Site-Specific Mutations in the TraI Relaxase and Upstream Region of Plasmid RP4

PLASMID

34, 236–239 (1995)

Site-Specific Mutations in the TraI Relaxase and Upstream Region of Plasmid RP4 S. P. COLE AND D. G. GUINEY Department of Medicine, University of California San Diego, 9500 Gilman Drive, La Jolla, California 92093-0640 Received April 11, 1995 The relaxase of RP4 nicks the double-stranded plasmid at the oriT site and binds covalently to DNA at the 5* end of the nick. The 80-kDa relaxase (TraI) is encoded on an operon with several overlapping open reading frames (ORFs). The importance in conjugation of a short ORF (traX) with a start site overlapping the 5* terminus of traI was investigated, as well as the effects of specific mutations in the relaxase. Elimination of TraX reduced the transfer efficiency by approximately 50% in several intergeneric matings, especially when Escherichia coli was the donor. While TraI was essential for transfer to occur, deletion of the C-terminus of TraI decreased, but did not eliminate plasmid transfer. Mutation of the active site tyrosine resulted in residual transfer associated with amino acid misincorporation. q 1995 Academic Press, Inc.

Plasmid RP4 carries multiple antibiotic resistance genes and can transfer to a wide variety of bacterial hosts by conjugation. This 60kb replicon has recently been sequenced (Pansegrau et al., 1994), and the genes essential for mobilization are located on two large segments of the plasmid (Guiney, 1993). During conjugation of RP4, one strand of the double-stranded DNA is nicked at the oriT site, delivered to the recipient cell, and the second strand is synthesized (Wilkins and Lanka, 1993). In preparation for transfer, the protein TraJ first binds to the oriT site. The relaxase enzyme TraI recognizes this binding, cleaves the DNA at this site, and attaches covalently to the 5* free phosphoryl group (Wilkins, 1993). A third protein associated with the relaxosome, TraH, was mutated by site-specific mutagenesis in whole RP4 (Cole et al., 1993). This knockout mutation reduced transfer efficiency up to sevenfold in a few gram-negative hosts, but did not affect Escherichia coli intraspecific matings. One of the essential motifs in TraI required for in vitro nicking is the tyrosine at position 22 (Y22), which has been shown to form a covalent bond through its hydroxyl group to 0147-619X/95 $12.00

the 5* phosphate of cytosine at the oriT (Pansegrau et al., 1993). A Y22F (Y22 replaced by phenylalanine) mutation was also tested for in vitro nicking and transfer in a plasmid consisting of the two large RP4 DNA Tra1 and 2 segments ligated to a ColD replicon (Balzer et al., 1994). Transfer efficiency of this plasmid was reduced with respect to wild-type plasmid, but some conjugation still occurred. However, protein N-terminal sequencing revealed that a small percentage of the total TraI protein in this mutant contained tyrosine at position 22 instead of phenylalanine (Balzer et al., 1994). Therefore, presence of a low level of wild-type protein from misincorporation allowed transfer of the ColD derivative. The authors concluded that this phenomenon could be an artifact due to the multiple copies of the ColD derivative. We report here the effects of transfer of whole RP4 specifically mutated at the Y22 site. For comparison, we have also constructed a TraI knockout mutant in RP4. TraI is a large protein (80 kDa), yet the domains required for nicking and covalent attachment to the DNA are located in the N-terminus of the protein (reviewed in (Cole et al., 1993),

236

Copyright q 1995 by Academic Press, Inc. All rights of reproduction in any form reserved.

