Peptides 25 (2004) 1377–1388
Review
Enterococcal peptide sex pheromones: synthesis and control of biological activity Josephine R. Chandler, Gary M. Dunny∗ Department of Microbiology, University of Minnesota Medical School, 1460 Mayo Bldg., 420 Delaware Street SE, Minneapolis, MN 55455-0312, USA Received 9 July 2003; accepted 31 October 2003 Available online 11 August 2004
Abstract The enterococcal pheromone-inducible plasmids such as pCF10 represent a unique class of mobile genetic elements whose transfer functions are induced by peptide sex pheromones. These pheromones are excreted by potential recipient cells and detected by plasmid-containing donor cells at the cell surface, where the pheromone is imported and signals induction of the plasmid transfer system. Pheromone is processed from a chromosomally encoded lipoprotein and excreted by both the donor and recipient cells, but a carefully controlled detection system prevents a response to self-pheromone while still allowing an extremely sensitive response to exogenous pheromone. © 2004 Elsevier Inc. All rights reserved. Keywords: pCF10; pAD1; pPD1; Enterococcus faecalis; Pheromone; Pheromone-inducible conjugative plasmid; PrgZ; PrgY; iCF10
1. Introduction The bacterial genus Enterococcus includes organisms with highly evolved systems for horizontal genetic transfer by conjugation [14,17,60]. This prolific exchange of genetic information contributes significantly to the medical problems associated with these organisms as opportunistic pathogens, since antibiotic resistance and virulence determinants can be exchanged readily among the enterococci and other organisms. Recent sequencing of a virulent strain of Enterococcus faecalis revealed that over one-fourth of the genome is composed of mobile elements [49], and it has been found that clinical isolates commonly harbor three to five coresident plasmids [21,57]. Several virulence traits can accumulate on a single plasmid, which transfer with high efficiency to quickly disseminate a trait through a population. The family of plasmids that shows the highest frequency of conjugative transfer among the enterococcal genetic elements are the pheromoneinducible plasmids. These plasmids are induced to transfer ∗ Corresponding author. Tel.: +1 612 625 9930; fax: +1 612 626 0623. E-mail address:
[email protected] (G.M. Dunny).
0196-9781/$ – see front matter © 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.peptides.2003.10.020
from donor cells by mating pheromones that are produced by potential recipient cells. Pheromone-inducible conjugative plasmids (referred to as pheromone plasmids in the remainder of this review) have evolved a fascinating and complex regulatory system to ensure their perpetuation and stable existence in a population. The transfer functions of the pheromone plasmids are induced by small (7–8 amino acid) peptides encoded by the chromosome of most if not all enterococcal strains. Each peptide is highly specific for a cognate plasmid, or for a family of closely related plasmids. All pheromones analyzed to date are produced by proteolytic processing of the cleaved signal sequences of secreted lipoproteins; in most cases the functions of the lipoproteins themselves are not well understood. The processed peptides are released into the growth medium and are utilized by the plasmid-containing donor cells to sense the presence of a nearby recipient cell. The best-studied representatives of this class of plasmids are pAD1, pPD1 and pCF10 [17,60]. These novel mobile elements share a similar mechanism of regulation that appears to be unique to Enterococci. The pheromone-inducible plasmids make use of an interesting combination of host and plasmid encoded gene products
1378
J.R. Chandler, G.M. Dunny / Peptides 25 (2004) 1377–1388
to sense exogenous pheromone, to subsequently activate the expression of transfer genes, and to avoid self-induction by pheromone that is encoded on the chromosome of the host cell. It has become clear that the pheromone plasmids have adapted a highly specific and sensitive method to conserve host cell metabolic energy without compromising their efficient dissemination through enterococcal populations. In this communication we will use results obtained from the analysis of pCF10, as well as pAD1 and pPD1, to illustrate some of the main features of pheromone-inducible conjugation. After providing an overview of the entire process in the following section, we will describe the synthesis of enterococcal pheromones and focus on the components of the regulatory machinery that interact specifically with the peptides, with emphasis on the molecular basis for the exquisite sensitivity and specificity of these systems.
2. An overview of pheromone-induced conjugation in the pCF10 system Fig. 1 illustrates the pheromone-dependent conjugative transfer of pCF10. Induction of plasmid transfer occurs when the recipient-produced pheromone is detected by the donor cell at the cell surface by the pCF10-encoded lipoprotein PrgZ [51]. PrgZ acts in concert with the chromosomally encoded oligopeptide permease system (Opp) to import the pheromone into the cytoplasm of the responder cell [36]. Import of the pheromone, rather than transduction of a signal across the membrane, is therefore necessary for pheromone response. After pheromone is transported into the cell it may
interact with several factors. The molecular interaction that initiates the induction process in donor cells is likely the binding of the imported pheromone to the pCF10-encoded cytoplasmic protein PrgX. PrgX is a critical negative regulator of the expression of conjugation functions. Binding of pheromone to PrgX abolishes its negative control functions so that conjugation genes are expressed. PrgX also acts as a positive regulator of its own expression and of the expression of an RNA regulator called Qa [5]. Both the expression of these regulators in the absence of exogenous pheromone, and the expression of the conjugation genes in response to pheromone are complex processes controlled at both the transcriptional and post-transcriptional levels. Pheromone-induced extension and stabilization of transcripts initiating from a strong promoter in the prgQ locus is responsible for expression of conjugation gene products encoded >5 kb downstream from the prgQ promoter. For more details on these regulatory circuits see [6,7]. As will be discussed later in this review, two additional pCF10-encoded gene products, PrgY and iCF10, are required to keep the conjugation system off in pCF10-containing cells grown in the absence of exogenous pheromone; these components likely block self-induction of donor cells by endogenous pheromone. An additional potential intracellular target of pheromone is the pCF10 encoded protein PrgW, which may be involved in the maintenance of pCF10. PrgW is a potential replication initiator protein that may interact with pheromone [36]. The most distinguishable effect of exposure of pCF10containing cells to exogenous cCF10 is visible aggregation of the culture that results from upregulation of the expression of aggregation substance (Asc10 or PrgB) from the
Fig. 1. Peptide-induced transfer of Enterococcus faecalis plasmid pCF10. (A) Peptide pheromone cCF10 (triangles) is expressed from the chromosome of both plasmid-containing donor cells and plasmid-free recipient cells. Inhibitor peptide (stars) is expressed from pCF10 and secreted into the medium to prevent pheromone from the donor cell from inducing itself, probably through competitive binding to the pheromone binding protein PrgZ. (B) Pheromone from a nearby recipient cell is detected by PrgZ. PrgZ imports the peptide pheromone using the chromosomally encoded Opp (Oligopeptide permease) system. (C) Imported cCF10 induces expression of transfer genes, including the cell–surface adhesin PrgB. (D) PrgB mediates aggregation of the donor and recipient cells. A mating channel is then formed and single-stranded pCF10 is transferred to the donor cell via a rolling circle mechanism. After pCF10 has established itself in the recipient cell, the inhibitor peptide and another mechanism of negative control, PrgY, is expressed to prevent self-induction by endogenous cCF10.
