Formation of Δtra F′ plasmids: Specific recombination at oriT

J. Mol. Biol. (1985) 186, 267-274

Formation of Atra F’ Plasmids: Specific Recombination at oriT Burton Horowitz and Richard C. Deoniert Molecular

Biology, D&artment of Biological Sciences of Southern California, University Park Los Angeles, CA 90089-1481, U.S.A.

University

(Received 8 March

1985)

Atra F’ plasmids can be isolated from matings between Hfr donors and recA- recipients, with selection for transfer of proximal chromosomal genes. Previous experiments indicate that F DNA from the neighborhood of the transfer origin up to the proximal junction with the chromosomal DNA is present on these plasmids, together with chromosomal segments. some of which belong to distinct size classes. We have sequenced across the novel joints contained in five Atra FproA+ plasmids and in five Atra FpurE’ plasmids, and we have compared these with the F sequence near oriT and with a chromosomal site near purE. The previously reported specificity in formation of some of these classes is confirmed at the nucleotide sequence level. The F DNA in nine of these novel joints extended beyond the nicking sites identified by others in do&T+ bacteriophages up to a position between two sequenced oriT- mutations. Small plasmids containing these novel joints are mobilized in trams by ~0x38 at frequencies less than 5 x IO-’ times the mobilization frequencies for similar plasmids that contain oriT. The relations of these findings to the location of the nicking site at oriT are discussed.

independent Atra F’ plasmids from Hfr strains in which F was integrated at one of the IS3 elements clockwise after proA or purE in the Escherichia coli K- 12 chromosome. Restriction analysis indicated that in many of these, Fsptra DNA up to the oriT region was replaced by distinct size classes of chromosomal DNA, indicating specificity in the formation of these plasmids. Here we present the nucleotide sequence at the novel joint between chromosomal DNA and F DNA near oriT for 10 Atra FpurE+ or Atra FproA+ plasmids. Mobilization analyses for small plasmids containing these novel joints have also been conducted to determine whether they are oriT+.

1. Introduction Some F’ plasmids are formed by reciprocal recombination events that excise F plasmid DNA fused to adjacent chromosomal markers (see Low, 1972, for a review). A second class of F’ plasmids (Atra F’ plasmids) results from matings between an Hfr donor and a recA-, F- recipient (Low, 1968). These plasmids have the tra operon of F deleted and carry proximal chromosomal genes from the Hfr parent (Guyer & Clark, 1976). Normal F transfer is initiated at a specific locus on F (oriT) and one of the DNA strands is t’ransferred to the recipient with the 5’ terminus leading (see Willetts & Wilkens, 1984, for a review). DNA complementary to the transferred strand is synthesized in the recipient. The transfer process ends with the circularization and establishment of the DNA in the recipient; these processes are known as repliconation (Clark & Warren, 1979). The four Atra FargG+ plasmids studied by Guyer & Clark (1976) and Guyer et al. (1977) each had different deletion endpoints on F, two of which fell near oriT. Hadley & Deonier (1980) isolated 23

2. Materials and Methods (a) Bacterial strains, bacteriophage and plasmids

The bacterial strains, plasmids and bacteriophages used are listed in Tables 1 and 2. Plasmid and bacteriophage constructs are shown diagrammatically in Fig. 2. Their construction is explained in the Results. (b) Media Ml3 phage were propagated in media described by Messing (1983). Matings were conducted in L broth (Lennox, 1955). For donor contraselection in matings promoted by ~0x38, L plates supplemented with 25pg streptomycin/ml or 50 pg ampicillin/ml and 25 pg streptomycin/ml were used.

t Author to whom reprint requests should be addressed at: Molecular Biology, ACBR 406me 1481, University of Southern California, University Park, Los Angeles. CA 90089-1481, U.S.A. 0022-2836/85/220267-08

$03.00/O

267

0

1985 Academic

Press Inc. (London)

Ltd.

Table 1 Bacterial Strain

strains Source

Genotype

N. Davidson

1~. purE, thi, ytr, recA

IGDllll

Meacock, A132463

thr-I, leu-6’3 thi-1. argEJ, his-4, proAZ, lac YI, galK2, ara-14, m&I, ~~1-5, rpsL31, tsx-33, i- supE44. rwrl Id

AR1 157

Same as AU2463

.JMlO3

H 11101

A(lac-proA, B), thi, rpsL, endA, sbcBI5, hsdR4, supE (F’ traD36, proAB, lacIZAMI5) thi, ku., recA, hsdF1, hsdM, pro

RI)17

thi-I, ~1-1 “?“, A(proB-lac),,

(Hroda 1971)

&

Guyer (Howard-Flanders & Theriot,. 1966)

U. Bachmann (Dewitt & Adelberg, 1962) BRL (Messing, 1983)

recA +

except

Ilr A-, supE44, rec,456

Insertional inactivation of the sd gene on pED961 was scored by using minimal medium (Clowes & Hayes, 1968) containing 100 pg sulfamethazine/ml (Brown & Willetts, 1981). (c) Chemicals,

M.

