Transformation of Rhizobium meliloti 41 with Plasmid DNA - Journal of ...

3 downloads 0 Views 1MB Size Report
Transformation of Rhizobium meliloti 41 with Plasmid DNA. GYORGY B. KISS* ... Plasmid pGV1106, a derivative of the wide-host-range plasmid S-a of the W ..... Clark, A. J., and G. J. Warren. 1979. ... Duncan, L. K., and A. B. Tlerney. 1973.
JOURNAL OF BACTERIOLOGY, May 1982,

p.

Vol. 150, No. 2

465-470

0021-9193/82/050465-06$02.00/0

Transformation of Rhizobium meliloti 41 with Plasmid DNA GYORGY B. KISS* AND ZSUZSA KALMAN Institute of Genetics, Biological Research Center, Hungarian Academy of Sciences, H-6701 Szeged, Hungary Received 18 September 1981/Accepted 10 December 1981

Plasmid pGV1106, a derivative of the wide-host-range plasmid S-a of the W incompatibility group, was introduced into Rhizobium meliloti 41 by plasmidmediated mnobilization to overcome the restriction offoreign DNA. The mobilized plasmid pKK2 differed from the original pGV1106 by an extra piece of DNA of 1.3 kilobase pairs which supposedly originated from pJB3JI used for mobilization. If pKK2 was isolated from R. meliloti 41, it could be successfully reintroduced by transformation. The transformation frequency was low (10 to 54 colonies per ,g of plasmid DNA) but reproducible, and several lines of evidence showed that it was the consequence of plasmid DNA uptake. The small size (10.3 kilobases) and elevated copy number (10 to 15 copies per cell) of pKK2 make it a potentially useful cloning vector for the study of symbiotic nitrogen fixation genes of R. meliloti 41.

Rhizobium meliloti is able to fix nitrogen in symbiotic association with its plant host Medicago sativa L. Both plant and bacterial genes are involved in the establishment and functioning of symbiosis. Until now, two plasmids, RP4 and pRK290, capable of autonomous replication in R. meliloti, were used for cloning R. meliloti nitrogen fixation genes (9, 29; t. Vincze, personal communication). In these cases, cloned DNA fragments could be reintroduced only indirectly, i.e., first transformation into Escherichia coli followed by conjugation back into R. meliloti. Another disadvantage of using RP4 and pRK290 as cloning vectors is their large size of 56 and 20 kilobase pairs (kb), respectively. A transformation system allowing direct introduction of a small plasmid vector would be of great benefit for the study of the symbiotic nitrogen fixation genes of R. meliloti. Transformation and transfection of different Rhizobium species has been reported earlier (11, 14, 22, 26). However, transformation with plasmid DNA in R. meliloti has not yet been achieved. The reason for the failure of plasmid DNA to transform might have been the lack of a suitable indigenous plasmid or the restriction of foreign DNA by Rhizobium restriction endonucleases (30). In this report we describe a transformation system for R. meliloti 41 using plasmid DNA. Essentially, a small derivative of a wide-hostrange plasmid, S-a, was introduced into R. meliloti 41 by plasmid-mediated mobilization to overcome the possible barrier of the restriction of foreign DNA. The mobilized plasmid, pKK2, isolated from a transconjugant, could be suc-

