Functional studies of E. faecalis RNase J2 and its role in virulence and fitness

Post-transcriptional control provides bacterial pathogens a method by which they can rapidly adapt to environmental change. Dual exo- and endonucleolytic activities of RNase J enzymes contribute to Gram-positive RNA processing and decay. First discovered in Bacillus subtilis, RNase J1 plays a key role in mRNA maturation and degradation, while the function of the paralogue RNase J2 is largely unknown. Previously, we discovered that deletion of the Enterococcus faecalis rnjB gene significantly attenuates expression of a major virulence factor involved in enterococcal pathogenesis, the Ebp pili. In this work, we demonstrate that E. faecalis rnjB encodes an active RNase J2, and that the ribonuclease activity of RNase J2 is required for regulation of Ebp pili. To further investigate how rnjB affects E. faecalis gene expression on a global scale, we compared transcriptomes of the E. faecalis strain OG1RF with its isogenic rnjB deletion mutant (ΔrnjB). In addition to Ebp pili regulation, previously demonstrated to have a profound effect on the ability of E. faecalis to form biofilm or establish infection, we identified that rnjB regulates the expression of several other genes involved in bacterial virulence and fitness, including gls24 (a virulence factor important in stress response). We further demonstrated that the E. faecalis RNase J2 deletion mutant is more sensitive to bile salt and greatly attenuated in in vivo organ infection as determined by an IV-sublethal challenge infection mouse model, indicating that E. faecalis RNase J2 plays an important role in E. faecalis virulence.


Introduction
Overwhelming evidence supports the importance of RNA metabolism, including RNA degradation and processing, in post-transcriptional gene regulation processes in many bacterial species [1,2]. In the Gram-negative bacterium Escherichia coli, it is well studied that endo-and exonucleases, including three main rate-determining enzymes RNase E, RppH, and PcnB, are involved in RNA degradation and processing [3][4][5]. In comparison, the enzymes that regulate RNA degradation and processing in Gram-positive bacteria have been studied largely in a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 by monitoring stress response and virulence of the mutant in a bile salt resistance assay and a mouse sublethal challenge model respectively.

Bacterial strains, media, chemicals and primers
The strains used in this study were OG1RF, its isogenic deletion mutant ΔrnjB [20], and insertion mutant TX10100 (gls24 disruption) [21]. Brain heart infusion (BHI) broth and tryptic soy broth without glucose (TSB) were purchased from Difco Laboratories (Detroit, MI). All chemicals were purchased from Sigma (St. Louis, MO). Oligonucleotides used in this study were purchased from Sigma (The Woodlands, TX) and are listed in Table 1.

Immuno-electron microscopy
Enterococci were grown in BHI broth, washed in 0.1 M NaCl solution and resuspended in PBS. Monoclonal antibody mAb 9 against EbpA was developed in house using recombinant EbpA protein as immunogen [22] following standard hybridoma generation methods previously described [20]. Immunogold labeling was performed using mAb 69 against EbpC [20], mAb 9 against EbpA, or mAb70 against Ace [23] for primary binding, followed by gold beadconjugated donkey anti-mouse IgG, using methods previously described [24]. Samples were viewed on a Jeol 1400 transmission electron microscope.

Cloning and mutagenesis
The 1689 bp DNA fragment containing the E. faecalis rnjB gene was PCR amplified from the OG1RF genome using primer pair 1185NdeF and 1185XhoR and inserted into vector pQE30 for protein expression. Point mutations of rnjB were introduced by site-directed mutagenesis using QuikChange mutagenesis kit (Agilent Technologies, La Jolla, CA) following the manufacturer's protocol. Wild type or mutant gene fragment containing the rnjB gene and its E. faecalis rnjA (ef2924) was PCR amplified from the OG1RF genome using primer pair 2924BamF and 2924KpnR.

Protein overexpression and purification
Recombinant His-tagged E. faecalis RNase J1, J2 and J2 mutants were overexpressed in E. coli BL21 and purified using Ni-NTA agarose (Qiagen, Valencia, CA) following the manufacturer's protocol.

