[qIGG

Molec. gen. Genet. 177, 177-184 (1979)

© by Springer-Verlag 1979

Genetic Determination of the Donor Properties in Escherichia coli K-12 Phenomena of Chromosome Mobilization and Integrative Suppression S.E. Bresler, S.V. Krivonogov, and V.A. Lanzov Leningrad Institute of Nuclear Physics, Leningrad, USSR

Summary. The chromosome mobilization is the ability of F + donors to introduce part of the chromosome besides F-plasmids into the recipient cells during conjugation. We studied the genetic determination of this phenomenon. Most efficient almost like true Hfr's are F+-cells of the genotype recBC-sbcB belonging to the RecF-recombination type. Their ability to chromosome mobilization is 50 fold higher comparing with wild type F + (of the RecBC-recombination type). This property is fully dependent on the recF gene but does not depend on recL. The donors reeBC-F + but without the mutation sbcB- act in mobilization about 4 times weaker than wild type, Hence we see two main levels of mobilization, quantitatively very different: a recF-dependent and recBCdependent. Both reveal an absolute requirment of the product of recA gene. The efficiency of mobilization of different markers along the chromosome was studied and mapped. The maps were identical, in spite of great difference in absolute frequencies for the RecF- and Rec BCpathways. They are not at all random. The sites of mobilization are coinsident with the points of interaction of the F-factor leading to stable Hfr's. Therefore it is suggested that these sites of predominant mobilization are IS-sequences and that during chromosome mobilization single-strand integration of the F-factor via a semichiasmus is effected. It gives a pulse to initiate D N A transfer into the recipient but is unstable and transient and does not yield true Hfr's. The suppression of the Dna ts phenotype in F + cells due to the integration of an F-plaslnid into the chromosome (integrative suppression) is increased manyfold on the RecF-pathway of recombination. Probably it is a manifestation of mentioned hot spots of recombination. Send offprint requests to: Prof. S.E. Bresler, Leningrad Nuclear Physics Institute, Gatchina, Leningrad district 188350, USSR

The regions fre described earlier (Bresler et al., 1978) and confirmed in this paper are regarded as substrates of some recF-dependent endonuclease of recombination. Probably they coinside with clusters of IS-sequences.

Introduction In our preceding paper (Bresler et al., 1978) we demonstrated the presence of specific sites on the E. coli chromosome with increased frequency of genetic exchanges per unit length. We called these D N A regions fre (frequent recombination exchange). The regionfre I is situated between the point of origin (PO) of the donor Hfr C and the PO of the donor R4. A second region fre H was found between the PO of the donors P72 and R4. We assumed that the fre-regions are rich in specific targets for the action of the recF-dependent endonuclease, which is possibly the initiating enzyme of recA-dependent recombination. The fre I and fre H regions contain clusters of hot points of recombination. In each fre-region we localized one of the hot sites of recombination - f i e 1 near to the gene tsx and fre 2 near to the metB. The enhanced recombinogenic effect in the freregions was observed by two experimental methods. The first one is just postconjugational recombination. The Fre-effect was slightly noticeable in crosses with a wild type (ree +) recipient. According to Clark's terminology the latter belonged to the RecBC-recombination pathway. The effect became very pronounced in recBC-sbcB- recipients (with deficient exonucleases V and I), i.e. on the RecF-recombination pathway. A second mean to detect the Fre-effect was the observation of plasmids instability. If these plasmids containedfre-regions these were totally lost during conjugation with a recBC-sbcB- recipient. Obviously it was due to a recombination event in a recip-

0026 8925/79/0177/0177/$01.60

178

S.E. Bresler et ai. : Genetic Determination of the D o n o r Properties in E. coli K-12