cessfully reintroduced into R. meliloti 41 by transformation. MATERIALS AND METHODS Strain. The microorganisms, plasmids, and bacteriophages used in this study are listed in Table 1. Media. Complete media (TA and GTA) and minimal media were described previously (21, 27) and were used for culturing both R. meliloti 41 and E. coli strains. Antibiotics. The concentration and source of antibiotics used were: kanamycin sulfate (Medexport, Moscow, U.S.S.R.), 200 Fg/ml; streptomycin sulfate (Biogal, Debrecen, Hungary), 200 ,ug/ml; tetracyclinehydrochloride (Sigma Chemical Co., St. Louis, Mo.), 15 ,ug/ml. For E. coli, the following concentrations were used: kanamycin sulfate, 5 ,ug/ml; streptomycin sulfate, 10 ,ug/ml; tetracycline-hydrochloride, 15 ,ug/ mi. Growth conditions. Both R. meliloti 41 and E. coli strains were cultured at 34°C. Matings. Matings were performed as described by Dixon et al. (10), except that the modified media described by Kiss et al. (21) were used. Isolatioa of plasmid DNA. Plasmids were prepared from bacterial cultures grown in complete medium without amplification. Cells were harvested by centrifugation at 6,000 x g for 10 min. Cleared lysates were prepared essentially as described by Hansen and Olsen (16), but the recovery of the plasmids was done by the method of Birnboim and Doly (1). Before electrophoresis, 5 pl of bromophenol blue (1.0 mg/ml), Tris-hydrochloride (50 mM, pH 8.0), and EDTA (10 mM, pH 8.0) were added to 20 ,il of plasmid preparations. For large-scale isolation (1.0 to 4.0 liters), the same method was used by increasing the volumes proportionally. Covalently closed circular plasmid DNA was purified by CsCl-ethidium bromide density gradient

465

KISS AND KALMAN

466

J. BACTERIOL.

TABLE 1. Bacterial strains, plasmids, and phages used Strain

E. coli J53

C600 JF1754 GY883 GY940 GY942 R. meliloti 41 AK631 GY887 GY888

Genotype/phenotypea

Source and

pro met (pJB3JI) pro (pGV1106) met hisB leuB hsdR pro (pGV1106, pJB3JI) JF1754 (pKK2) JF1754 (pKK2, pJB3JI)

J. E. Beringer (2) J. Leemans J. D. Friesen This study This study This study

Wild type AK631 (pJB3JI, pKK1, pKK2) AK631 (pJB3JI, pKK1) AK631 (pKK1) AK631 (pKK2)

A. Kondorosi This study This study This study This study

KK1 KK2 Plasmids Km Sm incW pGV1106 J. Leemans pJB3JI Tc Cma' incPl (2) Km Sm pKK1 This study pKK2 Km Sm This study cti5 Phage 16-3 tiS (27) a Cma+, Chromosome mobilization ability; cti5, gene(s) determining and regulating inducibiity (thermoinducible). Other chromosomal and plasmid markers are abbreviated according to Demerec et al. (7) and Novick et al.

(25). Transformation of E. coli was by the method of ultracentrfugation according to Radloff et al. (28) and Mandel and Higa (24) which was adapted to plasmid Clewell and Helinski (4). transformation by Cohen et al. (5). a c d ons and gel dctResic phorlss. Restriction endonucleases were purchased from Boehringer Mannheim Ltd., West Germany; RESULTS New England Biolabs, Beverly, Mass.; and Bethesda Research Laboratories, Rockville, Md. Digestions Mobilizaton of pGV1106 into R. meliloti 41. were carried out according to suppliers' specifications. Plasmid S-a, a representative of the W incomSingle and double digestions were done with BamHI, patibility group which is self-transmissible BamHI/EcoRI, BamHI/BglII, EcoRI, BgIl, BglIII (Tra+) and confers chloramphenicol (Cm'), sulPstI, PstI, Bglll/KpnI, KpnI, and HindIII enzymes. fonamide (Su'), kanamycin (Kmi), and streptoThe size of the fiagments was calculated from the mycin (Sm') resistance has a wide host range calibration curve obtained from the known Hindlll among gram-negative bacteria (17, 19). This digestion fragments of phage lambda (6). DNA fiagments were separated on horizontal or plasmid can be introduced and maintained in R. vertical slab gels of 0.7 to 1.5% agarose (Miles Labora- meliloti 41 (A. Kondorosi, unpublished results). tories Ltd., South Africa) followed by staining with A derivative of this plasmid was constructed in ethidium bromide and visualization with UV light (18). vitro by transformation of E. coli C600 with a Recording of gel patterns was with "Planfilm NP22" ligated BgIII digest of S-a (J. Leemans, personal (ORWO, East Germany). communication). This derivative S-a plasmid, Transonnatlon. Transformation of R. meliloti 41 pGV1106, was Tra- and conferred resistance was peirformed as follows. AK631 was inoculated into 200 ml of TA medium supplemented with 0.3 M of only to Km and Sm. Plasmid pGV1106 was sucrose and incubated for 4 to 6 h at 34°C (cell density introduced into R. meliloti 41 by mobilization 1 x I0V to 2 x 109). Cells were collected and suspend- using a helper plasmid, pJB3JI, a Km' Tcr ed in 20 ml of 100 mM MgCl2. After 10 min at 0°C, the deriyative of R68.45 (2, 15). For this purpose, suspension was pelleted and suspended in 0.5 nml of plB3JI was conjugated into E. coli C600 carryCaCl2 (150 mM). After 30 min at0°C, 100 pl of plasmid ing-pGV1106 (Table 2, cross 1). One Tcr, Kmr, DNA (1.0 og/0.1 ml) was added, and incubation was and Smr transconjugant, GY883, containing continued for 60 min. The cells were treated by heat at both pGV1106 and pJB3JI (Fig. 1, lane 3) was 40°C for 1.5 min and then incubated for 30 min at 0°C. then used as a donor in mating with R. meliloti Bacteria were plated on selective plates (GTA supple- 41 as recipient (Table 2, cross 2). Kmr and Smr mented with 200 FIg of kanamycin sulfate per ml and 200 .g of streptomycin sulfate per ml) after incubation transconjugants were screened for plasmid DNA in complete medium at 34WC for 2 h for phenotypic content. All contained the two indigenous plasexpression. Colonies appeared after 3 days of incuba- mids of R. meliloti 41 and pJB3JI. (We noted tion at 34°C. that pJB3JI had an unexpectedly higher mobility

