Microbiota attenuates chicken transmission-exacerbated campylobacteriosis in Il10−/− mice

Campylobacter jejuni is a prevalent foodborne pathogen mainly transmitting through poultry. It remains unknown how chicken-transmitted C. jejuni and microbiota impact on human campylobacteriosis. Campylobacter jejuni AR101 (Cj-P0) was introduced to chickens and isolated as passage 1 (Cj-P1). Campylobacter jejuni Cj-P1-DCA-Anaero was isolated from Cj-P0-infected birds transplanted with DCA-modulated anaerobic microbiota. Specific pathogen free Il10−/− mice were gavaged with antibiotic clindamycin and then infected with Cj-P0, Cj-P1, or Cj-P1-DCA-Anaero, respectively. After 8 days post infection, Il10−/− mice infected with Cj-P1 demonstrated severe morbidity and bloody diarrhea and the experiment had to be terminated. Cj-P1 induced more severe histopathology compared to Cj-P0, suggesting that chicken transmission increased C. jejuni virulence. Importantly, mice infected with Cj-P1-DCA-Anaero showed attenuation of intestinal inflammation compared to Cj-P1. At the cellular level, Cj-P1 induced more C. jejuni invasion and neutrophil infiltration into the Il10−/− mouse colon tissue compared to Cj-P0, which was attenuated with Cj-P1-DCA-Anaero. At the molecular level, Cj-P1 induced elevated inflammatory mediator mRNA accumulation of Il17a, Il1β, and Cxcl1 in the colon compared to Cj-P0, while Cj-P1-DCA-Anaero showed reduction of the inflammatory gene expression. In conclusion, our data suggest that DCA-modulated anaerobes attenuate chicken-transmitted campylobacteriosis in mice and it is important to control the elevation of C. jejuni virulence during chicken transmission process.

Scientific Reports | (2020) 10:20841 | https://doi.org/10.1038/s41598-020-77789-2 www.nature.com/scientificreports/ ex-GF mice 20 . We also found that DCA resisted against chicken colonization of C. jejuni human clinical isolate 81-176 and chicken isolate AR101 21 . The microbiota composition at the cecal of the infected birds transplanted with DCA-modulated microbiota was assessed at phylum level using real-time PCR, and the results showed that the microbiota compositions were different 21 . Campylobacter jejuni motility and adherence to cells aren't changed in the presence of DCA, although DCA induced virulence genes ciaB, cmeABC, dccR, and tlyA 22 . However, it is unclear whether DCA regulates C. jejuni chicken transmission and subsequent induction of campylobacteriosis.
In this study, we found that chicken-transmitted C. jejuni (passage 1, Cj-P1) induced more severe intestinal inflammation in Il10 −/− mice compared to the non-transmitted bacterium (passage 0, Cj-P0), while Cj-P1-DCA-Anaero induced less colitis, bacterial invasion, and inflammatory gene expression in Il10 −/− mice compared to Cj-P1. Thus, C. jejuni transmitted from birds raised in various conditions could behave differently in inducing enteritis. The outcome of this study will provide key information about the interplay between chicken microbiome, C. jejuni transmission, and host susceptibility and response, and could help the development of new prevention strategies against foodborne pathogens.

Material and methods
Campylobacter jejuni strains isolated from infected birds. Campylobacter jejuni strain AR101 (Cj-P0) was isolated from experimental chickens at Dr. Billy Hargis's laboratory at the University of Arkansas at Fayetteville and the bacterium was used in our recent report 21 . In the report, C. jejuni in the cecal digesta of the 28 days of age birds infected with Cj-P0 was cultured on C. jejuni selective blood plates with five antibiotics (cefoperazone, cycloheximide, trimethoprim, vancomycin and polymyxin B) for 48 h at 42 °C using the GasPak system (BD), and the isolated C. jejuni was named as Cj-P1 (C. jejuni passage 1). In the report, C. jejuni in the cecal digesta of 28 days of age birds fed DCA and infected with Cj-P0 was cultured on the selective C. jejuni plates and was named as Cj-P1-DCA. Cecal digesta from uninfected 28 days of age birds fed with DCA diet was collected and cultured on Brain Heart Infusion (BHI) plates under anaerobic or aerobic conditions, and the isolated bacteria were named anaerobic-microbiota (DCA-Anaero) or aerobic-microbiota (DCA-Aero), respectively. Campylobacter jejuni in cecal digesta of 28 days of age birds colonized with DCA-Anaero or DCA-Aero and infected with Cj-P0 was cultured on the selective C. jejuni plates and was named as Cj-P1-DCA-Anaero or Cj-P1-DCA-Aero, respectively. The microbiota at the cecal of those birds was assessed at phylum level using real-time PCR, and the results showed that the microbiota compositions were different 21 . The AR101 strains were routinely grown on the selective C. jejuni plates and examined under microscopy for size, morphology and motility".
Campylobacter jejuni motility assay. Cj-P0, Cj-P1, Cj-P1-DCA, Cj-P1-DCA-Anaero or Cj-P1-DCA-Aero was grown on the selective plates, collected, and diluted to an optical density at 600 nm (OD 600 ) of 1. Each bacterium of 1 μl was then stabbed into a 0.4% agar Brain Heart Infusion (BHI) plate without antibiotic cocktail. The less dense agar facilitated C. jejuni to easier move inside the agar and the bacterium formed a halo of growth around the inoculation point. Following microaerobic growth at 42 °C for 24 h, the radius of the ring was calculated relative to that of Cj-P0. Cj-P0 was grown on each plate to control plate-to-plate variation. Experiments were performed in triplicate and repeated three times.

