virtual microbiome summit research higlights

Highlights from the Virtual Microbiome Summit 2020

This post is long overdue, but I only recently had a chance to finish watching all of the recordings!

It was truly a blast to host this summit. I am so grateful to have been able to connect with so many incredible researchers and bring together such an amazing event.

In total, we had 25+ speakers, nearly 3,000 participants registered, 700 people on the live webinar at once, and 45 different countries represented.

This article will be a summary of the keynote and plenary talks from the summit. Fair warning that it is quite long, so feel free to skim for the topics that interest you.

If you enjoy this article, be sure to check out the full summit recordings, as there were a number of excellent students, postdoctoral fellows, and assistant scientists who spoke on everything from microbial drug metabolism to the infant microbiome to the effects of stress on gut barrier function!

The butyrate-producer Anaerostipes caccae is protective against food allergies

Dr. Cathryn Nagler from the University of Chicago kicked off the summit with a keynote talk on the role of gut bacteria in food allergies. Over 32 million Americans – about 1 in 10 adults and 1 in 13 children – now have at least one food allergy. Many people have multiple food allergies, along with other related diseases, including eczema, asthma, and allergic rhinitis. Several modern industrialized lifestyle factors are proposed to have contributed to the dramatic increase in food allergies, including increased use of antibiotics, a processed diet, increased urban living, birth by Cesarean section, formula feeding, and a lack of exposure to helminths and other infectious microbes. In a sensitized individual, a food allergic reaction occurs when an allergenic protein crosses the intestinal barrier and gets into the bloodstream, where mast cells and other components of the immune system launch an inflammatory response.

Dr. Nagler’s laboratory has identified a specific butyrate-producing bacterium, Anaerostipes caccae, that induces gut barrier protection by increasing the expression of the cytokine IL-22. Infants with cow’s milk allergy tend to have a much lower abundance of A. caccae in the ileum, the final third of the small intestine. Incredibly, colonizing germ-free mice with just this single species protected them against allergic responses. Dr. Nagler’s team is now working on a startup, ClostraBio, in the hopes of translating these findings into a treatment for human food allergies. They eventually hope to produce a live biotherapeutic of A. caccae, but this won’t be simple given the oxygen sensitivity of this obligate anaerobe. In the meantime, they are exploring ways to provide butyrate directly to the ileum using a polymer delivery platform. Initial animal studies in peanut-sensitized mice have shown promise for using this ileum-targeted butyrate in combination with oral immunotherapy. Dr. Nagler hopes to move on to clinical trials in young adults within the next year. Once developed, their biotherapeutic or ileum-targeted butyrate may have benefits for other conditions as well, including inflammatory bowel disease.

Low dietary fiber and specific gut microbes promote colonic inflammation

Dr. Eric Martens from the University of Michigan presented their work on the interplay between dietary fiber and the host mucus layer. Prior studies in the Martens lab had demonstrated that, in the absence of dietary fiber, mucin-degrading Bacteroides and Akkermansia species will expand in the gut and degrade the gut mucus layer. This causes low-grade inflammation and increases susceptibility to gut infection. (Importantly, this fiber-free diet is also quite high in simple sugars.) Most recently, the lab went on to test a fiber-free diet in the IL-10 genetic knockout model of inflammatory bowel disease (IBD). Fiber-deprived mice on this genetic background had a very thin mucus layer and a cytokine profile typical of IBD. Notably, adding fiber back into the diet blocked the progression of disease, suggesting that at least in this mouse model, fiber can be used to treat disease that has already been initiated.

Dr. Martens also presented some interesting data on exclusive enteric nutrition (EEN), a liquid formula diet that has been shown to be effective as a treatment for human Crohn’s disease. Given the absence of fiber in this diet, it is curious that is so effective for the treatment of Crohn’s disease. Interestingly, when genetically predisposed mice were put on EEN, many of the mice had elevated inflammatory markers. Most surprisingly, however, EEN increased the abundance of Eubacterium rectale, a butyrate producer typically thought to rely on dietary fiber. Future studies will seek to better understand the role of fiber in restoring balance in the gut and the role of mucus-degrading microbes in gut homeostasis.

Hyperglycemia is associated with gut barrier dysfunction and increased susceptibility to gut infection.

Dr. Cristoph Thaiss from the University of Pennsylvania presented on microbiome dynamics in metabolic disease. Several studies had previously found that a higher body mass index was associated with a higher risk of infection, but the underlying mechanism was unknown. Through a series of experiments, Dr. Thaiss and colleagues determined that hyperglycemia, rather than obesity itself, results in epigenetic reprogramming of gut epithelial cells, leading to gut barrier dysfunction, increased susceptibility to the gut pathogen Citrobacter rodentium, and an increase in microbial products entering systemic circulation. In a human cohort, they found that HbA1c, a marker of long-term blood glucose regulation, was strongly correlated with microbial molecules in circulation.

Dr. Thaiss and colleagues decided to look more closely at the various glucose transporters that are present in intestinal epithelial cells. On the apical side, facing the lumen, intestinal epithelial cells have a sodium-glucose symporter (SGLT1), which absorbs glucose from the intestinal lumen. On the basolateral side, closest to the bloodstream, GLUT2 facilitates glucose transport along its concentration gradient. Normally, this serves to move glucose from the epithelial cell into the bloodstream, where it will go to other tissues. However, in the case of hyperglycemia, the researchers hypothesized that glucose levels in the interstitial fluid of the gut are higher than inside intestinal epithelial cells, causing GLUT2 to facilitate a backward flow of glucose into the intestinal epithelium. Indeed, selectively knocking out GLUT2 or blocking glucose metabolism in intestinal epithelial cells led to at least partial restoration of intestinal barrier function and prevented the spread of microbial products into systemic circulation. This study suggests that addressing chronic hyperglycemia is essential to restoring gut homeostasis.

Diet as a lever to improve the industrialized microbiota: the case of fiber and fermented foods

Dr. Erica Sonnenburg from Stanford University kicked off the afternoon session with a talk on the industrialized microbiota. Over the last decade, the Sonnenburg lab has collected several lines of evidence that support a major change in the gut microbiome in recent history. Study of Hadza hunter-gatherers, Nepalese populations across a lifestyle gradient, and the irreversible loss of microbial diversity seen over generations of laboratory mice fed a low fiber diet all support the notion that the industrialized microbiota may be predisposing us to disease. Dr. Sonnenburg reasoned that the shift to a modern diet and lifestyle may have created a mismatch between our human genome and our microbiome. But could improving our diet repair the microbiota and improve health? Unlike many other means of manipulating the microbiota that require extensive testing in animal models, dietary changes are quite safe and can be studied directly in humans.

The researchers enrolled 18 healthy volunteers for a high-fiber intervention and 18 for a high fermented food intervention. Subjects in each group were instructed to slowly ramp up fiber or fermented foods over a 4-week period, maintain consumption for an additional 6 weeks, and then have a 4-week choice period. They found that fermented foods, but not fiber, resulted in a cohort-wide reduction in inflammatory markers and increased microbiota diversity. This was not simply attributed to species found in the fermented foods, suggesting that something about fermented foods makes the gut permissive to new species. The high fiber intervention, on the other hand, resulted in individualized immune responses, with some individuals having a reduction in inflammatory markers, others having no response, and still others having an increase in inflammation from the increased fiber intake. Most interestingly, those with the highest microbial diversity at baseline tended to have a more beneficial immune response to fiber. Dr. Sonnenburg emphasized that there are still many open questions, but that these diet intervention studies are important to inform public health and policy and create a food environment that is based on science and can serve to improve global health.

