Deletion analysis of Streptococcus pneumoniae late competence genes distinguishes virulence determinants that are dependent or independent of competence induction

Abstract

The competence regulon of Streptococcus pneumoniae (pneumococcus) is crucial for genetic transformation. During competence development, the alternative sigma factor ComX is activated, which in turn, initiates transcription of 80 'late' competence genes. Interestingly, only 16 late genes are essential for genetic transformation. We hypothesized that these late genes that are dispensable for competence are beneficial to pneumococcal fitness during infection. These late genes were systematically deleted, and the resulting mutants were examined for their fitness during mouse models of bacteremia and acute pneumonia. Among these, 14 late genes were important for fitness in mice. Significantly, deletion of some late genes attenuated pneumococcal fitness to the same level in both wild-type and ComX-null genetic backgrounds, suggesting that the constitutive baseline expression of these genes was important for bacterial fitness. In contrast, some mutants were attenuated only in the wild-type genetic background but not in the ComX-null background, suggesting that specific expression of these genes during competence state contributed to pneumococcal fitness. Increased virulence during competence state was partially caused by the induction of allolytic enzymes that enhanced pneumolysin release. These results distinguish the role of basal expression versus competence induction in virulence functions encoded by ComX-regulated late competence genes.

Fig. 1. ComX is important for bacteremia and acute pneumonia infections

(A) CD-1 mice (n=5) were intraperitoneally challenged with a suspension containing a 1:1 mixture of D39 and each mutant strain (1 × 10⁴ CFU/strain). After 24 hr, spleens were harvested, homogenized, serially diluted and plated onto THB agar plates with or without kanamycin to determine the competitive indexes (CI). Each dot indicates the CI from one mouse. Bar shows the mean CI. ^*p < 0.05 when comparing the CI of ΔcomX1ΔcomX2 against D39 using the GraphPad Prism statistical software. (B) CD 1 mice (n = 5) were intranasally challenged with 1:1 mixture of D39 and each mutant strain (5 × 10⁶ CFU/strain). After 48 hr, lungs were harvested for CI determination. ^*p < 0.05 when comparing the CI of ΔcomX1ΔcomX2 against D39 using the GraphPad Prism statistical software.

Fig. 2

ComX-regulated “late” competence gene loci important for host infection. Nineteen operons that contained the 54 late competence genes were non-polarly deleted by using the Janus cassette, and tested for their ability to compete against D39 in mouse model of bacteremia and acute pneumonia. Of these, deletion of 9 operons yielded mutants with reduced virulence. Subsequently, each gene within these 9 operons was deleted, and tested for virulence. Of these, 14 genes were identified to be important for infection in mouse models of bacteremia and/or acute pneumonia (See Table 1). Red boxes: Genes required for genetic transformation. Shaded boxes: Genes required for infections. “C” = combox in the promoter of each gene locus. Boxes with dotted lines: essential genes that were not included during the deletion process. spd_0974 and spd_0976 are transcribed as antisense RNA during competence.

Fig. 3. Virulence confirmation of five deletion mutants by single bacteremia and acute pneumonia infections

(A) CD-1 mice (n = 7) were intraperitoneally infected with 1 × 10⁴ of D39, ΔcomX1ΔcomX2, ΔrecA, Δspd_0981, ΔradC, or spd_0031. Mouse spleens were harvested 24 hr post-infection for bacterial enumeration. ^*p ≤ 0.05 when comparing the log10 CFU of individual mutants against D39 by using the GraphPad Prism statistical software. (B) CD-1 mice (n = 10) were intranasally infected with 5 × 10⁶ of D39, ΔcomX1ΔcomX2, ΔrecA, Δspd_0981, ΔradC, or spd_0031. Mouse lungs were harvested 48 hr post-infection for bacterial enumeration. ^*p ≤ 0.05 when comparing the log10 CFU of individual mutants against D39 by using the GraphPad Prism statistical software.

Fig. 4. Contribution of recA and its promoters to growth and to resistance to H₂O₂ stress

(A) The organization of the cinA-recA-dinF-lytA operon. p1 harbors a combox, which is active during competence. p2 is a constitutive promoter of recA. (B) The growth curves of D39 versus ΔrecA, Δp1, Δp2 and ΔrecAΔspxB. The growth experiments were performed in the THB medium in triplicates for three times. The mean ± standard deviation from one typical experiment is shown. (C) H₂O₂ sensitivity of D39 versus various mutants. 1 × 10³ CFU of each strain were seeded into 1 ml of THB supplemented with indicated concentrations of H₂O₂. Pneumococcal strains were then incubated 8 hr at 37°C with 5% CO₂. “−” indicated no growth. “+” indicated positive growth. The experiments were repeated independently three times with similar results. The results from a representative experiment are shown.

Fig. 5

Spd_1828 is truncated in pneumococcus. Protein alignment of Spd_1828 against various homologs from pneumococcal strains, and other Streptococcus and Neisseria species. DNA alignments were performed using a software program from DNA STAR. Boxed region indicate the ATP-binding domain.

Fig. 6

Δspd_0030 has growth defect in the absence of added CO₂. D39 and Δspd_0030 were grown to OD600 of 0.1, and then subjected to 10-fold serial dilutions. Each diluent (20 μl) was spotted onto the THB agar. The plates were incubated at 37°C overnight with or without 5% CO₂ and photographed. The experiments were repeated independently three times with similar results. The results from one representative experiment are shown.

