TY - JOUR
AU - Ma, Zhanshan, (Sam)
AB - ABSTRACT Human gut microbiome could translocate to other tissues, and the relocation triggered by HIV/SIV infection has received increasing attention. However, the underlying mode of this translocation, whether it is deterministic or random (passive) process, is not clear, not to mention quantitative estimation of the relocation probability and rates. Using multi-tissue microbiome datasets collected from SIV-infected macaques, originally reported by Klase et al. (2015), we apply Hubbell's unified neutral theory of biodiversity (UNTB) implemented by Harris et al. (2017) in the form of multi-site neutral (MSN) model to explore the translocation mode and rates of the gut microbiome. We found that (i) The translocation from gastrointestinal tract to tissues was driven by stochastic (neutral) forces as revealed by 100% neutrality-passing rates with MSN testing; (ii) The translocation probability from gastrointestinal tract to tissues is significantly larger than the baseline dispersal rates occurring within gastrointestinal tract (0.234 vs. 0.006 at the phylum level, P< 0.001). (iii) Approximately, 23% of phyla and 55% of genera were migrated from gastrointestinal tract to the tissues (liver and mesenteric lymph nodes). Our findings offer the first interpretation of the microbial translocation mode from gastrointestinal tract to tissues, and the first estimates of the translocation probability and level. microbiome translocation, multi-site neutral model (MSN), SIV infection, stochastic drift, translocation probability INTRODUCTION In the last few years, researchers have begun to explore the world of human tissue microbiome, i.e. the microbes distributed within body tissues including the lung, mammal breast gland, liver, colon, prostate and even blood that was considered as sterile not long ago (e.g. HMP Consortium 2012; Sze et al. 2012; Paisse et al. 2016). Obviously, a primary reason for the lag in tissue microbiome research behind the studies of microbiome distributed in or on our bodies has been ethical issues and to a less extent the technology to extract DNA from the tissues or blood. It was probably also for this reason, most reported tissue microbiome studies were conducted with pathological tissue samples of patients (e.g. lung or breast cancer), or with animal tissues. The delay in studying tissue microbiome does not mean it is less important than studying ‘regular’ human microbiome at all. In fact, studies have revealed that tissue microbiome is strongly related to the tissue microenvironment for cancer to develop. Indeed, many studies of tissue microbiome have been conducted with cancer tissue (e.g. Kim et al. 2016; Urbaniak et al. 2016; Yu et al2016; Cavarretta et al2017). The interrelationship between endogenous microbiota, the immune system, and tissue regeneration has been extensively studied in recent years due to its potential therapeutic applications (Arnold et al. 2016). Human gut microbiome could actually translocate to other parts of the human body, such as the liver, the brain, and other organs, where it could play a significant role in health and also lead to unprecedented damages by starting inflammatory cascades, leading to numerous metabolic diseases. In many cases, the translocation occurs abnormally when tissue was infected with pathogens such as HIV viruses. In this case, immune system would certainly act to resist the invasion by bacteria. The gastrointestinal (GI) tract is the largest mucosal organ in human body. It carries a very large percentage of the body's leukocytes. This concentration of immunological defense within the GI tract assumes the mission to contain, and potentially respond to, the large microbial mass within the lumen (Klase et al. 2015). During acute HIV (human immunodeficiency virus) and SIV (simian immunodeficiency virus) infection can lead to rapid loss of gut-resident CD4+ memory T cells and damage to epithelial barrier. The damage to epithelium allows for translocation of microbes and their microbial products from the intestinal lumen into the body. Consequently, immune activation triggered by translocated microbial products is associated with disease progression such as development of chronic inflammation. Through examining the bacterial communities both within the GI and systemic tissues of both healthy and experimentally SIV infected Asian macaques, Klase et al (2015) confirmed the translocation of GI microbiome into tissues including colon, liver, and MLNs (mesenteric lymph nodes). Klase et al. (2015) further revealed that the phylum Proteobacteria preferentially translocate and increased metabolic activity of Proteobacterial species within the colonic lumen of SIV-infected macaques. Despite the significant advances in research on the HIV/SIV infection triggered translocation, the underlying mode of the translocation is not clear. In this study, we investigate the translocation mode by applying Hubbell's (2001) unified neutral theory of biodiversity, which was developed to interpret the underlying mechanisms of community assembly and diversity maintenance in the community ecology of plants and animals, but have been successfully introduced to microbial community ecology in recent years. Specifically, by leveraging the advantage of the neutral theory, we estimate the migration rate among sites and the mode of translocation (whether it is random or deterministic). To the best of our knowledge, both the issues in the context of gut microbe translocation have not been addressed previously. Obviously, successfully addressing both questions may shed important light on the mode of the translocation and generate significant biomedical implications. In this study, besides addressing the mode of microbial relocation, we also test the applicability of the neutral theory to tissue microbiome. Previous neutral theory applications to human or animal microbiome have been limited to the microbiome distributed in or on the bodies (e.g. Harris et al 2017; Venkataraman et al. 2015; Zeng et al 2015; Li and Ma 2016; Li et al2018), but not within tissues. Furthermore, we apply a latest advance in neutral theory, a multi-site neutral (MSN) model by Harris et al. (2017), which allows for different migration rates among different microbiome sites (local communities such as liver and colon) and is particularly appropriate for estimating the migration rates in the study of microbiome translocation. MATERIAL AND METHODS The Data sets of SIV-infected stool and tissue microbiomes of non-human primates Klase et al (2015) conducted experimental infection of Asian macaques with SIV (simian immunodeficiency virus) to examine the microbiome changes occurred in the monkeys. Their study examined the bacterial communities both within the gastrointestinal (GI) tract and systemic tissues of both healthy and experimentally SIV-infected Asian macaques, including both pigtail macaques (PTMs) and rhesus macaques (RMs). Eight rhesus macaques (Macaca mulatta) were infected intravenously with 3000 TCID50 (50% tissue culture infectious dose) of SIVmac239 and were killed during the chronic phase of infection (>4 weeks). Three healthy uninfected control animals were killed as controls. At necropsy, tissue samples were harvested from the colon, MLNs, and liver, and flash frozen in liquid nitrogen for later DNA extraction. Twelve pigtail macaques (Macaca nemestrina) were infected intravenously with 3000 TCID50 of SIVmac239. Blood and stool were sampled longitudinally over the course (91 days) of infection. Metagenomic DNA samples were collected and amplicon-sequenced of the samples was performed to obtain sequencing reads (V1-V3 hyper-variable region) of the 16S-rRNA gene, which was processed with Mothur (v1.31.2) and supporting software packages for quality control and