ient specifically adapted to the RecF-pathway of recombination. We found this phenomenon at first in the ORF-1 plasmid bearing a fre I region. Later it was confirmed for a F'14 plasmid bearing a fre H region (results unpublished). The regions fre I and fre H contain clusters of points which serve as PO for different existing Hfrdonors. These are points of integration of the F-factor into the chromosome and it is accepted that they are structurally IS-sequences in the DNA. At least in case of 5 Hfr's this connection was proved. The integration of F proceeded via the sequences IS1, IS2, ~ 6 (Deonier and Davidson, 1977; ref. see Shapiro, 1977). The formation of Hfr's is a recA-dependent recombinational process (Callum and Broda, 1978). It was interesting to investigate if the fre-regions, the substrates of recF-dependent recombination were really identical with the sites of F-factor integration. To solve this question two genetic phenomena could be used - chromosome mobilization and integrative suppression. The former is manifested by some chromosome transfer in case when a population of F ÷ cells is mated with an F - strain. The mechanism of this phenomenon is not yet elucidated. It requires a functional recA gene in the donor (Clowes and Moody, 1966; M o o d y and Hayes, 1972). It was suggested that chromosome mobilization is caused by the formation of temporary unstable Hfr's in a population of F ÷ cells (Broda, 1967; Curtiss and Renshow, 1969; Curtiss and Stallions, 1969). It is quite natural to suppose that the integration of F-factors in these transient unstable Hfr's is effected in the same sites, where true stable Hfr's are built. Integrative suppression is a phenomenon related to chromosome mobilization. It is manifested by the ability of F÷-cells with prohibited initiation of D N A replication at nonpermissive temperatures because of a thermosensitive dnaA ts mutation to replicate at the expence of F-integration and the creation in this way of a new replicon with its initiation site independent of the dnaA-function. The resulting suppressed cells are true Hfr's (Nishimura et al., 1971). Chromosome mobilization and integrative suppression are in principle different from the phenomena used earlier for the demonstration of the Fre-effect because recombination proceeds in this case in the donor cells before conjugation and involves both the F-factor and the chromosome. As will be shown both effects are enhanced considerably on the RecF-pathway of recombination. Materials and Methods Symbols of markers are denoted according to B a c h m a n n et al. (1976). Bacterial strains used in this work are listed in Table 1. Bacteriophage MS2 was used for identification o f donor cells. -Me-

dia. A slightly modified minimal medium TPG, a maximal medium AP, selective agar, and conditions of mating were described earlier (Bresler et al., 1968). Transduction by phage Plvir was performed as earlier (Bresler et al., I978).

Experimental Results

a) Genetic Determination of Chromosome Mobilization First of all we compared the efficiency of chromosome mobilization for donors with rec-enzymes characteristic of different recombination pathways, i.e. the donor AB 1157 F ÷ (RecBC-pathway) and JC 7623 F ÷ (RecF-pathway). Both the donor and recipient X 5036 were grown on the medium AP/5, mixed in the ratio 1 : 10 and mated during 60 rain. Conjugation was interrupted by vibration and the titer of recombinants Pro ÷ was determined. Counterselection of donors was effected using the auxotrophic markers A r g - and His-. It must be noted that the donors used were occasionally P r o A - but ProC ÷. After conjugation we selected for the resulting recombinants Arg ÷ His ÷ ProA-- ProC ÷. The yield of recombinants was normalized to the number of viable donors. The efficiency of mobilization on the RecF-pathway was found almost 2 orders of magnitude higher than for common wild type strains (RecBC-pathway). The normalized yield of recombinants was comparable to that in a mating of regular Hfr's with the same recipient, i.e. 1 to 5%. Then we looked into the process when only the ProC character was used for selection. To this purpose we transduced the ProA ÷ marker into the donor AB 1157 F ÷ (strain ECK 061 F ÷) and into JC 7623 F ÷ (strain ECK 064 F ÷) and repeated the measurements of the efficiency of chromosome mobilization. The absolute yield of recombinants increased slightly but the relative efficiency of mobilization on both pathways of recombination remained t h e same. In the following work we used mainly matings with originally employed Hfr's according to the above mentioned scheme. The results for donors of different genotypes are listed in Table 2. We took arbitratily as 100% the efficiency of mobilization on the RecF-pathway of recombination. If we pay attention to the order of magnitude of the measured values we can denote 4 levels of mobilization efficiency. The first and highest one corresponds to the RecF-pathway. It is interesting that an additional mutation in recL, a gene of the same pathway, does not alter the high efficiency. A second level, 50 times lower, is characteristic of wild type cells or the RecBC-pathway. In this case additional recF- mutation is of secondary importance. A third level, 4 to 6 times lower than the second

S.E, Bresler et al. : Genetic Determination of the D o n o r Properties in E. coli K-12

