Functional Analysis of a Bacitracin Resistance Determinant Located on ICECp1, a Novel Tn916-Like Element from a Conjugative Plasmid in Clostridium perfringens

Abstract

Bacitracins are mixtures of structurally related cyclic polypeptides with antibiotic properties. They act by interfering with the biosynthesis of the bacterial cell wall. In this study, we analyzed an avian necrotic enteritis strain of Clostridium perfringens that was resistant to bacitracin and produced NetB toxin. We identified a bacitracin resistance locus that resembled a bacitracin resistance determinant from Enterococcus faecalis. It contained the structural genes bcrABD and a putative regulatory gene, bcrR. Mutagenesis studies provided evidence that both bcrA and bcrB are essential for bacitracin resistance, and that evidence was supported by the results of experiments in which the introduction of both the bcrA and bcrB genes into a bacitracin-susceptible C. perfringens strain was required to confer bacitracin resistance. The wild-type strain was shown to contain at least three large, putatively conjugative plasmids, and the bcrRABD locus was localized to an 89.7-kb plasmid, pJIR4150. This plasmid was experimentally shown to be conjugative and was sequenced. The sequence revealed that it also carries a tpeL toxin gene and is related to the pCW3 family of conjugative antibiotic resistance and toxin plasmids from C. perfringens. The bcr genes were located on a genetic element, ICECp1, which is related to the Tn916 family of integrative conjugative elements (ICEs). ICECp1 appears to be the first Tn916-like element shown to confer bacitracin resistance. In summary, we identified in a toxin-producing C. perfringens strain a novel mobile bacitracin resistance element which was experimentally shown to be essential for bacitracin resistance and is carried by a putative ICE located on a conjugative plasmid.

INTRODUCTION

Necrotic enteritis is a disease of increasing importance to the worldwide poultry industry, one of the major suppliers of protein for human consumption. The essential causative agent of this disease is the Gram-positive spore-forming bacterium Clostridium perfringens (1). Bacitracin (Bac) is one of the antimicrobial agents used to control necrotic enteritis (2, 3). Although its use as a growth promoter in animal feeds has been banned in the European Union since 1999, bacitracin is still used as a feed additive in some countries, including the United States (4). The extensive use of bacitracin appears to have led to the isolation of bacitracin-resistant C. perfringens strains (5, 6). However, little is known about the molecular basis of bacitracin resistance in this organism.

Bacitracin is a potent narrow-spectrum antibiotic that is active against Gram-positive bacteria. It is a mixture of high-molecular-weight cyclic polypeptides produced by members of the genus Bacillus and requires a divalent metal ion for its biological activity (7). Bacitracin forms a complex with undecaprenol pyrophosphate (UPP), the lipid carrier required for the translocation of cell envelope precursors across the membrane. By binding to UPP, bacitracin prevents its dephosphorylation, a process that is essential for recycling of the lipid carrier, thereby disrupting cell wall polymer biosynthesis (7, 8).

Several bacitracin resistance mechanisms have been reported (7, 9). In the bacitracin-producing organism Bacillus licheniformis, three genes (bcrABC) encoding a putative heterodimeric ATP-binding cassette (ABC) transporter have been proposed to mediate the efflux of bacitracin (10). Similar systems have been described in Bacillus subtilis (11), Streptococcus mutans (12), Enterococcus faecalis (9, 13, 14), and C. perfringens (5). A second mechanism involves the overproduction of undecaprenol kinase, which converts UPP to undecaprenol phosphate and overcomes the blocking of UPP recycling by bacitracin. This mechanism has been found in Escherichia coli (15) and in Streptococcus pneumoniae and Staphylococcus aureus (16). Other mechanisms of bacitracin resistance include the production of a membrane-bound phospholipid phosphatase in B. subtilis (17) and exopolysaccharide production in Xanthomonas campestris, Sphingomonas spp., E. coli, and S. mutans (12, 18).

In the bacitracin ABC transporter system from E. faecalis, the bcrA gene encodes an ATP-binding protein and bcrB encodes a membrane-bound permease. These genes, together with bcrD, form an operon that is induced by the transcriptional regulator BcrR (9, 13). The bcrD gene encodes an undecaprenol kinase, which is not required for bacitracin resistance in E. faecalis (9). BcrR, encoded by the gene immediately upstream of bcrABD, contains an N-terminal helix-turn-helix DNA-binding domain and C-terminal membrane-spanning α helices. It has been proposed that the membrane localization of BcrR is required for its function and that BcrR is always bound to its target DNA, two inverted repeats located upstream of P_bcrA. Upon sensing and binding to bacitracin, the conformation of this membrane-DNA-bound dimeric BcrR complex changes, which consequently stimulates the transcription of the bcrABD operon (13). In E. faecalis, the bcrRABD genes are located on an IS1216-like mobile element. In a bacitracin-resistant C. perfringens strain, c1261_A, isolated from a turkey (5), orthologous but seemingly truncated bcr genes were identified on a similar IS1216-like element on the chromosome.

In this study, we report the identification of a novel Tn916-like integrative conjugative element (ICE), ICECp1, which was located on a large conjugative plasmid in a necrotic enteritis-causing strain of C. perfringens. This putative ICE contained a bcrRABD bacitracin resistance determinant. Mutagenesis studies showed that both the bcrA and bcrB genes are essential for bacitracin resistance in C. perfringens.

MATERIALS AND METHODS

Strains, plasmids, and growth conditions.

The bacterial strains and plasmids used in this study are listed in Table 1. The E. coli cells used for plasmid propagation and cloning experiments were grown at 37°C on 2× YT medium supplemented with appropriate antibiotics (19). The C. perfringens strains were cultured at 37°C in heart infusion (HI) broth or agar (Oxoid), Trypticase-peptone-glucose (TPG) broth (20), fluid thioglycolate medium (FTG; Difco), or nutrient agar (21) in an atmosphere of 10% (vol/vol) H₂, 10% (vol/vol) CO₂, and 80% (vol/vol) N₂. When required, the media were supplemented with the following antibiotics at the indicated concentrations: for E. coli, chloramphenicol (Cm; 30 μg/ml) or erythromycin (Erm; 150 μg/ml), and for C. perfringens, bacitracin (Bac; 2 IU/ml), thiamphenicol (Tm; 10 μg/ml), tetracycline (Tc; 10 μg/ml), Erm (7 μg/ml or 50 μg/ml), rifampin (Rif; 10 μg/ml), nalidixic acid (Nal; 10 μg/ml), streptomycin (Str; 1 mg/ml), and/or saturated potassium chlorate (Chl; 1%, vol/vol).

TABLE 1.

