Contribution of Streptococcus mutans Strains with Collagen-Binding Proteins in the Presence of Serum to the Pathogenesis of Infective Endocarditis

Abstract

Streptococcus mutans, a major pathogen of dental caries, is considered one of the causative agents of infective endocarditis (IE). Recently, bacterial DNA encoding 120-kDa cell surface collagen-binding proteins (CBPs) has frequently been detected from S. mutans-positive IE patients. In addition, some of the CBP-positive S. mutans strains lacked a 190-kDa protein antigen (PA), whose absence strengthened the adhesion to and invasion of endothelial cells. The interaction between pathogenic bacteria and serum or plasma is considered an important virulence factor in developing systemic diseases; thus, we decided to analyze the pathogenesis of IE induced by S. mutans strains with different patterns of CBP and PA expression by focusing on the interaction with serum or plasma. CBP-positive (CBP⁺)/PA-negative (PA^-) strains showed prominent aggregation in the presence of human serum or plasma, which was significantly greater than that with CBP⁺/PA-positive (PA⁺) and CBP-negative (CBP^-)/PA+ strains. Aggregation of CBP⁺/PA^- strains was also observed in the presence of a high concentration of type IV collagen, a major extracellular matrix protein in serum. In addition, aggregation of CBP⁺/PA^- strains was drastically reduced when serum complement was inactivated. Furthermore, an ex vivo adherence model and an in vivo rat model of IE showed that extirpated heart valves infected with CBP⁺/PA^- strains displayed prominent bacterial mass formation, which was not observed following infection with CBP⁺/PA⁺ and CBP^-/PA⁺ strains. These results suggest that CBP⁺/PA^-S. mutans strains utilize serum to contribute to their pathogenicity in IE.

Keywords: Streptococcus mutans; blood; infective endocarditis.

FIG 1

S. mutans aggregation in the presence of serum or plasma obtained from a human volunteer. (A to C) Changes in the aggregation rates in the presence of serum (A) and plasma (B) and rates of self-aggregation (C) in the presence of S. mutans clinical strains with different patterns of CBP and PA expression at multiple time points. Times are in hours. There were significant differences between the strains, which were determined using ANOVA with the Bonferroni correction. *, P < 0.05 versus MT8148 at each time point; **, P < 0.01 versus MT8148 at each time point; ***, P < 0.001 versus MT8148 at each time point; ##, P < 0.01 versus NN2094 at each time point; ###, P < 0.001 versus NN2094 at each time point. (D) Aggregation rates in the presence of serum and plasma and self-aggregation rates for 45 S. mutans clinical strains classified by different patterns of CBP and PA expression (15 strains each displaying the CBP⁻/PA⁺, CBP⁺/PA⁺, and CBP⁺/PA⁻ phenotypes). There were significant differences between the groups, which were determined using ANOVA with the Bonferroni correction. **, P < 0.01; ***, P < 0.001. (E) Correlation between aggregation rates in serum and plasma analyzed by regression analysis. Each point represents the mean value for each bacterial strain. The colors of the symbols correspond to those in the keys to panels A to D. The bars in panels A to C represent standard deviations, and the bars in panel D represent standard errors.

FIG 2

Representative microscopic images of S. mutans strains mixed with whole blood obtained from a healthy volunteer. Arrowheads, bacterial masses. (Left) Low-magnification images; (right) high-magnification images. Bars = 50 μm (left) and 10 μm (right).

FIG 3

Aggregation of S. mutans strains in the presence of serum obtained from various subjects. (A) Representative images of S. mutans strains in the presence of serum obtained from a healthy volunteer. (Left) Macroscopic images; (middle) stereoscopic microscopic images (bars = 20 μm); (right) scanning electron microscopic images (bars = 5 μm). (B) Aggregation rates of CBP⁺/PA⁻ strains and their CBP-inactivated mutant strains in the presence of serum obtained from a healthy donor; (C) aggregation of S. mutans strains in the presence of serum obtained from five different volunteers; (D) aggregation of S. mutans strains in the presence of serum samples obtained from six different bovines; (E) aggregation of S. mutans strains in the presence of type IV collagen at various concentrations; (F) aggregation of TW295 and its mutant strains in the presence of type IV collagen at various concentrations; (G) aggregation of S. mutans strains in the presence of serum or complement-inactivated serum; (H) aggregation of TW295 and its mutant strains in the presence of serum or complement-inactivated serum. There were significant differences between the groups, which were determined using ANOVA with the Bonferroni correction. *, P < 0.05; **, P < 0.01; ***, P < 0.001. The bars in panels B to H represent standard deviations.

FIG 4

Evaluation of the ex vivo adherence model with bovine heart valve specimens. (A) Representative histopathological images following Gram staining of tissue sections of bovine heart valves infected with S. mutans strains in the presence of serum. Bars = 50 μm. (B) Rates of S. mutans adhesion to bovine heart valves in the presence of serum. There were significant differences in the rates, which were determined using ANOVA with the Bonferroni correction. **, P < 0.01 versus the TW295 group; ***, P < 0.001 versus the TW295 group; ###, P < 0.001 versus the NN2094 group. (C) Adhesion rates of S. mutans clinical strains classified by expression of CBP and PA. Each closed circle represents the mean value for each bacterial strain. Horizontal bars indicate the mean values for the groups. There were significant differences in the rates, which were determined using ANOVA with the Bonferroni correction. *, P < 0.05; ***, P < 0.001. (D) Representative histopathological images following Gram staining of tissue sections of bovine heart valves infected with TW295 in the presence of type IV collagen or serum. Bars = 50 μm. (E, F) Rates of TW295 adhesion to bovine heart valves in the presence of type IV collagen (E) and inactivated serum (F). There were significant differences in the rates, which were determined using ANOVA with the Bonferroni correction. ***, P < 0.001 versus the bacterium-only group; ##, P < 0.01 versus the serum group; ###, P < 0.001 versus the serum group. The bars in panels B, E, and F represent standard errors.

FIG 5

Evaluation of the rat model of IE. (A) Representative histopathological images following Gram staining of tissue sections from hearts extirpated from rats 7 days after S. mutans infection. Arrowheads, bacterial masses. Bars = 500 μm. (B) Representative histopathological images of bacterial mass formations in heart valves from rats infected with S. mutans TW295. The hearts were extirpated on days 1, 3, and 7 after infection. Bars = 100 μm. (C) Histopathological evaluation of bacterial mass formation for extirpated heart tissues infected with S. mutans TW295. The hearts were extirpated on days 1, 3, and 7 after infection. Each rectangle represents the mean value for each score. Bars represent standard errors. There were significant differences between the groups, which were determined using ANOVA with the Bonferroni correction. *, P < 0.05.