We also sought to establish colonies of outbred mice harboring distinct complex microbiota that is naturally occurring in contemporary rodent producers. Doing so would facilitate future CMTR studies as outbred mice have greatly improved reproductive indices. To this end, 8–10 week old female C57BL/6J (The Jackson Laboratory), C57BL/6NTac (Taconic Biosciences, Inc.), and Crl:CD1 (Charles River Laboratories) mice were purchased and allowed to acclimate for 1 week prior to use. Embryos were harvested from 8 week old Crl:CD1 (Charles River Laboratories) mice and implanted into recipient C57BL/6J (The Jackson Laboratory), C57BL/6NTac (Taconic Biosciences, Inc.), and Crl:CD1 (Charles River Laboratories) mice as previously described section Embryo Collection and Transfer. Offspring from surrogate dams were mated using an outbred mating scheme and individual colonies with targeted GM were maintained for two generations. Second generation 8–10 week old females were used for rederivation of B6.129P2-Il10^tm1Cgn/J (B6 IL-10^−/−) mice.

Helicobacter hepaticus (MU94) was grown on 5% sheep blood agar plates (Becton Dickinson, Franklin Lakes, NJ) overlaid with 5 ml Brucella broth (Becton Dickinson) supplemented with 5% fetal bovine serum (Sigma-Aldrich, St. Louis, MO) and incubated for 24 h at 37°C in a microaerobic chamber with 90%N₂, 5%H₂, and 5%CO₂ gas mixture. Twenty-four hours later, cultures were transferred to 250 ml Erlenmeyer flasks containing Brucella broth supplemented with 5% fetal bovine serum and incubated for an additional 48 h. Immediately prior to use, cultures were observed microscopically for viability. Experimental mice were intra-gastrically gavaged at 23 and 25 days of age with 10⁸ bacteria suspended in 0.5 ml sterile Brucella broth.

DNA was extracted from cecal contents to determine the relative abundance of H. hepaticus in rederived mice using qPCR analysis and previously described Helicobacter hepaticus primers (Beckwith et al., 1997). Briefly, samples were assayed in triplicate in 10 μL reactions that each contained 0.25 μL of each primer, 5 μL QuantiTect SYBR Green PCR Master Mix (BioRad, Hercules, CA), 2.75 μL nuclease-free water, and 2 μL sample DNA. Reactions were amplified in a real-time thermocycler (C1000 Touch Thermal Cycler CFX384 Real Time system, BioRad) using the following parameters: incubation at 95°C for 15 min, 15 s of denaturation at 94°C (40 cycles), 20 s of annealing at 58°C, and 30 s of extension at 72°C. Fluorescence was monitored at the end of each extension phase (BioRAD CFX Manageer 3.1, Biorad). Technical replicates were averaged and fold change of each GM group was compared to H. hepaticus expression in the GMCRL group.

At 111 ± 2 days of age, experimental mice were euthanized by CO₂ asphyxiation, and necropsy was performed. The entire gastrointestinal tract was removed and the colon length from cecal attachment to anus was measured. The cecum was incised longitudinally, and cecal contents were removed and placed in a sterile 2 ml round bottom tube. Samples were flash frozen in liquid nitrogen and stored at −80°C until processed for DNA extraction. Cecal tissue was rinsed with sterile saline and laid flat, serosal surface down, on a small section of white note card. The colon was removed and rinsed with sterile saline. Tissues were immersed in 10% buffered formalin for 48 h and processed for hematoxylin and eosin staining. Histological evaluation was performed by a rodent pathologist, blinded to the identity of the animal's experimental group.

Tissues were evaluated for disease using a scoring system previously adapted by our lab (Alvarado et al., 2015). Detailed evaluation of the severity of epithelial hyperplasia (0, none; 1, mild; 2, moderate, 3, severe with crypt branching, and/or herniation; 4, dysplasia) and inflammatory changes (0, no inflammation 1, mild inflammation limited to the mucosa; 2, moderate inflammation limited to the mucosa and submucosa; 3, severe inflammation with obliteration of normal architecture, erosions, crypt abscesses; and 4, severe inflammatory changes with ulceration). In addition, a separate score for the longitudinal extent of epithelial hyperplasia and inflammation was assigned (0, no significant changes; 1, one or two foci occupying <10% of the mucosa; 2, multifocal lesions occupying 10–60% of the mucosa; 3, diffuse lesions occupying >60% of the mucosa). The final lesion score was calculated as (hyperplasia score × longitudinal extent of hyperplasia score) + (inflammation score × longitudinal extent of inflammation score). Total disease scores ranged from 0 (no disease) to 24 (severe disease).