179

'Table 1, Bacterial strains used in the work Strain

Genotype

Source

F+~ W A 993 LC 179

gal met hsdM hsdR thr leu lac thyA mal dnaA46 ilv thi

Dr. B. Wolf Dr. L. Caro

F'-94

F'-lac + in F - prototroph

Dr. W. Hayes

thr ara leu proA tsx lae his str argE thi AB 1157 but reeA13 AB 1157 but recB21 recC22 sbeBI5 JC 7623 but reeL152 JC 7623 but argE + xyl + metB reeF143 E C K 033 but reeBC + JC 7623 but his + sbcB + JC 7523 but nalB ~ recA13 A B 1157 but proA + xyl + argE + dnaA46 JC 7623 but proA + xyl + argE + dnaA46 AB 1157 b n t p r o A ÷ AB 1157 but leu + AB 1157 but his + JC 7623 but proA ÷ JC 7623 but leu + JC 7623 but his + ara leu lac purE gal trp his argG str x y l mtl ilv metA thi as LC 102 but his + argG + as LC 102 but leu ~ as LC 102 but argG + l a c Z p r o C tsx trp str ara leu tonA tsx proC purE trp argA thyA str xyl mtl ilv metE thi

Dr. P. Howard-Flanders Dr. P. Howard-Flanders Dr. A.J. Clark Dr. A.J. Clark This laboratory This laboratory This laboratory S. T a m m , this laboratory LC 1 7 9 x E C K 061 LC 179 x E C K 064 Plvir(Hfr C) x AB 1157 Hfr C x A B 1157 Plvir(Hfr C) x AB 1157 Plvir(Hfr C) x JC 7623 Hfi" C x JC 7623 Plvir(Hfr C) x JC 7623 Dr. L. Caro

F : AB 1157 AB 2463 JC 7623 JC 8171 E C K 033 E C K 035 E C K 039 E C K 04I E C K 042 E C K 043 E C K 061 E C K 062 E C K 063 E C K 064 E C K 065 E C K 066 LC 102 E C K 067 E C K 068 E C K 069 X 5036 KS 28

as as as as as as as as as as as as as as as

PK Hfr PK Dr. Dr.

191 x L C 102 C × L C 102 191 x L C 102 F. Jacob G. Smirnov

Table 2. Relative yield of recombinants Pro + in crosses of F + donors of different rec-genotypes with the recipient X 5036 proC

Table 3. Relative yield of recombinants Pro + in crosses of donors F'-lac of different rec-genotype with the recipient X 5036 proC

Strain containing the plasmid F +

rec-geuotype

Relative yield of recombinants

Strain containing the plasmid F'lac

rec-genotype

Relative yield of recombinants

JC 7623 JC 8171 AB 1157 E C K 035 E C K 039 E C K 033 AB 2463 E C K 04l

recB recB ree + sbeB reeB reeB reeA reeB

100 (1-5%) 96

AB 1157 JC 7623 E C K 035 E C K 039 E C K 033

rec + recB recC sbcB sbcB reeF reeB recC recB reeC sbcB reeF

100 (0.5%) 26 152 22

recC sbcB recC sbcB recL reef recC recC sbcB r e e f recC sbcB recA

2 1.1

0.5 0.3 0.02 0.02

4

The n u m b e r in brackets shows the absolute yield of recombinants normalized to the n u m b e r of donors

The number in brackets shows the absolute yield of recombinants normalized to the n u m b e r of donors

one is found in case of a recB recC- mutations on the RecBC-pathway or alternatively in case of a recFmutation on the RecF-pathway. Finally a fourth very low level is observed in recA cells. It is 4 orders of magnitude lower than the highest one. This confirms the absolute dependence of the process on the recA function. The comparison of the levels shows that the highest of them is reeF-dependent and the

usual one depends on the integrity of genes recB and recC i.e. on exonuclease V. The four levels o f c h r o m o s o m e mobilization found for different rec-genotypes o f the donors are characteristic only of the integration of isolated F-factors. When we tried a plasmid F'lac we got drastically different results. They are presented in Table 3 and obtained according to the same experimental scheme.