J.R. Chandler, G.M. Dunny / Peptides 25 (2004) 1377–1388
pCF10-encoded prgB gene [31]. It is believed that the aggregation of the cells that is mediated by Asc10 initiates close contact between donor- and recipient cells, thus enabling efficient transfer of the plasmid, even in liquid medium. Interestingly, expression of Asc10 is induced in vivo during experimental infections by a host factor acting through the pheromone sensing system [29]. Expression of Asc10 on the cell surface increases virulence in several experimental systems [12,37,53]. Approximately 15 additional genes likely to encode components of the conjugation machinery are located 3 from prgB, and recent microarray experiments (H. Hirt and G. Dunny, unpublished) suggest that many of these are upregulated by pheromone. However, it is not clear whether all of these transcripts originate from prgQ, or whether there are additional promoters that drive expression of these genes. Relatively little detailed analysis of these genes has been completed and they will not be discussed further here.
3. Pheromone binding and import The PrgZ pheromone binding protein is critical in the first step of pheromone induction: recognition of pheromone and import into the cell cytoplasm. This protein is so important that most, if not all, of the pheromone-inducible conjugative plasmids encode a PrgZ-like protein (called TraC in nonpCF10 plasmids). PrgZ and the TraC proteins appear to have the same function and are highly similar (>70% amino acid identity). These proteins are homologs of the peptide-binding OppA proteins found in a wide range of bacterial species [55]. An OppA protein is also encoded by the chromosome of E. faecalis [36]. The TraC, PrgZ and OppA proteins are cell surface proteins that are anchored to a lipid moiety on the outer surface of the cytoplasmic membrane via covalent linkage to an N-terminal cysteine residue [51]. The PrgZ family of pheromone-binding proteins have been shown to increase the sensitivity of each system to its cognate pheromone [46,51]. This implies that the PrgZ family has a higher binding affinity for the cognate pheromone of its plasmid system than the OppA proteins, which are involved in a generally non-specific peptide-binding and uptake process. OppA imports peptides through the well-characterized Opp system, which is part of the ATP-binding cassette (ABC) transporter family. In addition to OppA, the Opp system is composed of a membranebound translocator with two transmembrane pore-forming subunits (OppB and OppC) and two cytoplasmic ATPases (OppD and OppF) [19,28,34]. Like OppA, PrgZ recruits the Opp system to import pheromone into the cell [36]. Several Opp systems are encoded in the E. faecalis chromosome. The low specificity of OppA renders it capable of binding and facilitating the import of cCF10 at very high concentrations in the absence of PrgZ, but at physiologically relevant cCF10 levels only PrgZ has a high enough affinity to bind and import cCF10 [36]. Although analysis of the structure and mechanism of action of the PrgZ family is still in progress, extensive work has
1379
been done on the OppA proteins of Escherichia coli (DppA), Salmonella typhimurium and Lactococcus lactis. Work on these proteins led to a model for the ligand binding mechanism that is believed to generally correspond to other OppAlike proteins, including PrgZ/TraC. Molecular modeling and X-ray crystallographic studies of S. typhimurium and E. coli OppAs indicate that the peptide ligands are bound by a Venus fly-trap mechanism where two lobes are connected by a flexible hinge and the ligand binds within the cleft between them [19]. After the ligand is bound, the hinge bends to bring the two lobes together and enclose the ligand [50], creating a ‘closed’ conformation. The closed, ligand-bound OppA must then dock onto the Opp translocator and release the ligand so it can be imported into the cytoplasm by the translocator. Binding kinetics analysis of the L. lactis OppA protein done by the Poolman group [18] indicate that two different classes of peptides (categorized into class I and class II) are released to the Opp transporter at different rates. They found that class I peptides with the amino acid sequence KYGK or AAAA are imported with normal Michaelis–Menton kinetics and class II peptides, which are all other peptides, are imported at a rate that increases sigmoidally with peptide concentration [18]. They propose that the difference in the pattern of rate constants is due to different binding affinities of the class I and II peptides for two existing forms of OppA, which they designate E1 and E2 . It is unclear at this point whether E1 and E2 represent two different molecular structures of OppA, or whether they represent unassociated, free OppA versus Opp-associated, ‘docked’ OppA. A model was proposed where class II peptides bind with a higher affinity to the E1 form of OppA, and this commences the conversion of OppA from the E1 to the E2 form. In this model, the conversion from E1 to E2 is the rate-limiting step, so the rate of peptide transport increases as the number of OppA proteins already converted to E2 increases. Class I peptides, which only bind the E2 form of OppA, are imported at a rate that increases proportionally with the concentration of available ligand. A similar phenomenon was seen for the maltosebinding protein of E. coli, in which different sugars bind distinct conformational isomers of the binding protein. These were termed the B and R modes, which correspond to the E1 and E2 forms, respectively [27]. Sugars that bind the R (E2 ) mode are transported, whereas sugars that bind the B (E1 ) mode are not. It was found for this system that only the R (E2 ) mode can activate ATP hydrolysis that is necessary for peptide transport by the Opp translocator [27], whereas the sugars that bind the B mode get trapped instead of converted to the R mode. The specificity of the Opp proteins of each system varies dramatically. The S. typhimurium OppA protein facilitates import of non-specific peptides 2–5 amino acids in length [55], and the L. lactis OppA protein facilitates import of non-specific peptides ranging from 4 to at least 18 residues [18]. The dipeptide binding protein, DppA, of E. coli and the S. typhimurium OppA protein both enclose the ligand completely in the binding pocket. Within the binding pocket, wa-
1380
J.R. Chandler, G.M. Dunny / Peptides 25 (2004) 1377–1388
ter molecules are used as flexible adaptors to broaden specificity and shield charges on the buried ligand, and hydrogen bonds and salt bridges are formed by the peptide backbone that facilitate binding [54]. The L. lactis OppA protein binds longer peptides with a slightly different mechanism. In this interaction, only the first six amino acids of the peptide ligand are enclosed by the protein while the C-terminus hangs out of the binding pocket, where it interacts with residues on the surface of the binding protein [33]. The interaction of the C-terminal residues with the binding protein appears to not only be favorable for closure but to also contribute to the binding specifity more so than the amino acids in the binding pocket itself [18,33]. There is some experimental evidence that the cCF10 pheromone binds PrgZ by its N-terminus [36] and the importance of the N-terminus in determining cCF10 activity is supported by some initial experimental results. Analogs of cCF10 with substituted amino acids at or near the N-terminus failed to induce activity at physiological concentrations (4 and MHA and GMD, unpublished results) whereas a tyrosine addition to the C-terminus exhibited normal biological activity [36]. Synthetic hybrid cPD1/cAD1 peptides also indicated that the N-terminus is important for specificity [14,32]. It is possible that the N-terminal interaction of pheromones within the binding pocket of their cognate peptide-binding proteins is uniquely specific. Alternatively, pheromone specificity may not be limited to