(Reference)

R. Baker (Dagert $ Ehrlich, 1979) Deonier & Mirels (1977)

(e) Restriction enzyme digestion and ligations All digestions and ligations were conducted in Trisacetate buffer (O’Farrell et al., 1980) using equimolal amounts of vector and insert termini and a total cohesive-end concentration of 0.04~~. DNA from the ligation solut,ion was transformed into either JM103 OI HBlOl by using the CaCl, treatment protocol of Dagert & Ehrlich (1979). Subcloning of novel joint fragments and derivatives was done in a similar manner.

reagents and enzymes

Antibiotics were purchased from Sigma Chemical Corp. (St, Louis). Restriction endonucleases, DNA, polymerases and bacteriophage T4 DNA ligase were purchased from New England Biolabs (Beverly, Mass.) and Pharmacia PL Biochemicals, Inc. (Piscataway, N.J.). Deoxynucleoside triphosphates and the 2,3’-dideoxynucleoside triphosphates used for sequencing were from Pharmacia PI, Biochemicals. [a-32P]dATP was purchased from New England Nuclear (Boston, Mass.). All chemicals and reagents were analytical grade.

(f) Hybridization

procedures

DNA from agarose gels was transferred to nitrocellulose filters by the method of Southern (1975). Hybridization probes were fragments containing novel joints or the BgZII fragment 14 of F cloned into M13mp9. Probe DNA was labeled by the primer extension method (Messing, 1983) using [a-32P]dATP as the labeling nucleotidn. Hybridization was performed at 42°C in 50% formamide and 10 x Denhardt’s solution (Denhardt,. 1966) with 3 x SSC (SSC is 0,15 M-NaCl, 0.015 M-sodium citrate, pH 7.0), 50 pg of salmon sperm DNA per ml, and O.lo/, sodium dodecyl sulfate.

(d) Plasmid DNA preparation F’ plasmid DNA and large quantities of small plasmids or Ml3 RF1 were prepared as described (Hadley & Deonier, 1979). Otherwise, a rapid microscale procedure was used (Klein et al., 1981).

Table 2 Plasmids Name

Property

and bacteriophages

or insert

Source (Reference)

l’lnsmids ~0x38

pRH126 pRH 127

pRHl29 pRH135 pRH128 pRHl12 pRH 113 pRHl14 pRH115 pRH 116

tru+, ysAh-a FpurE + Atra FpwE + Aha FpurE+

(Hadley & Deonier, 1980) (Ha&y (Hadley (Ha&y

FpurE+ Atra FpurE+

Atra Ah

(Hadley (Hadley (Hadley (Hadley

FproA +

Atra FproA + Alra FproA+ Atra

et al., 1980)

(Guyer

FproA+

& Deonier, & Deonier,

1980) 1980)

& & & & &

1980) 1980) 1980) 1980) 1980)

Deonier, Deonier, Deonier, Deonier, Deonier,

(Hadley & Deonier, 1980) (Hadley & Deonier, 1980) R. Skurray (Skurray et al., 1976) N. Willetts (Brown & Willetts. 1981)

pRS27

Atra FproA + pSClO1 + EcoRI f6 and f15 of F otiT+

pED961

pUR322

pUC8

ApR; multiple sequence cloning vector ApR; multiple sequence cloning vector

SRL (Vieira & Messing, 1982) BRL (Vieira & Messing, 1982)

M13mp8

Ml3 multiple sequence cloning vector

M13mp9

Ml3

BRL BRL

JAW9

+ sul’ gene on SalI-EcoRI

fragment

from

R46

Bacteriophages multiple

sequence

cloning

vector

(Messing (Mewing

& Vieira, & Vieira,

1982) 1982)