180

S.E. Bresler et al. : Genetic Determination of the Donor Properties in E. coli K-12

These data are in good agreement with those published by Willetts (1975) for the plasmid F'gal. It is well known that the level of chromosome mobilization caused by the plasmid F'lac on the RecBC re-. combination pathway is 2 orders of magnitude more important than that caused by the plasmid F (Evenchik et al., 1969). In this case mobilization occurs preferentially by integration of the plasmid via the homology region lac. In our case the yield of Pro + recombinants if using a F'lac plasmid is 0.5%, i.e. 25 times higher than during mobilization by means of the F-factor. In Table 3 this figure is arbitrarily taken as 100%. All other values characterize different damages in the genes responsible for recombination. The figures are somewhat levelled because of the competition of the nonspecific recBC-dependent and sitespecific recF-dependent integration (the latter is seemingly restricted to the IS or fre regions and is characteristic of the F-factor, as will be shown later).

b) Curing the Exconjugants from Free Plasmids by Means of Acridine Orange (A.O.) The high level of mobilization demonstrated by F + donors on the RecF-pathway of recombination may be taken as indication that a considerable part of the cell population became true Hfr's. Against this suggestion is the observation that the transmission of free F-plasmids from a F + recBC-sbcB- donor remains very frequent, the same as for rec + donors. Still it was important to verify that in case of strain JC 7623 F + there is no excessive Hfr-formation comparing to the control strain AB 1157 F +. For this purpose both strains were grown in the presence of different amounts of A.O., an agent that stops the replication of the F-plasmid but does not interfere with the replication of the cell chromosome with an integrated F-factor. Cultivation of bacteria was performed for 6-8 h on the medium AP (pH 7.6) containing a definite concentration of A.O. Afterwards the cells were cloned on agar AP/2. For each concentration of A.O. (30, 60, 90, 120, 150, 175 gg/ml) 30-50 clones were assayed for the presence of F-factor using the male-specific phage MS2. In the control (RecBCpathway) we obtained a gradual decrease of F + cells with increasing concentration of the dye. In the population of donors of the RecF-pathway up to 90% of the F-factors remained intact even at 120 gg/ml of A.O. But at 150 gg/ml an abrupt drop of the concentration of F + cells was observed (residual number about 4%) but the general viability of cells was not markedly reduced. At 175 pg/ml the growth of cells was strongly inhibited. It is not clear up to now why the conditions o f F-elimination are so criti-

cal in the latter case but in any way we conclude that the cells JC 7623 F + contain the episome in the free state because it may be eliminated by A.O.

c) Mobilization Maps of recFand recBC-Dependent Donor Cells The mobilization of the chromosome occurs unequally in different regions of the genetic map (Curtiss and Renshow, 1969). The efficiency of mobilization of different markers varied. Obviously this reflects the nonequivalence of different chromosome regions in the interaction with the F-plasmid. We remind that just this nonequivalence in the recombination frequency allowed to define the hot spots fre I and fre II. The question arises if there is some relation between the fre-regions and the distribution of prevailing mobilization sites. For this purpose we studied what can be called mobilization maps - the distribution of mobilization efficiency along the E. coli chromosome. The resulting maps are produced in Fig. 1 both for the donors F + of the RecF and RecBC type. The following strains were used as donors: J C 7 6 2 3 F +, ECK 0 6 4 F +, ECK 0 6 5 F +, E C K 066 F +, all of the RecF-pathway and AB 1157 F +, ECK 0 6 1 F +, ECK 0 6 2 F +, ECK 0 6 3 F + of the RecBC-pathway. As recipients served the strains: X 5036 with the markers P r o C - , T r p - ; ECK 067 with the markers P r o C - , P u r E - , T r p - , Ilv-, M e t A - , L e u - ; ECK 068 with P r o C - , A r g G - ; ECK 069 with P r o C - , His- and KS 28 with P r o C - , T h y A - , T r p - . The matings were performed for a short time (15 min) in the medium AP/5. Counterselection was effected using two amino acid auxotrophs (see legend to Fig. 1). For the evaluation of mobilization efficiency for a specific marker the yield of multiple recombinants was normalized to the yield of ProC + recombinants. This procedure corrected the data for some difference in conjugation ability of different parental strains. In some cases for control the efficiency of mobilization was estimated in different independent crosses (see legend to Fig. 1). The relative statistical error did not exceed 40% of the measured value. This is insignificant because the efficiencies change manyfold along genetic map. The mobilization maps for the RecF-type donors (Fig. l A) and the RecBC-type donors (Fig. 1 B) were found quite similar but only different in absolute scale. The mobilization efficiency drops from its maximal value near Ilv till its minimal value near ThyA. They differ more than 20 times. The map demonstrates the existence of at least two regions of the chromosome, where mobilization is extremely efficient. One of these regions coinsides with fre I and