Unveiling the “black box”: from precise microbial genomics to precision medicine

Dr. Ami Bhatt from Stanford gave an excellent talk on how we might use more precise microbial genomics to influence precision medicine. This was definitely one of the more technical of the talks presented at the summit but has major implications for the future of microbiome discovery. Currently, we can cure or treat disease using what Dr. Bhatt refers to as “black box” microbial therapies; for instance, fecal microbiota transplantation (FMT) for the treatment of refractory C. difficile infection. However, FMT donor samples are really quite mysterious, as most microbes in the human gut have yet to be discovered. The vast majority of microbial genes have been identified by isolating and culturing microbes and then sequencing their DNA. However, this approach is very time-consuming and low throughput.

The Bhatt lab has been particularly interested in how to generate microbial genomes from metagenomes. Metagenomics involves sequencing all of the DNA in a given sample. The resulting sequencing reads are stitched together to form longer reads, with the goal being to stitch them together into complete microbial genomes. In reality, we don’t actually get a full genome from these sorts of approaches. However, using newer bioinformatic approaches and technologies like read clouds and nanopore sequencing, Dr. Bhatt’s lab and others have been able to generate full genomes. This approach has even predicted novel, unculturable microbes that had never before been discovered or were previously unknown to inhabit the gut. Lastly, Dr. Bhatt shared an approach to predict and validate small proteins that may play a crucial role at the host-microbe interface. With these new tools, we may finally start to more completely understand what microbes are there, and what they are doing.

Beyond culture: time to focus on the small intestinal microbiome

Dr. Purna Kashyap from Mayo Clinic presented on the importance of going beyond culture to more accurately profile the small intestinal microbiome. Small intestinal bacterial overgrowth, or SIBO, was first described in 1897 by Faber in a patient with underlying intestinal strictures. Today, wide-ranging symptoms like abdominal pain and bloating are attributed to SIBO even in the absence of the classic symptoms of steatorrhea (fatty stools), malnutrition, and weight loss. Diagnosis is still often made using quantitative culture of small intestinal aspirates, despite the widespread availability of DNA sequencing. A recent clinical trial led by Dr. Kashyap’s laboratory that used next-generation sequencing along with traditional culture-based methodologies found that small intestinal dysbiosis, rather than bacterial overgrowth, underlies most functional GI symptoms.

I’ve written about this study before, so I won’t rehash all of the findings again here, but I was particularly fascinated by a few pieces of data that Dr. Kashyap highlighted in his talk. First were the specific taxa that were differentially abundant in the small intestine between symptomatic and asymptomatic patients. Aligned with previous findings, several members of the phylum Proteobacteria were more abundant in those with GI symptoms. However, Lactobacillus and Bifidobacterium, two genera commonly found in probiotics, were also overabundant in the small intestine of symptomatic patients. The second intriguing piece of data that I hadn’t seen before was in the latter part of the study, where they had asked a group of healthy individuals that habitually consume a high-fiber diet to adopt a low-fiber, high-sugar diet for 7 days. Not only did these individuals have a significant decrease in microbial diversity in the small intestine and an increase in intestinal permeability, but biopsies revealed a decreased response to the gut neurotransmitter serotonin on a low-fiber, high-sugar diet. Overall, this study suggests that we need to focus much more on microbial composition in individuals presenting with GI symptoms.

Host-pathogen interactions in the age of microbiome research

Dr. Andreas Baumler from the University of California Davis gave a talk describing several insights learned from studying enteric pathogens. I’ve written several articles inspired by Dr. Baumler’s research. Dr. Baumler began by describing what he sees as two distinct arms of the immune system: the classical sterilizing immunity, which detects and clears microbes in tissues, and microbiota-nourishing immunity, which does not remove microbes from the body but instead helps to balance microbial communities and maintain colonization resistance. Under conditions of homeostasis, the colon is predominantly obligate anaerobes (Bacteroidetes and Firmicutes). However, as soon as oxygen is present, facultative anaerobes will expand in the gut. This is a common signature seen in states of gut dysbiosis.

Using a mouse model of infection with Citrobacter rodentium, a pathogen similar to Enterohemorraghic E. coli in humans, Dr. Baumler and colleagues found that infection triggers hyperplasia of colonic crypts, which alters colonocyte metabolism and triggers an influx of oxygen into the gut lumen. This oxygen promotes the growth of Citrobacter and other facultative anaerobes. Early in infection, however, crypt hyperplasia and mucosal oxygenation are not yet present, leading the researchers to wonder what might be promoting early attachment of Citrobacter to the epithelium. By analyzing microbial gene expression, they found that many “oxygen-tolerant” species, including Citrobacter, have elevated expression of catalase, an enzyme that detoxifies hydrogen peroxide. Sure enough, epithelial NOX1 is constitutively expressed in the colon and is a continual source of hydrogen peroxide. Dr. Baumler proposed that we might therefore think of “oxygen-tolerant species” as more accurately “hydrogen-peroxide-tolerant species”.  In mouse models of NOX1 deficiency, bacteria have been found to reach all the way down into the crypts, suggesting that this hydrogen peroxide gradient is an important control mechanism of the host to maintain homeostasis.

Personalized medicine based on the microbiome and clinical data

Dr. Eran Segal from the Weizmann Institute of Science in Tel Aviv, Israel kicked off the third day of the summit with a fascinating talk on integrating microbiome and clinical data into personalized medicine. He first shared their published personalized nutrition project, in which they found that the microbiome could predict meal glucose responses with twice the accuracy of carb counting. Pre-diabetics put on a personalized diet determined by the algorithm were able to successfully lower their post-meal glucose responses. In a longer randomized controlled trial, the algorithm diet resulted in significantly better clinical outcomes than a standard of care diet at 6 months and maintained improvements up to one year.

In another project, Dr. Segal and colleagues wanted to examine what determines the level of metabolites in our blood. They found that over 92 percent of serum metabolites could be significantly predicted by known features, with diet having the largest predictive power and the microbiome coming in second. This could be particularly important clinically for the development of microbiome-based therapeutics. By targeted supplementation or depletion of a particular microbe, we could theoretically modulate levels of a health-promoting or disease-causing metabolite in the blood.

Dr. Segal next discussed bacterial SNPs as the next frontier in microbiome analysis. Using high-resolution sequencing techniques, we can now look at specific single nucleotide polymorphisms (SNPs) across bacteria, much like we can in human genes – in other words, performing metagenome-wide association studies (MWAS) in bacterial genomes – to study strain transmission between individuals or gut microbiota evolution over time within a single individual. Using a large microbiome dataset, the Segal lab found that host BMI is associated with several bacterial SNPs, including a number in polysaccharide degrading genes.

Dr. Segal finished by sharing a final pilot study of fecal microbiota transplants (FMT) in humans with atopic dermatitis (AD). They enrolled nine individuals with AD, each given placebo capsules for the first four weeks, and then the FMT capsules in four doses, separated by two weeks. They found that after FMT, many bacterial strains were engrafted, indicating real transmission from donor to recipient. Moreover, all nine patients responded quite favorably to the FMT, with an average 70 percent reduction in AD symptoms! This improvement persisted for many weeks. They’ll next be looking to repeat this in a larger cohort.