269

Recombination at oriT (g) C-test analysis Complementary sequences from segments inserted in Ml3 DNA in the opposite orientation were detected by the C-test (Messing, 1983). Approximately lo8 of each of the phage to be tested and a probe phage were combined, and the DNA was released from the virions by adding sodium dodecyl sulfate to O2o/0. The solution was incubated at 65°C for 1 h. and the DNA was resolved electrophoretically with a 0.7% agarose gel in Trisacetate buffer (0.04 M-Tris, 0.02 iv-sodium acetate. 0.005 M-EDTA, pH 82). Hybridization of the insert in the phage being tested to that of the probe phage produces a component that migrates more slowly than either of the parental circles. (h) DNA sepence

analysis

DKA sequences were determined by the dideoxy method of Sanger et al. (1977). Single-stranded DNA from Ml3 phage clones was prepared as described by Messing (1983). The pentadecamer universal sequence primer purchased from Bethesda Research Laboratories (Gaithersburg, Md.) was used. The DNA sequence of the opposite strand (Hong. 1981) was obtained by using a pentadecameric primer synthesized with a Syster 1450 DlVA synthesizer. (i) Mobilization analysis Overnight cultures from single colonies of strains to be mated were diluted 100 times with fresh L broth and allowed to grow at 37°C to a concentration of 2 x 10’ bacteria/ml. A 0.1 ml sample of the donor culture was mixed with 1 ml of the recipient (either AB1157 or AB2463) in L broth. This mixture was incubated for either 30 min or 90 min at 37°C with slow shaking in a 25 ml flask. The mating was interrupted by agitation, and the mixture was centrifuged at 10,000 revs/min in a Sorvall SS34 rotor and suspended in 1 ml of L broth. The mating mixture was again centrifuged as described above and the pellet was resuspended in L broth. The suspension was then diluted and plated on appropriate media. The total number of recipients was calculated from the viable count on L plates supplemented with streptomycin. Transfer of ~0x38 was determined by testing 15 to 30 Stf colonies for f2 sensitivity by the cross-streak method. Transfer of the small plasmid was evaluated from the number of AmpRStrs colonies on appropriately supplemented L plates. The mobilization frequency is expressed as the number of recipients per ml containing the small plasmid divided by the number of recipients per ml containing the conjugative plasmid.

chromosomal segments from the proA region (Hadley & Deonier, 1980). We refer to these two classes of plasmids as “specific” Atra F’ because the chromosomal segments within each class appear to be identical. The chromosomal segments are bounded by t,he chromosomal site for F integration and the chromosomal sites that become fused with the oriT region of F (see Fig. 1). Other plasmids analyzed in this study are the Atra FpurE+ plasmid pRH128 (derived from Hfr P3) and the Atra FproA+ plasmid pRH112 (derived from Hfr OR1 1). These plasmids contain chromosomal segments that differ from those of the specific class of Atra F’ in their respective phenotypic selection group (purE+ or proA ‘). We refer to these as a “non-specific” class of Atra F’ plasmids. Hybridization and restriction

analyses

as depict#ed in Figure

by the dideoxy primer.

(a) Construction of Ml3 phage and small plasmids containing novel joints from Atra F’ plasmids By restriction analysis, the Atra FpurE+ plasmids pRH126, pRH127, pRH129 and pRH135, which were derived from Hfr P3, were found to contain identical chromosomal segments from the purE region. Similarly, the independently isolated Atra FproA+ plasmids pRHll3, pRH114, pRH115 and pRH116, each of which was derived from an Hfr in which F is integrated at the chromosomal IS3 element a3p3 (cf. Fig. l), contain identical

that

the

2, and they

Rglll

were sequenced

method using the Ml3

%I11 fragments

containing

universal

the specific Atra

FpurE’ novel joints were cloned into the Bglll site in the plasmid pED961, and then they were subcloned into M13mp9 at its BamHI site and sequenced as described (Fig. 2). The chromosomal segment of the non-specific Atra FpurE+ plasmid pRH128 extends to a site further count,erclockwise from purE than do the segments contained on the other Atra FpurE’ plasmids (Fig. 1). Therefore, it includes the chromosomal site that participated in the formation of the Atra FpurE+ plasmids pRH126, pRH127, pRH129 and pRH135. The chromosomal sequence immediately

