OIG'G
Mol Gen Genet (1984) 195:9(~95
© Springer-Verlag 1984
Characterization of eight excision plasmids of Pseudomonas syringae pv. phaseolicola Les J. Szabo * and Dallicc Mills
Department of Botany and Plant Pathology, Oregon State University, Corvallis, Oregon 97331, USA Summary. Pseudomonas syringae pv. phaseolicola strain
LR719 contains a 150 kilobase pair (kb) plasmid pMC7105, stably integrated into its chromosome. Occasionally, single colony isolates of this strain contain an excision plasmid. Eight unique excision plasmids were selected and characterized by BamHI restriction endonuclease and blot hybridization analyses. These plasmids ranged in size from 35 to 270 kb; the largest contained approximately 130 kb of chromosomal D N A sequences. Restriction maps of pMC7105 were developed to deduce the site of integration and to identify the fragments in which recombination occurred to produce each excision plasmid. The eight excision plasmids were arranged into five classes based on the sites where excision occurs. A 20 kb region of pMC7105, which includes BamHI fragment 9 and portions of adjacent fragments, is present in all excision plasmids and thought to contain the origin of replication. The site of integration on pMC7105 maps within BamHI fragment 8. This fragment shows homology with seven other BamHI fragments of pMC7105 and with five chromosomal fragments identified among the excision plasmids. The data strongly suggest that the integration of pMC7105 may have occurred at a repetitive sequence present on the chromosome and on the plasmid.
Introduction
Indigenous plasmids have been reported in numerous plant pathogenic bacteria (for review see Lacy and Leary 1979) and some have been demonstrated to carry genes for pathogenicity. It is now well established that the oncogenic properties of Agrobacterium turnefaciens and A. rhizogenes are plasmid-determined traits (Bevan and Chilton 1982). Comai and Kosuge (1980) also recently showed that the tumor-like galls incited by Pseudomonas syringae pv. savastanoi on olive and oleander plants result from indoleacetic acid production which is encoded by plasmid-borne genes. Another recently determined property of a large 150 kilobase pair (kb) plasmid, pMC7105, in P. syringae pv. phaseolicola, the causal agent of halo blight of bean, is its ability to replicate autonomously or integrate into the bacterial chromosome * Present address: The Rockefeller University, Department of Cell Biology, 1230 York Avenue, New York, NY 10021, USA Offprints requests to: D. Mills
(Curiale and Mills 1982). The integrated form is stable in culture but single colony isolates are occasionally found which harbor a single excision plasmid (Szabo et al. 1981 ; Curiale and Mills 1982). Of four excision plasmids initially characterized by restriction endonuclease analysis, three were smaller than pMC7105 and each contained a subset of restriction fragments from this plasmid (Curiale and Mills 1982). The fourth plasmid (ca. 235 kb) was comprised of fragments of plasmid and chromosomal origin (Szabo et al. 1981). The stable integration of plasmids into bacterial chromosomes is a rare phenomenon with the exception of the Hfr form of F factor in Escherichia coli Kl2 and plasmids showing integrative suppression (for review see Holloway 1979). A consequence of recombination between plasmids and the host chromosome is the production of new gene combinations within the chromosome and on the new excision plasmids. Plasmids which carry chromosomal sequences resulting from imprecise excision are considered to provide a selective advantage to a natural bacterial population if they encode their own transfer functions or, alternatively, are mobilized by other resident plasmids. The mechanisms by which plasmids promote transfer of chromosomal sequences, referred to as chromosomal mobilization ability (Cma) (Haas and Holloway 1978), is not well understood primarily because of the difficulty in isolating stably integrated structures and F-prime-like excision plasraids. We report here the isolation and physical characterization of eight stable excision plasmids from the integrated form of pMC7105 in P. syringae pv. phaseolicola. Restriction maps of pMC7105 were developed and the site of integration and fragments involved in recombination during excision identified. The BamHI fragment which carries the site of integration on pMC7105 shows homology with seven other BamHI fragments from this plasmid as well as some chromosomal fragments detected in the excision plasmids.