PrgZ and these results may instead reflect a specific response mediated after import of the peptide occurs. In support of this alternative, there is evidence that the PrgZ family proteins are not entirely specific for their cognate pheromone. It is believed that they also bind the cognate inhibitors within each system with a lower affinity [43,45]. In addition, several of the pheromone binding proteins bind peptides of other systems, although the range is limited and cross-binding does not occur for every system [44,45,51,56]. These non-cognate peptides also do not signal [22,24]. The significance of these non-cognate interactions is not clear. In addition to the well-characterized OppA peptide binding proteins, recent sequencing projects have revealed a great number of putative peptide binding proteins within the chromosome of other systems. They are very similar to one another, but distinct differences can be seen even among the closely related E. faecalis TraC and PrgZ protein sequences [22]. Alignment of the three E. faecalis peptide binding proteins demonstrates that the first 300 amino acids are highly conserved whereas there is variability in the last ∼220 amino acids [22]. Some of these C-terminal residues line the binding cleft and may contribute to specificity. Many of the OppA proteins show sequence variability in the binding residues [19], indicating that the binding characteristics of each family member may diverge significantly. A highly conserved aspartate residue (D432 in PrgZ) is believed to lie within the binding cleft and interact with the N-terminal ␣-amino group of the bound peptide. Interestingly, the most highly conserved regions appear to be on the surface of the OppA binding protein [19], which is impressive considering that
surface residues typically have a higher freedom of mutation than buried residues. An important interaction with another protein would restrict this freedom of mutation, which may indicate that these highly conserved regions represent docking sites of the closed receptors onto the membraneassociated Opp translocator. Although much has been learned about the mechanism of the peptide binding proteins, there are clearly important unknowns that remain to be elucidated. The molecular details of the interaction between all OppA proteins and the membranebound Opp translocator, as well as the process of release of the peptide to this translocator, have thus far been primarily speculative. It has yet to be determined how the binding mechanism observed in L. lactis corresponds to pheromone binding in the case of the E. faecalis proteins. Given what is known about the specificity of the E. faecalis pheromone binding proteins and the analysis of their sequences it seems likely that their binding mechanism will differ from those observed with L. lactis OppA. In addition, the potential involvement of regulatory proteins in this process has been given very little attention. Such proteins could affect the specificity of binding, the rate of conformational change or the process of docking and release.
4. Pheromone synthesis and processing The pheromones themselves are encoded within the chromosome of E. faecalis, processed by host machinery and secreted in extremely small amounts into the culture supernatant. In the case of pCF10, the pheromone is secreted at ∼10−11 M and can induce a donor cell at concentrations as low as 2 × 10−12 M, which corresponds to less than five molecules per cell under the conditions tested [38]. Most strains produce multiple pheromones with different spectra of activities so that a single cell produces a battery of different active pheromones. Despite the diverse array of pheromones secreted by a single cell, each plasmid responds specifically to its cognate pheromone with surprising sensitivity, and does not respond to the pheromones of other plasmids [22,24]. Recent sequencing of the E. faecalis strain V583 [49] has revealed that the pheromone precursors lie within the Nterminal signal sequences of predicted surface lipoproteins [15]. The peptide sequences for several of these, and the surrounding amino acids that are part of the lipoprotein precursor, are listed in Table 1. The peptides for each plasmid system are very similar despite the specificity of each plasmid system’s response to its cognate pheromone. As can be seen in Table 1, the peptides are all 7–8 amino acids long and very hydrophobic. Proteolytic processing of the cCF10 precursor, as well as the cAD1 and cPD1 precursors, must occur for the release of the mature pheromone into the culture supernatant. All of the precursors have a cysteine residue downstream of the pheromone sequence (see Table 1) that is within a known lipoprotein signal peptidase processing site [15]. This
J.R. Chandler, G.M. Dunny / Peptides 25 (2004) 1377–1388
1381
Table 1 Pheromone precursors and processed peptide Plasmid
Pheromone/precursora b
Precursor gene namec (amino acid length)
Inhibitor peptide
pCF10 pAD1 pPD1 pAM373 pOB1
. . .magLVTLVFVlsacg. . . . . .aiaLFSLVLAGcg. . . . . .gllFLVMFLSGcv. . . . . .llgAIFILAScg. . . . . .tvaVAVLVLGAcg. . .
CcfA, 275 (4) Cad, 309 (2) 234 (15) 166 (15) 272 (15)
. . .vviAITLIFI . . .litLFVVTLVG . . .llfALILTLVS . . .iglSIFTLVA . . .SLTLILSA
a b c
Large bold letters represent amino acids in mature peptide. Bold C indicates cysteine residues upstream of predicted signal peptidase II cleavage site. The cPD1, cAM373, and cOB1 precursors have been identified by sequence analysis only and has not yet been named.
cysteine residue is the site where signal peptidase II cleaves, liberating the signal peptide from the lipoprotein [20]. In the case of cCF10, the cysteine residue is three amino acids downstream of the mature peptide, and therefore, may require further processing by a host exoprotease (illustrated in Fig. 2A). At least some of the prepeptides may be further processed by another protease that was recently identified as the membrane protein, Eep [3], that mediates cleavage at the upstream processing site. Eep has similarity to some zinc metalloproteases and may be a member of the “regulated intramembrane proteolysis” (Rip) family of proteins that are involved in a wide
variety of cellular processes [10]. Since Eep appears to be a membrane protein, the Eep-mediated cleavage of pheromone precursors most likely occurs there and may be accompanied by an active export across the membrane to the cell wall or extracellular environment. The peptide pheromones are extremely hydrophobic and may non-specifically bind to nonpolar cellular components such as the cell membrane, therefore potentially requiring active export in order to become disassociated from the cell. This type of simultaneous processing and export has been reported for peptide bacteriocins and related regulatory peptides [47], although to date there is
Fig. 2. Pheromone processing, release and sensing of peptides at the cell surface of donor cells. (A) Processing of cCF10 to mature peptide. The heptapeptide cCF10 is within the carboxy-terminal end of the signal sequence of the lipoprotein CcfA. Signal peptidase II cleaves before the cysteine residue contained within the conserved lipobox processing site, liberating the signal peptide. Further processing most likely occurs in the cell membrane by Eep, which cleaves at the amino-terminal end of the cCF10 peptide sequence. Final processing may be carried out by an exopeptidase, which cleaves off the remaining three C-terminal residues, resulting in mature cCF10. (B) Model of cCF10 export and endogenous control. As cCF10 is processed and excreted, a considerable amount remains associated with the cell wall. iCF10 neutralizes the pheromone released into the medium, whereas PrgY (see text and Fig. 4) controls wall-associated pheromone activity. PrgZ binds cCF10 that is not competitively inhibited by iCF10 in the extracellular medium. Components expressed from the plasmid are filled in with black, all the others are chromosomally encoded.
