Molec. gen. Genet. 143, 223-232 (1976) © by Springer-Verlag 1976

Isolation of Rec Mutants from an F-prime Merodiploid Strain of Escherichia coli K-12 K.A. Stacey* and Robert G. Lloyd** School of Biological Sciences, University of Sussex, Brighton, England

Summary. This paper describes a method of screening mutagenised populations of an E. coli gala/gal~ Fprime merodiploid for mutants defective in recombination. The method relies on scoring colonies on Eosin-Methylene Blue agar that have fewer than normal numbers of Gal + papillae. With a suitable choice ofgal- mutations most of the papillae arise by recombination and some of those colonies with less than normal numbers prove to be defective in some aspect of recombination or DNA repair. In addition to strains carrying mutations that can be ascribed to known loci, several novel mutant phenotypes were identified.

Introduction Clark and Margulies (1965) demonstrated that genetic analysis could be applied to the process of genetic recombination itself by isolating mutants of E. coIi that were deficient in their ability to yield recombinants after conjugation with an Hfr donor strain. Subsequent studies (see review by Clark, 1973) revealed that the majority of the mutants (Rec-) obtained from 'wild-type' strains could be classified into one of three categories defined by the location of the mutations on the chromosome, i.e. recA, recB or recC. Mutations at two further loci, recG and recH were isolated in one instance (Storm et al., 1971). Extensive studies have revealed that mutations in these recombination genes are pleiotropic (Clark, 1973). In the case of recB and recC the gene product is known and its in vitro properties defined (Oishi, 1969; Goldmark and Linn, 1972), but its role in the *

Present address:

BiologicalLaboratory, The University, Canter-

molecular process of recombination is not at all clear (Hall and Howard-Flanders, 1972; Birge and Low, 1974). Although the recA gene product appears essential for the formation of recombinants, there is no evidence to indicate whether it plays a direct role in recombination or is involved in some other cellular process necessary for recombination to proceed. In any attempt to extend the analysis of recombination mechanisms in 'wild-type' E. coli it would seem prudent therefore to search for mutant phenotypes other than those described by recA, recB and recC. We decided to look for Rec- mutants using a method which in the first instance does not rely upon mating a female strain with an Hfr donor. In stable F-prime merodiploid strains in which both copies of a gene in the duplicated region are mutated, at different sites, and do not complement one another, intragenic recombination between the mutant alleles can be monitored by observing the segregation of progeny that carry at least one wild-type allele for this locus. A mutation in a gene whose product is necessary for recombination might be expected to alter the segregation frequency of wild-type recombinants from such a merodiploid. By screening mutagenized populations of a merodiploid strain for clones that segregate a reduced number of wild-type progeny we thought it possible to isolate Rec mutants. We started this project with two aims in mind; one was to explore the potential of such a method as a means of isolating Rec- mutants and the other to search for novel mutant phenotypes. This report summarises our results which demonstrate the feasibility of the method and contains preliminary information on the mutants obtained. Materials and Methods

bury, Kent. ** P r e s e n t address: Genetics Department, University of Nottinglaam, University Park, Nottingham NG7 2RD.

Strains. The E. coli K-12 strains used are listed in Table la. Fig. 1 showsthe positionof relevantmarkersand Hfr origins of chromo-

224

K.A. Stacey and R.G. Lloyd: Isolation of Rec

Mutants

Table la. Bacterial strains Strain

Genotype

Source or reference

ABlI57

F thr-1 leu-6 proA2 his-4 argE3 thi-1 lacY1 galK2 ara-14 xyl-5 mtl-1 tsx-33 strA31 sup-37 2-

P. Howard-Flanders ~

JC2915

As AB1157 except cysC43

N.S. Willets a

AB2470

As AB1157 except recB21

P. Howard-Flanders a

JC5489

As AB1157 except recC22

M.H.L. Green ~

AB2494

As ABl157 except lex-1

P. Howard-Flanders

$2495

As AB2494 except arg + m e t A -

S. Sedgwick

PAM9951

As ABl157 except arg + m e t A - lon-

M . H i . Green

$9952

As PAM9951 except lac + pro +lon +

KL226 x PAM9951 Pro + Lac + (StrR), nonmucoid

AB2070

F - thi- his- proA2 trp- metE gal- mtl- tonA- tsx- strA-

Complex recombinant of W2915 (B. Bachman, personal communication)

$100 =K12S

F + thi? tonA24 or 27 Su + 2- 2S

N.D. Symondsa

$491

F - his p r o A - galT22 lac- uraP- tonA24 or 27 Str ~ Su 2 2s

Complex derivative of S100 (Lloyd, 1972)

$494

F' as $491 except F-gal- (T12?)

$491 x W4520 F-galazauracil R Lac- p2 s

$53

F+ 2

W4520

Fl gal+ /gal - or F8gal+ /W3104 galT12?

