Functional and Mutational Analysis of Conjugative Transfer Region 2 (Tra2) from the IncHI1 Plasmid R27

Abstract

The transfer 2 region (Tra2) of the conjugative plasmid drR27 (derepressed R27) was analyzed by PSI-BLAST, insertional mutagenesis, genetic complementation, and an H-pilus assay. Tra2 contains 11 mating-pair formation (Mpf) genes that are essential for conjugative transfer, 9 of which are essential for H-pilus production (trhA, -L, -E, -K, -B, -V, -C, -P, and -W). TrhK has similarity to secretin proteins, suggesting a mechanism by which DNA could traverse the outer membrane of donors. The remaining two Mpf genes, trhU and trhN, play an auxiliary role in H-pilus synthesis and are proposed to be involved in DNA transfer and mating-pair stabilization, respectively. Conjugative transfer abilities were restored for each mutant when complemented with the corresponding transfer gene. In addition to the essential Mpf genes, three genes, trhO, trhZ, and htdA, modulate R27 transfer frequency. Disruption of trhO and trhZ severely reduced the transfer frequencies of drR27, whereas disruption of htdA greatly increased the transfer frequency of wild-type R27 to drR27 levels. A comparison of the essential transfer genes encoded by the Tra2 and Tra1 (T. D. Lawley, M. W. Gilmour, J. E. Gunton, L. J. Standeven, and D. E. Taylor, J. Bacteriol. 184:2173-2183, 2002) of R27 to other transfer systems illustrates that the R27 conjugative transfer system is a chimera composed of IncF-like and IncP-like transfer systems. Furthermore, the Mpf/type IV secretion systems encoded by IncH and IncF transfer systems are distinct from that of the IncP transfer system. The phenotypic and ecological significance of these observations is discussed.

Type IV secretion systems play a central role in the pathogenesis of gram-negative bacteria. For example, Helicobacter pylori and Bordetella pertussis use type IV secretion systems to secrete virulence factors into their mammalian hosts (10, 40). In addition, conjugative plasmids transfer from donor to recipient bacteria, frequently between distantly related bacteria, using a type IV secretion system (8). Conjugation is particularly problematic since the transfer of plasmid DNA carrying multiple antibiotic resistance genes into pathogens renders the corresponding antibiotics ineffective in treating such pathogens. Despite the clinical importance of type IV secretion systems, the general mechanism by which they secrete macromolecules remains unknown (3).

The type IV secretion system used by bacterial plasmids is referred to as the mating-pair formation (Mpf) apparatus. The Mpf apparatus is a membrane-associated protein complex that functions in producing conjugative pili and subsequently transferring plasmid DNA (15). The Mpf apparatus consists of 12 or 13 proteins (15, 18), including the pilin subunit and usually a pilin-processing protein (i.e., acetylase [35] or peptidase [20]). The processed pilin subunits insert into the inner membrane before being assembled onto the donor surface by all, or some, of the Mpf apparatus components (32, 37). The IncF system produces F-pili, which initiate conjugative transfer by identifying suitable recipients cells and bringing them into close contact to form a stable mating pair (14). In other model systems, such as the IncP system, the exact role of the P-pili remains to be determined, although donors containing IncP plasmids also form stable mating pairs (47).

At the site of close contact between donor and recipient cells are regions of tightly appressed membranes, known as conjugative junctions (11, 47). Conjugative junctions are the most likely regions through which plasmid DNA would transfer between cells (27). It is thought that the conjugative junction consists of Mpf proteins, although the exact proteins remain to be identified (11, 47). The transferred DNA strand (T-strand), which exists as a relaxosome-T-strand nucleoprotein complex in the donor, is transferred into the recipient in the 5′-to-3′ direction (25). Transfer has been proposed to occur via a two-step mechanism, whereby the relaxosome-T-strand first interacts with the coupling protein and then is actively pumped into the recipient by the coupling protein through the Mpf apparatus (30). After plasmid transfer and establishment, the recipient becomes a donor.

Some of the more clinically important antibiotic resistance plasmids belong to the IncHI1 plasmid group (50). For over three decades, IncHI1 plasmids have been implicated as a significant factor in the persistence and reemergence of Salmonella enterica serovar Typhi, the causative agent of typhoid fever (22). The continued prevalence of IncHI1 plasmids can be partially attributed to the conjugative transfer functions of these plasmids, since horizontal transfer maintains a plasmid within a bacterial population. We have previously characterized the Tra1 region of derepressed R27 (drR27) and found that it contains the origin of transfer and genes encoding the relaxosome, coupling protein, and three Mpf proteins (26). drR27 is the prototypical IncHI1 plasmid that has an elevated transfer frequency compared to the wild type due to an insertion into htdA (48, 52). In addition, we have performed a preliminary analysis of the Tra2 region (45). In this study, we performed a functional and mutational analysis of the Tra2 region and compared the essential transfer components of R27 to other transfer systems. IncHI1 plasmids contain transfer region components that are chimeras composed of IncF-like and IncP-like plasmids transfer systems. We also note that the IncH and IncF plasmid Mpf systems are quite distinct from that of IncP. We discuss the phenotypic and ecological significance of the differences between IncH/F and IncP Mpf systems.

MATERIALS AND METHODS

Bacterial strains, growth conditions, and plasmids.