Gut epithelial metabolism links a high-fat Western diet, dysbiosis, and cardiovascular disease

Dr. Mariana Byndloss next presented data from her new lab at Vanderbilt University, where she studies the mechanisms by which gut inflammation alters host physiology to disrupt the gut microbiota, leading to disease. As a postdoc at UC Davis, Dr. Byndloss was involved in seminal research that identified gut epithelial cell metabolism as crucial to maintaining gut hypoxia and shaping the microbiota of the large intestine towards a predominance of obligate anaerobes. Gut dysbiosis is characterized by a loss of the anaerobic condition and a shift from obligate anaerobes to facultative anaerobes like Enterobacteriaceae. This shift can be induced using a single dose of the antibiotic streptomycin, which depletes butyrate-producers and forces the gut epithelium to rely on glucose for energy, rather than butyrate. I’ve written about this research before.

Most recently, the Byndloss lab found that a high-fat, Western-style diet also resulted in a loss of gut hypoxia and an expansion of this Enterobacteriaceae in mice. Rather than depleting gut butyrate levels, as the antibiotics had, the high-fat Western diet seemed to directly affect the gut epithelium. At least in vitro, it appeared that the fat content of the Western diet was directly damaging mitochondria, reducing beta-oxidation and increasing oxygen and nitrate availability in the gut. The expansion of Enterobacteriaceae may have implications for cardiovascular disease. Several species in this family are capable of converting choline to a compound called TMA, which is then oxidized by the liver to TMAO. Elevated serum levels of TMAO have been previously linked to atherosclerosis. Most notably, however, Dr. Byndloss and colleagues found that the ability of these microbes to grow on choline and produce TMA was dependent on nitrate! In other words, the presence of choline in the gut may only be an issue in the context of gut inflammation, where nitrate availability is increased. Future studies in the Byndloss lab may look at modulating the type of dietary fat and amount of fiber in this model.

Interactions between diet and the gut microbiome: plant lignans and ketogenic diets

Dr. Peter Turnbaugh from the University of California San Francisco gave a fascinating keynote talk on the bidirectional interactions linking diet and the gut microbiota. Diet is known to be a key factor that influences the microbiome and in turn, the microbiome can alter many dietary components in many interesting ways. For the first half of his talk, Dr. Turnbaugh shared their recent foray into understanding the metabolism of lignans, polyphenols that are found in many different plant foods. Notably, lignans themselves are biologically inert, requiring a series of transformations by human gut bacteria. The resulting phyto-estrogenic enterolignans have been shown to have anti-breast cancer effects in humans and animal models. Through a series of genomic experiments, the Turnbaugh lab identified a consortium of four bacterial strains that cooperate to convert the lignan pinoresinol to the active compound enterolactone. Further analyses of the gene content of human stool samples found that the majority of people do have this entire pathway, allowing for lignan conversion. Future work hopes to utilize gnotobiotic models to dissect the role of the bacterial interactions in relation to cancer and other diseases, and identify genetic markers that could potentially help to predict or promote this pathway in patients. In the meantime, these findings highlight the importance of crosstalk between the host and bacterial metabolites.

In the second half of his talk, Dr. Turnbaugh highlighted their recent work examining the effects of a ketogenic diet on the gut microbiome. They began by recruiting 17 obese human volunteers for a controlled dietary intervention in a metabolic ward. The subjects were given a baseline diet for four weeks, followed by four weeks on a ketogenic diet. Stool microbiome analysis revealed consistent shifts in the gut microbial community structure in response to the ketogenic diet and a significant depletion of Bifidobacterium species. To model this diet shift more closely, the lab next turned to mice. Of four dietary treatment arms – normal chow, low-fat, the classic high-fat diet, and a ketogenic diet – only a ketogenic diet increased ketone levels and, similar to their findings in humans, significantly depleted Bifidobacterium. Follow-up testing of a finer diet gradient ranging from 75 to 90 percent fat found that the relative abundance of Bifidobacterium in the gut decreased with increasing fat intake. Notably, they found that ketone bodies, particularly beta-hydroxybutyrate, are elevated in the lumen of the gut with consumption of a ketogenic diet or ketone ester and can directly inhibit the growth of Bifidobacterium. This may have implications for disease, as specific species in this genus are able to induce Th17 cells in the small intestine. Most interestingly, the ketogenic diet maintained the mucosal barrier, despite the lack of fermentable carbohydrates. Future studies in the Turnbaugh lab may look to identify whether liver ketogenesis contributes to intestinal ketone levels and examine the effects of a ketogenic diet or ketone esters on the gut in disease models.

For more on ketogenic diets, be sure to see my recent article that features the Turnbaugh paper and others.

Early-life microbial encounters and childhood asthma

Dr. Susan Lynch from UCSF gave a fascinating talk to kick end day three of the summit on the role of the early-life microbiome in the development of asthma and atopy. Asthma is the most prevalent disease among children in Westernized nations, characterized by a reduction in regulatory T cells, an increase in pro-inflammatory Th2 cells, and elevated circulating IgE antibodies. While there are risk genes known to be associated with asthma development, they don’t explain the full prevalence of the disease, and it is now accepted that the disease is strongly influenced by the environment and lifestyle. Drawing insight from the field of ecology, Dr. Lynch noted that it is well established that early colonizers of an ecosystem can dictate the rate and trajectory of species accumulation in various niches. She and her lab wondered whether this might also be true in the developing microbial ecosystem in the gut. They examined several large birth cohorts for patterns that might correlate with the risk of asthma and atopy. Indeed, they found that high-risk babies had a distinct gut microbiota at birth and acquired microbial diversity much more slowly than low-risk babies. At one month, high-risk infants had a lower abundance of various bacterial taxa and enrichment of the fungi Candida and Rhodotorula. High-risk infants in both cohorts also had altered metabolite profiles, with greater production of the linoleic acid metabolite 12,13-DiHOME. This metabolite was found to suppress regulatory T cells and their anti-inflammatory activity. Introduction of this metabolite or the bacterial genes that produce it, to the gut of mice was found to exacerbate airway inflammation, reduce regulatory T cell numbers, and increase circulating IgE.

Most recently, Dr. Lynch’s lab has taken a step further back to ask when we first encounter microbes. By examining prenatal sections of the small intestine, they found a very sparse but clear microbial signal in the fetal gut. Upon closer examination, a subset of the signals seemed to be dominated by a microbe called Micrococcus. Isolating this particular organism revealed that it exhibits a number of features that might allow its survival in utero, including growth on pregnancy hormones! This study suggests evidence for microbial encounters before birth, and as early as mid-gestation, that may potentially shape the maturation of the immune system.

Towards an ecological understanding of the role of the microbiome in human metabolic health: thinking in terms of microbial “guilds”

Dr. Liping Zhao from Rutgers University shared a new way of thinking about and analyzing the gut microbiome – in terms of functional groups, or “guilds”. While we typically look at individual microbial abundances, members of the gut microbiome work together as part of coherent functional groups. Members in the same guild exploit similar environmental resources and tend to become more or less abundant in response to the same environmental stimuli, but do not necessarily share genetic or taxonomic similarity. In other words, members of the same guild may be from completely different phyla in the gut! Inspired by his own weight loss journey, Dr. Zhao and his laboratory groups at Rutgers and in Shanghai have been examining the effects of a high-fiber dietary intervention in humans with obesity and diabetes. The diet includes a number of non-digestible polysaccharides from whole grains, traditional Chinese medicinal foods, and other prebiotics.