DNA extraction was performed as previously described (Ericsson et al., 2015b). Briefly, cecal contents were thawed at room temperature and a sterile 0.5 cm diameter stainless steel bead and 800 μl of lysis buffer were added to the 2 ml round-bottom tube. Samples were mechanically disrupted using a TissueLyser II (Qiagen, Venlo, Netherlands) for 3 min at 30 Hz, followed by incubation at 70°C for 20 min with periodic vortexing. Samples were centrifuged at 5,000 × g for 5 min, and the supernatant was transferred to a sterile 1.5 ml Eppendorf tube containing 200 μl of 10 mM ammonium acetate. Lysates were vortexed, incubated on ice for 5 min, and then centrifuged. Supernatant was transferred to a sterile 1.5 ml Eppendorf tube and one volume of chilled isopropanol was added. Samples were incubated on ice for 30 min and centrifuged at 16,000 × g at 4°C for 15 min. The DNA pellet was washed with 70% ethanol and resuspended in 150 μl Tris-EDTA, followed by addition of 15 μl of proteinase K and 200 μl AL buffer (DNeasy Blood and Tissue kit, Qiagen). Samples were incubated at 70°C for 10 min and 200 μl of 100% ethanol was added to the tubes. Samples were mixed by gentle pipetting and the contents transferred to a spin column from the DNeasy kit (Qiagen). The DNA was further purified following the manufacturer's instructions and eluted in 200 μl EB buffer (Qiagen). DNA concentrations were determined fluorometrically (Qubit dsDNA BR assay, Life Technologies, Carlsbad CA) and samples were stored at 20°C until sequencing.

Library construction and sequencing was performed at the University of Missouri DNA Core. Bacterial 16SrRNA amplicons were generated in a multiplexed (96-well) format using amplification of the V4 hypervariable region of the 16S rRNA gene, and then sequenced on the Illumina MiSeq platform as previously described (Ericsson et al., 2015b). Samples returning >10,000 reads were deemed to have successful amplification.

Assembly, binning, and annotation of DNA sequences were performed at the University of Missouri Informatics Research Core Facility. Briefly, contiguous DNA sequences were assembled using FLASH software (Magoc and Salzberg, 2011), and culled if found to be short after trimming for a base quality <31. Qiime v1.8 (Kuczynski et al., 2011) software was used to perform de novo and reference-based chimera detection and removal, and remaining contiguous sequences were assigned to operational taxonomic units (OTUs) using a criterion of 97% nucleotide identity. Taxonomy was assigned to selected OTUs using BLAST (Altschul et al., 1997) against the Greengenes database (DeSantis et al., 2006) of 16SrRNA sequences and taxonomy.

Statistical analysis was performed using Sigma Plot 13.0 (Systat Software Inc., Carlsbad CA). Differences in lesion score, phylum, and OTU relative abundance were determined using two-way ANOVA with sex and GM profile (i.e., GMJAX, GMTAC, GMCRL) as the independent variables, followed by Student Newman-Keuls post hoc test. To account for multiple testing, OTUs with a p < 0.01 were considered statistically significant. Statistical differences in Shannon index, colon length, and relative abundance of H. hepaticus following quantitative PCR were first evaluated using two-way ANOVA to assess main effects and interactions between sex and GM profile. No main effects were detected between male and female mice in any of those values, thus data from male and female mice were pooled and all subsequent analyses was performed using one-way ANOVA with Student Newman-Keuls post hoc test. Tests with a p < 0.05 were considered statistically significant. For gut microbiota analysis, OTUs with <10,000 reads were excluded from the data set. Bar graphs were generated with Microsoft Excel (Microsoft, Redmond WA) and principal component analysis (PCA) was generated using Paleontological Statistics Software Package (PAST) 3.12 (Hammer et al., 2001). All groups were visually inspected for descriptive analysis of consistency between animals (bar graphs) or clustering of animals within groups (PCAs). Statistical testing for differences in beta-diversity was performed via PERMANOVA, implemented using PAST 3.12.