RNase degradation assays
The exonuclease activities of purified recombinant RNase J1, J2 and the RNase J2 mutants were analyzed by RT-FeDEx assay as described [25]. Briefly, a 30 nt RNA was labeled with a carboxyfluorescein (FAM) group at its 3 0 -end and hybridized to a 17 nt DNA bearing a 5 0quenching group carboxymethylrhodamine (TAMRA). Degradation of the FAM-labeled RNA substrate releases the fluorophore from the proximity of its quencher and the rate of fluorescence accumulation is measured by a spectrofluorimeter in real time. Kinetic properties of wild-type RNase J1 and J2 were determined using 5 nM for RNase J1 and 200 nM for RNase J2, with substrate RNA concentrations ranging from 0.125 to 2 μM. For comparison of J1, wild-type and mutant J2 exonuclease activities, 5 nM of J1 and 200 nM of each J2 protein with a substrate RNA concentration of 250 μM were used for each assay.

Surface plasmon resonance (SPR) analysis
Protein-protein interactions of RNase J1/ J2 and J2 mutants were determined by surface plasmon resonance (SPR) on a Biacore T100 as previously described [11]. E. faecalis RNase J1 was immobilized on a CM5 sensor surface. K D values for these interactions were determined with various concentrations of J2 proteins (400 to 50 nM) run in duplicate and were fitted to the 1:1 binding model using BiaEvaluation software.

Complementation analysis
pMSP3535 vectors with wild type or rnjB derivatives with point mutations were transformed into the E. faecalis rnjB deletion mutant. Nisin was added to the culture medium at 25ng/ml to induce expression of the rnjB gene. The presence of EbpC on the cell surface after nisin induction was determined by flow cytometry as described previously [20].

Transcriptome analysis
The microarray slides printed with 3,160 primers for E. faecalis V583 ORFs were used for transcriptome analysis as described previously [26]. Cells grown in BHI to mid-exponential phase were suspended to an OD of 1 and used for RNA extraction. RNAs from three independent cultures for each strain were extracted from using NucleoSpin RNA II kit (Macherey-Nagel, Bethlehem, PA) according to the protocol provided by the supplier. Each RNA preparation was used in two separate dye swap hybridizations (one with parent-Cy3/mutant-Cy5 and the other with parent-Cy5/mutant-Cy3) by labelling with Cy3 or Cy5 during cDNA synthesis from total RNA using Superscript II Kit (Invitrogen) following manufacturer's instructions. Hybridized microarray slides were scanned (GenePix Pro 5.0; Axon Instruments, Inc.) with independent excitation of the fluorophores Cy3 and Cy5 at 10nm resolution. The signal and background fluorescence intensities were calculated for each DNA spot using the segmentation method of the GENPIXPRO software (Molecular Devices Corp., Union City, CA). For each open reading frame (ORF), ratios of OG1RF to ΔrnjB mutant were calculated ratios by averaging ratios for spots in all 6 chips. [27].

qRT-PCR
Cells grown in BHI to early-log phase, mid-log phase and stationary phase were suspended to an OD of 1 and used for RNA extraction. Total RNA was prepared using NucleoSpin RNAII kit (Macherey-Nagel, Bethlehem, PA) and reverse-transcribed to cDNA using the SuperScript II reverse transcriptase and random primers (Invitrogen, Carlsbad, CA) following the method provided by the manufacturer. Quantitative PCR on cDNA was performed using SYBR green PCR master mix kit and an ABI7900 Real-time PCR system (Invitrogen, Carlsbad, CA). The expression of 23S rRNA and pyrR was analyzed using primer pairs listed in Table 1. For each primer set, a reference curve was established using the genomic DNA purified from wild-type OG1RF cells. The amounts of gene transcripts (ng/ml) obtained for pyrR were normalized to the 23S rRNA transcript.

Extraction of Gls24 and Western blot analysis
Membrane-associated proteins were extracted from E. faecalis OG1RF and the rnjB deletion mutant using Zwittergent as described previously with minor modifications [28]. Briefly, E. faecalis cells grown to exponential phase were harvested and washed with PBS, and resuspended in 1% of the original culture volume in PBS with 0.2% Zwittergent 3-12 (Calbiochem, La Jolla, Calif.). The suspension was incubated at 25˚C for 1 h and dialyzing overnight at 4˚C against PBS. Sample was then concentrated by lyophilization, and aliquots stored at -70˚C. Equal amounts of total protein from the Zwittergent extracts were loaded on 4%-12% NuPAGE Novex Bis-Tris Gels (Invitrogen, Carlsbad, CA) under reducing conditions in MOPS buffer and transferred to an Immobilon-P PVDF membrane (Millipore, Billerica, MA) according to the manufacturer's protocol. Membranes were then probed with mouse anti-Gls24 mAb generated in-house against recombinant Gls24 using standard protocols previously described [20], followed by HRP-conjugated goat anti-mouse IgG antibody and developed using TMB substrate system (KPL, Gaithersburg, MD).

