It has long been recognized that the mammalian gut microbiota has a role in the development and activation of the host immune system. Much less is known on how host immunity regulates the gut microbiota. Here we investigated the role of adaptive immunity on the mouse distal gut microbial composition by sequencing 16 S rRNA genes from microbiota of immunodeficient Rag1−/− mice, versus wild-type mice, under the same housing environment. To detect possible interactions among immunological status, age and variability from anatomical sites, we analyzed samples from the cecum, colon, colonic mucus and feces before and after weaning. High-throughput sequencing showed that Firmicutes, Bacteroidetes and Verrucomicrobia dominated mouse gut bacterial communities. Rag1− mice had a distinct microbiota that was phylogenetically different from wild-type mice. In particular, the bacterium Akkermansia muciniphila was highly enriched in Rag1−/− mice compared with the wild type. This enrichment was suppressed when Rag1−/− mice received bone marrows from wild-type mice. The microbial community diversity increased with age, albeit the magnitude depended on Rag1 status. In addition, Rag1−/− mice had a higher gain in microbiota richness and evenness with increase in age compared with wild-type mice, possibly due to the lack of pressure from the adaptive immune system. Our results suggest that adaptive immunity has a pervasive role in regulating gut microbiota’s composition and diversity.
The mammalian gut is one of the most densely colonized habitats with trillions of microorganisms known as the microbiota (Ley et al., 2008). The microbiota have coevolved with the host and have a key role in host’s metabolism and immunity (Hooper et al., 2001; Backhed et al., 2004; Gill et al., 2006; Round and Mazmanian, 2009; Klaenhammer et al., 2012). Imbalance in the gut microbiota composition has been associated with diseases (Ley et al., 2005; Turnbaugh et al., 2008; Vijay-Kumar et al., 2010; Blaser, 2010). External factors, such as diet and antibiotics, have been shown to have an important role in shaping gut microbiota. For example, transition from milk to solid food in infants correlated an increase in Bacteroidetes (Koenig et al., 2010). In addition, gnotobiotic mice transplanted with human gut microbiota can be modulated with the diet (Turnbaugh et al., 2009). Antibiotic use, on the other hand, has been shown as a key factor to the perturbation of gut microbiota (Young and Schmidt, 2004; Dethlefsen et al., 2008).
Compared with modulation of gut microbiota by external factors, information is limited on how host immunity regulates its gut microbiota (Willing et al., 2011; Brown et al., 2013). Recent findings suggest an important role for the innate immune system. In particular, toll-like receptor (TLR5), which recognizes the bacterial flagellin to activate innate immune response, has been shown to alter colonic microbiota at the species level (Vijay-Kumar et al., 2010). In contrast, TLR2 and TLR4, which mainly recognize peptidoglycan and lipopolysaccharide, respectively, have no apparent effect on gut microbiota (Loh et al., 2008). In addition, deficiency of MyD88, a signaling adapter required for all toll-like receptors except TLR3, has been reported to alter gut bacterial diversity with an expanded population of segmented filamentous bacteria in the small intestine (Larsson et al., 2011). In non-obese diabetic mice, MyD88 deficiency was further shown to change gut microbiota that conferred protection against developing Type 1 diabetes (Wen et al., 2008). Another example of microbiota-regulating component of innate immunity is Interleukin-22-producing innate lymphoid cells, which were shown to selectively control colonization of Alcaligenes species to prevent systemic inflammation (Sonnenberg et al., 2012).
It has been hypothesized that the adaptive immune system evolved in vertebrate animals in order to memorize complex assemblages of beneficial microbiota (McFall-Ngai, 2007). It is thus conceivable that deficiencies in the host adaptive immune system would result in an altered microbiota. Indeed, such deficiency can lead to a markedly elevated level of bacterial flagellin (Cullender et al., 2013), which is normally kept low presumably by host immunity. In addition, mice deficient in recombination-activating gene 2 (Rag2) or activation-induced cytidine deaminase had increased segmented filamentous bacteria in the small intestine (Suzuki et al., 2004). Both mutations lead to the lack of IgA (immunoglobulin A) in the intestine, which provides ‘an immunological buffer’ between the host and microbiota (Brown et al., 2013). Besides IgA (Cebra, 1999; Suzuki et al., 2004; Macpherson and Uhr, 2004; Suzuki and Fagarasan, 2008), T-cell-mediated responses are also hypothesized to shape microbiota in the intestine but evidence is scarce. Importantly, the role of mature lymphocytes on the composition and diversity of gut microbiota is not known.
Microbiota vary with host age (Palmer et al., 2007; Koenig et al., 2010), diet (Cani et al., 2007) and luminal versus mucosal localization (Zoetendal et al., 2002; Eckburg et al., 2005). Environmental gradients created by water absorption and pH variability along the distal intestine could also affect microbiota composition (DiBaise et al., 2008). Thus, in this study, we aim to investigate the role of adaptive immunity on mouse distal gut microbiota with a carefully controlled study design that allows us to discern any interactions between age, intestinal location and immunological status. For this purpose, we employed the classic, non-leaky Rag1−/− mouse model that lacks all mature lymphocytes. Both wild-type (Rag1+/+) and mutant (Rag1−/−) mice were given the same sterile diet and water, and housed in the same clean environment. We took samples before and after weaning in order to detect possible interactions between immunological status and age. For each immunological status and age, we analyzed microbiota compositions from multiple anatomical sites including cecum, colon, colonic mucus and feces. Finally, to explicitly investigate the role of mature lymphocytes on microbiota, we performed adoptive transfer to Rag1−/− mice using Rag1+/+ bone marrow that replenished the adaptive immune system.