L. Theriot a

E~36

Fts114 lac+ /A (lac-pro)x m

$51

Hfr (Hayes) 2 + thi-1 tel-1

F. Jacob via J. Scaife b

$52

Hfr as $51 except P2 +

From P2 plaques on $51

$78

Hfr (Cavalli) rel-1 tonA22 met- 2 +

b

KL16

Hfr 2- thi-1 rel-1

N.S. Willetts"

KL96

Hfr 2- thi-1 rel-1

K.B. Low"

KL98

Hfr 2 +

N.S. Willetts a

KL983

Hfr as KL98 but 2-

K.B. Low c

PK191

Hfr A (proB-lac) thi-1 tel-1

K.B. Low °

$54

Hfr (Hayes) as $51 except )~-

M . H i . Green

JC5088

Hfr thr-300 ilv-318 spc-300 recA56

M.H.L. Green"

ilvE- lac-

Lloyd (1972)

2~ uraP

Bachmann, 1972. u Obtained from the original MRC Microbial Genetics Unit stocks. c Low, 1973. a

Table lb. Bacteriophages some transfer on the linkage map. Strain $491 was derived from S 100 by sequential selection of spontaneous mutations. The F-galepisome carried by $494 was obtained as a spontaneous segregant in W4520. Strain S100 is probably the Pasadena version of strain K12S (Bachmann, 1972) and we found that it carried a supE mutation. We isolated a spontaneous Su- derivative of this strain and discovered that it was now Gal- and a weak lactose fermenter. Therefore, it is probable that S100 carries the galT22 mutation of the ancestral strain W 1213 (Bachmann, 1972). The strain labelled W4520 in our collection was obtained from Yale via L. Theriot. There is a possibility that this strain is actually FSgal+/W3104 galT12 (Bachmann, 1972). Bacteriophage strains are listed in Table 1 b.

Strain

Characteristics

Source

#2 Plkc 2+ 2vir 2bio-1

Male specific from K12 strain Y10 int- red gain

2bio-ll

int- red- gain-

G.A. Maccacaro S.E. Luria N.D. Symonds S.W. Glover C. Radding (Manly et al., 1969) C. Radding (Manly et al., 1969) G. Lindahl

P2+

-

K.A. Stacey and R.G. Lloyd: Isolation of Rec

lex u v metA ' X ~ argE x 7 " polA~ ~"

thr leu . ~ r A ~ Hayes- -

Mutants

proA lac "~

$78 "~ n ~ l

16

.... PKI

rocC~/~'92E..~

rocB'/ I

(uraP "-~--

I~is

cySU recA

Fig.

1. E. coli K-12 linkage map, modified from Taylor and Trotter (1972). Hfr origins and direction of D N A transfer are indicated by arrowheads

Media. Nutrient broth and agar, tryptone broth and agar, EMB agar, minimal salts agar (Davis), M9 salts m e d i u m and phosphate buffer were described previously (Evenchick, Hayes and Stacey, 1969). PI saline, used to store P1 phage, contained 0.14 M NaC1, 0.01 M MgC12 and 0.005 M CaC12, supplemented with 10 mg gelatine per litre. CAS agar contained 10 g Difco casamino-acids and 12 g Davis agar per litre. Tetrazolium agar was nutrient agar supplemented with 10 g of the required sugar and 0.95 g of tetrazolium chloride per litre. EM9 m e d i u m was M9 glucose supplemented with 0.25% casamino-acids. Chemicals. Streptomycin sulphate was obtained from Glaxo L a b o r a t o r i e s . Mitomycin-C, acridine orange, 2, 3, 5-triphenyltetrazolium chloride, 6-azauracil, calf-thymus D N A and deoxynucleoside triphosphates were obtained from Sigma. Methyl-methane sulphonate (MMS) and N-methyl-N'-nitro-N-nitrozoguanidine (NG) were supplied by Ralph N. Emannel Ltd. 3H-thymidine (methyl-3H, sp. act. 18.4 Ci/m mole) and 2H-dCTP (sp. act. 12 Ci/ m mole) were obtained from the Radiochemical Centre, Amersham. Mating procedures, PI transduction and methods for the analysis of recombinant phenotype have been described previously (Lloyd and Barbour, 1974). In crosses with F - strains resistant to azauracil (uraP), counterselection of Hfr donors was achieved by incorporating 150 gg azauracil per ml of selective agar. Since the inhibitory effect of azauracil is overcome by the presence of uracil, its use was limited to media lacking uracil. General bacteriophage techniques were as described by A d a m s (1959). Zygotic Induction of 2. Zygotes from crosses with Hfr strain $52 were mixed with indicator strain $53 and plated on CAS agar containing azauracil. 2 plaques were counted after overnight incubation. The presence of a P2 phage in the donor ensured that no correction was needed for free 2 phage or spontaneous lysis, and in the absence of sufficient calcium, it was unnecessary to use P2 resistant strains. Spontaneous ;t Induction. Lysogens were obtained from the turbid centres of )o plaques. Strains sensitive to 2vir but resistant to 2 ÷ were assumed to be lysogenic for 2. The free phage assayed in exponential nutrient broth cultures of lysogens was taken as a measure of spontaneous induction. Curing F-prime Strains. Acridine orange curing (Hirota, 1960) was used where appropriate. W h e n this m e t h o d failed (as with

