Efficient Transfer of the Pheromone-Independent Enterococcus faecium Plasmid pMG1 (Gm^r) (65.1 Kilobases) to Enterococcus Strains during Broth Mating

Abstract

Plasmid pMG1 (65.1 kb) was isolated from a gentamicin-resistant Enterococcus faecium clinical isolate and was found to encode gentamicin resistance. EcoRI restriction of pMG1 produced five fragments, A through E, with molecular sizes of 50.2, 11.5, 2.0, 0.7, and 0.7 kb, respectively. The clockwise order of the fragments was ACDEB. pMG1 transferred at high frequency to Enterococcus strains in broth mating. pMG1 transferred between Enterococcus faecalis strains, between E. faecium strains, and between E. faecium and E. faecalis strains at a frequency of approximately 10⁻⁴ per donor cell after 3 h of mating. The pMG1 transfers were not induced by the exposure of the donor cell to culture filtrates of plasmid-free E. faecalis FA2-2 or an E. faecium strain. Mating aggregates were not observed by the naked eye during broth mating. Small mating aggregates of several cells in the broth matings were observed by microscopy, while no aggregates of donor cells which had been exposed to a culture filtrate of E. faecalis FA2-2 or an E. faecium strain were observed, even by microscopy. pMG1 DNA did not show any homology in Southern hybridization with that of the pheromone-responsive plasmids and broad-host-range plasmids pAMβ1 and pIP501. These results indicate that there is another efficient transfer system in the conjugative plasmids of Enterococcus and that this system is different from the pheromone-induced transfer system of E. faecalis plasmids.

Conjugative plasmids have been identified in many streptococci and enterococci (4). Two types of conjugative plasmids in enterococci have been reported. One type of plasmid, with a molecular size of approximately 30 kb, does not produce mating aggregates and does not transfer between donor and recipient cells in liquid broth. The plasmids are able to transfer at relatively low frequencies on a solid surface, such as during filter mating (4). These plasmids usually have a broad host range and transfer between a variety of streptococcal species, and even between different genera. Macrolide-lincosamide-streptogramin B (MLS) resistance is frequently associated with this type of plasmid (13, 22, 30). Of the MLS resistance-encoding plasmids, pIP501 (1, 13, 22, 28, 29, 46) and pAMβ1 (30) are representative; the transfer mechanism for these plasmids has been characterized in detail at the molecular level.

The other type of plasmid is mainly found in Enterococcus faecalis and is a pheromone-responsive plasmid (4–6, 10–12). The E. faecalis pheromone-related plasmid conjugation system is unique among the bacteria. The pheromone-responsive plasmid transfer between E. faecalis strains occurs at a high frequency of 10⁰ to 10⁻² per donor cell within a few hours of broth mating. A plasmid of this type usually has a molecular size greater than 45 kb and confers a mating response to the small sex pheromones secreted by potential recipient cells (4–6, 10–12). This mating signal induces the synthesis of a surface aggregation substance that facilitates the formation of mating aggregates. The sex pheromones also induce a series of genes required for plasmid transfer. Plasmid-free recipients secrete multiple sex pheromones, each specific for a donor harboring a related pheromone-responsive plasmid. Once a plasmid is acquired by the recipient, secretion of the related pheromone ceases, whereas other, unrelated pheromones continue to be produced. The pheromone (culture filtrate of a plasmid-free strain) induces self-aggregation of donor cells, as shown by suspension of the donor cells in a culture filtrate. Determinants encoded on pheromone-responsive plasmids include those for hemolysin, bacteriocin, and resistance to antibiotics and UV light (2, 4, 15, 24–26, 35, 41, 43, 44).

