An Escherichia coli plasmid-based, mutational system in which supF mutants are selectable: Insertion elements dominate the spontaneous spectra

Mutation Research, 270 (1992) 219 -231 © 1992 Elsevier Science Publishers B.V. All rights reserved 0027-5107/92/$05.00

2 iq

MUT 05176

An Escherichia coli plasmid-based, mutational system in w h i c h supF m u t a n t s are selectable: Insertion e l e m e n t s d o m i n a t e the s p o n t a n e o u s spectra Henry Rodriguez a, Elizabeth T. Snow b, Uppoor Bhat a and Edward L. Loechler a ,z Department of Biology, Boston Unirersity, Boston, MA 02215, USA and b Institute of Enrironmental Medicine, New York Unicersio, Medical Center, Tuxedo, NY 10987, USA (Received 24 December 1991) (Revision received I ! May 1992) (Accepted 29 May 1992)

Keywords: SupF mutants, selectable; Insertion elements

Escheriehia coil, plasmid-based,

mutational

system;

Spontaneous

spectra;

Summary A new system is described to determine the mutational spectra of mutagens and carcinogens in Escherichia coil; data on a limited number (142) of spontaneous mutants is presented. The mutational assay employs a method to select (rather than screen) for mutations in a supF target gene carried on a plasmid. The E. coli host cells (ES87) are lacl- (am26), and carry the IocZAMI5 marker for a-complementation in /3-galactosidase. When these cells also carry a plasmid, such as pUB3, which contains a wild-type copy of supF and lacZ-a, the lactose operon is repressed (off). Furthermore, supF suppression of lacl "m26 results in a lactose repressor that has an uninducible, lacl s genotype, which makes the cells unable to grow on lactose minimal plates. In contrast, spontaneous or mutagen-induced supF- mutations in pUB3 prevent suppression of lacl '."2~' and result in constitutive expression of the lactose operon, which permits growth on lactose minimal plates. The spontaneous mutation frequency in the supF gene is ~ 0.7 and ~ 1.0 x 10 -~ without and with SOS induction, respectively. Spontaneous mutations are dominated by large insertions (67% in SOS-uninduced and 56% in SOS-induced cells), and their frequency of appearance is largely unaffected by SOS induction. These are identified by DNA sequencing to be Insertion Elements; IS1 dominates, but IS4, ISS, gamma-delta and ISl0 are also obtained. Large deletions also contribute significantly (19% and 15% for - S O S and +SOS, respectively), where a specific deletion between a 10 base pair direct repeat dominates; the frequency of appearance of these mutations also appears to be unaffected by SOS induction, in contrast, SOS induction increases base pairing mutations (13% and 27% for - S O S and + SOS, respectively). The ES87/pUB3 system has many advantages for determining mutational spectra, including the fact that mutant isolation is fast and simple, and the determination of mutational changes is rapid because of the small size of supF.

Correspondence: Dr. Edward L. Loechler, Department of Biology, Boston University, Boston, MA 02215, USA.

220

Cancer appears to result from the activation of oncogenes from their normal counterparts, protooncogenes, via mutation (Bishop, 1991; Balmain and Brown, 1988; Hollstein et al., 1991). Mutagenic events can occur spontaneously or be induced by exogenous agents and both are likely to contribute to the carcinogenic process. A large body of work suggests that not all mutagens show the same mutagenic specificity; e.g., simple methylating and ethylating agents induce primarily GC AT mutations (Miller, 1980; Eckert eta!., 1988; Richardson et al., 1987; Burns et al., 1987, 1988; DuBridge et al., 1987; Horsfall et al., 1990), while GC ~ TA mutations are induced more often by bulky mutagens, such as benzo[a]pyrene (Eisenstadt et al., 1983; Mazur and Glickman, 1988; Yang et al., 1987, 1991; Bernelot-Moens et al., 1990; Carothers and Grunberger, 1990; Chert et al., 1990, 1991). Furthermore, mutations at (e.g.) G : C base pairs may occur at different rates depending upon sequence context, which has lead to the concept of mutational spectra (Miller, 1980). Both mutagenic specificity and sequence specificity of mutation may be important in the activation of certain oncogenes (Hollstein et al., 1991). A variety of systems have been developed to study mutagenic specificity and mutational spectra; the first comprehensive system had iacl as a target gene and was developed by Miller (Coulandre and Miller, 1977; Miller, 1980, 1983). One system developed recently is plasmid-based and involves the isolation of mutants in supF, a tRNA suppressor gene (Kraemer and Seidman, 1989). This system has many advantages (enumerated in Discussion) perhaps, most importantly, that at many base pairs all mutational changes are phenotypically detectable (e.g., GC --, AT, GC -~ TA and GC ~ CG). One of the drawbacks of the existing supF-based system has been that mutants are isolated via a genetic screen, where a small fraction of white (or light blue colonies), which contain supF- plasmids, must be identified in a field of blue colonies, which contain supF + plasmids. Herein we report the development of a system for E. coli that enables the selection of supF mutants; i.e., only cells containing supF- plasmids give rise to colonies. This facilitates mutant

isolation. A limited number of spontaneous mutants (142) were isolated using this system, where insertion elements dominate. Subsequent work will describe the use of this system to study carcinogen-induced mutagenesis, as well as the development of a similar system for use in human cells in culture. Materials and methods

Strains and plasmids. Strain ES87 has the genotype, (Apro-lac, strA)/F' (pro +, lacl °, lacl am26, lacZAM15). ES87 is derived from strain $90C, which is Apro-lac, strA; F' in ES87 is derived from F" (pro +, lacl °, lacI am26, lacZ +) carried in strain UVB124, via recombination with a lacZAM15 allele in strain, JM83. $90C, UVB124 and JM83 were all obtained from Dr. Reel Schaaper. pUB3 was constructed (Fig. 1) by the insertion of a 655-bp, blunt-ended fragment (HindlII to Ddel containing the supF gene and a small polylinker) isolated from the plasmid, PiAN7 (obtained from Dr. Brian Seed), into the AfllIl site at position 593 in the plasmid, pTZ19R, which was obtained from Pharmacia and is essentially pUC19 (Yanisch-Perron et al., 1985) with an fl origin of replication, pUB3 was numbered according to the numbering system for pTZI9R, which means that the promoter, pre-tRNA and supF gene are at 1197-1163, 1162-1122 and 1121-1029, respectively. However, when referring to supF, we use the numbering system from pZ189 (Kraemer and Seidman, 1989), to which all are more accustomed, pUB3 was purified by cesium chloride density gradient centrifugation in the presence of ethidium bromide (Sambrook et al., 1989). Reagents and media.