In order to determine the impact of differing complex GM communities on disease severity, B6 IL-10^−/− embryos were transferred to surrogate dams purchased from three different vendors (Figure S1). Surrogate dams were allowed to naturally deliver and raise pups. To induce disease, weanling pups were inoculated with Helicobacter hepaticus at 24 and 26 days of age. Experimental mice were euthanized at 111 days of age and the cecum and colon were removed for gross evaluation of colon length and histopathologic assessment of disease severity. Grossly, mice rederived in surrogate dams harboring Taconic GM (hereafter referred to as GMTAC) had significantly decreased colon length when compared to mice rederived in surrogate dams with either GMCRL (Charles River Laboratories) or GMJAX (The Jackson Laboratory) microbiota profiles (Figure 1A). To further evaluate disease severity, histopathologic scoring of lesions in the colon and cecum was performed. Rederived B6 IL-10^−/− mice harboring GMCRL had significantly lower mean lesion scores in both the colon (Figure 1C) and cecum (Figure 1E) as compared to mice colonized with GMJAX and GMTAC. In addition, mice rederived to surrogate dams with GMJAX had lower cecal lesion scores than mice rederived to surrogate dams with GMTAC. No differences in lesion severity were detected between male and female mice in any of the GM groups examined.

To determine if these findings were unique to the IL-10^−/− model on a C57BL/6 genetic background, C3H IL-10^−/− embryos were implanted in surrogate dams from the same vendors as used in the B6 IL-10^−/− experiment. In the C3H IL-10^−/− model, mice rederived in surrogate dams with either GMJAX or GMTAC had significantly shorter colon length than mice rederived in surrogate dams with GMCRL (Figure 1B), while no statistical difference in colon length was found between the GMJAX and GMTAC groups. As in the B6 genetic background, mice rederived in surrogate dams with GMCRL had significantly lower mean lesion scores in both the colon (Figure 1D) and cecum (Figure 1F) than either the GMJAX or GMTAC groups, with differences in mean lesion score also detected between mice harboring GMJAX and GMTAC. Again, no differences were found in lesion severity between males and females in the GM groups examined.

To evaluate the contribution of the GM in the differential disease severity observed between mice, cecal contents were characterized via sequencing of the V4 hypervariable region of the 16S rRNA gene. Decreases in microbial diversity have been associated with susceptibility to inflammation and may be indicative of IBD disease severity (Ott et al., 2004; Manichanh et al., 2006; Ott and Schreiber, 2006; Frank et al., 2007). In the B6 IL-10^−/− model, mice rederived in surrogate dams with GMCRL had a statistically higher mean Shannon index, indicating a more diverse cecal microbial composition, than either the GMJAX or GMTAC groups (Figure 2A). In addition, the GMTAC group had a greater mean Shannon index than the GMJAX group. However, in the C3H IL10^−/− model, the GMCRL group had a lower mean Shannon index than either the GMJAX or GMTAC groups respectively (Figure 2B). As pups of either genetic background were exposed to similar maternal microbes at birth, these differences in microbial diversity following disease induction suggest that there is an interaction between host genetic background and GM profile in shaping microbial diversity.

To evaluate β-diversity among rederivation groups, principal component analysis (PCA) was performed. In PCA, samples that are similar in microbial composition cluster together. In the B6 IL-10^−/− model, PCA demonstrated that mice harboring GMCRL and GMJAX clustered independently, indicating that these are distinct microbial populations (Figure 2C, Table 1). In addition, these two groups demonstrate tight within group clustering indicating that animals within individual groups were highly similar. In contrast, the GMTAC group demonstrated greater inter- and intra- group variability along PC1 and PC2 as compared to GMCRL and GMJAX groups, indicating a less uniform microbial composition in those mice. PCAs were also examined in the context of lesion scores and no differences in clustering patterns of high or low lesions scores were noted within any of the GM profile groups (data not shown). In the C3H IL-10^−/− model, independent clustering of GMCRL was noted while GMTAC and GMJAX clustered closely together indicating greater compositional similarity between the latter two groups (Figure 2D, Table 1). The GMCRL and GMTAC groups demonstrated tight intra-group clustering along PC2, with GMJAX mice having increased intra-group variability along PC2.