225 m a n y of the apparently Rec mutants) Ftsll4]ac + w a s substituted for the resident episome by mating with strain E'36 at 30 C. The lac + episome was then removed by growing the strains at 42 C overnight. Samples were then streaked out on tetrazolium lactose agar and colonies that looked Lac after incubation were tested for their sensitivity to phage/~2; resistant colonies were considered to be F - . Irradiations. A low pressure germicidal lamp (Hanovia) was used for UV irradiation. Survival was measured as described previously (Lloyd and Barbour, I974). G a m m a - r a y sensitivity was measured by exposing nutrient broth cultures (108 cells per ml) to a 6°Co source (4.6 Kr. per minute) at room temperature with constant aeration. Survivors were assayed on nutrient agar. Thymine Requiring Mutants. These were isolated by the method of Stacey and Simson (1965). Endogenous DNA Degradation. The procedure used was that of Howard-Flanders and Theriot (1966). DNA Polymerase-I Assay. The ability of crude cell extracts to incorporate 3H-dCTP into trichoroacetic acid insoluble material using sonicated calf-thymus D N A as primer was measured by the method of De Lucia and Cairns (1969). Screening Procedure for Rec- Mutants. Nitrosoguanidine mutagenesis was first attempted with 15 gg N G per ml in M9 m e d i u m at 30 C. A second mutagenesis was effected with 30 gg N G per ml in nutrient broth at 37 C. In each case the m u t a g e n was added when the cells were growing exponentially at a density of 108 per ml while the surviving fraction of cells was in the range 1 90%. Samples of the mutagenised cells were spread on E M B galactose agar to give, after incubation, approximately 50 colonies per plate. After 9 days incubation at 37 C those colonies that had developed few or no papillae (Rec- phenotype) were streaked on the same medium to check the phenotype. Each clone of this type was examined for the persistance of the episome by testing for its sensitivity to phage #2 and for additional growth requirements. The strains showing a Rec phenotype were then subjected to a second screening. Each one was grown to saturation in an attempt to produce F - phenocopies, mated with Hfr $78 and the number of Pro + (azauracil resistant) recombinants scored. These strains were also tested for the efficiency of transduction of the pro + allele using phage Plkc grown on strain S100. In addition, each strain was tested for its survival after receiving a UV dose of 500 ergs. Diploid strains that appeared Rec- after conjugation were made F by curing and then crossed with Hfr $78 to score recombination proficiency. F - strains that appeared Rec in this test were crossed with Hfr $52 and the extent of D N A transfer measured by the efficiency of zygotic induction of 2. Infertile strains were discarded. Those strains that showed a five-fold or greater reduction in recombination proficiency, after either conjugation or transduction, or that were UV sensitive were selected for further study.

Results

Isolation of Rec Mutants The rationale for isolating mutants relied on the expectation that a galA/gal~ merodiploid strain that is R e c - would segregate fewer Gal ÷ recombinants than a corresponding Rec + strain and that this difference could be detected by observing the number of Gal + papillae on the surface of colonies grown on EMB-galactose agar. Since defects other than those

226

K.A. Stacey a n d R.G. L l o y d : I s o l a t i o n of Rec

Preliminary Characterisation of the Selected Mutants

Table 2. Strains isolated f r o m $494 Strain

A

Rec P h e n o t y p e o n i s o l a t i o n a

Mutants

UV b

Conjugationx HfrS78

P1 transduction

R2721 R2881 R5421

+ ±

±

S S S

R7061 c R7151 R7721 c R8121

+ -

+ -

S S S S

AdditioI~al properties

Ilv GuaMetLeu-

A = m u t a n t s f r o m m u t a g e n e s i s in m i n i m a l m e d i u m ; B = m u t a n t s from b r o t h mutagenesis. Selection was m a d e for P r o + (azauracil resistant) r e c o m b i n a n t s ( + = R e c +, - = R e c , + = s l i g h t l y R e c - , c o m p a r e d w i t h $491). b The response to a U V dose of 500 ergs (S = sensitive c o m p a r e d with $491). F - on first isolation.

of recombination (e.g. loss of the episome or modification of the gal- alleles) would give the same phenotype, alternative tests for recombination proficiency were needed as a second stage in the screening procedure. The merodiploid strain $494 was therefore constructed. Exponential cultures of this strain grown in nutrient broth contained approximately 0.1% Gal + recombinant cells. Colonies grown on EMB agar became covered with Gal + papillae within ~ 5 days incubation at 37 C. Of 30,000 colonies examined, all developed papillae suggesting that spontaneous loss of the F-gal- episome was a rare event. Strain $494 was treated with nitrosoguanidine and the survivors were subjected to the screening procedure as described in the methods section. Approximately 66,000 colonies were screened, of which 834 showed a decrease in the number of Gal + papillae. The majority of these strains were eliminated during subsequent stages of testing as being of little interest. Most were F- or showed no reduction in recombination proficiency after conjugation or transduction: others did not allow any zygotic induction of 2 and were deemed unable to accept donor DNA during mating. In the initial experiments, all the clones isolated from the mutagenesis in minimal medium that showed an additional nutritional requirement or were resistant to phage ~t2 were discarded. For reasons that will become apparent, clones of this type derived from the broth mutagenesis were retained for further characterisation. The strains that were finally selected as being of interest are listed in Table 2 along with the properties used for their selection.