OTU binning. The OTUs at phylum, class, order, family and genus level were clustered respectively. The OTU abundance datasets of the three treatments, including stool, pigtail, and rhesus are used to test the MSN model. See Table S1 for a brief introduction on the samples and corresponding OTU data sets and Klase et al. (2015) for the detailed information. Multi-site neutral model (MSN) The computational procedures and program for fitting the MSN used in this study were from Harris et al. (2017). The following is a very brief, quantitative description on the MSN model, but the detailed description is referred to the original publication by Harris et al. (2017) given the complexity of the model. (i) Hubbell's Unified Neutral Theory of Biodiversity (UNTB) The UNTB (Hubbell 2001) conceptually distinguished the local community dynamics from metacommunity dynamics, but both are driven by similar neutral processes. Three key parameters (aspects) for UNTB can be indentified. The immigration rate (Ii), which links local community to the meta-community. The speciation rate, or the fundamental biodiversity number (θ), which can be considered as the rate at which new individuals are added to the meta-community due to speciation. A third aspect of the UNTB is to assume that the SAD (species abundance distribution) of each community sample can be described by the multinomial (MN) distribution. See Hubbell (2001); Harris et al. (2017) for more detailed introduction on the UNTB. (ii) HDP (Hierarchical Dirichlet Process) Limit to Neutral Metacommunities A fully general procedure for fitting multiple sites UNTB with different migration rates is computationally intractable even for a moderate number of sites, and approximate algorithms must be adopted (Harris et al. 2017). The extremely computational challenge has severely limited the utilization of multi-site sampling data, which is invaluable, in testing the UNTB and consequent applications, until Harris et al. (2017) developed an efficient Bayesian fitting framework by approximating the neutral models with the hierarchical Dirichlet process (HDP). Harris et al. (2017) approximation method captures the essential elements of the UNTB, i.e. neutrality, finite populations, and multiple geographically isolated populations linked by relatively rare migration—while little influenced by the specific details of the local community dynamics. This significant computational advance makes it possible to efficiently fit and test the neutral theory model with even the largest datasets in a reasonable amount of computational time. Harris et al. (2017) approach has additional advantages. One is that it can generate full posterior distributions over the parameters rather than just a maximum likelihood prediction. That is, it reconstructs the metacommunity distribution enabling us to separate the key question of whether a community appears neutral into two parts: testing of the neutrality at both metacommunity and local community levels. Sloan et al. (2006, 2007) showed that for large local population sizes, given a fixed finite-dimensional metacommunity distribution with S species present, then the local community distribution, |$\overline {{\pi _i}} $| can be approximated by a Dirichlet distribution (Sloan et al 2006, 2007). It was Harris et al. (2017) who developed the general framework for approximating the UNTB computationally efficiently. Assuming there is a potentially infinite number of species that can be observed in the local community, then the stationary distribution of observing local population i is a Dirichlet process (DP), i.e. $$\begin{eqnarray*} \left. {\overline {{\pi _i}} } \right|{I_i},\overline \beta \tilde{D}P({I_i},\overline \beta ) \end{eqnarray*}$$(1) where |$\overline \beta = ({\beta _1},...,{\beta _S})$| is the relative frequency of each species in the metacommunity. At the metacommunity level, a Dirichlet process is still applicable, but then the base distribution is simply a uniform distribution over arbitrary species labels, and the concentration parameter is the biodiversity parameter (θ) (Harris et al. 2017). This observation was also implicit in early fitting of the neutral models (Etienne 2005). The metacommunity distribution is purely the stick breaking process, i.e. $$\begin{eqnarray*} \overline \beta \tilde{S}tick(\theta ) \end{eqnarray*}$$(2) Since both local community and metacommunity are Dirichlet processes, it becomes a hierarchical Dirichlet process (HDP) in the domain of machine learning (Harris et al. 2017). (iii) Gibbs Sampler (MCMC algorithm) for the Neutral-HDP model The full neutral-HDP model is formed by combining previous equations (1-2) and the previously mentioned multi-nominal (MN) distribution of the community sample. Harris et al. (2017) devised an efficient Gibbs sampler for the UNTB-HDP approximation, which is a type of Bayesian Markov Chain Monte Carlo (MCMC) algorithm and can be summarized as the following four sampling steps, including sampling the biodiversity parameter, sampling the meta-community distribution, sampling the immigration rate, and sampling the ancestral states. Harris et al. (2017) discovered through experiments that to ensure sampling was from the stationary distribution, 50,000 Gibb samples for each fitted dataset were required with the first 25,000 iterations removed as burn-in. The results are reported as the median values over the last 25,000 samples with upper and lower credible limits (Bayesian confidence) given by 2.5% and 97.5% quantiles of those samples. (iv) Goodness-of-fitting test for the HDP neutral model To determine whether an observed dataset fits to the HDP-neutral model, Harris et al. (2017) proposed a similar Monte Carlo significance test to that used by Etienne (2007). Furthermore, Harris et al (2017) also developed a procedure to test for local neutral community assembly but with a fitted potentially non-neutral metacommunity because of the hierarchical nature of the model. The proportion of samples having likelihood not exceeding this forms the pseudo p-value, denoted by pL is then utilized for testing the neutrality of the local community assembly. We further explain the statistics and parameters involved in the neutrality test in the results and discussion section (legend for Table 1). Table 1. Test results of fitting the MSN (multi-site neutral) model to SIV-infected pigtail stool & tissue microbiome datasets*. ID . L0 . θ . M-value . m . Community Dominance . Species Richness . Meta-community . Local community . . . . . . . . LM . NM . N . pM . LL . NL . N . pL . Phylum level 2-PTA0P036-stool −81.931 2.259 72.309 0.221 4.406 9.0 −82.562 1231 2500 0.492 –80.334 1404 2500 0.562 2-PTA0P048-stool −97.083 2.352 179.083 0.302 4.677 13.0 −100.379 1160 2500 0.464 –100.659 892 2500 0.357 2-PTA0P048 −55.809 2.388 61.821 0.393 3.488 7.7 −58.480 1124 2500 0.450 −56.633 1141 2500 0.456 2-PTA2P032 −41.054 1.755 15.777 0.228 1.636 5.3 −33.411 1695 2500 0.678 −38.664 1590 2500 0.636 2-PTFR60-stool −78.245 2.039 129.945 0.274 4.475 9.7 −86.619 998 2500 0.399 −81.660 902 2500 0.361 2-PTFR60 −57.387 2.011 27.793 0.196 2.740 7.0 −50.357 1546 2500 0.618 −53.543 1684 2500 0.674 2-PTGC45-stool −57.116 1.790 97.974 0.347 3.462 8.7 −55.328 1337 2500 0.535 −57.638 1171 2500 0.468 2-PTGC45 −53.176 2.083 16.868 0.168 1.965 6.3 −42.891 1714 2500 0.686 −50.439 1569 2500 0.628 2-PTGD79-stool −51.473 1.993 144.168 0.313 3.920 10.5 −54.563 1120 2500 0.448 −52.180 1170 2500 0.468 2-PTGD79 −42.790 1.778 42.476 0.451 2.193 6.3 −38.321 1481 2500 0.592 −43.138 1201 2500 0.480 2-PTGD85-stool −82.604 1.939 155.632 0.260 4.243 10.0 −90.547 1033 2500 0.413 −86.368 841 2500 0.336 2-PTGD85 −60.936 3.842 5.569 0.049 2.821 6.0 −53.886 1604 2500 0.642 −56.235 1645 2500 0.658 2-PTGE34-stool −47.077 5.672 2.138 0.014 4.196 5.5 −38.810 1747 2500 