To identify specific microbial taxa that correlate with resistance or susceptibility to disease, relative abundance of the microbiota at both the phylum and OTU level was compared between GM profiles. In the B6 IL-10^−/− model, the phyla Deferribacteres and TM7 were enriched in mice of the GMCRL group as compared to mice in GMJAX and GMTAC groups (Figure 3A, Table S1). In the GMJAX and GMTAC groups, bacteria in the phylum Proteobacteria were present in greater relative abundance in mice rederived in GMTAC surrogate dams. Conversely, in the C3H IL-10^−/− model, a greater relative abundance of the phylum Proteobacteria was detected in mice of the GMCRL group (Figure 3B, Table S1) as compared to mice in the GMJAX and GMTAC groups. Moreover, the phylum Deferribacteres was detected at greater relative abundance in the GMTAC group as compared to GMCRL and GMJAX groups.

Resolved to the level of OTU, differences were detected in the relative abundance of several OTUs detected in all GM profiles (Figure 4A, Figure S5A, Table 2). In the B6 IL10^−/− model, several differences were found between GMCRL and the GMJAX and GMTAC groups; specifically, greater relative abundances of family Rikenellaceae, order Bacteroidales, Mucispirillum schaedleri, family Clostridiaceae, family Ruminococcaceae, order RF39, and family F16 were detected in mice colonized with the GMCRL profile (Table 2). In contrast, mice with the GMJAX profile had a higher relative abundance of Bacteroides acidifaciens and genus Clostridium. Similarly, in the C3H IL10^−/− model, the GMCRL profile had significantly higher levels of family Mogibacteriaceae, family Christensenellaceae, family Clostridiaceae, genus Sutterella relative to the other GM profiles (Figure 4B, Figure S5B, Table 3). In contrast, the GMJAX group had higher relative abundance of family Ruminococcaceae than either the GMCRL or GMTAC groups. Additionally, the GMTAC group both had higher relative abundance of Mucispirillum schaedleri, relative to GMJAX. Interestingly, the observed differences in the relative abundance of OTUs between B6 and C3H strains suggests a differential effect on colonization, as embryos from both strains were rederived on the same day using the same pools of surrogate dams.

While differential disease susceptibility may be a function of differential abundance of specific taxa, it may also be explained by the presence or absence of select taxa unique to each GM profile. Interestingly, in B6 IL10^−/− mice we found several unique OTUs including Alistipes massiliensis, Roseburia faecis, Clostridium saccharogumia, and genus Megamonas were found in mice rederived in GMCRL dams, whereas Eubacterium dolichum was detected only in the GMTAC rederived group. Aggregatibacter pneumotropica was only found in GMJAX and GMTAC groups with Propionibacterium acnes and genus Phascolarctobacterium detected in only GMCRL and GMJAX groups; Candidatus Arthromitus (segmented filamentous bacterium), genus Anaerotruncus and Oxalobacter formigenes were detected in both the GMTAC and GMCRL groups.

In contrast, in the C3H IL10^−/− mice, the GMCRL group harbored Prevotella copri, Alistipes massiliensis, family Paraprevotellaceae, Enterococcus casseliflavus, Faecalibacterium prausnitzii, genus Phascolarctobacterium, genus Fusobacterium, genus Anaerobiospirillum, genus Prevotella, and genus Anaerotruncus. Parabacteroides distasonis, family Porphyromonadaceae, genus Enterococcus, genus Roseburia, family Desulfovibrionaceae, and Flexispira rappini were found to overlap between the GMJAX and GMTAC groups; genus Turicibacter and family Peptostreptococcaceae were detected in both the GMJAX and GMCRL groups; genus Anaerotruncus and Candidatus Arthromitus were specific to the GMTAC and GMCRL groups. Differences in the presence of these OTUs and their interaction with other OTUs present at significantly different levels may account for observed differences in disease severity.