Mouse experiment.
Animal experiments were performed in accordance with the Animal Research: Reporting of In Vivo Experiments (https ://www.nc3rs .org.uk/arriv e-guide lines ). The experiments were approved by the Institutional Animal Care and Use Committee of the University of Arkansas. For C. jejuni infection experiments, cohorts of 5 to 9 SPF C57BL/6 Il10 −/− mice were orally gavaged daily with antibiotic clindamycin (Sigma-Aldrich) at 67 mg/kg body weight (BW) for 7 days. One day after the last antibiotic treatment, the mice were gavaged with a single dose (10 9 CFU/mouse) of Cj-P0 (5 mice), Cj-P1 (9 mice), Cj-P1-DCA-Anaero (5 mice), Cj-P1-DCA (9 mice), Cj-P1-DCA-Aero (5 mice), respectively as described above. Mice were followed clinically for evidence of diarrhea, failure to thrive, and mortality. At the end of experiments at 8 days post infection, tissue and stool samples from mouse colon were collected for protein, RNA, histology, and culture assay. For live C. jejuni counting, MLN and spleen were aseptically resected. Colon tissue was opened, resected, and washed three times in sterile PBS. Colonic luminal content (stool) was also collected. The freshly collected tissues and stool were weighed, homogenized in PBS, serially diluted, and cultured on selective C. jejuni plates supplemented with 5 antibiotics cocktail (cefoperazone, cycloheximide, trimethoprim, vancomycin, and polymyxin B) for 48 h at 37 °C using the GasPak system (BD Biosciences) as described before 21 . Campylobacter jejuni colonies were counted, and data were presented as CFU per gram tissue or stool. Histopathological images were acquired using a Nikon TS2 fluorescent microscope 23 . Intestinal inflammation was scored using a scoring system from 0-4 as showed before 20,24 . Fluorescence in situ hybridization (FISH). C. jejuni at intestinal tissue sections was visualized using FISH assay as previously described 24 . Briefly, tissue sections were deparaffinized and hybridized with the FISH probe for up to 48 h. The tissue sections were then washed, and the mammalian nuclei were visualized by staining with DAPI. The tissue sections were then imaged using the Nikon TS2 fluorescent Microscope system. Immunohistochemistry (IHC). Neutrophils in intestinal tissues were detected using anti-myeloperoxidase (MPO) IHC analysis as described previously 25 . Briefly, intestinal tissue sections were deparaffinized, blocked, and incubated with an anti-MPO antibody (1:400; Thermo Scientific) overnight at 4 °C. After incubation with anti-rabbit biotinylated antibody and three times washing, the tissue sections were incubated with avidin/biotin complex (Vectastain ABC Elite Kit, Vector Laboratories). After three times washing, the tissue sections were Scientific Reports | (2020) 10:20841 | https://doi.org/10.1038/s41598-020-77789-2 www.nature.com/scientificreports/ added with diaminobenzidine (Dako) within 2 min. The mammalian nuclei were visualized by staining with hematoxylin (Fisher Scientific). The tissue sections were then imaged using the Nikon TS2 Microscope system.