Bacterial strains and plasmids

Strain or plasmid	Characteristics	Source or reference
Strains
E. coli
DH5α	F− endA1 hsdR17(rK− mK−) thi-1 λ− recA1 gyrA96 relA1 rhoA supE44 deoR ϕ80dlacZΔM15 Δ(lacZYA argF)U169	Invitrogen
TOP10	F− mcrA Δ(mrr-hsdRMS-mcrBC) ϕ80dlacZΔM15 ΔlacX74 nupG recA1 araD139 Δ(ara-leu)7697 galU galK rpsL (Strr) endA1 λ−	Invitrogen
C. perfringens
CW504	CW362 Rifr Nalr	30
EHE-NE18	Australian chicken necrotic enteritis isolate	45
JGS4102	U.S. necrotic enteritis isolate	J. Glenn Songer (Iowa State University)
JIR39	CW362 Chlr Strr	21
JIR325	Strain 13 Rifr Nalr	46
JIR4394	Strain 13 Strr Chlr	32
JIR12635	CW504-derived transconjugant, Bacr Tcr	JGS4102 × CW504
JIR12637	CW504-derived transconjugant, Bacr Tcr	JGS4102 × CW504
JIR12642	JIR39-derived transconjugant, Bacr Tcs	JIR12635 × JIR39
JIR12703	JIR39-derived transconjugant, Bacr Tcs	JIR12637 × JIR39
JIR12708	CW504(pJIR4150) Bacr Tcs	JIR12703 × CW504
JIR12710	JIR325(pJIR4277) Bacr Tcs	Transformant
JIR12855	JIR12710 bcrA ΩTTa ermB RAM	Transformation with pJIR4106
JIR12858	JIR12710 bcrB ΩTT ermB RAM	Transformation with pJIR4107
JIR12930	JIR12855(pJIR4137) bcrA/bcrA+	Transformation with pJIR4137
JIR12931	JIR12858(pJIR4138) bcrB/bcrB+	Transformation with pJIR4138
JIR12932	JIR325(pJIR4139) (bcrAB+)	Transformation with pJIR4139
Plasmids
pJIR3422	E. coli-C. perfringens shuttle vector, catP+ lacZ α peptide	29
pJIR3562	pJIR3756 (MluI) harboring Pfdb ermB RAM (MluI; 1.23 kb)	28
pJIR3566	Clostridial TT derived from pJIR750ai, contains ermB RAM and lacZα, Cmr	28
pJIR4106	pJIR3566 retargeted to nucleotide 24/25 of bcrA	Recombinant, this study
pJIR4107	pJIR3566 retargeted to nucleotide 96/97 of bcrB (antisense strand)	Recombinant, this study
pJIR4137	pJIR3422 BamHI/KpnI Ω1.09-kb PCR product of bcrA	Recombinant, this study
pJIR4138	pJIR3422 BamHI/KpnI Ω0.82-kb PCR product of bcrB	Recombinant, this study
pJIR4139	pJIR3422 BamHI/KpnI Ω1.83-kb PCR product of bcrAB	Recombinant, this study
pJIR4140	pJIR3422 BamHI/KpnI Ω2.55-kb PCR product of bcrRAB	Recombinant, this study
pJIR4150	Conjugative 89.7-kb bacitracin resistance plasmid from JIR12708	Conjugation, this study
pJIR4277	40-kb bacitracin resistance plasmid from JIR12710	Transformation, this study

^aTT, TargeTron.

^bP_fd, Clostridium pasteurianum ferredoxin promoter.

DNA manipulation and molecular techniques.

Unless otherwise stated, standard procedures were used for DNA manipulation and molecular techniques (19). Competent E. coli (22) and C. perfringens (23) cells were prepared as described previously. DNA concentrations were quantified using a NanoDrop spectrophotometer (NanoDrop Technologies). The oligonucleotides used (Table 2) were synthesized by Sigma-Aldrich. Plasmid DNA from E. coli was extracted using a QIAprep miniprep kit (Qiagen) according to the manufacturer's instructions. Restriction endonucleases and other enzymes were used as specified by the manufacturers (Roche Diagnostics or New England BioLabs). Genomic DNA was extracted from C. perfringens cells grown in FTG or HI broth, as previously reported for Clostridium difficile (24). Plasmid DNA from C. perfringens was isolated and subjected to electrophoresis on a 0.8% MegaBase gel (Sigma) at 80 V for 4 h and visualized by use of a bioimaging system (Syngene) after being stained with GelRed (Biotium). Southern hybridizations were performed as described previously (25) with digoxigenin (DIG)-labeled PCR-amplified probes specific for the relevant antibiotic resistance genes or the target genes. DNA sequencing was performed by Micromon at Monash University using an Applied Biosystems 3730S genetic analyzer. Sequence data were compiled using the Sequencher (version 4.10.1) program (Gene Codes Corporation).

TABLE 2.

Oligonucleotides

Primer	Nucleotide sequence 5′-3′	Use or location
JRP3867	CGAAATTAGAAACTTGCGTTCAGTAAAC	TargeTron universal primer
JRP3939	CTCAGTACTGAGAGGGAACTTAGATGGTAT	catP PCR
JRP3940	CCGGGATCCTTAGGGTAACAAAAAACACC	catP PCR
JRP4555	GTTTACTTTGGCGTGTTTCATTGC	ermB PCR
JRP4632	AATAAGTAAACAGGTAACGTCT	ermB PCR
JRP4978	TAAAACAAGAATACCAAGCACAA	Amplification of bcrRABD and bcrD
JRP4981	TTTAATGAAAAGCTACAACAGCT	Amplification of bcrRABD and bcrR
JRP5335	CGAGCTCTTGATAAAATACCGAGTGCAGAGG	Cloning of bcrRABD upstream of bcrR
JRP5336	GGGGTACCCCCGATTTTTGCCTGCATATC	Cloning of bcrRABD downstream of bcrD
JRP5337	TGAACGCAAGTTTCTAATTTCGGTTTCAATCCGATAGAGGAAAGTGTCT	EBS2 primer to retarget intron to bcrA
JRP5338	AAAAAAGCTTATAATTATCCTTAATTGACACAGAAGTGCGCCCAGATAGGGTG	IBS primer to retarget intron to bcrA
JRP5339	CAGATTGTACAAATGTGGTGATAACAGATAAGTCACAGAAAATAACTTACCTTTCTTTGT	EBS1 primer to retarget intron to bcrA
JRP5340	TGAACGCAAGTTTCTAATTTCGATTATCTCTCGATAGAGGAAAGTGTCT	EBS2 primer to retarget intron to bcrB
JRP5341	AAAAAAGCTTATAATTATCCTTAGAGATCATCTTTGTGCGCCCAGATAGGGTG	IBS primer to retarget intron to bcrB
JRP5342	CAGATTGTACAAATGTGGTGATAACAGATAAGTCATCTTTACTAACTTACCTTTCTTTGT	EBS1 primer to retarget intron to bcrB
JRP5394	CGCGGATCCCGGAACTGAAAAATTTGGTCGT	Complementation upstream of bcrR
JRP5395	CGCGGATCCAACCCTTTAAAATAGGCTCTGAC	Complementation and probe upstream of bcrA
JRP5396	GGGGTACCTTATTTCAAATCCTCCTTTTTAAA	Complementation and probe for bcrB
JRP5403	GGGGTACCTTAAGCAATACCGCCACCTCCA	Complementation and probe for bcrA
JRP5404	CGCGGATCCTCGTGACAAGGATAAATGATAG	Complementation and probe upstream of bcrB

Conjugation experiments.