Bile salts resistance assay
The resistance of E. faecalis OG1RF and its mutants to bile salts was determined using an assay described previously [21]. One O.D. equivalent of overnight cultured cells were harvested and resuspended in 1 mL BHI broth, then diluted 50x in BHI broth with 0.3% bile salts (Fisher, Nazareth, PA)) and incubated at 37˚C for 30 min before plating on BHI-agar plates for CFU evaluation. The relative percent survival was determined by comparing the survival rates of mutants (CFU at 30min/CFU at time 0) to that of wild-type OG1RF.

IV sublethal challenge model
The intravenous challenge experiment was modified from a previously established model [29]. This model is part of a protocol approved by the Animal Welfare Committee, University of Texas Health Science Center at Houston, Houston, TX, USA. In this model, 1x10 8 bacterial cells of wild-type or ΔrnjB strains were injected into mice via the tail vein. After a period of 48 hours, the spleen and kidneys were collected from each mouse, homogenized and serially diluted before plating onto BHI agar plates with rifampicin. Colony growth was enumerated and CFU per gram of organ weight was calculated.

rnjB is required for normal growth and Ebp pili expression of Enterococcus faecalis OG1RF
While deletion of rnjB has no known phenotype in B. subtilis, it is lethal in Streptococcus pyogenes [8,30]. In our previous study [20], we constructed an E. faecalis OG1RF isogenic rnjB deletion strain. To determine the effect of the deletion of rnjB on E. faecalis growth, we compared the in vitro growth of the wild-type strain OG1RF and ΔrnjB. When grown in BHI broth, the ΔrnjB mutant had a delayed doubling time to that of wild type and demonstrated a lower saturation density (Fig 1A, 65 min vs. 52 min). However, both wild type and mutant demonstrated all phases of growth, and reached stationary phase around 5 hours. Essentially the same growth profiles were seen in TSBG, another rich medium (data not shown). We conclude that, unlike in other species, deletion of the E. faecalis rnjB gene is not lethal but does decrease the overall growth rate.
Beyond differences in growth rate or saturation density, the ΔrnjB mutant also showed phenotypic differences in surface protein production compared to wild type. We have previously demonstrated that surface expression of EbpC is attenuated in ΔrnjB via flow cytometry and Western blot analysis. As shown in Fig 1B using immune-electron microscopy, Ebp pili were broadly distributed on the cell walls distal to the division plane for the wild-type cells. In comparison, EbpC labeling was absent on the ΔrnjB mutant, demonstrating that deletion of rnjB has a profound effect on Ebp pilus surface display. In EM analysis using an in-house generated anti-EbpA monoclonal antibody (21,22) (Fig 1C), we demonstrated that the surface expression of another Ebp pilin component, EbpA, also could not be detected on the surface of the rnjB deletion mutant, further demonstrating that rnjB affects production of the Ebp pilin. To study whether rnjB specifically affects Ebp pilin or is broadly regulating cell-wall anchored protein display, we examined the surface expression levels in wild-type OG1RF and ΔrnjB of another E. faecalis LPXTG anchored protein, Ace. As shown in Fig 1D, no significant difference was observed in Ace surface expression between OG1RF and ΔrnjB as determined by TEM. These results suggest that rnjB regulation of surface protein expression is not general to surface anchored proteins, but rather Ebp pilin specific. Unlike B. subtilis, this clear phenotype of E. faecalis rnjB may offer complementary opportunities to understand gene function.