Materials and methods
Bacterial strains and plasmids. The bacterial strains and plasmids used in this study are listed in Table 1. The plasmid pMC7105, used as a reference standard in this study, was isolated from strain LR721. LR719 was isolated from LR716 after mitomycin C treatment (Curiale and Mills 1982).
91 Table l. Bacterial strains and plasmids
Strains P. syringae pv. phaseolicola LR716 LR719 LR721 PP806 PP807 PP808 PP809 PP810 PP812 PP813 PP814 Escherichia coli K-12 HB101 Plasmids pBR322 pAB0008
Genotype and/or phenotype ~
Plasmid content or insert
Source or reference
LcoCsStrr LcoCsStV
pMC7105 pMC7105 £2 chrom pMC7105 pEX8060 pEX8070 pEX8080 pEX8090 pEX8100 pEX8120 pEX8130 pEXS140
b b
Ser Rift Lcocs StV Lcoc~Stff Lcoc~StV Lcoc~StV Lcoc~Strr Lco¢~Str~ Lc&s Str~ Lcoc~StV
M. Curiale c this work this work this work this work this work this work this work this work
rn m B- leupro thi lacYrecA endol- Strr
H. Evans d
Apr Tcr Apr
H. Evans d this work
Barn-8 from pMC7105
a Apr, ampicillin resistance; Rift, rifampicin resistance; Strr, streptomycin resistance; Tcr, tetracycline resistance; Lco% low-cobalt, cold sensitive, morphology mutant; Ser, serine auxotroph b Curiale and Mills, 1982 c Auxotroph derived by ethyl methane sulfonate mutagenesis and spontaneous mutation to Rift a Nitrogen Fixation Laboratory, Oregon State University
Media and Culture Conditions. P. syringae pv. phaseolicola was cultured in MaNY medium, a modified MaSNY medium (Curiale and Mills 1977) which lacks sucrose. E. coli was grown in LB medium (Davis et al. 1980). Media were supplemented with ampicillin (100 lag/ml), chloramphenicol (100 ~tg/ml), rifampicin (100 lag/ml), streptomycin (50 lag/ ml) or tetracycline (12 lag/ml) when appropriate. DNA isolation. Large scale preparation of plasmid D N A from P. syringae pv phaseolicola was obtained by an adaptation of the procedure of Currier and Nester (1976), as modified by Szabo and Mills (1984). Recombinant plasmid DNA was isolated by a cleared lysate procedure of Kahn et al. (Kahn et al. 1979). Restriction endonuclease digestion. Restriction endonucleases EcoRI and Bamt-II were purified according to a procedure obtained from P. Myers (Cold Spring Harbor Laboratories, personal communication), and by the method of Wilson and Young (1980), respectively. All other restriction endonucleases were purchased from Bethesda Research Laboratories. The conditions used to digest D N A and for restriction mapping by double digestion were described by Davis et al. (1980). Typically, 2-3 lag of plasmid DNA were digested with 6-10 units of restriction endonuclease in a total volume of 40 gl.