The major reason for using this method was the belief that it would yield novel mutants. The first task therefore, was to compare the properties of these strains with those of known mutants (for a summary see Clark, 1973). Our results, which are summarised in Tables 3 and 4 and displayed in Figs. 2 and 3, show that some of the mutants were indeed novel. To establish the success of the method each mutant was investigated in further detail and, in particular, an attempt was made to locate the mutant alleles on the genetic map. R8121. The phenotype of this strain strongly suggested that it carried a mutation in the recA gene even though the absence of recombinants from conjugation was probably due to a failure in DNA transfer. P1 phage grown on R8121 was therefore used to transduce the cysC + allele into JC2915 (cysC). 7% (36/506) of the Cys + transductants were both UV sensitive and Rec-. One transductant was purified and its UV survival and recombination proficiency shown to be that expected or a recA- strain. When this strain was mated with Hfr strains KL98 and JC5088 (recA56), the yield of His+(StrR) progeny was less than 0.1% of that obtained with ABl157, but when the donor was KL16, for which recA + is an early marker, the yield was 7% of the control mating. These data point to the presence in strain R8121 of a recAmutation in addition to a defect that prevents chromosome transfer in mating. R2881. This strain was similar in many ways to a RecA- mutant but the severity of the defects was by no means as extreme. Furthermore, 2-1ysogens of R2881 could be induced by UV irradiation (results not shown), a feature that is usually not observed with recA- strains. Although the analysis of recombinants from crosses with Hfr's KL16 and KL98 showed that the allele (or alleles) responsible for the UV sensitive, Rec- phenotype of R2881 was located within the chromosomal region delineated by the origins of transfer of these Hfr strains, attempts to locate the mutation causing UV sensitivity by transduction were unsuccessful. However, when P1 phage grown on R2881 was used to transduce the cysC + allele to JC2915, 4.5% (37/812) of the Cys + transductants proved to have a Rec-, UV resistant phenotype. Five of these transductants were purified and shown to be Rec- (5% of the ABl157 control) when His + (StrR) recombinants were selected in a cross with Hfr KL98. Crosses with Hfr KL16, however, yielded normal numbers of His + recombinants. None of the five strains was significantly more sensitive to UV irradiation than ABl157. The 2 deletion mutants, 2bio-1 and 2bio-ll failed to form plaques on R2881 or any

K.A. Stacey and R.G. Lloyd: Isolation of Rec

227

Mutants

Table 3. Recombination proficiency Strain

Recombination proficiency b

Zygotic induction of 2"

R2721 R2881 R5421 R7061 R7151 R7721 R8121

Hfr Matings °

0.5 0.52 0.49 0.89 0.87 0.7 @.001

x $51

x $78

x KL98

x KL16

0.59 0.01 0.27 0.05 - 0 . 2 0.29 0.0016

0.43 0.02 0.16 0.05 - 0 . 2 0.24 0.0007 -

0.36 0.008 0.29 0.33 0.12 0.00015 -

1.0 0.2 0.37 1.0 0.6 0.06 -

P1 transduction d

Chromosome mobilization ~

0.0005 0.0056 0.5 0.059 0.33 0.0027 0.00005

0.7 0.0023 0.24 0.043 2.2 0.11 -

The number of 2 plaques observed after mating with Hfr $52, expressed as a fraction of that obtained with $491 as recipient. recombinants formed by the mutant b Obtained from the ratio recombinants formed by $491 c Mating times were 35 mins with Hfrs $78, $51 and KL98, and 45 mins with KLI6. Selection was for Pro + (xS51 and $78) or His + ( x KL98 and KL16) and samples were plated without interruption. a S100 was the PI donor and selection made for Pro +. e Chromosome mobilization was measured as the frequency of transfer of the thr + - l e u + region relative to g a l + transfer to ABl157 using F g a l + derivatives of each strain.

10-2 c

-¢ -g 6

_~10-4 "6 u.

Fig. 2. UV survival of exponential broth cultures (dose rate was in the range of 4 - 10 ergs/mmZ/sec.) • $491, * R2721, zx R2881, [] R5421, © R7061, • R7151, • R7721, o R8121

10-6 J

J

40

.J-

l

80

0

UV d o s e

40

in

80

seconds

o

80

o~o ~ ° f

"I"

,

of

A

60

. . . . .

O

JO

40 "O o

20

g

4m'~'-- n -

- -|

-

-

lm -

-

-

-"

°

~-__.=~.~.o__-o

0

I

I

I

1

2

3

0

Incubation (hours)

|

I

I

1

2

3

Fig. 3. D N A breakdown in exponential cultures grown in EM9 containing 5 btCi 3H-thymidine and 200 btg deoxyadenosine per ml. The cells were washed and resuspended in M 9 glucose. Samples were assayed for trichloroacetic acid-soluble and insoluble 3H activity after further incubation. Symbols as for Fig. 2. .......... unirradiated, 500 ergs UV

228

K.A. Stacey and R.G. Lloyd: Isolation of Rec- M u t a n t s

Table 4

Strain

Viabi- H C R ~c M M S ad lity b

N G "~

Spontaneous 2 induction "f

R2721 R2881 R5421 R7061 R7151 R7721 R8121

0.85 0.79 0.9 0.96 0.57 0.43 0.70

0.28 0.64 0•00012 0.62 1.0 1.0 0•0008

1.0 0•0036 2.2 5.5 13.0 11.8 0•00014

1.0 0.85 1.0 0.0007 0.1 (0.000001 1.0 0.0001 0.0015 1.0 1.0 0.03 1.0 0.000i

\O\o\"

,o-1 .; = 10-2

"

._o

"6 m 10-3

a All the numbers refer to the response of each strain expressed as a fraction of the $491 response. b Obtained from the ratio of the viable count on nutrient agar over the n u m b e r of cells observed microscopically in exponential broth cultures. The surviving fraction of a sample of 2vir irradiated with UV (survival on $491 was 0.2 of the unirradiated control). d The surviving fraction of nutrient broth cultures resuspended in phosphate buffer containing 0.025 M MMS, after 60 mins at room temperature ($491 survival was 0.6). The surviving fraction of M9 cultures exposed to 15 gg N G per ml for 20 mins at 30 °. f Obtained from the number of free 2phage per cell in broth cultures.

