Molec. gen. Genet. 150, 171- 181 (1977) © by Springer-Verlag 1977
Plasmid Replication and Hfr Formation in Strains of Escherichia coli Carrying seg Mutations A.F. Jamieson and P.L. Bergquist Department of Cell Biology, University of Auckland, Auckland, New Zealand
Summary. Several conditional-lethal mutations that do not permit the replication of F-factors of Escherichia coli K-12 are located at a site called seg. This gene is located on the E. coli chromosome between serB and thr. It is unrelated to other known genes involved in D N A replication. Strains carrying seg mutations were unable to replicate F'-lac ÷, several F'gal ÷ s, F'-his + and bacteriophage 2 at 42 °. However, neither phage T4, ColE1, nor any of the R factors tested were prevented from replicating at 42 ° C. When the kinetics of the loss of F-primes is studied in seg strains, it is found that the rate of curing depends on the size of the plasmid, larger F factors curing faster than smaller ones, and that Hfrs are formed at high frequencies. The Hfrs showed both F-genote enlargement and normal transfer of chromosomal markers. The F-genotes are unstable and segregate chromosomal markers at high frequencies. Some orthodox Hfrs were examined, and two that were known to revert to the F ÷ condition relatively frequently were found to generate enlarged F-genotes on mating, whereas two strains that were very stable with respect to reversion to the F÷ state did not show F-genote formation. F-genote formation from seg Hfr strains is dependent on a functional recA gene, as F-genote formation was not seen with a seg-2, recA-1 Hfr. This is in contrast to F-genote enlargement shown by both orthodox Hfrs and an Hfr strain constructed by integration of a temperature-sensitive F'-gal +, whose F-genote enlargement is Rec-independent. Thus there may be more than one mechanism for the formation of enlarged F-genotes.
Introduction A property of most of the plasmids of the Enterobacteriaceae is their ability to replicate autonomously
and be subsequently distributed to daughter cells at cell division. A n u m b e r of conditional-lethal mutations defective in replication functions have been recorded (reviewed by Willetts, 1972). In m a n y examples, these temperature-sensitive mutations have been assigned to sites on the plasmid, although some mutations are located on the bacterial chromosome (Cuzin and Jacob, 1967; Bergquist and Adelberg, 1972; H a t h a w a y and Bergquist, 1973). Jacob et al. (1963) first demonstrated that integration of F'ts-lac + could be observed in a fraction of the cells of a population as a result of exposure to the non-permissive temperature. Bergquist and Adelberg (1972) isolated a temperature-sensitive F'-gal + plasmid and showed that Hfr formation could occur in a similar manner to the integration of F'-lac + . Their strain, PB 15, showed an anomalous pattern of behaviour in several respects. It was unusually stable with respect to excision of the integrated F'-gal ÷ during vegetative growth. This strain carries two copies of the galactose operon and an hypothesis was put forward that postulated an inversion of one of the galactose operons was responsible for trapping the integrated F'-gal + and preventing its excision (Adelberg and Bergquist, 1972). On mating, PB 15 generated not only the expected chromosomal recombinants with the recipient but also gave rise to plasmid-like elements by a recA-independent event. This phenomenon was termed F-genote enlargement. The enlarged episomes showed instability of the extrachromosomal genetic markers that they carried, since segregation was observed at high frequency unless selection was maintained. H a t h a w a y and Bergquist (1973) described several other independently isolated conditional-lethal mutations that were carried on the plasmid that affected F-replication. Hfrs formed by integration of these mutant F'-gal ÷'s also showed F-genote enlargement, even though some of the strains they constructed did not have an inverted duplication of the galactose
172
A.F. Jamieson and P.L. Bergquist: Plasmid Replication and Hfr Formation m seg Strains
genes. Thus F-genote enlargement is not a phenomenon unique to PB15. Moreover, an Hfr was constructed by integration of wild-type F'-gal + into a strain carrying a chromosomally-located mutation that prevented F plasmid replication. This strain also showed F-genote enlargement. Hathaway and Bergquist (1973) concluded that F-genote enlargement occurred by way of a recombination event between the integrated F8 plasmid and a sex factor affinity site on the chromosome. They inferred that this behaviour was unique to Hfr's formed from F8. The chromosomal site, termed seg, has now been mapped with some precision (Jamieson and Bergquist, submitted for publication) and has been shown to be unrelated to several adjacent genes which affect chromosomal D N A replication. This paper reports an investigation of the ability of a number of alleles of the seg locus to maintain various plasmids of Escherichia coliand an examination of the phenomenon of F-genote enlargement in seg and wild-type Hfr strains. We conclude from our experiments that F-genote enlargement is unrelated to either the chromosomal or plasmid temperature-sensitive mutations and is a generalised phenomenon that is linked in some way to orthodox F' production in E. coli by either type I or type II excision events (Low, 1972).