Escherichia coli strains and plasmids used in this study are listed in Table 1. E. coli was grown at 27 or 37°C in Luria-Bertani broth (Difco Laboratories, Detroit, Mich.) with shaking or on Luria-Bertani agar plates. Antibiotics were added at the following concentrations when appropriate: ampicillin, 100 mg/liter; tetracycline, 10 mg/liter; nalidixic acid, 50 mg/liter; rifampin, 50 mg/liter; and chloramphenicol, 16 mg/liter. X-Gal (5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside) and IPTG (isopropyl-β-d-thiogalactoside) were each used at 100 mg/liter.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Relevant genotype, phenotype, or characteristic(s)a	Cloning or gene disruption primers (5′ to 3′)b	Source or reference
E. coli
DH5	supE44 lacU169 (φ80 lacZM15) hsdR17 recA1 endA1 gyrA96 thi-1 relA1
DY330	W3110 ΔlacU169 gal490 λc857 Δ(cro-bioA)		53
DY330N	Temperature-resistant revertant; Nalr		26
DY330R	Temperature-resistant revertant; Rif		26
DT1942	RG192 (Rif) with pDT1942
RG192	ara leu lac		26
Plasmids
R27	IncHI1 plasmid; Tcr		48
pDT1942	Derepressed R27; R27::TnlacZ		26
pMS119EH/HE	IPTG-inducible expression vector; Ampr		49
pGUN110	pDT1942 with cat inserted into trhA; Tcr Kanr Cmr	F: AACTGACATTGAATACTAACGTTGAAAGAGTAAAAACTAATGTGACGGAAGATCACTTC	This study
R: GATATCGCCGAAGGCTCCATCATCGGAACCCGCGTACGCATTATTCAGGCGTAGCACCAG
pGUN205	trhA cloned into pMS119EH; Ampr	F: TATAGAATTCAATGGAACTGACATTGAATAC	This study
R: TATAGGATCCTTAATGGTGATGGTGATGGTGCAGAGGAATACCAGCATCCAG
pMWG253	pDT1942 with cat inserted into trhL; Tcr Kanr Cmr	F: AGTACGAGATACCGCCACATACATATCGCTTTCCATACAGCTGTGACGGAAGATCACTTC	This study
R: CGCCACAACTACCGAAGTTATAAAGAAATCAAAAATATTATTATTCAGGCGTAGCACCAG
pMWG270	trhL cloned into pMS119EH; Ampr	F: TATAGAATTCAATGACAAGCCAGTACGAGATAC	This study
R: ATATGGATCCTTAATGGTGATGGTGATGGTGAGAATAGAGGTTTCTGATAAATG
pDT2970	pDT1942 with mini-Tn10 inserted into trhE; Tcr Cmr Kanr		45
pTDL242	trhE cloned into pMS119EH; Ampr	F: TATAGAATTCAATGAAACTTCTCAGCAGGTT	This study
R: TATAGGATCCTTAATGGTGATGGTGATGGTGGTTACCTATTGACGGCGG
pDT2981	pDT1942 with mini-Tn10 inserted into trhK; Tcr Cmr Kanr		45
pTDL243	trhK cloned into pMS119EH; Ampr	F: TATAGAATTCAATGACTTCAAAAAAGGTTTTTC	This study
R: TATAGGATCCTTAATGGTGATGGTGATGGTGTCTTTTAATCAGGCTTGGGC
pDOB3	pDT1942 with cat inserted into orf030; Tcr Kanr Cmr	F: TAGTTGTCTTTTTGGGTACCAGTTATTACTACCAAACTAACTGTGACGGAAGATCACTTC	This study
R: ATCCGCAAGCGTTGGGTCGCTTATAACAATCGTGCTTTTCTTATTCAGGCGTAGCACCAG
pDT2971	pDT1942 with mini-Tn10 inserted into trhB; Tcr Cmr Kanr		45
pTDL42	trhB cloned into pMS119EH; Ampr	F: ATATGAATTCATGGACATTAAAAAGGCCTC	This study
R: TATAGGATCCTCAGTGATGGTGATGGTGATGTGCGCGGCTTGTCAGGTTTG
pMWG263	pDT1942 with cat inserted into orf028; Tcr Kanr Cmr	F: GAGTGAGTCTGCTGATAGCTTAAATATCGACGCGGCTAAACTGTGACGGAAGATCACTTC	This study
R: TGTATCCATAGATCACGATAAGAACTGCCAGGCCTGCGACTTATTCAGGCGTAGCACCAG
pMWG264	pDT1942 with cat inserted into orf027; Tcr Kanr Cmr	F: AACGAAACCTTTAGACGGTGTATTGGCGAAAATGAAAAGCCTGTGACGGAAGATCACTTC	This study
R: GCGGTGACCACTGACGGACGGTATACAATCACGGGGGGGATTATTCAGGCGTAGCACCAG
pMWG265	pDT1942 with cat inserted into trhV; Tcr Kanr Cmr	F: GTTATTTTCAGCGCTCTGGTTGTTGGTTCATCTTCATACCCTGTGACGGAAGATCACTTC	This study
R: GAGAGATCATCACGATTGTTCGTCAGTTCCATAACTTCACTTATTCAGGCGTAGCACCAG
pMWG296	trhV cloned into pMS119EH; Ampr	F: TATAGAATTCAATGAACATTAAAAAAGTTATTTTCAG	This study