1382
J.R. Chandler, G.M. Dunny / Peptides 25 (2004) 1377–1388
no evidence for any active export. Interestingly, the peptides cOB1 and cAM373 do not appear to be processed by Eep [3], implying that an alternative processing system may be used by these peptides. There is nothing currently known about the regulation of expression of any the pheromone precursor proteins. There is very little known about the function of the lipoproteins from which the pheromones are processed. They appear to be unrelated, other than being lipoproteins with pheromone peptides in their signal sequences, and they are not closely located in the genome. The cPD1 and cCF10 precursors are an exception, however, in that they both have some similarity to SpoIIIJ, a protein related to sporulation in Bacillus subtilis. The cCF10 precursor was also found to share similarity with the protein YidC [4], a Gram-negative inner membrane translocase that is required for viability [52]. Some of the other lipoprotein pheromone precursors have similarity to putative lipoproteins of unknown function. Mutations in the cAD1 and cAM373 pheromone precursors do not appear to affect growth or viability [2,25]. Numerous efforts to genetically inactivate the cCF10 precursor have been unsuccessful [4], although the use of anti-sense RNA corresponding to the cCF10 precursor gene has recently been used to significantly reduce the production of active pheromone (MHA and GMD, unpublished results). Curiously, the plasmid pAM373 responds to a cAM373-like pheromone produced by Staphylococcus aureus and Streptococcus gordonii [16,40]. This is one of the only known cases of an E. faecalis sex pheromone plasmid that can respond to secreted peptide from a non-enterococcal host. Sequence analysis indicates that the pheromone produced by S. aureus is also processed from the signal sequence of a lipoprotein precursor, but this precursor protein appears to be unrelated to the E. faecalis cAM373 lipoprotein precursor. In addition to the function of the lipoproteins, the regulation of expression of these important precursors is also a mystery, although it was found that cAD1 activity is increased in anaerobic conditions [59]. Whether this is related to increased expression or stability is unknown. In the case of cCF10, a significant amount of pheromone can be found in the cell wall [11]. Titration of active cCF10 in the supernatant and cell fractions revealed that almost twice as much cCF10 activity can be found in the cell wall than what is in the supernatant [11]. It is difficult to tell whether this cell wall activity originates from self-pheromone that remains in the wall during export, or from exogenous pheromone in the growth media that associates with the cell wall. Either way, this cell-bound pheromone may serve as a signaling mechanism between closely associated cells that are on a surface, such as in biofilm formation. It is also a possibility that this cell-associated pheromone is playing an as-yet undefined role in the cell physiology of the producing cell. Whatever its role, cell-associated pheromone is probably a target of the plasmidencoded functions that control signaling by self-pheromone. This control of endogenous pheromone is critical in preventing aberrant induction by the endogenous pheromone and
maintaining the high sensitivity that is required for plasmid transfer. The mechanism of this control has not yet been elucidated, but the membrane protein PrgY, which appears to have an effect on the amount of pheromone associated with the cell wall [11], is most likely involved in this process, as discussed in Section 6.
5. Control of endogenous pheromone activity by inhibitor peptides Donor (plasmid-containing) cells can potentially continue to secrete pheromone even after acquisition of the plasmid. So how does the plasmid avoid a response to its own host’s endogenously produced pheromone? The pheromone plasmids have evolved two independent mechanisms of preventing this self-induction such that the transfer response is not induced unless a nearby recipient cell is detected. One mechanism to control self-induction by endogenous pheromone is through the production of a plasmid encoded inhibitor peptide (iCF10 in the case of pCF10), which essentially neutralizes endogenously produced pheromone in the culture medium. The inhibitor peptides are proposed to compete with the pheromone for binding to the surface binding protein PrgZ or TraC [42,43] (Fig. 2B). The balance of inhibitor and pheromone in the supernatant appears to be very carefully controlled. This ensures that selfinduction is blocked, but still enables a sensitive response to a slight increase in pheromone concentration brought about by a nearby recipient cell. In the case of the pCF10 plasmid, the amount of pheromone in the supernatant is the same for both plasmid-containing donor cells and plasmid-free recipient cells. The inhibitor peptide must, therefore, be produced at a level that is remarkably coordinated with the pheromone levels so that the pheromone activity of the donor cells is neutralized after acquisition of the plasmid. The level of iCF10 in donor cultures has been found to be 10–100-fold above the pheromone level (depending on how the inhibitor activity is assayed) in terms of the molar ratio, which is just sufficient to neutralize the cCF10 activity released by the same cells. If the inhibitor peptides function by competitive binding to the pheromone-receptor proteins, this is indicative that they bind more weakly than the pheromones themselves or have less affinity for the cell wall where the pheromone-receptor proteins are located. As shown in Table 1, the inhibitor peptides are very similar to each other and to the inducing pheromones. They are 7–8 amino acid hydrophobic peptides that appear to be processed from 22 to 23 amino acid precursors. These precursor peptides resemble signal peptides lacking a cognate exported protein polypeptide sequence. There is evidence that at least some of the inhibitor peptides may be processed by the Eep protease that is believed to be involved in the processing of the inducing pheromones [2]. Despite the apparent homology between the inhibitors and their peptides, their effective functions appear to be very specific. Pheromone cCF10
J.R. Chandler, G.M. Dunny / Peptides 25 (2004) 1377–1388
shows significant similarity to iAD1 and iPD1, yet it has been found that cCF10 has no significant inhibitory activity against cells carrying pAD1 or pPD1 [22]. Strangely, both cCF10 and iCF10 were found to competitively inhibit cPD1 binding to the pPD1 TraC surface binding protein at the same level as iPD1 [45]. These results imply that the function of the inhibitor may involve more than just competitive inhibition of the surface peptide binding proteins TraC and PrgZ, and may instead involve specificity at some other level, since the response of the systems to their cognate pheromones and inhibitors is still highly specific [22,24]. Whether the inhibitor peptide is imported into the cytoplasm of the cell is still speculative. Several different studies have looked at inhibitor binding to TraA, the PrgX-like intracellular negative regulator. Fujimoto and Clewell [26] found that in the case of pAD1, inhibitor binding to TraA is weak but present, and curiously it facilitated the TraA release of DNA in vitro in a similar manner as that of the pheromone. While it is possible that the inhibitor is imported, it seems very unlikely that it would activate the same intracellular response as the pheromone. Fujimoto and Clewell proposed that this binding and its effect is biologically relevant but would not occur in vivo if the inhibitor is never imported and does not exist in its processed form intracellularly. The idea that inhibitor is, therefore, not imported was proposed because their results suggested that this import would induce upregulation of the transfer genes and thus not function as an inhibitor. Their alternative hypothesis was that post-translational processing of TraA occurs that increases its specificity so that the processed form of TraA doesn’t bind inhibitor in vivo. Another formal possibility is that the in vitro binding assay employed in these studies did not reflect the biochemical activities of either TraA or the peptides in donor cells. The pPD1 TraA was also found to bind the iPD1 inhibitor. In this case, TraA was found to bind the inhibitor almost as strongly as the pheromone, and this binding interaction was even more specific than that of the pheromone with the surface pheromone binding protein TraC [45]. The effect of inhibitor binding in this system was not explored, so it is possible that in the pPD1 system the inhibitor may function as an intracellular antagonist of pheromone.