All reagents not mentioned explicitly were of the manufacturers' highest grade purity. TE is 10 mM Tris. HCI; 1 mM EDTA, pH 8.0. SOC media is 2% tryptone (Difco); 0.5% yeast extract (Difco); 10 mM NaCl; 2.5 mM KCI; 10 mM MgCI2; and 10 mM MgSO4; with 0.37% glucose (added after the other ingredients were mixed, autoclaved and cooled). M13 (10 × Salts) is 0.6 M K2HPO4; 0.33 M KH2PO4;

221

0/3553

i

FoXy

g// or±

"~

cm:tai'en;e~

250~p) ,i~~000

~transform

~ nomutation $mutation genotype

supF-/ 2acI-(am2 6)

off uninducible

-lactose +lactose growth

supr +/ 2ac~ - (am26)

on

on on

yes

no

lactose

10

20

•

30

•

40

.

50

,

60

,

70

•

80

.

90

.

100

o

GAATT~TTT~TcAA~GTAAcA~TTTAcAGcGG~GcGT~T2TGATATGA~G~G~GcT~GATAAGGGAGcAGGc~RGTA~A~G~TTA~TGT~ llO

120 •

130 °

140 •

150 •

160 •

170 •

180 .

190 •

200 •

TGGGGTTcccGAGcGGcCARAGGGAGcAGAcT~TRAAT~TG~CGTc&T~GA~TT~G~AGGTT~G~AT~TTcc~c~CACcAT~TT~TC~G

Fig. 1. Strategy for the ES87/pUB3 system to determine mutational spectra in a sup/: gene. pUB3 can either be adducted in vitro to study mutagen/carcinogen induced mutagenesis or transformed without adduetion to study spontaneous mutagenesis. The genes in pUB3 and the strategy to select supF mutants are described in the text. Polylinker I was derived from pTZIgR and in order consists of: EcoRl, Sacl, Kpnl, Smal, BamHI, Xbal. Sail, Pstl, Sphl and Hindlll. Polylinker !I was derived from PLAN7 and in order consists of: EcoRl, Smal, BamHi, Hinell, Pstl, Bglll and Xbal; the Hindlll site was lost during vector construction. The supF sequence is given at the bottom of the figure using the numbering system common to pZ189 (Kraemer and Seidman, 1989), where the promoter is between bp24 and bp58, the pre-tRNA is between ho59 and bp98, and the tRNA gene is between bp99 and bp183, in the pUB3 numbering system this corresponds to between bp1197 anti ;~p1163, between bp1162 and bp1122, and between bp1121 and bp1029, respectively.

222 76 mM (NH4)2SO4; and 17 mM sodium citrate (autoclaved). Lactose Minimal (LM) Plates are made as follows: to 900 ml of autoclaved agar (1.5%, Difco), 100 ml of M13 10 × Salts is added, and supplemented with 0.1 mM CaCIz, 0.1 mM MgSO4, 0.1% lactose, 40 /~g/ml ampicillin, 25 tzg/ml streptomycin, 0.025% thiamine-HCI, 20 ng/ml glucose, 20 mg/ml 5-bromo-4-chloro-3-indolyl-a-D-galactosidase (X-Gal; Gold Biotechnology). Lactose media is 50 mM Na2HPO4; 22 mM KH_,PO4; 8.5 mM NaCI; and 18.7 mM NH4CI; the mixture is autoclaved, cooled and supplemented with 2 mM MgSO4; 0.1 mM CaCI2; 0.2% glucose, and 0.1% lactose. Melibiose minimal (MM) plates are identical to LM plates except that 0.1% melibiose replaces 0.1% lactose. Melibiose media is identical to lactose media except that 0.1% melibiose replaces 0.1% lactose.

Preparation of transformation competent ES87 cells (Dower et al., 1988). A 2-liter flask containing 1 liter of Luria-Bertani (LB) media was inoculated with a 30-ml overnight culture of strain, ES87 (grown in LB supplemented with 25/zg/ml streptomycin), and was grown at 37°C with vigorpus shaking (New Brunswick G25 incubator) to OD.~.~,-- 0.65, then placed on ice for '10 rain. 1011 ml was placed into each of two centrifuge bottles on ice, which were set aside for experiments not involving SOS induction. (This is a typical experiment and the proportion of SOS-unindueed and SOS-induced cells was varied,) The remaining cells were centrifuged at 8000 rpm for 10 min at 4°C in six 250-ml centrifuge bottles. After decanting the supernatant, each pellet was resuspended in 200 ml of 10 mM MgSO4 (4oC) prior to SOSinduction via a slight modification of the procedure of KoffeI-Schwartz et al. (1984). 100 ml of these cells were put into a 150 × 15-mm petri dish (4°C) and irradiated with a 254-nm germicidal lamp (Ultra-Violet Products) at 12.6 J / m 2, which gave ~ 50% survival and corresponded to a dose known to induce the SOS functions (Witkin, 1976). This was repeated eleven times. Following the addition of 1200 ml of 2 × LB media (room temperature), the cells were distributed into two, 2-liter flasks, vigorously shaken for 20 min at 37°C to allow expression of SOS functions, then incubated on ice for 10 rain. The

cells were distributed into twelve 250-ml centrifuge bottles and they, as well as the two bottles from the non-SOS induced cells, were centrifuged at 8000 rpm for 10 min (4°C). Each pellet from the SOS-induced cells was resuspended in 100 ml of sterile distilled water (4°C), and combined into six, 250-ml centrifuge bottles (200 ml each), while each pellet from non-SOS induced cells was resuspended in 200 ml sterile distilled water (0°C). These 8 bottles were centrifuged at 8000 rpm for 10 min (4°C) and each resulting pellet was resuspended in 200 ml of sterile distilled water (0°C). Pelleting was repeated, except that, following decanting, the excess water was aspirated off and each pellet was resuspended in 200/zl of sterile 10% glycerol/H20 (0°C).