0.699 −40.530 1701 2500 0.680 2-PTGE34 −69.190 2.254 52.909 0.236 3.225 8.7 −64.459 1423 2500 0.569 −64.897 1633 2500 0.653 2-PTGE36-stool −74.295 1.829 133.537 0.301 3.809 9.3 −75.170 1205 2500 0.482 −77.247 904 2500 0.362 2-PTGE36 −65.465 3.216 12.863 0.085 3.498 7.0 −58.720 1543 2500 0.617 −61.040 1615 2500 0.646 2-PTGE40-stool −54.012 2.059 111.252 0.256 4.112 9.5 −54.767 1206 2500 0.482 −54.529 1195 2500 0.478 2-PTGE40 −39.748 2.207 29.801 0.485 1.894 6.0 −35.468 1531 2500 0.612 −38.812 1401 2500 0.560 2-PTGP27-stool −118.295 2.286 120.788 0.219 4.884 10.0 −132.632 940 2500 0.376 −120.554 1077 2500 0.431 2-PTGP27 −58.213 1.515 88.928 0.280 2.910 7.7 −53.952 1400 2500 0.560 −57.269 1388 2500 0.555 2-PTGR15-stool −87.019 1.342 113.617 0.262 3.430 8.0 −80.382 1426 2500 0.570 −91.354 753 2500 0.301 2-PTGR15 −58.021 2.122 13.878 0.085 2.732 6.0 −49.903 1593 2500 0.637 −53.229 1713 2500 0.685 Average −65.043 2.306 74.051 0.247 3.396 8 −63.255 1366 2500 0.546 −64.407 1300 2500 0.520 Passing rate 100% 100% Genus level 6-PTA0P036-stool −526.258 39.726 351.481 0.579 6.352 68.0 −535.607 1061 2500 0.424 −509.737 1788 2500 0.715 6-PTA0P048-stool −492.134 27.985 759.693 0.647 8.429 76.0 −542.063 472 2500 0.189 −505.627 699 2500 0.280 6-PTA0P048 −304.505 36.346 176.331 0.648 3.921 37.7 −291.763 1666 2500 0.666 −287.250 1917 2500 0.767 6-PTA2P032 −192.575 30.231 96.651 0.644 3.471 23.7 −169.774 2112 2500 0.845 −176.814 2007 2500 0.803 6-PTFR60-stool −504.056 31.988 666.747 0.659 7.879 75.7 −546.929 521 2500 0.208 −511.624 964 2500 0.386 6-PTFR60 −295.847 33.225 209.837 0.648 2.943 38.0 −281.344 1714 2500 0.686 −275.265 2090 2500 0.836 6-PTGC45-stool −276.386 22.235 344.911 0.651 5.054 41.0 −288.308 943 2500 0.377 −281.058 1008 2500 0.403 6-PTGC45 −308.692 61.828 94.992 0.533 2.190 30.0 −266.840 2350 2500 0.940 −284.957 2111 2500 0.844 6-PTGD79-stool −269.169 22.163 541.095 0.631 5.837 62.0 −276.861 1045 2500 0.418 −269.444 1231 2500 0.492 6-PTGD79 −186.193 22.340 156.346 0.752 2.048 27.0 −172.575 1816 2500 0.726 −180.535 1625 2500 0.650 6-PTGD85-stool −434.526 21.868 713.475 0.617 8.729 67.3 −484.371 475 2500 0.190 −451.787 582 2500 0.233 6-PTGD85** −362.903 153.189 63.940 0.372 2.372 27.3 −298.795 2443 2500 0.977 −322.027 2284 2500 0.914 6-PTGE34-stool −351.151 214.483 33.929 0.186 2.600 38.0 −282.458 2417 2500 0.967 −304.384 2294 2500 0.918 6-PTGE34 −314.646 27.072 283.980 0.624 4.323 41.7 −317.053 1171 2500 0.468 −298.127 1924 2500 0.770 6-PTGE36-stool −507.495 35.555 715.271 0.697 7.031 78.7 −551.547 469 2500 0.188 −518.860 820 2500 0.328 6-PTGE36 −496.030 180.459 87.402 0.388 2.785 38.3 −409.745 2471 2500 0.988 −441.977 2350 2500 0.940 6-PTGE40-stool −284.362 25.130 546.073 0.628 6.266 62.0 −300.999 819 2500 0.328 −281.508 1397 2500 0.559 6-PTGE40 −59.404 30.621 82.571 0.723 1.424 19.7 −136.837 2270 2500 0.908 −147.580 1956 2500 0.782 6-PTGP27-stool −633.246 27.238 590.541 0.578 9.108 67.3 −723.025 253 2500 0.101 −644.791 858 2500 0.343 6-PTGP27 −329.178 23.032 391.182 0.631 5.400 46.0 −334.395 1118 2500 0.447 −316.142 1851 2500 0.740 6-PTGR15-stool −511.331 22.661 581.787 0.645 7.299 58.3 −583.740 354 2500 0.142 −533.775 450 2500 0.180 6-PTGR15 −357.736 55.799 69.261 0.316 4.425 33.7 −332.687 1953 2500 0.781 −338.660 1833 2500 0.733 Average −368.083 52.053 343.523 0.582 4.995 48.064 −369.442 1360 2500 0.544 −358.270 1547 2500 0.619 Passing rate 100% 100% ID . L0 . θ . M-value . m . Community Dominance . Species Richness . Meta-community . Local community . . . . . . . . LM . NM . N . pM . LL . NL . N . pL . Phylum level 2-PTA0P036-stool −81.931 2.259 72.309 0.221 4.406 9.0 −82.562 1231 2500 0.492 –80.334 1404 2500 0.562 2-PTA0P048-stool −97.083 2.352 179.083 0.302 4.677 13.0 −100.379 1160 2500 0.464 –100.659 892 2500 0.357 2-PTA0P048 −55.809 2.388 61.821 0.393 3.488 7.7 −58.480 1124 2500 0.450 −56.633 1141 2500 0.456 2-PTA2P032 −41.054 1.755 15.777 0.228 1.636 5.3 −33.411 1695 2500 0.678 −38.664 1590 2500 0.636 2-PTFR60-stool −78.245 2.039 129.945 0.274 4.475 9.7 −86.619 998 2500 0.399 −81.660 902 2500 0.361 2-PTFR60 −57.387 2.011 27.793 0.196 2.740 7.0 −50.357 1546 2500 0.618 −53.543 1684 2500 0.674 2-PTGC45-stool −57.116 1.790 97.974 0.347 3.462 8.7 −55.328 1337 2500 0.535 −57.638 1171 2500 0.468 2-PTGC45 −53.176 2.083 16.868 0.168 1.965 6.3 −42.891 1714 2500 0.686 −50.439 1569 2500 0.628 2-PTGD79-stool −51.473 1.993 144.168 0.313 3.920 10.5 −54.563 1120 2500 0.448 −52.180 1170 2500 0.468 2-PTGD79 −42.790 1.778 42.476 0.451 2.193 6.3 −38.321 1481 2500 0.592 −43.138 1201 2500 0.480 2-PTGD85-stool −82.604 1.939 155.632 0.260 4.243 10.0 −90.547 1033 2500 0.413 −86.368 841 2500 0.336 2-PTGD85 −60.936 3.842 5.569 0.049 2.821 6.0 −53.886 1604 2500 0.642 −56.235 1645 2500 0.658 2-PTGE34-stool −47.077 5.672 2.138 0.014 4.196 5.5 −38.810 1747 2500 0.699 −40.530 1701 2500 0.680 2-PTGE34 −69.190 2.254 52.909 0.236 3.225 8.7 −64.459 1423 2500 0.569 −64.897 1633 2500 0.653 2-PTGE36-stool −74.295 1.829 133.537 0.301 3.809 9.3 −75.170 1205 2500 0.482 −77.247 904 2500 0.362 2-PTGE36 −65.465 3.216 12.863 0.085 3.498 7.0 −58.720 1543 2500 0.617 −61.040 1615 2500 0.646 2-PTGE40-stool −54.012 2.059 111.252 0.256 4.112 9.5 −54.767 1206 2500 0.482 −54.529 1195 2500 0.478 2-PTGE40 −39.748 2.207 29.801 0.485 1.894 6.0 −35.468 1531 2500 0.612 −38.812 1401 2500 0.560 2-PTGP27-stool −118.295 2.286 120.788 0.219 4.884 10.0 −132.632 940 2500 0.376 −120.554 1077 2500 0.431 2-PTGP27 −58.213 1.515 88.928 0.280 2.910 7.7 −53.952 1400 2500 0.560 −57.269 1388 2500 0.555 2-PTGR15-stool −87.019 1.342 113.617 0.262 3.430 8.0 −80.382 1426 2500 0.570 −91.354 753 2500 0.301 2-PTGR15 −58.021 2.122 13.878 0.085 2.732 6.0 −49.903 1593 2500 0.637 −53.229 1713 2500 0.685 Average −65.043 2.306 74.051 0.247 3.396 8 −63.255 1366 2500 0.546 −64.407 1300 2500 0.520 Passing rate 100% 100% Genus level 6-PTA0P036-stool −526.258 39.726 351.481 0.579 6.352 68.0 −535.607 1061 2500 0.424 −509.737 1788 2500 0.715 6-PTA0P048-stool −492.134 27.985 759.693 0.647 8.429 76.0 −542.063 472 2500 0.189 −505.627 699 2500 0.280 6-PTA0P048 −304.505 36.346 176.331 0.648 3.921 37.7 −291.763 1666 2500 0.666 −287.250 1917 2500 0.767 6-PTA2P032 −192.575 30.231 96.651 0.644 3.471 23.7 −169.774 2112 2500 0.845 −176.814 2007 2500 0.803 6-PTFR60-stool −504.056 31.988 666.747 0.659 7.879 75.7 −546.929 521 2500 0.208 −511.624 964 2500 0.386 6-PTFR60 −295.847 33.225 209.837 0.648 2.943 38.0 −281.344 1714 2500 0.686 −275.265 2090 2500 0.836 6-PTGC45-stool −276.386 22.235 344.911 0.651 5.054 41.0 −288.308 943 2500 0.377 −281.058 1008 2500 0.403 6-PTGC45 −308.692 61.828 94.992 0.533 2.190 30.0 −266.840 2350 2500 0.940 −284.957 2111 2500 0.844 6-PTGD79-stool −269.169 22.163 541.095 0.631 5.837 62.0 −276.861 1045 2500 0.418 −269.444 1231 2500 0.492 6-PTGD79 −186.193 22.340 156.346 0.752 2.048 27.0 −172.575 1816 2500 0.726 −180.535 1625 2500 0.650 6-PTGD85-stool −434.526 21.868 713.475 0.617 8.729 67.3 −484.371 475 2500 0.190 −451.787 582 2500 0.233 6-PTGD85** −362.903 153.189 63.940 0.372 2.372 27.3 −298.795 2443 2500 0.977 −322.027 2284 2500 0.914 6-PTGE34-stool −351.151 214.483 33.929 0.186 2.600 38.0 −282.458 2417 2500 0.967 −304.384 2294 2500 0.918 6-PTGE34 −314.646 27.072 283.980 0.624 4.323 41.7 −317.053 1171 2500 0.468 −298.127 1924 2500 0.770 6-PTGE36-stool −507.495 35.555 715.271 0.697 7.031 78.7 −551.547 469 2500 0.188 −518.860 820 2500 0.328 6-PTGE36 −496.030 180.459 87.402 0.388 2.785 38.3 −409.745 2471 2500 0.988 −441.977 2350 2500 0.940 6-PTGE40-stool −284.362 25.130 546.073 0.628 6.266 62.0 −300.999 819 2500 0.328 −281.508 1397 2500 0.559 6-PTGE40 −59.404 30.621 82.571 0.723 1.424 19.7 −136.837 2270 2500 0.908 −147.580 1956 2500 0.782 6-PTGP27-stool −633.246 27.238 590.541 0.578 9.108 67.3 −723.025 253 2500 0.101 −644.791 858 2500 0.343 6-PTGP27 −329.178 23.032 391.182 0.631 5.400 46.0 −334.395 1118 2500 0.447 −316.142 1851 2500 0.740 6-PTGR15-stool −511.331 22.661 581.787 0.645 7.299 58.3 −583.740 354 2500 0.142 −533.775 450 2500 0.180 6-PTGR15 −357.736 55.799 69.261 0.316 4.425 33.7 −332.687 1953 2500 0.781 −338.660 1833 2500 0.733 Average −368.083 52.053 343.523 0.582 4.995 48.064 −369.442 1360 2500 0.544 −358.270 1547 2500 0.619 Passing rate 100% 100% *The interpretation of each column, which also explains the procedure of MSN testing, is listed below: N = 2500 is the number of Gibb samples selected from 25 000 simulated communities (i.e. every tenth iteration of the last 25 000 Gibbs samples), it is chosen to compute the pseudo p-value below for conducting the neutrality test. L0 is the actual  ( observed) log-likelihood. θ is the median of biodiversity parameters computed from 25 000 times of simulations. M-value is the average medians of the migration rates of local communities in each meta-community, also computed from 25,000 times of simulations. LM is the median of the log-likelihoods of the simulated neutral meta-community samples; and NM is the number of simulated neutral meta-community samples with their likelihoods satisfying LM≤ L0. PM=NM/N is the pseudo p-value for testing the neutrality at meta-community level; if pM>0.05, the meta-community satisfies the MSN model. LL is the median of the log-likelihoods of the simulated local community samples, and NL is the number of simulated local community samples with their likelihoods not exceeding the L0. PL=NL/N, is the pseudo P-value for testing the neutrality at the local community level; if pL>0.05, the local community satisfies the neutral model. The definition of ‘community dominance’ follows Ma and Ellison (2018, 2019), which measures community unevenness and can be considered as proxy of community diversity. The species richness is simply the number of species in the community. ** A graph showing the goodness-of-fitting to the MSN neutral model is presented as Fig. 1 for this subject (6-PTGD85). Open in new tab Table 1. Test results of fitting the MSN (multi-site neutral) model to SIV-infected pigtail stool & tissue microbiome datasets*. ID . L0 . θ . M-value . m . Community Dominance . Species Richness . Meta-community . Local community . . . . . . . . LM . NM . N . pM . LL . NL . N . pL . Phylum level 2-PTA0P036-stool −81.931 2.259 72.309 0.221 4.406 9.0 −82.562 1231 2500 0.492 –80.334 1404 2500 0.562 2-PTA0P048-stool −97.083 2.352 179.083 0.302 4.677 13.0 −100.379 1160 2500 0.464 –100.659 892 2500 0.357 2-PTA0P048 −55.809 2.388 61.821 0.393 3.488 7.7 −58.480 1124 2500 0.450 −56.633 1141 2500 0.456 2-PTA2P032 −41.054 1.755 15.777 0.228 1.636 5.3 −33.411 1695 2500 0.678 −38.664 1590 2500 0.636 2-PTFR60-stool −78.245 2.039 129.945 0.274 4.475 9.7 −86.619 998 2500 0.399 −81.660 902 2500 0.361 2-PTFR60 −57.387 2.011 27.793 0.196 2.740 7.0 −50.357 1546 2500 0.618 −53.543 1684 2500 0.674 2-PTGC45-stool −57.116 1.790 97.974 0.347 3.462 8.7 −55.328 1337 2500 0.535 −57.638 1171 2500 0.468 2-PTGC45 −53.176 2.083 16.868 0.168 1.965 6.3 −42.891 1714 2500 0.686 −50.439 1569 2500 0.628 2-PTGD79-stool −51.473 1.993 144.168 0.313 3.920 10.5 −54.563 1120 2500 0.448 −52.180 1170 2500 0.468 2-PTGD79 −42.790 1.778 42.476 0.451 2.193 6.3 −38.321 1481 2500 0.592 −43.138 1201 2500 0.480 2-PTGD85-stool −82.604 1.939 155.632 0.260 4.243 10.0 −90.547 1033 2500 0.413 −86.368 841 2500 0.336 2-PTGD85 −60.936 3.842 5.569 0.049 2.821 6.0 −53.886 1604 2500 0.642 −56.235 1645 2500 0.658 2-PTGE34-stool −47.077 5.672 2.138 0.014 4.196 5.5 −38.810 1747 2500 0.699 −40.530 1701 2500 0.680 2-PTGE34 −69.190 2.254 52.909 0.236 3.225 8.7 −64.459 1423 2500 0.569 −64.897 1633 2500 0.653 2-PTGE36-stool −74.295 1.829 133.537 0.301 3.809 9.3 −75.170 1205 2500 0.482 −77.247 904 2500 0.362 2-PTGE36 −65.465 3.216 12.863 0.085 3.498 7.0 −58.720 1543 2500 0.617 −61.040 1615 2500 0.646 2-PTGE40-stool −54.012 2.059 111.252 0.256 4.112 9.5 −54.767 1206 2500 0.482 −54.529 1195 2500 0.478 2-PTGE40 −39.748 2.207 29.801 0.485 1.894 6.0 −35.468 1531 2500 0.612 −38.812 1401 2500 0.560 2-PTGP27-stool −118.295 2.286 120.788 0.219 4.884 10.0 −132.632 940 2500 0.376 −120.554 1077 2500 0.431 2-PTGP27 −58.213 1.515 88.928 0.280 2.910 7.7 −53.952 1400 2500 0.560 −57.269 1388 2500 0.555 2-PTGR15-stool −87.019 1.342 113.617 0.262 3.430 8.0 −80.382 1426 2500 0.570 −91.354 753 2500 0.301 2-PTGR15 −58.021 2.122 13.878 0.085 2.732 6.0 −49.903 1593 2500 0.637 −53.229 1713 2500 0.685 Average −65.043 2.306 74.051 0.247 3.396 8 −63.255 1366 2500 0.546 −64.407 1300 2500 0.520 Passing rate 100% 100% Genus level 6-PTA0P036-stool −526.258 39.726 351.481 0.579 6.352 68.0 −535.607 1061 2500 0.424 −509.737 1788 2500 0.715 6-PTA0P048-stool −492.134 27.985 759.693 0.647 8.429 76.0 −542.063 472 2500 0.189 −505.627 699 2500 0.280 6-PTA0P048 −304.505 36.346 176.331 0.648 3.921 37.7 −291.763 1666 2500 0.666 −287.250 1917 2500 0.767 6-PTA2P032 −192.575 30.231 96.651 0.644 3.471 23.7 −169.774 2112 2500 0.845 −176.814 2007 2500 0.803 6-PTFR60-stool −504.056 31.988 666.747 0.659 7.879 75.7 −546.929 521 2500 0.208 −511.624 964 2500 0.386 6-PTFR60 −295.847 33.225 209.837 0.648 2.943 38.0 −281.344 1714 2500 0.686 −275.265 2090 2500 0.836 6-PTGC45-stool −276.386 22.235 344.911 0.651 5.054 41.0 −288.308 943 2500 0.377 −281.058 1008 2500 0.403 6-PTGC45 −308.692 61.828 94.992 0.533 2.190 30.0 −266.840 2350 2500 0.940 −284.957 2111 2500 0.844 6-PTGD79-stool −269.169 22.163 541.095 0.631 5.837 62.0 −276.861 1045 2500 0.418 −269.444 1231 2500 0.492 6-PTGD79 −186.193 22.340 156.346 0.752 2.048 27.0 −172.575 1816 2500 0.726 −180.535 1625 2500 0.650 6-PTGD85-stool −434.526 21.868 713.475 0.617 8.729 67.3 −484.371 475 2500 0.190 −451.787 582 2500 0.233 6-PTGD85** −362.903 153.189 63.940 0.372 2.372 27.3 −298.795 2443 2500 0.977 −322.027 2284 2500 0.914 6-PTGE34-stool −351.151 214.483 33.929 0.186 2.600 38.0 −282.458 2417 2500 0.967 −304.384 2294 2500 0.918 6-PTGE34 −314.646 27.072 283.980 0.624 4.323 41.7 −317.053 1171 2500 0.468 −298.127 1924 2500 0.770 6-PTGE36-stool −507.495 35.555 715.271 0.697 7.031 78.7 −551.547 469 2500 0.188 −518.860 820 2500 0.328 6-PTGE36 −496.030 180.459 87.402 0.388 2.785 38.3 −409.745 2471 2500 0.988 −441.977 2350 2500 0.940 6-PTGE40-stool −284.362 25.130 546.073 0.628 6.266 62.0 −300.999 819 2500 0.328 −281.508 1397 2500 0.559 6-PTGE40 −59.404 30.621 82.571 0.723 1.424 19.7 −136.837 2270 2500 0.908 −147.580 1956 2500 0.782 6-PTGP27-stool −633.246 27.238 590.541 0.578 9.108 67.3 −723.025 253 2500 0.101 −644.791 858 2500 0.343 6-PTGP27 −329.178 23.032 391.182 0.631 5.400 46.0 −334.395 1118 2500 0.447 −316.142 1851 2500 0.740 6-PTGR15-stool −511.331 22.661 581.787 0.645 7.299 58.3 −583.740 354 2500 0.142 −533.775 450 2500 0.180 6-PTGR15 −357.736 55.799 69.261 0.316 4.425 33.7 −332.687 1953 2500 0.781 −338.660 1833 2500 0.733 Average −368.083 52.053 343.523 0.582 4.995 48.064 −369.442 1360 2500 0.544 −358.270 1547 2500 0.619 Passing rate 100% 100% ID . L0 . θ . M-value . m . Community Dominance . Species Richness . Meta-community . Local community . . . . . . . . LM . NM . N . pM . LL . NL . N . pL . Phylum level 2-PTA0P036-stool −81.931 2.259 72.309 0.221 4.406 9.0 −82.562 1231 2500 0.492 –80.334 1404 2500 0.562 2-PTA0P048-stool −97.083 2.352 179.083 0.302 4.677 13.0 −100.379 1160 2500 0.464 –100.659 892 2500 0.357 2-PTA0P048 −55.809 2.388 61.821 0.393 3.488 7.7 −58.480 1124 2500 0.450 −56.633 1141 2500 0.456 2-PTA2P032 −41.054 1.755 15.777 0.228 1.636 5.3 −33.411 1695 2500 0.678 −38.664 1590 2500 0.636 2-PTFR60-stool −78.245 2.039 129.945 0.274 4.475 9.7 −86.619 998 2500 0.399 −81.660 902 2500 0.361 2-PTFR60 −57.387 2.011 27.793 0.196 2.740 7.0 −50.357 1546 2500 0.618 −53.543 1684 2500 0.674 2-PTGC45-stool −57.116 1.790 97.974 0.347 3.462 8.7 −55.328 1337 2500 0.535 −57.638 1171 2500 0.468 2-PTGC45 −53.176 