All matings were performed as described previously (21, 26). The transconjugants were selected on nutrient agar supplemented with the appropriate antibiotics, and the conjugation frequency was calculated as the number of transconjugants per donor cell.

Bacitracin susceptibility tests.

Disc diffusion assays (Oxoid Limited) were initially used to screen for bacitracin-resistant strains. Bacitracin MICs were determined using the Etest system (AB Biodisk) per the manufacturer's instructions. C. perfringens cells grown on nutrient agar supplemented with the appropriate antibiotics were collected and diluted to an optical density at 600 nm equivalent to a 0.5 McFarland standard, and then 100 μl of the diluted culture was spread onto nonselective HI agar to yield confluent growth. Once the culture was dry, a bacitracin Etest gradient strip was placed onto the surface. The agar plates were then incubated anaerobically at 37°C for 24 h.

Sequencing and analysis.

High-throughput sequencing data for whole-genome DNA were generated by Micromon using an Illumina GAIIX genome analyzer. The resultant short read data were assembled de novo using the Velvet program (27). PCR amplification and Sanger sequencing were used to validate the assemblies.

Construction of bcrA and bcrB mutants and complementation.

bcrA and bcrB mutants were constructed as previously described (28) using derivatives of the TargeTron mutagenesis vector pJIR3566. To identify potential TargeTron insertion sites, the nucleotide sequences of bcrA and bcrB were submitted to an intron site finder (G. Carter and T. Seemann, Monash University). For the bcrA gene, the insertion site on the sense strand at a position 24/25 bp from the ATG start codon was selected from the predicted sites for TargeTron modification. For bcrB, a position 96/97 bp from the ATG start codon on the sense strand was selected. To retarget the group II intron, primer-mediated mutation by splicing by overhang extension PCR was carried out with the IBS, EBS2, EBS1, and EBS universal primers (Table 2) in accordance with the instructions of the Sigma-Aldrich TargeTron gene knockout system. The 350-bp retargeted PCR product was then digested with HindIII and BsrGI and ligated into the same sites of pJIR3566. The ligation mixture was used to transform E. coli TOP10 competent cells (Invitrogen), and plasmid DNA from the resultant transformants was purified and sequenced. The resultant plasmids, pJIR4106 and pJIR4107 (Table 1), were used for mutagenesis of bcrA and bcrB, respectively.

Introduction of the TargeTron plasmids, screening, and confirmation of mutants were carried out as previously described (28). Briefly, to demonstrate that the intron had inserted into the target gene (bcrA or bcrB), genomic DNA was purified and analyzed by PCR using the primer pairs JRP5395/JRP5403 (bcrA) and JRP5396/JRP5404 (bcrB), which flank the insertion sites. PCRs specific for ermB (with primer pair JRP4632/JRP4555) and catP (with primer pair JRP3939/JRP3940) were also performed to confirm that the ermB retrotransposition-activated marker (RAM) within the intron was inserted into the target gene and that the TargeTron vector was cured from the mutants, respectively. To confirm the PCR results, Southern hybridizations using probes specific for the target gene (bcrA or bcrB), ermB, and catP were carried out. Genomic DNA was digested with NdeI, separated by electrophoresis, transferred to a nylon membrane (GE Healthcare), and hybridized with each of the appropriate probes.

A bcrA complementation vector, pJIR4137, was constructed by cloning a 1.09-kb PCR product containing bcrA into the BamHI/KpnI sites of the shuttle vector pJIR3422 (Table 1), in which transcription of cloned genes is under TnpX control via the circular intermediate (CI) promoter (P_CI) of Tn4451 (29). To construct a bcrB complementation vector (pJIR4138), a 0.82-kb PCR product containing bcrB was cloned into the same sites of pJIR3422. In complementation experiments, pJIR4137 and pJIR4138 were introduced into JIR12855 and JIR12858, respectively, by electroporation.

To determine the minimal region required to confer bacitracin resistance in a bacitracin-susceptible strain, JIR325 (Table 1), PCR products containing bcrAB and bcrRAB were cloned into the BamHI/KpnI sites of pJIR3422, and the resultant constructs, pJIR4139 (bcrAB) and pJIR4140 (bcrRAB), respectively, together with pJIR4137 (bcrA) and pJIR4138 (bcrB), were individually introduced into JIR325 by electroporation. To ensure the integrity of the complementation plasmids in C. perfringens, each of the plasmids was isolated from the C. perfringens host, introduced back into E. coli, and then purified and sequenced.

RESULTS

The bcrRABD locus was identified in JGS4102, a bacitracin-resistant avian isolate of C. perfringens.

A bacitracin- and tetracycline-resistant strain, JGS4102, was identified from 21 avian isolates of C. perfringens examined using disc diffusion assays. JGS4102 had a bacitracin MIC of >256 μg/ml. To determine the genes responsible for bacitracin resistance, primers specific for bcrR and bcrD from E. faecalis (GenBank accession number AY496968) (Table 2) were used to amplify genomic DNA from JGS4102. A 3.2-kb amplicon was obtained and shown by sequence analysis to contain a bcrRABD gene locus that had 85% nucleotide sequence identity to its counterpart from E. faecalis. The genes were tandemly arranged, with 180 nucleotides between bcrR and bcrA, 40 nucleotides between bcrA and bcrB, and a 1-nucleotide overlap between bcrB and bcrD being detected. The arrangement of the bcrABD genes suggested that they form an operon. The genetic organization of these genes was also very similar to that observed in E. faecalis (9) and in a previously studied C. perfringens strain, c1261_A (5), although several of the genes in the latter strain appeared to be either incomplete or truncated (Fig. 1). The sizes of the deduced JGS4102-derived gene products were close to the sizes of their orthologues from E. faecalis (9) but larger than their orthologues from C. perfringens strain c1261_A (5) (Fig. 1). The deduced amino acid sequences of BcrR, BcrA, BcrB, and BcrD showed 77%, 89%, 92%, and 90% identity, respectively, to their E. faecalis orthologs.