In both children with Pardi-Willi Syndrome, a genetic form of obesity, and children with traditional obesity, the high-fiber intervention promoted weight loss, reduced inflammation, and reduced blood endotoxin. Guild-based analysis revealed that certain bacterial guilds correlated positively with disease, while others were negatively correlated with the disease phenotype. In another recent project on adult Type 2 diabetes, they looked at the effects of the high-fiber diet intervention, but also gave all of the participants acarbose, a drug that inhibits starch digestion by humans, thereby making more starch available for bacterial fermentation. The control group received the usual care plus acarbose. After three months, participants in the high-fiber diet group had greater improvements in HbA1c and other markers of glucose and insulin homeostasis. Transplanting the baseline or 3-month gut microbiota into germ-free mice demonstrated that the changes in the microbiome contributed to the effects on glycemic control. Microbes that responded positively to the intervention a higher genetic capacity for utilizing starch, inulin and arabinoxylan to produce SCFAs. Dr. Zhao closed by sharing that they have data to suggest that each person will have a different composition of such a “foundation guild” and will require different dietary fibers to support their growth. The lab hopes to be able to identify such avenues to target these foundation guilds for personalized nutrition and disease treatment.

Is diet the primary driver of microbiome variation?

Dr. Dan Knights from the University of Minnesota gave a fascinating talk on the degree to which diet might drive microbiome variation. While mouse studies lend support to the idea that diet is the primary driver of microbiome variation, this has been harder to pin down in human studies, largely because of the high variability of microbiota composition between subjects and within a single individual over time. Using ultra-dense longitudinal metagenomics data, they looked at the daily variation in the fecal microbiome and daily variation in dietary data in 34 subjects collected for two weeks. While most studies analyze dietary data using between 50-100 macro- and micronutrients, the Knights lab came to realize that these conventional nutrients ignore the “dark matter” of foods, including many compounds that our microbes utilize or respond to. By using a FoodTree that organizes foods with similar “dark matter”, the group was able to get a strong concordance between individual microbiomes and dietary intake. Overall, they found that diet explains about 20 percent of the microbiome variation between people. While there was no connection between dietary diversity and microbiome diversity, they did find that higher dietary diversity correlated with greater microbiome stability over time. Interestingly, their cohort included two computer scientists that were drinking nothing but Soylent! Despite the identical diet of these two individuals, their microbiomes did not converge and were not more stable on average than the other subjects.

Dr. Knight and colleagues next wanted to understand variation within a single person. Using the last few days of dietary history, they found that diet explains about 25 percent of a single person’s microbiome variation over time. Interestingly, some of the associations between particular foods and microbes were generalized across virtually all of the participants, while others were more personalized, meaning individuals had very different microbial responses to the same food. Dr. Knights suggested that this could be due to strain-level differences in the microbiome, slight differences in the food content that wasn’t captured by the FoodTree, and ecosystem-level differences, where small changes may reverberate through the entire gut in ways that are difficult for us to predict.

In a separate project, the lab looked at the diet and microbiome data of first- and second-generation U.S. immigrants from Thailand. They found that second-generation immigrants consumed a diet that was halfway between their traditional diet and a U.S. diet, yet their microbiome was already fully Westernized. In all, diet only explained about 16 percent of microbiome Westernization, suggesting that immigration-related changes in the microbiome are likely driven by a combination of factors associated with life in the U.S.

SCFA regulation of intestinal barrier function in health and disease

Dr. Sean Colgan from the University of Colorado Denver finished up day three with a talk on short-chain fatty acids. The intestine has a massively large surface area and must effectively maintain an effective barrier across this vast surface area to prevent the influx of microbes or their components into the bloodstream. Dr. Colgan’s laboratory has been focused on the role that short-chain fatty acids, and especially butyrate, play in maintaining this barrier function and gut homeostasis. The lab had previously found that the SCFA butyrate supports gut barrier function, maintains “physiologic hypoxia” in the gut, and also promotes epithelial wound healing.

In a recent set of experiments, the Colgan lab focused on a protein called synaptopodin (SYNPO) that is upregulated by butyrate. They found that SYNPO localizes to tight junctions and is a key component of a functional intestinal barrier. Mice that lack the SYNPO gene have almost no gut barrier and developed much more severe colitis when presented with a chemical challenge. In mice given antibiotics, SYNPO expression was lost, but administering butyrate restored SYNPO expression and the resulting barrier function. Thus, butyrate is a key signal for maintaining this important component of the gut barrier. Dr. Colgan believes that there will be many future opportunities to identify such microbiota-derived molecules that signal to the host in health and disease and to elucidate disease-specific changes inside and outside of the mucosa.

Domestication and the gut microbiota: implications for the human microbiome?

Dr. Aspen Reese from UCSD closed out the summit with a fascinating talk that delved into the impact of domestication on the gut microbiota from an evolutionary and ecological perspective. The microbiota has the potential to alter both genes and the environment, and therefore likely plays a significant role in host evolution. With its vast functional repertoire, rapid turnover, and greater capacity for horizontal gene transfer, the microbiome can collectively respond to environmental perturbations much faster than the host can. In the last several years, Dr. Reese and her colleagues have been using domestication as a model to explore this. The domestication of animals induced rapid genetic evolution as we selected for particular traits of interest, but also caused major ecological shifts, including a change in diet and environmental exposures. Sampling nine pairs of wild and domestic species of mammals, Dr. Reese found that about 15-16 percent of the variation of the microbial communities could be ascribed to domestication. But is this due to changes in host genes or the host environment?

To answer this question, Dr. Reese and colleagues caught wild mice and brought them into the lab environment. They then fed groups of lab mice or wild mice either a normal lab chow diet or a wild mouse diet for 30 days. Mice eating reciprocal diets saw consistent changes in the microbiota, suggesting that diet can overcome some of the genetic differences between wild-type animals and their domesticated counterparts. Repeating the experiment in a group of dogs and captive wolves eating reciprocal diets yielded similar results. To determine whether the domestic microbiota is a better match for the host, they again turned to lab mice and tried to colonize them with wild microbes. They found that lab mice can harbor a wild microbiota and that, in fact, a wild microbiota wins out over the lab microbial community without any antibiotic pretreatment, suggesting that these are quite plastic communities. Dr. Reese hypothesizes that local extinction events have probably driven a lot of these microbes out of the lab, but that they can readily and easily recolonize.

These experiments have a number of implications, not only for animal ecology, but also as a model for humans. In a lot of ways, we humans in industrialized countries are more like lab mice than wild mice – our environment and diet are relatively sterile, and we have significantly reduced pathogen exposure. Indeed, many of the differences between industrialized and non-industrialized human populations parallel the changes seen between domesticated and wild mammals! Dr. Reese closed by noting that domestication or industrialization may have induced functional tradeoffs – when we select for one trait, such as growth – it may come at the cost of another – like immunological health. I’ll be discussing this more in a future article on the industrialized microbiota.

Thanks to everyone who attended, presented, or otherwise helped make the Virtual Microbiome Summit happen! I hope you enjoyed this summary as much as I enjoyed hosting the event and watching the replay. If you did, be sure to share your favorite highlights in the comments below, subscribe to my blog newsletter, or become a patron for monthly gut research updates!

virtual microbiome summit research higlights

Highlights from the Virtual Microbiome Summit 2020

This post is long overdue, but I only recently had a chance to finish watching all of the recordings!

