Expression of Streptococcus pneumoniae Bacteriocins Is Induced by Antibiotics via Regulatory Interplay with the Competence System

Abstract

Pneumococcal bacteriocins (pneumocins) are antibacterial toxins that mediate intra-species competition within the human host. However, the triggers of pneumocin expression are poorly understood. Using RNA-sequencing, we mapped the regulon of the pneumocin cluster (blp) of Streptococcus pneumoniae D39. Furthermore, by analogy with pneumococcal competence, we show that several antibiotics activate the blp-genes. Using real-time gene expression measurements we show that while the promoter driving expression of the two-component regulatory system blpR/H is constitutive, the remaining blp-promoters that control pneumocin expression, immunity and the inducer peptide BlpC, are pH-dependent and induced in the late exponential phase. Intriguingly, competence for genetic transformation, mediated by the paralogous ComD/E two-component quorum system, is induced by the same environmental cues. To test for interplay between these regulatory systems, we quantified the regulatory response to the addition of synthetic BlpC and competence-stimulating peptide (CSP). Supporting the idea of such interplay, we found that immediately upon addition of CSP, the blp-promoters were activated in a comD/E-dependent manner. After a delay, blp-expression was highly induced and was strictly dependent on blpRH and blpC. This raised the question of the mechanism of BlpC export, since bioinformatic analysis showed that the genes encoding the putative exporter for BlpC, blpAB, are not intact in strain D39 and most other strains. By contrast, all sequenced pneumococcal strains contain intact comAB genes, encoding the transport system for CSP. Consistent with the idea that comAB mediate BlpC export, we finally show that high-level expression of the blp-genes requires comAB. Together, our results demonstrate that regulation of pneumocin expression is intertwined with competence, explaining why certain antibiotics induce blp-expression. Antibiotic-induced pneumocin expression might therefore have unpredictable consequences on pneumococcal colonization dynamics by activating genes that mediate intra-specific interference competition.

Fig 1. The blpAB genes are frameshifted in the majority of pneumococcal genomes.

(A) Distribution of blpAB ORF lengths in a set of 4,096 genomes. Amino acid residue numbers are shown at the top. Variants found in less than 0.5% of the genomes are not shown. Frameshifted regions encoding non-BlpA ORFs are shown as diagonal lines. (B) Distribution of bacteriocins between genomes. The proportion of genomes with different numbers of blp bacteriocins for genomes with a full-length blpA (n = 1,020) is compared to genomes with an interrupted blpA (n = 3,028). No blpA sequence was detected in 48 genomes.

Fig 2. Association between blpAB length and comAB phylotypes.

The phylogenetic tree of S. pneumoniae genomes is surrounded by three rings; moving outwards, these show the ORF lengths for blpAB, and the phylotypes for comA and comB, respectively. Partial length ORFs and missing data are also shown for comA and comB. comB phylotypes found in less than 0.5% of genomes are colored grey to indicate their status as full-length variants.

Fig 3. blp regulation in S. pneumoniae D39.

(A) Schematic representation of the blp regulated genes of S. pneumoniae D39. The genes are represented by blue arrows and drawn to scale. Promoters are indicated by flags. blpA and blpB have frameshift mutations, as marked in black. Note that the ORF named pncW in this figure, encoding a putative bacteriocin upstream of blpY, is not annotated in the published genome sequence of S. pneumoniae D39. (B) Alignment of blp promoters. Nucleotides conserved in all five promoters are marked with an asterisk (*), while nucleotides conserved in 4 out of 5 promoters are marked with two dots (:). The direct repeats are shown in blue and the extended -10 box in red. (C) Schematic overview of the tripartite reporter cassette integrated in the cep-locus of S. pneumoniae D39 between treR and amiF [20]. The blp promoters (denoted P_blp in the text) are inserted upstream of the three co-transcribed reporters luc (yellow), gfp (green) and lacZ (blue). The reporter cassette is transcriptionally isolated by terminators (lollipops). (D) Natural induction of promoter P_blpT. S. pneumoniae containing a fusion of promoter P_blpT to the reporter cassette luc-gfp-lacZ, grown in C+Y pH 8. Strains with or without deletion of the regulatory genes, blpSRHC, are shown. Natural expression from promoters P_blpA, P_pncW and P_blpK are shown in S1 Fig. (E) P_blpS is active in all growth stages and expression is independent of the blp regulatory genes. S. pneumoniae D39 containing a fusion of promoter P_blpS to the reporter cassette luc-gfp-lacZ, grown in C+Y pH 8. For D and E, gene expression as measured by luciferase activity (relative luminescence units per optical density, RLU/OD) is shown on the left axis and growth as measured by absorbance at 595 nm (OD₅₉₅) is shown on the right axis. Averages of 3 replicates with the standard deviation are plotted.

Fig 4. Conditions for induction of blp expression.

