Joshi
, R. P. Sharma &
H. C.
Bansal
), pp.
385
393
.
New Delhi
:
Oxford & I.B.H. Publishing Co
.
Bourre
,
F.
&
Sarasin
,
A.
(
1983
).
Targeted mutagenesis of SV40 DNA induced by UV-light
.
Nature, Land.
305
,
68
70
.
Calos
,
M. P.
,
Lebkowski
,
J. S.
&
Botchan
,
M. R.
(
1983
).
High mutation frequency in DNA transfected into mammalian cells
.
Proc. natn. Acad. Sci. U.S.A.
80
,
3015
3019
.
Cleaver
,
J. E.
&
Bootsma
,
D.
(
1975
).
Xeroderma pigmentosum: biochemical and genetic characteristics
.
A. Rev. Genet.
9
,
19
38
.
Cooper
,
G. M.
(
1984
).
Activation of transforming genes in neoplasms
.
Br.J. Cancer
50
,
137
142
.
Daya-Grosjean
,
L.
,
James
,
M. R.
,
Drougard
,
C.
&
Sarasin
,
A.
(
1987
).
An immortalized xeroderma pigmentosum, group C, cell line which replicates SV40 shuttle vectors
.
Mutât. Res.(in press)
.
Gentil
,
A.
,
Daya-Grosjean
,
L.
,
Margot
,
A.
&
Sarasin
,
A.
(
1985
).
Survival and mutagenesis of ultraviolet irradiated Simian virus 40 in foetal human fibroblasts
.
Biochimie
67
,
393
398
.
Gentil
,
A.
,
Margot
,
A.
&
Sarasin
,
A.
(
1982
).
Enhanced reactivation and mutagenesis after transfection of carcinogen-treated monkey cells with UV-irradiated simian virus 40 (SV40) DNA
.
Biochimie
64
,
693
696
.
Gentil
,
A.
,
Margot
,
A.
&
Sarasin
,
A.
(
1984
).
Apurinic sites cause mutations in simian virus 40
.
Mutât. Res.
129
,
141
147
.
Guerrero
,
L
,
Willasante
,
A.
,
Corces
,
V.
&
Pellicer
,
A.
(
1984
).
Activation of a c-K-ras oncogene by somatic mutation in mouse lymphomas induced by gamma radiation
.
Science
225
,
1159
1162
.
Gluzman
,
Y.
(
1981
).
SV40-transformed simian cells support the replication of early SV40 mutants
.
Cell
23
,
175
182
.
Lebkowski
,
J. S.
,
DuBridge
,
R. B.
,
Antelle
,
E. A.
,
Greisein
,
K. S.
&
Calos
,
M. P.
(
1984
).
Transfected DNA is mutated in monkey, mouse and human cells
.
Molec. cell. Biol.
4
,
1951
1960
.
Little
,
P. F. R.
,
Treisman
,
R.
,
Bierut
,
L.
,
Seed
,
B.
&
Maniatis
,
T.
(
1983
).
Plasmid vectors for the rapid isolation and transcriptional analysis of human β-globin gene alleles
.
Molec. Biol. Med.
1
,
473
488
.
Razzaque
,
A.
,
Mizusawa
,
H.
&
Seidman
,
M. M.
(
1983
).
Rearrangement and mutagenesis of a shuttle vector plasmid after passage in mammalian cells
.
Proc. natn. Acad. Sci. U.SA.
80
,
3010
3014
.
Sarasin
,
A.
(
1985
).
SOS response in mammalian cells
.
Cancer Invest.
3
,
163
174
.
Sarasin
,
A.
&
Benoit
,
A.
(
1986
).
Error-prone replication of ultraviolet-irradiated simian virus 40 occurs in mitomycin c-treated host cells only at low multiplicity of infection
.
Molec. cell. Biol.
6
,
1102
1107
.
Sarasin
,
A.
&
Hanawalt
,
P. C.
(
1978
).
Carcinogens enhance survival of UV-irradiated simian virus 40 in treated monkey kidney cells: induction of a recovery pathway
.
Proc. natn. Acad. Sci. U.SA.
75
,
346
350
.
Shaaper
,
R. M.
&
Loeb
,
L. A.
(
1981
).
Depurination causes mutations in SOS-induced cells
.
Proc, natn. Acad. Sci. U.SA.
78
,
1773
1777
.
Smith
,
K.
&
Sargentini
,
N. J.
(
1985
).
Metabolically-produced “UV-like” DNA damage and its role in spontaneous mutagenesis
.
Photochem. Photobiol.
42
,
801
803
.
Sukumar
,
S.
,
Notario
,
V.
,
Martin-Zanca
,
O.
&
Barbacid
,
M.
(
1983
).
Induction of mammary carcinomas in rats by nitroso-methyl-urea involves malignant activation of H-ra.s-1 locus by single point mutations
.
Nature, Land.
306
,
658
661
.
Thacker
,
J.
(
1985
).
The molecular nature of mutations in cultured mammalian cells: a review
.
Mutât. Res.
150
,
431
442
.
Yates
,
J. M.
,
Warren
,
N.
&
Sugden
,
B.
(
1985
).
Stable replication of plasmids derived from Epstein-Barr virus in various mammalian cells
.
Nature, Lond.
313
,
812
815
.