FIG 4.

The effect of timing on serotype success during competitive colonization. (A) To show mouse-to-mouse variation in serotype presence, URT lavages were obtained 35 days postinoculation with the six capsule-switch constructs. The proportion of each serotype was determined by colony immunoblot. Data from 10 individual mice are shown. (B) Schematic of the experimental timeline. In “head start” experiments, pups were inoculated IN at 4 days of age with a single capsule-switch construct (T4¹⁴). Six or 24 h later, as indicated, pups were then challenged with an even mixture of the other 5 capsule-switch constructs. In “delay” experiments, pups were inoculated IN at 4 days of age with an even mixture of 5 capsule-switch constructs. Six or 24 h later, as indicated, pups were challenged with the sixth capsule-switch construct (T4^23F). In both experiments, bacteria were recovered by URT lavages 3 days after the last inoculation and analyzed for serotype composition by colony immunoblotting. (C) The effect of a head start or delay on serotype success was determined by comparison to serotype proportions in standard 3-day colonization where all six constructs were given simultaneously. These differences in serotype proportion were compared to a hypothetical median of 0 using the Wilcoxon Signed Rank test. Red symbols (*) indicate a significant increase in serotype proportion, while black symbols (*) represent a significant decrease compared to the standard colonization. *, P < 0.05; **, P < 0.01. n = 6–10 mice/condition. (D) Pups were inoculated IN at 4 days of age with T4¹⁴ alone. Left-hand panels show the colonization density of T4¹⁴ recovered by URT lavage of individual mice at 2, 4, and 6 weeks postchallenge. Pups shown in right-hand panels were given a second inoculation with the other 5 capsule-switch constructs 2, 4, or 6 weeks after the first. Bacteria were recovered by URT lavage from these pups 3 days following the second infection, and analyzed for CFU and serotype composition by colony immunoblotting. Each bar denotes a single mouse. Dashed horizontal lines represent the limit of detection.