FIG 1.

Comparative genetic organization of the bcrRABD locus from JGS4102. Orthologous genes are highlighted with the same pattern. The nucleotide sequence identity between the orthologues of E. faecalis (AY496968), C. perfringens c1261_A (5), and C. perfringens JGS4102 is shown. The sizes of the genes and intergenic regions are shown to scale. For C. perfringens c1261_A, the vertical arrowheads represent sequences that are not available from the database.

Both bcrA and bcrB are essential for bacitracin resistance.

No functional studies on the bcrRABD locus from C. perfringens had been carried out. To determine if this locus is responsible for bacitracin resistance, we individually mutated both bcrA and bcrB by TargeTron-based insertional mutagenesis. The strain used for these studies was bacitracin-resistant strain JIR12710, which was derived from the transformation of bacitracin-susceptible strain JIR325 (MIC = 4 μg/ml) with plasmid DNA from a bacitracin-resistant JIR39-derived transconjugant (JIR12642; Table 1). The genotypes of the resultant bcrA (JIR12855) and bcrB (JIR12858) mutants were confirmed by PCR analysis and Southern hybridization (data not shown). Bacitracin susceptibility tests showed that the mutation of either bcrA (MIC = 16 μg/ml) or bcrB (MIC = 12 μg/ml) led to a dramatic reduction of resistance to bacitracin compared to that of parent strain JIR12710 (MIC > 256 μg/ml). To confirm that the susceptibility to bacitracin of the bcrA mutant was due to the inactivation of the bcrA gene instead of a polar effect or a secondary mutation, a bcrA complementation vector, pJIR4137 (Table 1), was constructed by cloning the wild-type bcrA gene into a shuttle vector, pJIR3422, so that expression of bcrA was under the control of the TnpX circular intermediate (CI) promoter and therefore was tightly regulated in E. coli (Table 1) (29). The resultant plasmid, pJIR4137, was subsequently introduced into the bcrA mutant JIR12855. The resultant bcrA/bcrA⁺ strain, JIR12930, had a significantly increased resistance to bacitracin (MIC = 128 μg/ml). Similarly, complementation of the bcrB mutant JIR12858 with a bcrB⁺ complementation vector, pJIR4138 (Table 1), yielded a bcrB/bcrB⁺ strain, JIR12931, that had wild-type resistance to bacitracin (MIC > 256 μg/ml). We also introduced the bcrB complementation vector into the bcrA mutant and the bcrA complementation vector into the bcrB mutant. Both of the resultant transformants were still susceptible to bacitracin, as expected.

bcrAB alone can confer bacitracin resistance.

To determine the minimal bcrRABD region that is required for bacitracin resistance, we cloned each of the bcrA, bcrB, bcrAB, and bcrRAB gene regions into pJIR3422 and then examined the ability of the resultant constructs to confer bacitracin resistance in C. perfringens. We excluded the bcrD gene, as data from E. faecalis indicated that bcrD is not essential for bacitracin resistance (9). JIR325-derived transformants were selected on thiamphenicol, resistance to which is encoded by the catP gene located on pJIR3422, and patched onto medium that contained bacitracin. The results showed that none of the pJIR4137 (bcrA⁺) or pJIR4138 (bcrB⁺) transformants were able to confer bacitracin resistance, whereas all of the JIR325 transformants derived from pJIR4139 (bcrA⁺ bcrB⁺) were bacitracin resistant (MIC = 128 μg/ml). On the basis of these results, it was concluded that both the bcrA and bcrB genes are required to confer bacitracin resistance in C. perfringens. We also attempted to transform JIR325 with the bcrR⁺ bcrA⁺ bcrB⁺ construct pJIR4140, but only low numbers of transformants were obtained. When the plasmids in these transformants were subsequently transferred into E. coli and analyzed, it was noted that the bcrR gene had been deleted and replaced by an IS1 element. It was concluded that pJIR4140 is unstable in C. perfringens.

Bacitracin resistance is transferable.

In addition to the bacitracin resistance locus, PCR analysis revealed that JGS4102 carries several toxin genes, specifically, the netB, cpb2, and tpeL genes, as well as the tetA(P) tetracycline resistance gene. Since previous studies (6, 26) demonstrated that the netB, cpb2, and tetA(P) genes are located on different conjugative plasmids in avian isolates of C. perfringens, we postulated that JGS4102 may carry more than one toxin or antibiotic resistance plasmid and asked whether bacitracin resistance is plasmid determined in this strain.

We carried out a series of conjugation experiments using the nontransformable C. perfringens strain CW504, which is a spontaneous Rif^r Nal^r mutant of CW362 and does not carry any plasmids (30), as the recipient in primary mixed plate mating experiments. From four independent mating experiments, we obtained 19 bacitracin-resistant transconjugants, 18 of which were also resistant to tetracycline. Only one transconjugant was Bac^r Tc^s. PCR analysis of several independently derived transconjugants confirmed that they carried the bcrRABD locus. These results suggest that the bacitracin resistance determinant is located on a transferable element.

To determine if the bcrRABD locus is located on a conjugative plasmid, secondary matings were undertaken to examine whether the primary transconjugants could act as donors in subsequent transfer experiments. Since intrastrain matings give more consistent conjugation frequencies (26), several independent CW504-derived bacitracin- and tetracycline-resistant transconjugants from the primary matings were used as donors in secondary matings using strain JIR39, an isogenic Sm^r Chl^r spontaneous mutant of CW362, as the recipient. The results showed that bacitracin resistance was transferred at a variable frequencies ranging from 4.3 × 10⁻⁶ to 2.9 × 10⁻¹ transconjugants per donor cell, depending on the donor. In donors that consistently yielded bacitracin-resistant transconjugants at a high transfer frequency, including JIR12637, we postulated that the bacitracin resistance determinant was located on a conjugative plasmid. We therefore decided to purify the putative plasmid and to determine its nucleotide sequence. To ensure that we obtained a strain with a single plasmid, we decided to transform JIR325, a highly transformable C. perfringens strain (Table 1), to bacitracin resistance. Plasmid DNA was purified from two JIR39-derived bacitracin-resistant, tetracycline-susceptible transconjugants (JIR12703 and JIR12642) that were derived from high- and low-transfer-frequency conjugation events, respectively, and used in an attempt to transform JIR325 to bacitracin resistance. However, bacitracin-resistant transformants were obtained only from the low-transfer-frequency transconjugant, probably reflecting the fact that the transformation of very large plasmids into C. perfringens is a very inefficient process. One of these transformants, JIR12710, was used for further analysis. We subsequently performed a tertiary mating, using the JIR39-derived high-transfer-frequency transconjugant JIR12703 as the donor and CW504 as the recipient, to obtain a bacitracin-resistant transconjugant that contained a single plasmid. As expected, the transfer of bacitracin resistance was detected at a high frequency (3.7 × 10⁻¹ transconjugants per donor cell). One of these transconjugants, JIR12708, was used for further analysis.

