Biofilm formation during pneumococcal carriage imprints naturally acquired humoral immunity

Abstract

Streptococcus pneumoniae (Spn) colonization of the nasopharynx is a prerequisite for transmission and invasive disease. To investigate how repeated asymptomatic colonization shapes immunity and influences bacterial traits, we developed the Repeated Asymptomatic Murine Pneumococcal Colonization (RAMPC₃) model using strains belonging to serotypes: 2 (D39), 3 (WU2), and 4 (TIGR4). Sequential colonization revealed strain- and exposure-order-dependent effects on bacterial burden, with initial colonization yielding robust carriage and subsequent exposures resulting in diminished burden and rapid clearance. Humoral profiling demonstrated antigenic imprinting: the first colonizing strain largely determined IgG and IgA specificity, with minimal diversification after repeated exposures. Reactivity was strongest for biofilm-associated antigens correlating with each strain's biofilm-forming capacity. Using TIGR4 mutants deficient in biofilm formation, we confirmed that in vivo aggregate formation drives humoral responses. Human sera from naturally colonized adults mirrored these findings, favoring biofilm antigens independent from capsule. Protection was demonstrated as triple-colonized mice exhibited reduced mortality and bacteremia following pneumococcal pneumonia challenge. Moreover, the initial colonizing strain influenced protection against heterologous infection, underscoring the lasting imprint of the biofilm phenotype on immunity. Finally, IgA responses in nasal-associated lymphoid tissue paralleled serum IgA patterns, validating systemic measurements as a proxy for mucosal immunity. Collectively, these results reveal that biofilm formation during colonization is a key determinant of humoral immunity and protection, providing insight into pneumococcal biology and informing strategies to design next-generation interventions.

Fig 1.. Spn burden in a murine repeated asymptomatic colonization model depends on strain and order exposure.

(A) Schematic of the Repeated Asymptomatic Murine Pneumococcal Colonization (RAMPC₃) model (see methods). 9-week-old C57BL/6J male and female mice were intranasally inoculated with 10⁶ CFU/mL of WU2 (Cohort A) or TIGR4 (Cohort B) for the first colonization event, followed by D39 for the second, and the final and third event with TIGR4 or WU2, respectively. (B) Bacterial burden was determined over a 2-week period post-inoculation by colony forming units (CFUs) obtained from nasal washes with saline. (C) Bacterial burden for each strain at Day 3 and Day 59 or Day 31. Each dot is one mouse sample. N=40-45 per group. Not detected (ND) ≤ 10² CFU/mL. Mann-Whitney t-test and median with 95% confidence interval (CI). *=p ≤0.0332; ****=p≤ 0.0001.

Fig 2.. The first colonization event imprints a mucosal and systemic antibody response to the proteins that persist following repeated colonization.

(A) Equal amounts of whole bacterial cell lysates grown planktonically (P) or in a biofilm (BF) from three Spn strains WU2 (serotype 3), D39 (serotype 2), and TIGR4 (serotype 4) were analyzed by immunoblot. Membranes were probed individually with mouse sera (1:1000) from Cohort A and Cohort B RAMPC₃ mice after the first, second, and third colonization events and secondary α-mouse IgG (1:10000). Representative blots shown. (B) The number of new protein antigens detected by IgG on immunoblots from Cohort A and Cohort B RAMPC₃ mice after the first, second, and third colonization events. N=15-16. One-way ANOVA and mean with standard deviation. **=p≤ 0.002; ****=p≤ 0.0001.

Fig 3.. Spn protein array identifies specific antigens following repeated colonization.

A pneumococcal protein array was constructed with 254 highly antigenic proteins. Proteins were selected from a panel of Spn strains and were conserved for recognition by IgG from healthy human adults (Croucher et al. 2017). Sera from RAMPC₃ colonized mice after the first and third colonization events were used to probe the protein array (1:100) (see methods). (A) Both cohorts and timepoints combined. (B) Cohorts and timepoints expanded. N=14. Mean with standard deviation.

Fig 4.. Repeated murine Spn colonization elicits a strain-dependent humoral response associated with biofilm formation.