VOL. 1SO, 1982

Cross no.

Donor

TRANSFORMATION OF R. MELILOTI WITH PLASMID DNA TABLE 2. Mobilization of pGV1106 and pKK2 by pJB3JIa Transfer frequency Counterf Recipient selectededm AA marker B marker marker

467

Tscoant isolated

GY883 1.0 1.0 Tcr 1.0 2.0 x 10-3 Tcr KMr SmT 3.8 x 10-7 1.9 x 10-4 GY887, GY888 GY942 1.0 1.0 Tcr 3 1.0 3.7 x 10-3 Tcr 4 Prototrophy Kmr Smr 2.1 X 10-6 5.6 x 10-4 a Transfer frequency was related to the total viable count of the recipients (A) and to the number of recipients that acquired Tcr phenotype (B). Mating experiments and selection for transconjugants were carried out as described in the text. Strains are listed in Table 1. 1 2

E. coli J53 E. coli C6W0 Kmr Smr Prototrophy AK631 GY883 Prototrophy Kmr Smr GY940 E. coli J53 Prototrophy AK631 GY942

in R. meliloti than in E. coli [Fig. 1, lanes 4 and 5].) Apart from these plasmids, most transconjugants acquired an additional plasmid with a slower mobility than pGV1106 (Fig. 1, lane 4). In one particular transconjugant, denoted GY887, two additional plasmid bands were detected both having a higher molecular weight than pGV1106. These two plasmids were named pKK1 (upper band, at the same position as that in the majority of the transconjugants) and pKK2 (lower band), respectively (Fig. 1, lanes 2 and 5). The restriction map of pKK1 and pKK2 revealed that they differed from pGV1106 only by an extra piece of DNA (see below). Transformation of R. melloti 41 with pKK1 and pKK2. Plasmid pKK2 had the smallest size among the plasmid derivatives in the transconjugants tested. Therefore, it was selected as a candidate for a new cloning vector in R. meliloti 41. This plasmid, however, appeared along with pKK1 and the mobilizing plasmid pJB3JI in the transconjugant. To obtain a strain carrying only pKK2, transformation seemed to be a plausible approach. Plasmids from strain GY887 were isolated by ultracentrifugation (see Materials and Methods). This plasmid preparation was used for the first transformation experiment. The conditions for transformation were essentially the same as those used for transfection of E. coli described by Mandel and Higa (24), except that R. meliloti 41 was cultured in complete medium supplemented with 0.3 M sucrose (see Materials and Methods). Eleven of the 26 colonies that appeared on selective plates on the third day after transformation were screened for plasmid DNA content. Although all contained one or two plasmids similar to those in strain GY887, only two clones, KK1 and KK2, carrying pKK1 and pKK2, respectively, were selected for further study (Fig. 1, lanes 6 and 7). In the subsequent transformation experiments, pure pKK2 plasmid DNA isolated from KK2 was used and the colonies that appeared

were thoroughly characterized. Several lines of evidence showed that the colonies grown under double selective pressure were R. meliloti 41 transformants as the consequence of plasmid DNA uptake. (i) No colony appeared on selective plates when plasmid DNA was omitted or treated with DNase before transformation,