R: ATATGGATCCTTAATGGTGATGGTGATGGTGGAGTTGTCTCACAATACCG
pGUN141	pDT1942 with cat inserted into trhZ; Tcr Kanr Cmr	F: CTGCCAGACTGCAAGTGCTTACCCTTATCGCATTTACACTCTGTGACGGAAGATCACTTC	This study
R: ACATTGAAATTCCGAAAGTTGATCCAGCTTCAGTATTGCTTTATTCAGGCGTAGCACCAG		/PICK>
pGUN206	trhZ cloned into pMS119EH; Ampr	F: TGAATTCAATGAAATATATATTGACCG	This study
		R: TGGATCCTTATTGAACAGGGTGATGGGC
pGUN139	pDT1942 with cat inserted into orf017; Tcr Kanr Cmr	F: GAGTTTCAGCATTTATTCTTACCTTACTTGTGACTGGATGCTGTGACGGAAGATCACTTC	This study
R: CTGAACTCTCAATCAGAAAAGTAACTTTTCCGTCGGTGTAGTTATTCAGGCGTAGCACCAG
pGUN142	pDT1942 with cat inserted into orf016; Tcr Kanr Cmr	F: GAACTGATAATTTTGCCGTTGCTGGCATTTTTAGGTCCATCTGTGACGGAAGATCACTTC	This study
R: ATTCCGTTGACGTAATTTGCAGCAATAGTAATAAATTCATTTATTCAGGCGTAGCACCAG
pGUN144	pDT1942 with cat inserted into trhO; Tcr Kanr Cmr	F: TCCGAGCACGTCCAGACAATTGATCTGACTAATTTTTCAGCTGTGACGGAAGATCACTTC	This study
R: AAGAGATTTCCTGTGGAGAAAACGATCAACTCCTCATCAGTTATTCAGGCGTAGCACCAG
pGUN77	trhO cloned into pMS119EH; Ampr	F: TGAATTCAATGGCATCCGAGCACGTCCAG	This study
		R: TGGATCCTTATAATGTGGCCTCTATATTATTATC
pTDL205	pDT1942 with cat inserted into htdA; Tcr Kanr Cmr	F: GCCCCAATTTCATCGCTTAAGTACGACTGGGAGTGGGCTCTGTGACGGAAGATCACTTC	This study
R: GCAGGCGGAATGACCACTTCGATGAGTATTGTATCAATTGTTATTCAGGCGTAGCACCAG
pTDL244	htdA cloned into pMS119HE; Ampr	F: TATAAAGCTTATGTTATCACGCAACTCACT	This study
R: TATATCTAGATTAATGGTGATGGTGATGGTGACTTTCATCCTCCAGTTCC
pTDL218	pDT1942 with cat inserted into htdF; Tcr Kanr Cmr	F: TCTTACTCGATCTTTAACGCAAAACCGTCGAAGCCTGACTCTGTGACGGAAGATCACTTC	This study
R: ACAGAAGTGGTTTTAAGAACTTTGCCAGTAAAATCGTACATTATTCAGGCGTAGCACCAG
pTDL206	pDT1942 with cat inserted into htdK; Tcr Kanr Cmr	F: CAATCTGTCGCGATGGCATTCGAAGTAAATCTCGACAAAACTGTGACGGAAGATCACTTC	This study
R: TTCAGTCTCAGTTACAAACGCCGAATTATCAGCAGAGCGATTATTCAGGCGTAGCACCAG
pDOB7	pDT1942 with cat inserted into orf009; Tcr Kanr Cmr	F: GTCGAGTACTCGAAGGAACCGACTATGCGAACCGTTGATACTGTGACGGAAGATCACTTC	This study
R: TTATAGCGGCCAACAGACACCTTCAATGGCAATTCCATCGTTATTCAGGCGTAGCACCAG
pGUN58	pDT1942 with cat inserted into trhP; Tcr Kanr Cmr	F: ACGTGCCAATAAATCTCGGAATCCTGATACTGGTTATGTTTGTGACGGAAGATCACTTC	This study
R: AGCTATGATATTCCCTGACCTTACCGATTTATCCTTCAGATTATTCAGGCGTAGCACCAG
pMWG114	trhP cloned into pMS119EH; Ampr	F: GAATTCATGAGAGAATACATACCTAAG	This study
		R: GGATCCTTAATGGTGATGGTGATGGTGCGCAAAGATCGCATATGTTTTC
pDT2985	pDT1942 with mini-Tn10 inserted into trhW; Tcr Cmr Kanr		45
pTDL245	trhW cloned into pMS119EH; Ampr	F: TATAGAATTCCATGGGACTCACGTTTTTGG	This study
R: TATAGGATCCTTAATGGTGATGGTGATGGTGATTCTCCTCTGCCGCGATTT
pDT2953	pDT1942 with mini-Tn10 inserted into trhU; Tcr Cmr Kanr		45
pTDL246	trhU cloned into pMS119HE; Ampr	F: TATAAAGCTTATGCTGTCGCTCTCGCTG	This study
		R: TATATCTAGATTAATGGTGATGGTGATGGTGGTTGGGTATATACCTGACGC
pDT2988	pDT1942 with mini-Tn10 inserted into trhN; Tcr Cmr Kanr		45
pTDL247	trhN cloned into pMS119HE; Ampr	F: TATAAAGCTTATGAAACATAGAAAAATAATTTCTTC	This study
		R: TATATCTAGATTAATGGTGATGGTGATGGTGATATTTTGGTATTTCCTTACCTG
pDOB5	pDT1942 with cat inserted into orf004; Tcr Kanr Cmr	F: GTTTCAGCTTTAGTTGCAAGTGCGTCTCTGGTAGGGTGTGCCTGTGACGGAAGATCACTTC	This study
R: CAAACCCGGCCAGAGAGTAAGCATTTTTGGCTACAGAATCTTATTCAGGCGTAGCACCAG
pDOB6	pDT1942 with cat inserted into trhI; Tcr Kanr Cmr	F: CAACAGATGAGCAAAGGGTCATCATTGAAAATGCAAATGCCTGTGACGGAAGATCACTTC	This study
		R: CGGGTAAATGTGACCATCCCCACTTTGGCATAAGGGAACTTATTCAGGCGTAGCACCAG

