Capsular branch-dependent KC capture of circulating Spn10A. (A and B) Bacteremia kinetics (A) and bacterial distribution at 30 min (B) of mice i.v. treated with 400 μg CPS10A or PBS before i.v. infection with 10⁶ CFU of TH860^10A and TH2912¹⁴. n = 6. (C and D) Bacteremia kinetics (C) and bacterial distribution at 30 min (D) of mice i.v. infected with 10⁶ CFU of TH860^10A and three serotype-10B pneumococcal strains. n = 6. (E) Biochemical structure of repeating units of CPS10A and CPS10B. (F) Schematic representation of the cps loci of Spn10A and Spn10B. (G) Representative images of TH860∆wcrG (left), TH860 (middle), and TH860∆cps (right) stained with Anthony’s staining method. Scale bar, 2 μm. (H and I) Bacteremia kinetics (G) and bacterial distribution at 30 min (H) of mice i.v. infected with 10⁶ CFU of TH860^10A or TH860^10A∆wcrG. n = 6. (J and K) Bacteremia kinetics (I) and bacterial distribution at 30 min (J) of mice i.v. treated with 400 μg CPS10A or CPS10A∆wcrG and i.v. infected with 10⁶ CFU of TH860^10A. n = 6. (L) Representative IVM images of liver sinusoids (left) and quantitation of bacteria immobilized on KCs (right) of mice i.v. treated with 400 μg CPS10A or CPS10A∆wcrG and i.v. infected with TH860^10A. n = 2. The clearance processes are demonstrated in Video 2. Data were representative results (G and L) or pooled from two independent experiments (A–D and H–K). Ordinary two-way ANOVA with Tukey’s multiple comparisons test (A, C, H, and J) and unpaired t test (K) were performed. ****, P < 0.0001; ns, not significant.