3. Results

indicat,ed

fragments containing the novel joints of these plasmids are partially homologous to BgZII fragment 14 of F (Hadley & Deonier. 1980). This indicates that the novel joint is located in the vicinity of osiT (Hadley & Deonier. 1980). The oriT contained on the Bglll fragment 14 of F was cloned from pRS27 into M13mp9 in the orientation depicted in MXRD606 (Fig. 2). The Atra FproA’ plasmids were digested with Bglll and the fragment’ mixtures were ligated to BamHldigested M13mp9 replicative form DNA. JM103 was transformed with the resulting ligation mixture and the transformants were screened for the presence of the novel joint fragment by the C-test procedure (described in Materials and Methods) using MXRD606 as the probe. The insertions detected by this method are oriented in the phage

clockwise

from

the recombination

site

extends rightward from the novel joints from these specific Atra FpurE+ plasmids. The chromosomal DKA

from

pRH128

contains

also

the

sequence

counterclockwise from the recombination site. This site is contained on a 3.8 kbt Bglll-EcoRI fragment from pRH128, and it was cloned into BamHIEcoRI-digested plJC8. The ligated DNA solution was transformed into JM103, and plasmid pXRD618 was identified by the presence of the

t Abbreviations bp. base-pairs.

used: kb, lo3 bases or base-pairs;

270

H. Horowitz

(a)

8

6 proA, 8 argF

-120

and R. C. Deorrirt

-80

tuc

-40

---

0

10

12

proc

40

min j?urE

80

120

160

200

240

kb

, pRHll3,114,115,116

, pRH126,127,129,135

, pRH112

I pRH128

(b) --

PWE

oriv

If-0 -

.

Figure 1. (a) Regions of the E. coli K-12 chromosome involved

in Atra

--

Hfr progenitor

F’ formation.

Chromosomal

sequences

are

line. IS2 and IS3 elements are represented by hatched and filled boxes, respectively. The F DNA in the Hfr progenitor of the Atru FpurE* plasmids is integrated at the IS3 element cQ4. Hfr progenitors of the Atra FproA+ plasmids contain F integrated at the IS3 t1& (Hadley & Deonier, 1980). The origin for the kb scale line is the EcoRI cleavage site in la&. Span lines indicate the lengths of chromosomal DNA on t,he plasmids indicated. (b) Formation of a Atra FpurE+ plasmid from an Hfr parent. represented

by a sawtooth

3.8 kb insert. From pXRD618 a smaller HindIIIEcoRI fragment containing the recombination site was identified by restriction analysis and Southern (1975) hybridization to a fragment containing the novel joint. The HindIII-EcoRI fragment was subcloned into M13mp8 replicative form DNA digested with Hind111 and EcoRI (Fig. 2) to produce MXRD618, from which the chromosomal recombination site for the specific class of Atra FpurE+ plasmids was determined. For mobilization studies BgZII fragments containing the specific Atra FpurE+ novel joints were cloned onto the mob-, pBR322-based vector pED961 (Brown & Willetts, 1981). EcoRI-Hind111 fragments containing a specific Atra FproA+ novel joint (pRHll3) and a non-specific novel joint (pRH128) were transferred from the Ml3 phage clones to EcoRI and HindIII-digested pUC8 DNA (Fig. 2).

(b) Mobilization

analysis

In the process of Atra F’ formation, DNA from the oriT region through the h-a operon up to the point of F integration is deleted (see Fig. 1). If a portion of the F oriT sequence is retained at the novel joint, it still may display limited function. We therefore attempted to mobilize small plasmids containing the novel joints by using ~0x38, which contains no insertion sequences or ~6, to supply the tra gene products. Such plasmids were not

mobilized in trans by ~0x38 (Table 3). Plasmid pXRD606, which contains oriT from F, was transferred at a frequency approximately equal to that of ~0x38. This is consistent with results Table 3 Mobilization eficiency of small plasmid constructs by ~0x38 from a recA- donor Mobilization Plasmid pXRD606 pXRD601 pXRD615 pXRD605 pXRD618 pED961 pUCt?

Insert

CWiT oriTA f oriTA oriTA Target4 None None

30 min mating 0.66 <2.7x lo-’ < 1.1 x lo-’ <2.5x lo-’ <3.9x lo-’ <4.2x lo-’ <‘&2x lo-’

efficiency? 90 min mating 0.95 54 x 10-6 <3.9x 10-s WI <5.5x 10-8 <3.5x 10-8 NT

RD17 (~0x38) was the recA- donor strain. t Mobilization efficiency is the ratio of the number of recipients containing the small plasmid construct to the number of recipients containing ~0x38. The values were obtained using a recA - recipient (AB2463). Values for matings using a recA+ recipient (AB1157) were similar. All values are the average from 2 experiments. When no transconjugants contained the small plasmid construct, the value reported is the reciprocal of the number of transconjugants containing ~0x38. 1 oriTA represents the novel joint segment from a Atra F plasmid. For exact identification, see Fig. 2. 4 pXRD618 contains the chromosome1 target sequence used for specific Atra FpuvE+ formation. 11NT, not tested.

at oriT

Recombination

271

(a) (1.08kb.