The bacitracin MICs of JIR12708 and JIR12710 were identical to the bacitracin MIC of the wild-type strain JGS4102 (>256 μg/ml), and PCR analysis of the bcrRABD region showed that both strains carried the entire locus. The transfer frequencies of JIR12708 and JIR12710 were consistent in subsequent conjugation experiments using JIR39 and a Sm^r Chl^r strain 13 derivative, JIR4394, respectively, as recipients, with a high transfer frequency of 11.4 ± 2.1 (JIR12708) and a low transfer frequency of 1.1 ×10⁻⁵ ± 5.3 × 10⁻⁶ (JIR12710) being observed, as expected.

The bcrRABD locus is located on a conjugative plasmid.

To determine the genetic location of the bcrRABD region, we purified plasmid DNA from JGS4102, JIR12708, and JIR12710. The resultant plasmids were separated by electrophoresis on 0.8% MegaBase gels and analyzed by Southern hybridization using bcrB- and tcpCF-specific probes. The tcpC and tcpF genes are located in the conjugation locus of all conjugative plasmids in C. perfringens (31) and are essential for efficient conjugative transfer (32, 33). The region from tcpC to tcpF can therefore be utilized as a conjugation-specific genetic marker. The results showed that there were at least three large plasmids in JGS4102 (Fig. 2). Note that we could not exclude the possibility that there were other plasmids of similar size that were not separated under these gel electrophoresis conditions. By comparison with the three sequenced conjugative plasmids (82 kb, 71 kb, and 49 kb, respectively) from the C. perfringens necrotic enteritis strain EHE-NE18 (26) (Fig. 2), we concluded that the three plasmid bands in JGS4102 were >82 kb, ∼75 kb, and ∼55 kb in size, respectively. In contrast, JIR12708 had a single >82-kb band that was of the same size as the largest plasmid from JGS4102, while JIR12710 contained a plasmid of ∼40 kb (Fig. 2), which was different from any of the plasmids found in its parental strain, JGS4102, and an ∼54-kb plasmid of lower copy number, which was presumably the native pCP13 plasmid of JIR325.

FIG 2.

Plasmid analysis by gel electrophoresis and Southern hybridization. Plasmid DNA extracted from strains EHE-NE18, JGS4102, JIR12708, and JIR12710 was separated by electrophoresis on 0.8% MegaBase gels and visualized after staining with GelRed (A) and then hybridized with a DIG-labeled probe specific for bcrB or tcpCF (B). Arrows, the known sizes of the three large plasmids from strain EHE-NE18 (26). Other size standards represent those for HindIII-digested λcI857 DNA.

Southern hybridization analysis showed that the single plasmid in JIR12708 and the largest plasmid in JGS4102 hybridized with a bcrB-specific probe and the tcpCF-specific probe, providing evidence that the bcrRABD locus is located on a conjugative plasmid that is >82 kb in size. This plasmid was designated pJIR4150. In addition, it was found that the other two plasmids in JGS4102 also hybridized with tcpCF, indicating that they are probably conjugative. Finally, in JIR12710 the bcrRABD locus was present on a 40-kb plasmid (pJIR4277) that did not carry the tcpCF locus and presumably was not conjugative.

The gene for bacitracin resistance is carried by a novel Tn916-like element, ICECp1, that may also be conjugative.

The observations that bacitracin resistance could be transferred only at a low frequency from JIR12710 and that pJIR4277 did not carry the tcp conjugation locus suggested that bcrRABD may be located on an ICE that has the ability to insert into both conjugative and nonconjugative plasmids or into the chromosome. To examine this hypothesis, we decided to determine the nucleotide sequence of pJIR4150 from JIR12708. We used this strain as it contained a single bacitracin resistance plasmid of the same size as the equivalent plasmid from JGS4102, which therefore guaranteed the simplest process for assembling the sequence data. Previous studies have shown that the assembly of plasmid sequences from C. perfringens strains that contain more than one closely related conjugative plasmid is a very difficult process (26). Genomic DNA from JIR12708 was purified and sequenced on an Illumina GAIIx instrument using 36-bp paired-end chemistry, yielding 413 Mbp of reads. The plasmid sequence reads were obtained by subtracting the chromosomal sequence from the entire genomic sequence short reads using an unpublished CW504 genome sequence. The plasmid sequence was then assembled using the sequence of the prototype plasmid pCW3 (32) as a scaffold. Primer walking and conventional Sanger sequencing were used to fill several sequence gaps and to validate the assembled data. The result was the determination of the complete sequence of the 89,692-bp circular plasmid pJIR4150 (Fig. 3).

FIG 3.

Comparative alignment of pJIR4150 and pCW3 sequences. The genetic organizations of pJIR4150 (GenBank accession number LN835295) and pCW3 (GenBank accession number DQ366035) are compared. Each arrow represents an ORF. The color codes are as follows: red, conserved tcp locus; dark blue, other conserved ORFs shared by the large conjugative C. perfringens plasmids; yellow, plasmid replication region; pink, tetracycline resistance gene region; light blue, regions unique to pJIR4150.The location of the ICECp1 insertion is shown by the large white arrow.

Detailed analysis of pJIR4150 led to the identification of a novel, potentially conjugative ca. 30-kb element, which we have designated ICECp1 (Fig. 4) and which contained the bcrRABD gene region. ICECp1 contained orthologues of most of the conserved open reading frames (ORFs) found in the tetracycline resistance element Tn916 from E. faecalis and other ICEs in this family (Table 3; Fig. 4); these orthologues included ORFs 13 to 23, ORFs 7 to 9, xis, and int. ICECp1 also carried two genes encoding putative transposases that are not encoded by genes on Tn916 and several ORFs encoding hypothetical proteins and a putative collagen adhesin-like protein. The bcrRABD region was found to be inserted between ORF19 and ORF18 (Table 3; Fig. 4). We compared the sequence of the adjacent region with the corresponding area of the bcrRABD locus from E. faecalis (GenBank accession number AY496968). Apart from the sequence of bcrRABD and its upstream 481 nucleotides, which contained a putative gene encoding a conserved hypothetical protein, there was no other region of sequence identity. It was reported that, like the bcrRABD locus from E. faecalis, the bcrRABD operon in C. perfringens strain c1261_A is flanked by IS1216-like sequences (5). However, no IS1216-like sequences were present on ICECp1. Interestingly, we found that in an unpublished E. faecalis ATCC 35308 genome (GenBank accession number ASDE01000007) there was a 30-kb region with 41 ORFs that had a high level of sequence similarity to ICECp1 (Fig. 4), some of which had amino acid sequences identical to those of their counterparts from ICECp1. These putative proteins included Wmk_01073 to Wmk_01076, the orthologues of BcrRABD (Fig. 4). Moreover, some of the non-Tn916 ORFs from ICECp1 also had orthologues in another ICE, Tn5386.

