How inflammation has shaped the microbiome over millions of years
CMIT researchers have found that hundreds of strains of bacteria have evolved to thrive in the inflamed gut. A better understanding of those adaptations could help us treat disease.
By Lindzi Wessel
We tend to think of our organs as wholly ours, a part of us that’s made of our cells, and dedicated in purpose to our healthy functioning. But to the trillions of bacteria crowded into our digestive tract, the organs that comprise it are also their habitat — an ecosystem that offers the shelter and nutrients they need to survive and that, when it changes, forces them to adapt.
Understanding how specific bacteria might influence our health has long been a priority of scientists who study conditions like inflammatory bowel disease, a lifelong and sometimes debilitating ailment associated with chronic inflammation of the gut. But few studies have investigated how such a disease might, in turn, drive the evolution of the bacteria that call the gut home.
However, a uniquely comprehensive analysis led by members of MIT’s Center for Microbiome Informatics and Therapeutics (CMIT) identified hundreds of bacteria strains that are linked to inflammatory bowel disease (IBD) and pinpointed specific attributes of those bacteria that allow them to thrive in an inflamed gut. The research, published in Cell Host & Microbe, estimates that some of these adaptations have been around for millions of years and are much more prevalent than previously thought, suggesting that relationships between gut inflammation and bacterial communities have been entrenched throughout human history.
“This is the first study to look at the evolutionary relationships of gut bacteria on a large scale in order to find lineages that are associated with inflammation,” says Christopher Smillie, a CMIT member and assistant professor at Harvard Medical School, and senior author on the paper. “That approach gives us a higher resolution lens to understand disease.”
An evolutionary perspective
Determining how the microbiome influences disease is complicated. A complex, ever-shifting community of trillions of interacting cells, the gut microbiome poses unique analytical challenges. Traditionally, most researchers have analyzed the microbiome by cataloging bacteria in the gut by species, using a specific snippet of DNA (known as “16S rDNA”) that is found in all bacteria as a barcode to distinguish one species from another. But this technique doesn’t account for the vast diversity across strains — genetic variants within species which, in some cases, differ in more than half the genes they harbor.
It stands to reason that these expansive genetic differences mean strains of the same species might have substantially different effects on the environment around them. And, indeed, microbiome researchers long ago identified a handful of strains within well-studied bacterial species whose presence in the gut seems to be a harbinger of developing IBD. But it’s never been clear to researchers just how these strains influence disease progression and how they might have arisen from species that — apart from those anomalous strains — are thought to be harmless.
Smillie and his team wondered if the answer to such questions might lie in how different strains have evolved. Inflammation, they reasoned, drives a lot of big changes in the gut, including shifts in oxygen levels, acidity, and nutrient availability. Could those conditions be responsible for the emergence of disease-linked strains?
“We figured that maybe these factors, together, created a selective pressure that led to the well-studied strains that we know are linked to IBD,” says Adarsh Kumbhari, a postdoctoral fellow in the Smillie lab and first author of the paper. “And if that selective pressure did exist, we figured there’d be no reason for it to be unique to those strains.”
To gain a better understanding of how that pressure might work, the team reconstructed the genomes of all the strains within a few key bacterial species and mapped them into evolutionary trees, which estimate the relationships between strains and when they likely diverged from common ancestors. To their surprise, once the trees were built, the team found lineages associated with IBD in every species they looked at. They tried looking at a few more species, and, again, IBD-associated strains popped out.
“We wondered whether we would find any evolutionary associations with IBD,” says Kumbhari. “And we quickly realized ‘oh, wow, we may actually have something really important here and we should keep looking.’ We decided to look across the whole microbiome.”
Leveraging public data
Sorting a person’s entire gut microbiome by strain is not a simple task. While a specific snippet of genetic material might be enough to identify different species from one another, further subdividing those species into strains requires a deeper mapping of each of the thousands of bacterial genomes contained within a person’s gut and then figuring out where each one came from — akin to sorting through millions of puzzle pieces to reassemble thousands of distinct puzzles.
As a graduate student in the lab of CMIT Co-Director Eric Alm, Smillie worked on computational strategies to tackle this problem and has continued to fine-tune these approaches with his team since launching his own lab. With such techniques at their disposal, his team realized they had the power to evaluate strains across the human gut microbiome using massive amounts of data already available to the public.
The team began exploring existing datasets for genetic information collected from the gut microbiomes of people with IBD and found many cases where other researchers had sequenced the entirety of genes found in stool samples, but hadn’t sorted those genes into the genomes to which they belonged. Using cutting-edge computational analyses, the team reconstructed the genomes of more than 140,000 bacterial strains. Prior to their study, a commonly used database contained only 500 strain genomes collected from IBD patients. Smillie’s study added more than 56,000.
Using this extensive data, the researchers showed that hundreds of strains were associated with IBD – many of them from species that were not previously associated with disease.
“What people have done in the past is just not high enough resolution,” says Smillie. “Looking at the strain level is showing us that these disease-associated bacteria are everywhere, we just haven’t been seeing them.”
The gut as an ecosystem
Further analysis showed that these strains aren’t new. The team estimated that they diverged from other lineages between 3.6 and 7.6 million years ago, possibly even before the emergence of modern humans.
To better understand this evolutionary persistence, the researchers turned their attention to the genetic differences between health-associated and IBD-associated strains. Wanting to know more about where disease-associated strains come from, the team looked at 59 different species and compared the genes in health-associated and IBD-associated strains within them. Looking across all species, the team found frequent changes in genes related to cell movement, antibiotic resistance, oxygen use, and toxicity towards human cells, among other traits that could prove useful in adapting to inflammatory conditions.
“Inflammation is an immune response that’s been around for millions of years,” says Kumbhari. “It makes sense that, over those millions of years, these bacteria would respond to deeply conserved elements of inflammation and develop traits to survive or even thrive in that context.”
In experiments that examined how the microbiome changed during bouts of inflammation, the team also observed that IBD-associated strains outcompeted other strains, and did so especially well in patients with IBD. When an inflammation bout was over, however, the prevalence of IBD-associated strains did not return to normal. Instead, levels stayed high for months or even years after an inflammation episode had receded — potentially setting the stage to exacerbate future inflammatory flares.
The study also revealed that certain beneficial bacteria likely play a critical role in shaping the IBD gut ecosystem as well. These strains progressively disappeared in experiments that tracked IBD as it got worse, suggesting that losing them also could mean the loss of protective factors against disease. The loss of certain strains of a bacterium called Eggerthella lenta showed a particularly strong link to IBD symptoms, which, the researchers say, could make them a powerful biomarker to help identify patients at risk of severe IBD.
“We’ve got this new understanding now of how chronic inflammatory diseases, like IBD, might shift the microbiome into a new state,” says Smillie. “Disease itself might influence the microbiome, and then through that effect, the microbiome might reinforce the disease. It gives us a new way to look at the problem.”