6. Control of endogenous pheromone activity by the PrgY family of proteins PrgY is the other pCF10-encoded element involved in control of endogenous pheromone. While the inhibitors control endogenous pheromone in the supernatant, the membrane protein PrgY controls endogenous pheromone activity that remains associated with the cell. A recent analysis of pheromone production and distribution in cCF10 producing E. faecalis indicated that a significant amount of cCF10 remains cell associated [11]. The average plasmid-free recipient cell appears to have twice as much cCF10 in its cell wall than the amount that is released into the supernatant
1383
[11]. When pCF10 is acquired, the concentration of cCF10 in the supernatant is not affected [43], whereas the cell wallassociated cCF10 is reduced 8-fold from that of plasmid-free recipient cells [11]. PrgY was found to be involved in this reduction of wall-associated cCF10 after acquisition of the plasmid, but its mechanism of action is far from clear. The first evidence for involvement of PrgY, and its homologue TraB in the pAD1 plasmid, in control of pheromone activity were the observations that mutations in these genes resulted in a constitutively clumpy phenotype [13,30]. Interestingly, the amount of pheromone present in the medium of pAD1- or pPD1-containing cells is reduced as compared with plasmid-free recipient cells, and subsequent analysis of the effect of TraB on supernatant pheromone levels indicated that this protein is responsible for this effect [1,46]. These results led to the conclusion that TraB is a “pheromone shutdown” protein that somehow blocks pheromone expression or degrades active pheromone once a cell acquires plasmid. This is clearly not the function of PrgY, since the supernatant pheromone concentration is not reduced in pCF10 containing donor cells. The disparate functions of PrgY and the TraB proteins is curious, especially considering their high sequence homology and conserved location in the plasmid (see Fig. 3A). In addition, the pPD1 TraB protein can complement a prgY mutation in pCF10 (E. Bryan, J.R. Chandler and G.M. Dunny, unpublished results), indicating that at least some function is shared between the proteins. It is possible that the TraB and PrgY proteins share a common primary function that is unrelated to pheromone shut-down. The pheromone shut-down effect by pAD1 was found to vary depending on the strain [41], and if the primary function of TraB is to shut-down pheromone production, then it is unclear how the variable pheromone levels in the different strains would be accounted for by the plasmid system. Also, work on pAD1 regulation by Muscholl-Silberhorn et al. [39] indicated that although TraB partially shuts off pheromone production, the pheromone binding protein TraC is the most important for this shut off, irrespective of the presence of TraB. Whether or not the PrgY and the TraB proteins share a primary function, the mechanism of endogenous control by these proteins is still a complete mystery. It is possible that the TraB proteins and/or PrgY function as a membrane protease that degrades pheromone on its way out of the cell or trapped in the cell wall (see Fig. 4). This could potentially occur through a PrgY interaction with the membrane protease Eep that is involved in processing pheromone during export. It is also possible that PrgY functions through an interaction with the pheromone binding protein PrgZ, since PrgZ is involved in pheromone uptake and has also been found to be co-transcribed with PrgY (B.A. Buttaro and G.M. Dunny, unpublished results). PrgY may somehow regulate PrgZ so that it no longer detects self-cCF10 but is still able to detect exogenous cCF10. Although, there is no evidence for a plasmid-mediated modification of self-cCF10, PrgY could do this through a physical relocation of PrgZ. In this way PrgZ could only bind to extracellular, and not cell-wall associated,
1384
J.R. Chandler, G.M. Dunny / Peptides 25 (2004) 1377–1388
Fig. 3. Genetic organization of the regulatory region of pCF10, pPD1 and pAD1. PrgY has significant similarity to the TraB proteins in pPD1 and pAD1. (A) Diagram of the PrgY and TraB operons in pCF10, pPD1 and pAD1. TraC has significant similarity to PrgZ. Numbers below each gene represent the predicted size in kDa of the protein product and the percent identical residues of the protein product compared with the PrgY protein. iCF10, ipd and iad represent the genes encoding the inhibitor peptides. The genes encoding aggregation substance are downstream of the inhibitor peptides. (B) A multiple sequence alignment of PrgY with the TraB genes from pPD1 and pAD1. Putative transmembrane domains are highlighted in grey. Residues that are conserved within the entire PrgY family of proteins are in bold.
pheromone. Although this does not explain the increase in cell wall pheromone that is seen in a PrgY mutant [11], it is a distinct possibility. PrgY may alternatively play a role in increasing the sensitivity of PrgZ to the slight increase in cCF10 concentration that occurs in the presence of a nearby recipient cell. Under all circumstances, pCF10-containing donor cells have a static pheromone concentration, both in the cell wall and in the supernatant. It is very possible that PrgY interacts with PrgZ so that it remains unresponsive to this static concentration of cCF10, but still highly sensitive and responsive to a change in concentration brought about by a nearby recipient cell. In this model, PrgY associates with PrgZ so that its response
to the same peptide signal differs depending on the change in peptide concentration, similar to the adaptation response in bacterial chemotaxis. In chemotaxis, the chemoattractant receptor proteins are believed to exist as a heterogeneous population with differing affinities for chemoattractant [9], so that a highly sensitive response to a slight change in attractant concentration occurs. Given what is known about the Opp system and import of pheromone peptides in the E. faecalis plasmid systems, it is likely that PrgY would impart this type of regulation through a precise control of the PrgZ binding kinetics, rather than a non-specific reduction of binding or import, so that the high sensitivity required for responding to a recipient cell can still be upheld. For example, PrgY
J.R. Chandler, G.M. Dunny / Peptides 25 (2004) 1377–1388
1385
Fig. 4. Model of endogenous pheromone control by PrgY. PrgY may regulate endogenous induction a number of ways, including interference with pheromone export, sequestration or degradation of pheromone in the cell wall, or by an interaction with PrgZ. Interference with export may occur through an interaction with Eep or at another processing or export site (steps 1 and 2). Alternatively, endogenous pheromone may be blocked from signaling in the cell wall by PrgY through sequestration or degradation (step 3) but this must occur in such a way that the pheromone level in the supernatant is not affected. PrgY may block signaling by endogenous pheromone through an interaction with PrgZ involving either a physical relocation of PrgZ (step 4), or by regulating the PrgZ response through a potential conformational change so that it responds only to a change in pheromone concentration introduced by a nearby recipient cell (step 5). See the text for more details.