2.083 16.868 0.168 1.965 6.3 −42.891 1714 2500 0.686 −50.439 1569 2500 0.628 2-PTGD79-stool −51.473 1.993 144.168 0.313 3.920 10.5 −54.563 1120 2500 0.448 −52.180 1170 2500 0.468 2-PTGD79 −42.790 1.778 42.476 0.451 2.193 6.3 −38.321 1481 2500 0.592 −43.138 1201 2500 0.480 2-PTGD85-stool −82.604 1.939 155.632 0.260 4.243 10.0 −90.547 1033 2500 0.413 −86.368 841 2500 0.336 2-PTGD85 −60.936 3.842 5.569 0.049 2.821 6.0 −53.886 1604 2500 0.642 −56.235 1645 2500 0.658 2-PTGE34-stool −47.077 5.672 2.138 0.014 4.196 5.5 −38.810 1747 2500 0.699 −40.530 1701 2500 0.680 2-PTGE34 −69.190 2.254 52.909 0.236 3.225 8.7 −64.459 1423 2500 0.569 −64.897 1633 2500 0.653 2-PTGE36-stool −74.295 1.829 133.537 0.301 3.809 9.3 −75.170 1205 2500 0.482 −77.247 904 2500 0.362 2-PTGE36 −65.465 3.216 12.863 0.085 3.498 7.0 −58.720 1543 2500 0.617 −61.040 1615 2500 0.646 2-PTGE40-stool −54.012 2.059 111.252 0.256 4.112 9.5 −54.767 1206 2500 0.482 −54.529 1195 2500 0.478 2-PTGE40 −39.748 2.207 29.801 0.485 1.894 6.0 −35.468 1531 2500 0.612 −38.812 1401 2500 0.560 2-PTGP27-stool −118.295 2.286 120.788 0.219 4.884 10.0 −132.632 940 2500 0.376 −120.554 1077 2500 0.431 2-PTGP27 −58.213 1.515 88.928 0.280 2.910 7.7 −53.952 1400 2500 0.560 −57.269 1388 2500 0.555 2-PTGR15-stool −87.019 1.342 113.617 0.262 3.430 8.0 −80.382 1426 2500 0.570 −91.354 753 2500 0.301 2-PTGR15 −58.021 2.122 13.878 0.085 2.732 6.0 −49.903 1593 2500 0.637 −53.229 1713 2500 0.685 Average −65.043 2.306 74.051 0.247 3.396 8 −63.255 1366 2500 0.546 −64.407 1300 2500 0.520 Passing rate 100% 100% Genus level 6-PTA0P036-stool −526.258 39.726 351.481 0.579 6.352 68.0 −535.607 1061 2500 0.424 −509.737 1788 2500 0.715 6-PTA0P048-stool −492.134 27.985 759.693 0.647 8.429 76.0 −542.063 472 2500 0.189 −505.627 699 2500 0.280 6-PTA0P048 −304.505 36.346 176.331 0.648 3.921 37.7 −291.763 1666 2500 0.666 −287.250 1917 2500 0.767 6-PTA2P032 −192.575 30.231 96.651 0.644 3.471 23.7 −169.774 2112 2500 0.845 −176.814 2007 2500 0.803 6-PTFR60-stool −504.056 31.988 666.747 0.659 7.879 75.7 −546.929 521 2500 0.208 −511.624 964 2500 0.386 6-PTFR60 −295.847 33.225 209.837 0.648 2.943 38.0 −281.344 1714 2500 0.686 −275.265 2090 2500 0.836 6-PTGC45-stool −276.386 22.235 344.911 0.651 5.054 41.0 −288.308 943 2500 0.377 −281.058 1008 2500 0.403 6-PTGC45 −308.692 61.828 94.992 0.533 2.190 30.0 −266.840 2350 2500 0.940 −284.957 2111 2500 0.844 6-PTGD79-stool −269.169 22.163 541.095 0.631 5.837 62.0 −276.861 1045 2500 0.418 −269.444 1231 2500 0.492 6-PTGD79 −186.193 22.340 156.346 0.752 2.048 27.0 −172.575 1816 2500 0.726 −180.535 1625 2500 0.650 6-PTGD85-stool −434.526 21.868 713.475 0.617 8.729 67.3 −484.371 475 2500 0.190 −451.787 582 2500 0.233 6-PTGD85** −362.903 153.189 63.940 0.372 2.372 27.3 −298.795 2443 2500 0.977 −322.027 2284 2500 0.914 6-PTGE34-stool −351.151 214.483 33.929 0.186 2.600 38.0 −282.458 2417 2500 0.967 −304.384 2294 2500 0.918 6-PTGE34 −314.646 27.072 283.980 0.624 4.323 41.7 −317.053 1171 2500 0.468 −298.127 1924 2500 0.770 6-PTGE36-stool −507.495 35.555 715.271 0.697 7.031 78.7 −551.547 469 2500 0.188 −518.860 820 2500 0.328 6-PTGE36 −496.030 180.459 87.402 0.388 2.785 38.3 −409.745 2471 2500 0.988 −441.977 2350 2500 0.940 6-PTGE40-stool −284.362 25.130 546.073 0.628 6.266 62.0 −300.999 819 2500 0.328 −281.508 1397 2500 0.559 6-PTGE40 −59.404 30.621 82.571 0.723 1.424 19.7 −136.837 2270 2500 0.908 −147.580 1956 2500 0.782 6-PTGP27-stool −633.246 27.238 590.541 0.578 9.108 67.3 −723.025 253 2500 0.101 −644.791 858 2500 0.343 6-PTGP27 −329.178 23.032 391.182 0.631 5.400 46.0 −334.395 1118 2500 0.447 −316.142 1851 2500 0.740 6-PTGR15-stool −511.331 22.661 581.787 0.645 7.299 58.3 −583.740 354 2500 0.142 −533.775 450 2500 0.180 6-PTGR15 −357.736 55.799 69.261 0.316 4.425 33.7 −332.687 1953 2500 0.781 −338.660 1833 2500 0.733 Average −368.083 52.053 343.523 0.582 4.995 48.064 −369.442 1360 2500 0.544 −358.270 1547 2500 0.619 Passing rate 100% 100% *The interpretation of each column, which also explains the procedure of MSN testing, is listed below: N = 2500 is the number of Gibb samples selected from 25 000 simulated communities (i.e. every tenth iteration of the last 25 000 Gibbs samples), it is chosen to compute the pseudo p-value below for conducting the neutrality test. L0 is the actual  ( observed) log-likelihood. θ is the median of biodiversity parameters computed from 25 000 times of simulations. M-value is the average medians of the migration rates of local communities in each meta-community, also computed from 25,000 times of simulations. LM is the median of the log-likelihoods of the simulated neutral meta-community samples; and NM is the number of simulated neutral meta-community samples with their likelihoods satisfying LM≤ L0. PM=NM/N is the pseudo p-value for testing the neutrality at meta-community level; if pM>0.05, the meta-community satisfies the MSN model. LL is the median of the log-likelihoods of the simulated local community samples, and NL is the number of simulated local community samples with their likelihoods not exceeding the L0. PL=NL/N, is the pseudo P-value for testing the neutrality at the local community level; if pL>0.05, the local community satisfies the neutral model. The definition of ‘community dominance’ follows Ma and Ellison (2018, 2019), which measures community unevenness and can be considered as proxy of community diversity. The species richness is simply the number of species in the community. ** A graph showing the goodness-of-fitting to the MSN neutral model is presented as Fig. 1 for this subject (6-PTGD85). Open in new tab RESULTS Table 1, Tables S1 and S2 (Supporting Information) exhibited the results from fitting the multi-site neutral (MSN) model to SIV-infected pigtail stool & tissue (Fig. 1), SIV-infected pigtail stool (Fig. S1, Supporting Information), and SIV-infected rhesus tissue (Fig. S2, Supporting Information) microbiome data sets, respectively. In each table, the fitting results at phylum and genus level were reported respectively. From these results, we observed the following findings: All tested data sets (100%) passed the neutrality test with the MSN model (P-value > 0.10), both in multiple time-point stool data (Table S2, Supporting Information) and multiple-tissue data (Table 1; Table S3, Supporting Information). The larger P-value represents better fitting to the MSN model. The success of the neutrality tests with the MSN model suggests that the community assembly and diversity maintenance after SIV infection were driven by stochastic neutral forces. Dispersal and drift were the major driving forces that shaped the observed community structure and diversity. We postulate that the mode of microbial dispersal (translocation) from GI tract (colon) to tissues (liver and MLN), once triggered by SIV infections, is a stochastic neutral process, governed by the MSN neutral model. Figure 1. Open in new tabDownload slide A graph exhibiting the goodness-of-fitting to the MSN neutral model for the subject ‘6-PTGD85’ in the treatment of pigtail stool & tissue (Table 1) at the microbial genus level. Figure 1. Open in new tabDownload slide A graph exhibiting the goodness-of-fitting to the MSN neutral model for the subject ‘6-PTGD85’ in the treatment of pigtail stool & tissue (Table 1) at the microbial genus level. It should be clarified that although the translocation from GI tract to tissues was driven by stochastic (neutral) forces as revealed by 100% neutrality passing rates with MSN testing, the stochastic nature of the translocation does not imply that the event that triggers (initiates) the translocation is random, which is obviously not in the case of HIV/SIV infection. Instead, it means that, once the translocation is triggered, the process is stochastic or random without a need for further external driving force. The average