Electroporation, 100 p,! of freshly prepared competent ES87 cells were mixed with ~ 250 pg of pUB3 (in 2.5 p,I) and placed on ice for 1 min. Following transfer to an electroporation cuvette (0.2 cm gap; Bio-Rad), the sample was pulsed at 2.5 kV/200 0 / 2 5 ~ F in an electroporator (BioRad Gene Pulser, Model 1652076). Immediately thereafter, 1 ml of SOC media was added and the mixture divided equally into three disposable culture tubes, which already contained 666 ~l of SOC media (total volume, ~ 1.033 ml). Each sample was allowed to recover tbr I h at 370C. Mutant selection. Total transformants were determined as follows. An aliquot of cells from each transformation was removed, diluted 1 : 100, and 100 p,I plated onto LB plates containing ampicillin (40 ~,g/ml; Sigma). Cells containing supF ÷ or supF- plasmids will grow on these plates, in order to select for cells containing pUB3 with a mutant (supF-) gene, the cells were transferred to microfuge tubes, centrifuged (Brinkmann 5415) at 10000 rpm for 1 rain at room temperature, and the media aspirated off (being cautious not to disrupt the pellets). Cells were resuspended in 200 btl of 0.1% lactose media: 50 ~l were plated on LM plates containing X-Gal (20 rag/liter), ampicillin (40 #.g/ml), and streptomycin (25/xg/ml). Plating this volume ensured that a significant fraction ( ~ 30%) of the plates will have no mutants (nulls). Blue colonies

223

were picked after two days and streaked onto new LM plates. The majority (90-95%) of these colonies were true mutants (i.e., ES87 cells with supF- pUB3), but some were "false mutants" and gave wild-type supF + sequences. Many of the false mutants appeared to be l a d - revertants, which were screened out as follows. A colony growing on a lactose minimal plate was picked with a sterile toothpick and dispersed into 500 ~l of melibiose media; 3 gl were spotted onto a MM plate containing X-Gal (20 rag/l), ampicillin (40 #g/ml), and streptomycin (25 t~g/ml). After incubation for 18 h at 370C, true mutants grew and were blue, while most false mutants showed little or no growth and were white. Following this screening, some false mutants remained, although the frequency was low ( < 5%); this is addressed in the Discussion. DNA sequencing. DNA for single-stranded DNA sequencing was isolated as follows. LB (3 ml) containing ampicillin (40 p.g/ml) and streptomycin (25/zg/ml) was inoculated with 100/zi of an overnight culture (in LB media) of a true mutant, and grown for 1 h at 37°C with vigorous shaking. Thereafter, 50 ~1 of a stock of MI3K07 helper phage (at a titer of I x 10tt/ml; Russel et al., 1986) was added, and incubated for 12 h at 37*(2 with vigorous shaking. A l-ml aliquot was removed and centrifuged in a microcentrifuge at 10000 rpm for 5 rain at room temperature. The supernatant was transferred to a new microfuge tube, and was centrifuged again. The supernatant was mixed with 300 #1 of 2.5 M NaCI/20% PEG-8000 (J.T. Baker) and placed on ice for 1 h. Following centrifugation (10000 rpm for 15 min at 4°C), the supernatant was discarded and the pellet was resuspended in 100/.d of TE (pH 8.0). Single-stranded DNA was deproteinized with TE-saturated phenol and, subsequently, extracted with chloroform (100/zl). Following ethanol precipitation (120 mM sodium acetate; pH 5.2) the pellet was washed with 150 #1 of 70% EtOH, repelleted, dried and resuspended in 10/~1 of TE (pH 8.0). This DNA was suitable for DNA sequencing. Single-stranded and double-stranded pUB3 from ES87 cells did not always give readable

DNA sequences (see Discussion); thus, ds pUB3 from ES87 cells was isolated by a rapid procedure and transformed into DH5-~ cells, from which unambiguous DNA sequences were obtained. An overnight culture (1.5 ml) of ES87 cells containing a true mutant was pelleted, and resuspended by vortexing in 200/~l of solution I plus (50 mM glucose; 25 mM Tris; 10 mM EDTA; 7 mg/ml lysozyme; pH 8.0). Following incubation for 5 min at room temperature, 400/zi of solution II (0.2 N NaOH; 1% SDS) was added, and incubated 5 rain on ice. Subsequently, 300/~l of 7.5 M ammonium acetate (pH 7.8) was added, mixed, incubated 10 min on ice, and centrifuged (10000 rpm for 15 min at room temperature). The supernatant was added to 0.5 ml of isopropanol (-17°C), and incubated for 15 min at -70°C. Nucleic acids were pelleted (10000 rpm for 15 rain at 4°C), washed with 150/zl of 70% ethanol, repelleted, dried, and resuspended in 20/zl of TE (pH 8.0) and incubated for 1 h at 37°C with RNAase (0.2 mg/ml; Sigma). Double-stranded pUB3 from ES87 cells was retransfonned into competent DH5-a cells prepared by the standard CaCI2/heat shock method (Sambrook et ai., 1989). Following heat shock (2 min at 42°C), 1 ml of 2 x LB was added, and recovery conducted (1 h at 370C with shaking). Transformants were obtained by plating 100 g! onto LB plates containing ampicillin (40 #g/ml), and X-Gal (20 mg/I). A DH5-a transformant containing pUB3 was picked, grown overnight in 10 ml of LB containing ampicillin (40 #g/ml), and then poured into a snap-cap plastic tube (14 ml; Sarstedt Model 55.538) and centrifuged (3000 rpm for 10 min at 4°C). The pellet was resuspended (100/~l of 0.73 mM sucrose; 50 mM Tris; 150 mM NaCi; pH 8.0), mixed with lysozyme (10 mg/ml in 50 #l), gently vortexed, and incubated for l0 rain (0°C). Thereafter, 15/.d of 0.5 M EDTA (pH 8.0) was added, gently vortexed, and incubated for 10 rain (0°C). To this, 150 #! of lysis buffer (0.5% Triton X-100; 50 mM Tris; 62.5 mM EDTA; pH 8.0) was added, gently vortexed, and incubated for 10 min at (0°C) then centrifuged (10000 rpm for 20 min at 4°C). The viscous layer was removed and the remaining cell lysate heated for 1 rain at 90°C, and centrifuged at 10000 rpm for 5 min at 4°C. This