Drug-resistant enterococci have been an increasingly significant cause of nosocomial infection in recent years (34). Enterococcus faecium clinical isolates usually have multiple-drug resistance. Of the drug-resistant strains, those resistant to high levels of penicillin and those resistant to gentamicin and glycopeptide (vancomycin) cause serious nosocomial infections (31, 34, 45). Among the resistance traits acquired by E. faecium, glycopeptide (vancomycin) resistance has been shown to be encoded on a conjugative plasmid (9, 31, 32, 38). One of the vancomycin resistance plasmids, pIP819 (34 kb), which is a relatively small plasmid, transfers by filter mating to other E. faecium and E. faecalis strains, a variety of streptococcus species, and Listeria monocytogenes (9, 32). The other plasmid, pHKK100 (58 kb), which is a relatively large plasmid, is a pheromone-responsive plasmid and transfers efficiently (10⁻⁴ per donor cell) from E. faecium to E. faecalis JH2-2 in broth mating (19). However, a highly efficient system for plasmid transfer during broth mating between E. faecium strains has not been described to date. In this report, we show that a plasmid conferring resistance to high levels of gentamicin isolated from an E. faecium clinical strain transferred efficiently to E. faecium or E. faecalis during broth mating and that the plasmid was different from the pheromone-responsive plasmid.

MATERIALS AND METHODS

Bacterial strains, plasmids, and media.

The bacterial strains and plasmids used in this study are listed in Table 1. E. faecalis strains were grown in Todd-Hewitt broth (THB) (Difco Laboratories, Detroit, Mich.) or N2GT broth (nutrient broth no. 2; Oxoid Ltd., London, England) supplemented with 0.2% glucose and 100 mM Tris-HCl (pH 7.5). N2GT broth was also used in the mating experiments. Escherichia coli strains were grown in Luria-Bertani medium. Agar plates were prepared by adding 1.5% agar to broth media. All bacterial strains were grown at 37°C. Antibiotics were used at the following concentrations: ampicillin, 100 μg/ml; erythromycin, 12.5 μg/ml; chloramphenicol, 20 μg/ml; fusidic acid, 25 μg/ml; rifampin, 25 μg/ml; tetracycline, 10 μg/ml; streptomycin, 500 μg/ml; kanamycin, 500 μg/ml; gentamicin, 500 μg/ml.

TABLE 1.

Bacterial strains and plasmids

Strain or plasmid	Genotype or phenotype	Description and source or reference
Strains
E. faecalis FA2-2	rif fus	Derivative of JH2; 8
E. faecalis JH2SS	str spc	Derivative of JH2; 42
E. faecium KTRF	rif fus	Derivative of plasmid-free E. faecium KT; this study
E. faecium KTSS	str spc	Derivative of plasmid-free E. faecium KT; this study
E. hirae 9790RF	rif fus	Derivative of E. hirae ATCC 9790
E. hirae 9790SS	str spc	Derivative of E. hirae ATCC 9790
E. coli DH1	F−recA1 endA1 gryA96 thi-1 relA1 hsdR17 supE44	36
Plasmids
pMG1	Gmr	65.1-kb conjugative plasmid from E. faecium strain; this study
pAD1	hly/bac uvr	59.6-kb pheromone-responsive conjugative plasmid from DS16; 8
pAM714	hly/bac erm	pAD1::Tn917, wild-type transfer; 23
pPD1	bac	59-kb pheromone-responsive conjugative plasmid from E. faecalis 39-5; 49
pAM373	tet	36-kb pheromone-responsive conjugative plasmid; 7
pAMβ1	erm	26.5-kb broad-host-range conjugative plasmid from DS5; 30
pIP501	erm cat	39.2-kb broad-host-range conjugative plasmid; 13
pAM120	amp tet	pBR322-derived vector pGL101 carrying pAD1 EcoR1 fragment F::Tn916; 18
pMG326	amp lacZ	E. coli vector pMW119 (amp lacZ) (Nippon Gene Co., Ltd.) carrying a part (11.5 kb) of pPD1 EcoRI fragment A; 37
pAM401	cat tet	E. coli-E. faecalis shuttle vector; 48

Mating procedures.