^aAbbreviations: Nal^r, nalidixic acid resistance; Rif^r, rifampin resistance; Tc^r, tetracycline resistance; Km^r, kanamycin resistance; Amp^r, ampicillin resistance.

^bUnderlined regions of cloning primers designate restriction endonuclease cleavage sites; bold bases designate a His tag and an engineered stop codon. Underlined regions of disruption primers designate cat-specific bases.

DNA manipulations.

R27 DNA was isolated using either ultracentrifugation in a CsCl-ethidium bromide gradient (52) or the large-construct kit (Qiagen Inc., Mississauga, Ontario, Canada). Standard recombinant DNA methods were performed as described by Sambrook and Russell (46). Restriction endonucleases were used as specified by the manufacturer (Invitrogen, Carlsbad, Calif.), and digested DNA was analyzed by agarose gel electrophoresis.

Computer analysis.

Laser gene software (DNASTAR Inc., Madison, Wis.) was used for nucleotide sequence analysis. The predicted protein sequence for each open reading frame (ORF) was compared to the GenBank nonredundant database using PSI-BLAST. We identified conserved motifs manually or with ScanProsite (http://ca.expasy.org/tools/scnpsite.html) and obtained predictions for molecular weight and pI values with Compute pI/M_w (http://ca.expasy.org/tools/pi_tool.html).

Mutagenesis.

Mutants with mutations of trhE, trhK, trhB, trhC, trhW, trhU, and trhN were created prior to this study using random mini-Tn10 mutagenesis, as described previously (45). Mini-Tn10 consists of a chloramphenicol resistance cassette flanked by Tn10 inverted repeats (23). To identify the insertion position for each mutant, we sequenced the region flanking the mini-Tn10 insertions by using the dideoxy method (Sequenase version 2; United States Biochemical Co., Cleveland, Ohio) with primers PNE11 (5′TATTCTGCCTCCCAGAGCCT) and PNE12 (5′TGGTGCGTAACGGCAAAAGC). Sequence results were compared to the complete nucleotide sequence of R27 (accession no. NC_002305). All remaining Tra2 genes (trhA, trhL, orf030, orf028, orf027, trhV, trhZ, orf017, orf016, trhO, htdA, htdF, htdK, orf009, trhP, orf004, and trhI) were mutated by gene disruption using the E. coli recombination system as described previously (26, 53). Gene disruptions were created by insertion of a chloramphenicol resistance cassette (cat from mini-Tn10) into each of the above-mentioned ORFs in a sequence-specific fashion. DNA substrates were generated through PCR with primers (∼60 nucleotides) that produced a linear chloramphenicol acetyltransferase (CAT) cassette with 40-bp terminal arms homologous to the desired target site (Table 1). To screen presumptive mutants, the target gene was PCR amplified (using cloning primers in Table 1) and analyzed by agarose gel electrophoresis (1% agarose). An increase in the size of the ORF by 900 bp demonstrated that the CAT cassette had been inserted into the target gene

Cloning of transfer genes.

The primer sequences used for cloning Tra2 transfer genes are listed in Table 1. PCR products, which included terminal EcoRI and BamHI restriction endonuclease sites for cloning into pMS119EH or terminal HindIII and XbaI for cloning into pMS119HE, were cloned into pGEM-T (Promega) and then subcloned into either pMS119EH or pMS119HE. A His₆ tag was engineered into the C terminus of each protein to allow detection of the recombinant protein by immunoblot analysis with anti-His₆ antibodies (data not shown).

Conjugation assay.

Conjugal transfer of drR27 was performed as previously described (26). For complementation experiments, E. coli donors (DY330R) which contained the R27 Tra2 mutant and an expression vector encoding the appropriate transfer protein were mated with recipients (DY330N). To determine the effect of overexpressing transfer proteins on the transfer frequency, each transfer protein was expressed in trans within a donor containing drR27 during conjugation experiments. Transfer frequencies were expressed as transconjugants per donor.

Phage plaque assays.

To determine the Hgal infectivity of Tra2 mutants, phage spot tests were performed as previously described (26, 31).

RESULTS

Nucleotide analysis of the Tra2 region.

The Tra2 region is 36 kb in length and contains 28 ORFs, 4 of which code for two separate partitioning modules (Fig. 1) (T. D. Lawley and D. E. Taylor, unpublished data). All ORFs are transcribed in the same direction, with the exception of ORF004 and TraO to TraZ, which appear to constitute an operon. Previous analysis indicated that the Tra2 region encodes mating-pair formation (Mpf) proteins that are related to Mpf proteins of IncF transfer systems, as determined by Basic Local Alignment Search Tool (BLAST) analysis (45). Prior to our mutational and functional analysis of the Tra2 region, we performed position-specific iterated BLAST (PSI-BLAST) analysis, a more sensitive version of BLAST analysis, and protein alignments with the predicted transfer protein products of the Tra2 region and our findings are described quantitatively (Table 2). TrhL, -E, -B, -V, -C, -P, -W, -U, and -N are homologous to Mpf proteins from the F factor (45), with levels of identity ranging from 20 to 32% (Table 2). The N terminus of TrhW is similar to TrbC of the F factor, suggesting that there is a gene fusion in R27 and that the functions of these two proteins are coupled in the F factor. TrhP (previously named TrhF) is homologous to the signal peptidase I protein family, which includes TraF, a transfer peptidase of IncP plasmids (48).

FIG. 1.

ORF map of Tra2 of R27. Black ORFs indicate genes required for conjugative transfer and Hgal infection. Stippled ORFs indicate genes which regulate R27 transfer frequency. White ORFs are not essential for conjugative transfer. Grey ORFs are partitioning genes, and the Inc region is a 3-kbp intergenic region involved in IncHI1 plasmid incompatibility (16). Black bent arrows represent the insertion location of mini-Tn10 and the direction of CAT transcription. White bent arrows represent the insertion location of the CAT cassette and the direction of transcription. The ORF map corresponds to coordinates 0 to 40 kbp and ORF003 to ORF034 on R27 (48).

TABLE 2.

