Fig. 2.

Purified recombinant SodB-based vaccine, but not a CjaA-based vaccine, was protective against homologousCampylobacterchallenge in chickens. Data represent the arithmetic mean of log transformed caecal counts of C. jejuni from a minimum of 3 independent trials where each antigen was tested concomitantly, each using 3–6 birds per time interval, ± the standard error of the mean (SEM). Data for the GST and GST–CjaA control groups was available from an additional experiment that did not test GST–SodB concomitantly but which had an identical design to the three experiments that tested GST–SodB in parallel. Samples were collected at weekly intervals at post-mortem examination and as a result the lines are inferred to reflect the course of excretion, and are not longitudinal values from the same animals. Values in the table are mean CjaA- and SodB-specific serum IgY and bile secretory IgA levels in vaccinated and control groups measured by ELISA at the time intervals on the x-axis. Changes compared to the GST control group were determined using a generalised linear model (R² = 0.47) and post-hoc Dunnet's tests. Statistically significant differences to the GST control group are denoted with and *** for P ≤ 0.001.