Antimicrobial Susceptibility of Corynebacterium bovis Isolates from Immunodeficient Rodents

Corynebacterium bovis, the causative agent of hyperkeratotic dermatitis in immunodeficient mice, is a significant problem in preclinical oncology research. Infection results in lifelong skin colonization and a decrease in successful engraftment of patient-derived xenograft tumor models. The use of antimicrobial agents for C. bovis is controversial in light of reports of poor efficacy and the possibility of selection for resistant strains. The purpose of this study was to describe the antimicrobial susceptibilities of C. bovis isolates obtained exclusively from immunodeficient rodents in order to aid in antimicrobial dose determination. Between 1995 and 2018, 15 isolates were collected from 11 research institutions across the United States. Antimicrobial susceptibility testing was performed for 24 antimicrobials commonly used against gram-positive bacteria. Our results provide an updated understanding of the susceptibility profiles of rodent C. bovis isolates, indicating little variability between geographically and temporally distant isolates. These results will facilitate appropriate antimicrobial use to prevent and treat C. bovis infections in immunodeficient rodents.

Corynebacterium bovis is a gram-positive, facultatively anaerobic pleomorphic bacillus that infrequently causes infections in humans⁵ but is more clinically relevant in veterinary medicine. Veterinary interest in this bacterium originated in the dairy industry, where it causes subclinical mastitis in infected animals and is the most common Corynebacterium spp. isolated from infected udders. When present as a primary infection, C. bovis can cause decreases in milk quality with no significant decrease in milk yield.^9,12 Despite being considered a minor pathogen, the impact of C. bovis on milk quality remains economically important to the dairy industry.C. bovis was first recognized in the mid1970s in athymic nude mice with hyperkeratotic dermatitis, a condition that would later be termed ‘scaly skin disease.’⁶ Once genetically characterized in the mid1990s and confirmed to have an association with clinical disease, C. bovis emerged as an important pathogen of immunodeficient mice in the laboratory animals.⁷ Historically, C. bovis infections of research mice primarily occurred in athymic nude mice. However, as the number of transgenic immunodeficient strains has expanded, C. bovis is no longer considered an infection exclusively of athymic nude mice, as infections have been reported in immunodeficient and ‘immune-vague’ research rodents around the world.^{3,10,11,15,21}Antimicrobial susceptibility testing is used to identify the minimum inhibitory concentration (MIC) of specific antimicrobials that prevents the growth of an individual bacterial isolate in vitro. By including many isolates of the same organism into a test population, the MIC can be calculated that inhibits the growth of 50% (MIC₅₀) or 90% (MIC₉₀) of the isolates.²² MIC have been published for C. bovis isolates obtained from dairy cows.²⁵ In the dairy industry, dry cow therapy (the administration of antibiotics at the end of lactation) is highly effective at eliminating subclinical mastitis caused by Corynebacterium spp.¹ However, elimination of C. bovis from immunodeficient mouse populations is much more challenging.^15,17 To date, the dose of amoxicillin used to treat C. bovis-infected immunodeficient mice has been informed by MIC data from dairy cows isolates²⁵ and in vivo pharmacokinetic data in the form of blood plasma concentrations of amoxicillin administered in the drinking water.¹⁶ However, our group and others have demonstrated the reemergence of infection in immunodeficient mice after the discontinuation of antibiotic administration in a C. bovis-free environment. These findings suggest that the MIC for C. bovis isolates from mice may differ from that of cows.²Recently, the genomes of C. bovis isolates obtained from humans, cows, mice, and rats were sequenced. Subsequent genomic comparisons assessing the average nucleotide identity between isolates identified sequence divergence obtained from humans and cows as compared with isolates from rodents.⁴ In particular, the number of genomic islands and virulence factors were significantly higher in the rodent isolates than in the human and cow isolates. However, whether phenotypic changes in antimicrobial susceptibility accompany this genetic divergence is unknown. Considering the prior observations and new developments in our understanding of C. bovis across multiple species, the purpose of this study is to describe antimicrobial susceptibility profiles of C. bovis isolates obtained exclusively from immunodeficient rats and mice.