Materials and methods
Both Rag1+/+ and Rag1−/− mice were of C57BL/6 background. After being purchased from The Jackson Laboratory (Bar Harbor, ME, USA), both strains were housed in the same room and on the same shelf of a maximum barrier, specific pathogen-free facility at College of Veterinary Medicine, Virginia Tech, Blacksburg, VA, USA. To ensure adaptation to identical housing environment, both strains were bred in-house for at least three generations with sterilized individually ventilated caging, and received sterilized drinking water and feed (2918-Irradiated Teklad Global 18% Protein/6% Fat Rodent Diet, Harlan Laboratories, Indianapolis, IN, USA) prior to the initiation of the study. The same housing condition was used in the entire study. For adoptive transfer, the bone marrow of 6-week-old female Rag1+/+ or Rag1−/− mice was injected intravenously into age-matched female Rag1−/− mice of the same litter. Using recipient mice of the same litter was intended to reduce the confounding effect of maternal microbiota transmission on pups’ gut flora (Ubeda et al., 2012). The experiment was repeated once to achieve a sample size of six in each group. Both litters of recipient mice were from the same Rag1−/− dam. Briefly, total bone marrow cells were collected from femurs and tibias of Rag1+/+ or Rag1−/− mice, removed of red blood cells, washed and resuspended in sterile phosphate-buffered saline. Recipient Rag1−/− mice from the same litter were anesthetized with isoflurane carried in oxygen and retro-orbitally injected with Rag1+/+ or Rag1−/− bone marrow cells (2 × 106 cells in 150 μl phosphate-buffered saline per mouse). The recipient mice were housed for 8 weeks until sample collection at 14 weeks of age. To retain gut microbiota, the standard procedure of antibiotics treatment in bone marrow transplantation was omitted.
Microbiota sampling, DNA extraction and PCR
Microbiota samples were collected from cecum, colon and colonic mucus within 20 mins after euthanasia. Cecal content was collected by manual extrusion using a sterile pipette tip. The colon was carefully opened longitudinally with a pair of sterile scissors. Colonic content was gently picked up with tweezers. Colonic mucus was then collected by gently scraping off the colonic wall. To avoid cross-contamination, each microbiota sample was collected using a new pair of sterile tweezers. For T2 only, feces were picked up directly from individual cages. (For T1, pups were housed in the same cage with their mothers, making the collection of feces from individual pups impossible.) For adoptive transfer, microbiota samples were collected from the lumen of the colon. All samples were stored at −80 °C till being processed at the same time. DNA was extracted with the QIAamp DNA Stool Mini Kit (Qiagen, Valencia, CA, USA). The V4 region (252 bp) of 16 S rRNA gene was PCR-amplified with 515 F and 12-base GoLay barcoded 806 R primers (Caporaso et al., 2012). The purified amplicons were sequenced with a MiSeq sequencer (Illumina, San Diego, CA, USA).
Taxonomy assignments and community structure analyses
Sequencing reads were processed with Quantitative Insights Into Microbial Ecology (QIIME) (Caporaso et al., 2010). High-quality reads with Phred score ⩾20 (corresponding to an sequencing error rate ⩽0.01) were first checked for chimeras with USEARCH61 (Edgar, 2010). Chimeric sequences were removed from further analysis. Nonchimeric sequences were mapped into operational taxonomic units (OTUs) against the Greengene reference sequences (version 2013.5) with the program UCLUST (Edgar, 2010). Bacterial taxonomy was assigned by using a naive Bayes classifier (Wang et al., 2007) trained with the Greengenes taxonomy with over 200 000 reference sequences (McDonald et al., 2012). Abundant taxa were classified to the species level by using BLAST (Altschul et al., 1997) against the Greengenes data set. A phylogenetic tree was constructed (Price et al., 2010) from PyNAST-aligned sequences representing each OTU. For beta diversity calculations, all microbiota were randomly sampled with 13 000 sequences to achieve even sampling. Principle coordinates were calculated from unweighted UniFrac metrics (Lozupone and Knight, 2005). Alpha diversity metrics, including observed species defined as OTUs with 97% or higher similarity, phylogenetic diversity (Faith, 1992) and Shannon index, and equitability were calculated using QIIME.
Nucleotide accession numbers
Sequences determined in this study were deposited in the Metagenomics RAST server (Meyer et al., 2008) under the accession number 4547766.3.
Total bacteria measurement by real-time PCR
Total bacteria were measured by using real-time PCR with domain-specific primers F340 and R514 (Wlodarska et al., 2011). Five-point standard curves were generated by using Lactobacillus rhamnosus (ATCC 7469) genomic DNA. Real-time PCR was performed using iTaq Universal Supermixes (Bio-Rad, Hercules, CA, USA) on an Applied Biosystems (Foster City, CA, USA) 7500 cycler with the program: one cycle at 95 °C for 5 mins, followed by 40 cycles of 94 °C for 15 s and 63 °C for 45 s.
Flow cytometry analysis
Mononuclear cells were isolated from the spleen and colonic lamina propria as described previously (Hur et al., 2012). Cells after staining with fluorochrome-labeled antibodies (eBioscience, San Diego, CA, USA) were analyzed with the FACSAria (BD Biosciences, San Jose, CA, USA).
The significance of microbiota grouping by Rag1 status and by age was determined by permutational multivariate analysis methods PERMANOVA and PERMDISP (Anderson, 2001). Unpaired Student’s t-test was performed in Microsoft Excel 2011. Mann–Whitney U-test, one-way and two-way analysis of variance were performed by using Prism (GraphPad, La Jolla, CA, USA) and R version 3.0.2.