It was truly a blast to host this summit. I am so grateful to have been able to connect with so many incredible researchers and bring together such an amazing event.

In total, we had 25+ speakers, nearly 3,000 participants registered, 700 people on the live webinar at once, and 45 different countries represented.

This article will be a summary of the keynote and plenary talks from the summit. Fair warning that it is quite long, so feel free to skim for the topics that interest you.

If you enjoy this article, be sure to check out the full summit recordings, as there were a number of excellent students, postdoctoral fellows, and assistant scientists who spoke on everything from microbial drug metabolism to the infant microbiome to the effects of stress on gut barrier function!

The butyrate-producer Anaerostipes caccae is protective against food allergies

Dr. Cathryn Nagler from the University of Chicago kicked off the summit with a keynote talk on the role of gut bacteria in food allergies. Over 32 million Americans – about 1 in 10 adults and 1 in 13 children – now have at least one food allergy. Many people have multiple food allergies, along with other related diseases, including eczema, asthma, and allergic rhinitis. Several modern industrialized lifestyle factors are proposed to have contributed to the dramatic increase in food allergies, including increased use of antibiotics, a processed diet, increased urban living, birth by Cesarean section, formula feeding, and a lack of exposure to helminths and other infectious microbes. In a sensitized individual, a food allergic reaction occurs when an allergenic protein crosses the intestinal barrier and gets into the bloodstream, where mast cells and other components of the immune system launch an inflammatory response.

Dr. Nagler’s laboratory has identified a specific butyrate-producing bacterium, Anaerostipes caccae, that induces gut barrier protection by increasing the expression of the cytokine IL-22. Infants with cow’s milk allergy tend to have a much lower abundance of A. caccae in the ileum, the final third of the small intestine. Incredibly, colonizing germ-free mice with just this single species protected them against allergic responses. Dr. Nagler’s team is now working on a startup, ClostraBio, in the hopes of translating these findings into a treatment for human food allergies. They eventually hope to produce a live biotherapeutic of A. caccae, but this won’t be simple given the oxygen sensitivity of this obligate anaerobe. In the meantime, they are exploring ways to provide butyrate directly to the ileum using a polymer delivery platform. Initial animal studies in peanut-sensitized mice have shown promise for using this ileum-targeted butyrate in combination with oral immunotherapy. Dr. Nagler hopes to move on to clinical trials in young adults within the next year. Once developed, their biotherapeutic or ileum-targeted butyrate may have benefits for other conditions as well, including inflammatory bowel disease.

Low dietary fiber and specific gut microbes promote colonic inflammation

Dr. Eric Martens from the University of Michigan presented their work on the interplay between dietary fiber and the host mucus layer. Prior studies in the Martens lab had demonstrated that, in the absence of dietary fiber, mucin-degrading Bacteroides and Akkermansia species will expand in the gut and degrade the gut mucus layer. This causes low-grade inflammation and increases susceptibility to gut infection. (Importantly, this fiber-free diet is also quite high in simple sugars.) Most recently, the lab went on to test a fiber-free diet in the IL-10 genetic knockout model of inflammatory bowel disease (IBD). Fiber-deprived mice on this genetic background had a very thin mucus layer and a cytokine profile typical of IBD. Notably, adding fiber back into the diet blocked the progression of disease, suggesting that at least in this mouse model, fiber can be used to treat disease that has already been initiated.

Dr. Martens also presented some interesting data on exclusive enteric nutrition (EEN), a liquid formula diet that has been shown to be effective as a treatment for human Crohn’s disease. Given the absence of fiber in this diet, it is curious that is so effective for the treatment of Crohn’s disease. Interestingly, when genetically predisposed mice were put on EEN, many of the mice had elevated inflammatory markers. Most surprisingly, however, EEN increased the abundance of Eubacterium rectale, a butyrate producer typically thought to rely on dietary fiber. Future studies will seek to better understand the role of fiber in restoring balance in the gut and the role of mucus-degrading microbes in gut homeostasis.

Hyperglycemia is associated with gut barrier dysfunction and increased susceptibility to gut infection.

Dr. Cristoph Thaiss from the University of Pennsylvania presented on microbiome dynamics in metabolic disease. Several studies had previously found that a higher body mass index was associated with a higher risk of infection, but the underlying mechanism was unknown. Through a series of experiments, Dr. Thaiss and colleagues determined that hyperglycemia, rather than obesity itself, results in epigenetic reprogramming of gut epithelial cells, leading to gut barrier dysfunction, increased susceptibility to the gut pathogen Citrobacter rodentium, and an increase in microbial products entering systemic circulation. In a human cohort, they found that HbA1c, a marker of long-term blood glucose regulation, was strongly correlated with microbial molecules in circulation.

Dr. Thaiss and colleagues decided to look more closely at the various glucose transporters that are present in intestinal epithelial cells. On the apical side, facing the lumen, intestinal epithelial cells have a sodium-glucose symporter (SGLT1), which absorbs glucose from the intestinal lumen. On the basolateral side, closest to the bloodstream, GLUT2 facilitates glucose transport along its concentration gradient. Normally, this serves to move glucose from the epithelial cell into the bloodstream, where it will go to other tissues. However, in the case of hyperglycemia, the researchers hypothesized that glucose levels in the interstitial fluid of the gut are higher than inside intestinal epithelial cells, causing GLUT2 to facilitate a backward flow of glucose into the intestinal epithelium. Indeed, selectively knocking out GLUT2 or blocking glucose metabolism in intestinal epithelial cells led to at least partial restoration of intestinal barrier function and prevented the spread of microbial products into systemic circulation. This study suggests that addressing chronic hyperglycemia is essential to restoring gut homeostasis.

Diet as a lever to improve the industrialized microbiota: the case of fiber and fermented foods

Dr. Erica Sonnenburg from Stanford University kicked off the afternoon session with a talk on the industrialized microbiota. Over the last decade, the Sonnenburg lab has collected several lines of evidence that support a major change in the gut microbiome in recent history. Study of Hadza hunter-gatherers, Nepalese populations across a lifestyle gradient, and the irreversible loss of microbial diversity seen over generations of laboratory mice fed a low fiber diet all support the notion that the industrialized microbiota may be predisposing us to disease. Dr. Sonnenburg reasoned that the shift to a modern diet and lifestyle may have created a mismatch between our human genome and our microbiome. But could improving our diet repair the microbiota and improve health? Unlike many other means of manipulating the microbiota that require extensive testing in animal models, dietary changes are quite safe and can be studied directly in humans.

The researchers enrolled 18 healthy volunteers for a high-fiber intervention and 18 for a high fermented food intervention. Subjects in each group were instructed to slowly ramp up fiber or fermented foods over a 4-week period, maintain consumption for an additional 6 weeks, and then have a 4-week choice period. They found that fermented foods, but not fiber, resulted in a cohort-wide reduction in inflammatory markers and increased microbiota diversity. This was not simply attributed to species found in the fermented foods, suggesting that something about fermented foods makes the gut permissive to new species. The high fiber intervention, on the other hand, resulted in individualized immune responses, with some individuals having a reduction in inflammatory markers, others having no response, and still others having an increase in inflammation from the increased fiber intake. Most interestingly, those with the highest microbial diversity at baseline tended to have a more beneficial immune response to fiber. Dr. Sonnenburg emphasized that there are still many open questions, but that these diet intervention studies are important to inform public health and policy and create a food environment that is based on science and can serve to improve global health.