FIG 4.

Comparative genetic organization of ICECp1. The genetic organization of ICECp1 is compared to that of Tn916 (GenBank accession number U09422), Tn5397 (GenBank accession number AF333235), Tn5801 (GenBank accession numbers BA000017 and NC002758) (44), Tn5386 (GenBank accession number DQ321786), and an unknown element from E. faecalis (GenBank accession number ASDE01000007). The arrows represent the ORFs. The color coding for the functional module is as follows: orange, conjugation; purple, transcriptional regulation; green, recombination (insertion and excision); red, tetracycline resistance gene; dark red, lantibiotic resistance locus; pink, bacitracin resistance locus; blue, genes not present in Tn916; light blue, genes present in ICECp1 and their orthologues in Tn5386 and the unknown element; asterisks, ORFs whose products have amino acid sequences identical to those of the products from ICECp1. The ORFs and intergenic regions are shown to scale.

TABLE 3.

ORFs identified in pJIR4150

Locus_tag	Gene or locus name	Coding sequence positions	Features
pJIR4150_001		1–471	Conserved hypothetical protein, homologue of pCW3_0008
pJIR4150_002	regB	540–1070	Putative LexA family transcriptional regulator, homologue of pCW3_0009
pJIR4150_003		1356–1583	Conserved hypothetical protein, homologue of pCW3_0010
pJIR4159_004		1649–2146	Conserved hypothetical protein, homologue of pCW3_0011
pJIR4150_005	parRD	2206–2499, Ca	Plasmid segregation protein ParRD, homologue of pCW3_0012
pJIR4150_006	parMD	2502–3380, C	Plasmid segregation protein ParMD, homologue of pCW3_0013
pJIR4150_007	rep	3934–4764	Plasmid replication protein, homologue of pCW3_0014
pJIR4150_008	regC	5032–6087, C	Putative LexA family transcriptional regulator, homologue of pCW3_0015
pJIR4150_009	regD	6257–6490	Putative DNA-binding transcriptional repressor, homologue of pCW3_0016
pJIR4150_010		6586–6960	Conserved hypothetical protein, homologue of pCW3_0017
pJIR4150_011		7021–8067	Putative SCP-like extracellular protein, homologue of pCW3_0018
pJIR4150_012	pemK	8138–8512	Putative PemK-like RNase toxin, homologue of pCW3_0019
pJIR4150_013	cna	8651–11566	Putative collagen adhesin precursor, homologue of pCW3_0020
pJIR4150_014		11637–11765	Conserved hypothetical protein, homologue of pCW3_0021
pJIR4150_015		11781–12023	Conserved hypothetical protein, homologue of pCW3_0022
pJIR4150_016		12035–12406	Conserved hypothetical protein
pJIR4150_017	srtA	12422–13072	Putative sortase family protein, homologue of pCW3_0023
pJIR4150_018		13085–13447	Conserved hypothetical protein
pJIR4150_019		13492–13794	Conserved hypothetical protein, homologue of pCW3_0024
pJIR4150_020		13798–14016	Conserved hypothetical protein, homologue of pCW3_0025
pJIR4150_021	dam	14080–14844	Putative DNA adenine methylase, homologue of pCW3_0026
pJIR4150_022		14852–15154	Conserved hypothetical protein, homologue of pCW3_0027
pJIR4150_023		15155–15463	Conserved hypothetical protein, homologue of pCW3_0028
pJIR4150_024	intP	15856–16656	Putative relaxase function, tyrosine site-specific recombinase-like protein, homologue of pCW3_0029
pJIR4150_025	tcpA	16711–18330	DNA translocase/coupling protein, putative hexameric ATPase, homologue of pCW3_0030
pJIR4150_026	tcpB	18377–19372	Variant of DNA translocase/coupling protein, homologue of pCW3_0031
pJIR4150_027		19356–19526	Conserved hypothetical protein
pJIR4150_028	tcpC	19519–20595	VirB8-like conjugation protein, homologue of pCW3_0032
pJIR4150_029	tcpD	20607–20954	Transmembrane conjugation protein, homologue of pCW3_0033
pJIR4150_030	tcpE	20967–21335	Transmembrane conjugation protein, homologue of pCW3_0034
pJIR4150_031	tcpF	21393–23918	ATPase conjugation protein, homologue of pCW3_0035
pJIR4150_032	tcpG	23919–24815	Cell wall-associated peptidoglycan hydrolase conjugation protein, homologue of pCW3_0036
pJIR4150_033	ltrA	25370–27172	Putative LtrA-like reverse transcriptase/maturase
pJIR4150_034	tcpH	27501–29999	VirB6-like transmembrane conjugation protein, homologue of pCW3_0037
pJIR4150_035	tcpI	30002–30496	Putative membrane-bound metal-dependent hydrolase, homologue of pCW3_0038
pJIR4150_036	tcpJ	30505–31245	Conserved hypothetical protein, homologue of pCW3_0039
pJIR4150_037		31313–31585	Putative transposase
pJIR4150_038		31639–32460	Putative transposase
pJIR4150_039		33504–33809	Conserved hypothetical protein, homologue of pCW3_0040
pJIR4150_040		33824–34843	Methyltransferase pseudogene, homologue of pCW3_0041
pJIR4150_041		34860–35033	Conserved hypothetical protein, homologue of pCW3_0042
pJIR4150_042		35030–36052	Conserved hypothetical protein, homologue of pCW3_0043
pJIR4150_043		36147–37553	Putative amido ligase, homologue of pCW3_0044
pJIR4150_044	tpeL	37827–43166, C	Monoglycosyltransferase toxin
pJIR4150_045	tpeE	43377–43574, C	UviB-like holin
pJIR4150_046	tpeR	43618–44184, C	UviA-like alternative sigma factor
pJIR4150_047		44378–45406	Putative autolysin/peptidoglycan aminohydrolase
pJIR4150_048		45755–45967	Conserved hypothetical protein, pseudogene, similar to part of ISCpe3 transposase
pJIR4150_049		46419–47321, C	Conserved hypothetical protein, homologue of pCW3_0048