Broth matings were performed as previously described (11) with a donor/recipient ratio of 1:10. Overnight cultures of 0.05 ml of the donor and 0.45 ml of the recipient were each added to 4.5 ml of fresh N2GT broth, and the mixtures were incubated at 37°C with gentle agitation for the appropriate times and then vortexed. Portions of the mixed cultures were then plated on solid media with appropriate selective antibiotics. Colonies were counted after 48 h of incubation at 37°C. Pheromone induction and the mating experiments were performed as previously described (5, 6, 10, 23).

Clumping assay.

Detection of aggregation (clumping) was as previously described (4, 10, 11). The pheromone corresponded to a culture filtrate of plasmid-free strain FA2-2. Generally, 0.5 ml of culture filtrate from late-log-phase cells was mixed with 0.5 ml of fresh N2GT broth and 20 μl of the overnight-cultured cells to be tested for their ability to respond. The mixtures were cultured for 2 to 4 h at 37°C with shaking and were examined for clumping.

Isolation and manipulation of plasmid DNA.

Plasmid DNA was isolated by the alkaline lysis method (36). Plasmid DNA was treated with restriction enzymes and subjected to agarose gel electrophoresis for analysis of DNA fragments, etc. Restriction enzymes were obtained from Nippon Gene (Toyama, Japan), New England Biolabs, Inc., and Takara (Tokyo, Japan) and were used in accordance with the supplier’s specifications. Agarose was obtained from Wako Chemicals, Osaka, Japan. Gels with a 0.8% agarose concentration were used for size determination of large DNA fragments (greater than 0.5 kb), and 2.0% agarose gels were used to determine the sizes of smaller fragments (less than 0.5 kb) (15, 43). A “glass milk” kit (Gene Clean II kit; Bio 101, Inc., La Jolla, Calif.) or low-melting-point agarose and β-agarase I (Nippon Gene) were used for the elution of the DNA fragments from agarose gels. The eluted fragments were ligated to dephosphorylated, restriction enzyme-digested vector DNA with T4 DNA ligase and were then introduced into E. coli by electrotransformation (14). Transformants were selected on Luria-Bertani agar medium containing suitable antibiotics.

Determination of the pMG1 restriction map.

Initially, pMG1 DNA was digested with EcoRI, and the sizes of the EcoRI fragments were determined. The molecular sizes of EcoRI fragments A to E were 50.2, 11.5, 2.0, 0.7, and 0.7 kb, respectively. To determine the order of the EcoRI fragments, pMG1 DNA was partially digested with a low concentration of EcoRI or was completely digested with EcoRI, KpnI, XbaI, EcoRV, or NdeI and one of the other enzymes examined (double digestion). The digested DNAs were examined by agarose gel electrophoresis (data not shown), and the order of the fragments determined by these methods was ACDEB.

To determine the order of EcoRI fragments, the relational clone method was also used and a relational clone set was obtained by methods previously described (15). After agarose gel electrophoresis of an EcoRI partial digest of pMG1 DNA, the fragments were eluted and used for cloning. The cloning vector and host strain were pAM401 and E. coli DH1, respectively.

Southern hybridization.

Southern hybridization was performed with the digoxigenin-based nonradioisotope system of Boehringer GmbH (Mannheim, Germany), and all procedures were mainly based on the manufacturer’s manual and common protocols (36). Hybridization was performed at 42°C overnight in the presence of 50% formamide. Signals were detected with the DIG chemiluminescence detection kit (Boehringer GmbH). CSPD (Boehringer, GmbH) was used as a substrate for alkali phosphatase conjugated to antidigoxigenin antibody.

RESULTS AND DISCUSSION

Isolation of plasmid pMG1 from an E. faecium clinical isolate.

