9±5.5 (P=0.003). To determine whether the advantage of B. parapertussis was manifest at earlier stages of infection, mice (four per group) were inoculated as in the previous experiment with a mixed 1 : 1 inoculum of B. pertussis and B. parapertussis and euthanized at days 1, 2, 4 and 7 postinoculation. Doxorubicin research buy The competitive advantage of B. parapertussis was observed as early as 24 h postinoculation (mean CI=7) and was maintained through the peak of infection (Fig. 1b). Together, these data indicate that B. parapertussis not only outcompetes B. pertussis in a mixed infection, but also that it benefits

from the presence of B. pertussis in the infection. To further explore the competition between these two organisms in a mixed infection, mice (four per group) were infected with mixed inocula Pirfenidone in vitro at ratios (B. pertussis to B. parapertussis) of 10 : 1, 3 : 1, 1 : 3 and 1 : 10 (106 total CFU). Mice were euthanized 7 days postinoculation, and the bacterial load and ratio of the two organisms were determined as before. Bordetella parapertussis outcompeted B. pertussis

at all inoculum ratios (Fig. 2). In mice where the inoculum contained B. parapertussis as the predominant strain (1 : 3 and 1 : 10), B. pertussis was at a significant disadvantage, with relatively low CFU recovered from the mice (no B. pertussis was recovered from two mice in the 1 : 10 group). Remarkably, even when the inoculum contained 10-fold excess of B. pertussis (10 : 1), greater CFU of B. parapertussis were recovered from the host, with a mean CI of 31 (Fig. 2). Overall, these data show that B. parapertussis is able to outcompete B. pertussis in a mixed infection over a range of input ratios and apparently gains a greater advantage (higher CI) when the initial inoculum contains higher numbers of B. pertussis. To determine

whether the advantage to B. parapertussis occurs early in the process of infection, experiments were conducted staggering the inoculation with the two organisms. Two different staggered inoculations were tested: (1) mice were inoculated first with 2.5 × 105 CFU B. pertussis (t0) and were inoculated new 3 h later with 2.5 × 105 CFU B. parapertussis (t3), or (2) mice were inoculated initially with 2.5 × 105 CFU B. pertussis (d0) and inoculated 24 h later with 2.5 × 105 CFU B. parapertussis (d1). Mice from each group were euthanized on day 2 post-B. pertussis inoculation, and the bacterial loads and the ratio of the two organisms recovered were determined. In mice inoculated at t0 and t3, greater numbers of B. parapertussis than B. pertussis were recovered (CI=1.82, P<0.05, data not shown). When the inoculations were staggered by 24 h, neither strain had a significant advantage (CI=0.9, data not shown), even though B. pertussis colonization began 24 h earlier than that of B. parapertussis. These data indicate that B. parapertussis can outcompete B. pertussis during early infection even when B.