/ m4330f1216

11-20-95 07:22:29

plasas

AP: Plasmid

237

SHORT COMMUNICATION

tion was constructed using pDB12m, which contains TraI truncated by a translational terminator in all three frames (MURF1) at isoleucine-11 (kindly provided by E. Lanka) (Balzer et al., 1994). This pMB1 derivative was recombined into RP4 in E. coli C2110, and the cointegrate was resolved to mutant (RP4-12m) and wild-type (RP4) plasmids. The mutant RP4-12m was compared to wildFIG. 1. Organization of the relaxase operon, indicating type RP4 in its transfer efficiency in E. coli TraJ–OrfX–TraI overlapping sequences. Lollipops indi- intraspecific mating experiments (Table 1). cate mutation positions. The TraH mutation was described Frequencies are reported as the number of previously (Cole et al., 1993). transconjugates divided by the number of recipients, in keeping with our previous determinations (Cole et al., 1993). In 35-min filter (Balzer et al., 1994)). We have previously mates, no transfer was detected between E. found by deletion mutagenesis in whole RP4 coli JA221 and JA221nalr with RP4-12m. This that the C-terminus of the protein is not re- experiment confirms the central role of TraI quired for transfer, but the efficiency decreases in conjugation of RP4. in most intergeneric matings with this mutant Single base pair alterations were made us(Cole et al., 1993). We have reconstructed this ing the long primer unique site mutagenesis mutation in RP4 and determined its frequency (LP-USE) procedure (Ray and Nickoloff, of transfer in a variety of hosts. 1992). Mutations were performed on an AccIThe relaxase operon is complex, with sev- PpuM1 fragment containing the oriT, traJ, eral overlapping open reading frames (Zie- and the 5* end of traI cloned in pUC9Cm. gelin et al., 1991). As shown in Fig. 1, traJ The target region for mutation is in the center is separated from traI by a small open reading of this fragment, which is necessary for effiframe (ORF) encoding traX. The TGA stop cient cointegrate formation and resolution. As site of traJ overlaps the ATG start site of traX for RP4-12m, the site-specific mutations genby one base. Additionally, the TGA termina- erated were recombined into RP4 to make a tion codon of TraX overlaps the GTG start cointegrate and then resolved to mutant and site of traI by one base. traX may exist solely wild-type plasmids. as a means to translationally couple traJ to The mutation at Y22 of TraI was made ustraI. Alternatively, the peptide itself may be ing the mutagenic oligomer GCGGAGCTC important for transfer, as TraX expression can G T G A A G T T C A T C A C C G A C G A G C A be detected as a LacZ fusion protein (Panse- AGGC. In addition to modifying the Y22 to grau et al., 1994). To test if orfX is important F, this oligomer also generated a new SacI for conjugation, a site-specific mutation was site. Because of degeneracy of the third base constructed in orfX on whole RP4, and the of the codon at this position, formation of the transfer phenotype of the mutant was com- new restriction site did not further alter the pared to wild-type RP4. amino acid sequence of TraI. The mutation Site-directed mutations in RP4 were first generated in pUC9Cm (pSC34) was conmade on fragments cloned on pUC9Cm and firmed by DNA sequencing. Following rethen recombined into RP4. This procedure in- combination of the mutagenized plasmid into volves cointegrate formation in a polA0 strain RP4 and resolution, the mutant (RP4-34) was (C2110) of E. coli, followed by resolution of tested for transfer in E. coli JA221 intraspethe cointegrate using cycloserine enrichment cific matings as before (Cole et al., 1993) (Ta(Cole et al., 1993). The TraI knockout muta- ble 1). Interestingly, the Y22F mutation de-

/ m4330f1216

11-20-95 07:22:29

plasas

AP: Plasmid

238

SHORT COMMUNICATION TABLE 1 TRANSFER EFFICIENCIES OF TraI MUTANT DERIVATIVES OF RP4

Plasmid in trans

(Wild-type) RP4

(D TraI) RP4-12m

(Y22F TraI) RP4-34

— pACYC184 pSC16c

2 1 1002 (6)a 2 1 1002 (3) 2 1 1002 (3)

õ8 1 1007 (3) n.d.b n.d.