F (OfiT)

+ pXRD606(Bg)

---YT-Bg (1*2kb,

OffaFpf0A

Bg

+ pXRD601-604 pED96 l(Bg)

pRH113,114.115,1

16) ___j M 13mp9(B)

cm*mNy‘;;g

Bg

MXRD601-604

7

OPw

specific

MXRD606

pRHl26.127,129.135) /G’

specific

__l M 13mp9(B)

pED96 1 (Bg)

Bg

(0.73kb,

&raFpurE

MXRD613-616 (E-H)

A pUC8(E-H)

pXRD613-616

(2.7kb. pRHl28)

fh?FpurE

-.-P--J1-----? Bg

non-specific

(1.6kb,

&raFproA

target

(3.8kb.

MXRD605(E-H)

E M 13mp803-E)

p, pUCB(E-H)

pXRD605

pRH112)

vr,,z/a, Bg

non-specific

specific

bl4)

Bg-

-

MXRD612

Bg

M 13mp9(B)

pRHl28) F

pXRD618(E-H)

___f M 13mp8(E-Ii)

MXRD616

(b) ..,...) --

-

H

B

E ----

Figure 2. Plasmid and bacteriophage constructs containing novel joints from Atra F’ plasmids (+). oriT from F (---+----), or chromosomal target sequence for Atra FpurE+ formation (v). F DKA is represented as a straight line. and chromosomal Dir;A is depicted as a sawtooth line. Plasmid constructs are named with a prefix p. and Ml3 caonstructs are named with a prefix M. An arrow indicates derivation, and the vector used is named underneath. The enzyme used to cleave the vector is indicated in parenthesis. The sizes and sources of each insert. are indicated above the relevant segments. Enzymes used to generate termini are indicat,ed as follows: E. EcoRI: B. BarnHI: Bg, BgZII; H. HindTTT. (b) Cloning sites on pUC8 or phage vector M13mp8. The multiple cloning sequence is in the opposite orientation in Ml3mp9. The dotted arrow depicts the direction of dideoxy sequencing. For details of the constructions. see Results. previously for similar small plasmids (Everett & Willetts, 1982). Mobilization of small plasmids in trana was not detected in either 30-minute or go-minute matings. (<5x 10e7 per pOX38+ transconjugant, see Table 3). Therefore, the deletion of F DNA accompanying Atra F’ formation has removed a portion of oriT, or a segment required for its function. The ApR transconjugants that were obtained contained plasmids obtained

t’hat

were

transferred

via

cointegration

mediated

by recombination between the novel joint and oriT of ~0x38. Mobilization of pXRD601 by ~0x38 yielded four pXRD601+ exconjugants. Three were from mating with the recA +recipient AB1157 and one was from mating with AB2463, which is recA-. All four transconjugants contained fusions of the small plasmid with ~0x38. Since all of the small plasmids containing novel joints were found to be non-mobilizable, the fused cointegrates presumably were formed in the donor, which is recA-. The structure of the cointegrates was analyzed by digestion with EcoRI. The small plasmid consists of a single 7.3 kb EcoRI fragment. In the three chimeras isolated from AB1157, this fragment and the 8.0 kb fragment 6 of F, which includes oriT, is missing and two additional fragments of 4.2 kb and

11.1 kb are present. This result suggests that small plasmids containing Atra FpurE’ novel joints recombined in a recA-independent fashion with oriT on ~0x38. Digestion of these plasmids with BgEII produced the 1.1 kb BgZII fragment 14 of F, which includes oriT, and the 0.72 kb Atra FpwE’ fragment that contains the novel joint. Recombination between F DNA adjacent to the novel joint and that adjacent to oriT would produce this structure, which would retain a functional con jugat,ive

oriT. The cointegrate (data not shown).

plasmids

are

The fourth cointegrate (isolated from AB2463) displays BgZII and EcoRI restriction digest patterns consistent with recombination between oriV of ~0x38 and a site on the vector DNA of pXRD601. Such ori V-mediated site-specific recombination has been reported in other contexts (Kilbane & Malamy, 1980; O’Connor & Malamy, 1984). (c) Nucleotide

sequence of regions Atra F’ formation

involved

in

Regions surrounding the novel joints of the Atra F’ plasmids were sequenced clockwise relative to the F map starting from the BgEIT site counter-

27”

H. Horowitz 120

I50

5~CTCACCACCA

AAAGCACCAC

ACCCCACGCA

AAAACAAGTT3’