!

i

80 120 UV dose to phage(sec.) 40

Fig. 4. Survival of plaque formation of UV irradiated 2vir on $491 (e), R5421 (n) and R7151 ( , )

e

m 10-1 ._~

of the Rec transductants of JC2915, whereas 2 + grew normally on these strains. Failure to yield plaques with 2 mutants deleted for int, red and gain functions is characteristic of strains that are recA (Manly, Signer and Radding, 1969; Zissler, Signer and Schaefer, 1971)• It was therefore concluded that R2881 carries an unusual defect at the recA locus. Further studies on this mutation will be reported in a later publication. R7721. The phenotype of this strain corresponded closely with that of strains that are recB- or recC . Assuming that it belonged to this group of mutants, P1 phage was used to transduce the thyA + allele from R7721 to thyA derivatives of ABl157 (rec+), AB2470 (recB21) and JC5489 (recC22). N o Thy + transductants were obtained with AB2470, but 43 were obtained with JC5489 and 257 with ABl157. 50 of the ABl157 transductants were also UV sensitive and R e c - . R7721 was therefore regarded as having a mutation in the recB gene. R7151. Although the recombination proficiency of this strain was only slightly reduced, it was very sensitive to UV irradiation and had a reduced capacity to support the growth of UV irradiated 2vir (Fig. 4). An F-gal ÷ version of R7151 was used as a donor to transfer the argE + allele to AB1157 (argE). 80% (83/104) of the Arg + (Str R) recombinants proved to be as UV sensitive as R7151. P1 phage grown on R7151 was used to transfer metA + to $9952 (metA-), Ten out of 146 Met + transductants were UV sensitive• These results suggested that R7151 car-

\

c 10"2 0

"6 U.

10-3

" ~ ,

0

,

2;

40

K. rads Fig. 5. G a m m a - r a y survival of $491 ( e ) and R715l (~')

ried a uvrA- mutation. However, it was unusually sensitive to UV light and showed some recombination deficiency. One possible explanation was that it also carried a lex- mutation. Some support for this could be derived from a comparison of the UV survival curves of R7151 and $491 in that the final slopes of these curves had about the same relationship to one another as found for ABl157 (uvr + lex +) and ABl157 uvrA- lex (Howard-Flanders, 1968). If R7151 were lex- it should be sensitive to ionising radiation. Figure 5 shows the gamma-ray survival of this strain and, as predicted, it proved more sensitive than $491. Furthermore, P1 grown on R7151 failed to transduce UV resistance into $2495 (lex-). This strain was tentatively concluded to be mutant at both the uvrA and lex loci. R5421. The main features of this strain were its UV sensitivity, the extensive D N A degradation after UV irradiation and its partial H C R - phenotype (fig-

K.A. Stacey a n d R.G. L l o y d : I s o l a t i o n of Rec

229

Mutants

located clockwise from metE as shown on the linkage map. Two of the Met +, UV sensitive transductants were purified and shown to be as sensitive to UV irradiation as R542l. Since the polA gene is also linked to metE, strain JG122 (metE-, polA1) was transduced with P1 grown on R5421. None of 170 Met + transductants became UV resistant. It seemed likely, therefore, that R5421 carried a mutation in the poIA gene. Support for this came when attempts to isolate recB- or recA derivatives ofR5421 failed (polA-recB- and polA- recA- double mutants are i n v i a b l e - M o n k and Kinsross, 1972). When the metE linked mutation was transferred to a strain that is temperature sensitive for recA, the double mutants are i n v i a b l e - M o n k and Kinross, 1972). temperature (Lloyd et al., 1974). To confirm that R5421 is polA-, the D N A polymerase-I activity in crude extracts was measured. The results displayed in Fig. 6 show that R5421 and one of the UV sensitive transductants of AB2070 were deficient in this enzyme activity although preliminary experiments indicated that partially purified extracts of R5421 did contain normal levels of polymerase-I associated 5'-3' exonuclease activity. It was concluded that R5421 carried a polA mutation similar to polA1. R7061. When first isolated, this strain was already F . However, it proved to have a R e c - , UV sensitive phenotype. The recombination deficiency of this strain was more reliably scored when used as a recipient in transduction (partly because spontaneous reversion of the pro- and his- mutations was considerably less than in $491), or when chromosome mobilisation was measured using an F-gaP derivative of R7061 as a donor strain. The yield of recombinants from crosses with Hfr donors proved variable. Table 5 summarises the results of a number of crosses and the analysis of unselected markers. The phenotypesthat could be most reliably scored by replica-plating methods were UV sensitivity, Rec, and, in particular,

C

4

3

/

/

{1.

o

1

?