pJIR4150_050		47450–47857	Conserved hypothetical protein, homologue of pCW3_0049
pJIR4150_051		47977–48225	Conserved hypothetical protein, homologue of part of pCW3_0050
pJIR4150_052		49593–49904, C	Conserved hypothetical protein, homologue of pCpb2-CP1_58
pJIR4150_053		50308–52023	Putative cell wall surface anchor family protein with two Cna_B-like domains
pJIR4150_054		52023–53582	Putative cell wall surface anchor family protein with one Cna_B-like domain
pJIR4150_055	int	54367–55563, C	Tyrosine recombinase or integrase, 67% identical over 391 aa to the Tn916 Int
pJIR4150_056	xis	55643–55849, C	Excisionase, 83% identical over 62 aa to Tn916 Xis
pJIR4150_057	ORF8	56249–56497, C	Putative transcriptional regulator, 46% identical over 69 aa to Tn916 ORF8
pJIR4150_058	ORF7	56484–56912, C	Alternative sigma factor, 38% identical over 136 aa to Tn916 ORF7
pJIR4150_059		57352–57717, C	Conserved hypothetical protein
pJIR4150_060		57737–57961, C	Conserved hypothetical protein
pJIR4150_061	ORF9	58113–58475	Putative HTHb-XRE family transcriptional regulator, 36% identical over 76 aa to Tn916 ORF9
pJIR4150_062		58515–58850, C	Putative transcriptional regulator
pJIR4150_063		58985–59551, C	Putative acetyltransferase
pJIR4150_064		59613–60521, C	Putative transcriptional regulator with potential HTH domain
pJIR4150_065	ORF13	61471–62364, C	VirB8-like conjugation protein from TcpC superfamily, 53% identical over 293 aa to Tn916 ORF13
pJIR4150_066	ORF14	62382–63389, C	Peptidoglycan hydrolase, 68% identical over 336 aa to Tn916 ORF14
pJIR4150_067	ORF15	63386–65512, C	Hypothetical conjugation protein, 58% identical over 671 aa to Tn916 ORF15
pJIR4150_068	ORF16	65512–67959, C	AAA-like ATPase conjugation protein, 77% identical over 813 aa to Tn916 ORF16
pJIR4150_069	ORF17	67946–68335, C	TcpE superfamily family conjugation protein, 72% identical over 126 aa to Tn916 ORF17
pJIR4150_070	ORF18	68446–68949, C	ArdA antirestriction superfamily protein, 59% identical over 166 aa to Tn916 ORF18
pJIR4150_071		68967–69464, C	Putative ArdA family protein, 35% identical over 165 aa to Tn916 ORF18
pJIR4150_072		69799–70107, C	Conserved hypothetical protein
pJIR4150_073	bcrD	70109–70936, C	Undecaprenyl pyrophosphate phosphatase, 90% identical over 273 aa to Enterococcus faecalis BcrD (GenBank accession no. AAS78449)
pJIR4150_074	bcrB	70936–71637, C	Bacitracin ABC-2 family transporter permease, 89% identical over 233 aa to Enterococcus faecalis BcrB (GenBank accession no. AAS78450)
pJIR4150_075	bcrA	71678–72595, C	Bacitracin ABC transporter ATPase, 92% identical over 305 aa to Enterococcus faecalis BcrA (GenBank accession no. AAS78451)
pJIR4150_076	bcrR	72776–73990, C	Helix-turn-helix-type XRE family transcriptional regulator, 77% identical over 204 aa to Enterococcus faecalis BcrR (GenBank accession no. AAS78452)
pJIR4150_077		73380–73706, C	Putative HTH domain protein
pJIR4150_078		73743–74096	Putative DNA binding protein with HTH domain
pJIR4150_079		74177–74365, C	Conserved hypothetical protein
pJIR4150_080		74392–74838, C	Putative acetyltransferase family protein
pJIR4150_081		74855–75265, C	Putative NTF-2/SnoaL-like protein
pJIR4150_082	ORF19	75460–75681, C	Conjugation protein, 91% identical over 73 aa to Tn916 ORF19
pJIR4150_083		75678–75818, C	Small conserved hypothetical protein
pJIR4150_084	ORF20	75834–76946, C	Relaxase, 64% identical over 328 aa to Tn916 ORF20
pJIR4150_085	ORF21	77215–78609, C	Coupling protein FtsK/SpoIIIE-like DNA translocase, AAA-ATPase superfamily, 63% identical over 460 aa to Tn916 ORF21
pJIR4150_086		78713–79138, C	Conserved hypothetical protein
pJIR4150_087	ORF22	79264–79650, C	Conserved hypothetical protein, 63% identical over 127 aa to Tn916 ORF22
pJIR4150_088	ORF23	79666–79992, C	Conserved hypothetical, 67% identical over 102 aa to Tn916 ORF23
pJIR4150_089		80406–81227, C	Putative transposase
pJIR4150_090		81281–81553, C	Putative transposase with HTH domain
pJIR4150_091		81626–84499, C	Putative cell wall protein with two potential Cna_B domains
pJIR4150_092		84911–85336, C	Conserved hypothetical protein
pJIR4150_093		85848–86201	Conserved hypothetical protein, homologue of pCW3_0049
pJIR4150_094		86417–87274	Putative transposase, homologue of pCW3_0050
pJIR4150_095		87476–88195, C	Conserved hypothetical protein
pJIR4150_096		88270–88590	Conserved hypothetical protein
pJIR4150_097		888740–89611	Putative resolvase, homologue of pCW3_0006

^aC, complementary strand.

^bHTH, helix-turn-helix.

pJIR4150 had the 40-kb core region (Fig. 3) that is conserved among the members of the large conjugative toxin and antibiotic resistance plasmid family from C. perfringens; this region contains the genes essential for plasmid replication, maintenance, and conjugative transfer (31). A 49-kb region, including ICECp1, was unique to pJIR4150 (Fig. 3). Other genes in this region included bcr_00044, which encodes a large clostridial glycosylating toxin, TpeL (1,779 amino acids [aa]); bcr_00053, bcr_00054, and bcr_00091, which encode putative surface proteins (571 aa, 519 aa, and 957 aa, respectively); and others genes encoding hypothetical proteins (GenBank accession number LN835295).