Real-time RT-PCR.
Total RNA from colonic tissue was extracted using TRIzol (Invitrogen). cDNA was prepared using M-MLV (Invitrogen). mRNA levels of proinflammatory genes were determined using the SYBR Green PCR Master Mix (Bio-Rad) on a Bio-Rad 384-well Real-Time PCR System and normalized to Gapdh. The primer sequences of the genes Gapdh, 17a, Il1β, and Cxcl1 were reported before 20 .
White blood cell isolation and migration assay. Blood was collected from Il10 −/− mice and the red blood cells were lysed in the buffer of 8.3 g/l NH 4 Cl in 0.01 M Tris-HCl buffer of pH 7.5. The collected white blood cells were resuspended in 1% FBS RPMI 1640 medium. Cells were plated at 10 4 per insert in 24-well Transwells (Corning) with 3 μm pores and incubated at 37 °C and 5% CO 2 . Cj-P0, Cj-P1, Cj-P1-DCA, Cj-P1-DCA-Aero, and Cj-P1-DCA-Anaero at 10 5 CFU/well were inoculated into the bottom wells. White blood cells migrated into the bottom well were imaged and counted one hour later using the Nikon TS2 Microscope system ,similar to previously described 25 . Cells in six fields per well were counted.

Statistical analysis.
Values were displayed as mean ± standard error of the mean as reported before 20 . Differences between groups were analyzed using the nonparametric Mann-Whitney U test with Prism 7.0 software. Experiments were considered statistically significant if P value was < 0.05.
Ethics approval and consent to participate. All animal protocols were approved by the Institutional Animal Care and Use Committee of the University of Arkansas at Fayetteville.

Results
Chicken-transmitted C. jejuni induced more severe intestinal inflammation. To address whether the asymptomatic C. jejuni transmission in chickens influenced its induction of intestinal inflammation in susceptible hosts such as human or Il10 −/− mice, we cultured infected bird cecal content, isolated C. jejuni, and labeled it as passage 1 or Cj-P1. We reasoned that C. jejuni transmitting through chickens altered virulence and would induce worse intestinal inflammation. To examine this hypothesis, SPF Il10 −/− mice were orally gavaged daily with antibiotic clindamycin for 7 days. The mice were then infected with Cj-P0 or Cj-P1 with a single oral gavage dose of 10 9 CFU/mouse. Interestingly, after 6 days post-infection, Il10 −/− mice infected with chickentransmitted Cj-P1 showed clinical sign of enteritis. After 8 days post infection, Il10 −/− mice infected with Cj-P1 demonstrated severe morbidity and bloody diarrhea and the experiment had to be terminated. At cellular level, Cj-P0 induced mild intestinal inflammation in the Il10 −/− mice, shown as crypt hyperplasia, mild goblet cell depletion, and mild immune cell infiltration into lamina propria (Fig. 1A). Remarkably, Cj-P1 induced more severe intestinal inflammation and higher histopathological score compared to Cj-P0 ( 2.8 vs. 0.8, P < 0.05) , shown as crypt abscesses, extensive immune cell infiltration and massive goblet depletion (Fig. 1A,B). These results suggest that C. jejuni transmission through chickens enhances infection capacity and induces more severe intestinal inflammation. and invasion are essential for its successful induction of campylobacteriosis 26 . To investigate the mechanism of how Cj-P1 induced more intestinal inflammation compared to Cj-P0, we next evaluated C. jejuni colonization and invasion in colon. Colon content and tissue were weighed, homogenized, serially diluted, and cultured on selective plates. Notably, the luminal colonization level of C. jejuni Cj-P1 was significantly denser in mouse colon compared to that of mice infected with Cj-P0 (3.24 × 10 7 vs. 6.22 × 10 6 CFU/g stool, P = 0.049) ( Fig. 2A).
We then examined C. jejuni invasion into colon tissue. In consistent with luminal C. jejuni colonization levels, Cj-P1 significantly invaded into colon tissue compared to Cj-P0 (6.47 × 10 5 vs. 8.12 × 10 4 CFU/g tissue, P = 0.016) (Fig. 2B). To further detect the C. jejuni spatial distribution in colon tissue, we visualized C. jejuni DNA using fluorescence in situ hybridization (FISH) and fluorescence microscopy imaging. Notably, while Cj-P1 DNA was detected widely in the inflamed crypts and the lamina propria section of the mouse intestine, Cj-P0 was seldomly detectable in the mouse colon (Fig. 2C). These results indicate that Cj-P1 gains virulence ability to invade more aggressively into colonic crypts.
Transmitted Cj-P1 induced severe crypt abscesses in colon. Since C. jejuni infection induced strong intestinal inflammation in Il10 −/− mice, we examined the histopathology of the infected mice at higher magnification to have more detailed assessment. Notably, Cj-P1 depleted the majority of goblet cells and induced massive immune cell infiltration into lamina propria and numerous crypt abscesses compared to Cj-P0 (Fig. 3A).
Since crypt abscesses were observed in histopathology slides, we then detected the neutrophils by targeting neutrophil marker myeloperoxidase (MPO) using immunohistopathology (IHC). As showed in Fig. 3B, Cj-P0 induced a few MPO positive cells into crypt and formed fewer crypt abscesses, whereas Cj-P1 showed stronger induction of crypt abscesses compared to Cj-P0. These results indicate that Cj-P1 gains virulence capacity to induce more infiltration of immune cells such as neutrophils.