Spn lab strain’s (WU2, D39, TIGR4) ability to form biofilms as measured by crystal violet assay (see methods). N=8-10. Equal amounts of whole bacterial cell lysates grown (B) planktonically (P) or in a (C) biofilm (BF) from three Spn strains WU2 (serotype 3), D39 (serotype 2), and TIGR4 (serotype 4) were run on ELISAs and individually probed with serum (1:1000) from RAMPC₃ mice in both Cohort A and Cohort B after the third colonization event. Secondary antibody α-mouse IgG (1:10000). Each dot is one mouse sample. N=24-30. (D) Linear regression correlation between ability of Spn strains to form biofilms and a ratio of antibody recognition to biofilm antigens. Each dot is one mouse sample. N=10-12. (E) Three isogenic TIGR4 mutant’s (ΔpsrP, ΔcbpA, ΔspxB) ability to form biofilms as measured by crystal violet assay (see methods). N=6. (F) Equal amounts of whole bacterial cell lysates grown planktonically (P) or in a biofilm (BF) from TIGR4 and three isogenic TIGR4 mutants were run on ELISAs and individually probed with serum (1:1000) from mice colonized once with each strain for 21 days. TIGR4 mutants are pooled together. Secondary antibody α-mouse IgG (1:10000). Each dot is one mouse. N=2. (G) Linear regression correlation between ability of Spn TIGR4 mutant strains to form biofilms and a ratio of antibody recognition to biofilm antigens. One-way ANOVA or Mann-Whitney t-test and mean with standard deviation. *=p≤0.0332; **=p≤ 0.002; ***=p ≤0.0002; ****=p≤ 0.0001.

Fig 5.. Serum antibodies from asymptomatic colonized human adults recognize Spn antigens depending on strain.

Equal amounts of whole bacterial cell lysates (WCL) grown planktonically (P) or in a biofilm (BF) from (A, B) three Spn lab strains WU2 (serotype 3), D39 (serotype 2), and TIGR4 (serotype 4) and (C, D) corresponding serotype clinical isolates were run on ELISAs and individually probed with serum (1:1000) from asymptomatically colonized adults (aged 40-82) and secondary antibody α-human IgA and IgG (1:10000). Each dot is one human sample. N=17. (E, F) Spn lab strains (WU2, D39, TIGR4) and their corresponding clinical isolate’s ability to form biofilms as measured by crystal violet assay (see methods). N=8-10. (G) Linear regression correlation between ability of lab Spn strains and their (H) corresponding clinical isolates to form biofilms and antibody recognition to biofilm antigens. Mann-Whitney t-test, One-way ANOVA, and mean with standard deviation. *=p≤0.0332; **=p≤ 0.002; ***=p ≤0.0002; ****=p≤ 0.0001.

Fig 6.. Repeated asymptomatic colonization with Spn protects against pneumococcal pneumonia.

A cohort of age-matched female naïve mice and the RAMPC₃ mice from both Cohorts A and B were intratracheally challenged with 10⁵ CFU/mL of Spn strain 6A-10 (serotype 6A) (see methods). (A) Biofilm formation for D39, WU2, TIGR4, and 6A-10 strains as determined by crystal violet assay. (B) Survival over time and (C) bacterial burden in the blood 48 hours post-infection (hpi) was recorded. Not detected (ND) ≤ 10² CFU/mL. Kaplan-Meier survival curve with log-rank test. One-way ANOVA and mean with standard deviation or standard error of the mean (SEM). *=p≤0.0332; ****=p≤ 0.0001.

Fig 7.. Protection against pneumococcal pneumonia is dependent on the colonizing Spn strain.

9-week-old C57BL/6J female mice were intranasally inoculated with 10⁶ CFU/mL of (A) WU2 or (B) TIGR4. After one month, mice were intratracheally challenged with 10⁵ CFU/mL of a different Spn strain: WU2 (serotype 3), D39 (serotype 2), and TIGR4 (serotype 4). Survival was recorded. N=5-6 per group. Kaplan-Meier survival curve with log-rank test. (C, D, and E) Equal amounts of whole bacterial cell lysates grown planktonically (P) or in a biofilm (BF) from three Spn strains (WU2, TIGR4, or D39) were run on ELISAs and were individually probed with mouse sera (1:1000) from surviving mice before and after colonization and infection and secondary α-mouse IgG (1:10000). Each dot is one mouse sample. N=2-4. Two-way ANOVA and mean with standard deviation. *=p≤0.0332; **=p≤ 0.002; ***=p ≤0.0002; ****=p≤ 0.0001.