DISCUSSION

In this study, we have shown that bacitracin resistance in an avian necrotic enteritis isolate of C. perfringens was determined by a bcrRABD locus that was carried by an 89.7-kb conjugative plasmid, pJIR4150, which was completely sequenced. Mutagenesis studies showed that two genes in this locus, bcrA and bcrB, were both essential for bacitracin resistance and that they were sufficient to confer bacitracin resistance in a bacitracin-susceptible C. perfringens strain. Next-generation sequencing led to the identification of a novel Tn916-like element, ICECp1, that carried the bcrRABD locus and was located on pJIR4150. Further analysis of pJIR4150 revealed that apart from this putative ICE, this plasmid contained the 40-kb core region that is conserved among large conjugative plasmids in C. perfringens, a tpeL toxin gene, and other genes that were unique to pJIR4150.

A recent study (5) reported the identification of a similar bacitracin resistance region on the chromosome of a bacitracin-resistant, turkey isolate of C. perfringens, although the sequence of the locus lodged in the database appears to be incomplete and no genetic studies were carried out. The findings of our studies are in agreement with those of the previous work (5) and provide experimental validation for the role of the bcrA and bcrB genes in conferring bacitracin resistance. Taken together, these studies provide evidence that bcrAB-mediated bacitracin resistance can be either chromosomal or plasmid determined in C. perfringens and that the resistance genes may be located on different mobile genetic elements, ICECp1 in JGS4102 and an IS1216-derived element in strain c1261_A.

Extensive studies have shown that an equivalent ABC transporter system is responsible for bacitracin resistance in several bacteria; these resistance determinants also contain a regulatory gene(s). In some bacteria, this bacitracin resistance locus also encodes a two-component regulatory system that responds to exposure to bacitracin (10–12). In contrast, in E. faecalis there is only one regulatory gene, bcrR, the product of which functions as a sensor kinase and as a response regulator (13). The putative bacitracin resistance system identified in our study was genetically closest to that of E. faecalis. We have presented evidence that BcrR is not essential for bacitracin resistance in C. perfringens strain JIR325, which appears to be different from the findings obtained in E. faecalis and other organisms (9–13, 34). However, other workers have shown that in different Enterococcus spp. the bcrAB genes are expressed constitutively in the absence of bcrR, which is sufficient for bacitracin resistance (35).

To our knowledge, ICECp1 is the first Tn916-like putative ICE shown to harbor a bacitracin resistance determinant. Tn916 was originally found in E. faecalis (36) and comprises 24 ORFs, the products of which are involved in conjugative transfer, recombination (excision and insertion reactions), transcriptional regulation, and tetracycline resistance (36, 37) (Fig. 4). Over 35 different bacterial genera and often multiple species within a single genus have been shown to carry Tn916-like elements (37). Despite the high level of similarity between ICECp1 and Tn916, there were some fundamental differences. For example, ORFs 5, 6, 10, 12, and 24 of Tn916 were not present in ICECp1 (Fig. 4, Table 3). The absence of ORF12 is not surprising, as it encodes a tet(M) leader peptide that appears to be involved in the regulation of tet(M) expression (38). The other ORFs are not conserved in all Tn916-like elements (39, 40). Other differences included the absence of the tet(M) gene, which was replaced by a bacitracin resistance cassette inserted at a different location. Two ORFs (pJIR4150_090 and pJIR4150_089) that were located upstream of ORF23 and that were not orthologous to any Tn916 ORFs may also be part of ICECp1. These ORFs encode putative transposases that have high levels of sequence identity to transposases from Streptococcus entericus. The next ORF, pJIR4150_091, encodes a putative collagen adhesion-like protein, an orthologue of which is present in Tn5386 (40). In addition, there were 14 other ORFs in ICECp1 that encoded putative hypothetical proteins that did not show similarity to proteins encoded by genes located on Tn916 (Table 3; Fig. 4).

The presence of accessory antibiotic resistance genes and other ORFs in Tn916-like ICEs has been reported previously. For example, Tn5397 from Clostridium difficile carries the gene for the TndX large serine recombinase instead of the genes for the Xis and Int proteins and also contains a self-splicing group II intron (Fig. 4). Tn5801 from Staphylococcus aureus has accessory ORFs located before the conjugation module; one of these, sav0415, is predicted to encode a transposase (Fig. 4). In Tn5386, in addition to the lantibiotic immunity genes that are found instead of tet(M), a group II intron and a gene encoding a putative cell surface-associated protein are inserted (40), as discussed previously. In S. aureus, Tn6009 carries the mercury resistance mer operon incorporated into a Tn916-like element (41). These findings suggest that the broad host range of Tn916-like elements has provided a good opportunity to acquire new genetic material from diverse sources, as also now exemplified by ICECp1 from C. perfringens.

Based on the Southern hybridization results, we concluded that pJIR4150 is the largest plasmid carried by JGS4102. The high conjugation frequency observed in JIR12708, which only carried pJIR4150, provided good evidence that this plasmid is conjugative and that in this strain the transfer of bacitracin resistance is mediated by the tcp conjugation machinery carried by pJIR4150. Conversely, we found that in strain JIR12710 the bcrRABD locus is localized on the 40-kb nonconjugative plasmid pJIR4277, which was not present in the parental strain, JGS4102. We postulate that this plasmid was derived by recombination and deletion events from one or more of the plasmids present in JGS4102. The low level of transfer observed when JIR12710 was used as a donor suggests that ICECp1, like Tn916, may be conjugative, but further studies are required to validate this hypothesis.

In conclusion, in this study we have identified a bacitracin resistance determinant, bcrRABD, from a necrotic enteritis-causing avian isolate of C. perfringens, JGS4102, and have shown that both bcrA and bcrB are essential for bacitracin resistance. Conjugation experiments showed that bcrRABD is transferable in mixed plate matings but that the transfer frequency is variable. By Southern hybridization and genomic sequencing, we found that in a high-transfer-frequency derivative, the bacitracin resistance determinant is located within a putative Tn916-like element, ICECp1, on a conjugative toxin plasmid, whereas in a derivative with a low frequency of transfer of bacitracin resistance, bcrRABD is localized on a nonconjugative plasmid that appears to have arisen by unknown recombination events. These findings represent the first time that a conjugative bacitracin resistance plasmid has been identified in C. perfringens and the first time that a conjugative C. perfringens plasmid that carries both antibiotic resistance genes and a toxin gene (tpeL) has been identified and fully sequenced, although a fragment from plasmid pF262A that carries a tetracycline resistance gene and that contains both the tcp locus and the cbp2 gene has been reported (42). Although the TpeL toxin has no defined role in disease, these results have significant implications for the poultry industry since they imply that treatment with bacitracin may lead to the selection of C. perfringens strains that carry a conjugative plasmid that carries both bacitracin resistance and toxin genes.