may restrict the conversion of PrgZ from the E1 or an ‘open’ (freeform) conformation to the E2 , or ‘closed’ (docked and ready for import) conformation so that this conversion only occurs specifically in response to increased levels of cCF 10 binding. Although the specificity of certain molecules for distinct forms of a binding protein have been described [18,27], the regulated response of the binding protein to the type or concentration of peptide has not to our knowledge been explored. A question that remains is whether iCF10 binding or import is affected by PrgY. In experiments done with a pCF10 derivative that does not express PrgY, which is constitutively clumpy, exogenous addition of iCF 10 did not restore the non-clumpy phenotype of a wild-type plasmid [11], implicating that the mechanisms of these two regulatory elements are completely separate. However, if the function of PrgY effectively alters the function of iCF10, then adding more iCF10 would not alleviate the problem. PrgY may cause PrgZ to respond differently to the two peptides, or block import of iCF10 to prevent aberrant signaling by this molecule. It is interesting to note that in the absence of iCF10 expression, exogenous cCF10 is still required for the induction of PrgB expression that confers a clumpy phenotype [7]. This indicates that iCF10, which is expressed from the same promoter as prgB [8], does not function in cis. This data also indicates the importance of PrgY in controlling the transition from the uninduced (non-clumpy) to induced (clumpy) state, whether or not iCF10 is present. It has been found that addition of the pPD1 inhibitor iPD1 has been found to inhibit the constitutively induced phenotype of the TraB mu-
tant [46], although the significance of this finding has not been explored. Whether PrgY and the TraB proteins are involved in some sort of pheromone sequestration or degradation or whether they are involved in an interaction with PrgZ has yet to be elucidated. It is also possible that PrgY, and perhaps the TraB proteins, play some other as yet undefined role in the physiology of the cell or interfere in some other way with the re-internalization of endogenous pheromone. Interestingly, other PrgY-like proteins have been identified through recent genome sequencing efforts (see Table 2 for a representative list). The prospect that the function of these proteins is conserved, possibly even throughout all of the animal kingdoms, is supported by conserved residues (see Fig. 3B) and transmembrane domains that are found through this family of proteins. In addition, a recent mutational analysis of PrgY has mapped point mutations that abolish function without destroying the stability of the protein to residues that are conserved throughout the entire protein family (J.R. Chandler and G.M. Dunny, unpublished results). The role of other PrgY-like proteins outside of the pheromone-inducible plasmid systems is an intriguing question. Interestingly, the bacterial species that have been found to encode a PrgY-like protein do not have characterized peptide-signaling systems and are quite distantly related to E. faecalis. The most closely related homologues were found in Oceanobacillus and Borrelia species. The most surprising of the species that encode a PrgY-like protein are the spirochetes. It is also surprising that more closely related species that have extensively characterized peptide-signaling systems, such as
1386
J.R. Chandler, G.M. Dunny / Peptides 25 (2004) 1377–1388
Table 2 Representatives of PrgY-like proteins throughout the biological kingdom Genus and species
Amino acid length
Percentage identity
No. of transmembrane domainsa
E. faecalis pCF10 (PrgY) E. faecalis pPD1 (TraB) E. faecalis pAD1 (TraB) Oceanobacillus iheyensis HTE831 Borrelia burgdorferi Archaeoglobus fulgidus Arabidopsis thaliana Caenorhabditis elegans Mus musculus
383 384 388 390 404 396 371 453 422
N/A 77 44 39 35 32 32 25 28
4 4 4 4 4 4 2 1 2
a Transmembrane domains were predicted using the HMMTOP server by Tusnady and Simon [58], version 2.0 (http://www.enzim.hu/hmmtop/html/ document.html).
S. aureus and B. subtilis, do not encode any such PrgY family protein. Whatever the function of the PrgY and TraB proteins, it clearly appears to be important for a diverse range of organisms. The role this protein may play in these other species is entirely unknown. The PrgY-like proteins in other species could regulate peptide-signaling systems that have not yet been revealed. To date, the only peptide-signaling system identified in a Gram-negative organism is that of the cyanobacterium Anabaena [61], which is not among the organisms that encode a PrgY-like protein. Alternatively, the mechanism of the PrgY-like proteins may not be specific for peptide-signaling systems but may instead regulate a peptide uptake system or some type of intracellular signaling system. Clearly the finding that PrgY is not unique to the pheromone plasmid systems further deepens the mystery surrounding the role of this protein in regulation of pheromone induction.
7. Conclusion The pheromone plasmids are a fascinating example of a highly evolved peptide-based system for cell–cell communication. As mentioned previously, pCF10 is only one of a family of conjugative pheromone-responsive plasmids. The size of these plasmids varies from 36 to 107 kb, but to the extent that they have been analyzed they are surprisingly similar in their genetic makeup and organization. The regulatory regions of some of these plasmids have been particularly well characterized (see Fig. 3) because of their importance in plasmid spread and the unique mechanism of the gene products. The PrgX family proteins and the inhibitor peptides appear to be unique to these plasmid systems, although the sequence homology of the PrgX family proteins is low. There are several other peptide-signaling systems that have been characterized in bacteria, including competence development in B. subtilis and virulence regulation in S. aureus (for recent reviews on these subjects see [14,23,34,35,48]). Like the pheromone-inducible conjugation system of E. faecalis, these systems involve an induced cellular response to a critical concentration of pheromone that is present in the extracellular medium. The pheromone-
inducible conjugation system of E. faecalis is unique, however, in that two cell types, the donor and the recipient, are communicating rather than a uniform population of cells. For example, a plasmid-free cell secretes pheromone but is not induced by it, even at a high cell density, while the conjugative functions of a nearby plasmid-containing donor cell are induced. Similarly, both plasmid-containing and plasmid-free cells secrete the same pheromone, but a plasmid-containing cell is carefully regulated so that it only responds when pheromone secreted by a nearby recipient cell is detected. For reasons discussed in the previous sections, the inhibitor peptides and PrgY family regulatory proteins appear to play an important role in regulating this sensitive response. Although these proteins are not unique to the conjugative plasmids, they are not present on the chromosome of the sequenced E. faecalis V583 strain [15] and thus far they have not been found on the chromosomes of any other bacterial species with a well characterized peptide-inducible signaling system. The specificity of pheromone induction is also a novel feature of the pheromone-inducible conjugative plasmids. Clinical strains often carry several different pheromone plasmids, but the degree of cross-induction of the response systems is very low. To ensure such high specificity there are clearly several distinct control circuits that ensure this specific pheromone response. These include the specific pheromone binding and import that occurs at the cell surface, the interaction of pheromone with regulatory proteins in the cell cytoplasm, the regulated expression of transfer functions associated with intracellular pheromone, and the DNA processing associated with transfer of the plasmid. This high specificity that is involved in pheromone response is another fascinating mechanism that has evolved by these systems and is crucial to ensure propagation of these plasmids with a minimal expenditure of host cell metabolites.
Acknowledgments The authors thank all of the members of the Dunny laboratory for providing much of the data on which this paper is based. This research is supported by NIH grant no. GM49530.