(A) Expression from P_blpT is pH-dependent. Similar pH-dependent expression from promoters P_blpA, P_pncW and P_blpK are shown in S3 Fig. (B) P_blpS activity is not affected by pH. For A and B, S. pneumoniae containing P_blpT and P_blpS reporter fusions, respectively, were grown in C+Y with different pH-values. LacZ production can be observed as blue colonies on plates. (C) Expression from P_blpT is induced by exposure to sub-lethal concentrations of competence-inducing antibiotics. Similar antibiotic-induced expression from promoters P_blpA, P_pncW and P_blpK are shown in S4 Fig. (D) P_blpS activity is not significantly affected by antibiotic exposure. For C and D, S. pneumoniae containing P_blpT and P_blpS reporter fusions were grown in C+Y pH 7.4 with or without sub-lethal concentrations of antibiotics. HPUra (0.15 μg/ml), ciprofloxacin (0.4 μg/ml) and streptomycin (6 μg/ml) were added from the start of the growth curve. For all plots, gene expression as measured by luciferase activity (RLU/OD) is shown on the left axis and growth as measured by absorbance at 595 nm (OD₅₉₅) is shown on the right axis. Averages of three replicates with the standard deviation are plotted.

Fig 5. blp expression is induced by CSP.

(A-B) Competence and blp expression are both stimulated by CSP addition, although with different kinetics. Competence reporter strain P_ssbB-luc (A) and blp reporter strain P_blpT-luc (B) grown in C+Y pH 7 with or without addition of 100 ng/ml CSP after 100 min (indicated by an arrow). Similar CSP induction of promoters P_blpA, P_pncW and P_blpK are shown in S5 Fig. (C-D) CSP causes a weak but immediate activation of the blp promoters independent of blpSRHC. Reporter strains for P_blpT (C) and P_blpA (D), in which blpSRHC are deleted, grown in C+Y pH 7. (E) The immediate inducing effect is dependent on the ComCDE regulatory system. Expression from P_blpT in a strain with ΔblpSRHC and ΔcomCDE grown in C+Y pH 7 is not induced by CSP addition. For extended plots of C-E, see S5 Fig. For all plots, gene expression as measured by luciferase activity (RLU/OD) is shown on the left axis and growth as measured by absorbance at 595 nm (OD₅₉₅) is shown on the right axis. Averages of three replicates with the standard deviation are plotted. (F) blp expression is activated after competence in all cells. Time-lapse fluorescence microscopy of dual reporter strains for competence (P_ssbB-rfp) and blp expression (P_blpK-gfp) grown on C+Y agarose pH 8. PC; phase contrast.

Fig 6. Interplay between blp and com.

(A) The absence of blpAB does not abolish natural blp activation. Strains (wild type or ΔblpAB) carrying the P_blpT reporter were grown in C+Y pH 8 (similar results for P_blpA, P_pncW and P_blpK in S6 Fig). (B) Induction of blp expression in late exponential phase by CSP can only occur if comAB is intact. Strains with the P_blpT reporter were grown in C+Y pH 7.4, and CSP (100 ng/ml) was added as indicated by the arrow. (C) Natural induction of blp expression can only occur if comAB is intact. Strains with the P_blpT reporter were grown in C+Y pH 8. (D) Ectopic, inducible expression of comAB, but not comA or comB alone, activate blp expression. Strains with the P_blpT reporter were grown in C+Y pH 7.4 with 0.1 mM ZnCl₂ and 0.01 mM MnCl₂ for induction. (E) Complementation of blp expression in a ΔcomAB mutant by ectopic expression of comAB from a Zn-inducible promoter. (F) comCDE is required for blp activation in late exponential phase. Ectopic expression of comAB cannot activate blp expression when comCDE is absent suggesting that the weak activation of blpC transcription by comE is necessary for blp expression. For E and F, reporter strains were grown with or without 0.1 mM ZnCl₂ and 0.01 mM MnCl₂ in C+Y pH 7.1. For all plots, gene expression as measured by luciferase activity (RLU/OD) is shown on the left axis and growth as measured by absorbance at 595 nm (OD₅₉₅) is shown on the right axis. Averages of three replicates with the standard deviation are plotted.

Fig 7. Model for competence-mediated blp activation.

When the concentration of mature ComC (CSP) in the extracellular milieu reaches a threshold, CSP binds the membrane-located histidine kinase ComD (1), which autophosphorylates and transfers the phosphate group to activate the response regulator ComE (2). Activated ComE-P can bind to com-promoters to activate normal expression (3, black triangles), which includes comC and creates an autocatalytic loop. Expression of comX, which is an alternative sigma factor, leads to expression of the late competence genes encoding proteins involved in DNA transformation and repair. The current study suggests that activation of ComE-P might also directly induce expression of the blp promoters (4, grey triangles). This weak induction of BlpC production is necessary for triggering blp activation. BlpA’B’ is probably inoperative and pre-BlpC is instead processed and transported by ComAB (5). When the extracellular concentration of BlpC reaches a threshold, it binds to histidine kinase BlpH (6), which autophosphorylates and phosphorylates response regulator BlpR (7). Activated BlpR-P can then bind to and activate normal expression of the blp promoters (8, black triangles), which activate the genes encoding putative bacteriocins and their immunity proteins.