ISTOCK, TLFURRERIf we think of our bodies as walking, talking ecosystems, it stands to reason that microbial collections may change in response to the foods we eat and places we live. Similarly, disease-related shifts in the human environment are expected to influence our resident microbes, even if those microbes themselves don’t cause the changes.

Investigators and investors are increasingly banking on the possibility that microbes finding safe harbor in human habitats may be useful for detecting disease—from acute infections to chronic inflammatory conditions.

“It makes sense that the microbiome could be an indicator of lots of different things going on in the body,” said University of California, Davis, evolution and ecology researcher Jonathan Eisen, whose lab studies microbial contributions to ecosystem health. “The community of microbes is a pretty good readout of availability of carbon and nitrogen and activation of the immune system and oxygen levels and moisture...

From that point of view, it’s not all that big a leap to think that monitoring microbial community members, genes, or outputs—be they proteins or metabolites—could offer methods for measuring disease status or perhaps predicting individuals’ responses to specific treatments. A number of companies are working on developing tools to do just that.

WIKIMEDIA, PHOTO BY ERIC ERBE, DIGITAL COLORIZATION BY CHRISTOPHER POOLEY, BOTH OF USDA, ARS, EMUOne intense focus on using gut microbes in diagnostics is to track inflammatory bowel diseases (IBD) such as Crohn’s or ulcerative colitis, because the inflammation that marks those conditions may dramatically alter the types of microbes that survive and thrive in the large intestine.

Past studies have unraveled some of the most telling gut microbiome features in IBD — and they go well beyond the presence or absence of specific bugs in the gut. Individuals with these and other maladies often have lower-than-usual gut microbe diversity and decreased levels of harmless, or “commensal,” microbes as well, providing the potential for multiple test outcomes to explore.

San Francisco-based uBiome, for instance, offers a “SmartGut” test that involves targeted sequencing of a 16S ribosomal RNA bacterial barcode gene to measure microbial composition. That test is performed in a CLIA-certified and College of American Pathologists-accredited lab and reports on microbial patterns that have been linked to conditions ranging from IBD to type 2 diabetes to obesity. (Eisen recently stepped down the company’s scientific advisory board to avoid potential conflicts with another project he’s pursuing.)

The firm’s website suggests the approach has high accuracy when it comes to profiling the gut microbes it targets—boasting 99 percent sensitivity for picking them up, on average, and 100 percent specificity for the targeted bugs. Where it gets trickier is translating such patterns into clinical clues.

To that end, uBiome researchers published a recent PLOS ONE study aimed at defining the normal variability in the gut microbiome. Using a 16S rRNA sequencing strategy similar to that underlying the SmartGut test, they profiled stool samples from nearly 900 healthy individuals to track gut levels of 28 gut microbial species or genera, including five pathogens, 20 commensal organisms, and three species that are believed to be beneficial. They also outlined the associations documented between these microbes and 13 disease types.

“The combinatorial information of which organisms are outside of the healthy range can be used by a physician to augment a treatment plan,” they wrote, noting that the test panel considered in the PLOS ONE paper “reports on some microorganisms that are not usually interrogated in the clinic but provide additional insight into the overall gut health of a patient in a clinical setting.”

WIKIMEDIA, UNITED STATES DEPARTMENT OF AGRICULTURELike other clinically focused firms, uBiome markets its SmartGut test as a screening tool rather than a definitive diagnostic, providing results that are usually closer to a cloud of microbial clues than a single data point depicting the presence of a single pathogen. The UK-based Map My Gut is also using 16S sequencing for its tests, which are ordered through registered health professionals under the National Health Service for individuals with health concerns such as gastrointestinal conditions or obesity as well as healthy individuals who are curious about their gut microbiome and diet. Again, microbiome patterns are the predominant outcome and the company’s site emphasizes that these tests “are not medical diagnostic tools.”

Another company is seeking even firmer clinical evidence of its proposed diagnostic test. Metabiomics, a subsidiary of BioSpherex, just wrapped up a clinical trial aimed at understanding whether its bacterial DNA barcode sequencing method can predict colon cancer from stool and/or colonic mucosal samples. Investigators enrolled 260 individuals between the ages of 45 and 80 years old, who had also been tested by colonoscopy, for the trial. The study wrapped up in March and results have not yet been released.

Still other firms, such as Second Genome, are more focused on mining microbiome data to develop new disease treatments for metabolic disease or conditions such as IBD, though corresponding diagnostic tests may be used to target such treatments to individuals most likely to benefit from them, explained Second Genome co-founder and senior bioinformatics director Todd DeSantis.

While there is “enormous potential” in microbiome-based diagnostics, said Eamonn Quigley, a clinician and researcher at Houston Methodist Hospital, “we need to tread carefully” in interpreting microbiome data and avoid confounding clues when trying to use these data to make diagnoses.

Quigley wrote a perspectives article published in Nature Reviews: Gastroenterology & Hepatology in May that outlines a series of considerations—from microbiome dynamics to an individual’s diet, exercise habits, age, health status, and more—that may affect microbial test interpretation. He chairs the advisory board for an Irish company called Alimentary Health, which taps microbial data for insights into treating inflammatory and infectious disease.

At the moment, established microbiome-based screening tests are one piece in a patient's clinical puzzle, rather than a method for landing on a firm diagnosis. But the tests are being aggressively pursued in the diagnostic realm because they could obviate more-invasive forms of diagnostic testing.

In a recent Cell Metabolism study, for instance, investigators documented differences in gut microbial communities between individuals with non-alcoholic fatty liver disease (NAFLD) who had either mild to moderate or advanced liver fibrosis. The team got a glimpse at these gut microbes using bacterial DNA extracted and sequenced from subjects’ stool, a sample type that’s a lot easier to get ahold of than the liver biopsies typically used for confirming NAFLD status in the study’s 86 participants.

Using metagenomic sequencing, the researchers saw higher-than-usual levels of Escherichia coli in gut microbiomes from the 14 NAFLD sufferers with advanced fibrosis, along with differences in representation by three dozen other microbial species.

By combining information on these advanced fibrosis-associated microbes with clues about individuals’ gut microbial diversity, age, and body mass index, the team came up with a model to predict advanced fibrosis across individuals with NAFLD. These results—coupled with preliminary validation tests in other individuals with cirrhosis or advanced fibrosis—suggest it might eventually take nothing more than excrement and expertise to find advanced liver disease.

“This is a relatively young field, so you always have to be careful with the possibilities for translating these findings,” said Mount Sinai Icahn School of Medicine microbiome researcher Jose Clemente, whose lab is in discussions with a number of companies interested in using microbes to treat IBD. He believes it may not be enough to consider gut microbes alone, but also interactions with their hosts, which may be driven by individuals’ genetics and susceptibilities to certain conditions.

“Just by looking at any of the factors [in isolation], I don’t think you’re going to have very accurate diagnostic tools,” Clemente explained. “You’re going to get much better accuracy and diagnoses when you combine the factors.”

Interested in reading more?

The Scientist ARCHIVES

Become a Member of

Receive full access to more than 35 years of archives, as well as TS Digest, digital editions of The Scientist, feature stories, and much more!