Escherichia coli is the leading cause of severe mastitis in dairy farms. As E. coli mastitis is refractory to the hygienic control measures adapted to contagious mastitis, efficient vaccines are in demand. Existing mastitis vaccines, based on the use of killed rough E. coli J5 as the antigen, aim at inducing phagocytosis by neutrophils. We assessed the binding of J5-induced antibodies to isogenic rough and smooth strains along with a panel of mastitis-associated E. coli. Analysis by enzyme-linked immunosorbent assay revealed that antibodies to OmpA or killed J5 bind readily to rough E. coli but poorly to smooth strains. Flow cytometry analysis indicated that immunization with J5 induced antibodies that cross-reacted with rough E. coli strains but with only a small subpopulation of smooth strains. We identified type 1 fimbriae as the target of most antibodies cross-reacting with the smooth strains. These results suggest that the O-polysaccharide of lipopolysaccharide shields the outer membrane antigens and that only fiber antigens protruding at the bacterial surface can elicit antibodies reacting with mastitis-associated E. coli. We evaluated J5-induced antibodies in an opsonophagocytic killing assay with bovine neutrophils. J5 immune serum was not more efficient than preimmune serum, showing that immunization did not improve on the already high efficiency of naturally acquired antibodies to E. coli. In conclusion, it is unlikely that the efficiency of J5 vaccines is related to the induction of opsonic antibodies. Consequently, other research directions, such as cell-mediated immunity, should be explored to improve E. coli mastitis vaccines.
IMPORTANCE Despite intensive research, mastitis remains an important disease in dairy cattle with a significant impact on animal welfare, use of antibiotics, and, in the end, the economy of dairy farms. Although vaccines available so far have shown limited efficacy against coliform mastitis, vaccination is considered one of the measures that could limit the consequences of mastitis. One reason for the lack of efficiency of current vaccines likely stems from the current evaluation of vaccines that relies mostly on measuring antibody production against vaccine antigens. This report clearly shows that vaccine-induced antibodies fail to bind to most mastitis-associated E. coli strains because of the presence of an O-antigen and, thus, do not allow for improved phagocytosis of pathogens. As a consequence, this report calls for revised criteria for the evaluation of vaccines and suggests that cell-mediated immunity should be targeted by new vaccinal strategies. More generally, these results could be extended to other vaccine development strategies targeting coliform bacteria.