A Study of a Bacteriophage-Resistant Salmonella Mutant and Factors Influencing Prevalence of Salmonella in a Multi-Species Animal Facility

Abstract

Bacteriophage (phage) treatment for the reduction of multiple drug resistant Salmonella Newport in dairy calves is being examined in our lab from a clinical disease and food safety perspective. An unintended consequence of phage treatment could be the emergence of fully virulent but phage resistant Salmonella. Working from the hypothesis that resistance would attenuate virulence in S. Newport, we generated a spontaneous mutant resistant to 4 out of 5 lytic phages used in our treatment regimen. Two pairs of 8-10 week old calves were challenged orally with the phage resistant mutant, with one pair administered a dose of 7.45x109 colony forming units (CFU’s) and the second pair 1.47x1010 CFU’s. A third pair of calves was challenged with a total dose of 1.96x1010 CFU’s composed of a 1:1.3 ratio of parent:mutant in a competition assay designed to determine how well the mutant competed against the parent strain in vivo. The four calves inoculated with the phage resistant S. Newport mutant strain alone showed much reduced Salmonella fecal shedding and no clinical disease signs. The calves in the competition assay showed severe diarrhea and high fecal shedding of both parent and mutant Salmonella strains, along with signs of fever and lethargy. The decreased shedding and absence of signs in the calves which received the phage resistant strain alone indicates that the spontaneous generation of multi-phage resistance in S. Newport had an attenuating effect on virulence perhaps by decreasing the mutant’s fitness and/or its ability to colonize its host. The results of the competition experiment suggest that the dual inoculation with parent and mutant strain restored virulence to the mutant strain and may increase fitness in the parent strain. Future experiments designed to examine the mechanism of phage resistance in this attenuated S. Newport mutant are planned.

Diseases caused by serotypes from the bacterial genus Salmonella can have a major impact on animal and human health. Little research has been performed to examine factors contributing to Salmonella incidence in multi-species animal production facilities such as veterinary teaching hospitals and the ability of Salmonella serovars to move to adjacent facilities. We hypothesize that Salmonella can move between proximally located animal facilities and pastures and that specific factors increase the likelihood of isolating environmental Salmonella serotypes in these locations. Over two years, 631 samples were collected from various large animal facilities and pastures within a veterinary school. Data was recorded to assess factors that contribute to increased prevalence of environmental Salmonella contamination. Samples were processed for Salmonella isolation as described in the USDA FSIS Microbiology Laboratory Guidebook. Salmonella isolates were submitted to the NVSL, Ames, IA, or Biovet, Inc., for serotyping. Data was analyzed with Statistical Analysis System (SAS). Of the 631 samples obtained, 230 (36%) samples were positive for at least one Salmonella serotype. Salmonella was recovered from the majority of facilities and areas sampled regardless of animal species exposure. A Risk Ratio and Fisher’s Exact Test was used for bivariate analysis of factors associated with Salmonella isolation. The factors shown to be significant were season, resident species, and environment. Variables associated with the isolation of Salmonella from environmental samples included summer season, water and drain samples, samples exposed to bovine residential animals, indoor environments such as buildings, and the dairy and food animal barns locations with isolation of Salmonella serovars. The significantly increased frequency of Salmonella isolation from environmental samples exposed to dairy cattle indicates that this species is either the source of this pathogen, or is serving as an amplifying host for Salmonella. The recovery of cattle-associated serotypes supports this conclusion.