S.E. Bresler et al. : Genetic Determination of the D o n o r Properties in E. coli K-12

t00-

181

A

025

0.50 o c:, -,=,

P. 0.25 N

E

O

10

20

30

/~0

50

rio

70

BO

90

100

0.020

B

"C,

= o

11015

"6 N

:--T. O

~:

0.010

0.005

I

I

I

I

_eu PFoCPurE (fFe I) IS3 IS2

I

Trp (f re Ill)

,

His

ThyA AFgG

IS3

R5 R4 10R21i0R7C OR10RSZ,

Itv MetA {Ire IT)

IS2 KL99 B~

PKI91;. iB311 /d,

~a

Leu C

KL10AB312i PK3KL'251i72 liZ, all AB313 Ra-2 D

•

•

Ib

,I

•

2

7

1

g

7

E

Fig. 1 A - E . Dependence of the mobilization efficiency of donors belonging to the RecF (A) and RecBC (B) pathways of recombination on the position of the markers on the E. coli chromosome. Below the points are mapped where the PO of different stable Hfr's are situated. C The m a p of all known Hfr's (Low, 1972; B a c h m a n n et al., 1976). D The same according to Curtiss and Stallions (1969). E The same according to Tresguerres et al. (1975). IS-elements are shown according to Shapiro (1977), Deonier and Davidson (1977) and Deonier (see Guyer and Clark, 1977). All markers used for analysis and markers used for counterselection are listed in the following Table containing all crosses used for the estimation of mobilization efficiency according to the yield of corresponding recombinants

Crosses Donors

Markers used for counterselection

ProC + PurE + Trp +

His

ArgE-

ProC ProC ProC ProC ProC

His Thr ThrHis His-

ThrLeuArgEArgEThr

Recipients

RecFpathway

RecBCpathway

E C K 064

E C K 061

ECK ECK ECK ECK ECK

ECK ECK ECK ECK ECK

064 064 066 065 064

Selected markers

061 061 063 062 061

X5036 E C K 067 KS 28 E C K 067 E C K 068 E C K 069 E C K 067 KS 28

+ + + ÷ +

Ilv + MetA + ArgG ÷ His + Leu + ThyA ÷

182

S.E. Bresler et al. : Genetic Determination of the Donor Properties in E. coti K-12

fre III (the latter was localized very approximately in our earlier experiments between genes Trp and His (Bresler et al., 1978)). The second region isfre H. The similarity of both maps A and B in spite of the great difference in the absolute frequencies of recombination shows that mobilization is initiated in specific spots of the chromosome notwithstanding of what recombination type is the donor strain. In the lower part of Fig. 1 we quote the data for the localization of various Hfr's (the PO are given). Diagram C is the map of all known Hfr's, D is the map of the most frequent types found in experiments on the efficiency of Hfr-formation, E is the same but observed by a different group of authors. The figures on the diagrams under the arrows refer to the numbers of detected Hfr's of a specific type. The comparison of these maps with the mobilization map shows the identity of the regions of efficient mobilization with clusters of sites of frequent integration. We may assume that unstable Hfr's, which cause chromosome mobilization arize in the same points of the chromosome where stable Hfr's are formed. Taking into account that the integration of F-factors proceeds via the IS sequences we assume that the temporary transient products of F-recombination, which are the origin of chromosome mobilization, occur in the same hot spots or substrates of the recF-dependent recombination endonuclease.

d) Sites of Prevailing Mobilization Inside the Regions fre I and fre II We found a hot point fre 1 in the region fre I near to the gene tsx. It was located on the basis of following observations. 1) It is transmitted by Hfr ORT, which is able to transfer only a minor part of fre I. 2) The genetic instability of the plasmid F ' O R F - t is connected with this point. A more thorough analysis of plasmid instability performed in our laboratory by S.E. T a m m showed the presence of a second hot point fre 3 in the same region located near gene lac. Both hot points correspond quite accurately to the most probable points of integration of F-factors into the chromosome with the formation of stable Hfr's OR7 and OR21. It was interesting to localize the site of prevailing mobilization with higher precision to compare with the Hfr-map. For this purpose we studied the genetic linkage of the markers ProC + and PurE + situated on both sides of Tsx s. Among 400 mobilization products with the phenotype ProC ÷, selected after the mating ECK 064 F + x ECK 067 proC purE, 82% were phenotypically PurE +. Hence the linkage coefficient ~tp,,.e/p,.oC= 0.82 (400). If we perform primary selection for PurE + we obtain gproC/