Fig. 7. Allolysis genes are important for bacteremia and acute pneumonia infections

(A) CD-1 mice (n = 6) were intraperitoneally infected with 1 × 10⁴ of D39, ΔcbpD, ΔcibAB, ΔlytA, ΔcbpDΔcibAB, ΔcbpDΔlytA, ΔcibABΔlytA, or ΔcbpDΔcibABΔlytA. Mouse lungs were harvested 24 hr post-infection for bacterial enumeration. ^*p ≤ 0.05 when comparing the log₁₀ CFU of individual mutants against D39 by using the GraphPad Prism statistical software. (B) CD-1 mice (n = 7) were intranasally infected with 5 × 10⁶ of pneumococcal strains described in A. Mouse lungs were harvested 48 hr post-infection for bacterial enumeration.^*p ≤ 0.05 when comparing the log₁₀ CFU of individual mutants against D39 by using the GraphPad Prism statistical software.

Fig. 8. A factor that regulates exit from competence development is important for virulence. Analysis of ΔdprA competitiveness against D39 or ΔcomB, as well as ΔcoiA and ΔssbB competitiveness against D39

(A) CD-1 mice (n = 5) were intranasally infected with 1:1 ratio (1 × 10⁴ bacteria/strain) of D39 versus ΔdprA, or ΔcomB versus ΔdprAΔcomB. After 24 hr, spleens were isolated for CI determination. Each dot indicates the CI from an individual mouse. Bar shows the mean CI. ^*p < 0.05 when comparing the CI of ΔdprA against D39 by using the GraphPad Prism statistical software. (B) CD-1 mice (n = 5) were intranasally infected with 1:1 ratio (2.5 × 10⁶ bacteria/strain) of D39 versus ΔdprA, or ΔcomB versus ΔdprAΔcomB. After 48 hr, lungs were harvested for CI determination. Each dot indicates the CI from an individual mouse. Bar shows the mean CI. ^*p < 0.05 when comparing the CI of ΔdprA against D39 by using the GraphPad Prism statistical software.

Fig. 9

In vivo competition assays differentiate the importance of basal versus competence-induced expression in genes required for bacteremia infection in CD-1 mice (n = 5). Competition assays of late competence gene mutants in D39 and comX-null genetic backgrounds. Columns 1–2: D39 versus ΔcbpD, ΔcomX1ΔcomX2 vs. ΔcbpDΔcomX1ΔcomX2; columns 3–4: D39 vs. ΔcibAB, ΔcomX1ΔcomX2 vs. ΔcibABΔcomX1ΔcomX2; columns 5–6: D39 vs. Δp1, ΔcomX1ΔcomX2 vs. Δp1ΔcomX1ΔcomX2; columns 7–8: D39 vs. ΔlytA, ΔcomX1ΔcomX2 vs. ΔlytAΔcomX1ΔcomX2; columns 9–10: D39 vs. ΔradC, ΔcomX1ΔcomX2 vs. ΔradCΔcomX1ΔcomX2; columns 11–12: D39 vs. Δspd_0030, ΔcomX1ΔcomX2 vs. Δspd_0030ΔcomX1ΔcomX2; columns 13–14: D39 vs. Δspd_0031, ΔcomX1ΔcomX2 vs. Δspd_0031ΔcomX1ΔcomX2; columns 15–16: D39 vs. Δspd_0981, ΔcomX1ΔcomX2 vs. Δspd_0981ΔcomX1ΔcomX2; columns 17–18: D39 vs. Δspd_1778, ΔcomX1ΔcomX2 vs. Δspd_1778ΔcomX1ΔcomX2; columns 19–20: D39 vs. Δspd_1828, ΔcomX1ΔcomX2 vs. Δspd_1828ΔcomX1ΔcomX2. ^*p < 0.05 when comparing the CI of each mutant derived from D39 against D39 versus the CI of each mutant derived from ΔcomX1ΔcomX2 against ΔcomX1ΔcomX2 by using the GraphPad Prism statistical software.

Fig. 10

Competence-dependent virulence is partially mediated through the enhanced release of pneumolysin. (A) Hemolytic analysis of cell-free culture supernatants from the wild-type D39, Δply, ΔcomX1ΔcomX2, ΔdprA, ΔlytA, ΔcbpD, ΔcibAB, and Δp1. Hemolytic assays were independently performed six times in triplicates. Results from one typical experiment is shown. (B) Western blot analysis of concentrated culture supernatants from (A) by using an anti-pneumolysin antibody. Experiments were performed independently three times. Westren blots from one typical experiment are shown. Densitometry analysis of pneumolysin release can be found in Fig. S1. (C) In vivo bacteremia infection in CD-1 mice (n = 5 – 9) between D39 against ΔlytA, ΔcbpD, and ΔcibAB, and between Δply against ΔplyΔlytA, ΔplyΔcbpD, and ΔplyΔcibAB. ^*p < 0.05 when comparing the CI of ΔcbpD and ΔcibAB against D39 versus the CI of ΔcbpDΔply and ΔcibABΔply against Δply by using the GraphPad Prism statistical software.