Unveiling the “black box”: from precise microbial genomics to precision medicine

Dr. Ami Bhatt from Stanford gave an excellent talk on how we might use more precise microbial genomics to influence precision medicine. This was definitely one of the more technical of the talks presented at the summit but has major implications for the future of microbiome discovery. Currently, we can cure or treat disease using what Dr. Bhatt refers to as “black box” microbial therapies; for instance, fecal microbiota transplantation (FMT) for the treatment of refractory C. difficile infection. However, FMT donor samples are really quite mysterious, as most microbes in the human gut have yet to be discovered. The vast majority of microbial genes have been identified by isolating and culturing microbes and then sequencing their DNA. However, this approach is very time-consuming and low throughput.

The Bhatt lab has been particularly interested in how to generate microbial genomes from metagenomes. Metagenomics involves sequencing all of the DNA in a given sample. The resulting sequencing reads are stitched together to form longer reads, with the goal being to stitch them together into complete microbial genomes. In reality, we don’t actually get a full genome from these sorts of approaches. However, using newer bioinformatic approaches and technologies like read clouds and nanopore sequencing, Dr. Bhatt’s lab and others have been able to generate full genomes. This approach has even predicted novel, unculturable microbes that had never before been discovered or were previously unknown to inhabit the gut. Lastly, Dr. Bhatt shared an approach to predict and validate small proteins that may play a crucial role at the host-microbe interface. With these new tools, we may finally start to more completely understand what microbes are there, and what they are doing.

Beyond culture: time to focus on the small intestinal microbiome

Dr. Purna Kashyap from Mayo Clinic presented on the importance of going beyond culture to more accurately profile the small intestinal microbiome. Small intestinal bacterial overgrowth, or SIBO, was first described in 1897 by Faber in a patient with underlying intestinal strictures. Today, wide-ranging symptoms like abdominal pain and bloating are attributed to SIBO even in the absence of the classic symptoms of steatorrhea (fatty stools), malnutrition, and weight loss. Diagnosis is still often made using quantitative culture of small intestinal aspirates, despite the widespread availability of DNA sequencing. A recent clinical trial led by Dr. Kashyap’s laboratory that used next-generation sequencing along with traditional culture-based methodologies found that small intestinal dysbiosis, rather than bacterial overgrowth, underlies most functional GI symptoms.

I’ve written about this study before, so I won’t rehash all of the findings again here, but I was particularly fascinated by a few pieces of data that Dr. Kashyap highlighted in his talk. First were the specific taxa that were differentially abundant in the small intestine between symptomatic and asymptomatic patients. Aligned with previous findings, several members of the phylum Proteobacteria were more abundant in those with GI symptoms. However, Lactobacillus and Bifidobacterium, two genera commonly found in probiotics, were also overabundant in the small intestine of symptomatic patients. The second intriguing piece of data that I hadn’t seen before was in the latter part of the study, where they had asked a group of healthy individuals that habitually consume a high-fiber diet to adopt a low-fiber, high-sugar diet for 7 days. Not only did these individuals have a significant decrease in microbial diversity in the small intestine and an increase in intestinal permeability, but biopsies revealed a decreased response to the gut neurotransmitter serotonin on a low-fiber, high-sugar diet. Overall, this study suggests that we need to focus much more on microbial composition in individuals presenting with GI symptoms.

Host-pathogen interactions in the age of microbiome research

Dr. Andreas Baumler from the University of California Davis gave a talk describing several insights learned from studying enteric pathogens. I’ve written several articles inspired by Dr. Baumler’s research. Dr. Baumler began by describing what he sees as two distinct arms of the immune system: the classical sterilizing immunity, which detects and clears microbes in tissues, and microbiota-nourishing immunity, which does not remove microbes from the body but instead helps to balance microbial communities and maintain colonization resistance. Under conditions of homeostasis, the colon is predominantly obligate anaerobes (Bacteroidetes and Firmicutes). However, as soon as oxygen is present, facultative anaerobes will expand in the gut. This is a common signature seen in states of gut dysbiosis.

Using a mouse model of infection with Citrobacter rodentium, a pathogen similar to Enterohemorraghic E. coli in humans, Dr. Baumler and colleagues found that infection triggers hyperplasia of colonic crypts, which alters colonocyte metabolism and triggers an influx of oxygen into the gut lumen. This oxygen promotes the growth of Citrobacter and other facultative anaerobes. Early in infection, however, crypt hyperplasia and mucosal oxygenation are not yet present, leading the researchers to wonder what might be promoting early attachment of Citrobacter to the epithelium. By analyzing microbial gene expression, they found that many “oxygen-tolerant” species, including Citrobacter, have elevated expression of catalase, an enzyme that detoxifies hydrogen peroxide. Sure enough, epithelial NOX1 is constitutively expressed in the colon and is a continual source of hydrogen peroxide. Dr. Baumler proposed that we might therefore think of “oxygen-tolerant species” as more accurately “hydrogen-peroxide-tolerant species”.  In mouse models of NOX1 deficiency, bacteria have been found to reach all the way down into the crypts, suggesting that this hydrogen peroxide gradient is an important control mechanism of the host to maintain homeostasis.

Personalized medicine based on the microbiome and clinical data

Dr. Eran Segal from the Weizmann Institute of Science in Tel Aviv, Israel kicked off the third day of the summit with a fascinating talk on integrating microbiome and clinical data into personalized medicine. He first shared their published personalized nutrition project, in which they found that the microbiome could predict meal glucose responses with twice the accuracy of carb counting. Pre-diabetics put on a personalized diet determined by the algorithm were able to successfully lower their post-meal glucose responses. In a longer randomized controlled trial, the algorithm diet resulted in significantly better clinical outcomes than a standard of care diet at 6 months and maintained improvements up to one year.

In another project, Dr. Segal and colleagues wanted to examine what determines the level of metabolites in our blood. They found that over 92 percent of serum metabolites could be significantly predicted by known features, with diet having the largest predictive power and the microbiome coming in second. This could be particularly important clinically for the development of microbiome-based therapeutics. By targeted supplementation or depletion of a particular microbe, we could theoretically modulate levels of a health-promoting or disease-causing metabolite in the blood.

Dr. Segal next discussed bacterial SNPs as the next frontier in microbiome analysis. Using high-resolution sequencing techniques, we can now look at specific single nucleotide polymorphisms (SNPs) across bacteria, much like we can in human genes – in other words, performing metagenome-wide association studies (MWAS) in bacterial genomes – to study strain transmission between individuals or gut microbiota evolution over time within a single individual. Using a large microbiome dataset, the Segal lab found that host BMI is associated with several bacterial SNPs, including a number in polysaccharide degrading genes.

Dr. Segal finished by sharing a final pilot study of fecal microbiota transplants (FMT) in humans with atopic dermatitis (AD). They enrolled nine individuals with AD, each given placebo capsules for the first four weeks, and then the FMT capsules in four doses, separated by two weeks. They found that after FMT, many bacterial strains were engrafted, indicating real transmission from donor to recipient. Moreover, all nine patients responded quite favorably to the FMT, with an average 70 percent reduction in AD symptoms! This improvement persisted for many weeks. They’ll next be looking to repeat this in a larger cohort.