Agarose gel electrophoresis. Agarose gel electrophoresis of restriction fragments was carried out on a horizontal slab gel of 0.7-1.4% agarose which was submersed in Tris-acetate running buffer (Davis et al. 1980). Plasmid and whole cell D N A fragments were separated by electrophoresis through 0.7% agarose at 1.5 V/cm for 26-30 h. Agarose gels were stained with ethidium bromide, visualized under UV light and photographed with Polaroid-type 55 Land film and a Kodak yellow filter (No. 15). HindIII and XbaI digests of lambda D N A were used as molecular weight standards. D N A fragments were recovered from agarose gels by electroelution into dialysis bags (Maniatis et al. 1982). DNA blot hybridization. D N A fragments were transferred from agarose gels to diazobenzyloxymethyl (DBM) paper as described by Alwine et al. (1979). Purified DNA's were labeled with [32p] _ dCTP (800 gci/mM) to a specific activity of 2 to 7 x 107 cpmAtg using nick translation kits purchased from Bethesda Research Laboratories or New England Nuclear. D N A blot hybridizations were carried out in 14 ml of 50% formamide buffer at 42 ° C, containing 0.5 to 1.5 × 10 v cpm of probe, for 16-20 h and washed as previously described (Alwine et al. 1979). Autoradiography was performed using Kodak No-Screen or X-Omat X-ray film. Molecular cloning. BamHI fragments of pMC7105 were cloned into pBR322. Equal molar concentrations of completely digested pMC7105 and pBR322 D N A (Bolivar and Backman 1979) were ligated together in L-buffer (66 mM Tris, pH 7.6, 6.6 mM MgCI2, 10 mM dithiothreitol, 0.5 mM ATP) (Dugaiczyk et al. 1975) with 50 units/ml of T4 ligase (Bethesda Research Laboratories) at 12.5 ° C for 10-20 rain, diluted 20-50 fold with L-buffer and incubated an additional 18-20 h. The partially digested products of pMC7105 were cloned into the BamHI site of pBR322, which had been treated with calf intestinal alkaline phosphatase (Boehringer Mannheim). BamHI fragment 8 of pMC7105 was subcloned into pBR322. Competent cells of E. coli HBI01 were prepared and transformed with recombinant plasmids as described by Morrison (1979).
Results
Restriction endonuclease analysis of excision plasmid DNA. Excision plasmids can readily be detected by analysis of random single colony isolates of LR719 (Szabo et al. 1981). These plasmid-containing colonies are stable and harbor a single plasmid species. Eight of these isolates were selected for use in this study and designated PP806-810, 812-814 (Table 1). Purified plasmid D N A from each strain was digested with BamHI endonuclease and analyzed by agarose gel electrophoresis. A summary of the composition and size of each excision plasmid is presented in Table 2. BamHI digestion ofpMC7105 produced 19 fragments ranging from 41 to 0.5 kb in size. The smallest fragment was not detected in previous studies (Szabo et al. 1981; Curiale and Mills 1982). The excision plasmids ranged in size from 35 kb (pEX8080) to 268 kb (pEX8120). Four of the plasmids (pEXS080, pEXSI40, pEXS090 and pEX8130) were smaller than pMC7105 and composed of a subset of the 19 BamHI fragments. One plasmid, pEX8070, was indistinguishable
92 Table 2. Composition and size of 8 excision plasmids derived from pMC7105 Plasmid
pMC7105 pEX8080 pEX8140 pEX8090 pEX8130 pEX8100 pEX8060 pEX8120 pEX8070
Origin of BamHI fragments pMC7105
chromosomaP
kb
no. fragments
kb
no. fragments
151 35 52 73 103 36 143 134 151
19 3 3 7 11 6 18 17 19
116 91 134 -
18 18 16 -
Total kb
151 35 52 73 103 152 234 268 151
a The amount of chromosomal DNA contained in each excision plasmid was calculated by summing the size of fragments which were different from pMC7105. This results in an over-estimation of chromosome sequences because sequences from pMC7105 involved in integration and excision are also included
from pMC7105 in size and exhibited an identical BamHI banding pattern. In order to verify that pEX8070 was identical to pMC7105, EcoRI and PstI restriction endonuclease analyses were performed. The banding patterns of these plasmids were identical; each had 47 EcoRI and 33 PstI fragments (data not shown) which suggested that pEX8070 resulted from precise excision of pMC7105 from the chro-
mosome. The remaining plasmids (pEX8100, pEX8060 and pEX8120) contain fragments not found in pMC7105. In the case of pEX8100, these extra fragments comprise more than 60% of the plasmid sequences.