Rag1 status as a source of variation in gut microbiota community structure
We compared Rag1+/+ and Rag1−/− mice of the same genetic background to determine the effect of Rag1 and adaptive immune system on gut microbiota. Mice lacking the adaptive immune system are often housed at commercial vendors in maximum barrier facilities. Wild-type mice, on the other hand, are usually raised in standard barrier rooms. This may result in distinct baseline microbiota for these strains, known as the cage effect. Several experimental approaches could be used to eliminate the cage effect as a potential confounding factor in our study. The first option was to breed heterozygous (Rag1+/−) animals and compare Rag1+/+ and Rag1−/− littermates. The second option was to co-house Rag1+/+ and Rag1−/− mice to allow coprophagy, which could be used to equilibrate gut microbial diversity in co-housed animals (Antonopoulos et al., 2009). The third option was to adapt Rag1+/+ animals to the same housing conditions as Rag1−/− mice (maximum barrier and sterilized caging, water and feed). Because we planned to involve preweaning pups and their gut microbiota might be affected by immunoglobulins in milk and placenta, the third option was chosen as it restricted Rag1−/− pups to Rag1−/− dams (that is, no immunoglobulins from milk or placenta). Therefore, herein we will describe the changes of gut microbiota in the absence of the entire adaptive immunity that includes maternal immunoglobulins.
Microbiota samples from cecal, colonic and mucosal contents at two different ages, 14 days post birth (T1) and 28 days postweaning (T2), were analyzed. For T2 when mice were individually caged, fecal samples were also collected and analyzed. High-throughput sequencing of the 42 samples yielded 9 342 844 reads. After quality trimming and chimera removal, the final data set contained 7 330 205 nonchimeric high-quality sequences and was used for taxonomy assignments and diversity analyses.
Differences between microbiota samples were calculated by using the UniFrac metrics, which measures phylogenetic dissimilarities between microbial communities (Lozupone and Knight, 2005). Principal coordinate analysis based on UniFrac metrics showed a separation of Rag1+/+ and Rag1−/− samples along the first two axes that explained 17% and 13% of data variation, respectively (Figure 1a). Rag1 status as a significant source of variation was confirmed by a nonparametric permutational multivariate analysis, PERMANOVA (P<0.001). We further confirmed that Rag1+/+ and Rag1−/− samples differed because of their coordinates and not their relative dispersions by the PERMDISP test (P=0.438) that showed the two groups had similar dispersion. In contrast, although samples from two different ages differ from each other based on the PERMANOVA test (P<0.001), dispersion was a significant factor that caused this difference (PERMDISP, P<0.001). Samples from different anatomical locations of the intestine were not significantly different (P=0.099). Importantly, the average UniFrac distance between Rag1+/+ and Rag1−/− samples was larger than that within Rag1+/+ samples (Figures 1b, P<0.0001), indicating significantly greater phylogenetic difference between Rag1+/+ and Rag1−/− samples than within Rag1+/+ themselves. Likewise, larger average UniFrac distance between T1 and T2 samples than within T1 samples supported change of communities with sampling age (Figures 1c, P<0.0001).
Bacterial taxonomy comparisons
Bacterial taxa showed variations with treatment, age and anatomical site (Figure 2a). Firmicutes, Bacteroidetes and Verrucomicrobia, collectively representing over 90% of total sequences in each subject, dominated the gut microbiota. Actinobacteria, Proteobacteria and Tenericutes were present as minor constituents. The major components at the taxonomic Order level were Clostridiales, Lactobacillales, Bacteroidales and Verrucomicrobiales (Figure 2a). It is notable that total bacterial numbers did not vary significantly with immunodeficiency (Supplementary Figure S1). Regardless of Rag1 status, 1 g of cecal or colonic content contained 1010–1011 copies of 16 S rDNA. One gram of mucus had 107.5−109.5 copies of 16 S rDNA. One gram of feces contained 108.6−1010 copies of 16 S rDNA.
Several taxa were markedly affected by Rag1 deficiency. Rag1−/− mice had significantly lower colonic Lactobacillales and Enterobacteriales than the controls had at T1 (Figure 2b). The phenomenon was absent at T2 (Supplementary Table S1), suggesting that the association was age-dependent. At T2, Rag1−/− mice had significantly more Verrucomicrobiales (P<0.05) in colonic content and feces compared with Rag1+/+ mice (Figure 2b). The same trend was observed at T1 for cecal content, colonic content and colonic mucus (Figure 2a, shown in pink color), but the difference was not statically significant (Supplementary Table S1) due to variations within the Rag1−/− group.
The age of the mice also affected microbiota taxa. Lactobacillales decreased from T1 to T2 in the colon of Rag1+/+ mice (Figure 2c), consistent with previous reports (Pantoja-Feliciano et al., 2013) and due to the diet change from milk to solid food before and after weaning. Because Rag1−/− colons did not have much Lactobacillales to begin with (Figure 2b), the decrease from T1 to T2 was less obvious (Figure 2a and Supplementary Table S2). Analysis of other major components such as Clostridiales and Bacteroidales also revealed changes over time (Supplementary Table S2), reflecting major changes in microbiota before and after weaning. In particular, a shift from Bacteroidales to Clostridiales was noted in colonic mucus of Rag1−/− mice (Figure 2c). The lower level of Bacteroidales in Rag1−/− mice was consistent with the reported higher counts of bacterial flagellin in Rag1−/− mice than wild-type mice (Cullender et al., 2013), as Bacteroidales do not have flagellin (Lozupone et al., 2012). In Rag1−/− mice, a decrease in the abundance of Verrucomicrobiales with age was also observed (Figure 2a) but it was not statistically significant (Supplementary Table S2).
