Search

Enterococcus faecalis is an increasingly significant pathogenic bacterium in humans, as well as a commensal member of the intestinal flora. Pheromone-responsive plasmid systems play an important role in the dissemination of antibiotic resistance among this and other species. pCF10 is a well-characterized member of this class of plasmids, and allows for transmission of tetracycline resistance in E. faecalis. The experiments described in this thesis were designed to assess the effects of three genes thought to be important for the response of plasmid pCF10 to the peptide mating pheromone cCF10 in E. faecalis: 1) the chromosomal locus responsible for pheromone production- ccfA, 2) the plasmid-encoded negative regulator prgY thought to be responsible for sequestration of endogenous pheromone at the cell surface, and 3) the plasmid-encoded positive regulator prgZ, an oligopeptide permease subunit A homolog responsible for specific import of pheromone. In this study transcription emanating from regulatory regions of this plasmid was characterized over time using reverse transcription quantitative polymerase chain reaction (RT-qPCR). Results from these assays were confirmed using phenotypic analysis and Northern blotting. Expression of ccfA was shown to be important to moderate the pheromone response. This moderation effect is believed to be the result of basal induction of donor cells, resulting in production of the peptide inhibitor iCF10 from the plasmid. As the first study to probe PrgY function at the level of transcription, RNA quantitation revealed that levels of endogenous pheromone cCF10 resulting from processing of the CcfA lipoprotein precursor are insufficient to fully induce a conjugation response in the absence of negative regulator PrgY, suggesting secondary negative regulation. Finally, PrgZ was shown to be important for regulated shutdown of the conjugation response in addition to initiation. This lends evidence to previous research suggesting specific import of inhibitor iCF10 by PrgZ. Probing the effects of these regulators over time in strains that differ in their ability to secrete pheromone allows us to paint a more complete regulatory picture of this complex and dynamic system.

Regulation of conjugation of pCF10, a pheromone-response plasmid of Enterococcus faecalis, is a well characterized process that serves as a model for the control of gene expression in bacteria by intercellular signaling. The genes encoded in the pCF10 prgQ operon mediate conjugative transfer of the plasmid in response to a peptide pheromone secreted by plasmid-free E. faecalis cells. The products of the prgX operon, the small RNA Anti-Q and PrgX, the transcriptional repressor of the prgQ promoter, negatively regulate expression of the prgQ operon. Transcription of prgX is initiated at a promoter within the prgQ operon, but oriented in the opposite direction, making transcription of the two operons overlapping and convergent. Each operon encodes a small RNA within its 5' terminus, which is complementary to 5' sequences of the opposing operon. The orientation of the operons and RNA species derived from the 5' terminus of each operon suggested that, in addition to regulation of the prgQ operon by Anti-Q, there may be unrecognized regulatory interactions between the two operons. This led me to undertake a focused study of potential transcriptional and posttranscriptional regulatory interactions between the prgQ and prgX operons. I found that the operons encode mechanisms to negatively regulate each other. Both use a small RNA to reciprocally regulate the expression of downstream genes from the other operon. Notably, each RNA acts via a different mechanism; Qs, derived from the 5' terminus of the prgQ operon, directs posttranscriptional processing of the prgX mRNA by the host factor RNase III. Anti-Q, an RNA derived from the 5' terminus of the prgX operon, negatively regulates transcription elongation of the prgQ operon without the assistance of host-encoded proteins. Additionally, I gained understanding of cis-acting mechanisms that regulate expression of the prgX operon. Specifically, transcription from the prgQ promoter represses the activity of the prgX promoter via a mechanism termed transcriptional interference, and Anti-Q sequences act as an intrinsic terminator, attenuating the expression of prgX mRNA. In addition to elucidating some of the complexity of regulation between the prgQ and prgX operons, this work revealed the high degree of functional efficiency in the RNA sequences of both Anti-Q and Qs. Anti-Q directs its own genesis by acting as an unusual factor-independent termination signal to RNA polymerase and the mature RNA regulates gene expression from the prgQ operon. Qs, for its part, fills at least three roles in the regulation of conjugation, acting as a leader sequence that can attenuate downstream gene expression, a mRNA for the prgQ gene, and a regulatory RNA that directs endonucleolytic processing of its target. Despite the fact that the overlapping sequences between prgQ and prgQ are shared by both operons, the sequences necessary for Anti-Q and Qs to mediate each of their activities are distinct. This raises the possibility that these regulatory functions evolved separately and that the combination of two ancestral regulatory pathways gave rise to the pheromone-response circuit of pCF10. This work has a number of broader implications. It deepens our understanding of how bacteria use RNAs to regulate gene expression. Additionally, the mobile genetic elements of E. faecalis contribute to the evolution of multi-drug resistant E. faecalis and other pathogens. Understanding the molecular mechanisms used by conjugative plasmids may allow the development of novel interventions to prevent this evolution or target bacteria carrying these plasmids.