1 2 3 4 5 6 7 8 9

chr

RNA FIG. 1. Plasmid patterns of R. meliloti 41 and E. coli strains in 0.7% agarose gel. Plasmid preparations and electrophoresis were carried out as described in the text. Plasmids were isolated from the following strains: E. coli J53 (pJB3JI), lane 1; E. coli C600 (pGV1106), lane 2; GY883, lane 3; GY888, lane 4; GY887, lane 5; KK1, lane 6; KK2, lane 7; and AK631, lane 8. Lane 9 shows the HindIII fragments of phage lambda (5). Arrows show (from top to bottom) the position of plasmid bands of pRme4la, pJB3JI, pKK1, pKK2, and pGY1106. Only the smaller and indigenous plasmid, pRme4la of R. meliloti 41 (Banfalvi et al., Mol. Gen. Genet., in press) could be detected by this procedure. Chr, Linear chromosomal fragments detected in all cases.

468

KISS AND

KALMAN

whereas successful transformation could be repeated in eight independent experiments using pure pKK2 preparations. (ii) Both colony morphology and bacterial shape of the transformants under light microscope were indistinguishable from that of the wild-type R. meliloti 41. The transformants grew at the same rate as the parent strain in GTA medium. (iii) All transformants tested (51 colonies) were sensitive to phage 16-3 specific for R. meliloti 41. (iv) Two large plasmids, pRme4la and pRme4lb, found in the wild-type R. meliloti 41 (Z. Banfalvi, V. Sakanyan, C. Koncz, A. Kiss, I. Dusha, and A. Kondorosi, Mol. Gen. Genet., in press) could also be detected in the transformants by using the method of Eckhardt (12; data not shown). (v) Several representatives of the transformants (KK1 and KK2) were able to nodulate and fix nitrogen as effectively as the wild type in symbiotic association with their host plant Medicago sativa L. (vi) Plasmid loss went together with the loss of resistance to Km and Sm. If KK2 was grown under nonselective conditions for 10 generations, 13% of the colonies (36 out of 281) had spontaneously lost their Kmr and Smr phenotype. No plasmid similar to pKK2 could be detected in these Kms and Sms segregants. (vii) Plasmid pKK2 could always be retrieved from all transformants tested (42 instances). If transformation was performed with plasmids pGV1106 and pKK2 isolated from E. coli, no colony appeared on selective plates. Reintroduction of pKK2 into R. melilod 41 by conjugation. Plasmid pKK2 isolated from KK2 was introduced into E. coli JF1754 by transformation, with a frequency of 6.0 x 103 transformants per ,ug of plasmid DNA. One purified transformant, GY940, was selected into which pJB3JI was conjugated (Table 2, cross 3). A transconjugant (GY942) containing both pJB3JI and pKK2 was crossed with R. meliloti 41 (Table 2, cross 4). Eleven of 12 transconjugants possessed the original pKK2. This result showed that pKK2 could be introduced into R. meliloti 41 not only with transformation but by conjugation as well. Copy number of pKK2 in R. meliloti 41. The pattern and intensity of the plasmid bands of GY887 and KK2 (Fig. 1, lanes 5 and 7) indicated that more than one plasmid derivative of pGV1106 per cell could be present. This was supported by the estimation of the minimal copy number of pKK2 in R. meliloti 41. Plasmid DNA was isolated from three independent cultures of strain KK2 of known viable count. The amount of DNA loaded on an agarose gel of these three isolates was estimated by a calibration curve of known amounts of purified pKK2 run in the same el. Using the Loschmidt constant (6.023 x 102 mol/liter) and the molecular weight of