Analysis of different anatomical sites showed distinct distribution patterns of the three abundant taxa. Based on a weight ratio of 2.3:1 for cecal versus colonic contents (total weight minus wall weight in rodents, estimated according to Pan et al. (2009) and Campbell et al. (1997)) and relative abundance of taxa, we calculated the bacterial biomass distribution between cecal and colonic contents (Figure 2d). The cecum had consistently higher Bacteroidales and Clostridiales than the colon. However, Lactobacillales, which were more abundant before weaning (Figure 2c), concentrated in the colon (Figure 2d). Immunological status (Rag1+/+ or Rag1−/−) had no effect on the anatomical distribution of the three most abundant taxa. Members of Verrucomicrobiales, on the other hand, were most abundant in the colonic content (Figure 2a and Supplementary Table S3).
Analysis of microbial community diversity
We investigated microbial diversity as a function of immunological status, age and location in the intestine. Microbiota diversity was assessed for both richness (species abundance) and evenness (species distribution). Richness was measured as the number of observed species, phylogenetic diversity (Faith, 1992) and the Shannon index. Evenness was measured by equitability. Rag1+/+ and Rag1−/− microbiota had comparable richness and evenness (Figure 3a and Figure 3d). However, all three richness metrics significantly increased with age (Figure 3b). At T1, 331 observed species were identified, and this number increased (30%) to 427 at T2. The phylogenetic diversity, which adds total branch lengths from a site and thus reflects evolutionary divergence of different species, was 28% higher at T2 than T1. The Shannon index also showed the same trend as observed species and phylogenetic diversity. Importantly, we detected an interaction between age and Rag1 status on the Shannon index (two-way analysis of variance, P=0.028 for the interaction), which was higher for Rag1−/− than Rag1+/+ at T2 but not T1 (Supplementary Figure S2). The evenness also increased with age (Figure 3e). Because equitability (E) and Shannon index (H) are mathematically related (E=H/log2(Sobs), where Sobs is the number of observed species), an interaction between age and Rag1 status was again observed (two-way analysis of variance, P=0.048 for the interaction; Supplementary Figure S2). These results suggest that Rag1−/− mice had a higher gain in microbiota diversity with increase in age compared with wild-type mice, possibly due to the lack of control from the adaptive immune system. Interestingly, the colonic mucus had significantly higher richness and evenness than the colonic content and contained the most diverse microbial community among the four anatomical sites (Figures 3c and f).
Regulation of Akkermansia muciniphila colonization by adaptive immunity
Because the abundance of Verrucomicrobiales was higher in mice lacking the adaptive immune system (Figures 2a and b and Supplementary Table S1), we decided to further analyze this lineage. In our data set, a total of 117 Verrucomicrobiales-affiliated OTUs were classified at the species level as A. muciniphila (Supplementary Figure S3 and Supplementary Table S4), which represented 99.996% of all sequences in the Verrucomicrobia phylum (403 717 sequences affiliated with the 117 OTUs versus a total of 403 733 sequences in the phylum; the rest 16 sequences were unclassified species in Prosthecobacter (four sequences), Luteolibacter (four sequences), Chthoniobacter (one sequence) or unclassified genera (seven sequences)). The distances between these OTUs were generally less than 5%, as shown in a neighbor-joining phylogenetic tree (Supplementary Figure S4). Importantly, A. muciniphila was enriched in Rag1−/− samples compared with significantly lower abundance in Rag1+/+ samples (Figure 4a, Mann–Whitney U-test, P<0.0001). In order to explicitly investigate the effect of adaptive immunity on A. muciniphila, we performed transfer of bone marrow from Rag1+/+ to Rag1−/− mice. Adoptive transfer from Rag1−/− to Rag1−/− was also performed as the control. Pre- and post-transfer antibiotics were intentionally avoided while recipient mice were closely monitored. As expected, Rag1+/+ bone marrow replenished the spleen and large intestine of Rag1−/− mice with mature T and B cells (Figure 4b). A majority of engrafted T cells were CD4+ using the transplantation method. Both early and mature B cells (both B220+) were present, as the recipient mice contain endogenous early-stage B cells that are not V(D)J rearranged. As the colonic content had the highest percentage of Verrucomicrobiales in postweaning adult mice (Figure 2a and Supplementary Table S3), colonic contents were collected from recipient mice on which sequencing was performed. A total of 12 samples yielded 48 991 nonchimeric high-quality sequences. As observed in the colonic content of adult Rag1+/+ and Rag1−/− mice (Figure 2a), Bacteroidales and Clostridiales dominated colonic microbiota of the recipient mice regardless of the donor bone marrow (Figure 4c). However, the colonic microbiota with Rag1+/+ bone marrow transplantation clearly separated from those with Rag1−/− bone marrow transfers on the first principal coordinate PCo1, which explained 34% of variations (Figure 4d). Indeed, the introduction of Rag1+/+ bone marrow into Rag1−/− mice significantly decreased the relative abundance of Bifidobacteriales, Bacteroidales and Erysipelotrichales (Supplementary Figure S5). Importantly, the relative abundance of A. muciniphila in the colon was repressed more than fivefold in Rag1−/− mice receiving Rag1+/+ bone marrow compared with those receiving Rag1−/− bone marrow (Figure 4e), which, together with the observation that the colon of Rag1+/+ contained significantly less A. muciniphila than that of Rag1−/− mice (Figure 4a), suggests possible control of the colonization of A. muciniphila by components of the adaptive immune system that included gut-resident B cells and CD4+ T cells.