Summary of computer analysis of nucleotide sequence of Tra2 of R27

Gene or ORFa	Length (aa)	Mol wt (103)	Motifb	PSI-BLAST results (protein, source, and length of homolog)	Degree of relatedness (identical residues/range of identity = % identity)	Accession no. of homolog
trhA	117	12.4	None	HtdZ of R478 (126 aa)	77/117 = 65%	AAL27017
trhL	105	12.4	None	TraL of F (103 aa)	21/105 = 20%	BAA97945
trhE	261	29.2	None	TraE of F (188 aa)	30/147 = 20%	BAA97946
trhK	410	44.4	None	TraK of F (242 aa)	48/230 = 21%	BAA97947
				HrpH of P. syringae (701 aa)	33/166 = 20%	AAC05014
orf30	146	16.4	None	HtdO of R478 (153 aa)	60/138 = 43%	AAL27021
trhB	452	48.3	None	TraB of F (475 aa)	96/391 = 24%	BAA97948
orf28	170	19.0	None	HtdV of R478 (112 aa)	39/75 = 52%	AAL27016
orf27	250	28.0	None	DsbC of SXT (230 aa)	53/196 = 27%	AAL59715
trhV	316	33.8	Lipoprotein motif	TraV of F (171 aa)	24/96 = 24%	BAA97952
trhC	893	102.0	Walker A and Walker B	TraC of F (875 aa)	200/899 = 22%	BAA97954
trhZ	130	14.1	None	None
orf17	173	19.0	Lipoprotein motif	None
orf16	279	30.8	None	None
trhO	315	34.9	None	None
htdA	150	16.7	None	HtdA of R478 (150 aa)	126/149 = 84%	AAB05912
htdF	144	15.8	None	HtdF of R478 (141 aa)	71/142 = 50%	AAB37116
htdK	187	20.8	None	HtdK of R478 (177 aa)	88/167 = 52%	AAB37117
orf9	1,406	155.5	None	None
trhP	170	19.4	Signal peptidase 1	TraF of RP4 (177 aa)	28/124 = 22%	AAA98335
trhW	502	56.5	None	TraW of F (210 aa) in C terminus	49/193 = 25%	BAA97956
				TrbC of F (212 aa) in N terminus	34/146 = 23%	AAB61940
trhU	335	37.0	None	TraU of F (330 aa)	104/320 = 32%	AAB61938
trhN	1,058	116.7	None	TraN of F (602 aa)	25/112 = 22%	BAA97959
					27/132 = 20%
orf4	290	31.1	Lipoprotein motif	None
trhI	594	67.6	DNA helicase motifs	DNA helicase II of Serratia marcescens (720 aa)	167/640 = 26%	AAC46278

^aGene designation and coordinates as in reference 48.

^bConserved Domain Database (http://www.ncbi.nlm.nih.gov/) and ScanProsite (http://ca.expasy.org/tools/scnpsite.html).

ORF031 is similar to TraK of the F factor and has been named TrhK (Fig. 2; Table 2). Interestingly, TrhK also has a significant level of similarity to the HrcC/HrpH secretin family present in the type III secretion system of organisms such as Pseudomonas fluorescens (42). The similarity shared between TrhK-like proteins and HrcC-like proteins is present in the C terminus (Fig. 2), the defining domain of secretin proteins (39). The alignment between TrhK-like proteins, HrpH-like proteins, and PulD (prototypical secretin) suggests that TrhK-like proteins contain the β-domain, predicted to form the outer membrane ring, and the S-domain, which interacts with a lipoprotein (Fig. 2) (19).

FIG.2.

Alignment of type IV secretion secretin-like proteins of the TrhK family with the HrcC TTSS secretins and the PulD type II secretion system secretin. Represented below the alignment are the β-domain and S-domain from PulD. The β-domain is made up of the transmembrane regions which are embedded into the outer membrane, and the S-domain interacts with the lipoprotein. Alignment was performed with ClustalW using the BLOSUM scoring matrix, and shading was performed using GeneDoc in conservative mode with shading to four levels. Proteins (in the order in which they are listed in the alignment): TraK of F factor (GenBank accession no. AAC44189), TraK of R391 (AAM08021), TraK of Rts1 (BAB93771), HtdP of R478 (AAL27020), TrhK of R27 (AAD54050), TraK of pED208 (AAM90706), TraK of pNL1 (AAD03962), TraK of SXT (AAL59718), HrcC of Pantoea stewartii (AAG01463), HrcC of Erwinia amylovora (AAB49179), HrcC of Erwinia chrysanthemi (AAC31975), HrcC of Pectobacterium carotovorum (AAK97280), HrcC of Pseudomonas syringae (AAC05014), RscC of Pseudomonas fluorescens (AAK81929) and PulD of Klebsiella oxytoca (P15644).

Besides being similar to IncF plasmid Mpf proteins, the Tra2 Mpf proteins are similar to the Mpf proteins from the IncHI2 plasmid R478 from Serratia marcescens (accession no. AF030442) (41), pNL1 from Novosphingobium aromaticivorans (NC_002033) (44), the IncJ element R391 from Providencia rettgeri (AY090559) (5), the IncT plasmid Rts1 from Proteus vulgaris (NC_003905) (38), and the SXT element from Vibrio cholerae (AY055428) (reference 4 and data not shown).

The presence of nine Mpf proteins in the Tra2 region and three Mpf proteins in the Tra1 region (26) which are homologous to Mpf proteins from the F factor means that R27 contains an equivalent to each of the essential Mpf proteins from the F factor (15), with the exception of the pilin subunit. Given the presence of trhP, whose gene product is homologous to the pilin-processing protein TraF from IncP plasmids (12), we therefore aligned the putative core regions of TrhA and HtdZ (R478) with the core region of TrbC, the pilin of RP4 and R751, to identify conserved regions (Fig. 3). The core region of TrbC is processed with the removal of the N-terminal 37 residues by the host-encoded LepB and the removal of the C-terminal 27 residues by an unknown host peptidase (12). This 82-amino-acid (aa) peptide, which contains the four residues to be removed by TraF, was aligned with the processed form of TrhA and HtdZ (the leader peptide, predicted by SignalP, has been removed) using ClustalW. This alignment identified 15% identity and 34% similarity among these four proteins. Within the conserved regions are several residues which are highly conserved between TrbC and several of its homologs, including VirB2 from the Ti plasmid (13), particularly in the region which is cleaved by the transfer peptidase (Fig. 3).

