Modeling pathogen-driven neonatal late-onset sepsis: a modification to the murine cecal slurry

Neonatal sepsis is a major cause of neonatal morbidity and mortality. Reliable animal models are essential to our understanding of late-onset sepsis, but notable limitations exist in the current standard murine cecal slurry model. We sought to refine the existing model by using an injection of known stock slurry ("NEC'teria") cultured from an infant who died of necrotizing enterocolitis to better mimic sepsis following the translocation of neonatal specific bacterial pathogens from the intestine into the peritoneum.

To induce sepsis, neonatal mice (P7 and P14 - P16) were given an intraperitoneal injection of varying concentrations of NEC'teria, while sham controls received an injection of PBS. Mice were monitored for survival and tissue samples, serum, and peritoneal washes were collected for further assessment of inflammation, immune response, and intestinal injury. Ceca were collected for microbiome analysis.

While the polymicrobial cecal slurry from adult mice contained common healthy gut microbes, NEC'teria is composed of bacteria, primarily from the Enterobacteriaceae and Enterococcaceae families, that are common causes of late-onset sepsis. NEC'teria exposure significantly increased serum inflammatory cytokines, resulted in intestinal injury, altered the microbiome composition, and induced significant changes in local and systemic immune cell expression. Sepsis-induced mortality, inflammation, and intestinal injury were live-bacteria dependent and could be attenuated by administration of an antibiotic one hour after bacterial injection.

Our modification to the cecal slurry neonatal sepsis model resulted in a consistent sepsis-related mortality and phenotypic changes in neonatal mouse pups that resembled the changes that occur in human preterm infants who develop late-onset sepsis. Our pathogenic slurry is highly relevant to neonatal sepsis, as it is comprised of bacterial families found commonly in septic neonates. We expect our model to be highly reproducible between institutions, due to the standardized bacterial dose and characterized stock solution.