Sequence homology between pMC7105, pEX8060, pEX8100 and pEX8120. It was presumed that the extra fragments in pEX8060, pEX8100 and pEX8120 were of chromosomal origin. However, it was possible that these fragments could have been derived by rearrangement ofpMC7105. To ascertain their origin, pMC7105, pEXS060, pEX8100 and pEX8120 were digested with BamHI and the fragments were separated by agarose gel electrophoresis and blotted to D B M paper (Fig. 1). The blot was hybridized with labeled pMC7105 D N A and the results are shown in Figure 1 B. The probe hybridized to all of the pMC7105 fragments contained within these excision plasmids as expected, but also to eight of the 34 fragments which were thought to be o f chromosomal origin. The majority of these extra fragments showed no homology with pMC7105, indicating that they contain chromosomal sequences. Two of the fragments which did show homology with this probe should be plasmid-chromosome juncture fragments formed by the integration of pMC7105. The remaining six fragments most likely represent chromosomal sequences which contain regions homologous to pMC7105. A comparison of the chromosomal fragments among the three largest excision plasmids revealed that, with the exception of a single fragment, pEX8100 and pEX8060 appeared to be identical, whereas pEX8120 contains a completely different set of fragments (Fig. 1 A). To characterize further the degree of sequence homology among the chro-
Fig. 1. Agarose gel electrophoresis of BamHI-digested pMC7105 (1), pEX8100 (2), pEX8060 (3), and pEX8120 (4). (A) Ethidium bromidestained 0.7% agarose gel. Autoradiograms of the DNA blotted fragments from the gel shown (A) after hybridization with 32p-labeled: (B) pMC7105; (C)pEX8100; (D)pAB0008. Fragments smaller than 0.9 kb migrated off the gel and were not detected. Numbers refer to BamHI fragments from pMC7105. Chromosomal fragments in pEX8100 and pEX8060 are indicated by (o), and those in pEX8120 by (A). Lambda DNA and HindIII-digested lambda DNA were used as molecular weight standards
93 mosomal fragments of these three excision plasmids, pEX8100 DNA was labeled and hybridized to the DBM blot of the fragments shown in Figure 1 A. All of the chromosomal fragments of pEXS060, but only three fragments from pEX8120, hybridized to pEX8100 probe (Fig. 1 C). In addition, eight fragments (2, 3, 5, 6, 12, 13, 14 and 18) from pMC7105, which were not contained within pEX8100, also hybridized. These results confirm that pEX8060 and pEX8100 essentially contain an identical set of chromosomal sequences, but they are different from those ofpEX8120. These results are readily explained if the site of excision which produced pEX8060 and pEX8100 is within the chromosome on one side (left) of the site of integration of pMC7105, and the chromosomal site for excision of pEX8120 is on the other side (right). Restriction map of pMC7105. To obtain a better understanding of the map position of BamHI fragments from pMC7105 within the excision plasmids, and to identify the site of integration and the fragments that recombine to form fusion fragments, a restriction map of pMC7105 was constructed by cloning partially restricted BamHI fragments into pBR322. Twenty clones containing two or more BarnHI fragments were selected and used to construct a restriction map by identifying overlapping segments (Fig. 2). Occasionally, noncontiguous, partially-digested BamHI fragments were cloned. These clones were readily identified by comparison with the restriction :fragments present in the various excision plasmids. To confirm the order of the BamHI fragments, an J(baI map was constructed by single and double digestion of pMC7105 and the cloned, partially restricted BamHI fragments. During the construction of the XbaI map, the number of sites were inconsistent between the double digests of purified pMC7105 and the cloned, partially digested BamHI fragments. Two of the 12 J(baI sites identified by single and double digestion were unrestrictable when pMC7105 DNA was extracted from E. coli clones. These sites were positively identified in individual BarnHI fragments from pMC7105 which were isolated by electroelution from agarose gels and digested with XbaI (data not shown). Identification of the site of integration and occurrence of a repetitive sequence on pMC7105. Upon integration, one of the 19 BamHI fragments of pMC7105 is disrupted and two new juncture fragments are formed which contain plasmid and chromosomal sequences. Two pieces of information indicate that the site of integration is located within BamHI fragment 8 (Barn-8) of pMC7105. First, this is the only fragment which has not been detected in any excision plasmid, with the exception of pEX8070, a plasrnid which was formed by precise excision of pMC7105. Although pEX8100 and pEX8060 each contains a fragment similar in size to Barn-8 (Fig. 1, lanes 1, 2 and 3), we have shown by Southern hybridization using subcloned region,; of Barn8 as probe that this fragment is not Barn-8 (Szabo and Mills, submitted). Second, excision plasmids pEX8100 and pEX8120 contain overlapping regions of pMC7105, as well as chromosomal sequences to the left and right, respectively, of the site of integration within the chromosome. Neither plasmid contains Barn-8, but each contains one of the fragments (either Barn-7 or Barn-14) known to be adjacent to Bam-8 in pMC7105 (Fig. IA, lanes 2 and 4, and Fig. 2). In pEX8100 and pEX8060, Bam-7 is expected to be contigu-
Fig. 2. Restriction endonuclease map of pMC7105 constructed with BamHI and XbaI
pEXS080 ....191 , pEXSI40 171il4
l-o{9l
pEX8090
191
,
I~IIM 2 {I 5
3~
Ill4 ~19l
!