(a)

GAGTGGTGGT

TTTCGTGGTG

TGGGGTGCGT

TTTTGTTCAA

tb,

%TCACCACCA GAGTGGTGGT

AAAGCACCAC TTTCGTGGTG

ACCACACGTA TGGTGTGCAT

TCCAGACGGT3’ AGGTCTGCCA

icj

%ACGACGGT A~TGCTG~~A

GCAGACTGAC CGTCTGACTG

ACCACACGTA TGGTGTG~AT

TCCAGACGGT~’ AGGTCTGCCA

td,

%TCACCACCA GAGTGGTGGT

AAAGCACCAC TTTCGTGGTG

ACCACACGGA TGGTGTGCCT

ACGAGCCATT3’ TGCTCGGTAA

(e,

%TCACCACCA GAGTGGTGGT

AAAGcAccAC TTTCGTGGTG

ACCTCTTGCA TGGAGAACGT

GGATACTTG?’ CCTATGAACG

5’ CTCACCATCA GAGTGGTAGT

AAGGCGGGAATATGTTAAGA TTCCGCCCTT ATACAATTCT

(‘I

and R. C. Lkoniet

TTCATAGATT3’ AAGTATCTAA

Figure 3. Kucleotide sequences of o&T, novel joint and chromosomal target regions involved in Atra F formation. In most cases, both strands were sequenced. Xucleotides that were not sequenced directly but were inferred from reading the complementary, strand are underlined. (a) oriT region of F. (b) Kovel Joint of specific (4 independent isolates). Al?YZ FpurE + plasmids (v) (:hromosomal target region for specific Atra FpurE+ plasmids. (d) Kovel joint of specific Ah-a Fproil’ plasmids (4 independent isolates). (e) Xovel joint of the non-specific Ah-u FpurE+ pRH128. (f) Novel joint of the non-specific Atra FproA+ pRHl12. Xumbers above the first and last nucleotides shown indicate distance in bp clockwise from the center of t,he BgZIT site at 66.6 on t,he F map. Sequence hyphens have been omitted for rlarity in Figs 3 and 4. clockwise from oriT on F (Fig. 2). The sequences of the oriT region from F and from the A&a F’ plasmids are’ presented in Figure 3, and Figure 4 indicates the inferred sequences of the sites before the novel joints were formed. F sequence extends approximately 142 bp from this BgZII site up to the novel joint,. The exact location of the novel joint cannot be determined for the Atra FpurE+ specific class because the F and chromosomal DNA are homologous for the 5 bp counterclockwise of the 3’ F

(site A)

LlfroFproA

c

site

v%!ii%#B

XXXXXA

purE (chrom.) pRti 128 site

XXXXXT 3

F bite

2

1

6)

pRH 112 site

Figure 4. Homologies between the F ori?’ region and chromosomal DNA involved in Atra F’ formation. Sequences are numbered clockwise from the BglII site as described in Fig. 3. Ho&ogies are boxed. X indicates that the ident’ity of a nucleotide is not’ known. The which the first recognized base of positions at chromosomal sequence appears (A) and opposite which nicks were found by Thompson et al. (1984) (77) are point indicated, as are the positions of the oriTmutations (*) (Thompson et al., 1984). F (site A) participated in the formation of 9 of the novel joints, and F (site B) is implicated in one of the formation events.