-1"0

E

O. o

i

°f/

× 2

0Vo--0

i

I

I

;

I

I

2 minutes

4

I

ii I

6

incubation

Fig. 6. D N A p o l y m e r a s e activity in the s u p e r n a t a n t fraction of s o n i c a t e d cells (3 x 108/ml). R e a c t i o n m i x t u r e s (1.0 ml) c o n t a i n e d 0.8 mI extract, 50 Ixg (0.1 ml) s o n i c a t e d c a l f - t h y m u s D N A a n d 0.1 ml ' h o t - m i x ' ( 1 0 0 0 n . m o l e s each of d A T P , d G T P , T T P a n d 50Ix Ci 3H-dCTP). I n c u b a t i o n was at 30 ° C a n d 0.1 m l samples were assayed for t r i c h l o r o a c e t i c a c i d - i n s o l u b l e material. • $491, [] R5421, o M e t + U V R t r a n s d u c t a n t o f A B 2 0 7 0 , A M e t + U V s t r a n s d u c t a n t of AB2070 (P1 g r o w n on R5421

ure 4). Recombination proficiency was only slightly altered. However, colonies of the F-gal- merodiploid showed no Gal + papillae after 9 days incubation on EMB-galactose agar. When this was subsequently checked, papillae were observed if incubation was continued for at least 12 days at 37 C. P1 phage grown on R5421 was used to transfer metE + to AB2070 (metE- ilvE-). Of 435 Met + transductants examined, 297 were Ilv + and 16 were UV sensitive. All the UV sensitive clones were Ilv- suggesting that the allele responsible for UV sensitivity in R5421 was

Table 5. A n a l y s i s of R7061 r e c o m b i n a n t s Donor strain

KL96 KL226 $54 PK191

Selected marker a

His + Pro + Pro + His +

R e c o m b i n a t i o n U n s e l e c t e d m a r k e r analysis ° proficiency b No. tested U V R MC R 0.4 0. l 0.25 0.001

80 . 136 .

80 .

1 .

1 .

UVRq-MC R

-

1

136

1

. 1

.

Rec +

.

~' M a t i n g was for 4 0 m i n s w i t h KL96, K L 2 2 6 a n d PK191, a n d 6 0 m i n s with $54. Zygotes were p l a t e d w i t h o u t i n t e r r u p t i o n on selective m e d i u m c o n t a i n i n g azauracil. u Expressed as a f r a c t i o n of the proficiency of $491 c o n t r o l crosses. c D e t e r m i n e d as described by L l o y d a n d B a r b o u r (1974), U V R = r e s i s t a n t to U V light; M C R = r e s i s t a n t to m i t o m y c i n - C ; Rec t = r e c o m b i n a t i o n proficient in a plate m a t i n g w i t h Hfr P K 1 9 I a n d selecting His + .

230

Table

Strain

K.A. Stacey and R.G. Lloyd: Isolation of Rec

6. A s u m m a r y o f t h e p r o p e r t i e s o f t h e R e c Recombination

proficiency

Mutants

mutants DNA

repair capacity

Location of the m u t a n t allele(s)

Mating a

Transduction a

F-prime a

UV b

MMS a

HCR ~

DNA

breakdown c

R2721

+

-

+

-

+

+

ND

ND

R2881

-

-

-

-

-

+

excessive

R5421

±

±

±

-

-

-

excessive

R7061

-

-

-

-

-

+

reduced

R7151

±

±

+

-

+

-

ND

R7721

-

-

±

-

-

+

much reduced

R8121

-

-

ND

-

-

+

excessive

cysC polA n e a r his a n d n e a r pro i n uvrA a n d lex i n recB i n recA linked to in

a Data from Table 2; b Fig. 2; c Fig. 3 (after UV irradiation). + = wild-type, - ~ defective, ND = not determined.

mitimycin-C sensitivity (R7061 was discovered to be highly sensitive to mitomycin-C). The results indicated that R7061 carried at least two mutations, one near proA and the other near his. When a mitomycinC sensitive, UV sensitive, Rec + (Pro +) recombinant of R7061 (derived from a cross with Hfr $54) was mated with KL96, His + recombinants that inherited the donor allele for UV resistance were also fully mitomycin-C resistant. Furthermore, when a UV resistant, mitomycin-C sensitive (His +) recombinant of R7061 (from a cross with KL96) was crossed with Hfr $54, 60 out of 90 Pro + recombinants also became fully resistant to mitomycin-C. These results are best explained if R7061 carried two mutations each giving rise to a mitomycin-C sensitive phenotype. Some support for this was obtained when it proved impossible to isolate mitomycin-C resistant revertants of R7061. The his linked allele (or another closely linked mutation) seems to be responsible for the UV sensitive phenotype. Significant recombination deficiency was observed only when both mutations were present. Experiments are in progress to obtain a more precise definition of the nature of the defective genes in R7061. R2721. This strain was not studied in detail but we report it here because it proved UV sensitive and did not yield any transductants, although sensitive to P1 phage. The absence of transductants was made more obvious by the very low level of reversion of the auxotrophic mutations in this strain. Subsequent work on this strain (Oliver and Stacey, 1973) has shown the low rate of reversion and the absence of transduction to be due to the same mutation. In addition, P1 phage grown on R2721 appeared not to contain any transducing particles (P. Oliver-personal communication). It, too, is a multiple mutant; the mutation responsible for the UV sensitivity is separate from that which interferes with transduction.