migration probability (m) reported in Table S2 (Supporting Information) (m = 0.006 at phylum level, m= 0.01 at genus level), representing the natural changes occurred among three time points, is significantly lower than those (m) reported in Table 1 (m = 0.247 for phylum 0.582 for genus) and Table S3 (Supporting Information) (m = 0.221 for phylum 0.538 for genus), both the tables representing migration/dispersal from GI colon to liver and MLNs (i.e. the translocation). The translocation probability is approximately 37–41 times (at the phylum level) or 54–58 times (at the genus level) larger than the normal time series changes as exhibited in Table S2 (Supporting Information) and the difference is statistically significant based on non-parametric test (P-value < 0.001). This is, of course, as expected, revealing that the magnitude of translocation is far more significant than the natural migration (diffusion among three time points within GI tract). Fig. 2 shows the significantly higher migration probability (m) from GI tract to tissues (estimated in Table 1 and Table S3 (Supporting Information) for the pigtail stool & tissue and rhesus tissue, respectively), compared with the base-level dispersal occurring within GI tract (estimated from time series samples of stool in Table S2, Supporting Information) (P-value < 0.001). Figure 2. Open in new tabDownload slide The box plot showing significantly higher migration probability (m) from GI tract to tissues (estimated in Table 1 and Table S3, Supporting Information, for the pigtail stool & tissue and rhesus tissue, respectively) (P-value < 0.001), compared with the base-level dispersal occurring within GI tract (estimated from the time-series samples of pigtail stool in Table S2, Supporting Information). Box red lines, blue lines, edges, whiskers, and bigger red points signify the median, mean, interquartile range (IQR), 1.5× IQR and > 1.5× IQR, respectively. The smaller points in each box are the real values of m of each sample. Figure 2. Open in new tabDownload slide The box plot showing significantly higher migration probability (m) from GI tract to tissues (estimated in Table 1 and Table S3, Supporting Information, for the pigtail stool & tissue and rhesus tissue, respectively) (P-value < 0.001), compared with the base-level dispersal occurring within GI tract (estimated from the time-series samples of pigtail stool in Table S2, Supporting Information). Box red lines, blue lines, edges, whiskers, and bigger red points signify the median, mean, interquartile range (IQR), 1.5× IQR and > 1.5× IQR, respectively. The smaller points in each box are the real values of m of each sample. As a side note, while the migration probability (or translocation probability in our study) from GI tract to tissues, estimated with either pigtail (Table 1) or rhesus (Table S3, Supporting Information) microbiomes are significantly higher than the naturally occurring migration (diffusion) within GI tract (as estimated from the time-series stool samples in Table S2, Supporting Information) (P <0.001), the migration probability between estimated for pigtail and rhesus datasets showed no significant difference (P = 0.488 at the phylum level and P = 0.308 at the genus level). This lack of difference between the two Asian macaques in their translocation probability from GI tract to tissue should be expected because it simply cross-validate the same mechanism of translocation between both species of macaques. Using the natural migration probability within GI tract (i.e. three time series points of stool samples) as a baseline migration rate, we postulate that the translocation probability from GI tract to tissue should be [(0.247 + 0.221)/2–0.006]=0.228 at the phylum level, and [(0.538 + 0.582)/2–0.01]=0.55 at the genus level. From these probabilities, we further estimate that approximately 23% of phyla and 55% of genera were migrated from GI tract to the liver and MLNs. Fig. 3 draws the plots of average community dominance and species richness for stool and each of the three tissues (colon, liver and MLNs) based on the parameters in Table 1. The size order (stool>colon>liver>MLNs) of either dominance or richness metric reflects the potential to dispersal. That is, the higher the dominance (richness) is, the higher the potential for microbes to disperse (migrate) to other tissues. This observation simply confirms the fact that translocation occurred from GI tract to other tissues. Figure 3. Open in new tabDownload slide A graph showing community dominance (phylum level & genus level) and species richness (phylum and genus level): the magnitude order (stool>colon>liver>MLNs) of either dominance or richness metric reflects the potential to disperse from high to low. Figure 3. Open in new tabDownload slide A graph showing community dominance (phylum level & genus level) and species richness (phylum and genus level): the magnitude order (stool>colon>liver>MLNs) of either dominance or richness metric reflects the potential to disperse from high to low. DISCUSSION Microbial translocation is an important phenomenon in many diseases, such as inflammatory bowel disease and HIV/SIV infection (Alzahrani et al. 2019; Kumar et al. 2020), which is triggered by damage to the structural and/or immunological barrier of the GI tract (Cosata et al. 2016; Stanley et al. 2016; Manfredo Vieira et al. 2018). Specifically, investigations demonstrated that HIV/SIV-associated microbial translocation results from a series of immunopathological events, including (i) rapid depletion of GI tract mucosal CD4+ T cells, with a significant loss of Th17 cells; (ii) mucosal immune hyperactivation or persistent inflammation, such as decreased production of IL-17 and IL-22 effector cytokines by Th17 cells; (iii) epithelial cell apoptosis and tight junction disruption with loss of epithelial integrity and (iv) dysbiosis of the GI tract microbiome, with a predominance of opportunistic bacteria and depletion of probiotic species (Marchetti, Tincati and Silvestri 2013; Shan and Siliciano 2014; Hel et al. 2017; Neff et al. 2018; Alzahrani et al. 2019; Crakes and Jiang 2019; Vinto et al. 2020). For example, Hensley-McBain et al. (2018) found that mucosal damage, including epithelial barrier damage and microbial translocation, preceded mucosal immune dysfunction in pathogenic SIV infection. Epithelial barrier damage occurred 3–14 days post-SIV infection, and microbial translocation was observed 14 days post-SIV (Hensley-McBain et al. 2018). For another example, Giron et al. (2020) found that the alteration of gut glycosylation plays roles in HIV-associated microbial translocation. When microbes escape the gut barrier and get access to the systemic circulation, MLNs and liver would act as further ‘firewalls’ to control the translocation of microbial components into blood stream (Balmer et be al. 2014; Spadoni et al. 2015; Manfredo Vieira et al. 2018). Studies have shown that gut microbial translocation to liver, MLNs and other systemic tissues could trigger and promote autoimmunity (Ma et al. 2018; Manfredo Vieira et al. 2018), which plays an important role in several non-AIDs comorbiditis, such as liver fibrosis, non-alcoholic fatty liver disease, cirrhosis, autoimmune hepatitis and systemic lupus erythematosus (Marchetti, Tincati and Silvestri 2013; Swanson et al. 2016; Ma et al. 2018; Manfredo Vieira et al. 2018; Reid et al. 2018). In summary, "leaky gut" may lead to microbial translocation, and microbes or their products entering the circulation may directly stimulate the immune responses, making the host susceptible to various types of diseases via inducing chronic or acute inflammatory response. Given a close association with chronic inflammation during HIV/SIV infection, microbial translocation has become a hot topic in the clinical