Host–microbe interactions are important for health and disease. How mucosal-colonizing commensal bacteria affect host immunity has been a primary focus of attention. The present study aimed to dissect the opposite function, where host immunity is considered the cause, while microbiota the effect. Using the classic Rag1−/− model with non-leaky deficiency in adaptive immunity, we showed that the lack of mature lymphocytes led to decreased representation of Lactobacillales and Enterobacteriales in neonates (Figure 2b), and increased amount of Verrucomicrobiales (A. muciniphila at the species level) in both neonatal and adult mice (Figure 2a, pink color; Figures 2b and 4a). Analysis of microbial diversity showed that the gut microbiota of wild-type and immunodeficient animals had similar community diversity according to the three metrics compared (Figures 3a and d). In addition, contrary to previous prediction (Willing et al., 2011), the total number of bacteria did not change with immunodeficiency when external factors, such as diet and caging, were carefully controlled (Supplementary Figure S1). Adoptive transfer of Rag1+/+ bone marrow, which generated mature B and T cells in the spleen and intestinal lamina propria of postweaning adult Rag1−/− mice (Figure 4b), reversed the phenotype seen in Rag1−/− mice and suppressed the colonization of A. muciniphila in the colon (Figure 4e). Our results suggest that components of the adaptive immune system can directly alter gut microbiota.
Mice lacking adaptive immunity require different husbandry than wild-type mice, and such difference may directly affect gut microbiota. Several experimental approaches could be used to eliminate the cage effect as a potential confounding factor in our study. We chose to adapt Rag1+/+ animals to the same housing conditions as Rag1−/− mice, instead of heterozygous breeding or litter co-housing, to avoid the exposure of Rag1−/− pups to maternal immunoglobulins from Rag1+/− or Rag1+/+ dams. Therefore, we described herein the changes of gut microbiota in the absence of the entire adaptive immunity. The effect of maternal immunoglobulins on the microbiota of Rag1−/− mice will be investigated in the future. It has been noted, however, after the completion of this study, that the combination of housing condition adaption and dam co-housing (prior to breeding) could further equilibrate the baseline microbiota of the two mouse strains. Another feasible approach to eliminate the cage effect could have been to breed heterozygous animals (F0) to generate Rag1+/+ and Rag1−/− littermates (F1). Within F1 mice, siblings with the same homozygous genotype could be then mated to generate the wild-type and knockout animals involved in the analysis.
Deficiencies of IgG and IgA, achieved by deleting an enzyme that mediates class switch recombination, were shown to expand segmented filamentous bacteria in the small intestine (Suzuki et al., 2004). However, their effect on the large intestine was not known. A recent study compared adult fecal microbiota of wild-type versus Rag1−/− mice (Dimitriu et al., 2013), but a direct effect of adaptive immunity on microbiota was not demonstrated through adoptive transfer. In addition, whether the microbiota differences between mouse strains were affected by animal age and sampling site was unknown. Using high-throughput sequencing, we have shown a significant lower level of Lactobacillales and complete absence of Enterobacteriales in the colon of neonatal Rag1−/− mice compared with the wild type (Figure 2b). Decreased colonization of these microbes may affect the integrity of gut microbiota as a physical barrier against enteric pathogens. Intriguingly, a consistent increase of A. muciniphila was found in the colon of neonatal and adult Rag1−/− mice (Figures 2a, b, and 4a), suggesting that the lack of mature lymphocytes in the intestinal tract and/or the lack of immunoglobulin in the gut and milk may allow the overgrowth of this bacterium. We confirmed adaptive immunity as the cause of this change using an adoptive transfer experiment, where Rag1+/+ bone marrow was shown to suppress the colonization of A. muciniphila (Figure 4e). The adoptive transfer experiment also excluded the possibility that maternal immunoglobulins were the only cause of microbiota change, as the recipient mice were postweaning adults. Which components of the adaptive immune system could have altered the gut microbiota is yet to be determined. However, by transferring Rag1+/+ bone marrow to Rag1−/− mice, we showed that restoration of CD4+ T cells and mature B cells in the gut reversed the change of A. muciniphila colonization (Figures 4b and e). Whether the gut-resident lymphocytes affect microbiota through producing cytokines or immunoglobulins will be investigated further. Finally, it was noted that the relative abundance of A. muciniphila in Rag1−/− mice decreased with age (T1: 0.177±0.060; versus T2: 0.042±0.010; P=0.021). The lower level of A. muciniphila in bone marrow transfer experiments (Figure 4e) was likely due to the age effect, as these mice were 14 weeks of age and much older than mice at T1 and T2.
Interestingly, the introduction of Rag1+/+ bone marrow into Rag1−/− mice did not appear to shift the microbiota profile of the recipient mice towards that of Rag1+/+ mice (Supplementary Figure S5) except for the relative representation of A. muciniphila (Figures 4a and e). The lack of resemblance between the gut microbiota of Rag1−/− mice receiving wild-type bone marrow and that of wild-type mice may be due to the fact that only B cells and CD4+ T cells were well engrafted (Figure 4b). Very few CD8+ T cells were present in the spleen and colonic lamina propria of reconstituted Rag1−/− mice. In addition, a lower percentage of dendritic cells expressed CD103 in mice receiving Rag1+/+ bone marrow than those receiving Rag1−/− bone marrow (Supplementary Figure S6). Both CD103+ dendritic cells and CD8+ T cells are important for the integrity of intestinal mucosa (Scott et al., 2011; Fleissner et al., 2010). Therefore, the microbiota changes observed in the bone marrow transplantation experiment may pertain to the effects of B cells and/or CD4+ T cells, and not CD8+ T cells or CD103+ dendritic cells.