first divergent base at 143. However. all of thr plasmids in the specific classes. whether isolated from the purE or the proA region of the chromosome, possess the same 142 bp of F DNX. The non-specific Atra FpurE+ plasmid pRH 128 also shares this 142 bp of t’he F sequence. but the F sequence on the non-specific Atrcx F~uJ~ + plasmid pRHl12 extends only 126 bp clockwise of the HglII site. The sequences across the novel joints of’ the specific Atrn FpurE+ plasmids were identical. The sequences across the novel joints of the specific Atm FproA + plasmids were also identical. but thej differed from those seen with the Atra FpurE+ plasmids. They therefore appear to hare been formed by a site-specific procaess that displays nucleotide specificity for both F and chromosomal sequences. The chromosomal target’ sequenct’ used in the formation of the specific Ah F,vwrE+ plasmids includes a S/l0 bp homology to F sequence in t’he oriT region. 4. Discussion Although considerable information is available on the processing of F DNA during conjugation (see Willetts & Wilkins, 1984, for a review), fundamental biochemical issues remain to be addressed. The F oriT might operate vin nicking and ligation steps analogous to those associated with the 4X174 c&4 gene product (Willetts & Wilkens, 1984), or oriT might’ be a complex genetic containing several subregions. For structure example, in addition to a domain that determines the position of the nick, there also might be separate or overlapping domains t,hat determine the t)hat are required for response to triggering, repliconation, or that recognize structures near the base of the pilus. Present evidence indicates that monomer length units of F are transferred, which implies that transfer terminates at the 3’ OH produced by the initiating nick (see Willetts & Wilkins. 1984, for discussion and references). However, for ColEl and for F. repliconation can occur at sequences other than the origin at which t,ransfer was initiated, t’hus reducing multimeric molecules to monomers (Warren & Clark. 1980: Everett & Willetts, 1982). Because detailed information about the normal oriT biochemistry is unavailable, it is impossible at this time to state whether Atra F’ formation proceeds by normal biochemical steps associated with oriT or whether it is a totally aberrant, process. Several features of Atra F’ plasmids indicate to UH that they may have been formed by “normal” steps. First, Atra F’ plasmids form from Hfr strains at frequencies comparable to the frequency of type II F’ formation (Hadley $ Deonier. 1980; Buysse &, Palchaudhuri, 1984). In the Zac-purE region this means that the recombination event at oriT is almost as efficient as recA-mediated IS5 x IS5 recombination that produces the predominant type II F’ class (Timmons et al., 1983). Moreover, IO to

Recombination at oriT ZOO/bof the transconjugants from mating a recAtype II F’ strain with a recA- recipient contain 1984). Atra F’ variants (Buysse & Palchaudhuri, Second, nine of the ten Atra F’ novel joints examined here reveal F DNA joined to bacterial DNA at, the same position on F, indicating a high order of site specificity. Finally, the novel joint in nine of the ten examples lies between t,he t’wo sites identified with oriT point mutants (Thompson et al., 1984). If the Atra F’ plasmids that we examined were indeed formed by the normal F oriT biochemistrv. then our data suggest that the initiat’ing nick occurs in the vicinit’y of bp 142 (F (site A), Fig. 4). Direct biochemical analysis of nicked oriT segments carried by ;i vectors identified several nick sites. with primary sites located between bp 122 and bp 12X (Thompson et al.. 1984). One of the ten novel joints that, we examined (i.e. from pRH 112) displayed a discontinuity in F sequence in this region (F (site R), Fig. 4). The discrepancy between these loci and the primary locus that we find near bp 142 can be explained in a number of ways: (1) the site near bp 142 results from aberrant biochemical steps related or unrelated to oriT; (2) the loci counterclockwise from bp 128 result from DNA processing subsequent to the initial nicking and are recovered in altered form as a result of 2 packaging biochemistry (Thompson et al., 1984); or (3) there are two regions in the oriT segment that can be nicked by F gene products. There is insufficient information to prove or disprove any of these possibilit’ies. We have completely sequenced only the chromosomal site near purE. If the oriT segments of all of the Hfr progenitors of the Atra F’ plasmids are identical to oriT of F, then the clockwise portions of the other chromosornal sites are as shown in Figure 4. Partial homology to the F oriT region is evident. The nucleotides at positions 144. 147 and 149 are conserved in all novel joints near site A). The non-specific example (from pRH128) shows the least homology. If the chromosomal site near purE is typical of sites producing specific novel join&, then the chromosomal sites are short,er than the minimum homology that can be sensed in recAmediated recombination (Gonda & Radding, 1983). The ext’ent of homology between the chromosomal sequence and F in the region from bp 143 to 150 is as great as the homology between int junction type recognition sequences in attR and the secondary attachment sites for bacteriophage 1 (Ross & Landy, 1983). This and the repeated use of the same chromosomal site in Atra F’ formation makes more plausible the suggestion that Atra F’ plasmids can result from recognition by the F nicking/ ligating complex of chromosomal sequences similar to oriT. The chromosomal sites may be recognized by the repliconat,ion machinery, or perhaps these sequences in the purl% and proA regions of the chromosomes are preferentially nicked by a hostencoded enzyme. Once nicked, these sites might then be available for the ligation reaction t,hat is