medicine. If the microbial translocation can be detected and blocked early in the early stage of HIV infection, it would be helpful to prevent the complications of HIV infection and improve the therapeutic effect. In the present study, via neutral-theoretic modeling analysis, we determined the mode of microbial translocation triggered by HIV/SIV infection, and estimated the relocation probability and level. The distribution of microbiome diversity, like other biomes on the earth planet, is shaped by the four processes, i.e. drift, dispersal, speciation and selection (Vellend 2010; Hanson et al. 2012; Rosindell et al. 2012; Li & Ma 2019). However, in different biomes (microbiomes), the relative importance, especially the balance between stochastic drift and deterministic selection can vary. Although the topic has been investigated in several studies on the microbiomes distributed in or on animal/human bodies, to the best of our knowledge, it has not been studied for any tissue microbiome. Using the microbiome datasets of colon, liver, and MLNs obtained from SIV-infected Asian macaques, we apply Hubbell's unified neutral theory of biodiversity (UNTB) for determining the role of stochastic drift and dispersal in shaping the distribution of microbiome in the monkey body, especially the translocation of gut microbiome to various tissues after SIV infection. Methodologically, we use the multi-site neutral (MSN) model developed by Harris et al. (2017), which allows for more accurate and efficient estimations of the migration rates among multiple sites. Harris et al. (2017) MSN model is particularly advantageous for our objectives in this study for two obvious reasons. One is that it allows for different migration rates among different local communities, i.e. translocations of microbes among different tissues or sites can be different, which is simply a biological reality. The MSN model therefore enables us to achieve more realistic estimations of the translocations rates. Second, previous applications of the neutral theory to human microbiomes (Li and Ma 2016; Li et al2018) have only used single site datasets, i.e. one microbiome sample from each individual subject. In contrast, the microbiome datasets (Table S1, Supporting Information) we used in this study are truly multi-site samples, i.e. multiple microbiome samples were taken from each individual subject, and constitute a true metacommunity with realistic dispersal (migration) occurring at ecological time scale. Thus, Harris et al. (2017) MSN model offers an ideal neutral theory model to fully take advantages of the datasets and should produce more reliable testing results of the translocation mode – whether it is stochastic or deterministic. A stochastic mode would indicate that the translocation is likely a passive dispersal, and a deterministic mode would indicate that some selective forces most likely drive the translocation. Our previous neutral theory analysis with the MSN modeling show that (i) All macaque tissue microbiome samples (100%) we tested passed the neutrality test, indicating that stochastic drift and dispersal (migration) play a predominant role in shaping the distribution of monkey microbiomes. The 100% of neutrality is somewhat exceptional given that majority existing studies of the human microbiomes showed limited influence of neutral forces in shaping the microbiome distribution (e.g. Harris et al 2017, Venkataraman et al. 2015, Zeng et al 2015, Li and Ma 2016, Li et al2018). (ii) The bacteria translocation probability from GI tract to tissues is approximately 37–41 times (at the phylum level) or 54–58 times (at the genus level) larger than the normal diffusion naturally occurring in the gut microbiome (within GI tract). It is further estimated that approximately 23% of phyla and 55% of genera were migrated from GI tract to other tissues (liver and MLNs). Our findings therefore provide the first interpretation of the translocation mode from gut to tissues (i.e. the translocation was likely induced by SIV infection, but the process is largely random or stochastic), and the first estimates of the translocation probability and the translocated portions of microbiomes. Obviously, these estimates, although derived from rigorous testing of the neutral theory model, are subject to future experimental validation, which is beyond the reach of the previous article, but certainly should be conducted to truly advance the field. Our study is only a preliminary step in proposing a potentially meaningful model hypothesis to explore this important topic. Finally, we would like to reiterate some practical limitation of our analysis. First, we realized that the two discoveries about translocation probability and transferred fraction of the microbes we made require further experimental validation with animal models, because our findings strongly leveraged the theoretical modeling. Nevertheless, given the extreme difficulty to estimate those numbers in practice, our theoretical derivations still offer a valuable starting point to further quantify those important numbers, which is of extremely important biomedical significance in our opinion. Second, although we mentioned HIV in previous sections, all the data sets and analyses in the manuscript were limited to SIV, and it is further acknowledged that the data sets were originally collected and published by Klase et al. (2015). DATA ACCESSIBILITY The raw sequencing data sets were originally collected and published by Klase et al. (2015). Detailed access information was available in Klase et al. (2015) dysbiotic bacteria translocate in progressive SIV infection. Mucosal Immunology, Doi:10.1038/mi.2014.128’ ACKNOWLEDGEMENTS We are deeply indebted to Dr Z Klase & Dr JM Brenchley, NIH for their generous support and help in providing the microbiome datasets (16s-rRNA OTU tables) used in this study. We appreciate Mr Lianwei Li, the Computational Biology and Medical Ecology Lab, the Chinese Academy of Sciences for his significant support in the data analysis. AUTHOR CONTRIBUTIONS ZM designed the study, interpreted the results and wrote the paper. WL performed the data analysis and participated in the interpretations. All authors agreed the submission. Conflicts of interest None declared. REFERENCES Alzahrani J , Hussain T, Simar D et al. Inflammatory and Immunometabolic Consequences of Gut Dysfunction in HIV: Parallels With IBD and Implications for Reservoir Persistence and non-AIDS Comorbidities . EBioMedicine , 2019 ; 46 : 522 – 31 . Google Scholar Crossref Search ADS PubMed WorldCat Arnold CP , Merryman MS, Harris-Arnold A et al. Pathogenic shifts in endogenous microbiota impede tissue regeneration via distinct activation of TAK1/MKK/p38 . eLife . 2016 ; 5 : e16793 . Google Scholar Crossref Search ADS PubMed WorldCat Balmer ML , Slack E, de Gottardi A et al. The Liver May Act as a Firewall Mediating Mutualism Between the Host and Its Gut Commensal Microbiota . Sci Transl Med . 2014 ; 6 : 237ra66 . Google Scholar Crossref Search ADS PubMed WorldCat Cavarretta I , Ferrarese R, Cazzaniga W et al. The Microbiome of the Prostate Tumor Microenvironment . Eur Urol . 2017 ; 72 : 625 – 31 . Google Scholar Crossref Search ADS PubMed WorldCat Costa FRC , Françozo MCS, de Oliveira GG et al. Gut Microbiota Translocation to the Pancreatic Lymph Nodes Triggers NOD2 Activation and Contributes to T1D Onset . J Exp Med . 2016 ; 213 : 1223 – 39 . Google Scholar Crossref Search ADS PubMed WorldCat Crakes KR , Jiang G. Gut Microbiome Alterations During HIV/SIV Infection: Implications for HIV Cure . Front Microbiol . 2019 ; 10 : 1104 . Google Scholar Crossref Search ADS PubMed WorldCat Etienne RS . A neutral sampling formula for multiple samples and an ‘exact’ test of neutrality . Ecol Lett . 