Helicobacter spp. are thought to be a provocateur of intestinal inflammation and dysbiosis (Fox, 2007; Jergens et al., 2007) and as such are necessary for the induction of disease in the IL-10^−/− model. To determine if the observed differences in lesion score among GM groups could be due to differences in H. hepaticus colonization, we performed semi-quantitative real time PCR on cecal bacterial contents. In both B6 IL-10^−/− (Figure 5A) and C3H IL-10^−/− (Figure 5B) mice, mice colonized with GMCRL had significantly lower relative abundance of H. hepaticus than mic colonized with GMJAX or GMTAC. All mice received the same H. hepaticus inoculum post-weaning and are genetically identical, suggesting differences between GM profiles resulted in differing colonization resistance against H. hepaticus.

Because the initial complex microbiota targeted rederivation (CMTR) was performed using surrogate dams with varying genetic backgrounds, it is possible that differences in maternal care, in utero environment, or epigenetic factors contributed to the observed differential disease severity. To address this consideration, we established breeding colonies of CD1 dams harboring either the GMCRL, GMJAX, or GMTAC profiles. To do so, CMTR was used to transfer CD1 embryos to three surrogate dams (Crl:CD1, C57BL/6J, and C57BL/6NTac), and thus generate pups with the previously studied GM profiles (Figure S2). Offspring were then used to establish three separate breeding colonies using an outbred mating scheme. Colonies were monitored for two generations using next generation sequencing to confirm expected differences in GM composition and complexity. PCA of first and second generation females at 8–10 weeks of age confirmed three distinct GM profiles (p ≤ 0.001; F = 12.19; Figure S3). Second generation females were used as surrogate dams for CMTR of the B6 IL-10^−/− model (Figure S4). As in the previous studies, pups were inoculated with H. hepaticus at 24 and 26 days of age and disease severity was assessed at 111 days of age. Grossly, B6 IL-10^−/− pups harboring GMCRL had a significantly greater colon length than the GMJAX or GMTAC groups (Figure 6A). In addition, GMCRL-colonized pups had lower cecal and colonic histopathologic lesion scores compared to GMJAX and GMTAC groups, in agreement with findings from studies using inbred surrogate dams (Figures 6B,C).

Again confirming the previous findings in the B6 IL-10^−/− model, sequencing of cecal contents revealed that the GMCRL group had increased diversity when compared to the GMJAX and GMTAC groups (Figure 7A). Moreover, the GMTAC group was more diverse than the GMJAX group. PCA demonstrated that each group maintained three distinct microbial profiles (Figure 7B, Table 1). To further investigate GM composition, relative abundance at the phylum and OTU level were examined. As in our initial experiments using the B6 IL-10^−/− model, there was increased relative abundance of phylum Proteobacteria in both the GMJAX and GMTAC groups (Figure 8A, Table S1). In contrast to our previous findings, subjectively there was lower relative abundance of Deferribacteres in GMJAXmice, with increased relative abundance observed in GMTAC males. At the OTU level, we found several OTUs which differed between groups and had similar relative abundance patterns as seen in the previous study including B. acidifaciens, family Rikenellaceae, order Bacteroidales, Mucispirillum schaedleri, family Clostridiaceae, genus Clostridium, and family F16 (Figure 8B, Figure S5, Table 4). In addition, we found several species within the phyla Bacteroidetes, Firmicutes, and Proteobacteria that differed between groups including genus Bacteroides, family Mogibacteriaceae, family Christensenellaceae, genus rc4-4, order Clostridiales, family Enterobacteriaceae, and genus Helicobacter.

To determine if the observed differences in lesion score among GM groups could be explained by differences in H. hepaticus colonization, we performed semi-quantitative real time PCR on cecal contents. As observed in our previous experiments, we found that B6 IL-10^−/− mice with GMCRL were colonized with H. hepaticus to a lesser degree than mice colonized with either GMJAX or GMTAC profiles (Figure 9).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.