2~ supernatant was transferred to a new microfuge tube containing 200 pl of 7.5 M ammonium acetate and 900 #l of EtOH. Following incubation for l0 min ( - 70°C), the DNA was pelleted (10 000 rpm for 20 min at 4°C), the supernatant removed via aspiration and the pellet dried. Plasmid DNA was resuspended in 400/zl of TE (pH 8.0), deproteinized with TE-saturated phenol, and was extracted with chloroform (200 p,I). Finally, the DNA ( ~ 30 /~g) was precipitated in ethanol as before, and resuspended in 20 #l of TE (pH 8.0) and incubated for 1 h at 37°C with RNAase (0.2 mg/mi). DNA sequencing was based upon the Sequenase kit (United States Biochemical, version 2.0). Double-stranded pUB3 I,,-, 3 p,g) from strain DH5-a was diluted to 20/.tl with distilled water. Thereafter, 4 p,1 ef 2 N NaOH, 20 p,i of distilled water, l0/~l of 3 M sodium acetate (pH 4.8) ano 130 pl of EtOH were added sequentially; nucleic acid precipitation proceeded overnight at -70°C. Following pelleting, the DNA was resuspended in 7 t~l of distilled water, and 2 Izl of 5 x Reaction Buffer and l p,l ( -,, 100 ng) of primer (5'-TTCTTTCTCAACGTAACACT-3') were added. This primer is complementary to bases 4 to 23 by the standard supF numbering system (Fig. 1) and is immediately upstream of the promoter. The mixture was warmed for 2 min at 65°C in a sealed tube, then cooled to room temperature over a period of 31) rain. Subsequently, 1 ~1 of 0.1 M DTT, 2 ~tl of Labelling Mix (diluted 15-fold with distilled water), 0.4 Ixl of [a-3:p]-dATP (10 mCi/ml and 800 (i/rumple; New England Nuclear), and 2 #1 of Sequenase (diluted 1:8 with Enzyme Dilution buffer) were added. After incubation for 2 min at room temperature, 3.5 p,I was transferred to each termination tube (containing 2.5 #1 of the correspor, ding G, A, T and C mixes), and incubated for 20 min at 37°C, Finally, 4/~1 of Stop Solution was added to each tube. Samples were stored at - 17°C, then heated for 2 rain at 91)°C, placed on ice and loaded (2 ~1) onto an 8% denaturing polyacrylamide gel. For ss DNA sequencing, 1-2/zg of DNA in a volume of 7 ~tl was used. The rest of the sequencing procedure was identical (see above) beginning with the addition of 2 p.I of 5 × Reaction Buffer.

Mutation frequency calculation. It is possible during the recovery period after transformation for some, albeit limited, cell division to occur. Consequently, when 2 or more mutants appear on a plate, some could in principle be siblings. This lead us to calculate mutation frequency based upon the Poisson Distribution. The volume plated in each transformation was adjusted such that ~ 30% of all plates had no mutant colonies (nulls). The average number of mutant colonies per plate was calculated from the Poisson Distribution according to the formula, A = -In[f(0)] where A is the average number of mutant colonies per plate and f(0) is the fraction of null plates. Mutation frequency (MF) is merely,

MF = [A x P ] / T where P is the total number of plates and T is total transformants. The reported mutation frequency is the average of 4 Expts. Statistical analyses to determine if there is a difference between the mutants collected from SOS-uninduced vs. SOS-induced cells was performed using a hypergeometric test (Adams and Skopek, 1987).

Analysis o]' the large insertion mutants. The first ~ 30 nucleotides of a large insertion were read, and analyzed through Genbank (Pearson and Lipman, 1988). The sequence with maximum homology is reported. Results

The strategy outlined in Fig. 1 (described below) was devised by one of us (E.T.S.) and involves a positive selection for mutations induced in the supF gene of the plasmid, pUB3, using ES87 cells, pUB3 can be either used to study spontaneous mutagenesis or adducted prior to transformation, as shown in Fig. l, in order to study carcinogen-induced mutagenesis, pUB3 transformation by both electroporation and CaCl 2 procedures have been used, but the former requires smaller quantities of DNA and is the method of choice. Both cells uninduced and in-

225 TABLE 1 SPONTANEOUS MUTATIONS DETECTED IN THE supF GENE of pUB3 FOLLOWING TRANSFORMATION INTO SOS-UNINDUCED AND SOS-INDUCED ES87 CELLS Type of mutation

- SOS

+ SOS

Insertions Insertion elements No sequence d Other insertions

47" (0.44) b 38 c (0.36) 6 (0.06) 4 c (0.03)

41 (0.58) 35 (0.49) 1 (0.01) 5 r (0.07)

Deletions Deletion of bp68 to bp119 Deletion of bpl50 to bpl71 Other deletions

13 ( 0 . 1 4 ) ! 1 (0.10) 1 (0.01) I g (0.01)

11 ( 0 . 1 6 )

9 (0.09)

20 (0.28)

Base-pairing mutations Individual mutations G 99 ---, T G 99~A G105~C GII5~A AI35 ~ C A137~G TI40~C TI40~ A G144 --* T C155 ~ G G160 ~ A C169 ~ T TI71 -~ G C174 --' G C!74 ..~ A C179 --* G Mutagenic specificity GC --, AT GC --* TA GC .~ CG AT -* GC AT ...) TA AT -=. CG Total mutants Mutation frequency h (S.D.) a t, c d

8 (0.11 ) 3 (0.04) 0 ( < 0.01)

2 I I

3

1 I i 1 5 1 1

I

2 1 2 i 2 1

1 2 (0.02) 2 ((}.02) 2 (0.02) 3 (0.03) 0 ( < 0.01 ) 0 ( < 0.{Ill 69 0.66 × ! O - ~, (0.10x 10 -t')

4 (0.{)6) 2 ((o.03) 5 (0.07) 6 t0.08) i t0.01 ) 2 (0.03) 73 1.03 × I 0 - t,

(0.43 x l0 -t,)

Number of mutant isolated of the indicated type. Mutational frequency ( x 10 +6) as determined by procedures outlined in Materials and methods. One mutant, which gave a plasmid of 3.0 kb, had an ISI insertion and a deletion. In several instances, plasmids with molecular weights greater than pUB3 itself (i.e. > 3.5 kb) were isulated that could not be sequenced. We presume that these result from insertion elements that disrupt the binding of the DNA sequencing primer. ¢ One insertion was identified with three G : C base pairs added at bpi74. Two large insertions were identified with no significant homology to any E. coil gene reported in Genbank. t One mutant, which gave a plasmid of 6.0kb, was isolated with an insertion after bpl37 that sho~ved homology to rpsM: there was no obvious homology between the inserted sequence and supF. One insertion was identified that duplicated the sequence 5'-TTCCCC-3', which is located between bplT0 and bp175. Three large insertions were indentified with no significant homoh)gy to any E. coil gent reported in Genbank. One large deletion from bp167 in supF to bpl019 in pUB3 was isolated. h Two significant figures are reported for the mutation frequency, although fewer are warranted. The indicated mutants were isolated in 9 Expts. The mutation frequent.T (Materials and methods) is the average of the 4 most reliable Expts.: S.D. is standard deviation.