DCA-Anaero attenuated chicken transmitted Cj-P1 invasion. To investigate why Cj-P1-DCA-
Anaero induced less intestinal inflammation, we then evaluated C. jejuni colonization and invasion in colon. In consistent with previous results, Cj-P1 colonized in colon was in the trend but not significant compared to Cj-P0 and Cj-P1-DCA-Anaero (Fig. 6A). Notably, Cj-P1 invades more than 100 folds in colon compared to Cj-P0, an effect attenuated by 99% in Cj-P1-DCA-Anaero (Fig. 6B). Further visualization of C. jejuni spatial invasion by FISH showed that Cj-P1 DNA was detected deeply in the inflamed crypts and the lamina propria, while Cj-P1-DCA-Anaero was mostly absent in the mouse colon (Fig. 6C). These results suggest that DCA-Anaero attenuates C. jejuni transmission-increased virulence on invasion.

DCA-Anaero attenuated chicken transmitted Cj-P1 induction of inflammatory response.
Because of the reduced bacterial invasion and colitis in Cj-P1-DCA-Anaero compared to Cj-P1, we reasoned that the former strains would induce fewer inflammatory responses. To examine this possibility, colon tissue was collected, and RNA was extracted. Gene expression of inflammatory cytokines and chemokines were measured using real-time PCR. Cj-P0 induced inflammatory genes of Il17a, Il1β, and Cxcl1 at 10, 4, and 2 folds, respectively, compared to uninfected mice (Fig. 7A). Remarkably, Cj-P1 induced the gene expression by 19, 10, and 11 folds, respectively, compared to Cj-P0, which was attenuated by 86, 74, and 86%, respectively, by Cj-P1-DCA-Anaero. Since the three pro-inflammatory cytokines mediated immune cell activity, we then visualized inflammatory neutrophil distribution in colon by IHC of MPO. Notably, Cj-P1 induced strong infiltration of MPO positive neutrophil and crypt abscesses, whereas Cj-P1-DCA-Anaero barely elicited neutrophil migration and crypt abscesses (Fig. 7B). These results suggest that DCA-Anaero attenuates C. jejuni transmissionincreased virulence on the induction of inflammatory responses.