The Streptococcus faecalis transposon Tn917 encodes inducible resistance to the macrolide, lincosamide, and streptogramin B antibiotics, and is inducible for transposition when exposed to low levels of erythromycin. To understand the mechanisms of induction the complete nucleotide sequence of Tn917 was determined. The element is 5,257 base pairs in length and is flanked by 38 base pair inverted repeats. Six open reading frames (ORF) were identified, all on the same DNA strand. They are designated ORF1 through ORF6 starting with the 5' end of the transposon. A number of similarities were observed between Tn917 and the Escherichia coli transposon Tn3. The inverted repeats show strong homology, the predicted amino acid sequence of ORF4 shows homology to the amino acid sequence of the Tn3 resolvase, and a region upstream of ORF4 is homologous to the res site of Tn3. The MLS resistance determinant (ORF2) and the related leader sequence (ORF1) of Tn917 is virtually identical to the corresponding region on the S. sanguis plasmid pAM77. MLS resistance has been previously shown to be regulated post-transcriptionally (translational attenuation); a similar mechanism probably regulates Tn917. Deletions within ORF5 prevent Tn917 transposition. A frameshift was introduced into ORF6 to determine whether this region is necessary for transposition. Transposons carrying this mutation could insert but were incapable of resolving. On this basis it was determined that ORF6 is necessary for a complete transposition event. A model is proposed to explain the induction of transposition. In the absence of inducing antibiotics, transcription from a promoter preceding ORF1 terminated upstream from the transposition genes. When the inducing antibiotic is present, transcription continues and allows for the expression of downstream genes. Northern blot analyses identified a 4.8 kilobase mRNA transcript from cells induced with erythromycin. Both in the presence and absence of erythromycin a small mRNA band hybridized to the probes; and, as seen with other MLS determinants, this mRNA species was enhanced in the presence of erythromycin. A region resembling a factor-independent transcription termination site is located between ORF3 and ORF4. Upstream from this potential transcription termination site are sequences homologous to those identfied in bacteriophage lambda that are essential to anti-termination.

The conjugative transfer frequency of the Streptococcus faecalis plasmid pAD1 increases 10,000-fold sex pheromone (cAD1) induction. Before the increase in plasmid transfer, donor cells synthesize a proteinaceous adhesin that facilitates the formation of mating aggregates. Four novel surface proteins appearing after exposure of pAD1-containing cells to sex pheromone were identified by Western blot analysis a 74 kilodalton (kDa) protein, the major species present, and three of higher molecular weight: 130, 153, and 157kDa . Induced cells containing another conjugative S. faecalis plasmid, pPD1, gave rise to three high molecular weight proteins of the same size (130, 153, and 157 kDa) which cross-reacted with antisera raised against induced pAD1-containing cells. However, the major protein species produced by pPD1-containing cells had a molecular weight of 78,000 and did not cross-react with the corresponding band of the pAD1 system. Pheromone-induced transfer of the two plasmids, when both were present in the same cell, was shown to be independent; induction was limited to the pheromone-specified plasmid. The possibility that lipoteichoic acid might act as a receptor (binding substance) for the induced adhesin protein was suggested by the ability of free lipoteichoic acid to inhibit specifically the clumping of induced cells. A 31.2 kb contiguous region of the plasmid was found to be necessary for the mating response. Insertions of the transposon Tn917 in two previously identified regions (traA and traB) resulted in increased transfer frequencies, while insertions in five new regions (D, E, F, G, and H) decreased the ability of pAD1 to transfer. Alterations in the normal protein profiles of induced cells were observed for cells carrying plasmids with insertions in a number of these regions. Insertions in region H and G allowed the formation of visible mating aggregates even though the plasmid transfer frequency was decreased in broth matings (to levels below detection for region H insertions). No aggregate formation was observed with insertions in regions D, E, or F; however, plasmid transfer in broth matings was observed for region F mutants. Insertions in region C decreased the cell's sensitivity to exogenous cAD1 and increased its production of the pheromone inhibitor, iAD1. A map giving the location of these regions is shown, and related aspects of the regulation of the pAD1 mating response are discussed.