--..6

duced for SOS were used, where the latter was accomplished by UV-irradiation of the cells prior to transformation using a standard protocol (Koffel-Schwartz et al., 1984). In separate work using pUB3 adducted with either the (+)-anti-diol epoxide of benzo[a]pyrene or the (+)-anti-diol epoxide of dibenz[a,j]anthracene (data not

shown), this treatment was shown to dramatically change the mutational specificity and increase mutation frequency ( > 5-fold), which suggests that SOS is indeed induced. pUB3 (Fig. 1) was constructed from the plasmid, pTZ19R, and the plasmid, PLAN7, and has: (i) a ColE1 plasmid origin of replication; (ii) an

TABLE 2 INSERTION ELEMENT MUTANTS

supF

supF

upstream of insert h

downstream of insert b

33 34 44 66

TTACAGCGG TACAGCGGC CGTCATTTG CGCTTC CC6

CGCGTCATT GCGTCATTT ATATGATGC ATAAGGGA6

1 (4.3) 2 (4.3, 7.2) l (4.3)

8 (3.0, 4.3, 4.3, 4.3, ~t.3. 4.3, 4.3, 8.0)

5 (4.3, 4.3, 4.3, 4.3, 5.0)

91

GTAAAAGCA

TTACCTGTG

3 (4.3, 4,3, 8.7)

125 13h 157 162 165

GGTTTCCCT CAGACTCTA TCGACTTCG TTCGAAGGT GAAGGTTCG

CGTCTGAGA AATCTGCCG AAGGGTCGA TCGAATCCT AATCCTTCC

7 (4.3,4.3, 4.3, 4.3, 4.3, 5.0, 8.9) ! (4.3)

1 (5,7) 5 (4.3. 4,3,

I (8.0) I (7.2) 2 (4.3,4.3)

171

TCfiAATCCT

TCCCCCACC

4,3, 4,3, 8,1) 2 (4.3,4.3)

I (4.3)

103

CCTGTGGTG

GGGTTCCCG

I (4,0)

I (6.0)

IS5 137

A(IACTC TAA

ATCTGCC6T

2 (4,6, 4,6)

2 (4.6, 4,6)

TACA6CGGC AGCG6CGC6 GCGGCGCGT GTGGTfif66

GCGTCATTT TCATTTGAT CATTTGATA TTCC CGAGC

2(4.8, 7,7)

CGATAAGG6 GCCAGTAAA CAGTAAAAG AAGCATTAC

AGCAGGCCA AGCATTACC CATTACCTG CT6TGGTGG

2(9.1,9.1)

Site "

Number (size) - SOS c

Number (size) + SOS ¢

ISI d

I (4.3)

IS4

ISIO 34 37 38

116

I (4.6) 2 (4.6, 4.6)

1 (6.0) 2 (4.6, 4,6) 10 (4.6, 4,6, 4,6, 4.9, 4.9, 4.9, 5.7, 5.7, 7,3, 8.8)

gamma .della

73 87 80 95

I (7.6) 3 (8.1,8.1, 8.9)

1 (7.5)

" The number of the last base in supF prior to the site of the insertion, h Sequence of the 9 bases in supF immediately upstream and downstream of the site of the insertion, The number of occurrences at the indicated site; the numbers in parentheses indicate the approximate size of the insertion based upon a comparison to standards in an agarose gel, d The likely source of the inserted sequences based upon sequence homology (Pearson and Lipman, 1988),

227 M13 origin of replication making it a phagemid; (iii) a bla gene encoding ampicillin resistance; (iv) a lacZ-a fragment, which permits a-complementation for/~-galactosidase activity; and (v) a supF gene, which is the mutational target. A similar plasmid, pZStet, has also been constructed and uses tetracycline rather than ampiciIIin as the antibiotic resistance marker (E.T.S., unpublished). ES87 cells are l a d - (am26), which is suppressed when the plasmid, pUB3, carrying supF +, is present. As a result, cells that are lacI(am26)/supF + have a functional lactose repressor. SupF suppression !cads to the replacement of glutamine, the wild-type amino acid at position 248 (Miller et al., 1978), with tyrosine; this change generates an i s repressor (Klcina and Miller, 1990). Furthermore, ES87 cells were constructed to be i Q, which results in the production of greater quantities of the lactose repressor. The consequence of the presence of both i s and i ° changes is that ES87 cells have a lactose operon that is uninducible by lactose (or IPTG). Thus, ES87 ceils harboring a pUB3 plasmid that is supF + do not grow on lactose as the sole carbon source. In contrast, ES87 cells harboring a pUB3 plasmid that contains a mutant supF- gene do not suppress the l a d - (am26) mutation, such that the lactose repressor is defective and the lactose operon is constitutively on. Thus, ES87 cells harboring a supF- pUB3 plasmid grow on lactose as the sole carbon source. There are several additional features of this system that should be noted. ( 1 ) T h e replication of pUB3 generates two progeny plasmids; thus, if a mutation occurs during replication in one of the two strands, then only one half of the progeny plasmids (on average) will contain the mutation. Initially, each cell containing mutant progeny will be supF+/supF - and will be phenotypically SupF +. To permit the segregation of pure supFcells, a small amount of glucose is included on the lactose minimal plates to allow some growth. (2) The following protocol was used to ensure that no mutants included in the data set were siblings. Immediately following a transformation of pUB3 into ES87 cells, the mixture was divided into 3 samples. Transformation was repeated numerous times. No more than one supF- mutant was isolated from any sample. (3) During the

recovery period, cell growth is possible such that in principle siblings could be generated. To avoid over-estimating the number of independent mutants, experiments were designed so that ~ 30% of them had no colonies (nulls). The fraction of nulls was used to calculate mutation frequency based upon the Poisson Distribution (Materials and methods). Using this strategy, mutant plasmids were isolated and sequenced (Materials and methods). Table 1 lists the 69 and 73 spontaneous mutants isolated from SOS-uninduced and SOS-induced cells, respectively. Large insertions, whose sequences suggested they were insertion elements, dominated the spontaneous spectra. Table 2 provides more detailed information about the nature of the insertion elements. The mutation frequency was approximately 0.66 and 1.03 × 10 -6 without and with SOS induction, respectively. (In all cases, we report two significant figures, but fewer are probably warranted.) The range of mutation frequencies from 4 Expts. was 0.54-0.75 × 10 -6 without SOS induction and 0.68-1.65 × 10 -6 with SOS induction. Discussion

