Microbiota-metabolites interactions in non-human primate gastrointestinal tract

Background The microbiota has been recognised as an important part for maintaining human health. Perturbation to its structure has been implicated in many diseases, such as obesity and cancers. The microbiota is highly metabolically active and fills in many niche metabolic pathways absent from the human host. Diseases such as obesity, cardiovascular disease and colorectal cancer has been linked to altered microbiota metabolism. However, there is a gap in the current knowledge on how mucosal-associated microbiota and colon mucosa interact. Here we performed an integrated analysis between the mucosal-associated microbiota and the mucosal tissue metabolites in healthy non-human primates. Results We found that the overall microbiota composition is influenced by both the tissue location as well as the host. We also identified bacteria signatures for different intestinal locations. The distal colon bacterial signature includes Ruminococcaceae, Bacteroidales, Christensenellaceae, Clostridiales, Sphaerochaeta, Victivallaceae, GMD14H09, CF231, ML615J-28, RF39 and R4-45B taxa. In the cecum, the signatures include Prevotella, Anaerovibrio, Roseburia, and Anaerostipes. Desulfovibrionaceae family is the only taxon that may be a signature for the duodenum. We also found an intricate global relationship between the microbiota and the host tissue metabolome that is mainly driven by the distal colon. Most importantly, we found microbial-centric tissue metabolites clusters that may have potential implications to studying host-microbiota metabolic interactions.


Background
The human gastrointestinal tract harbors trillions of microorganisms, including thousands of species of bacteria, termed microbiota [1] . It has become evident that the gut microbiota is important in regulating and maintaining the health of the host and is implicated in many diseases, including cancers [1][2][3][4][5] . Despite numerous studies indicating important roles of microbiota in diseases, many of these studies have largely focused on the taxonomic composition of the microbiota. The mechanism of the host-microbiota interaction, however, still remains unclear.
Previous studies suggest, the gut microbiota produces a vast amount of metabolites.
This metabolic interaction between the host and its microbiota has widespread implications around the body [7] . For example, the obesity associated microbiota has been shown to possess increased metabolic capability to harvest energy from food [9,10] . The metabolism of L-carnitine by the gut microbiota has been shown to promote atherosclerosis [11] . These studies suggest a potential metabolic adaptations of the microbiota in response to the host metabolic change [10] .
The most direct metabolic interaction between the host and its microbiota, however, is at the colon. In fact, the majority (~70%) of energy source required by the normal colon epithelium come from butyrate produced by the microbiota through fermentation of complex carbohydrates [12] . Without a functional microbiota, the colon epithelia will undergo autophagy and fail to maintain its normal structure and function [13] . More importantly, this metabolic interactions may have important implications in colorectal cancer, the 2nd most deadly cancer in the United States [1,[5][6][7][14][15][16] .
Most previous studies in microbiota used fecal samples or biopsy samples due to the ease of sampling. Thus, the mucosal host-microbiota metabolic interactions along healthy gastrointestinal tract is largely unknown. Here, we investigate such metabolic interactions in 10 healthy baboons, a family of Old World monkeys belong to the Papio genus. We collected tissue samples from 6 small and large intestine segments and sequenced the 16S rRNA gene to identify the mucosal-associated microbiota composition. We also performed untargeted metabolomics on the immediate adjacent tissues to profile the tissue metabolites composition. To our knowledge, this is one of the first analyses to comprehensively establish the intestinal host-microbiota metabolic interactions in NHPs.