Gut epithelial metabolism links a high-fat Western diet, dysbiosis, and cardiovascular disease

Dr. Mariana Byndloss next presented data from her new lab at Vanderbilt University, where she studies the mechanisms by which gut inflammation alters host physiology to disrupt the gut microbiota, leading to disease. As a postdoc at UC Davis, Dr. Byndloss was involved in seminal research that identified gut epithelial cell metabolism as crucial to maintaining gut hypoxia and shaping the microbiota of the large intestine towards a predominance of obligate anaerobes. Gut dysbiosis is characterized by a loss of the anaerobic condition and a shift from obligate anaerobes to facultative anaerobes like Enterobacteriaceae. This shift can be induced using a single dose of the antibiotic streptomycin, which depletes butyrate-producers and forces the gut epithelium to rely on glucose for energy, rather than butyrate. I’ve written about this research before.

Most recently, the Byndloss lab found that a high-fat, Western-style diet also resulted in a loss of gut hypoxia and an expansion of this Enterobacteriaceae in mice. Rather than depleting gut butyrate levels, as the antibiotics had, the high-fat Western diet seemed to directly affect the gut epithelium. At least in vitro, it appeared that the fat content of the Western diet was directly damaging mitochondria, reducing beta-oxidation and increasing oxygen and nitrate availability in the gut. The expansion of Enterobacteriaceae may have implications for cardiovascular disease. Several species in this family are capable of converting choline to a compound called TMA, which is then oxidized by the liver to TMAO. Elevated serum levels of TMAO have been previously linked to atherosclerosis. Most notably, however, Dr. Byndloss and colleagues found that the ability of these microbes to grow on choline and produce TMA was dependent on nitrate! In other words, the presence of choline in the gut may only be an issue in the context of gut inflammation, where nitrate availability is increased. Future studies in the Byndloss lab may look at modulating the type of dietary fat and amount of fiber in this model.

Interactions between diet and the gut microbiome: plant lignans and ketogenic diets

Dr. Peter Turnbaugh from the University of California San Francisco gave a fascinating keynote talk on the bidirectional interactions linking diet and the gut microbiota. Diet is known to be a key factor that influences the microbiome and in turn, the microbiome can alter many dietary components in many interesting ways. For the first half of his talk, Dr. Turnbaugh shared their recent foray into understanding the metabolism of lignans, polyphenols that are found in many different plant foods. Notably, lignans themselves are biologically inert, requiring a series of transformations by human gut bacteria. The resulting phyto-estrogenic enterolignans have been shown to have anti-breast cancer effects in humans and animal models. Through a series of genomic experiments, the Turnbaugh lab identified a consortium of four bacterial strains that cooperate to convert the lignan pinoresinol to the active compound enterolactone. Further analyses of the gene content of human stool samples found that the majority of people do have this entire pathway, allowing for lignan conversion. Future work hopes to utilize gnotobiotic models to dissect the role of the bacterial interactions in relation to cancer and other diseases, and identify genetic markers that could potentially help to predict or promote this pathway in patients. In the meantime, these findings highlight the importance of crosstalk between the host and bacterial metabolites.

In the second half of his talk, Dr. Turnbaugh highlighted their recent work examining the effects of a ketogenic diet on the gut microbiome. They began by recruiting 17 obese human volunteers for a controlled dietary intervention in a metabolic ward. The subjects were given a baseline diet for four weeks, followed by four weeks on a ketogenic diet. Stool microbiome analysis revealed consistent shifts in the gut microbial community structure in response to the ketogenic diet and a significant depletion of Bifidobacterium species. To model this diet shift more closely, the lab next turned to mice. Of four dietary treatment arms – normal chow, low-fat, the classic high-fat diet, and a ketogenic diet – only a ketogenic diet increased ketone levels and, similar to their findings in humans, significantly depleted Bifidobacterium. Follow-up testing of a finer diet gradient ranging from 75 to 90 percent fat found that the relative abundance of Bifidobacterium in the gut decreased with increasing fat intake. Notably, they found that ketone bodies, particularly beta-hydroxybutyrate, are elevated in the lumen of the gut with consumption of a ketogenic diet or ketone ester and can directly inhibit the growth of Bifidobacterium. This may have implications for disease, as specific species in this genus are able to induce Th17 cells in the small intestine. Most interestingly, the ketogenic diet maintained the mucosal barrier, despite the lack of fermentable carbohydrates. Future studies in the Turnbaugh lab may look to identify whether liver ketogenesis contributes to intestinal ketone levels and examine the effects of a ketogenic diet or ketone esters on the gut in disease models.

For more on ketogenic diets, be sure to see my recent article that features the Turnbaugh paper and others.

Early-life microbial encounters and childhood asthma

Dr. Susan Lynch from UCSF gave a fascinating talk to kick end day three of the summit on the role of the early-life microbiome in the development of asthma and atopy. Asthma is the most prevalent disease among children in Westernized nations, characterized by a reduction in regulatory T cells, an increase in pro-inflammatory Th2 cells, and elevated circulating IgE antibodies. While there are risk genes known to be associated with asthma development, they don’t explain the full prevalence of the disease, and it is now accepted that the disease is strongly influenced by the environment and lifestyle. Drawing insight from the field of ecology, Dr. Lynch noted that it is well established that early colonizers of an ecosystem can dictate the rate and trajectory of species accumulation in various niches. She and her lab wondered whether this might also be true in the developing microbial ecosystem in the gut. They examined several large birth cohorts for patterns that might correlate with the risk of asthma and atopy. Indeed, they found that high-risk babies had a distinct gut microbiota at birth and acquired microbial diversity much more slowly than low-risk babies. At one month, high-risk infants had a lower abundance of various bacterial taxa and enrichment of the fungi Candida and Rhodotorula. High-risk infants in both cohorts also had altered metabolite profiles, with greater production of the linoleic acid metabolite 12,13-DiHOME. This metabolite was found to suppress regulatory T cells and their anti-inflammatory activity. Introduction of this metabolite or the bacterial genes that produce it, to the gut of mice was found to exacerbate airway inflammation, reduce regulatory T cell numbers, and increase circulating IgE.

Most recently, Dr. Lynch’s lab has taken a step further back to ask when we first encounter microbes. By examining prenatal sections of the small intestine, they found a very sparse but clear microbial signal in the fetal gut. Upon closer examination, a subset of the signals seemed to be dominated by a microbe called Micrococcus. Isolating this particular organism revealed that it exhibits a number of features that might allow its survival in utero, including growth on pregnancy hormones! This study suggests evidence for microbial encounters before birth, and as early as mid-gestation, that may potentially shape the maturation of the immune system.

Towards an ecological understanding of the role of the microbiome in human metabolic health: thinking in terms of microbial “guilds”

Dr. Liping Zhao from Rutgers University shared a new way of thinking about and analyzing the gut microbiome – in terms of functional groups, or “guilds”. While we typically look at individual microbial abundances, members of the gut microbiome work together as part of coherent functional groups. Members in the same guild exploit similar environmental resources and tend to become more or less abundant in response to the same environmental stimuli, but do not necessarily share genetic or taxonomic similarity. In other words, members of the same guild may be from completely different phyla in the gut! Inspired by his own weight loss journey, Dr. Zhao and his laboratory groups at Rutgers and in Shanghai have been examining the effects of a high-fiber dietary intervention in humans with obesity and diabetes. The diet includes a number of non-digestible polysaccharides from whole grains, traditional Chinese medicinal foods, and other prebiotics.