~IM
2 115
3 N
z I{5
3 N_6A_~
pEX8130 pEX8120
A-P)
c~ 171114 1~19[
pEX8100
c, I;, Ill 4 {_ol9l
,
~ll~l
81 711l 4 I-o{9i
,
~ll~l 2 II 5
pEX8060
(A-P)
11_3~ (A-P)
Ill I/ (a-p)
i~,fls = 17111 4 N9I
pEXSOTO
3 1~6118
' Integratedform of pMC7105
(o-p) 0
20 (kb)
SCALE:
Fig. 3. A proposed BamHI linear map of the eight excision plasmids and the integrated form of pMC7105. Fragments involved in the excision of these plasmids are indicated by the open-ended bars (~=). The juncture fragments formed by integration of pMC7105 are designated ~ and #. One of the sites for excision of pEX8060 is located within Barn #, which produced a modified juncture fragment, #'. The 16 chromosomal fragments which were determined to map to the left of the chromosomal site for integration are identified collectively by upper case letters A-P. Sixteen other chromosomal fragments which map to the right of this site are identified by the letters a-p. The order of the chromosomal fragments has not been determined. Chromosomal sequences not found in the excision plasmids are indicated by solid bars ( I )
ous with the left plasmid-chromosome juncture fragment, which henceforth is referred to as c~. In pEX8120, Bam-14 is expected to be contiguous with the right plasmid-chromosome juncture fragment, which henceforth is referred to as ft. A proposed linear BamHI restriction map of each of the eight excision plasmids is presented in Fig. 3. A region approximately 20 kb in size, which includes Bam-9 and portions of adjacent fragments Barn-10 and Barn-l, is the only region common to all excision plasmids.
94 In an attempt to identify the plasmid-chromosome juncture fragments in these large excision plasmids, Bam-8 of pMC7105 was subcloned into pBR322 and the resulting plasmid, pAB0008, was nick translated and used to probe the DBM blot of fragments shown in Figure 1 A. This probe hybridized to four chromosomal fragments from pEX8100; five from pEX8060; and three from pEX8120 (Fig. 1D). These same chromosomal fragments from each plasmid also hybridized with pMC7105 probe (Fig. 1 B). Moreover, eight fragments (1, 2, 4, 5, 7, 8, 12 and 14) of pMC7105 hybridized with the pAB0008 probe. No hybridization was seen when pBR322 was used as a probe (data not shown). It was apparent from these results that Bam-8 contains a repetitive DNA sequence and it could not be used to positively identify the plasmid-chromosome juncture fragments. Discussion
The stable integration of pMC7105 into the bacterial chromosome of P. syringae pv. phaseolicola (Curiale and Mills 1982) and subsequent detection of excision plasmids has provided a unique opportunity to study several aspects of genetic recombination in phytopathogenic pseudomonads. We have used restriction endonuclease analysis to characterize eight excision plasmids which ranged in size from 35 kb to 268 kb. A proposed linear map of these eight plasraids and their relationship to the integrated form of pMC7105 is shown in Fig. 3. These excision plasmids have been arranged into five classes on the basis of the pMC7105 fragments they harbor and the position at which excision from the chromosome has occurred. The plasmids were derived by: (i) excision entirely within pMC7105 sequences (pEX8080, pEX8090, pEXS130, pEX8140); (ii) excision at a site within pMC7105 and at a site in the chromosome to the right of the site of integration (pEX8]20); (iii) excision at a site within pMC7105 and at a site in the chromosome to the left of the site of integration (pEXS100); (iv) excision at sites within chromosomal sequences on either side of the Site of integration (pEXS060); and (v) excision precisely at the site of integration resulting in the formation of a plasmid identical to pMC7105 (pEX8070). Although pEX8060 (class iv) is thought to have resulted from excision within chromosomal sequences on either side of the site of