J.R. Chandler, G.M. Dunny / Peptides 25 (2004) 1377–1388
J.R.C. is a trainee funded by the NIH MinnCResT training grant no. T32 DE07288. [19]
References
[20] [21]
[1] An FY, Clewell DB. Characterization of the determinant (traB) encoding sex pheromone shutdown by the hemolysin/bacteriocin plasmid pAD1 in Enterococcus faecalis. Plasmid 1994;31:215–21. [2] An FY, Clewell DB. Identification of the cAD1 sex pheromone precursor in Enterococcus faecalis. J Bacteriol 2002;184:1880–7. [3] An FY, Sulavik MC, Clewell DB. Identification and characterization of a determinant (eep) on the Enterococcus faecalis chromosome that is involved in production of the peptide sex pheromone cAD1. J Bacteriol 1999;181:5915–21. [4] Antiporta MH, Dunny GM. ccfA, the genetic determinant for the cCF10 peptide pheromone in Enterococcus faecalis OG1RF. J Bacteriol 2002;184:1155–62. [5] Bae T, Clerc-Bardin S, Dunny GM. Analysis of expression of prgX, a key negative regulator of the transfer of the Enterococcus faecalis pheromone-inducible plasmid pCF10. J Mol Biol 2000;297:861–75. [6] Bensing BA, Dunny GM. Pheromone-inducible expression of an aggregation protein in Enterococcus faecalis requires interaction of a plasmid-encoded RNA with components of the ribosome. Mol Microbiol 1997;24:295–308. [7] Bensing BA, Manias DA, Dunny GM. Pheromone cCF10 and plasmid pCF10-encoded regulatory molecules act post-transcriptionally to activate expression of downstream conjugation functions. Mol Microbiol 1997;24:285–94. [8] Bensing BA, Meyer BJ, Dunny GM. Sensitive detection of bacterial transcription initiation sites and differentiation from RNA processing sites in the pheromone-induced plasmid transfer system of Enterococcus faecalis. Proc Natl Acad Sci USA 1996;93:7794–9. [9] Bornhorst JA, Falke JJ. Quantitative analysis of aspartate receptor signaling complex reveals that the homogeneous two-state model is inadequate: development of a heterogeneous two-state model. J Mol Biol 2003;326:1597–614. [10] Brown MS, Ye J, Rawson RB, Goldstein JL. Regulated intramembrane proteolysis: a control mechanism conserved from bacteria to humans. Cell 2000;100:391–8. [11] Buttaro BA, Antiporta MH, Dunny GM. Cell-associated pheromone peptide (cCF10) production and pheromone inhibition in Enterococcus faecalis. J Bacteriol 2000;182:4926–33. [12] Chow JW, Thal LA, Perri MB, Vazquez JA, Donabedian SM, Clewell DB, et al. Plasmid-associated hemolysin and aggregation substance production contribute to virulence in experimental enterococcal endocarditis. Antimicrob Agents Chemother 1993;37:2474–7. [13] Christie PJ, Dunny GM. Identification of regions of the Streptococcus faecalis plasmid pCF-10 that encode antibiotic resistance and pheromone response functions. Plasmid 1986;15:230–41. [14] Clewell DB. Sex pheromone systems in enterococci. In: Dunny GM, Winans SC, editors. Cell–cell signaling in bacteria. Washington, DC: ASM Press; 1999. p. 47–65. [15] Clewell DB, An FY, Flannagan SE, Antiporta M, Dunny GM. Enterococcal sex pheromone precursors are part of signal sequences for surface lipoproteins. Mol Microbiol 2000;35:246–7. [16] Clewell DB, An FY, White BA, Gawron-Burke C. Streptococcus faecalis sex pheromone (cAM373) also produced by Staphylococcus aureus and identification of a conjugative transposon (Tn918). J Bacteriol 1985;162:1212–20. [17] Clewell DB, Dunny GM. Conjugation and genetic exchange in enterococci. In: Gilmore MS, editor. The enterococci: pathogenesis, molecular biology, and antibiotic resistance. Washington, DC: ASM Press; 2002. p. 265–300. [18] Detmers FJ, Kunji ER, Lanfermeijer FC, Poolman B, Konings WN. Kinetics and specificity of peptide uptake by the oligopeptide trans-
[22]
[23] [24]
[25]
[26]
[27]
[28] [29]
[30]
[31]
[32]
[33]
[34] [35]
[36]
[37]
[38]
1387
port system of Lactococcus lactis. Biochemistry 1998;37:16671– 9. Detmers FJ, Lanfermeijer FC, Poolman B. Peptides and ATP binding cassette peptide transporters. Res Microbiol 2001;152:245–58. Dev IK, Ray PH. Signal peptidases and signal peptide hydrolases. J Bioenerg Biomembr 1990;22:271–90. Dunny G, Funk C, Adsit J. Direct stimulation of the transfer of antibiotic resistance by sex pheromones in Streptococcus faecalis. Plasmid 1981;6:270–8. Dunny GM, Antiporta MH, Hirt H. Peptide pheromone-induced transfer of plasmid pCF10 in Enterococcus faecalis: probing the genetic and molecular basis for specificity of the pheromone response. Peptides 2001;22:1529–39. Dunny GM, Leonard BA. Cell–cell communication in Gram-positive bacteria. Annu Rev Microbiol 1997;51:527–64. Ehrenfeld EE, Kessler RE, Clewell DB. Identification of pheromoneinduced surface proteins in Streptococcus faecalis and evidence of a role for lipoteichoic acid in formation of mating aggregates. J Bacteriol 1986;168:6–12. Flannagan SE, Clewell DB. Identification and characterization of genes encoding sex pheromone cAM373 activity in Enterococcus faecalis and Staphylococcus aureus. Mol Microbiol 2002;44:803–17. Fujimoto S, Clewell DB. Regulation of the pAD1 sex pheromone response of Enterococcus faecalis by direct interaction between the cAD1 peptide mating signal and the negatively regulating DNAbinding TraA protein. Proc Natl Acad Sci USA 1998;95:6430–5. Hall JA, Ganesan AK, Chen J, Nikaido H. Two modes of ligand binding in maltose-binding protein of Escherichia coli. Functional significance in active transport. J Biol Chem 1997;272:17615–22. Higgins CF. ABC transporters: from microorganisms to man. Annu Rev Cell Biol 1992;8:67–113. Hirt H, Schlievert PM, Dunny GM. In vivo induction of virulence and antibiotic resistance transfer in Enterococcus faecalis mediated by the sex pheromone-sensing system of pCF10. Infect Immun 2002;70:716–23. Ike Y, Clewell DB. Genetic analysis of the pAD1 pheromone response in Streptococcus faecalis, using transposon Tn917 as an insertional mutagen. J Bacteriol 1984;158:777–83. Kao SM, Olmsted SB, Viksnins AS, Gallo JC, Dunny GM. Molecular and genetic analysis of a region of plasmid pCF10 containing positive control genes and structural genes encoding surface proteins involved in pheromone-inducible conjugation in Enterococcus faecalis. J Bacteriol 1991;173:7650–64. Kitada C, Fujino M, Mori M, Sakagami Y, Isogai A, Suzuki A, et al. Synthesis and structure/activity relationships of Streptococcus faecalis sex pheromones cPD1 and cAD1. In: Izumiya N, editor. Peptide chemistry. Osaka: Protein Research Foundation; 1985. p. 43–8. Lanfermeijer FC, Detmers FJ, Konings WN, Poolman B. On the binding mechanism of the peptide receptor of the oligopeptide transport system of Lactococcus lactis. EMBO J 2000;19:3649–56. Lazazzera BA. The intracellular function of extracellular signaling peptides. Peptides 2001;22:1519–27. Lazazzera BA, Palmer T, Quisel J, Grossman AD, Dunny A. Celldensity control of gene expression and development in Bacillus subtilis. Washington, DC: ASM Press; 1999. Leonard BA, Podbielski A, Hedberg PJ, Dunny GM. Enterococcus faecalis pheromone binding protein, PrgZ, recruits a chromosomal oligopeptide permease system to import sex pheromone cCF10 for induction of conjugation. Proc Natl Acad Sci USA 1996;93:260–4. McCormick JK, Hirt H, Dunny GM, Schlievert PM. Pathogenic mechanisms of enterococcal endocarditis. Curr Infect Dis Rep 2000;2:315–21. Mori M, Sakagami Y, Ishii Y, Isogai A, Kitada C, Fujino M, et al. Structure of cCF10, a peptide sex pheromone which induces conjugative transfer of the Streptococcus faecalis tetracycline resistance plasmid, pCF10. J Biol Chem 1988;263:14574–8.