J. BACTERIOL.

pKK2 (6.6 megadaltons), it was calculated that the minimum copy number was 10 to 15 per cell. (The same theoretical considerations were used by Hansen and Olsen [16] for calculating plasmid yields.) This result is in good agreement with the finding that the W-type plasmids have a larger number of copies per cell (13). Restriction map of pKK2. A physical map of pKK2 was constructed from the size of fragments generated by restriction endonuclease digestions. Single and double digestions of pure pKK2 plasmid preparations were carried out as described in Materials and Methods. The restriction map of pKK2 revealed that it differed from pGV1106 only by an extra piece of DNA of 1.3 kb (Fig. 2). The foreign DNA fragment of pKK2 contained an additional KpnI and two PstI sites as compared to the original plasmid pGV1106. The 0.8-kb PstI fragment originated from the insert comigrated with the 0.8-kb PstI fragment of pJB3JI (data not shown). A physical map of pKK1 was also constructed. It was similar to pKK2 except having an insert of 2.5 kb. The insert contained the characteristic 0.8-kb PstI fragment and the KpnI site but in the opposite orientation compared to that of pKK2.

DISCUSSION In this report we have demonstrated the successful transformation of R. meliloti 41 with plasmid DNA. The frequency of transformation

Bgl !'

FIG. 2. Restriction map and size of pKK2. These were constructed from the size of fragments generated by single and double restriction endonuclease digestions (see text). Plasmid pKK2 is a conjugative derivative of pGY1106 found in a Kmr and Smr R. meliloti 41 transconjugant. External arrows show the position of the insert and the pGV1106 part of plasmid pKK2. The position of the Km and Sm resistant determinants are shown by internal arrows.

VOL. 150, 1982

TRANSFORMATION OF R. MELILOTI WITH PLASMID DNA

was low (10 to 54 colonies per ,ug of DNA) but reproducible, and seven different experimental lines of evidence showed that the Kmr and Smr colonies were R. meliloti 41 transformants as the consequence of plasmid DNA treatment (see Results). Among these, one was a direct proof, namely the recovery of the unchanged transforming plasmid, pKK2, from the transformants in all cases tested. The low frequency of transformation probably was not due to a low plating efficiency because in a reconstruction experiment 78% of the bacteria containing pKK2 could form colonies on selective plates. There are indications that the conditions used for transformation in this study were suboptimal and the cells could be made more competent for plasmid DNA uptake. In fact, preliminary results showed that the frequency of transformation could be increased by a factor of 10 when the cells were pregrown in complete medium supplemented with glycine (data not shown). This beneficial effect of glycine on transformation was demonstrated by Drozanska and Lorkiewicz (personal communication). Transformants could not be detected when plasmid pKK2 or pGV1106 was isolated from E. coli and used for transformation. This result suggests the restriction of foreign DNA by R. meliloti 41. This is in contrast with the finding of Dunican and Tierney (11) and O'Gara and Dunican (26) who reported transformation of R. trifolii Ti with R-factors isolated from Pseudomonas aeruginosa and E. coli with high frequency. In our case, the low frequency of transformation probably did not permit the detection of those cells in which the entering plasmid escaped the restriction system of the host. The failure of pGV1106 in transformation could also be explained by the absence of the extra piece of DNA present in pKK2 (Fig. 2). Whether or not this insert had an effect on transformation or on the maintenance and phenotypic expression of pKK2 is not known at present. The lack of transformants, when plasmid DNA isolated from E. coli was used (see above), showed that the introduction of pGV1106 by mobilization into R. meliloti 41 was a crucial step. However, none of the transformants tested received the original plasmid pGV1106. The restriction map of pKK2 (the smallest derivative of pGV1106 found in the transconjugants) revealed an extra piece of DNA of 1.3 kb inserted into pGV1106 (Fig. 2). This is not surprising because the appearance of an extra piece of DNA in the mobilized plasmid is characteristic of mobilization (20; for review see reference 3). There are two indications that this 1.3-kb insert in pKK2 originated from the plasmid pJB3JI responsible for mobilization. (i) Plasmid pKK2 harbors a 0.8-kb PstI fragment comigrating with