integration of pMC7105, the precise site of excision to the right is uncertain. It is possible that excision could have occurred within plasmid sequences in the plasmid-chromosome juncture fragment. Evidence was presented which indicates that the site of integration is within Barn-8 of pMC7105. This fragment has homology with seven other BamHI fragments of pMC7105 and with five chromosomal fragments identified among the large excision plasmids (Fig. 1D). There is strong evidence that both integration of F factor and the formation of F-prime plasmids in E. coli results from recombination between homologous insertion sequences (Hadley and Deonier 1980; Ohtsubo et al. 1981). Whether a similar mechanism is attributed to the formation of these plasmids remains to be determined. It is noteworthy, however, that of the eight excision plasmids described here (Fig. 3) and four described previously (Szabo et al. 1981; Curiale and Mills 1982), approximately 75% of the excision events resulted from recombination between four (1, 8, 10 and 12) of the 19 BamHI fragments of pMC7105. Three of these fragments (1, 8 and 12) have a common repetitive
sequence (Fig. 1 D). These results suggest that some of these excision plasmids may have been formed by recombination within a common repetitive sequence rather than within random sequences in pMC7105. A recent analysis (Quant and Mills 1984) of five P. syringae phaseolicola strains from diverse geographic areas revealed two plasmids of approximately 122 and 127 kb which contained some fragments in common with pMC7105. It was significant that one of these plasmids also contained a 20 kb fragment which showed no homology with pMC7105 probe DNA. Although the origin of these plasmids is uncertain, it seems plausible from evidence presented in this study that they could have evolved through recombination with the chromosome. If these are naturally occurring excision plasmids and if integration of pMC7105 occurs at a single site in the chromosome, any fragment which is not homologous with pMC7105 sequences would be expected to have homology with chromosomal sequences present on one or more of the excision plasmids characterized here (Fig. 3). If integration occurs at multiple sites in the chromosome, the 20 kb fragment could be used as a probe to identify other sites for integration of pMC7105. Many phytopathogenic bacteria harbor plasmids, but with the exception of the Ti and Ri plasmids of Agrobacterium species (Bevan and Chilton 1982), and pIAAI of P. syringae pv. savastanoi (Comai and Kosuge 1980), the functions encoded by these plasmids remain to be determined. Although no gene functions were mapped to pMC7105 in this study, certain sequences essential for its maintenance may have been fortuitously identified among the excision plasmids. Restriction analysis has identified a single region which is common to all excision plasmids. This region is approximately 20 kb in size and contains all of BamHI fragment 9 and a portion of fragments 1 and 10 (Fig. 3). This conserved region is expected to contain the origin of replication for pMC7105, and possibly genes for plasmid incompatibility. The discovery of the stable integrated form of pMC7105, which occasionally produces recombinant excision plasmids, should prove useful in determining whether pMC7105 can promote chromosome transfer. Cma has been attributed to a variety of plasmids detected in a large number of species of bacteria (Holloway 1979). The potential for genetic variability in natural populations is greatly enhanced in bacteria which have evolved genetic mechanisms for Cma. If pMC7105 is determined to be a conjugatire plasmid, it should be extremely useful in genetic studies and in studies of plasmid evolution among natural populations of phytopathogenic pseudomonads. Acknowledgments. We thank M. Cantrell for his assistance during preliminary experiments, M. Curiale and H. Evans for bacterial strains, P. Myers, L. Comai and T. Kosuge for communicating unpublished procedures, and the members of the laboratory of D.M. for critical discussion of this work. Journal Article No. 6952 of the Oregon Agricultural Experiment Station. This research was supported by Science and Education Administration (U.S. Department of Agriculture) grants no. 59-2411-0-1-443-0and 80CRGO-10443 from the Competitive Research Grants Office. References