1388
J.R. Chandler, G.M. Dunny / Peptides 25 (2004) 1377–1388
[39] Muscholl-Silberhorn A, Rozdzinski E, Wirth R. Aggregation substance of Enterococcus faecalis: a multifunctional adhesin. Adv Exp Med Biol 2000;485:75–83. [40] Muscholl-Silberhorn A, Samberger E, Wirth R. Why does Staphylococcus aureus secrete an Enterococcus faecalis-specific pheromone? FEMS Microbiol Lett 1997;157:261–6. [41] Nakayama J, Dunny GM, Clewell DB, Suzuki A. Quantitative analysis for pheromone inhibitor and pheromone shutdown in Enterococcus faecalis. Dev Biol Stand 1995;85:35–8. [42] Nakayama J, Ono Y, Suzuki A. Isolation and structure of the sex pheromone inhibitor, iAM373, of Enterococcus faecalis. Biosci Biotechnol Biochem 1995;59:1358–9. [43] Nakayama J, Ruhfel RE, Dunny GM, Isogai A, Suzuki A. The prgQ gene of the Enterococcus faecalis tetracycline resistance plasmid pCF10 encodes a peptide inhibitor, iCF10. J Bacteriol 1994;176:7405–8. [44] Nakayama J, Suzuki A. Genetic analysis of plasmid-specific pheromone signaling encoded by pPD1 in Enterococcus faecalis. Biosci Biotechnol Biochem 1997;61:1796–9. [45] Nakayama J, Takanami Y, Horii T, Sakuda S, Suzuki A. Molecular mechanism of peptide-specific pheromone signaling in Enterococcus faecalis: functions of pheromone receptor TraA and pheromone-binding protein TraC encoded by plasmid pPD1. J Bacteriol 1998;180:449–56. [46] Nakayama J, Yoshida K, Kobayashi H, Isogai A, Clewell DB, Suzuki A. Cloning and characterization of a region of Enterococcus faecalis plasmid pPD1 encoding pheromone inhibitor (ipd), pheromone sensitivity (traC), and pheromone shutdown (traB) genes. J Bacteriol 1995;177:5567–73. [47] Nes IF, Eijsink GH. Regulation of group II peptide bacteriocin synthesis by quorum-sensing mechanisms. In: Dunny GM, Winans SC, editors. Cell–cell signaling in bacteria. Washington, DC: ASM Press; 1999. p. 175–92. [48] Novick RP. Regulation of pathogenicity in Staphylococcus aureus by a peptide-based density-sensing system. In: Dunny GM, Winans SC, editors. Cell–cell signaling in bacteria. Washington, DC: ASM Press; 1999. p. 129–46. [49] Paulsen IT, Banerjei L, Myers GS, et al. Role of mobile DNA in the evolution of vancomycin-resistant Enterococcus faecalis. Science 2003;299:2071–4.
[50] Quiocho FA, Ledvina PS. Atomic structure and specificity of bacterial periplasmic receptors for active transport and chemotaxis: variation of common themes. Mol Microbiol 1996;20:17– 25. [51] Ruhfel RE, Manias DA, Dunny GM. Cloning and characterization of a region of the Enterococcus faecalis conjugative plasmid, pCF10, encoding a sex pheromone-binding function. J Bacteriol 1993;175:5253–9. [52] Samuelson JC, Chen M, Jiang F, Moller I, Wiedmann M, Kuhn A, et al. YidC mediates membrane protein insertion in bacteria. Nature 2000;406:637–41. [53] Schlievert PM, Gahr PJ, Assimacopoulos AP, Dinges MM, Stoehr JA, Harmala JW, et al. Aggregation and binding substances enhance pathogenicity in rabbit models of Enterococcus faecalis endocarditis. Infect Immun 1998;66:218–23. [54] Sleigh SH, Seavers PR, Wilkinson AJ, Ladbury JE, Tame JR. Crystallographic and calorimetric analysis of peptide binding to OppA protein. J Mol Biol 1999;291:393–415. [55] Tame JR, Murshudov GN, Dodson EJ, Neil TK, Dodson GG, Higgins CF, et al. The structural basis of sequence-independent peptide binding by OppA protein. Science 1994;264:1578– 81. [56] Tanimoto K, An FY, Clewell DB. Characterization of the traC determinant of the Enterococcus faecalis hemolysin-bacteriocin plasmid pAD1: binding of sex pheromone. J Bacteriol 1993;175:5260– 4. [57] Tomich PK, An FY, Damle SP, Clewell DB. Plasmid-related transmissibility and multiple drug resistance in Streptococcus faecalis subsp. zymogenes strain DS16. Antimicrob Agents Chemother 1979;15:828–30. [58] Tusnady GE, Simon I. The HMMTOP transmembrane topology prediction server. Bioinformatics 2001;17:849–50. [59] Weaver KE, Clewell DB. Control of Enterococcus faecalis sex pheromone cAD1 elaboration: effects of culture aeration and pAD1 plasmid-encoded determinants. Plasmid 1991;25:177–89. [60] Wirth R. The sex pheromone system of Enterococcus faecalis. More than just a plasmid-collection mechanism? Eur J Biochem 1994;222:235–46. [61] Yoon HS, Golden JW. Heterocyst pattern formation controlled by a diffusible peptide. Science 1998;282:935–8.