Thirty-three E. faecium clinical isolates were examined for resistance to gentamicin. Two of the 33 E. faecium clinical isolates showed resistance to concentrations of gentamicin greater than 1,000 μg/ml. The two strains were designated GF112 and GF113. GF112 was resistant to erythromycin, tetracycline, and gentamicin, and GF113 was resistant to ampicillin, tetracycline, and gentamicin. To examine the transferability of the gentamicin resistance trait, broth mating experiments were performed between each of the Gm^r E. faecium strains and an E. faecium KTRF recipient strain. The mixture of donor and recipient was incubated at 37°C with gentle agitation for 4 h and then vortexed. A 0.1-ml aliquot of the mixed culture was plated out on THB agar containing 500 μg of gentamicin per ml for selection of the plasmid plus 25 μg of rifampin and fusidic acid per ml for counterselection. The two strains were found to transfer the gentamicin resistance trait. GF112 was used for further analysis. E. faecium GF112 harbored several plasmids, and the E. faecium KTRF transconjugant also harbored a number of plasmids (data not shown). To identify the Gm^r plasmid, repeated transfer experiments between E. faecium KTRF and KTSS were performed by short mating (30-min mating). To examine the plasmid content of the transconjugant, plasmid DNA was isolated from 30 transconjugants obtained from each mating experiment; the DNA was digested with EcoRI and examined by agarose gel electrophoresis. After repeated transfer experiments, one gentamicin-resistant transconjugant of E. faecium KTRF was found to harbor a single plasmid. The plasmid was designated pMG1 and conferred gentamicin resistance. The restriction map of pMG1 was constructed by the relational clone method described in Materials and Methods. EcoRI restriction of pMG1 gave five fragments, A through E, with molecular sizes of 50.2, 11.5, 2.0, 0.7, and 0.7 kb, respectively. The clockwise order of the fragments was ACDEB (Fig. 1). The cleavage sites for KpnI, XbaI, EcoRV, and NdeI are also shown in Fig. 1.

FIG. 1.

Physical map of pMG1. Fragments produced by restriction endonuclease restriction of pMG1 DNA are denoted by letters.

Conjugative transfer in Enterococcus strains.

Plasmid pMG1 was examined for conjugative transfer in broth. As shown in Table 2, pMG1 transferred between E. faecium strains, between E. faecalis strains, and between E. faecium and E. faecalis strains. The transfer frequencies were around 10⁻⁴ per donor cell after 3 h of mating (Table 2). pMG1 also transferred from E. faecium to E. hirae 9790 at a frequency of about 10⁻⁴ per donor cell (data not shown). Plasmid pAM714 is an erythromycin resistance-conferring derivative of pAD1::Tn917 (23). Plasmid pAD1 (58 kb) (5, 6, 8) is a pheromone-responsive conjugative hemolysin/bacteriocin plasmid found in E. faecalis. Pheromone-responsive conjugative plasmid pAM714 transfers between E. faecalis strains at a frequency of 10⁻² per donor cell. However, the plasmid did not transfer to E. faecium in broth.

TABLE 2.

Transfer frequency of plasmid pMG1 in Enterococcus^a

Donor	Recipient	Type of mating	Transfer frequency after 3 h of mating (no. of transconjugants per donor cell)
E. faecium KTSS(pMG1)	E. faecium KTRF	Intraspecies	6 × 10−4
E. faecium KTSS(pMG1)	E. faecalis FA2-2	Interspecies	2 × 10−4
E. faecalis FA2-2(pMG1)	E. faecalis JH2SS	Intraspecies	4.5 × 10−4
E. faecalis JH2SS(pMG1)	E. faecium KTRF	Interspecies	4 × 10−5
E. faecalis JH2SS(pAM714)	E. faecalis FA2-2	Intraspecies	2 × 10−2
E. faecalis JH2SS(pAM714)	E. faecium KTRF	Interspecies	<10−7

^aBroth matings were performed as described in Materials and Methods. Fresh N2GT broth and mixtures of donor and recipient strains were incubated at 37°C with gentle agitation for 3 h and then vortexed. Aliquots of the mixed culture were then plated on solid medium with appropriate antibiotics for selection. The selective medium was THB agar containing gentamicin or erythromycin for the selection of pMG1 or pAM714 transconjugants, respectively, plus the appropriate drugs for the selection of the recipient strains. The plates were incubated for 2 days at 37°C, and the colonies were counted. 