Adt'antages of the ES87 /pUB3 system There are several features of the ES87/pUB3 cell system that make it attractive for studying mutational spectra in E. coil, including the fact that a lot of data can be generated rapidly. Points (1) and (2) refer to attributes of supF that have been discussed previously (e.g., see IOaemer and Seidman, 1989). (1) supF is a small gene (85 bp), which makes the task of locating a mutation by DNA sequencing relatively easy. (2) Most base pairs within the supF gene are targets for mutation; thus, in spite of its small size the number of base pairs that are phenotypically sensitive to mutatiqn is actually relatively large (Kraemer and Seidman, 1989). Furthermore, many of these base pairs detect all mutational changes (e.g., GC --* AT, GC ~ TA and GC --, CG) such that few mutations are silent. To date out of 255 possible base pairing mutations (i.e., 3

228

for each of the 85 bp in supF), 170 (67%) have been detected phenotypically based upon the compilation of Kraemer and Seidman (1989) and our own work. In fact all three base pairing mutational changes have been detected at 37 of 85 sites (44%). (3) The fact that at any given base pair, frequently all mutational changes can be detected means that, if a single mutational change is observed (e.g., GC---> AT), then this can be attributed to the mutagenic specificity of the mutagen/carcinogen at that site and not to the fact that the other mutational changes (e.g., GC ~ TA and GC ~ CG) are phenotypically silent. (4) The ES87/pUB3 system is a forward mutational assay involving a selection, which makes the collection of mutants simnle. A number of other systems involving target genes in the lactose operon rely upon phenotypic screens and require the isolation of white or light blue mutant colonies (or plaques) in a field of dark blue colonies (or plaques). (5) The fact that pUB3 can be manipulated in vitro has several advantages. For example, pUB3 is a phagemid allowing the isolation of ss DNA. Thus, it is straightforward to construct pUB3 heteroduplexes with ass gap that spans the supF gene, especially because the supF gene is surrounded by two polylinkers (Fig. 1). This makes (among other things) polymerase fidelity studies feasible.

Difficulties with the ES87 /pUB3 system (1) ES87 cells are repair-proficient. Thus, studies in repair-deficient strains will require either derivatives of ES87 cells or pUB3 to be transformed into another strain of E. coil initially where mutations are fixed, then the plasmid DNA isolated and screened for supF- mutants in ES87 cells. (2) pUB3 is a phagemid, so ss DNA can be isolated directly from ES87 cells for sequencing. However, we have found that ~ 10-20% of the time, this ss pUB3 DNA does not give a DNA sequence that can be read unambiguously. (Work is in progress to perfect this.) Furthermore, ds pUB3 isolated from ES87 cells does not give DNA sequences that can be read unambiguously ~50% of the time. Thus, for most of these

studies, ds pUB3 was isolated from ES87 cells by a rapid procedure and re-transformed into DH5-a cells, from which pUB3 DNA was isolated that gave reliably readable DNA sequences. (3) As outlined in Materials and methods, ~ 5-10% of the cells that were selected from lactose minimal plates are referred to as "false mutants" because they proved to have a wild-type supF + sequence. Many of these false mutants could be screened out using a simple screen. True mutants, which are supF-, readily grow and give blue colonies on melibiose minimal plates containing X-gal, while false mutants (e.g., lacl-(am26) revertants) are white or show little growth (Materials and methods). That false mutants were not isolated more frequently is perhaps surprising given that any cell containing a supF + plasmid and acquiring a second mutation that converted the lacl-(am26) chromosomal allele into a non-suppressible laclmutant (e.g., most missense or frameshift mutations) would have grown on lactose minimal plates containing ampieillin. On occasion false mutants were obtained following the melibiose screen, but the fraction was always low ( < 5%). Thus, this issue is not a practical prob'em, which is rationalized below. It has been reported that the spontaneous mutation frequency for conversion of laci + to lacl- is 2-4 x 10 -t' (Schaaper et al., 1986), which is likely to be an upper limit of the mutation frequency for conversion of laci-(am26) to a non-suppressible lacl-. This is in the range of the mutation frequency that we report herein ( ~ 10-6; Table 1). However, there is one important difference between these studies in that we collected mutants after 2 days of growth, while Schaaper et al. (1986) collected mutants after up to 4 days of growth. Phenomenologically, it appears that initially colonies arise from mutants that pre-existed in a culture, while with time another set of mutants appear following plating by a process called "directed mutation" (reviewed in Foster, 1992). We have principally collected pre-existing mutants, in that we see no mutant colonies until after 2 days of growth on lactose minimal plates. (We chose to collect mutants after 2 days, because, ultimately, we are primarily interested in studying carcinogen/mutagen-in-

229

duced mutations, which is expected to occur relatively early.) Furthermore, if we wait 4 days, a much larger number of mut*m colonies do grow. It seems reasonable to imagine that the apparent mutation frequency for the generation of lacicells would be lower than 10 -6 if mutant collection by Schaaper et al. (19861 had been done after 2 days. If true, then this would explain why we do not isolate a high fraction of "false mutants" from cells that are supF+/lacl - (non-suppressible), because they are present at a lower frequency than cells that are supF-/lacl(am26).

Ecaluation of the spontaneous mutational spectra Although our data set is relatively small (142 mutants), some analysis seems warranted. Insertion elements (Schaaper et al., 1986; Shapiro, 1988; reviewed in Galas and Chandler, 19891 and large deletions between direct repeats (AIbertini et al., 1982; Schaaper et al., 1986; Schaaper and Dunn, 19911 often contribute significantly to the spontaneous mutational spectra in E. coll. Insertion elements have been reported to contribute between < 2% and ,,, 60% depending upon the system (reviewed in Galas and Chandler, 1989). One rationale for this variability could be that the fraction of directed mutants, which tend to be base pairing (Foster, 1992), varied in these different studies. The most common insertion element is ISI (Table 2), and the approximate size of these mutant plasmids ( -., 4.3 kb) is consistent with the insertion of IS1 (768 bp) into pUB3 (3553 bp). The sequence in supF that is downstream of the site of insertion almost invariably begins with an A : T base pair, and the hot-spots at bp66, bp91 and bp165 have three A : T base pairs in a row. in fact, these are the only sites in the supF region that have three or more A : T base pairs in a row except for several sites with runs of 5'-(A:T),,-3' sequences (n >i 3). The latter sequences may be poor sites for insertion because they are associated with bent DNA (Olson et al., 1988). Our results are consistent with what has been noted previously about IS1 insertion sites (reviewed in Galas and Chandler, 1989). The site of our IS4 insertion (Table 2) does not appear to conform to the reported consensus sequence (Mayaux et al., 1984), while our IS5