273

presumed to accompany repliconation (BroomeSmith, 1980; Everett & Willetts, 1982). The failure of the chromosomal site near purl3 to confer mobilizability on small plasmids does not contradict this hypothesis because nicking may be necessary but not sufficient for transfer. It is gratifying that the fragments containing the novel joints also fail to confer the oriT+ character on small plasmids. since they cont,ain sequence alterations in the immediate vicinity of previously identified oriT- point mut)ations (Thompson et nl.. 1984). The authors thank Alicia M. Bogardus and Kim Spear for excellent technical assistance. We thank Xeil Willetts for providing pED961. and Pierre Prrntki for assistance in synthesizing oligonucleotide primers. This research was supported by Public Health Service grant GM24589 from the National Tnstit’ut,es of Health. References Broda. I’. & Meacock. I’. (1971). ,Vol. CPU. Chnuf. 113. 166- 173. Broome-Smith. J. (1980). Plasmid, 4. 51-63. Brown. A. M. C’. & Willetts. N. S. (1981). Plnw~id. 5. 188 201. Buyssr. J. M. & Palchaudhuri. 8. (1984). Nol. Co)/. (bet. 193, 543-553. Clark. A. J. & Warren. G. J. (1979). .3 nnrl. ll~/s. (lenet. 13. 99-125. (:lowes. R. C. & Hayes, W. (1968). Experiments in :Mol~culnr Genetics, John Wiley 8: Sons. Inc.. Kew York. Dagert, M. & Ehrlich, S. D. (1979). Gene, 6. 23-28. Denhardt. D. T. (1966). Biochem. Biophys. Res. Pommun. 23. 641-646. Deonier, R. C. & Mirels, L. (1977). Proc. Xuf. Acad. Sci.. l’.cS.,4. 74. 396553969. DeWitt. 8. K. & Adelberg, E. A. (1962). Genetics, 47. 877-585. Everett, R. & Willet,ts. I\;. S. (1986). EMBO .J. 1. 74i$53. Gonda. D. K. & Radding, C. M. (1983). Cell. 34, 6477654. Guyrr, M. S. & Clark. A. J. (1976). J. Bacterial. 125. 233% 247. Buyer, M. S.. Davidson, N. & Clark, A. *J. (1977). J. Bacterial. 131, 970-980. Buyer. M. S., Reed, R. R., Steitz. J. A. & Low. K. B. (1980). Crold Spring Harbor Symp. Quan,t. Biol. 45. 135-140. Hadlev. R. G. & Deonier, R. C. (1979). J. Ba,cterioZ. 139, 9631-976. Hadley. R. G. & Deonier. R. C. (1980). J. Bacterial. 143. 680-692. Hong. G. F. (198 1). Biosci. Rep. 1, 243-252. Howard-Flanders, P. & Theriot. L. (1966). Genetics, 53. 1137m1148. Kilbanr, ,J. tJ. & Malamy. M. H. (1980). J. -Vol. Biol. 143. 73-93. Klein. R. D., Selsing, E. & Wells, R. D. (1981). Plasmid, 3, 88.-91. Lennox, E. S. (1955). Virology, 1, 190-206. Low, K. B. (1968). Proc. Nat. Acad. Sci., IT.S.A. 60, 160-167. Low, K. 13. (1972). Bacterial. Rev. 36, 587-607. Messing, ,J. (1983). In Methods in Enzymology (Wu. R., Grossman, L. & Moldave, K., eds). vol. 101, pp. P6-78, Academic Press, Pu’ewYork.

Messing, J. & Vieira, J. (1982). Gene, 19, 269-276. O’(~onnor. M. B. & Malamy. M. H. (1984). J. Mol. Biol. 175, 263-284. O’Farrell, I’. H., Kutter, E. & Nakaniski. M. (1980). L!fol. (/en,. &net. 179. 421-43.5. Ross. W. & Landg, A. (1983). Cell, 33, 261-272. Sanger. F.. Nicklen. S. & Coulsen, A. R. (1977). Proc. Nat.

Acad. Sk, lT.S.A. 74. 5463-5467. Skurray,

R. A.. Nagaishi, H. & Clark, A. J. (1976). Proc.

AM. Acad. A’ci., i!.iS.A 73. 64-68.

Southern, E. M. (1975). J. Mol. Biol. 98> 503--517. Thompson. R.. Taylor, L.. Kelly, K., Everett. R. h Will&tjs. K’. S. (1984). h’MBO J. 3, 1175-l 180. Timmons. M. S., Kogardus. A. M. QL Deoniw. I?. (‘. (1983). J. BacterioZ.153, 395--407. Vieira. J. & Messing, J. (1982). Gene, 19. 2599268. \Varren. t:. ,J. & Clark, A. J. (1980). Proc. ‘Vat. Acad. Lki., I’.S.A. 77. 6724.-6728. Willt%s, 9. S. bi Wilkins. H. (1984). Microhiol. Rely. 48, “4-41.

Edited by J. Miller