Figure 2 shows the kinetics of pMG1 and pAM714 plasmid transfer. pMG1 showed normal transfer between E. faecium strains and between E. faecalis strains. Approximately 60 min of mating was required before a significant level of transfer occurred, and an increase in the number of transconjugants was observed between 2 and 3 h after the start of the mating experiment. In the early stage of mating, e.g., between 15 and 30 min, a few transconjugants grew on the selective agar plates. Pheromone-responsive plasmid pAM714 also showed normal transfer (23). The maximum transfer frequency of pMG1 was about 5 × 10⁻⁴ after 3 h of mating, which was about a factor of 1 × 10⁻² less than that of pAM714. For the pheromone-responsive plasmid, a sex pheromone secreted into the broth culture by the recipient induces the synthesis of an aggregation substance on the donor cell surface, and donor and recipient then form a mating aggregate (4, 10, 11). Usually, the transconjugant of the pheromone-responsive plasmid is not obtained in a short mating period, such as 15 min.

FIG. 2.

Kinetics of transfer for pheromone-independent plasmid pMG1 and pheromone-dependent plasmid pAM714. Matings were carried out with mixtures that each consisted initially of 0.05 ml of the donor, 0.45 ml of the recipient, and 4.5 ml of fresh N2GT broth. The mixtures were incubated at 37°C with gentle shaking. At the indicated time points, 0.1-ml samples were removed, diluted appropriately, and plated on THB agar plates containing gentamicin or erythromycin for the selection of pMG1 or pAM714 transconjugants, respectively, plus drugs for selection of the recipient strains. Symbols: ○, transfer from E. faecalis FA2-2(pMG1) to E. faecalis JH2SS; ▵, transfer from E. faecium KTSS(pMG1) to E. faecium KTRF; □, transfer from E. faecalis JH2SS(pAM714) to E. faecalis FA2-2.

The mating kinetics of pMG1 suggest that mating aggregates occur by the random collision of donor and recipient cells, with little or no transfer of the plasmid occurring in the first 60 min of mating. Increased transfer occurs after this period. In the early stage of the mating experiment, a few mating cells survived and completed mating on the selective agar plate, and the resulting transconjugants then grew. These observations suggest that some factor necessary for plasmid transfer is induced during coincubation of the donor and recipients and that induction of this factor might require contact between the donor and recipient.

Pheromone response.

The pheromone-responsive plasmids of E. faecalis induce the synthesis of a proteinaceous surface adhesion substance (i.e., aggregation substance) on the donor cell surface, which facilitates the formation of mating aggregates upon random collision between donor and recipient cells (4, 10, 11). The donor cells can be induced to self-aggregate by induction with the pheromone of recipient cells (4, 10, 11). Mating aggregates are observed by the naked eye during broth matings between E. faecalis donor cells containing pAM714 and an E. faecalis recipient cell. The self-aggregates of the donor cells were also observed by induction with a culture filtrate (pheromone) of FA2-2 (23).

In mating experiments between donor cells containing pMG1 and recipient cells, mating aggregates were not observed by the naked eye during broth mating. During broth mating, small mating aggregates of several cells were observed by microscopy (Fig. 3A), while aggregates of donor cells were not even observable by microscopy after exposure to a culture filtrate of the recipient strain, E. faecalis FA2-2 or E. faecium KTRF (Fig. 3B).

FIG. 3.

Microscopic observation. The mating mixture (A) and pheromone-exposed donor cells (B) were observed by microscopy (Nikon Optiphoto 2 at ×1,000 magnification) after 2 h of mating between E. faecalis FA2-2(pMG1) donor cells and E. faecalis JH2SS recipient cells, as shown in Fig. 2 (A), and after 2 h of exposure of E. faecalis FA2-2(pMG1) to FA2-2 culture filtrate (pheromone), as shown in Table 3 (B).

