Function of BriC peptide in the pneumococcal competence and virulence portfolio

Abstract

Streptococcus pneumoniae (pneumococcus) is an opportunistic pathogen that causes otitis media, sinusitis, pneumonia, meningitis and sepsis. The progression to this pathogenic lifestyle is preceded by asymptomatic colonization of the nasopharynx. This colonization is associated with biofilm formation; the competence pathway influences the structure and stability of biofilms. However, the molecules that link the competence pathway to biofilm formation are unknown. Here, we describe a new competence-induced gene, called briC, and demonstrate that its product promotes biofilm development and stimulates colonization in a murine model. We show that expression of briC is induced by the master regulator of competence, ComE. Whereas briC does not substantially influence early biofilm development on abiotic surfaces, it significantly impacts later stages of biofilm development. Specifically, briC expression leads to increases in biofilm biomass and thickness at 72h. Consistent with the role of biofilms in colonization, briC promotes nasopharyngeal colonization in the murine model. The function of BriC appears to be conserved across pneumococci, as comparative genomics reveal that briC is widespread across isolates. Surprisingly, many isolates, including strains from clinically important PMEN1 and PMEN14 lineages, which are widely associated with colonization, encode a long briC promoter. This long form captures an instance of genomic plasticity and functions as a competence-independent expression enhancer that may serve as a precocious point of entry into this otherwise competence-regulated pathway. Moreover, overexpression of briC by the long promoter fully rescues the comE-deletion induced biofilm defect in vitro, and partially in vivo. These findings indicate that BriC may bypass the influence of competence in biofilm development and that such a pathway may be active in a subset of pneumococcal lineages. In conclusion, BriC is a part of the complex molecular network that connects signaling of the competence pathway to biofilm development and colonization.

Fig 1. Expression of briC is induced by cognate CSP.

β-galactosidase assay measuring PbriC-LacZ activity in pneumococcal R6 cells grown to exponential phase in Columbia Broth at pH 6.6 followed by no treatment or treatment with CSP1 or CSP2 for 30 minutes. Y-axis denotes PbriC-lacZ expression levels in Miller Units. Activity is expressed in nmol p-nitrophenol/min/ml. Error bars represent standard error of the mean for biological replicates (at least n = 3); “ns” denotes non-significant, **** p<0.0001 using ANOVA followed by Tukey’s post-test.

Fig 2. CSP-induction of briC is ComE-dependent.

(A) Genomic organization of the briC locus, displaying a ComE-binding box. Green: ComE-binding box within the briC promoter region. The expanded region denotes a logo of ComE-binding box generated from thirty-four pneumococcal genomes represented in Fig 5. This consensus is aligned with the published ComE-binding box consensus sequence [36]. The putative -10 region, the transcription start site (TSS) as determined by Cappable-Seq [37], the ribosome binding site (RBS) and the transcriptional terminator are labeled. The downstream gene is predicted to be a pseudogene in R6D, R6 and D39. In TIGR4, this region encodes two coding sequences (SP_0430 and SP_0431). The R6D sequence corresponds to the C-terminal of SP_0430. (B) mRNA transcript levels of briC (solid black) and comE (dashed black lines) as measured by qRT-PCR in R6D WT & R6DΔcomE cells. Cells were grown in Columbia broth at pH 6.6 to an OD₆₀₀ of 0.3, and then treated with CSP1 for either 0’, 10’ or 15’. Data was normalized to 16S rRNA levels. Y-axis denotes normalized concentrations of mRNA levels in arbitrary fluorescence units as calculated from LinRegPCR. Error bars represent standard error of the mean calculated for biological replicates (n = 3); “ns” denotes non-significant, * p<0.05 using ANOVA followed by Tukey’s post-test relative to the respective 0’ CSP treatment. Further, briC levels are also significantly higher in WT relative to ΔcomE cells for the same time points post-CSP treatment (p<0.05).

Fig 3. BriC does not influence early biofilm development.

(A) Representative confocal microscopy images showing top view of the reconstructed biofilm stacks of WT, ΔbriC and ΔbriC::briC cells of strain R6D stained with SYTO59 dye at 24h. Images are pseudo-colored according to depth (scales shown). (B) COMSTAT2 quantification of 24h biofilm images. Y-axis denotes units of measurement: μm³/μm² for biomass, and μm for maximum thickness and average thickness over biomass. Error bars represent standard error of the mean calculated for biological replicates (n = 3); “ns” denotes non-significant comparisons, **** p<0.0001 using ANOVA followed by Tukey’s post-test.

Fig 4. BriC stimulates late biofilm development.

(A) Representative confocal microscopy images showing top view of the reconstructed biofilm stacks of WT, ΔbriC and ΔbriC::briC cells of strain R6D stained with SYTO59 dye at 72h. Images are pseudo-colored according to depth (scales shown). (B) COMSTAT2 quantification of 72h biofilm images. Y-axis denotes units of measurement: μm³/μm² for biomass, and μm for maximum thickness and average thickness over biomass. Error bars represent standard error of the mean calculated for biological replicates (n = 3); “ns” denotes non-significant comparisons, **** p<0.0001 using ANOVA followed by Tukey’s post-test.