The Gram-negative bacterium A. muciniphila has been recently found to be enriched following antibiotics treatment (Dubourg et al., 2013; Hansen et al., 2013) and have the capability to control inflammation (Everard et al., 2013). Initially considered as a host mucin-degrading bacterium, A. muciniphila was found to colonize the mouse gut without consuming much host-derived compounds (Berry et al., 2013). Although the physiological function of this bacterium in immunocompromised hosts is still unclear, our results suggest that A. muciniphila may be used as a biomarker for immunodeficiency. Identification of such biomarker is especially valuable for infants with severe combined immunodeficiency (SCID), commonly known as the ‘bubble boy disease’. Young children without a family history of SCID are often not diagnosed until 6 months old or older. If untreated, the disease is almost always fatal within the first year of age. In addition, immunodeficient infants should not receive live virus vaccinations. Therefore, early diagnosis of SCID, potentially through measuring the abundance of A. muciniphila in the feces of infants not treated with antibiotics, can be beneficial. Whether or not immunodeficient babies indeed have A. muciniphila enriched in their feces will be determined in the future.
Our results also showed that distal gut microbiota’s community structure (Figure 1c) and microbial diversity (Figures 3b and e) differed with animal age. As young pups started to consume solid food, the abundance of Lactobacillales decreased (Figure 2c). In Rag1−/− animals, more Clostridiales and fewer Bacteroidales were identified postweaning. This observation is consistent with the previous report on the expansion of Clostridiales-related segmented filamentous bacteria in mice lacking IgA (Suzuki et al., 2004). We also observed that microbial diversity increased with age (Figures 3b and e), a finding that is consistent with a recent 2.5-year longitudinal monitoring of human infant microbiota (Koenig et al., 2010). Despite the high number of sequences obtained in this study and others, the number of OTUs usually does not reach a plateau, indicating that microbiota have been under-sampled. Thus, the increased diversity at T2 was likely a result of increased representation of rare OTUs that were below the detection limit at T1 but became detectable at T2. It cannot be ruled out, however, that the increase of microbial diversity might have come from acquisition of bacteria in the dust and aerosols as the housing environment was not completely germ-free.
Interestingly, we found that among our four sampling sites, colonic mucus contained the highest microbial diversity (Figures 3c and f). The outer layer of colonic mucus is known to harbor commensal bacteria (Johansson et al., 2008). The higher observed mucosal microbial diversity in our study are likely to be an effect of multiple factors, among which, the availability of renewable mucin glycoprotein reduced axial movement compared with the lumen (Deplancke and Gaskins, 2001), and increased colonization of aerobic bacteria that can utilize oxygen diffused from the tissues surrounding the intestinal lumen (Pedron et al., 2012). Furthermore, we found that Rag1−/− mice had a higher gain in microbiota richness and evenness with increase in age compared with wild-type mice (Supplementary Figure S2), possibly due to the lack of control from the adaptive immune system.
Taken together, we provided evidence that adaptive immunity could alter the composition and diversity of gut microbiota. Defects of adaptive immune system, such as in primary immunodeficiency (for example, SCID) or acquired immunodeficiency (for example, HIV infection), may change gut microbiota that would in turn modulate the immune response to deteriorate or compensate for the defects. We identified in this study a potential fecal biomarker for early diagnosis of infantile SCID. An increased level of fecal A. muciniphila, which normally exists at a negligible level in the feces of healthy young children (Costello et al., 2013), may prompt timely examinations and potentially prevent pediatric emergencies associated with SCID.
Altschul SF, Madden TL, Schaffer AA, Zhang JH, Zhang Z, Miller W et al. (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389–3402.
Anderson MJ . (2001). A new method for non-parametric multivariate analysis of variance. Austral Ecol 26: 32–46.
Antonopoulos DA, Huse SM, Morrison HG, Schmidt TM, Sogin ML, Young VB . (2009). Reproducible community dynamics of the gastrointestinal microbiota following antibiotic perturbation. Infect Immun 77: 2367–2375.
Backhed F, Ding H, Wang T, Hooper LV, Koh GY, Nagy A et al. (2004). The gut microbiota as an environmental factor that regulates fat storage. Proc Natl Acad Sci USA 101: 15718–15723.
Berry D, Stecher B, Schintlmeister A, Reichert J, Brugiroux S, Wild B et al. (2013). Host-compound foraging by intestinal microbiota revealed by single-cell stable isotope probing. Proc Natl Acad Sci USA 110: 4720–4725.
Blaser MJ . (2010). Helicobacter pylori and esophageal disease: wake-up call? Gastroenterology 139: 1819–1822.
Brown EM, Sadarangani M, Finlay BB . (2013). The role of the immune system in governing host-microbe interactions in the intestine. Nat Immunol 14: 660–667.
Campbell JM, Fahey GC, Wolf BW . (1997). Selected indigestible oligosaccharides affect large bowel mass, cecal and fecal short-chain fatty acids, pH and microflora in rats. J Nutr 127: 130–136.
Cani PD, Neyrinck AM, Fava F, Knauf C, Burcelin RG, Tuohy KM et al. (2007). Selective increases of bifidobacteria in gut microflora improve high-fat-diet-induced diabetes in mice through a mechanism associated with endotoxaemia. Diabetologia 50: 2374–2383.
Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK et al. (2010). QIIME allows analysis of high-throughput community sequencing data. Nat Methods 7: 335–336.
Caporaso JG, Lauber CL, Walters WA, Berg-Lyons D, Huntley J, Fierer N et al. (2012). Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms. ISME J 6: 1621–1624.
Cebra JJ . (1999). Influences of microbiota on intestinal immune system development. Am J Clin Nutr 69: 1046s–1051s.