Discussion The idea of using partially diploid strains of E. coli to look for Rec- mutants is not in itself new. Symonds and Erskine used specialised transducing particles of bacteriophage 2 to create diploids of the galactose region, with both donor and recipient gal-, and looked for clones that did not give Gal ÷ recombinants (N. Symonds-personal communication). Meselson (personal communication) replicated colonies from mutagen treated cells of an F - , gal- strain onto a lawn of an F'gal- donor strain and looked for clones that gave no Gal ÷ recombinants on the replica plate. In each case, Rec- mutants were found amongst those strains that gave no Gal ÷ recombinants under these conditions. Our method was somewhat different in that the experimental strain was a stable F-prime merodiploid. Although the method was simple to use and allowed in the first instance for a visual screening for mutants, it is clear, in retrospect, that it could be made less time consuming and possibly more sensitive. The procedure we followed in identifiying strains of interest proved rather tedious for three reasons: a) the high rate of curing as a result of mutagen treatment, b) the probability that many of the F-gal- clones that did not show papillae had acquired additional defects in the gal alleles, and c) failure of the F- phenocopying procedure to give reliable estimates of recombination proficiency. It occurred to us that perhaps the method could be modified so that false Rec clones, due to loss of the F-prime, might be eliminated, e.g. by using and F-prime that carried another gene that would allow selection of merodiploid clones on the indicator plates. We were worried that certain mutants of interest were eliminated before the second stage of the screening, which did not rely on recombination in F-prime merodiploids (such mutants

K.A. Stacey and R.G. Lloyd: Isolation of Rec

Mutants

might be detected by testing recombination at another duplication). We decided to retain the F- segregants from the broth mutagenesis for further testing, a measure that was later justified since two of these (R7061 and R7721) were subsequently shown to be Rec-. It is likely that loss of the episome reflected defective repair of mutagen damage, since repair of the chromosome is obligatory for survival whereas repair of the episome is dispensible (Cerda-Olmeda and Hanawalt, 1967; Takahashi and Barnard, 1967). It is clear that the mutants we obtained were the result of extensive mutagenesis, judging by the high frequency of strains with additional auxotrophic requirements, and in two instances at least (R7061 and R7151) by the presence of more than one mutation affecting D N A repair and recombination. For these and other reasons, the association of phenotypes based on the original mutants with the mutant alleles that have been described should be regarded with caution; their confirmation requires the examination of the phenotypes obtained after transfer of the mutated alleles to unmutagenised strains. The success of the method is illustrated by the mutants we obtained (Table 6). We were not surprised to find recA- strains since recA is needed for recombination between an F-prime and the chromosome (Wilkins, 1969; Hall and Howard-Flanders, 1972). Strain R2881 was considered to carry an unusual allele at the recA locus and subsequent work has tended to support this view (Lloyd; manuscript in preparation). Characterisation of unusual recA mutations might prove useful in defining the role of recA in recombination; it is already clear that this particular mutation causes a delay in the formation of recombinants after conjugation (Lloyd and Low, manuscript in preparation). A mutation in the recB gene was also to be expected since the gene product is required for the recovery of viable recombinants (Birge and Low, 1974). The isolation of a l e x - strain demonstrated the sensitivity of our method because recombination is reduced by a factor of only 3-4 fold in such mutants (Mount, Low and Edminston, 1972). However, we did not investigate the possibility that the Rec- phenotype of the original mutant (R7151) was enhanced by the presence of the additional mutation in uvrA. The polA- strain (R5421) proved somewhat of a surprise; Gross found no interference with recombination by polA1 (Gross and Gross, 1969). Recently Konrad and Lehman reported that another polA mutation (polexl) causes increased recombination in an F-prime merodiploid (Konrad and Lehman, 1974). polexl however abolishes the 5'-3' exonuclease activity associated with the Pol-1 enzyme, leaving the polymerase activity intact, whereas the reverse was observed for the R5421 polA mutation.

231

The two remaining strains (R272l and R7061) carry mutations which we have so far been unable to attribute to any of the known recombination or DNA repair genes. The phenotype of R7061 suggests that it ought to be possible to isolate such a mutant by the method of Clark (Clark and Margulies, 1965; Storm et al., 1971). That it was not may be attributed to the need for two mutational events in widely separated genes, both of which seem necessary to give the Rec- phenotype. These new mutants are currently under investigation since they add to the number of genes that are known to play some part in the complex mechanisms involved in DNA repair and recombination. Acknowledgments. We wish to thank Professor N. Symonds for his encouragements and also the many people who kindly gave us bacterial and bacteriophage strains. This work was carried out in the M.R.C. Group of Bacterial Recombination while one of us (R.G. Lloyd) was an M.R.C. Scholar, and in part at the Microbiology Unit, Oxford University while R.G. Lloyd was supported by a Guiness Fellowship for postdoctoral research.

