Why Do Peptidyl Arginine Deiminase (PAD) 4-Deficient Mice Have Improved Survival Following Cecal Ligation and Puncture (CLP) Sepsis? Implications for the Role Neutrophil Extracellular Trap (NETs) in Immune Homeostasis

Background:

NETs are webs of decondensed chromatin that capture and kill bacteria. However, when released in a dysregulated manner, NETs can cause collateral tissue injury to the host. In prior studies designed to understand the role of NETs in sepsis, mice deficient in PAD4 (Padi4-/-), a crucial mediator of NET release (NETosis), were subjected to the CLP model of polymicrobial sepsis, and found to have improved survival compared to WT controls.

Aims:

We proposed that this result might be due to critical differences in the inflammatory environment and gastrointestinal (GI) microbiome in Padi4-/- mice compared to WT animals. To test this hypothesis, we replicated the CLP studies using the alternative cecal slurry (CS) model of polymicrobial sepsis, which allowed us to treat WT and Padi4-/- mice with bacterial inoculum from either genotype-matched or genotype-switched donors.

Methods:

Complete blood counts (CBCs), plasma, stool, and tissue samples were obtained from 9-16 week WT and Padi4-/- C57Bl6 mice at baseline. Plasma proteomic analysis was performed with the Olink 96 mouse exploratory panel, immunohistochemistry and flow cytometry were used to quantify neutrophil infiltration in the GI tract, and 16S rRNA sequencing was employed to assess the relative abundance of different bacterial species in fecal samples of WT to Padi4-/- mice prior to infection. WT and Padi4-/ mice had polymicrobial sepsis induced by intraperitoneal injections of CS derived from cecal samples of either pooled WT mice or Padi4-/- donor mice. 24 hours post-injection severity of illness was assessed with the mean sepsis score (score), CBCs were obtained, bacterial dissemination was quantified in blood and liver homogenates, and these samples were also analyzed with 16S rRNA sequencing to quantify differences in which microbial species had proliferated. Plasma was as subjected to proteomic analysis. Survival studies were performed on a parallel subset of mice.

Results:

Compared to WT mice at baseline, Padi4-/- mice had elevated circulating neutrophils, increased tissue neutrophil infiltration, most prominently in the jejunal villi, and higher plasma levels of proinflammatory cytokines including TNFα, IL5, and IL6, and reduced levels of IL23r. Principle component analysis (PCA) of 16S rRNA sequencing of fecal samples showed distinct clustering of samples from WT and Padi4-/- mice with higher levels of Bacteroidetes species in samples from Padi4-/- mice with a trend towards higher species diversity in the Padi4-/- mice compared to WT controls, suggesting that PAD4 expression influences composition of the GI microbiome. In CS studies, WT mice that received WT CS had higher MSS, lower temperatures, lower platelet counts, higher inflammatory cytokines, and 50% survival. In contrast, Padi4-/- mice that received Padi4-/- CS, had lower bacterial dissemination, reduced inflammatory cytokines, and 100% survival. Comparative CS studies were performed in which animals were injected with genotype-switched CS. Padi4-/- mice injected with WT CS had elevated MSS, higher plasma inflammatory cytokines, and 75% survival, whereas WT mice treated with Padi4-/- CS had mild disease course with 100% survival. PCA analysis of 16S rRNA sequencing of bacteria in blood and livers from septic mice showed clustering patterns affected by the genotype of the recipient mouse, but influenced more prominently by the genotype of the CS donor mice, with higher levels of Enterococcaceae in mice that received WT slurry and higher levels of Lactobacillaceae in mice treated with Padi4-/- CS.

Conclusion:

Deficiency of PAD4 in mice influences baseline inflammation, neutrophil mobilization, and GI microbiome composition. These changes are likely the result of compensatory mechanisms that must be accounted for in understanding the role of NETs in the immune response. Results of our CS experiments suggest that improved outcomes in Padi4-/- mice following CLP reflect changes in baseline inflammatory signaling as well as reduced pathogenicity of the Padi4-/- GI microbiome. Results of 16S rRNA sequencing of blood and liver samples from septic mice demonstrate that mouse PAD4 expression as well as CS donor genotype influence the survival and proliferation of different bacterial species in our polymicrobial sepsis model. Further studies are underway to examine how the complex interplay between PAD4-dependent NETs and bacteria influences immune homeostasis.