Human Colonoids as a Model for Studying Host‐Pathogen Interactions with Campylobacter jejuni

Background Campylobacter is a leading cause of bacterial enterocolitis in the U.S. and around the world. In addition to the acute morbidity, it is associated with delayed sequelae such as environmental enteropathy and post‐infection irritable bowel syndrome. Host responses to Campylobacter infection are poorly studied due to colonization resistance in mice thereby restricting virulence studies to immortalized cell lines. These models lack the complexity of the host environment. Recent studies have highlighted the use of organoids for studying enteric pathogens. We propose that human organoids provide a unique platform for studying pathogenic mechanisms incorporating unique host (human) and infection (strain) characteristics. Objective To demonstrate human colonic organoids (colonoids) as a model for studying C. jejuni infection. Methods Sigmoid colonic biopsies were collected from healthy volunteers between 18–35 years of age (n=3). Intestinal crypts were isolated, washed and plated in matrigel domes for colonoid formation. Mature colonoids were harvested after 7 days, and dissociated using a 0.05% trypsin solution. These were plated on collagen IV coated plates, generating 2D‐monolayers for infection experiments. Three different C. jejuni strains (Sequence Types 45, 22, 8) were grown to an OD 600 of 1 and used to infect the colonoid monolayers. To assess adhesion and internalization, colonoids were infected for a total of 3 hours after which monolayers were either washed and subsequently lysed (to determine attached plus invaded bacteria) or treated with gentamycin and then lysed (to determine invaded bacteria). Bacterial quantification was done by plating on Mueller Hinton agar. Additional colonoid monolayers were grown on permeable transwell inserts. Baseline transepithelial resistance (TER) of confluent monolayers was recorded and then continuously assessed every hour after the addition of the C. jejuni strains (using impedance spectroscopy in cellZscope) to evaluate changes in barrier integrity. Results All C. jejuni strains attached and invaded colonoid monolayers derived from all three volunteers. Two strains (ST 45 and 22) caused greater adhesion and invasion in all three colonoids compared to the third strain (Adhesion mean (SEM): 4.32 x 10 −2 (2.33 x 10 −3 ), 9.36 x 10 −3 (4.82 x 10 −4 ) vs 7.79 x 10 −3 (3.51 x 10 −3 ); Invasion mean (SEM) 3.47 x 10 −4 (2.13 x 10 −5 ), 2.33 x 10 −4 (1.36 x 10 −4 ) vs 4.57 x 10 −5 (1.73 x 10 −5 ) (ANOVA, p<0.05, n=3/group for both analyses). Additionally, greater drop in TER was seen in colonoid monolayers exposed to the two strains with higher adhesion and invasion (Mean % of baseline TER at 30 hr post‐infection 2.58% (1.21), 14.14% (3.39) vs 42.29% (18.49), p<0.01). These results correlated with virulence of these strains in conventional T84 monolayers. Transcriptomic and proteomic characterization of the C. jejuni infection in these colonoids is in progress. Conclusions Human colonoids are useful to assess the role of the host in infectivity and epithelial barrier responses to C. jejuni infection. Future work involves determining responses in colonoids derived from volunteers with specific characteristics to create unique host‐bacterial infection models. Support or Funding Information Supported by funding from Department of Defense W81XWH‐18‐1‐0074