Fig 5. Distribution of the genomic region encoding BriC across streptococcal strains.

Distribution of briC alleles in fifty-five streptococcal genomes. The briC alleles are visualized against a maximum likelihood tree of streptococcal genomes generated from the core genome, where the numbers on the branches represent bootstrap values. Different species in the tree are color-coded as follows: S. pneumoniae (blue), S. pseudopneumoniae (pink), S. mitis (green), S. oralis (beige), and S. infantis (grey). The shapes at the tip of the branches illustrate briC alleles. Types 1A and 1B represent variants of the alleles widespread across pneumococcal strains; types 3–5 denotes alleles outside the species. The red tick denotes strains that have a long briC promoter due to a RUP insertion. In PMEN1 strains, this variant leads to an increase in basal levels of briC in a CSP-independent manner.

Fig 6. Alignment of pneumococcal BriC alleles.

Alignment of 19 BriC alleles identified in the database of 4,034 pneumococcal genomes. Alleles are labeled 1A-1S followed by the number of representatives in the database (total 3,976). Sequences are colored based on percent identity to highlight the variability between alleles. Black arrow denotes the predicted cleavage site.

Fig 7. Long briC promoter is associated with an increase in the basal levels of briC.

β-galactosidase assay comparing the LacZ activity of the R6 (short promoter, PbriC-lacZ) and PN4595-T23 (long promoter with RUP, PbriC_long-lacZ) promoters. Both promoter activities were tested in (A) strain R6 and (B) strain PN4595-T23. Cells were grown in Columbia broth at pH 6.6 until mid-log phase, followed by either no treatment or treatment with CSP for 30 minutes. Y-axis denotes promoter activity in Miller Units expressed in nmol p-nitrophenol/min/ml. Error bars represent standard error of the mean for biological replicates (n = 3); ** p<0.01, & **** p<0.0001 using ANOVA followed by Tukey’s post-test.

Fig 8. BriC plays a pivotal role in regulating biofilm development.

(A) Representative confocal microscopy images showing top view of the reconstructed biofilm stacks of WT, ΔcomE, ΔbriC::PbriC_{Shuffled ComE-box}-briC, ΔcomE::PbriC_long-briC and WT::PbriC_long-briC cells of strain R6D stained with SYTO59 dye at 72h. Images are pseudo-colored according to depth (scale shown). (B) COMSTAT2 quantification of 72h biofilm images. Y-axis denotes units of measurement: μm³/μm² for biomass, and μm for maximum thickness and average thickness over biomass. Error bars represent standard error of the mean calculated for biological replicates (at least n = 3); “ns” denotes non-significant comparisons, and **** p<0.0001 using ANOVA followed by Tukey’s post-test.

Fig 9. ComAB plays a role in the export of BriC.

(A) Representative confocal microscopy images showing top view of the reconstructed biofilm stacks of WT, ΔcomE, ΔcomE::PbriC_long-briC and ΔcomEΔcomAB::PbriC_long-briC cells of strain R6D stained with SYTO59 dye at 72h. Images are pseudo-colored according to depth (scale shown). (B) COMSTAT2 quantification of 72h biofilm images. Y-axis denotes units of measurement: μm³/μm² for biomass, and μm for maximum thickness and average thickness over biomass. Error bars represent standard error of the mean calculated for biological replicates (at least n = 3). (C) Extracellular Nano-Glo HiBiT activity of the BriC reporter produced by WT and ΔcomAB cells (whole cells). The HiBiT activity was measured by recording luminescence with an integration time of 2000 milliseconds. Error bars represent standard deviation calculated for biological replicates (n = 3); “ns” denotes non-significant comparisons, *** p<0.001, and **** p<0.0001 using ANOVA followed by Tukey’s post-test.

Fig 10. BriC contributes to pneumococcal colonization of the mouse nasopharynx.

CD1 mice were infected intranasally with 20μl PBS containing approximately 1 X 10⁵ CFU of (A) WT (grey circles), ΔbriC (orange squares), and ΔbriC::briC (black triangles) (B) WT (grey circles), ΔcomE (blue diamonds), and ΔcomE::PbriC_long-briC (black crosses), ΔbriC::PbriC_{Shuffled ComE-box}-briC (yellow triangles), and WT::PbriC_long-briC (green circles) cells of the pneumococcal strain D39. At predetermined time points (0, 3 & 7 days post-infection), at least five mice were culled, and the pneumococcal counts in the nasopharyngeal washes were enumerated by plating on blood agar. Y-axis represents Log₁₀ counts of CFU recovered from nasal washes. X-axis represents days post-inoculation. Each data point represents the mean of data from at least five mice. Error bars show the standard error of the mean. **** p<0.0001 relative to the WT strain, and # p<0.0001 relative to the ΔcomE strain, calculated using ANOVA and Tukey post-test.