Microbiota along the non-human primate gastrointestinal tract.
We first assessed the NHP GI tissue-associated microbiota composition in 10 baboons using 16S rRNA gene sequencing method ( Supplementary Table 1 ). At the phyla level, the NHP GI tissue-associated microbiota is dominated by the bacterial phyla of Firmicutes, Bacteroidetes and Proteobacteria, regardless of the tissue location ( Figure   1A, B ). This microbiota composition is similar to that observed in human GI tissue-associated microbiota, however dissimilar to that observed in mouse fecal samples ( Supplementary Figure 1 ). [5] In the NHP samples, most phyla level composition remain unaltered along the GI tract except for Tenericutes and Lentisphaerae (p < 0.005, ANOVA with Tukey post-hoc test). Both of these two phyla are more abundant in the distal colon compared to all other locations. Since many important information were masked at the phyla level, we next analyzed diversity metrics to better understand the microbiota composition difference at the OTU resolution. We first assessed the beta-diversity between tissue locations by performing Principal Coordinate Analysis (PCoA) using both weighted and unweighted UniFrac distance metrics. The unweighted UniFrac distance only consider the presence and absence of a certain OTU, while weighted UniFrac distance will consider the abundance, thus these metrics can give an overview on the microbial structure difference of different tissue locations [17] . The PCoA of unweighted UniFrac distance show clusterings mainly based on the tissue location (p < 0.01, PERMANOVA) as well as the sample origin (p < 0.001), the weighted UniFrac distance also show a same clustering based on the tissue location (p < 0.05) and sample origin (p < 0.001) ( Figure   1C, Supplementary Figure 2 ). This suggests both host and tissue location may have an impact on the mucosal-microbiota structure in the intestines [1] .
We then analyzed the alpha-diversity to discern the microbial diversity within each sample. Consistent with previous reports, we found the small intestinal microbiota has significantly lower phylogenetic diversity (p < 1 x 10 -7 , two-tailed t-test, Figure 2A ), lower Shannon index (p < 0.0005, Figure 2B ) and lower chao1 index (p < 1 x 10 -6 , Figure 2C ) [18,19] . This observation is likely due to the microbial concentration gradient along the GI tract, where the small intestine harbor less bacteria due to the high pH environment. We then assessed the microbiota differences at the genus and the OTU level to discover site specific bacterial community signature. We found 17 bacterial taxa to significantly differ in relative abundance in at least 1 tissue site (p < 0. We then analyzed the microbiota-metabolites relationships using spearman's ranked correlation on the metabolites with assigned identity. The spearman's ranked correlation is a nonparametric test that can be used to reveal subtle relationship between microbiota and metabolites [20] . We first summarized the microbiota OTU data to the genus level, then performed the correlation with the corresponding metabolites data. Globally, a total of 4,410 significant correlations (q < 0.1, False discovery rate adjusted p-value) were found between the microbiota and the known metabolites  Figure 6B ). Additionally, the correlation network for the small intestine is more interconnected as compared to the large intestine. One explanation is that the large intestine harbors more bacterial species compared to the small intestine, thus there could be more functional redundancies in the large intestine which means less correlations at the lower levels of the OTU hierarchy. Indeed, when we performed a similar analysis using the phyla level data, the large intestine had 78 significant correlations (q < 0.1) compared to 39 significant correlations (q < 0.1) for the small intestine ( Supplementary Table 3 ). Interestingly, both networks appears to have a microbiota-centric correlation, where most metabolites only correlation to few bacteria and such correlations tend to be at the same direction.

Discussion
Currently, there is limited knowledge in the microbiota composition along different sections of the GI tract in either human or NHP samples [23,24] . Studies in human subjects usually require prior bowel preparation, which have been shown to alter the microbiota [25] . In this study, we collected tissue samples from healthy non-human primates without prior bowel preparation, thus providing an unaltered view of the healthy microbiota. Previous studies have analyzed the GI tract microbiota compositions in mouse, chicken, dog, cow and horse [19,[26][27][28][29] . However, due to the anatomical differences, in addition to the dietary and genetic difference, these animals may have different microbiota along the GI tract. To our knowledge, this study is the first report to comprehensively analyze the mucosa-associated microbiota-metabolites interactions along the GI tract in NHPs.
In this study, we performed untargeted metabolomics on the intestinal tissues. Although it is able to identify over 3,395 compounds, we were only able to assign identities to 292 compounds. This lack of positive identification is mainly due to the lack of database available. It is conceivable that when such database becomes available, we will be able to extract additional information from the data. Another caveat of the current study is that we were not able to identify the metabolites origin. Future studies should aim to separate metabolites originated from the host, microbiota or food source.
Similar to previous reports in humans, we found variations to the microbiota composition between different NHP subjects. In addition, we found that the microbiota composition along the GI tract is also influenced by the host. Previous studies suggest that this variation between individuals can be attributed to factors such as genetics, dietary preferences and other factors. [1,30,31] This study also characterized the microbiota signatures along different sections of the GI tract, and supported the previous hypothesis that the small intestines harbor less diverse microbiota compared to the large intestine.
This study also found that the host-microbiota metabolic interactions are microbiota-centric. Most metabolites are correlated with few microbes and in the same direction. It is important to note that this analysis could not identify the directionality of such interactions. Surprisingly, only 4 microbiota-metabolome correlation pairs were found in common between the small and large intestine. As speculated previously, it is possibly due to the higher level of functional redundancy present in the large intestine due to the larger species present. Another possible explanation is that the pH, as well as the nutrient composition are different between the small and large intestines, and these factors together may have additional effects on the microbiota and the tissue metabolome.
Moreover, the location specific correlation may suggest a potential strategy to target beneficial bacteria in different intestinal locations. Notably, 3H-1,2-Dithiole-3-thione has been previously shown as a potent antioxidant and potential chemopreventive agent, by targeting the transcription factor NRF2. [32]

Conclusions
In the present study, we report the host-microbiota interactions along the healthy non-human primate lower gastrointestinal tract. Our study provided a global view of the microbiota landscape of healthy NHPs. Our analysis suggests an intricate global relationship between the microbiota and the metabolites along the GI tract. Further functional validation is warranted to establish the directionality of such interactions. low quality reads were first trimmed using Trimmomatic v0.33. Then, the forward and reverse read pairs were merged using PandaSeq v2.8 [33] . OTUs were then picked using QIIME v1.9.1 "pick_open_reference_otus.py" script against Greengenes 16S database (May, 2013 release), allowing 97% similarity [34,35] . The unfiltered OTU

Consent for publication
Not applicable

Availability of data and material
All data generated or analysed during this study are included in this published article and its supplementary information files.  Shannon Index, and c. Chao1 Index.