purE=0.55 (400). Hence we conclude that the site of prevailing mobilization is situated somewhat to the right of purE gene and probably coincides with the PO of donor HfrC. This confirms very well the main idea (see map D in Fig. 1). A similar analysis was performed in case offre II. By means of the cross ECK 064 F + x ECK 067 ilv metA the linkage coefficients ~titv/metA=0.9 (200) and I.tmetA/ilv=0.52 (200) were determined. They showed that the predominant site of mobilization is situated to the left side of gene ilv, probably coinsiding with PO of Hfr K L 25 or AB 312 (see the maps C and D o f Fig. 1). The sitesfre 1 andfre 2 described earlier do not coinside fully with the sites of prevailing F-mobilization but are situated near enough. In general the data confirm the suggestion that clusters of sites exist on the chromosome which guarantee a highly efficient recombination with the exogenote. Evidently all these sites are substrates but differ somewhat in their affinity to the initiating enzyme (endonuclease) of recombination.

e) Integrative Suppression on the RecF-Recombination Pathway In contrast to chromosome mobilization the phenomenon of integrative suppression is the result of the formation of stable Hfr's. In the work of Tresguerres et al. (1975) mild conditions were established for the selection of suppression products of the phenotype Hfr Dna + (minimal medium, 40 ° C). The formation of Hfr's in a population of cells F + dnaA t~ occurs in these conditions in the usual sites of predominant integration of the F-factor (see map E in Fig. 1). An alteration of conditions (maximal medium, 42 ° C) yields products of suppression with unusual PO (Nishimura et al., 1971). This is due to low viability of some types of Hfr's. We emphasize that in our experiments (mating of ECK 064 F ÷ x ECK 067 with subsequent selection for markers ProC +, Trp +, Ilv +) the efficiency of chromosome mobilization did not depend on the nutritional medium (we tried the minimal medium TPG, the maximal media AP, AP/5, L-broth). If the processes of mobilization and integrative suppression are connected, the latter must be strongly enhanced on the RecF-pathway of recombination. We verified this conclusion comparing two strains: E C K 0 4 3 dnaA46 (RecF-pathway) and E C K 0 4 2 dnaA46 (RecBC-pathway). They were selected from the parental strains ECK 064 and E C K 061 correspondingly. Integrative suppression was assayed on minimal media at 40 ° C. We measured the number of viable Dna + cells in the culture of ECK 043 F + dnaA

S.E, Bresler et al. : Genetic Determination of the Donor Properties in E. coli K-12

©

46 and ECK 042 F + dnaA46 and normalized the data to the percentage of viable cells in a population of corresponding F - dnaA46, i.e. recipients ECK 043 and E C K 042. For the RecF-pathway of recombi6.10-3 nation this ratio amounted to 3.10 - 6 - 2 0 0 0 , for the 8- 10 - 4 RecBC-pathwaY9.10_6=90.

It follows that

183

F÷ ce[/

the

RecF-pathway activates integrative suppression about 20 times. Evidently in this case we deal with another recF-dependent phenomenon.

Semichiasmus

*1

\

unstable

(~--.

Hfr

Mobiiisation

Discussion We shall try to express a hypothesis about the mechanism of chromosome mobilization. For this purpose we must sum up the main facts and points of view. 1. Chromosome mobilization is due to the formation of transient unstable Hfr's. F + strains were described totally unable to form stable Hfr's. Nonetheless mobilization proceeds with the usual efficiency (Curtiss and Renshow, 1969, 1969a). 2. Mobilization is in principle a recombination process. Therefore it is dependent on the genes recA, recB, recC, and recF. 3. The formation of unstable Hfr's occurs in the same sites, where stable Hfr's are built, i.e. via the IS-sequences. This follows from the good correspondence of the mobilization map with the map of stable Hfr-formation. Integrative suppression, a process mediated by stable Hfr's, has the same genetic determination, it is enhanced on the RecF-pathway. 4. For mobilization there is a strict requirement for the integration of the marker oriT of the F-plasmid into the chromosome (Willetts, 1972). We may assume that the transfer of D N A into the recipient starts from the site oriT like in conventional conjugation (Guyer and Clark, 1977). The sequence of tragenes is transferred the last after the entire cell chromosome. 5. During conjugation only one D N A strand is transferred into the recipient (Ohki and Tomizawa, 1968; Rupp and Ihler, 1968). It is quite natural to suppose that recombination of the episome with the chromosome starts with the physical assimilation of one D N A strand, the formation of a semi-chiasmus. 6. The most probable scheme of recombination assumes at first the formation of a semi-chiasmus which may be eventually transformed into a full chiasmus giving a stable recombinant or cleaved into parental cells (see review of Holliday, 1974). 7. Analysis shows that the formation of semichiasmus is by one order of magnitude more frequent than that of full chiasmus (Kushev, 1974).