Materials and Methods Bacterial Strains and F-primes The bacterial strains employed are listed in Table 1. They are all derivatives of Escherichia coli K-12 except for Escherichia coli C, which was used as an indicator in some experiments with bacteriophage 7. The origin of the various F'-gal + plasmids is confused. The experiments of Bergquist and Adelberg (1972) and Hathaway and Bergquist (1973) used an F'-gal + present in strain AB2605, which was called " F 8 of Hirota in strain W3104 of Echols". However, this F'-gal + has a substantially lower molecular weight than F8 and does not carry tolAB, nadA or aroG. In this paper it is called F8-4 (allotted to us by B. Bachmann, Coli Genetic Stock Center, Yale University). It m a y be identical with the deleted derivative of F8, F8-13, which also lacks the same genetic markers, and which was obtained by Ohtsubo from Fukasawa (our interpretation of information received from B. Bachmann). F8-4 has a molecular weight of about 63 x 106 daltons. The plasmid called F8 in this paper is the F8 of Hirota which carries tolAB,nadA and aroG. It is the same F'-gal that has been comprehensively investigated by Sharp et al. (1972) and has a molecular weight of 78 x 106 daltons. The F'-gal ÷, 2 att, bio ÷, previously called F100 in H a t h a w a y and Bergquist (1373), is F100-2 of Ohtsubo and Hsu, since it does not carry the genetic markers in the fep, leuS, lip region (B. Bachmann, personal communication). It has a molecular weight of about 150x 106 daltons. The F'-lac + used is identical to the F42 studied previously by Sharp et al. (1972) and has a molecular weight of about 97 x 106 daltons (Hu et al., 1975). All F-primes and Hfrs used, as well as the locations of the genetic markers discussed in this paper are shown in Figure 1.
hsmdnaClsegthr..polB ""'L/"
°
In~,~ dnaH.F100
"
' fepi ~ nadA F8 ;.~aroG ~"gal ~chlA
xyl• strA~60 ~
recA~~,~_~i tryAguaA
F8-4
/Irp
his F103
Fig. I. Genetic m a p of Escherichla coli K-12 based on Taylor and Trotter (1972). The relative positions of the genetic loci used in this study are indicated. The F-primes and the points of origin and transfer directions of the Hfr strains employed in this paper are shown inside the circle representing the chromosome of E. coli
Media, Mating Conditions and Genetic Techniques Media, mating conditions, acridine orange curing and the scoring of unselected markers was described in Bergquist and Adelberg (1972). Temperature-induced loss of F's was carried out as described by Hathaway and Bergquist (1973). The scoring of the seg mutation was as described by Jamieson and Bergquist (submitted for publication). The burst size of 2 was determined as described by Paul and Inouye (1974). The seg mutations were transferred from the original strains in which they had been isolated by mutagenesis with N-methyl-N'nitro-N-nitrosoguanidine in the following manner. A spontaneous thr m u t a n t of the recipient strain was isolated by ampicillin selection (Miller, 1972). A suitable F' (usually F8-4 or F42) was transferred into the thr strain, and it was then transduced to Thr + with a P1 lysate prepared on the seg strain. After purification, the Thr ÷ transductants were screened for seg by their inability to maintain F' replication at 42 °.
Col and R Factors ColE1 was transferred into PB966 (seg-2, galT12) by mobilisation with F'-gal + (F8-4). The strain, which now carried both F'-gal + and ColE1 was incubated with PB966 for 20 h at 34 ° C. The exconjngants were spread onto Luria plates and Gal + colonies that carried a colicin factor were identified by using the triple layer technique (Ozeki et al., 1962). The presence ofseg-2 was confirmed by the demonstration that F'-gai + did not replicate an the selected clones at 42 ° C.
Construction of Hfr Strains Hfr strains carrying seg mutations were constructed in the following manner. F', seg strains were spread to the appropriate M a c C o n k e y
A.F. Jamieson and P.L. Bergquist: Plasmid Replication and Hfr Formation in seg Strains
173
Table 1. List of bacterial strains Strain
Mating type
C h r o m o s o m a l markers a
Other Properties
Derivation
PB741
Hfr
thi-28, A(proB-lac)XIII
sup-56 (unmapped amber), p.o.66;
*PK191, B. Low
transfers~his, tyrA, lys ... p.o. 53; transfers,--his, trp, lac ...
