New Insights into the Enterococcus faecalis CroRS Two-Component System Obtained Using a Differential-Display Random Arbitrarily Primed PCR Approach

RAP-PCR procedure.

E. faecalis strain JH2-2 (29) was cultured in liquid GM17 medium (22), and cells were collected during mid-exponential phase (at an optical density at 600 nm of 0.4) or entering stationary phase. Total RNA was extracted as previously described (16) and used as a template for reverse transcription with a 1st Strand cDNA synthesis kit (Roche) using the random primer Rap1 (Table 1). The resulting cDNA was purified with a QIAquick purification kit (QIAGEN) to remove Rap1 and used as a template for PCR in the presence of [α-³²P]dATP. For the first cycle (5 min at 94°C, 15 min at 36°C, and 1 min at 72°C), only one random primer, Rap2 (Table 1), was added. The reaction mixture was incubated for 5 min at 50°C, the primer Rap1 was added to the mixture, and 30 additional cycles (30 s at 94°C, 30 s at 50°C, and 1 min at 72°C) were performed. The resulting RAP-PCR products were purified, heat denatured, and analyzed with a polyacrylamide sequencing gel (18). The gel was then transferred onto 3MM Chr paper (Whatman), dried, and exposed to Kodak BioMax MS film. The resulting autoradiogram is presented in Fig. 1A.

• Open in new tab
• Download powerpoint

FIG. 1.

RAP-PCR fingerprints of RNA samples from E. faecalis. (A) RAP-PCR fingerprints of RNA samples from E. faecalis JH2-2 strain. RNA was extracted from exponentially growing cells (lane 1) and from cells harvested at the beginning of the stationary phase (lane 2). The RAP-PCR was performed with primer Rap1 for the first cDNA strand synthesis and Rap2 for the second cDNA strand synthesis. (B and C) RAP-PCR fingerprints of RNA samples from exponentially growing cells of the E. faecalis JH2-2 wild-type strain (lanes 1) and of the err05 mutant (lanes 2). In panel B, the first cDNA strand was obtained using primer Rap3 and the second cDNA strand using Rap4, while in panel C, the primer order was inverted (Rap4 for the first cDNA strand synthesis and Rap3 for the second cDNA strand synthesis). Arrows indicate differentially expressed RAP-PCR products.

RAP-PCR methods usually use only one random primer or a mixture of primers to create both strands of the cDNAs (1, 19, 26). Identification of the RAP-PCR products then requires the additional step of cloning the corresponding cDNA before sequencing. In this work, by using two distinct primers, one for the synthesis of the first strand of cDNA and the other for the second strand, the RAP-PCR procedure allowed us to bypass the cloning step. After precisely aligning the RAP-PCR gel and the autoradiogram, gel plugs containing the differentially amplified RAP-PCR products were excised and resuspended in Tris-EDTA buffer, and this mixture was used as a template for PCR amplification using primers Rap1 and Rap2. The resulting PCR products were purified and used as templates for DNA sequencing using the primer Rap1.

Identification of growth-dependent-expressed genes.

The gel autoradiogram presented in Fig. 1A depicts four differentially amplified RAP-PCR products: the RAP-PCR product A1 was more intensely amplified with RNA extracted from cells entering stationary phase, whereas the three other RAP-PCR products, A2, A3, and A4, were more intensely amplified with RNA from exponentially growing cells. The RAP-PCR product A1 was identified as an internal region of the pyrAb gene (NCBI locus EF_1716 [14]). As part of the 10-cistron operon encoding enzymes for pyrimidine biosynthesis, pyrAb encodes the large subunit of the carbamoyl-phosphate synthase. RAP-PCR product A3 corresponds to an internal fragment of the rplC gene (NCBI locus EF_0206 [14]) encoding the ribosomal protein L3. Northern hybridizations confirmed the pyr operon induction and the rplC repression during stationary phase (data not shown). These results are similar to those previously observed with a different gram-positive bacterium, Bacillus subtilis, which showed pyr operon induction and rplC repression during stationary phase (4). RAP-PCR products A2 and A4 were identified as internal segments of croS and croR, respectively, two genes organized as a bicistronic operon encoding the CroRS two-component system. CroR is an OmpR response regulator able to bind to DNA. This regulatory system has been shown to be involved in control of the salB gene expression (10, 12). SalB is a secreted protein first identified as a major antigen expressed during human endocarditis infection (28) and also shown to be involved in E. faecalis cell shape maintenance and stress response (9, 10, 12) and in cell adherence and biofilm formation (11). Northern hybridizations confirmed the higher level of croRS mRNA during mid-exponential growth phase than during the stationary phase (data not shown). croRS expression has been previously tested with exponentially growing E. faecalis cells subjected to physicochemical treatments (10), and no obvious differences were observed under these conditions. This suggested that CroRS function was not essential for the cell's response to the presence of bile salts, hyperosmotic challenge, acid stress, and hypo- or hyperthermic shock. Our results show that the croRS operon is repressed when cells enter the stationary phase, implying that its function might be more important during exponential growth than during stationary phase. To gain a better comprehension of this two-component system, we next examined the nature of the genes under CroRS control.