Costello EK, Carlisle EM, Bik EM, Morowitz MJ, Relman DA . (2013). Microbiome assembly across multiple body sites in low-birthweight infants. mBio 4: e00782–13–e00782–13.
Cullender TC, Chassaing B, Janzon A, Kumar K, Muller CE, Werner JJ et al. (2013). Innate and adaptive immunity interact to quench microbiome flagellar motility in the gut. Cell Host Microbe 14: 571–581.
Deplancke B, Gaskins HR . (2001). Microbial modulation of innate defense: goblet cells and the intestinal mucus layer. Am J Clin Nutr 73: 1131S–1141S.
Dethlefsen L, Huse S, Sogin ML, Relman DA . (2008). The pervasive effects of an antibiotic on the human gut microbiota, as revealed by deep 16S rRNA sequencing. PLoS Biol 6: e280.
DiBaise JK, Zhang H, Crowell MD, Krajmalnik-Brown R, Decker GA, Rittmann BE . (2008). Gut microbiota and its possible relationship with obesity. Mayo Clin Proc 83: 460–469.
Dimitriu PA, Boyce G, Samarakoon A, Hartmann M, Johnson P, Mohn WW . (2013). Temporal stability of the mouse gut microbiota in relation to innate and adaptive immunity: mouse gut microbiota dynamics. Environ Microbiol Rep 5: 200–210.
Dubourg G, Lagier J-C, Armougom F, Robert C, Audoly G, Papazian L et al. (2013). High-level colonisation of the human gut by Verrucomicrobia following broad-spectrum antibiotic treatment. Int J Antimicrob Agents 41: 149–155.
Eckburg PB, Bik EM, Bernstein CN, Purdom E, Dethlefsen L, Sargent M et al. (2005). Diversity of the human intestinal microbial flora. Science 308: 1635–1638.
Edgar RC . (2010). Search and clustering orders of magnitude faster than BLAST. Bioinformatics 26: 2460–2461.
Everard A, Belzer C, Geurts L, Ouwerkerk JP, Druart C, Bindels LB et al. (2013). Cross-talk between Akkermansia muciniphila and intestinal epithelium controls diet-induced obesity. Proc Natl Acad Sci USA 110: 9066–9071.
Faith DP . (1992). Conservation evaluation and phylogenetic diversity. Biol Conserv 61: 1–10.
Fleissner D, Hansen W, Geffers R, Buer J, Westendorf AM . (2010). Local Induction of Immunosuppressive CD8+ T Cells in the Gut-Associated Lymphoid Tissues. PLoS One 5: e15373.
Gill SR, Pop M, DeBoy RT, Eckburg PB, Turnbaugh PJ, Samuel BS et al. (2006). Metagenomic analysis of the human distal gut microbiome. Science 312: 1355–1359.
Hansen CHF, Holm TL, Krych Ł, Andresen L, Nielsen DS, Rune I et al. (2013). Gut microbiota regulates NKG2D ligand expression on intestinal epithelial cells. Eur J Immunol 43: 447–457.
Hooper LV, Wong MH, Thelin A, Hansson L, Falk PC, Gordon JI . (2001). Molecular analysis of commensal host-microbial relationships in the intestine. Science 291: 881–884.
Hur EM, Patel SN, Shimizu S, Rao DS, Gnanapragasam PNP, An DS et al. (2012). Inhibitory effect of HIV-specific neutralizing IgA on mucosal transmission of HIV in humanized mice. Blood 120: 4571–4582.
Johansson ME, Phillipson M, Petersson J, Velcich A, Holm L, Hansson GC . (2008). The inner of the two Muc2 mucin-dependent mucus layers in colon is devoid of bacteria. Proc Natl Acad Sci USA 105: 15064–15069.
Klaenhammer TR, Kleerebezem M, Kopp MV, Rescigno M . (2012). The impact of probiotics and prebiotics on the immune system. Nat Rev Immunol 12: 728–734.
Koenig JE, Spor A, Scalfone N, Fricker AD, Stombaugh J, Knight R et al. (2010). Colloquium paper: succession of microbial consortia in the developing infant gut microbiome. Proc Natl Acad Sci USA 108: 4578–4585.
Larsson E, Tremaroli V, Lee YS, Koren O, Nookaew I, Fricker A et al. (2011). Analysis of gut microbial regulation of host gene expression along the length of the gut and regulation of gut microbial ecology through MyD88. Gut 61: 1124–1131.
Ley RE, Backhed F, Turnbaugh P, Lozupone CA, Knight RD, Gordon JI . (2005). Obesity alters gut microbial ecology. Proc Natl Acad Sci USA 102: 11070–11075.
Ley RE, Hamady M, Lozupone C, Turnbaugh PJ, Ramey RR, Bircher JS et al. (2008). Evolution of mammals and their gut microbes. Science 320: 1647–1651.
Loh G, Brodziak F, Blaut M . (2008). The Toll-like receptors TLR2 and TLR4 do not affect the intestinal microbiota composition in mice. Environ Microbiol 10: 709–715.
Lozupone C, Faust K, Raes J, Faith JJ, Frank DN, Zaneveld J et al. (2012). Identifying genomic and metabolic features that can underlie early successional and opportunistic lifestyles of human gut symbionts. Genome Res 22: 1974–1984.
Lozupone C, Knight R . (2005). UniFrac: a new phylogenetic method for comparing microbial communities. Appl Environ Microbiol 71: 8228–8235.
Macpherson AJ, Uhr T . (2004). Induction of protective IgA by intestinal dendritic cells carrying commensal bacteria. Science 303: 1662–1665.