2007 ; 10 : 608 – 18 . Google Scholar Crossref Search ADS PubMed WorldCat Etienne RS . A new sampling formula for neutral biodiversity: A new sampling formula . Ecol Lett . 2005 ; 8 : 253 – 60 . Google Scholar Crossref Search ADS WorldCat Giron LB , Tanes CE, Schleimann MH et al. Sialylation and fucosylation modulate inflammasome-activating eIF2 Signaling and microbial translocation during HIV infection . Mucosal Immunol . 2020 ; 1 – 14 . OpenURL Placeholder Text WorldCat Hanson CA , Fuhrman JA, Horner-Devine MC et al. Beyond biogeographic patterns: Processes shaping the microbial landscape . Nat Rev Microbiol . 2012 ; 10 : 497 – 506 . Google Scholar Crossref Search ADS PubMed WorldCat Harris K , Parsons TL, Ijaz UZ et al. Linking Statistical and Ecological Theory: Hubbell's Unified Neutral Theory of Biodiversity as a Hierarchical Dirichlet Process . Proc IEEE . 2017 ; 105 : 516 – 29 . Google Scholar Crossref Search ADS WorldCat Hel Z , Xu J, Denning WL et al. Dysregulation of Systemic and Mucosal Humoral Responses to Microbial and Food Antigens as a Factor Contributing to Microbial Translocation and Chronic Inflammation in HIV-1 Infection . PLoS Pathog . 2017 ; 13 : e1006087 . Google Scholar Crossref Search ADS PubMed WorldCat Hensley-McBain T , Berard AR, Manuzak JA et al. Intestinal Damage Precedes Mucosal Immune Dysfunction in SIV Infection . Mucosal Immunol . 2018 ; 11 : 1429 – 40 . Google Scholar Crossref Search ADS PubMed WorldCat HMP Consortium . A framework for human microbiome research . Nature . 2012 ; 486 : 215 . Crossref Search ADS PubMed WorldCat Hubbell SP . The unified neutral theory of biodiversity and biogeography . Princeton, NJ. : Princeton University Press , 2001 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Kim HJ , Kim YS, Kim KH et al. The microbiome of the lung and its extracellular vesicles in nonsmokers, healthy smokers and COPD patients . Exp Mol Med . 2016 ; 49 : e316 . Google Scholar Crossref Search ADS WorldCat Klase Z , Ortiz A, Deleage C et al. Dysbiotic bacteria translocate in progressive SIV infection . Mucosal Immunol . 2015 , DOI:10.1038/mi.2014.128 . OpenURL Placeholder Text WorldCat Kumar M , Coria AL, Cornick S et al. Increased intestinal permeability exacerbates sepsis through reduced hepatic SCD-1 activity and dysregulated iron recycling . Nat Commun . 2020 ; 11 : 483 Google Scholar Crossref Search ADS PubMed WorldCat Li LW , Ma ZS. Testing the Neutral Theory of Biodiversity with Human Microbiome Datasets . Sci Rep . 2016 ; 6 , Article No. 31448 . OpenURL Placeholder Text WorldCat Li W , Ma ZS.. Diversity scaling of human vaginal microbial communities . Zool Res . 2019 ; 40 : 587 – 94 . Google Scholar Crossref Search ADS PubMed WorldCat Li W , Yuan Y, Xia Y et al. A cross-scale neutral theory approach to the influence of obesity on community assembly of human gut microbiome . Front Microbiol . 2018 ; 2018 . OpenURL Placeholder Text WorldCat Ma HD , Zhao ZB, Ma WT et al. Gut Microbiota Translocation Promotes Autoimmune Cholangitis . J Autoimmun . 2018 ; 95 : 47 – 57 . Google Scholar Crossref Search ADS PubMed WorldCat Manfredo Vieira S , Hiltensperger M, Kumar V et al. Translocation of a Gut Pathobiont Drives Autoimmunity in Mice and Humans . Science . 2018 ; 359 : 1156 – 61 . Google Scholar Crossref Search ADS PubMed WorldCat Marchetti G , Tincati C, Silverstri G. Microbial Translocation in the Pathogenesis of HIV Infection and AIDS . Clin Microbiol Rev . 2013 ; 26 : 2 – 18 . Google Scholar Crossref Search ADS PubMed WorldCat Ma ZS , Ellison AM. Dominance network analysis provides a new framework for studying the diversity-stability relationship . Ecological Monographs . 2019 . https://esajournals.onlinelibrary.wiley.com/doi/full/10.1002/ecm.1358 . OpenURL Placeholder Text WorldCat Ma ZS , Ellison AM. A unified concept of dominance applicable at both community and species scale . Ecosphere . 2018 . OpenURL Placeholder Text WorldCat Neff CP , Krueger O, Xiong K et al. Fecal microbiota composition drives immune activation in HIV infected individuals . EBioMedicine . 2018 ; 30 : 192 – 202 . Google Scholar Crossref Search ADS PubMed WorldCat Paisse S , Valle C, Servant F et al. Comprehensive description of blood microbiome from healthy donors assessed by 16S targeted metagenomic sequencing . Transfusion . 2016 ; 56 : 1138 – 47 . Google Scholar Crossref Search ADS PubMed WorldCat Reid M , Ma Y, Scherzer R et al. Contribution of Liver Fibrosis and Microbial Translocation to Immune Activation in Persons Infected With HIV and/or Hepatitis C Virus . J Infect Dis . 2018 ; 217 : 1289 – 97 . Google Scholar Crossref Search ADS PubMed WorldCat Rosindell J , Hubbell SP, He F et al. The case for ecological neutral theory . Trends Ecol Evol . 2012 ; 27 : 203 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat Shan L , Siliciano RF. Unraveling the relationship between microbial translocation and systemic immune activation in HIV infection . J Clin Invest . 2014 ; 124 : 2368 – 71 . Google Scholar Crossref Search ADS PubMed WorldCat Sloan WT , Lunn M, Woodcock S et al. Quantifying the roles of immigration and chance in shaping prokaryote community structure . Environ Microbiol . 2006 ; 8 : 732 – 40 . Google Scholar Crossref Search ADS PubMed WorldCat Sloan WT , Woodcock S, Lunn M et al. Modeling taxa-abundance distributions in microbial communities using environmental sequence data . Microb Ecol . 2007 ; 53 : 443 – 55 . Google Scholar Crossref Search ADS PubMed WorldCat Spadoni I , Zagato E, Bertocchi A et al. A Gut-Vascular Barrier Controls the Systemic Dissemination of Bacteria . Science . 2015 ; 350 : 830 – 4 . Google Scholar Crossref Search ADS PubMed WorldCat Stanley D , Mason LJ, Mackin KE et al. Translocation and Dissemination of Commensal Bacteria in Post-Stroke Infection . Net Med . 2016 ; 22 : 1277 – 84 . Google Scholar Crossref Search ADS WorldCat Swanson S , Ma Y, Scherzer R et al. Association of HIV, Hepatitis C Virus, and Liver Fibrosis Severity With the Enhanced Liver Fibrosis Score . J Infect Dis . 2016 ; 213 : 1079 – 86 . Google Scholar Crossref Search ADS PubMed WorldCat Sze MA , Dimitriu PA, Hayashi S et al. The Lung Tissue Microbiome in Chronic Obstructive Pulmonary Disease . Am J Respir Crit Care Med . 2012 ; 185 : 1073 – 80 . Google Scholar Crossref Search ADS PubMed WorldCat Urbaniak C , Gloor GB, Brackstone M et al. The Microbiota of Breast Tissue and its Association with Breast Cancer . Appl Environ Microbiol . 2016 ; 82 : 5039 – 48 . Google Scholar Crossref Search ADS PubMed WorldCat Vellend M . Conceptual synthesis in community ecology . Q Rev Biol . 2010 ; 85 : 183 – 206 . Google Scholar Crossref Search ADS PubMed WorldCat Venkataraman A , Bassis C M, Beck J M et al. Application of a neutral community model to assess structuring of the human lung microbiome . Mbio . 2015 ; 6 . OpenURL Placeholder Text WorldCat Vinton CL , Starke CE, Ortiz AM et al. Biomarkers of Cellular Stress Do Not Associate With sCD14 in Progressive HIV and SIV Infections in Vivo . Pathog Immun . 2020 ; 5 : 68 – 88 . Google Scholar Crossref Search ADS PubMed WorldCat Yu GQ , Gail MH, Consonni D et al. Characterizing human lung tissue microbiota and its relationship to epidemiological and clinical features . Genome Biol . 2016 ; 17 : 163 . Google Scholar Crossref Search ADS PubMed WorldCat Zeng Q et al. Neutral Models of Microbiome Evolution . PLoS Comput Biol . 2015 ; 11 : e1004365 . Google Scholar Crossref Search ADS PubMed WorldCat © FEMS 2020. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - A theoretic approach to the mode of gut microbiome translocation in SIV-infected Asian macaques
JF - FEMS Microbiology Ecology
DO - 10.1093/femsec/fiaa134
DA - 2020-08-01
UR - https://www.deepdyve.com/lp/oxford-university-press/a-theoretic-approach-to-the-mode-of-gut-microbiome-translocation-in-KsGn0tpFjg
VL - 96
IS - 8
DP - DeepDyve
ER -