Spatially nested pathobionts in cecal crypts as a potential source of sepsis

Author(s):
Alexander Zaborin; Beatriz Penalver Bernabe; Anukriti Sharma; Sanjiv Hyoju; Neil Gottel; Jack Gilbert; Olga Zaborina; John Alverdy

Background:

Genetic tracking of bacteria has revealed that many healthcare-associated infections arise from the patient’s intestinal microbiota. These infections can be caused by pathobionts-normal microbial community members that become detrimental under stress. The microscopic niches for pathobionts remain ill-defined.  We have previously demonstrated that cecal crypts are densely colonized by microbiota that become disrupted when the host is exposed to major surgical injury.

Hypothesis:

Here we hypothesize that pathobionts can remain undetectable when camouflaged within cecal crypt sites, from which they systemically disseminate and propagate when the core crypt microbiota are disrupted. Thus, the aim of this study was to define the composition and structure of the normal cecal crypt microbiota and the fate of crypt pathobionts following surgical injury in mice.

Methods:

C57/B6 mice underwent pre-operative antibiotic treatment and overnight starvation followed by 30% hepatectomy. Cecum and liver were collected for microbiota analysis by 16S rRNA. Laser capture microdissection was used to isolate crypt and luminal contents of the cecum from histological samples. DNA was extracted and 16S rRNA amplified and sequenced with Illumina MiSeq. Exact sequence variants (ESV) were identified with DADA2, and abundance differences between the conditions was determined with generalized linear models. Bacterial presence was corroborated using Fluorescence in Situ Hybridization (FISH).

Results:

Before surgery, the microbiota in cecal crypts compared to lumen had lower diversity (p<0.05) and significantly distinct composition (beta-diversity p<0.001). Notably, cecal crypts were enriched with Mucispirilum schaedlery (Relative abundance (RA), 18% in crypts (C) vs 0.4% in lumen (L), p<0.001), where they formed tightly bundled packages (as visualized by FISH), incorporated with different potentially pathogenic species of Gammaproteobacteria  (RA, 3%C vs 0.6%L, p<0.001). In contrast, the lumen was enriched in genera Turicibacter (16%L vs 0.12%C, p<0.001) and Lactobacillus (9%L vs 0.7%C, p<0.001). Following surgery, microbiota differentiation constraint was lost; M. schaedlery became undetectable in crypts, and Proteobacteria became dominant (86%, p<0.001). The luminal microbiota resembled the Proteobacteria-enriched crypts (rho>0.1, p<0.05) and was inversely associated with the pre-surgery lumen composition (p<10^-5), as 100 Clostridia-related ESVs were highly depleted (p<0.001).  Some potentially pathogenic Gammaproteobacteria species (e.g., genus Serratia) were observed to propagate and disseminate to distant tissues such as the liver (32%, p=0.04).

Conclusions:

Cecal crypts may represent a hidden niche for pathobionts that propagate and disseminate following surgical stress. Identification of spatially nested microbes as potential sources of sepsis requires microscopic sampling and genetic speciation.