6 1 1004 (6) 2 1 1003 (3) 3 1 1002 (3)

a

Numbers in parentheses represent number of times experiment was repeated. n.d., not done. c Plasmid pSC16 has the 3-kb AccI fragment containing wild-type RP4 oriT, traJ and tralI cloned in pACYC184. b

creased transfer only 33-fold in E. coli intraspecific matings, or to 3% of wild-type RP4, but did not completely eliminate transfer. This result can be explained by the small portion of TraI protein that contains a misincorporated tyrosine at amino acid 22 (Balzer et al., 1994). The relatively low copy number of RP4-34, however, refutes the theory (Balzer et al., 1994) that misincorporation of tyrosine at this position is due to a multicopy artifact. We conclude, therefore, that misincorporation of tyrosine occurs in this mutant in vivo, and that perhaps a strong positive pressure exists for tyrosine at this position. The mutation could be fully complemented with wild type TraI in trans (Table 1). For the knockout mutation at TraX, the mutagenic primer GTCCGCCCTAGGGCA-

GAGCCATAACTTTTTTAGCC was used. The ATG of TraX was modified to ATA, while the stop codon of TraJ remained intact (TGA modified to TAA). Therefore, this oligomer did not alter TraJ, but eliminated synthesis of TraX. With this oligomer, a new AvrII site was generated upstream of the TraX modification, without additional alteration of the TraJ amino acid sequence. The mutation in the pUC9Cm derivative (pSC52) was recombined into RP4 to make RP4-52. As shown in Table 2, RP4-52 reduced transfer frequency by about 50% in most recipients when E. coli was the donor (except for Shigella flexneri), but had almost no effect when E. coli was used as recipient. We conclude from this experiment that TraX is more important for transfer from E. coli

TABLE 2 BROAD HOST RANGE MATING FREQUENCIES OF RP4

AND

MUTANT PLASMIDS

Donor Å E. coli JA221 RP4 K. pneumonia P. stutzeri P. putida P. maltophilia P. aeruginosa S. flexneri E. coli JA221 a b

5.2 3.0 5.0 4.7 5.6 1.9 1.4

1 1 1 1 1 1 1

1001 1002 1001 1002 1003 1001 1001

RP4-14 (5)a (5) (4) (4) (3) (3) (20)

2.8 6.8 3.8 7.2 4.0 4.3 1.8

1 1 1 1 1 1 1

1001 1003 1001 1002 1003 1002 1002

(3) (3) (4) (3) (3) (3) (18)b

Recipient Å E. coli JA221 RP4-52

2.9 1.2 3.4 2.2 3.9 1.8 7.9

1 1 1 1 1 1 1

1001 1002 1001 1002 1003 1001 1002

(3) (3) (3) (3) (3) (3) (16)b

RP4 3.0 3.0 5.5 5.4

1 1 1 1

1002 1002 1001 1002

RP4-14 (3) (5) (3) (4)

6.4 1.1 1.2 1.7

1 1 1 1

1003 (3) 1003 (3) 100 (3) 1002 (4)

11-20-95 07:22:29

4.7 2.2 5.6 5.0

1 1 1 1

1002 1002 1001 1002

(3) (4) (3) (3)

1.2 1 1001 (3) 7.0 1 1003 (3) 9.6 1 1002 (3)

Numbers in parentheses represent number of times experiment was repeated. % Difference from RP4 is significant based on paired sample t test.