E. faecalis rnjB encodes an active ribonuclease
Sequence analysis revealed that the proteins coded by E. faecalis rnjA (ef2924) and rnjB (ef1185) are orthologs of the B. subtilis RNase J1 and J2, respectively. Both B. subtilis enzymes display endoribonuclease activity, while the exo-ribonuclease activity of J2 is much weaker than that of J1 [10]. In an effort to study E. faecalis RNase Js' enzymatic activity, we cloned Due to the high concentration of enzyme required for this assay, we could not accurately measure Kcat and K M for RNase J2. However, RT-FeDex assay clearly demonstrated that RNase J2 is an active enzyme with much lower exonuclease activity as compared to J1 (Fig 2B). For Crystal structures of RNase J from T. thermophilus and B. subtilis J1 revealed residues that are involved in catalytic zinc ion coordination and RNA recognition [31,32]. In E. faecalis RNase J2, three of these conserved residues (His 69, His 71 and Asp 73) confirmed by multiple sequence alignments of E. faecalis RNase J2 with diverse bacteria and archaea were selected for structure-function studies. Amino acid substitution variants of RNase J2 at these three conserved residues (H69A, H71A and D73A) were introduced by site-directed mutagenesis, and the exo-ribonuclease activity of the three point-mutation proteins were determined by RT-Fe-DEx assay described above. When comparing the exonuclease activities of wild-type and mutant RNase J2s, the wild-type enzyme and one of the mutations (D73A) exhibited enzymatic activities while the other two mutants H69A and H71A were inactive (Fig 2C). SPR analysis of E. faecalis RNase J1 and J2 interaction (Fig 2D) demonstrated that wild-type RNase J2 and the two inactive mutants (H69A and H71A) bind to E. faecalis RNase J1 in vitro. The K D s of these interactions were determined to be 74 nM for the J1-J2 interaction, 81 nM for J1-J2H69A, and 89 nM for J1-J2H71A, all similar to that previously reported in B. subtilis J1J2 interaction (K D of 80 nM) [32], suggesting that the ability of RNase J1 and J2 to form heteromers is conserved between Bacillus and Enterococcus. The interactions between E. faecalis RNase J1 and RNase J2 inactive mutants were comparable to that of the wild-type RNase J2, suggesting that the disruption of RNase J2 active sites does not affect the interaction between J1 and J2.

E. faecalis RNase J2 activity is required for Ebp pilus production
We previously demonstrated that E. faecalis rnjB positively regulates the abundance of ebp transcript. To test whether the ribonuclease activity is related to the function of RNase J2 in Ebp pilin gene regulation, we examined production of Ebp pili on the rnjB deletion mutant strain complemented with wild-type and mutant RNase J2 genes. As shown in Fig 3, the introduction of the wild-type RNase J2 gene was able to restore pilus expression while mutant genes encoding inactive RNase J2 enzymes failed to do so. This suggests that the ribonuclease activity of RNase J2 is important for its function in pilin regulation.

Specific transcripts of E. faecalis affected by absence of RNase J2
To globally characterize the role of RNase J2 in gene expression, we compared the transcriptome of ΔrnjB and OG1RF grown in BHI to exponential phase using whole genome DNA microarray analysis. Table 2 lists genes whose transcripts were influenced by deletion of rnjB. In summary, compared to OG1RF, mRNA transcripts of 36 genes from the deletion mutant demonstrated a 3-fold decrease with P value less than 0.05, or were part of an operon in which at least one member achieved this level of differentiation. In addition, compared to OG1RF, there were 26 mRNA transcripts in the mutant which demonstrated a 2 -fold increase with a P value less than 0.05, or were part of a gene cluster in which at least one member achieved this level of differentiation. The dataset of this microarray study was deposited in Gene Expression Omnibus (GEO) with the series accession number GSE95005.
Among the genes that are down-regulated in ΔrnjB are the 5 genes in the ebp clusters, (ebpR, A, B, C and srtC), which is consistent with previous qRT-PCR results [20], and adjacent genes ef1096 and ef1097, the latter encoding a putative antimicrobial peptide [33]. In addition, the virulence related gene cluster, gls24 operon (ef0077-ef0080), is significantly down regulated in ΔrnjB (Table 2).
Gls24 (EF0079), first identified as a general stress protein involved in glucose starvation response and bile-salts resistance, also has been shown to play an important role in E. faecalis virulence, with a gls24 mutant significantly attenuated in establishing infection in a mouse peritonitis model. [13,21,22,34]. For further exploration of the down-regulation of gls24 in our transcriptome analysis, we performed Western blot analysis to assess the level of Gls24 expression in an RNase J2 deletion mutant by probing with a Gls24 specific monoclonal antibody we generated using techniques previously described [35,36]. We found that the Gls24 expression level is significantly down regulated in the RNase J2 deletion mutant (Fig 4), consistent with observed decreases in transcriptome analysis.
In contrast to genes down regulated in ΔrnjB mentioned above, nine genes of the pyrimidine nucleotide biosynthesis (Pyr) operon, ef1712-1720, pyrP,B,Aa,Ab,C,DII,D-2,F,E, demonstrated 3-8 fold increases in expression levels in ΔrnjB compared to wild type. In Bacillus subtilis the pyr operon is regulated by the first gene of the operon, pyrR, through an autogenous transcriptional attenuation mechanism [37]. However, the E. faecalis pyrR (ef1721) gene was not represented on the microarray chip, preventing analysis of whether pyrR transcript is affected by rnjB. To investigate the role of rnjB on pyrR transcription, we tested the effect of ΔrnjB on RNA levels of pyrR by qRT-PCR. As shown in Fig 5, mRNA levels of the pyrR gene in ΔrnjB were 2, 15 and 7.8 -fold higher than those of wild type in early log, mid-log and stationary phase cultures respectively. This result, together with the transcriptome analysis, demonstrated that rnjB affects the whole pyr operon, likely through the regulation of pyrR transcription.