We examined whether the mating aggregate is dispersed by the addition of EDTA (49). Mating aggregates of E. faecalis JH2SS containing pheromone-responsive plasmid pAM714 and the recipient FA2-2 were dispersed by the addition of 25 mM EDTA, whereas mating aggregates of E. faecalis JH2SS containing pMG1 and the recipient FA2-2 were not dispersed by the addition of 25 mM EDTA (data not shown). These results suggest that a mating substance encoded by pMG1 may be produced and may be located on the donor cell surface; this mating substance is different from the aggregation substance produced by the E. faecalis strain containing the pheromone- responsive plasmid.

Pheromone induction and a mating experiment were performed to determine whether plasmid pMG1 responded to the pheromone of E. faecalis (5, 6, 10, 23). Donor cells of E. faecalis FA2-2(pMG1) or E. faecium KTRF(pMG1) and E. faecalis FA2-2(pAM714) were exposed (for 120 min) to an FA2-2 culture filtrate (pheromone) to induce aggregation and mating functions before a short (10-min) mating period. The short mating was carried out between the induced or uninduced donor cells and the recipient cells. As shown in Table 3, the transfer frequency for the induced and uninduced plasmid pMG1 was less than 10⁻⁷ per donor cell. On the other hand, the transfer frequency of induced pheromone-responsive plasmid pAM714 was 10⁻³ per donor cell and that of the uninduced plasmid was less than 10⁻⁷ per donor cell. These results indicated that plasmid pMG1 did not respond to the pheromone.

TABLE 3.

Transferability of plasmid pMG1 during short mating after exposure to the E. faecalis pheromone^a

Donor	Recipient	Exposure to pheromone induction for 2 hb	Transfer frequencyc after a 10min mating
E. faecalis FA2-2(pMG1)	E. faecalis JH2SS	+	≦10−7
		−	≦10−7
E. faecium KTRF(pMG1)	E. faecium KTSS	+	≦10−7
		−	≦10−7
E. faecalis FA2-2(pAM714)	E. faecalis JH2SS	+	10−3
		−	≦10−7

^aFor induction with pheromone, 0.1 ml of an overnight culture of the donor strain was diluted with 0.9 ml of a 1:1 mixture of pheromone (culture filtrate of strain FA2-2) and fresh N2GT broth. The overnight culture was similarly diluted with N2GT broth without induction as a control. Each culture was incubated for 2 h with gentle agitation at 37°C. After incubation, 0.1 ml of each donor strain was mixed with 0.9 ml of the indicated recipient strain and the mixture was incubated for 10 min at 37°C. The mixtures were plated on selective medium containing either 500 μg of gentamicin, 500 μg of streptomycin, and 500 μg of spectinomycin per ml for the selection of pMG1 transconjugants or 12.5 μg of erythromycin, 500 μg of streptomycin, and 500 μg of spectinomycin for the selection of pAM714 transconjugants. 

^b+, induction was performed; −, induction was not performed. 

^cNumber of transconjugants per donor cell. 

DNA-DNA hybridization.

Several pheromone-responsive plasmids have been reported. Among these plasmids, the pheromone-related conjugation systems are well studied for pAD1 (5, 6, 16, 17, 39, 40, 47), pCF10 (3, 12, 20, 33), and pPD1 (15, 17), which confer responses to the sex pheromones cAD1, cCF10, and cPD1, respectively. The genes involved in the regulation of the pheromone response have been identified and are known to be clustered in a 7-kb region on each plasmid. There is homology between the genes of these plasmids (15, 17, 21). There is also homology between the genes of the putative surface exclusion protein and those of the aggregation substance, which are located downstream of the regulatory region (15, 17). For pAD1, the genes of the pheromone response regulatory region are located in EcoRI fragment B and the genes for the aggregation substance are located in EcoRI fragments A and E (5, 17, 40). For pPD1, the genes of the pheromone response regulatory region are located in EcoRI fragment A and the genes of the aggregation substance are located in EcoRI fragments A, C, and G (15, 17). In a previous study, 18 pheromone-responsive plasmids, including pAD1, pPD1, and pCF10, were analyzed for homology by Southern blotting (21). In that study, plasmid pAM373 did not show homology to the aggregation substance genes of the other pheromone-responsive plasmids (21).