/ m4330f1216

RP4-52

plasas

AP: Plasmid

239

SHORT COMMUNICATION

and may give it some selective advantage in multiple mating events. The mutation truncating TraI at the C-terminus was described previously (Cole et al., 1993) and was reconstructed and tested for broad host range effects. A deletion of 38 bp truncates TraI from 731 to 481 amino acids, and leaves TraH intact. The new construct in RP4 (RP4-14) was confirmed by multiple restriction digests and exhibited somewhat higher transfer frequencies in several hosts than those previously reported (Cole et al., 1993). A more significant effect was noted when the mutated plasmid was mated into E. coli than when E. coli was used as the donor. This suggests that the C-terminus of TraI is important for optimal transfer into E. coli and that it assists this recipient in procuring or processing RP4 DNA. The exception to this rule was Pseudomonas putida, which displayed slightly increased transfer frequencies with truncated TraI in matings to E. coli. In this case, the C-terminus of TraI may be slightly deleterious for optimal transfer of RP4. Transfer efficiencies of RP4-14 and RP452 in E. coli JA221 intraspecific matings at 427 or 217C were similar to those obtained at 377C (data not shown). We deduce that the Ctermini of TraI and orfX are not required for transfer at either physiological or environmentally relevant temperatures. In conclusion, we have created a variety of site-directed mutations in the relaxase (TraI) and upstream region (TraX), recombined them into RP4, and determined their importance in transfer. While TraI is clearly required for transfer, a mutation at the tyrosine at amino acid 22 is misincorporated significantly enough to allow transfer to occur at reduced, but substantial, levels. This may indicate a strong pressure to maintain the tyrosine at this position. The C-terminus of TraI is not required for transfer to occur, but assists transfer into E. coli. This portion of the enzyme may function in DNA processing events in the re-

/ m4330f1216

cipient, with host proteins being able to substitute in many but not all cases. Finally, the importance for transfer of the overlapping ORF traX was found to improve efficiencies for donor E. coli. All of these minor improvements in transfer efficiency may work in concert to improve RP4 transmission among different bacterial populations over long periods of time in the natural environment. ACKNOWLEDGMENT This work was supported by Public Health Services Grant GM28924 from the National Institutes of Health.

REFERENCES BALZER, D., PANSEGRAU, W., AND LANKA, E. (1994). Essential motifs of relaxase (TraI) and TraG proteins involved in conjugative transfer of plasmid RP4. J. Bacteriol. 176, 4285–4295. COLE, S. P., LANKA, E., AND GUINEY, D. G. (1993). Sitedirected mutations in the relaxase operon of RP4. J. Bacteriol. 175, 4911–4916. GUINEY, D. G. (1993). Broad host range conjugative and mobilizable plasmids in Gram-negative bacteria. In ‘‘Bacterial Conjugation’’ (D. B. Clewell, Ed.), pp. 75– 103. Plenum, New York. PANSEGRAU, W., SCHRO¨DER, W., AND LANKA, E. (1993). Relaxase (TraI) of IncPa plasmid RP4 catalyzes a sitespecific cleaving-joining reaction of single-stranded DNA. Proc. Natl. Acad. Sci. USA 90, 2925–2929. PANSEGRAU, W., LANKA, E., BARTH, P. T., FIGURSKI, D. F., GUINEY, D. G., HAAS, D., HELINSKI, D. R., SCHWAB, H., STANISCH, V. A., AND THOMAS, C. M. (1994). Complete nucleotide sequence of Birmingham IncPa plasmids—Compilation and comparative analysis. J. Mol. Biol. 239, 623–663. RAY, F. A., AND NICKOLOFF, J. A. (1992). Site-specific mutagenesis of almost any plasmid using a PCR-based version of unique site elimination. BioTechniques 13, 342–348. WILKINS, B., AND LANKA, E. (1993). DNA processing and replication during transfer between gram-negative bacteria. In ‘‘Bacterial Conjugation’’ (D. B. Clewell, Ed.), pp. 105–136. Plenum, New York. ZIEGELIN, G., PANSEGRAU, W., STRACK, B., BALZER, D., KRO¨GER, M., KRUFT, V., AND LANKA, E. (1991). Nucleotide sequence and organization of genes flanking the transfer origin of promiscuous plasmid RP4. DNA Sequence J. DNA Sequencing Mapping 1, 303–327. Communicated by D. R. Helinski

11-20-95 07:22:29

plasas

AP: Plasmid