Alwine JC, Kemp DJ, Parker BA, Reiser J, Renart J, Stark GR, Wahl GM (1979) Detection of specific RNA's or specific fragments of DNA by fractionation in gels and transfer to diazo-
95 benzyloxymethylpaper. In: Wu R (ed) Methods in enzymology, vol 68. Academic Press, New York, pp 220-242 Bevan MW, Chilton MD (1982) T-DNA of the Agrobacterium Ti and Ri plasmids. Ann Rev Genet 16:357 384 Bolivar F, Backman K (1979) Plasmids of Escherichia coli as cloning vectors. In: Wu R (ed) Methods in enzymology, vol 68. Academic Press, New York, pp 245~67 Comai L, Kosuge T (1980) Involvement of plasmid deoxyribonucleic acid in indolaecetic acid synthesis in Pseudomonas savastanoi. J Bacteriol 143:950-957 Curiale MS, Mills D (1977) Detection and characterization of plasraids in Pseudomonas glycinea. J Bacteriol 131:224-228 Curiale MS, Mills D (1982) Integration and partial excision of a cryptic plasmid in Pseudomonas syringae pv. phaseolicola. J Bacteriol 152:797-802 Currier TC, Nester EW (1976) Isolation of covalently closed circular DNA of high molecular weight from bacteria. Anal Biochem 76:431 441 Davis RW, Botstein D, Roth JR (1980) In: Advanced bacterial genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York Dugaiczyk A, Boyer HW, Goodman HM (1975) Ligation of EcoRI endonuclease-generated DNA fragments into linear and circular structures. J Mol Biol 96:171 184 Hadley RG, Deonier RC (1980) Specificity in the formation of A tra F-prime plasmids. J Bacteriol 143:680-692 Haas D, Holloway BW (1978) Chromosome mobilization by the R plasmid R6:45: A tool in Pseudomonas genetics Mol Gen Genet 158 : 22%237 Holloway BW (1979) Plasmids that mobilize bacterial chromosome. Plasmid 2:1-19 Kahn M, Kolter R, Thomas C, Figurski D, Meyer R, Remaut E, Helinski DR (1979) Plasmid cloning vehicles derived from plasmids ColE1, F, R6K, and RK2. In: Wu R (ed) Methods in enzymology, vol 68. Academic Press, New York, pp 268 280
Lacy GH, Leary JV (1979) Genetic systems in phytopathogenic bacteria. In: Grogan RG, Zentmyer GA, Cowling EB (eds) Ann Rev Phytopathol 17:181 202 Maniatis T, Fritsch EF, Sambrook J (1982) Molecular cloning. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York Morrison DA (1979) Transformation and preservation of competent bacterial cells by freezing. In: Wu R (ed) Methods of enzymology, vol 68. Academic Press, New York, pp 326-331 Ohtsubo E, Zenilman M, Ohtsubo H, McCormick M, Machida C, Machida Y (1981) Mechanism of insertion and cointegration mediated by IS1 and Tn3. Cold Spring Harbor Symp Quant Biol 45 : 283-296 Quant RL, Mills D (1984) An integrative plasmid and multiplesized plasmids of Pseudomonas syringae pv. phaseoIicola have extensive homology. Mol Gen Genet 193:459-466 Szabo LJ, Mills D (1984) Involvement of repetitive sequences in the integration of pMC7105 and formation of excision plasmids in Pseudomonas syringae pv. phaseolicola. J Bacteriol 157:821-827 Szabo LJ, Volpe J, Mills D (1981) Identification of F'-like plasmids of Pseudomonas syringae pv. phaseolicola. In: Lozano JC (ed) Proc 5th Int Conf Plant Path Bact, CIAT, Call, Colombia, pp 403-411 Wilson GA, Young FE (1980) Purification and properties of the BamHI endonuclease. In: Grossman L, Moldave K (ed) Methods of enzymology, vol 65. Academic Press, New York, pp 147153
Communicated by A. B6ck
Received October 19, 1983 / February 2, 1984