FIG. 3.

Alignments of the processed pilin subunits from the IncHI (R27 and R478) and IncP (R751 and RP4) transfer systems. The protein sequences of the pilin subunits have the leader sequences removed and the cleavage site shown (gray arrow). The protein sequences of the IncP pilin subunits are missing the 27 C-terminal residues, which are cleaved by an unknown peptidase (black arrow 1), and the site of cleavage by the TraF peptidase is shown (black arrow 2). Alignments were performed in ClustalW and shaded using GeneDoc in conservative mode and three levels. Accession numbers are as follows: TrhA, AAD54053; HtdZ, AAL27017; TrbC R751, NP_044241; and TrbC RP4, AAA26429.

To identify conserved sequence motifs, the predicted protein sequence of each of the Tra2 ORFs was compared to both the Conserved Domain database and the Prosite database (Table 2). A lipoprotein motif was detected in TrhV, which is characteristic of the TraV protein family (21). Lipoprotein motifs were also found in ORF017 and ORF004. TrhC contains Walker A and Walker B motifs, suggesting that this protein binds ATP and possibly energizes some aspect of conjugation. TrhP contains the signal peptidase 1 motif, consistent with the PSI-BLAST analysis. TrhI contains DNA helicase II motifs, also consistent with the PSI-BLAST analysis.

Gene disruptions and identification of transfer mutants.

Seven mini-Tn10 transfer mutants have been identified within the Tra2 region (trhE, -K, -B, -C, -W, -U, and -N) (45). To identify a role in transfer for the remaining 17 Tra2 genes (the entire Tra2 excluding 4 partitioning genes), we systematically created gene disruptions of each of these ORFs in drR27, as previously described (26). Each mutant was then tested for its ability to transfer (Table 3). Gene disruptions of trhA, -L, -V, and -P abolished conjugative transfer of drR27. Disruptions of trhO and trhZ reduced transfer to 0.0002 and 0.006% of the transfer frequency of drR27, respectively, whereas disruption of most of the remaining genes had a minimal effect on transfer. Since mutations were made in drR27 (wild-type R27 with TnlacZ inserted into htdA), we inserted a CAT cassette into htdA within wild-type R27 to see if a different disruption of this gene (CAT [1 kb] versus TnlacZ [8.7 kb]) had the same effect on transfer. An insertional disruption of htdA by CAT increased the transfer frequency of R27 by 6,000-fold, as did the insertion by TnlacZ (52). These data, combined with those from the random transposon mutagenesis experiments, indicate that 11 genes within the Tra2 region are essential for conjugative transfer. In addition, trhO and trhZ play a role in conjugative transfer by enhancing the transfer frequency of drR27, whereas htdA represses the transfer frequency of wild-type R27.

TABLE 3.

Effect of mutations in Tra2 on drR27 conjugation

drR27 mutationa in the donor strain	Transferb phenotype of mutantc	Relative conjugation frequency of drR27 mutant (%)d	Relative conjugation frequency during complementation (%)e	Relative conjugation frequency of drR27 with overexpressed transfer proteinf (%)	Plaque with Hgalg
trhA	−	0	0.2	93	−
trhL	−	0	0.7	208	−
trhE	−	0	2.2	150	−
trhK	−	0	15.9	130	−
orf030	+	55	ND	ND	+
trhB	−	0	14	79	−
orf028	+	10	ND	ND	+
orf027	+	100	ND	ND	+
trhV	−	0	0.2	72	−
trhC	−	0	90	95	−
trhZ	+	0.006	73	23	−
orf017	+	43	ND	ND	+
orf016	+	95	ND	ND	+
trhO	+	0.0002	77	83	+
htdAh	+++	600,000	120	90	+++
htdF	+	7	ND	ND	+
htdK	+	7	ND	ND	+
orf009	+	0.6	ND	ND	+
trhP	−	0	131	54	−
trhW	−	0	144	88	−
trhU	−	0	12	90	±
trhN	−	0	133	82	±
orf004	+	34	ND	ND	+
trhI	+	76	ND	ND	+

^aGene designation according to complete sequence of R27 (48) or this study. The donor strain was DY330R.

^bAbility of mutant to transfer via conjugation. Detectable matings are scored as +, and nondetectable (<10⁻⁸ transconjugants/donor) matings are scored as −.

^cCAT cassette introduced into gene by random transposon mutagenesis or homologous recombination as described in Materials and Methods.

^dTransfer frequency of drR27 mutant relative to drR27 [(drR27 mutant frequency/drR27 frequency) × 100%]. Due to the inherent variation of drR27 transfer frequency, the relative frequency was determined using the drR27 transfer frequency for each experiment. drR27 transferred with an average frequency of 5 × 10⁻¹ transconjugant/donor. Values represent the average values of two independent experiments.

^eRestoration of conjugation by complementation in trans using a cloned gene (see Table 1); ND, not determined. Transconjugants per donor cell determined after 18 h of liquid mating with recipient DY330N, as described in Materials and Methods.

^fThe designated transfer protein was overexpressed in donor containing drR27 during the conjugation assay. Relative transfer frequency of drR27 with transfer protein/drR27 for each experiment. Values represent the average value of two independent experiments. ND, not determined.

^gAbility of H-pilus-specific RNA bacteriophage to form plaques with E. coli harboring the designated mutant of drR27 (agar spot test). E. coli containing drR27 forms plaques when infected with Hgal.