RAP-PCR for the analysis of the CroRS regulon.

Two additional RAP-PCR experiments were performed using RNA from exponentially growing cells of E. faecalis wild-type strain JH2-2 and from the mutant strain err05 (10), which is unable to express CroR. For the first assay, we used the primer Rap3 (Table 1) for the synthesis of the first strand of cDNA and the primer Rap4 (Table 1) for the second strand, whereas we inverted the order of these two primers (i.e., use of Rap4 for the synthesis of the first cDNA strand and Rap3 for the second cDNA strand) for the second RAP-PCR. This led to the observation of four RAP-PCR products, B1, B2, B3 (Fig. 1B), and C1 (Fig. 1C). These products were amplified to a higher level by using RNA extracted from the err05 mutant than by using RNA from the wild-type JH2-2 strain. The RAP-PCR products B2 and B3 were identified as internal fragments of the croR gene. Even if the CroRS system has been shown to be subject to autoregulation (3, 12), we interpret our result as the consequence of an aberrant transcription of the disrupted croR gene from the integrated plasmid in the mutant err05 strain. RAP-PCR product B1 was identified as an internal segment of a gene encoding a 243-amino-acid (aa) protein homologous to ATP-binding subunits of ABC transporters (data not shown) such as GlnQ, the ATP-binding subunit of the Bacillus stearothermophilus Gln/Glu ABC transporter (59% identity; 79% homology) (27). The gene was therefore renamed glnQ. RAP-PCR product C1 corresponds to a gene that encodes an 825-aa protein homologous to amino acid binding proteins of ABC transporters (data not shown) such as GlnH, the Gln-binding subunit of the Bacillus subtilis Gln/Glu ABC transporter (56% identity; 73% homology) (17). The gene was therefore renamed glnH.

CroR is directly involved in the regulation of the gln operon.

The genes glnQ and glnH are located at the same locus on the E. faecalis chromosome (Fig. 2A). Genes encoding ABC transporter components are usually organized as an operon. Downstream of glnH, we found two other genes encoding putative permease subunits of an ABC transporter. These two genes were renamed glnM and glnP according to their homologies. Chromosome sequence analysis identified an inverse repeat (IR) structure (where ΔG₀ is −34.8 kcal mol⁻¹) 173 bp downstream of glnP. This IR is directly followed by a stretch of T residues, suggesting that the IR could act as a Rho-independent transcription terminator (Fig. 2A). No such structure was found in the intergenic region between the genes glnQ, glnH, glnM, and glnP, suggesting that they might form a four-cistron operon. Northern hybridizations using a glnQ probe (PCR amplified using primers GlnQ1 and GlnQ2; Table 1) revealed a 3.1-kb transcript when using RNA extracted from the wild-type E. faecalis strain JH2-2 (data not shown). This shows that glnQ, glnH, glnM, and glnP are likely cotranscribed as an operon, hereafter referred to as the gln operon.

• Open in new tab
• Download powerpoint

FIG. 2.