insertion site has a sequence (5'-CTAA-3'), which corresponds to the reported consensus (5'CTAG-3'; Engler and van Bree, 1981; Kleckner, 1981; Galas and Chandler, 1989). Gamma-delta appears to insert in A + T rich regions (Liu et al., 1987), which is consistent with what we obtained. With the possible exception of IS10 insertions at bpll6, the relative frequency at which insertion element mutagenesis occurs does not appear to be significantly influenced by SOS-induction (Table 1). Excluding all IS10 insertions, MF-0.32 x 10 -6 and = 0.32 x 10 -6 is calculated for insertion element mutagenesis without and with SOS induction, respectively. ISI0 (and TNI0) insertion appears to induce the SOS response, so there is precedent for an association between these two phenomena; however, the sites where we observe ISI0 insertions bear little resemblance to the loose consensus sequence, which is 5'-GCTNAGC-3' where the 5'-TNA-3' element is most essential (Kleckner, 1989). The second-most prevalent mutation is a deletion between a direct repeat of the sequence, 5'-'VFCCCTCGTC-3', which is found between bp 68-77 in the control region of supF and bp 120-129 in the supF gene. These mutations appear not to be SOS-inducible, where MF = 0.10 × l0 -~' and = 0.11 x l0 -r' is calculated without and with SOS induction, respectively. Base-pairing mutations appear to increase following SOS induction, where MF--0.09 × 10 -~' and -- 0.28 x l(I -~' is calculated without and with SOS induction, respectively. This difference appears to be statistically significant using a hypergeometric test ( p = 0.04; Adams and Skopek, 1987). (The other cases where SOS induction did not appear to affect insertion element and deletion mutagenesis were confirmed statistically.) SOS induction of spontaneous base pairing mutagenesis has been noted by others (Miller and Low, 1984; Yatagai et al., 19911. The fact that large insertions and deletions dominate the spontaneous mutational spectra may prove helpful, because point mutations often dominate the mutational spectra of mutagens/ carcinogens (see Introduction). This makes it less likely that a significant fraction of spontaneous base-pairing mutations will be found in mutagen/carcinogen-induced mutational spectra,

23O which

is

another

attractive

feature

of

the

ES87/pUB3 system.

Acknowledgements We thank Dr. Thomas Adams (Chemical Industry Institute of Technology) for executing the statistical analyses. We acknowledge the helpful comments of both referees. This work was supported by NIH Grant ES03775 and ACS Grant CN-54.

References Adams, W.T., and T.R. Skopek (1987) Statistical test for the comparison of samples from mutational spectra, J. Mol. Biol., 194, 391-396. AIbertini, A.M., M. Hofer, M.P. Calos and J.H. Miller (1982) On the formation of spontaneous deletions: The importance of short sequence homologies in the generation of large deletions, Cell, 29, 319-328. Balmain, A., and K. Brown (1988) Oncogen¢ activation in chemical carcinogenesis, Adv. Cancer Res., 51, 147-182. Bernelot-Moens, C., B.W. Glickman and A.J.E. Gordon (1990) Induction of specific frameshift and base substitution events by benzo[a]pyrene diol epoxide in excision-repairdeficient Escherichia colt, Carcinogenesis, 11,781-785. Bishop, J.M. (1991) Molecular themes in oncogenesis, Cell, 64, 235-248. Barns, P.A., A.J.E. Gordon and B.W. Glickman (1987) Influence of neighboring base sequence on N-methyI-N'-nilro. N-nitrosoguanidine matagenesis in the lacl gene of Escherichla colt, J, Mol, Biol., 194, 385-390. Burns, P.A., A.J.E, Gordon, K. Kunsmann and B.W, Glick. man (1988) Influence of neighboring base sequence on the distribution and repair of N.ethyI.N'.nitrosourea.induced lesions in Escherichia colt, Cancer Res., 48, 4455-4458. Carothers, A.M., and D. Grunberger (1990) DNA base changes in benzo[a]pyrene diol epoxide-induced dihydrorelate reductase mutants of Chinese hamster ovary cells, Carcinogenesis, 11, 189-192. Chen, R.-H., V.M. Maher and JJ. McCormick (1990) Effects of excision repair by diploid human fibroblasts on the kinds and locations of mutations induced by ( + / - )-7/L 8~t.dihydroxy.gu, I0a-epo~-7,8,9, I 0-dihydrobenzo[a]pyrene in the coding region of the HPRT gene, Proc, Natl. Acad. Sci. (U.S.A.), 87, 8680-8684. Chen, R.-H., V.M. Maher and J.J. McCormick (1991) Lack of a cell cycle-dependent strand bias for mutation induced in the HPRT gene by (+/-)-7~8,8a-dihydroxy-9a,10aepoxy-7,%9,10-dihydrobenzo[a]pyrene in excision repairdeficient human cells, Cancer Res., 51, 2587-2592. Coulondr¢, C., and J.H. Miller (1977) Genetic studies of the lac represser, IV. Mutagenic specificity in the lacl gene of Escherichia coli, J. Mol. Biol., 117, 577-606.