f

I

Chiasmus stobie

Conjugation

Fig. 2. Scheme of the formation of unstable Hfr's with subsequent chromosome mobilization or of stable Hfr's with subsequent chromosome transfer during conjugation

Based on these facts is the model presented in Fig. 2. A similar model was proposed earlier (Evenchik et al., 1969) to explain UV-induced mobilization. The interaction of the F-plasmid with the chromosome starts through a IS-element. It begins with a nick effected by a recF-dependent enzyme. A semichiasmus is formed and the crossover is able to migrate in the range of the IS sequences, which are homologous in both the chromosome and epiSome. In this state the transfer of one D N A strand from the site oriT of the F-factor may be initiated and draws alone the chromosome genes of the cell. But the semi-chiasmus is unstable, in most cases it is reversed to the initial state of separated chromosome and F-factor. In some cases it is transformed into a full chiasmus leading to a stable Hfr. In this paper we considered two phenomena chromosome mobilization and integrative suppression in addition to two studied earlier - hot spots of recombination (regions fre) and unstability of plasmids F' containing fre sites (Bresler et al., 1978). All four effects are enhanced on the RecF-recombination pathway. All depend on the integrity of the reeF gene. All are based on the site-specific recombination. It is natural to conclude that the fre-sites, the substrate

184

S.E. Bresler et al. : Genetic Determination of the Donor Properties in E. coil K-12

for the recF-dependent endonuclease, are clusters of IS-sequences on the cellular chromosome. It is i n t e r e s t i n g to n o t e t h a t t h e e f f i c i e n c y o f t h e s e p r o c e s s e s is i n d e p e n d e n t o f t h e g e n e r e c L , w h i c h is important for the recF-dependent branch of repair ( R o t h m a n a n d C l a r k , 1977; R o t h m a n et al., 1979) a n d s e e m s t o be f u n c t i o n a l l y c o n n e c t e d w i t h t h e g e n e p o l A ( S m i r n o v et al., 1973; M a t t e r n a n d H a u t m a n , 1974; R o t h m a n a n d C l a r k , 1977). T h e f a c t t h a t t h e p r o d u c t o f r e c L is n o t r e q u i r e d f o r t h e s i t e - s p e c i f i c r e c o m b i n a t i o n e m p h a s i z e s its p e c u l i a r f e a t u r e s . P r o b a b l y in t h e s e cases c h r o m o s o m e c l e a v a g e is p r e c i s e l y h o m o l o gous and does not involve subsequent repair DNA synthesis.