The gln operon and CroR. (A) Schematic representation of the gln operon. Open reading frames are represented by large arrows; their orientation shows the transcriptional direction. The nucleotide sequence of the transcriptional initiation nucleotide (+1), determined by 5′ RACE PCR, and of the putative Rho-independent terminator are indicated. (B) Quantitative reverse transcription-PCR analysis of the expression of the gln operon. Bars represent the ratios of abundance of 23S rRNA to abundance of gln mRNA in exponentially growing cells of E. faecalis JH2-2 (1) or of the CroR-deficient err05 mutant ± 2 standard deviations. The corresponding ratio was calculated using the formula n = E_gln^CTgln/E_23S^CT23S, where E_23S and E_gln represent the efficiency of the PCR (the factor by which the number of copies for the corresponding gene [23S rRNA or the gene glnQ] increases at each cycle), and C_T represents the corresponding cycle threshold determined. (C) Analysis of the binding of CroR to the promoter Pgln. Electrophoretic mobility shift assays were performed with 0.25 ng of radiolabeled Pgln DNA, corresponding to the gln operon promoter, and different amounts of phosphorylated His₆-CroR. Specificity of the His₆-CroR binding to Pgln was tested by adding either 50× excess of unlabeled competitor (cp, corresponding to Pgln) DNA or an unlabeled no-competitor (nc [11]) DNA to the binding mixture. The positions of the free DNA (F) and the His₆-tagged CroR/DNA complexes (C1, C2, and C3) are indicated on the left.

Real-time PCR experiments performed with cDNA synthesized from RNA extracted from exponentially growing cells of E. faecalis JH2-2 and from the CroR-deficient err05 mutant and with oligonucleotides p23SF, p23SR, pglnQF, and pglnQR (Table 1) revealed that the gln transcript is 4.3 times more abundant in the err05 mutant (Fig. 2B). This result confirmed the RAP-PCR observations and suggested that CroRS might be involved in the repression of the gln operon.

Using RNA extracted from exponentially growing E. faecalis JH2-2 cells, we performed 5′ rapid amplification of cDNA ends (RACE) PCR experiments, as previously described (16), and mapped the gln transcript start site (+1) at a base A localized 108 nucleotides upstream of the start codon of the glnQ open reading frame (Fig. 2A). Gel retardation assays were then performed as previously described (12) using His₆-tagged CroR (His₆-CroR) protein and a radiolabeled DNA fragment that contains the promoter region of the gln operon (Pgln, 299 bp) from nucleotide −143 to nucleotide +156 relative to the transcriptional start point (PCR amplified with primers Pgln1 and Pgln2; Table 1). When Pgln is incubated with different amounts of phosphorylated His₆-CroR, bands with reduced mobility are observed, demonstrating that CroR binds to the Pgln promoter, since this interaction is specific (Fig. 2C). This result suggests that the CroR protein regulates transcription of the gln operon via a direct DNA binding interaction.

Conclusions.

The RAP-PCR procedure described in this work was successfully used to characterize the modifications of E. faecalis gene expression. This novel RAP-PCR approach, which uses two distinct primers for the first and second strand DNA and consequently allows the suppression of the cloning step, rapidly led to differentially amplified RAP-PCR product sequences of excellent quality. Consequently, compared to conventional RAP-PCR procedures, the present method offers the advantages of (i) a reduction of the time required for identification of differentially expressed genes and (ii) a reduction of the number of sequencing reactions, as only one sequence per RAP-PCR product is needed with our procedure, while other methods required sequencing of several clones.

We showed that during the transition from exponential growth to the beginning of the stationary phase, the pyrAb gene was induced, whereas the rplC gene and the croRS operon were repressed. The CroRS system found in E. faecalis is probably required during exponential growth as a pleiotropic regulator of genes involved in the stress response (10, 12; unpublished data) and for cell shape maintenance (12), biofilm formation (11), cell adherence (11), intrinsic β-lactam resistance (3), and glutamine/glutamate transport.

It is interesting to note that two of the identified CroR targets, salB and glnQ, were first characterized as genes encoding antigenic determinants expressed during E. faecalis human endocarditis infections (28). Future studies examining the role of the CroRS two-component system in the E. faecalis infection process seem warranted.