RNase J2 plays important role in E. faecalis fitness and virulence
The transcriptome analysis revealed that the rnjB gene is required for the expression of multiple virulence-related gene loci, including ebpABC, gls24, and ef1097, a putative antimicrobial peptide [33]. We have previously shown that ΔrnjB is attenuated in biofilm formation analyzed by a microplate-binding adherence assay, which correlates with decreased Ebp pili surface display. Gls24, on the other hand, has been reported to be important for E. faecalis bile salts resistance and virulence [21]. To test whether deletion of rnjB has an effect on E. faecalis stress resistance, a bile-salt resistance assay was performed on ΔrnjB, Δgls24, and wild-type cells. As demonstrated in Fig 6, the gls24 deletion strain showed significant reduction in bile-salt resistance. In contrast, the rnjB deletion showed modestly reduced resistance to the bile salts compared to wild type (45%) with a P score of 0.03. This result is consistent with the results of The change represents mRNA expression levels in OG1RF relative to those in the ΔrnjB mutant and corresponds to averages of experiments done with three independent cultures. Minus indicates that the expression was lower than in the ΔrnjB mutant than in the wild type. c The P value from a one-sample t test, testing whether the grand mean log e ratio was different from 0.0, was significant at the 0.05 or better level for each of the genes tested.
To evaluate the impact of RNase J2 on bacterial virulence in vivo, we utilized a sublethal challenge model in mice [29] to study the capacity of wild type vs. ΔrnjB E. faecalis to establish infection. Two rounds of independent experiments were performed and the combined results are shown in Fig 7. Mice infected with ΔrnjB (n = 16) show a 50-fold reduction in mean CFUs in the spleens (ΔrnjB 1.2 x10 3 vs. wild type 6.8x10 4 CFU/gm) and a 400-fold reduction kidneys in BHI medium containing 0.3% bile salt relative to that of wild-type OG1RF was calculated as described in the Materials and Methods section. Each experiment was performed in triplicate and the experiments were repeated 3 times. Error bar represent the standard error of these repeats.
(ΔrnjB 2.2x10 3 vs. wild type 9.3x10 5 CFU/gm) compared to mice infected with wild type bacteria (n = 16), demonstrating a significant decrease in the ability of the mutant to establish infection (Fig 7). While these results are consistent with the observed inhibition of virulence factor expression in ΔrnjB data presented, the described growth delay of the ΔrnjB is also a potential factor in overall attenuation. Nonetheless, in totality, these observations are further supportive evidence of rnjB playing an important role in Enterococci fitness.