^hExperiments and mutagenesis on htdA were performed on wild-type R27, which transfers at a frequency of 5 × 10⁻⁵ transconjugant/donor. All values in this row are relative to wild-type R27.

Genetic complementation of transfer-deficient mutants.

The wild-type gene corresponding to each of the transfer-deficient mutants was cloned into the expression vector pMS119EH (or HE) (Table 1). In addition, the trhO, trhZ, and htdA genes were cloned to test for an effect on the corresponding mutants. Each clone was transformed into E. coli containing the appropriate drR27 transfer mutant. When the wild-type version of each transfer gene was expressed in trans with each of the transfer mutants, conjugative transfer was restored to various degrees for each of the 11 transfer mutants (Table 3), demonstrating that each of these genes is essential for conjugative transfer. When trhZ and trhO were expressed in trans within donors containing the corresponding mutations, the transfer frequency was restored to drR27 levels, indicating that these gene products have a positive effect on transfer. When htdA was expressed in donors with drR27 (wild-type R27 within CAT inserted into htdA), the transfer frequency was reduced to wild-type R27 levels, indicating that htdA has a negative effect on the transfer frequency.

During the complementation experiments, the conjugation frequency varied between 144% (trhW) and 0.2% (trhV) of the transfer frequency of drR27. The variation could be due to polar transcriptional effects because of the CAT insertions and/or that the overexpressed transfer proteins reduce the conjugation frequencies. To address these possibilities, we compared the complementation frequencies for each mutant to the transfer frequency of drR27 when the corresponding transfer protein was overexpressed in donors during conjugation experiments (Table 3). Any difference in the conjugation frequency between the complementation and the overexpression experiments reflects polar effects on downstream genes. Overexpression of transfer proteins had a minimal effect on transfer frequencies, especially for trhL and trhV, which had the lowest complementation frequencies. This suggests that both mini-Tn10 and CAT insertions resulted in partially polar transfer mutations of several genes. This observation has previously been noted for the Tra1 transfer genes (26). Nevertheless, this approach allows the identification of essential transfer genes.

Hgal plaque assay of R27 transfer-deficient mutants.

Hgal is an H-pilus-specific bacteriophage that lyses E. coli harboring drR27 to form distinctive plaques, allowing the application of a simple assay for H-pili production (26). Each Tra2 mutant was tested using an Hgal plaque assay to determine which mutants were resistant or sensitive to Hgal lysis (Table 3). E. coli cells containing transfer mutants trhA, -L, -E, -K, -B, -V, -C, -S, -P, and -W were resistant to Hgal. Hgal sensitivity was restored for each of these mutants when each mutant was complemented with the wild-type gene (data not shown). These observations suggest that these transfer genes are involved in H-pilus biosynthesis. The ability of the trhZ mutant to resist Hgal is interesting since this mutant is capable of transferring at very low levels, suggesting that R27 is capable of transferring at a low frequency in the absence of H-pili. E. coli strains containing drR27 with mutations in transfer genes trhU and trhN were capable of forming plaques, although they were notably smaller and not as clear as drR27 plaques. Disruption of htdA in wild-type R27 resulted in plaques which were larger than those produced by wild-type R27; this is attributed to the increase in H-pilus production by donors (52). E. coli cells containing an insertional disruption of orf030, orf028, orf027, trhO, orf017, orf016, htdF, htdK, orf009, orf004, or trhI were all capable of forming plaques, suggesting that these genes are not essential for H-pilus biosynthesis.

DISCUSSION

R27 contains two regions which contribute to conjugative transfer functions, Tra1 and Tra2, since mutations in either region abolish transfer. Tra1 contains the origin of transfer and nine essential transfer genes encoding Mpf proteins, a coupling protein, and relaxosome proteins (26). In this study, mutational and genetic analysis demonstrated that Tra2 contains 11 Mpf genes that are essential for conjugative transfer. The 20 essential transfer genes identified within both Tra1 (26) and Tra2 (this study) probably represent the entire conjugative transfer apparatus encoded by R27, since our comparison to IncF and IncP conjugative transfer systems has identified an equivalent to each essential transfer component present in these systems (see below).

Of the 11 Mpf proteins encoded within the Tra2 region, 9 are homologous to Mpf proteins of the IncF transfer system. Of these, TrhL, -E, -K, -B, -V, -C, and -W are essential for conjugative pilus biosynthesis, the same function assigned to their F factor Mpf counterparts (15). TrhU and TrhN are not essential for H-pilus biosynthesis but do appear to play an auxiliary role in H-pilus assembly. Based on their homology and shared mutant phenotypes with TraU (36) and TraN (33) of the F factor, we propose that TrhU and TrhN also play a role in DNA transfer and mating-pair stabilization, respectively. Including the 3 Mpf genes from Tra1, R27 contains an equivalent to 12 of the 13 essential F factor Mpf proteins, with the exception of the pilin subunit, illustrating that the IncH Mpf/type IV secretion system has a common ancestry with the IncF system. Although these proteins are responsible for conjugal piliation and subsequent DNA transfer, their exact roles, collectively and individually, remain unknown. TrhC, a putative ATPase, was recently shown to form membrane-associated complexes, and complex formation was dependent on the presence of TrhB, TrhE, and TrhL (17). This suggests that these proteins form a multiprotein complex that may function in H-pilus synthesis or R27 transfer or both.