469

the 0.8-kb PstI fragment of pJB3JI (data not shown). This PstI fragment of pJB3JI sits in the middle of a region responsible for R-prime and cointegrate formation as well as conduction of non-self-transmissible plasmids (8, 23; G. Reiss, B. W. Holloway, and A. Puihler, Genet. Res., in press). The molecular mechanism by which pJB3JI mobilizes pGV1106 is under investigation. (ii) pKK2 could be mobilized by pJB3JI with a frequency of about three times that of pGV1106 (see Table 2, cross 4). In addition, 11 of 12 transconjugants of this cross contained unaltered pKK2. In this case, the mobilization could be accomplished through homology with pJB3JI. Plasmid pKK2 is a new potential cloning vector for the study of symbiotic nitrogen fixation genes in R. meliloti 41. It displays the following useful features. (i) The transformation system described here makes possible the direct introduction of pKK2 into R. meliloti 41 (ii) As an alternative way, pKK2 can be mobilized from E. coli without structural alterations as compared with pGV1106. Thus, cloning can be performed in E. coli, and the selected recombinant plasmids can be mobilized by pJB3JI. (iii) pKK2 has at least three unique restriction cleavage sites for BglII, BamHI, and EcoRI. EcoRI is particularly useful since cloning into this site destroys the structural gene of the Smr determinant (J. Leemans, unpublished results), allowing a rapid screening for inserts. (iv) High plasmid yield can be obtained in consequence of the elevated copy number of pKK2 (see Results). (v) The relatively small size makes the cloning of large restriction fragments easier. Recently, a 13.0-kb BamHI fragment carrying the structural genes of nitrogenase of R. meliloti 41 (unpublished) has been successfully inserted into the BamHI site of pKK2. The study of the biological function of this fragment by complementation of ineffective mutants of R. meliloti 41 is in progress. ACKNOWLEDGMENTS We thank J. Leemans for providing pGV1106 and data prior to publication, B. Sain and P. Venetianer for encouragement and for helpful discussion of the manuscript, K. Dob6 and T. T6th for technical assistance, Zs. Racz for secretarial assistance, and B. Dusha for making the gel photo. This work was supported in part by OMFB grant 13017/FPI. LITERATURE CITED 1. Birnboim, H. C., and J. Doly. 1979. A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Res. 7:1513-1523. 2. Brewfn, N. J., J. E. Beringer, and A. W. B. Johnston. 1980. Plasmid-mediated transfer of host-range specificity between two strains of Rhizobium leguminosarum. J. Gen. Microbiol. 120:413-420. 3. Clark, A. J., and G. J. Warren. 1979. Conjugal transmission of plasmids. Annu. Rev. Genet. 13:99-125. 4. Clewell, D. B., and D. R. Helnkdl. 1970. Properties of a

470

5.

6. 7.

8. 9.

10.

11.

12.

13. 14.

15. 16.

17.

18.

J. BACTERIOL.

KISS AND KALMAN

supercoiled deoxyribonucleic acid-protein relaxation complex and strand specificity of the relaxation event. Biochemistry 9:4428-4440. Cohen, S. N., A. C. Y. Chaug, and L. Shu. 1972. Nonchromosomal antibiotic resistance in bacteria: genetic transformation of Escherichia coli by R-factor DNA. Proc. Natl. Acad. Sci. U.S.A. 69:2110-2114. Danel, D. L., J. R. de Wet, and F. R. Blatter. 1980. New map of bacteriophage lambda DNA. J. Virol. 33:390-400. Denerec, M., E. A. Adelberg, A. J. Clark, and P. E. Hartan. 1966. A proposal for a uniform nomenclature in bacterial genetics. Genetics 54:61-76. Depklcker, A., M. De Block, D. Inz6, M. Van Montag, and J. Schdl. 1980. IS-like element IS8 in RP4 plasmid and its involvement in cointegration. Gene 10:329-338. Dltta, G., S. Sbtod, D. Corbin, an D. R. Hellaski. 1980. Broad host range DNA cloning system for Gramnegative bacteria: construction of a gene bank of Rhizobium meliloti. Proc. Natl. Acad. Sci. U.S.A. 77:7347-7351. Dixon, R. A., F. C. Cannon, and A. Kondoros. 1976. Construction of a P plasmid carrying nitrogen fixation genes from Klebsiella pneumoniae. Nature (London) 260:268-271. Duncan, L. K., and A. B. Tlerney. 1973. Transformation of an R-factor from Pseudomonas aeruginosa into Rhizobaum trifolii. Mol. Gen. Genet. 126:187-190. E dt, T. 1978. A rapid method for the identification of plasmid desoxyribonucleic acid in bacteria. Plasmid 1:584-588. Falkow, S. 1975. Infectious multiple drug resistance. Pion, London. Gdbor-Hotcbklm, M. 1972. Biological properties of Rhizobia and their information transfer system, p. 212-241. In L. Ledoux (ed.), Uptake of informative molecules by living cells. North-Holland Publishing Co., Amsterdam. Eam, D., and B. W. Holoway. 1976. R factor variants with enhanced sex factor activity in Pseudomonas aeruginosa. Mol. Gen. Genet. 144:243-251. Hane , J. B., and R. H. Obe. 1978. Isolation of large bacterial plasmids and characterization of the P2 incompatibility group plasmids pMG1 and pMGS. J. Bacteriol. 135:227-238. Hedges, R. W., and N. Datta. 1971. fi R factors giving chloramphenicol resistance. Nature (London) 234:220221. HelUng, R. B., H. M. Goodman, and H. W. Boyer. 1974.