Discussion
Since discovered over a decade ago [8], many studies have demonstrated an important role for RNase J1 in RNA degradation and maturation. In comparison, the function of RNase J2 has remained largely unknown. Due to lack of an identified substrate and lower enzymatic activity compared to that of RNase J1, it has been speculated that RNase J2 is functionally a duplication of J1 and acts as a substitute or possible regulator of J1 [10,11]. However, our study of E. faecalis RNase J2 suggests that the enzyme may have a very specific role in regulating biological cues for fitness and virulence: First, the ribonuclease activity of RNase J2 is required for Ebp pilus production, suggesting the involvement of RNA degradation or maturation in the process. Second, deletion of RNase J2 in E. faecalis specifically altered the RNA abundance of 62 genes. Many of these genes are clustered and thus only approximately 20 distinct loci are affected, including several virulence and fitness related genes clusters, indicating specificity of RNase J2 function. Third, among the affected genes are several genes with known roles in gene regulation. For example, PyrR functions as an antiterminator that acts on the leader of the Pyr operon. Thus the effect of RNase J2 on the PyrP, B, Aa, Ab, C, D, F and E genes is likely explained by an effect on PyrR. This role of RNase J2 in antitermination resembles that of B. subtilis RNase J1 in thrS and trp operon post-transcription attenuation [38,39]. Similarly, The EbpABC operon is regulated by EbpR and the effect of RNase J2 on EbpABC is likely through EbpR. Therefore, we suspect that RNase J2 directly acts on a small number of mRNAs and/or sRNAs that in turn affect the 62 genes we identified.
The exonuclease activity of B. subtilis RNase J2 is much lower than RNase J1 [11]. This difference between J1 and J2 activities suggests that J1 and J2 have distinct functions in RNA processing. Crystal structure of B. subtilis RNase J1 reveals that the Zn 2+ ion in the J1 catalytic site is coordinated by H74, H76, H142 and D164, while the other Zn 2+ ion is coordinated by residues D78, H79, D164 and H390 [32]. These residues that coordinate the second Zn 2+ ion are not conserved in RNase J2s, suggesting a different active site structure. We reported in this Functions of Enterococcus faecalis RNase J2 manuscript that the exonuclease activity of E. faecalis RNase J2 is also much lower than its J1 paralog. As shown in Fig 2A, the three residues we chose for amino acid substitution (H69, H71 and D73) are comparable to H74, H76 and D78 in B. subtilis RNase J1. Mutants of H69 and H71 (equivalent to H74, H76 of the B. subtilis RNase J1, respectively) significantly reduced E. faecalis RNase J2 activity, suggesting that these two residues are involved in Zn 2+ ion coordination. Mutation of D73(equivalent to D78 of the B. subtilis RNase J1), however, resulted in a mild increase of enzyme activity, suggesting that this residue is not involved in Zn 2+ ion coordination in J2, further confirming structural differences between RNase J1 and J2. It has been reported that the B. subtilis RNase J1 and J2 can form a complex with altered enzymatic properties compared to the individual enzymes [11], suggesting a role of RNase J2 as a regulator of J1 through protein-protein interaction. In our study, the inactive mutants of RNase J2 failed to restore Ebp pilus expression, yet the protein-protein interaction with RNase J1 was not affected. This indicates that RNase J2 activity, although much weaker compared with the RNase J1, has unique function in E. faecalis either through altering RNase J1 activity or independent of RNase J1 activity.
The most interesting observation from the transcriptome analysis is that the rnjB deletion mutant showed decreased expression of several known virulence factors including the Ebp proteins and Gls24. Disruption of the ebp operon has previously been shown to significantly reduce E. faecalis biofilm formation and the ability to form vegetation in a rat endocarditis model [22]. Gls24 is an important E. faecalis stress responder. A gls24 deletion strain has reduced bile-salt resistance [21]. Additionally, gene disruption experiments have shown that an E. faecalis mutant lacking a functional gls24 gene was highly attenuated in a mouse peritoneal lethal challenge model [21] and that disruption decreased virulence in an experimental rat endocarditis model [40]. In our studies of fitness and virulence, we must acknowledge the difficulty in distinguishing attenuation due to a change in pathogenicity factors from those that might be related to the slower growth of ΔrnjB mutant seen in vitro studies. However, the decrease in virulence in concert with the observed decreased expression of several virulence factors in the RNase J2 deletion mutant, supports a link between RNase J2 and bacterial virulence. This in turn could offer a novel means with which to elucidate regulation mechanisms and design strategies to manipulate virulence factor production. It's also worthwhile to point out that the Fsr regulon is down-regulated in the rnjB deletion mutant by microarray analysis. However, we did not observe the same results in either early-log phase or late phase cultured cells using qRT-PCR analysis. In addition, a gelatin plate assay using overnight culture of OG1RF and ΔrnjB showed no difference in end-point gelatinase production (data not shown), further suggesting that the Fsr quorum signal in the ΔrnjB mutant is not abolished. Although it has been reported that the S. aureus Agr system (Fsr analog) plays an important role in RNA regulation, whether E. faecalis rnjB is involved in Fsr regulation requires further investigation. Altogether, due to the importance of the virulence factors that are affected by rnjB in bacterial infection, it can be viewed as promising targets for novel antimicrobial agents. More broadly, the involvement of RNase J2 in bacterial virulence should be further explored in other Grampositive species, including S. pyogenes and S. aureus.