References Adams, M.H. : Bacteriophages. New York: Interscience Publishers Inc. 1959 Bachmann, B.J.: Pedigrees of some mutant strains of Escherichia coli K-12. Bact. Rev. 36, 525-557 (1972) Birge, E.A., Low, K.B. : The detection of transcribable recombination products following conjugation in rec +, recB- and recCstrains of Escherichia coli K-12. J. molec. Biol. 83, 447-457 (1974) Cerda-Olmeda, E., Hanawalt, P.D.: Repair of DNA damaged by N-met~yl-N'-nitro-N-nitrosoguanidinein Escherichia coll. Mutation Res. 4, 369-371 (1967) Clark, A.J. : Recombination deficient mutants of E. coli and other bacteria. Ann. Rev. Genet. 7, 6 7 - 8 6 (1973) Clark, A.J., Margulies, A.D.: Isolation and characterisation of recombination deficient mutants of Escherichia coli K-12. Proc. nat. Acad. Sci. (Wash.) 53, 451-459 (1965) De Lucia, P., Cairns, J.: Isolation of an E. coli strain with a mutation affecting DNA polymerase. Nature (Lond.) 224, 1164-1166 (1969) Evenchik, Z., Stacey, K.A., Hayes, W.: Ultraviolet induction of Chromosome transfer by autonomous sex factors in Escherichia coll. J. gen. Microbiol. 56, 1 - 14 (1969) Goldmark, P.J., Linn, S. : Purification and properties of the recBC DNase of Escherichia coli K-12. J. biol. Chem. 247, 1849 - 1860 (1972) Gross, G., Gross, M.: Genetic analysis of an E. coli strain with a mutation affecting DNA polymerase. Nature (Lond.) 224, 1166 - 1169 (1969) Hall, J.D., Howard-Flanders, P.: Recombinant F' factors from Escherichia coli K-12 strains carrying recB or recC. J. Bact. 110, 578-584 (1972) Hirota, Y.: The effect of acridine dyes on mating type factors in Escherichia coli. Proc. nat. Acad. Sci. (Wash.) 46, 5 7 - 6 4 (I960) Howard-Flanders, P. : Genes that control DNA repair and genetic recombination in Escherichia coll. Adv. viol. med. Physics 1, 299 (i968)

232 Howard-Flanders, P., Theriot, L. : Mutants ofEscherichia eoli K-12 defective in DNA repair and genetic recombination. Genetics 53, 1137-1150 (1966) Konrad, E.B., Lehman, I. : A conditional lethal mutant of Escherichief coli K 12 defective in the 5'-3' exonuclease associated with DNA polymerase I. Proc. nat. Acad. Sci. (Wash.) 71, 2048-2051 (1974) Lloyd, R.G. : Ph.D. Thesis, Sussex University (1972) Lloyd, R.G., Barbour, S.D.: The genetic location of the sbcA gene of Escherichia coli. Molec. gen. Genet. 134, 157-171 (1974) Lloyd, R.G., Low, B., Godson, G.N., Birge, E.A.: Isolation and characterisation of an Escherichia coli K-12 mutant with a temperature-sensitive RecA- pheotype. J. Bact. 120, 407-415 (I974) Low, K.B.: Rapid mapping of conditional and auxotrophic mutations in Escherichia coli K-12. J. Bact. 113, 798-812 (1973) Manly, K.F., Signer, E.R., Radding, C.M. : Nonessential function of bacteriophage 2. Virology 37, 177-188 (1969) Monk, M., Kinross, J.: Conditional lethality of recA and recB derivatives of a strain of Escherichia coli K-12 with a temperature-sensitive deoxyribonucleic-acid polymerase 1. J. Bact. 109, 971-978 (1972) Mount, D.W., Low, K.B., Edminston, S.J.: Dominant mutations (lex) in Escherichia coli K-12 which affect radiation sensitivity and frequency of ultraviolet light-induced mutations. J. Bact. 112, 886-893 (1972)

K.A. Stacey and R.G. Lloyd: Isolation of Rec- Mutants 0liver, P., Stacey, K.A. : A mutant of E. coli K-12 with an impairment in the formation of stable Pl-mediated transductants. Soc. gen. Microbiol. Proc. 1, 12 (1973) Stacey, K.A., Simson, E.: Improved method for the isolation of thymine-requiring mutants of Escherichia coli. J. Bact. 90, 5 5 4 555 (1965) Storm, P.K., Hoekstra, W.P.M., De Haan, P.G., Verhoef, C.: Genetic recombination in Escherichia coli. IV. Isolation and characterisation of recombination-deficient mutants of E. coli K-12. Mutation Res. 13, 9 - 1 7 (1971) Takahashi, I., Barnard, R.A. : Effect of N-methyl-N'-Nitro-N-Nitrosoguanidine on the F factor of Eseherichia coli. Mutation Res. 4, 111-117 (1967) Taylor, A., Trotter, C.D.: Linkage map of Escherichia coli strain K-12. Bact. Rev. 36, 504-524 (1972) Wilkins, B.M. : Chromosome transfer from F-lac + strains of Escherichia coli K-12 mutant at recA, recB, or recC. J. Bact. 98, 599-604 (1969) Zissler, J., Signer, E.R., Schaefer, F.: In A.D. Hersley (ed.), the bacteriophage lambda, p. 455. (1971)

Communicated by R. Devoret

Received October 3, 1975