McDonald D, Price MN, Goodrich J, Nawrocki EP, DeSantis TZ, Probst A et al. (2012). An improved Greengenes taxonomy with explicit ranks for ecological and evolutionary analyses of bacteria and archaea. ISME J 6: 610–618.
McFall-Ngai M . (2007). Adaptive Immunity: care for the community. Nature 445: 153–153.
Meyer F, Paarmann D, D’Souza M, Olson R, Glass EM, Kubal M et al. (2008). The metagenomics RAST server–a public resource for the automatic phylogenetic and functional analysis of metagenomes. BMC Bioinformatics 9: 386.
Palmer C, Bik EM, DiGiulio DB, Relman DA, Brown PO . (2007). Development of the human infant intestinal microbiota. PLoS Biol 5: 1556–1573.
Pan X, Chen F, Wu T, Tang H, Zhao Z . (2009). Prebiotic oligosaccharides change the concentrations of short-chain fatty acids and the microbial population of mouse bowel. J Zhejiang Univ Sci B 10: 258–263.
Pantoja-Feliciano IG, Clemente JC, Costello EK, Perez ME, Blaser MJ, Knight R et al. (2013). Biphasic assembly of the murine intestinal microbiota during early development. ISME J 7: 1112–1115.
Pedron T, Mulet C, Dauga C, Frangeul L, Chervaux C, Grompone G et al. (2012). A crypt-specific core microbiota resides in the mouse colon. mBio 3: e00116–12–e00116–12.
Price MN, Dehal PS, Arkin AP . (2010). FastTree 2—approximately maximum-likelihood trees for large alignments. PLoS One 5: e9490.
Round JL, Mazmanian SK . (2009). The gut microbiota shapes intestinal immune responses during health and disease. Nat Rev Immunol 9: 313–323.
Scott CL, Aumeunier AM, Mowat AM . (2011). Intestinal CD103+ dendritic cells: master regulators of tolerance? Trends Immunol 32: 412–419.
Sonnenberg GF, Monticelli LA, Alenghat T, Fung TC, Hutnick NA, Kunisawa J et al. (2012). Innate lymphoid cells promote anatomical containment of lymphoid-resident commensal bacteria. Science 336: 1321–1325.
Suzuki K, Fagarasan S . (2008). How host-bacterial interactions lead to IgA synthesis in the gut. Trends Immunol 29: 523–531.
Suzuki K, Meek B, Doi Y, Muramatsu M, Chiba T, Honjo T et al. (2004). Aberrant expansion of segmented filamentous bacteria in IgA-deficient gut. Proc Natl Acad Sci USA 101: 1981–1986.
Turnbaugh PJ, Bäckhed F, Fulton L, Gordon JI . (2008). Diet-induced obesity is linked to marked but reversible alterations in the mouse distal gut microbiome. Cell Host Microbe 3: 213–223.
Turnbaugh PJ, Ridaura VK, Faith JJ, Rey FE, Knight R, Gordon JI . (2009). The effect of diet on the human gut microbiome: a metagenomic analysis in humanized gnotobiotic mice. Sci Transl Med 1: 6ra14.
Ubeda C, Lipuma L, Gobourne A, Viale A, Leiner I, Equinda M et al. (2012). Familial transmission rather than defective innate immunity shapes the distinct intestinal microbiota of TLR-deficient mice. J Exp Med 209: 1445–1456.
Vijay-Kumar M, Aitken JD, Carvalho FA, Cullender TC, Mwangi S, Srinivasan S et al. (2010). Metabolic syndrome and altered gut microbiota in mice lacking toll-like receptor 5. Science 328: 228–231.
Wang Q, Garrity GM, Tiedje JM, Cole JR . (2007). Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Env Microbiol 73: 5261–5267.
Wen L, Ley RE, Volchkov PV, Stranges PB, Avanesyan L, Stonebraker AC et al. (2008). Innate immunity and intestinal microbiota in the development of Type 1 diabetes. Nature 455: 1109–1113.
Willing BP, Russell SL, Finlay BB . (2011). Shifting the balance: antibiotic effects on host–microbiota mutualism. Nat Rev Microbiol 9: 233–243.
Wlodarska M, Willing B, Keeney KM, Menendez A, Bergstrom KS, Gill N et al. (2011). Antibiotic treatment alters the colonic mucus layer and predisposes the host to exacerbated citrobacter rodentium-induced colitis. Infect Immun 79: 1536–1545.
Young VB, Schmidt TM . (2004). Antibiotic-associated diarrhea accompanied by large-scale alterations in the composition of the fecal microbiota. J Clin Microbiol 42: 1203–1206.
Zoetendal EG, von Wright A, Vilpponen-Salmela T, Ben-Amor K, Akkermans ADL, de Vos WM . (2002). Mucosa-associated bacteria in the human gastrointestinal tract are uniformly distributed along the colon and differ from the community recovered from feces. Appl Environ Microbiol 68: 3401–3407.
We thank Melissa Makris for the use of Flow Cytometry Core Facility at College of Veterinary Medicine at Virginia Tech. This study was partially funded by Virginia Bioinformatics Institute & Fralin Life Science Institute Small Grants Program (VBI/Fralin-GRL-01).
The authors declare no conflict of interest.
Supplementary Information accompanies this paper on The ISME Journal website
About this article
Cite this article
Zhang, H., Sparks, J., Karyala, S. et al. Host adaptive immunity alters gut microbiota. ISME J 9, 770–781 (2015). https://doi.org/10.1038/ismej.2014.165
This article is cited by
Nature Reviews Microbiology (2023)
Nature Cell Biology (2023)
BMC Veterinary Research (2022)
Timing matters: age-dependent impacts of the social environment and host selection on the avian gut microbiota
Pediatric Research (2022)