A novel observation is that TrhK and TraK-like homologs have similarity to secretins. Secretins are outer membrane pores that are found in type II and type III secretion systems and allow the passage of macromolecules across the outer membranes. Secretins form high-molecular-weight multimers (29), and preliminary results suggest that TrhK does as well (Lawley and Taylor, unpublished). Secretins consist of two domains, a nonconserved N-terminal specificity domain and a conserved C-terminal domain which defines the secretin family (19, 39). The C-terminal domain contains the β-domain, which inserts into the outer membrane, and a S-domain, which interacts with a stabilizing lipoprotein (19). Both the β- and S-domains appear to be conserved in TrhK. TraK of the F factor was recently shown to be present in the outer membrane of donor cells, and the C-terminal region interacts with TraV, a transfer lipoprotein (21). The presence of secretins in type IV secretion systems would suggest a mechanism by which DNA and pili could transverse the outer membrane of donors.

The two remaining Tra2-encoded Mpf proteins, TrhA and TrhP, are homologous to the pilin (TrbC) and peptidase (TraF), respectively, of the IncP transfer system. Although the similarity between the pilin subunits is weak, the presence of the peptidase implies a maturation process for the H-pilus which is analogous to that observed for the P-pilus. This observation points to a common ancestry for pilus subunit processing, but further work is required to demonstrate H-pilus cyclization.

Our analysis of Tra1 and Tra2 indicated that the transfer components of R27 have a common ancestry with the IncP and IncF systems (26). The relaxosome, pilin, and peptidase appear to have a common lineage with IncP plasmids, whereas the Mpf/type IV secretion system is of the IncF lineage (Fig. 4). There is no evidence of gene redundancy in any of these transfer systems. If the chimeric nature of these transfer systems did occur through DNA recombination between plasmids, the recombinant genomes subsequently underwent loss of redundant DNA, resulting in the current transfer systems. The separation of Tra1 and Tra2 by 63 kb may be a remnant of such a recombination-loss event. It has been proposed that the conjugative transfer regions of the Ti plasmid have evolved in a similar fashion (1).

FIG. 4.

Comparison of the complete F, R27, and RP4 conjugative transfer regions (some regulatory genes are excluded). Essential transfer genes are presented with color and pattern, with the same color and pattern representing homologous gene products, while nonessential transfer genes are white. The Mpf genes are solid in color, and the coupling protein, relaxosome components, and regulatory genes are patterned. Light gray genes represent gene products with no shared homology. Thin lines above the R27 map indicate homologs to the IncF system, while thick lines represent homologs to the IncP system. Lipo, lipoprotein motif; red box, Walker A motif; green box, origin of transfer; capital letters for gene names, Tra; lowercase letters for gene names, Trb (F and RP4) or Trh (R27). Double slash indicates noncontiguous regions. Arrows representing ORFs are proportional to the length of the ORFs.

Mpf systems were once viewed as highly conserved protein machines; however, with the expanding genomic databases and the careful analysis of several conjugative systems, it is becoming evident that there are key differences between the essential components of IncF/H Mpf and IncP (IncW/N) Mpf. Although the Mpf systems encode approximately the same number of Mpf proteins, there appears to be a subset of conserved core components and a subset of nonconserved proteins. Based on homology, there are core components shared between the IncF/H and IncP Mpf systems (IncF/IncP): TraL/TrbD, TraE/TrbJ, TraB/TrbI, TraC/TrbE (ATPase), and TraK/TrbG (7, 9, 15). Additional core components can also be assigned based on functional analogy: TraA/TrbC (pilin and their associated maturation proteins, peptidase or acetylase) and TraV/TrbH (lipoprotein). The nonshared components, and therefore the defining components, are a subset of essential Mpf proteins that are unique to each subfamily. The IncF/H subfamily contains (IncF nomenclature) TraW, TrbC, TraU, TraN, TraF, TraH and TraG, which are not present in the IncP subfamily. Other transfer systems which contain an equivalent to each of the essential IncF/H Mpf components include R478 (IncHI2) (41), the SXT element (4), R391 (IncJ) (5), Rts1 (IncT) (38), and pNL1 (44). The IncP subfamily contains TrbB (ATPase), TrbF, TrbG, TrbK (Eex), and TrbL, which are not present in the IncF/H subfamily (18) (Fig. 4). A careful phylogenetic analysis will be required to fully delineate the evolutionary relationship between the IncF/H and IncP Mpf systems.

The core components reveal the common ancestry between the Mpf systems, whereas the nonconserved components probably reflect divergent evolution of these subfamilies to optimize the conjugation machinery for transfer in specialized environments. Conjugative transfer systems of the IncF/H subfamily (IncF/H/T/J) are capable of transferring both in liquid and on solid surfaces with approximately equal efficiencies, and they produce thick flexible pili, whereas the transfer systems of the IncP subfamily (IncP/N/W) transfer more efficiently on solid surfaces than in liquid and produce short rigid pili (6). It is likely that the conjugative pilus type dictates the mating capabilities of each subfamily (6). Furthermore, the differences in pilus structure may be attributed to the differences in the Mpf/type IV systems. In addition, the nonconserved Mpf proteins are probably responsible for the differences in mating capabilities. For example, the TraN (24) and TraG (34) proteins of the IncF subfamily, which are not found in the IncP subfamily, have been implicated in mating-pair stabilization, a phenotype that would enhance mating in liquid by stabilizing the mating pair(s) against the shearing forces of the fluid environment.

The above observations are consistent with those made in ecology-based studies on conjugative plasmid transfer. IncF/H Mpf systems are usually present in enteric bacteria, which are likely to encounter both the mammalian digestive tract and raw sewage, environments which would contain both solid and liquid milieux and are generally nutrient rich (51). Plasmids containing IncF/H Mpf systems transfer in the human digestive tract (2) and are known to transfer in aquatic environments more efficiently than those containing IncP Mpf systems (28). In comparison, IncP transfer systems are generally associated with soil organisms, like Pseudomonas and Klebsiella, and transfer efficiently in nutrient-limited soil environments. Plasmids containing IncF/H secretion systems did not transfer under the same conditions (43). It is therefore possible that the differences in the Mpf systems could reflect the differences in the ecological niches of the transfer systems.