References Bachmann, B.J., Low, B.K., Taylor, A.L.: Recalibrated linkage map of Escherichia coli K-12. Bacteriol. Rev. 40, 116 (1976) Bresler, S.E., Krivonogov, S.V., Lanzov, V.A. : Scale of the genetic map and genetic control of recombination after conjugation in Escherichia coli K-12. Hot spots of recombination. Molec. Gen. Genet. 166, 337 (1978) Bresler, S.E., Lanzov, V.A., Lukjaniec-Blinkova, A.A.: On the mechanism of conjugation in Escherichia coli K-12. Molec. Gen. Genet. 102, 269 (1968) Broda, P.: The formation of Hfr strain in Escherichia coli K-12. Genet. Res. Camb. 9, 35 (1967) Callum, J., Broda, P. : Chromosome transfer from "recombinationdeficient" strains of Escherichia coli K-12. Int. Cong. Genet. C-10, Moscow, 1978 Clowes, R.C., Moody, E.E.: Chromosomal transfer gene and Hfr formation by F in Escherichia coli K-12. Genetics 53, 717 (1966) Curtiss, R., Stallions, D.R. : Probability of F integration and frequency of stable Hfr donors in F + populations of Escherichia coli K-12. Genetics 63, 27 (1969) Curtiss, R., Renshaw, J.: Kinetics of F transfer and recombinant production in F + x F - matings in Escherichia coli K-12. Genetics 63, 39 (1969) Curtiss, R., Renshaw, J. : F ÷ strains of Escherichia coil K-12 defective in Hfr formation. Genetics 63, 7 (1969a) Deonier, R.C., Davidson, N.: The sequence organization of the integrated F-factor in two Hfr strains of Escherichia coli. J. Mol. Biol. 107, 207 (1977) Evenchik, Z., Stacey, K.A., Hayes, W.: Ultraviolet induction of chromosome transfer by autonomous sex factors in Escherichia coIi. J. Gen. Microbiol. 56, 1 (1969) Holliday, R. : Molecular aspects of genetic exchange and gene conversion. Genetics 78, 273 (1974)

Horii, Z.L., Clark, A.J.: Genetic analysis of the RecF pathway of genetic recombination in Escherichia coti K-12: isolation and characterization of mutants. J. Mol. Biol. 80, 327 (1973) Guyer, M.S., Clark, A.J.: Early and late transfer of F genes by Hfr donors ofE. coli K-12. Molec. Gen. Genet. 157, 215 (1977) Kalyaeva, E.S., Danilevskaya, O.N. : Fertility factor of Escherichia coli K-12 and its role in bacterial chromosome mobilization. Molec. Biol. (USSR) 11, 957 (1977) Kushev, V.V. : Mechanisms of genetic recombination. New York : Plenum Press 1974 Low, B.K.: Escherichia coli K-12 F-prime factors, old and new. Bacteriol. Rev. 36, 587 (1972) Mattern, I.E., Houtman, P.C.: Properties of uvrE mutants of Escherichia coli K-12. II. Construction and properties ofpol and rec derivatives. Mutation Res. 25, 281 (1974) Moody, E.E., Hayes, W.: Chromosome transfer by autonomous transmissible plasmids: the role of the bacterial recombination (rec) system. J. Bacteriol. 111, 80 (1972) Nishimm-a, Y., Caro, L., Berg, C.M., Hirota, Y.: Chromosome replication in Escherichia coli. IV. Control of chromosome replication and cell division by an integrated episome. J. Mol. Biol. 55, 441 (1971) Ohki, M., Tomizawa, J.: Asymmetric transfer of DNA strands in bacterial conjugation. Cold Spring Harbor Symp. Quant. Biol. 33, 651 (1968) Rothman, R.H., Clark, A.J. : Defective excision and postreplication repair of UV-damaged DNA in a recL mutant strain of E. coli K-12. Molec. Gen. Genet. 155, 267 (1977) Rothman, R.H., Margessian, L.J., Clark, A.J. : W-reactivation of phage lambda in recF, recL, uvrA and uvrB mutants of E. coli K-12. Molec. Gen. Genet. 169, 279 (i979) Rupp, W.D., Ihler, G. : Strand selection during bacterial mating. Cold Spring Harbor Symp. Quant. Biol. 33, 647 (1968) Shapiro, J.A.: Maps: F, the E. coli sex factor, in DNA insertion elements, plasmids and episomes, Bukhari, A.I., Shapiro, J.A., Adhya, S.L. (eds.), New York: Cold Spring Harbor Lab. 1977 Smirnov, G.B., Filkova, E.V., Skavronskaya, A.L., Saenko, A.D., Sinzinis, B.I. : Loss and restoration of viability of E. coli due to combinations of mutations affecting DNA polymerase I and repair activities. Molec. Gen. Genet. 121, i39 (1973) Tresguerres, E.F., Nandadasa, H.G., Pritchard, R.H. : Suppression of initiation-negative strains of Escherichia coli by integration of the sex factor F.J. Bacteriol. 121, 554 (1975) Willetts, N.S. : Location of the origin of transfer of the sex factor F.J. Bacteriol. 112, 773 (1972) Willetts, N.S.: Recombination and the Escherichia coli K-12 sex factor F.J. Bacteriol. 121, 36 (1975)

Communicated

b y D. G o l d f a r b

Received July 3, 1979