Plasmid pMG326, which contains a part (11.5 kb) of EcoRI fragment A of pPD1, contains the regulatory region for the pheromone response of pPD1, a possible surface exclusion protein gene, and the gene encoding the N-terminal region of the aggregation substance (37). Plasmid pMG326 DNA was studied for homology with that of plasmid pMG1 (Fig. 4B). The pMG326 DNA probe hybridized to specific EcoRI fragments of pheromone-responsive plasmids pAD1, pPD1, and pAM373 and broad-host-range plasmid pAMβ1. The pMG326 DNA probe hybridized to EcoRI fragments B and E of pAD1 and to EcoRI fragment A of pPD1. But the pMG326 DNA probe did not hybridize with any EcoRI fragments of pMG1 and broad-host-range plasmid pIP501. These results indicate that pMG1 did not contain any sequences homologous with the consensus sequence found in the pheromone-responsive plasmids.

FIG. 4.

Agarose gel electrophoresis of restriction endonuclease-digested plasmid DNAs and hybridization with plasmid pMG326, which contains a part (11.5 kb) of EcoRI fragment A, or the pMG1 probe. (A) Agarose gel electrophoresis of restriction endonuclease-digested plasmid DNAs. Plasmid pIP501 was digested with KpnI, and the other plasmids were digested with EcoRI. Lanes: 1, pAMβ1; 2, pAM373; 3, pIP501; 4, pMG1; 5, pAD1; 6, pPD1; 7, HindIII-digested lambda DNA. Duplicate gels were Southern blotted and hybridized to pMG326 DNA (B) or pMG1 DNA (C).

Plasmid pMG1 DNA was also studied for homology with that of pheromone-responsive plasmids pAD1, pPD1, and pAM373 and broad-host-range plasmids pIP501 and pAMβ1 (Fig. 4C). The pMG1 DNA probe hybridized to all EcoRI fragments of pMG1 DNA, whereas the pMG1 DNA probe did not hybridize with any EcoRI fragments of other plasmid DNAs.

The hybridization studies indicate that pMG1 did not contain any sequences homologous with those of the pheromone-responsive plasmids or the broad-host-range plasmids; thus plasmid pMG1 was different from these conjugative plasmids with respect to its DNA sequence.

Conjugative transposon Tn916 (16.5 kb; encodes tetracycline resistance) can be transferred from a donor to a recipient strain during conjugative transposition. The transfer of Tn916 occurs via mating on a solid surface (filter mating). Jaworski and Clewell (27) reported that mutants of Tn916 can transfer efficiently via liquid mating. Thus, conjugative transposon Tn916 was studied for homology with plasmid pMG1. Plasmid pAM120 (18), consisting of Tn916 cloned into pBR322-derived vector pGL101, was used for the hybridization study with EcoRI fragments of pMG1 DNA. pAM120 did not hybridize with any EcoRI fragment of pMG1 (data not shown), indicating that pMG1 did not contain any sequence homologous with that of conjugative transposon Tn916.

Among the gram-positive bacteria, a highly efficient transfer system which functions during broth mating has only been found in pheromone-responsive plasmids (4–6, 10, 12). The majority of pheromone-responsive plasmids have been found in E. faecalis and transfer efficiently in broth between E. faecalis strains (4, 6). One pheromone-responsive plasmid is found in E. faecium and also transfers efficiently via broth mating from E. faecium to E. faecalis (38). pMG1 is a relatively large plasmid (65.1 kb), which transfers efficiently in broth between E. faecium strains, between E. faecalis strains, between E. faecium and E. faecalis strains, and also between E. faecium and Enterococcus hirae 9790, indicating that pMG1 has great ability to transfer among the enterococci.

In contrast to what was found for the pheromone-responsive plasmids, E. faecalis or E. faecium strains carrying pMG1 were not induced to form mating aggregates or to transfer the plasmid by the E. faecalis pheromone (culture filtrate of E. faecalis FA2-2), indicating that plasmid pMG1 does not require a pheromone.