Dower, W.J., J.F. Miller and C.W. Ragsdale (1988) High efficiency transformation of E. colt by high voltage electroporation, Nucl. Acids Res., 16, 6127-6145. DuBridge, R.B., P. Tang, H.C. Hsai, P.-M. Leong, J.H. Miller and M.P. Calos (1987)Analysis of mutation in human cells by using an Epstein-Barr virus shuttle system, Mol. Cell. Biol., 7, 379-387. Eckert, K.A., C.A. Ingle, D.K. IOinedinst and N.R. Drinkwater (1988) Molecular analysis of mutations induced in human cells by N-ethyl-N-nitrosurea, Mol. Carcinogen., 1, 50-56. Eisenstadt, E., AJ. Warren, J. Porter, D. Atkins and J.H. Miller (1982) Carcinogenic epo~ides of benzo[a]pyrene and cyclopenta[cd]pyrene induce base substitutions via specific transversions, Prec. Natl. Acad. Sci. (U.S.A.), 82, 1945-1949. Engler, J.A., and M.P. van Bree (1981) The nucleotide sequence and protein-coding capability of the transposable element IS5, Gene, 14, 155-163. Foster, P.L. (1992) Directed mutation: between unicorns and goats, J. Bacteriol, 174, 1711-1716. Galas, D.J., and M. Chandler (1989) Bacterial insertion sequences, in: D.E. Berg and M.M. Howe (Eds.), Mobile DNA, American Society for Microbiology, Washington, DC, pp. 109-192. Hollstein, M., D. Sidransky, B. Vogelstein and C.C. Harris (1991) p53 mutations in human cancers, Science, 253, 49-53. Horsfall, M.J., A.J.E. Gordon, P,A. Burns, M. Zielenska, G.M.E. van der Vliet and B.W. Glickman (1990) Mutational specificity of alkylating agents and the influence of DNA repair, Environ. Mol. Mutagen., 15, 107-122. gleckner, N. (1981) Transposable elements in prokaryotes, Annu, Rev. Oenet,, 15, 391-4(¼. Kleekner, N. (1989) Transposon Tnl0, in: D.E. Berg and M.M. Howe (Eds,), Mobile DNA, American Society for Microbiology, Washington, DC, pp, 109-192. Kleina, L.G,, and J.H. Miller (1990) Genetic studies of the la¢ represser, XIlI. Extensive amino acid replacements generated by the use of natural and synthetic suppressors, J. Mol, Biol., 212, 295-318. KoffeI-Schwartz, N., J.M. Verdier, M. Biehara, A.M, Freund, M.P. Daune and R.P.P. Fuchs (1984) Carcinogen-induced mutation spectrum in wild-type, urea and umuC strains of Escherichia colt, J, Mol. Biol,, 177, 35-51. Kraemer, K.H., and M.M. Seidman (1989) Use of supF, an Escherichia coli tyrosine suppressor tRNA gene, as a mutagenic target in shuttle-vector plasmids, Mutation Res., 22O, 61-72. Levin, D.E., and B.A. Ames (1986)Classifying mutagens as to their specificity in causing the six possible transition and transversions: A simple analysis using the Salmonella mutagenicity assay, Environ, Mutagen., 8, 9-28. Liu, L., W. Whalen, A. Das and C.M. Berg (1987) Rapid sequencing of cloned DNA using a transposon for bidirectional priming: sequence of the Escherichia colt K-12 avtA gene, Nucleic Acids Res., 15, 9461-9569.

231 Mayaux, J.-F., M. Springer, M. Graffe, M. Fromant and G. Fa,vat (1984) IS4 transposition in the attenuator region of the Escherichia coil pheS, T operon, Gene, 30, 137-146. Mazur, M., and B.W. Glickman (1988) Sequence specificityof mutations induced by benzo[a]pyrene-7,8-diol-9,10epoxide at endogenous aprt gene in CHO cells, Som. Cell Mol. Genet., 14, 393-400. Miller, J.H. (1980) The lacl gene, in: J.H. Miller and W.S. Reznikoff (Eds.), The Operon, 2nd edn., Cold Spring Harbor Press, Cold Spring Harbor, NY, pp. 31-88. Miller, J.H. (1983) Mutational specificity in bacteria, Annu. Rev. Genet., 17, 215-238. Miller, J.H., and K.B. Low (1984) Specificity of mutagenesis resulting from the induction of the SOS system in the absence of mutagenic treatment, Cell, 37, 675-682. Miller, J.H., C. Coulondre and P.J. Farabaugh (1978) Correlation of nonsense sites in the lacl gene with specific codons in the nucleotide sequence, Nature (!_x)ndon)~274) 770775. Olscn, W.K., M.H. Sarma, R.H. Sarma and M. Sundaraiingam (Eds.) (1988) Structure and Expression, Vol. 3, Adenine Press, Schenectady, NY. Pearson, W.R., and D.J. Lipman (1988) Improved tools for biological sequence comparisons, Proc. Natl. Acad. Sci. (U.S.A.), 85, 2444-2448. Richardson, K.K., F.C. Richardson, R.M. Crosby, J.A. Swenberg and T.R. Skopek (1987) DNA base changes and alkylation following in vivo exposure of Escherichia coil to N-methyI-N-nitrosourea or N-ethyI-N-nitrosourea, Proc. Natl. Acad. Sci. (U.S.A.), 84, 344-348. Russel, M., S. Kidd and M.R. Kelley (1986) An improved filamentous helper phage for generating single-stranded plasmid DNA, Gene, 45, 333-338. Sambrook, J., E.F. Fritsch and T. Maniatis (1989), Chapter 13, in: Molecular Cloning: A Laboratory Manual, 2nd

edn., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N-Y, pp. 70--75. Schaal,e,, R.M., and R.L. Dunn (1991) Spontaneous mutation in the Escherichia coil lacl gene, Genetics, 129, 317-326. Schaaper, R.M., B.N. Danforth and B.W. Glickman (1986) Mechanisms of spontaneous mutagenesis: An analysis of the spectrum of spontaneous mutation in the Escherichia coil lacl geue, J. Mol. Biol., 189, 273-284. Shapiro, J.A. (1988) What transposable elements do in bacteria, in: M.E. Lambert, J.F. McDonald and I.B. Weinstein (Eds.), Banbury Report 30: Eukaryotic Transposable Elements as Mutagenic Agents, Cold Spring Press, Cold Spring Harbor, NY, pp. 3-16. Witkin, E.M. (1976) Ultraviolet mutagenesis and inducible DNA repair in Escherichia coil, Bacteriol. Rev., 40, 869907. Yang, J.L., V.M. Maher and J.J. McCormick (1987) Kinds of mutatinn~ formed when a human shuttle vector containing adducts of ( + / - )-7/3,ga-dihydroxy-gc,,10a-epoxy7,8,9,10-dihydrobenzo[a]pyrene replicates in human cells, Proc. Natl. Acad. Sci. (U.S A), 84, 3787 -3791. Yang, J.-L., R.-H. Chen, V.M. Maher and J.J. McCormick (1990) Kinds and locations of mutations induced by ( + / - )-7[3,8a-dihydroxy-9a,lOa-epoxy-7,8,9,10-dihydro-

benzo[a]pyrcne in the coding region of the hypoxanthine(guanine) phosphoribosyltransferase gene in diploid human fibroblasts, Carcinogenesis, 12, 71-75. Yanisch-Perrun, C., J. Vierira and J. Messing (1985) Improved MI3 phage cloning vectors and host strains: nucleotide sequences of M13mpl9 and pUC19 vectors, Gene, 33, 103-l!9. Yatagai, F., M,J. Horsfall and B.W. Glickman (1991) Specificity of SOS mutagenesis in native Ml31acl phage, J. Bacleriol., 173, 7996-7999.