*KL983, B. Low
T2-r, Cavalli-type, transfers
*KL226, B. Low
PB753
Hfr
PB756
Hfr
xyl-7, lacYlorZ4, mglP4 rel-1, tonA22
Plasmid Replication and Hfr Formation in Strains of Escherichia coli Carrying seg Mutations A.F. Jamieson and P.L. Bergquist Department of Cell Biology, University of Auckland, Auckland, New Zealand
Summary. Several conditional-lethal mutations that do not permit the replication of F-factors of Escherichia coli K-12 are located at a site called seg. This gene is located on the E. coli chromosome between serB and thr. It is unrelated to other known genes involved in D N A replication. Strains carrying seg mutations were unable to replicate F'-lac ÷, several F'gal ÷ s, F'-his + and bacteriophage 2 at 42 °. However, neither phage T4, ColE1, nor any of the R factors tested were prevented from replicating at 42 ° C. When the kinetics of the loss of F-primes is studied in seg strains, it is found that the rate of curing depends on the size of the plasmid, larger F factors curing faster than smaller ones, and that Hfrs are formed at high frequencies. The Hfrs showed both F-genote enlargement and normal transfer of chromosomal markers. The F-genotes are unstable and segregate chromosomal markers at high frequencies. Some orthodox Hfrs were examined, and two that were known to revert to the F ÷ condition relatively frequently were found to generate enlarged F-genotes on mating, whereas two strains that were very stable with respect to reversion to the F÷ state did not show F-genote formation. F-genote formation from seg Hfr strains is dependent on a functional recA gene, as F-genote formation was not seen with a seg-2, recA-1 Hfr. This is in contrast to F-genote enlargement shown by both orthodox Hfrs and an Hfr strain constructed by integration of a temperature-sensitive F'-gal +, whose F-genote enlargement is Rec-independent. Thus there may be more than one mechanism for the formation of enlarged F-genotes.
Introduction A property of most of the plasmids of the Enterobacteriaceae is their ability to replicate autonomously
and be subsequently distributed to daughter cells at cell division. A n u m b e r of conditional-lethal mutations defective in replication functions have been recorded (reviewed by Willetts, 1972). In m a n y examples, these temperature-sensitive mutations have been assigned to sites on the plasmid, although some mutations are located on the bacterial chromosome (Cuzin and Jacob, 1967; Bergquist and Adelberg, 1972; H a t h a w a y and Bergquist, 1973). Jacob et al. (1963) first demonstrated that integration of F'ts-lac + could be observed in a fraction of the cells of a population as a result of exposure to the non-permissive temperature. Bergquist and Adelberg (1972) isolated a temperature-sensitive F'-gal + plasmid and showed that Hfr formation could occur in a similar manner to the integration of F'-lac + . Their strain, PB 15, showed an anomalous pattern of behaviour in several respects. It was unusually stable with respect to excision of the integrated F'-gal ÷ during vegetative growth. This strain carries two copies of the galactose operon and an hypothesis was put forward that postulated an inversion of one of the galactose operons was responsible for trapping the integrated F'-gal + and preventing its excision (Adelberg and Bergquist, 1972). On mating, PB 15 generated not only the expected chromosomal recombinants with the recipient but also gave rise to plasmid-like elements by a recA-independent event. This phenomenon was termed F-genote enlargement. The enlarged episomes showed instability of the extrachromosomal genetic markers that they carried, since segregation was observed at high frequency unless selection was maintained. H a t h a w a y and Bergquist (1973) described several other independently isolated conditional-lethal mutations that were carried on the plasmid that affected F-replication. Hfrs formed by integration of these mutant F'-gal ÷'s also showed F-genote enlargement, even though some of the strains they constructed did not have an inverted duplication of the galactose
172
A.F. Jamieson and P.L. Bergquist: Plasmid Replication and Hfr Formation m seg Strains
genes. Thus F-genote enlargement is not a phenomenon unique to PB15. Moreover, an Hfr was constructed by integration of wild-type F'-gal + into a strain carrying a chromosomally-located mutation that prevented F plasmid replication. This strain also showed F-genote enlargement. Hathaway and Bergquist (1973) concluded that F-genote enlargement occurred by way of a recombination event between the integrated F8 plasmid and a sex factor affinity site on the chromosome. They inferred that this behaviour was unique to Hfr's formed from F8. The chromosomal site, termed seg, has now been mapped with some precision (Jamieson and Bergquist, submitted for publication) and has been shown to be unrelated to several adjacent genes which affect chromosomal D N A replication. This paper reports an investigation of the ability of a number of alleles of the seg locus to maintain various plasmids of Escherichia coliand an examination of the phenomenon of F-genote enlargement in seg and wild-type Hfr strains. We conclude from our experiments that F-genote enlargement is unrelated to either the chromosomal or plasmid temperature-sensitive mutations and is a generalised phenomenon that is linked in some way to orthodox F' production in E. coli by either type I or type II excision events (Low, 1972).