In both children with Pardi-Willi Syndrome, a genetic form of obesity, and children with traditional obesity, the high-fiber intervention promoted weight loss, reduced inflammation, and reduced blood endotoxin. Guild-based analysis revealed that certain bacterial guilds correlated positively with disease, while others were negatively correlated with the disease phenotype. In another recent project on adult Type 2 diabetes, they looked at the effects of the high-fiber diet intervention, but also gave all of the participants acarbose, a drug that inhibits starch digestion by humans, thereby making more starch available for bacterial fermentation. The control group received the usual care plus acarbose. After three months, participants in the high-fiber diet group had greater improvements in HbA1c and other markers of glucose and insulin homeostasis. Transplanting the baseline or 3-month gut microbiota into germ-free mice demonstrated that the changes in the microbiome contributed to the effects on glycemic control. Microbes that responded positively to the intervention a higher genetic capacity for utilizing starch, inulin and arabinoxylan to produce SCFAs. Dr. Zhao closed by sharing that they have data to suggest that each person will have a different composition of such a “foundation guild” and will require different dietary fibers to support their growth. The lab hopes to be able to identify such avenues to target these foundation guilds for personalized nutrition and disease treatment.

Is diet the primary driver of microbiome variation?

Dr. Dan Knights from the University of Minnesota gave a fascinating talk on the degree to which diet might drive microbiome variation. While mouse studies lend support to the idea that diet is the primary driver of microbiome variation, this has been harder to pin down in human studies, largely because of the high variability of microbiota composition between subjects and within a single individual over time. Using ultra-dense longitudinal metagenomics data, they looked at the daily variation in the fecal microbiome and daily variation in dietary data in 34 subjects collected for two weeks. While most studies analyze dietary data using between 50-100 macro- and micronutrients, the Knights lab came to realize that these conventional nutrients ignore the “dark matter” of foods, including many compounds that our microbes utilize or respond to. By using a FoodTree that organizes foods with similar “dark matter”, the group was able to get a strong concordance between individual microbiomes and dietary intake. Overall, they found that diet explains about 20 percent of the microbiome variation between people. While there was no connection between dietary diversity and microbiome diversity, they did find that higher dietary diversity correlated with greater microbiome stability over time. Interestingly, their cohort included two computer scientists that were drinking nothing but Soylent! Despite the identical diet of these two individuals, their microbiomes did not converge and were not more stable on average than the other subjects.

Dr. Knight and colleagues next wanted to understand variation within a single person. Using the last few days of dietary history, they found that diet explains about 25 percent of a single person’s microbiome variation over time. Interestingly, some of the associations between particular foods and microbes were generalized across virtually all of the participants, while others were more personalized, meaning individuals had very different microbial responses to the same food. Dr. Knights suggested that this could be due to strain-level differences in the microbiome, slight differences in the food content that wasn’t captured by the FoodTree, and ecosystem-level differences, where small changes may reverberate through the entire gut in ways that are difficult for us to predict.

In a separate project, the lab looked at the diet and microbiome data of first- and second-generation U.S. immigrants from Thailand. They found that second-generation immigrants consumed a diet that was halfway between their traditional diet and a U.S. diet, yet their microbiome was already fully Westernized. In all, diet only explained about 16 percent of microbiome Westernization, suggesting that immigration-related changes in the microbiome are likely driven by a combination of factors associated with life in the U.S.

SCFA regulation of intestinal barrier function in health and disease

Dr. Sean Colgan from the University of Colorado Denver finished up day three with a talk on short-chain fatty acids. The intestine has a massively large surface area and must effectively maintain an effective barrier across this vast surface area to prevent the influx of microbes or their components into the bloodstream. Dr. Colgan’s laboratory has been focused on the role that short-chain fatty acids, and especially butyrate, play in maintaining this barrier function and gut homeostasis. The lab had previously found that the SCFA butyrate supports gut barrier function, maintains “physiologic hypoxia” in the gut, and also promotes epithelial wound healing.

In a recent set of experiments, the Colgan lab focused on a protein called synaptopodin (SYNPO) that is upregulated by butyrate. They found that SYNPO localizes to tight junctions and is a key component of a functional intestinal barrier. Mice that lack the SYNPO gene have almost no gut barrier and developed much more severe colitis when presented with a chemical challenge. In mice given antibiotics, SYNPO expression was lost, but administering butyrate restored SYNPO expression and the resulting barrier function. Thus, butyrate is a key signal for maintaining this important component of the gut barrier. Dr. Colgan believes that there will be many future opportunities to identify such microbiota-derived molecules that signal to the host in health and disease and to elucidate disease-specific changes inside and outside of the mucosa.

Domestication and the gut microbiota: implications for the human microbiome?

Dr. Aspen Reese from UCSD closed out the summit with a fascinating talk that delved into the impact of domestication on the gut microbiota from an evolutionary and ecological perspective. The microbiota has the potential to alter both genes and the environment, and therefore likely plays a significant role in host evolution. With its vast functional repertoire, rapid turnover, and greater capacity for horizontal gene transfer, the microbiome can collectively respond to environmental perturbations much faster than the host can. In the last several years, Dr. Reese and her colleagues have been using domestication as a model to explore this. The domestication of animals induced rapid genetic evolution as we selected for particular traits of interest, but also caused major ecological shifts, including a change in diet and environmental exposures. Sampling nine pairs of wild and domestic species of mammals, Dr. Reese found that about 15-16 percent of the variation of the microbial communities could be ascribed to domestication. But is this due to changes in host genes or the host environment?

To answer this question, Dr. Reese and colleagues caught wild mice and brought them into the lab environment. They then fed groups of lab mice or wild mice either a normal lab chow diet or a wild mouse diet for 30 days. Mice eating reciprocal diets saw consistent changes in the microbiota, suggesting that diet can overcome some of the genetic differences between wild-type animals and their domesticated counterparts. Repeating the experiment in a group of dogs and captive wolves eating reciprocal diets yielded similar results. To determine whether the domestic microbiota is a better match for the host, they again turned to lab mice and tried to colonize them with wild microbes. They found that lab mice can harbor a wild microbiota and that, in fact, a wild microbiota wins out over the lab microbial community without any antibiotic pretreatment, suggesting that these are quite plastic communities. Dr. Reese hypothesizes that local extinction events have probably driven a lot of these microbes out of the lab, but that they can readily and easily recolonize.

These experiments have a number of implications, not only for animal ecology, but also as a model for humans. In a lot of ways, we humans in industrialized countries are more like lab mice than wild mice – our environment and diet are relatively sterile, and we have significantly reduced pathogen exposure. Indeed, many of the differences between industrialized and non-industrialized human populations parallel the changes seen between domesticated and wild mammals! Dr. Reese closed by noting that domestication or industrialization may have induced functional tradeoffs – when we select for one trait, such as growth – it may come at the cost of another – like immunological health. I’ll be discussing this more in a future article on the industrialized microbiota.

Thanks to everyone who attended, presented, or otherwise helped make the Virtual Microbiome Summit happen! I hope you enjoyed this summary as much as I enjoyed hosting the event and watching the replay. If you did, be sure to share your favorite highlights in the comments below, subscribe to my blog newsletter, or become a patron for monthly gut research updates!