Discussion
Although C. jejuni is a prevalent foodborne pathogen mainly transmitted from chickens, few approaches available to control the bacterial chicken colonization. Moreover, the microbiota and cellular events responsible for host resistance or susceptibility to C. jejuni infection remain largely elusive 27,28 . Foodborne C. jejuni is mainly monitored through enumerating the bacteria in food such as chickens. The limitation of the practice is the lack of understanding of the virulence change after chicken colonization/transmission. In previous studies, we infected chickens with C. jejuni AR101 (labeled as Cj-P0) and the chickens didn't show any clinical signs and grew comparably to uninfected birds 21 . The result was consistent with previous researches that C. jejuni asymptomatically colonizes chickens as a commensal-like pathogen 29 . Interestingly, C. jejuni passaging through chickens increases colonization potential in chickens 30 and mice 31 , however, it hasn't been well studied if the asymptomatic transmission in chickens influenced C. jejuni induction of intestinal inflammation in susceptible hosts such as human or Il10 −/− mice. In this study, we investigated how C. jejuni transmitted from birds raised in different husbandry www.nature.com/scientificreports/ influenced its virulence in subsequent infection using Il10 −/− mice. The results reveal new insights regarding C. jejuni chicken transmission, pathogen virulence, and husbandry conditions. One of the notable findings in this study was that chicken-transmitted C. jejuni (Cj-P1) increased virulence to induce stronger campylobacteriosis in mice, although chickens colonized with C. jejuni were healthy, shown with comparable body weight gain between uninfected and infected birds 21 . Campylobacter jejuni colonizes 95% flock of 20,000 chickens within 7 days after the initial one bird infected with the bacterium 32 . Transmitted C. jejuni increases phase-variable controlled flagellar 33 , which may contribute to its fast horizontal transmission rate in the flock. Furthermore, C. jejuni passaging through chicken reservoir promotes phase variation in specific contingency genes, and the populations with the variations colonize mice 31 . In consistent with these previous reports, in this study, the chicken-transmitted C. jejuni colonized mice with more number than the original pathogen Cj-P0. A new observation in this study is that beyond colonization ability increase, the chicken-transmitted C. jejuni also increased virulence to induce more severe colitis. At the host cellular level, the transmitted C. jejuni induced more immune cell infiltration in colon and increased immune cell migration in in vitro assay. Immune cell migration is one of the important steps in inducing campylobacteriosis 25 . Innate immune cells can directly sense bacterial cellular molecules with various receptors such as Toll-like receptors (TLR) and Nod-like receptors (NLR) 34 . It would be helpful to investigate in the future whether the chicken-transmitted C. jejuni had phase variations in cell surface virulence genes, such as lipooligosaccharide. In addition, it remains to be determined whether the chicken-transmitted C. jejuni increased virulence genes on mobility, growth, and toxin production. Together, these findings suggest that an equal count of C. jejuni may induce quite a different campylobacteriosis, and monitoring chicken-transmitted C. jejuni virulence is important for preventing foodborne campylobacteriosis.
The remarkable finding in this study is that DCA-modulated anaerobes in chickens reduced the transmitted C. jejuni virulence on inducing campylobacteriosis in mice (Cj-P1-DCA-Anaero vs. Cj-P1). DCA-modulated anaerobes reduce C. jejuni chicken colonization 21 . Conventionalized anaerobic microbiota reduces C. jejuniinduced intestinal inflammation in gnotobiotic mice 20 . It is well documented that microbiota prevents intestinal pathogen colonization through competitive exclusion such as virulence expression 35 , nutrition exclusion 36 , altered pH 37 , and bactericidal products 38 . However, it remains largely elusive whether microbiota influences co-inhabited pathogens on their-transmitted infection capacity. Here, we showed that DCA-modulated anaerobes in chickens reduced the transmitted C jejuni colonization and invasion in the large intestine of mice. Mechanistically, DCA-modulated anaerobes reduced transmitted C jejuni induction of immune cell migration in vitro and into the intestine as well as decreased the bacterial induction of proinflammatory mediator expression. It will be important to conduct future research on which C. jejuni virulence factors were down-regulated by DCA-Anaero and were responsible for the reduction of campylobacteriosis. These findings suggest that the increased chicken-transmitted C. jejuni virulence is feasible to be controlled with select microbiota (e.g. DCAmodulated anaerobes). www.nature.com/scientificreports/ Interestingly, DCA-modulated C. jejuni (Cj-P1-DCA) failed to reduce intestinal inflammation, while DCA in the diet reduces C. jejuni chicken colonization 21 . Campylobacter jejuni multidrug efflux transporter gene, CmeABC, is significantly up-regulated during the initial exposure to bile acids such as DCA 39 . In consistent with the elevated production of CmeABC, bile salts in culture media promote C. jejuni resistance to multiple antimicrobials 39 . DCA induces C. jejuni virulence gene expression and invasion into epithelial cell in vitro 22 . DCA failed to reduce C. jejuni in vitro growth 21 . Consistently, Cj-P1-DCA showed similar motility and induction of immune cell migration compared to Cj-P1, suggesting comparable virulence on activating immune cells.
Taken together, our data revealed that DCA-modulated anaerobic microbiota not only reduced C. jejuni colonization in chickens, but also reduced chicken-transmitted campylobacteriosis. The reduction of virulence reflected on immune cell migration and infiltration into the intestine, C. jejuni invasion, pro-inflammatory response, and collectively intestinal inflammation. These findings highlighted the importance to monitor chickentransmitted C. jejuni virulence in addition to colonization counts. Although it is a new concept, using select microbiota in poultry production is a key step to successfully prevent foodborne campylobacteriosis.

Data availability
Data sharing is not applicable to this article as no datasets were generated or analyzed during the current study.