19.

20.

21.

22. 23.

24.

25.

26.

27. 28.

29.

30.

Analysis of endonuclease R-EcoRI fragments of DNA from lambdoid bacteriophages and other viruses by agarose-gel electrophoresis. J. Virol. 14:1235-1244. Hr tn, J. P., VIllarroed-M_anlola, M. Montae, and J. Sdhel. 1977. Transposition of Tnl to a broad host range drug resistance plasmid, p. 179-183. In A. I. Bukhari, J. A. Shapiro, and S. L. Adhya (ed.), DNA insertion elements, plasmids and episomes. Cold Spring Harbor Laboratory, Cold Spring Harbor, N.Y. Hooykm, P. J. J., H. den-DWuk-Ras, ad R. A. Scd_pesort 1980. Molecular mechanism of Ti plasmid mobilization by R-plasmids: isolation of Ti plasmids with transposon-insertions in Agrobacterium tumefaciens. Plasmid 4:64-75. ms, G. B., E. Vhne, Z. KImin, T. Forrai, and i. Koadoroi. 1979. Genetic and biochemical analysis of mutants affected in nitrate reduction in Rhizobium meliloti. J. Gen. Microbiol. 113:105-118. Ko , A., and A. W. B. Johnston. 1981. The genetics of Rhizobium. Int. Rev. Cytol. 13:191-224. Leeman, J., R. Vilaroel, B. SR", M. Van Montap, and J. Schell. 1980. Direct repetition of a 1.2 Md DNA sequence is involved in site-specific recombination by the P1 plasmid R68. Gene 10:319-328. Mandel, M., and A. Hip. 1970. Calcium-dependent bacteriophage DNA infection. J. Mol. Biol. 53:159-162. Novick, R. P., R. C. Cloves, S. N. Cohen, R. Ca HI, N. Datta, and S. Falkow. 1976. Uniform nomenclature for bacterial plasmids: a proposal. Bacteriol. Rev. 40:168189. O'Gara, F., and L. K. Duwcn. 1973. Transformation and physical properties of R-factor RP4 transferred from Escherichia coli to Rhizobium trjfolii. J. Bacteriol. 116:1177-1180. Oros, L., and T. Sik. 1970. Genetic mapping of rhizobiophage 16-3. Acta Microbiol. Acad. Sci. Hung. 17:185-194. Railff, R., W. Baer, and J. V_nogad. 1967. A dyebuoyant density method for the detection and isolation of cloned circular duplex DNA: the closed circular DNA in HeLa cells. Biochemistry 57:1514-1521. Rawkan, G. B., and F. M. Ausbel. 1981. A general method for site-directed mutagenesis in procaryotes. Nature (London) 289:85-88. Sch er, E. A. 1966. Factors affecting phage-restricting ability in Rhizobium leguminosarum strain LA. Can. J. Microbiol. 12:395-407.