Materials and Methods Bacterial Strains and F-primes The bacterial strains employed are listed in Table 1. They are all derivatives of Escherichia coli K-12 except for Escherichia coli C, which was used as an indicator in some experiments with bacteriophage 7. The origin of the various F'-gal + plasmids is confused. The experiments of Bergquist and Adelberg (1972) and Hathaway and Bergquist (1973) used an F'-gal + present in strain AB2605, which was called " F 8 of Hirota in strain W3104 of Echols". However, this F'-gal + has a substantially lower molecular weight than F8 and does not carry tolAB, nadA or aroG. In this paper it is called F8-4 (allotted to us by B. Bachmann, Coli Genetic Stock Center, Yale University). It m a y be identical with the deleted derivative of F8, F8-13, which also lacks the same genetic markers, and which was obtained by Ohtsubo from Fukasawa (our interpretation of information received from B. Bachmann). F8-4 has a molecular weight of about 63 x 106 daltons. The plasmid called F8 in this paper is the F8 of Hirota which carries tolAB,nadA and aroG. It is the same F'-gal that has been comprehensively investigated by Sharp et al. (1972) and has a molecular weight of 78 x 106 daltons. The F'-gal ÷, 2 att, bio ÷, previously called F100 in H a t h a w a y and Bergquist (1373), is F100-2 of Ohtsubo and Hsu, since it does not carry the genetic markers in the fep, leuS, lip region (B. Bachmann, personal communication). It has a molecular weight of about 150x 106 daltons. The F'-lac + used is identical to the F42 studied previously by Sharp et al. (1972) and has a molecular weight of about 97 x 106 daltons (Hu et al., 1975). All F-primes and Hfrs used, as well as the locations of the genetic markers discussed in this paper are shown in Figure 1.
hsmdnaClsegthr..polB ""'L/"
°
In~,~ dnaH.F100
"
' fepi ~ nadA F8 ;.~aroG ~"gal ~chlA
xyl• strA~60 ~
recA~~,~_~i tryAguaA
F8-4
/Irp
his F103
Fig. I. Genetic m a p of Escherichla coli K-12 based on Taylor and Trotter (1972). The relative positions of the genetic loci used in this study are indicated. The F-primes and the points of origin and transfer directions of the Hfr strains employed in this paper are shown inside the circle representing the chromosome of E. coli
Media, Mating Conditions and Genetic Techniques Media, mating conditions, acridine orange curing and the scoring of unselected markers was described in Bergquist and Adelberg (1972). Temperature-induced loss of F's was carried out as described by Hathaway and Bergquist (1973). The scoring of the seg mutation was as described by Jamieson and Bergquist (submitted for publication). The burst size of 2 was determined as described by Paul and Inouye (1974). The seg mutations were transferred from the original strains in which they had been isolated by mutagenesis with N-methyl-N'nitro-N-nitrosoguanidine in the following manner. A spontaneous thr m u t a n t of the recipient strain was isolated by ampicillin selection (Miller, 1972). A suitable F' (usually F8-4 or F42) was transferred into the thr strain, and it was then transduced to Thr + with a P1 lysate prepared on the seg strain. After purification, the Thr ÷ transductants were screened for seg by their inability to maintain F' replication at 42 °.
Col and R Factors ColE1 was transferred into PB966 (seg-2, galT12) by mobilisation with F'-gal + (F8-4). The strain, which now carried both F'-gal + and ColE1 was incubated with PB966 for 20 h at 34 ° C. The exconjngants were spread onto Luria plates and Gal + colonies that carried a colicin factor were identified by using the triple layer technique (Ozeki et al., 1962). The presence ofseg-2 was confirmed by the demonstration that F'-gai + did not replicate an the selected clones at 42 ° C.
Construction of Hfr Strains Hfr strains carrying seg mutations were constructed in the following manner. F', seg strains were spread to the appropriate M a c C o n k e y
A.F. Jamieson and P.L. Bergquist: Plasmid Replication and Hfr Formation in seg Strains
173
Table 1. List of bacterial strains Strain
Mating type
C h r o m o s o m a l markers a
Other Properties
Derivation
PB741
Hfr
thi-28, A(proB-lac)XIII
sup-56 (unmapped amber), p.o.66;
*PK191, B. Low
transfers~his, tyrA, lys ... p.o. 53; transfers,--his, trp, lac ...
*KL983, B. Low
T2-r, Cavalli-type, transfers
*KL226, B. Low
PB753
Hfr
PB756
Hfr
xyl-7, lacYlorZ4, mglP4 rel-1, tonA22
