Deciphering the Microbiome

For all its health potential, the microbiome won’t give up its secrets without model systems

by Erik Ness

Grow, Fall 2013


Jim Steele used to be one of the skeptics. He’d be at a conference, listening to early research on the health benefits of probiotics. Steele scoffed at the small experiments. “We would literally try not to laugh in the audience, but we’d laugh pretty hard when we went out that night,” he admits.

But slowly the punch lines gave way to revelation. Steele, a professor in CALS’ Department of Food Science, conducts research on lactic acid bacteria, with a focus on Lactobacillus species. They’re important for human gut health, critical for the production of cheese and yogurts, and are the most common probiotic genus. He knew how incredibly useful they were, but still watched with a humbling disbelief as the data on the health potential of these microbes kept getting broader, deeper and more intriguing.

Our microbiota—what we call the totality of our bacterial companions—is ridiculously complex. Each human harbors a wildly diverse ecosystem of bacteria, both in the gut and elsewhere on the body. They have us completely outnumbered: where the typical body may contain a trillion human cells, your microbial complement is 10 trillion. They have 100 times more genes than you, a catalog of life potential called the microbiome. (The terms “microbiome” and “microbiota” are often used interchangeably in the popular press.)

While our initial, germaphobic impulse may be to freak out, most of these bacterial companions are friendly, even essential. On the most basic level they aid digestion. But they also train our immune system, regulate metabolism, and manufacture vital substances such as neurotransmitters. All of these things happen primarily in the gut. “In many ways the gut microbiota functions like an organ,” says Steele. “It’s extraordinarily important for human health,” with as much as 30 percent of the small molecules in the blood being of microbial origin.

Early research has suggested possible microbiota links to protecting against gastric cancers, asthma, numerous GI disorders, autoimmune disease, metabolic syndrome, depression and anxiety. And the pace of discovery seems to be accelerating; these headlines broke in just a few months last spring:

• Mouse studies suggested that the microbe Akkersmania muciniphila may be a critical factor in obesity;

• Kwashiorkor, a form of severe malnutrition that causes distended bellies in children, was linked to a stagnant microbiota;

• Risk of developing Type 2 diabetes was linked to an altered gut microbiota.

The catch: For all the alluring promise of microbes for human health—and it’s now clear they’re critically important—we have almost no idea how this complex system works.

The human gastrointestinal (GI) tract is a classic black box containing hundreds or thousands of species of bacteria (how many depends on how you define a species). There are viruses, fungi and protozoans. Add to that each person’s distinct DNA and their unique geographic, dietary and medical history—each of which can have short- and long-term effects on microbiota. This on-board ecosystem is as unique as your DNA.

Beyond these singularities, the action is microscopic and often molecular, and even depends on location in the GI tract. Most microbiota studies are done with fecal material. “Is that very informative of what’s going on in the ileum?” asks Steele, referring to the final section of the small intestine, which is thought to be the primary site where immunomodulation occurs. “From an ecosystem perspective, fecal material is many miles away from the ileum. Is it really reflective of the ileum community?”

Developing the tools to unlock this black box begins with simply accepting the idea that these bacteria—germs!—are part of us. It’s a fundamental shift in how we think about health, which has evolved for centuries around the prism of disease development, or pathogenesis. For centuries we had no idea that microorganisms caused plague, cholera, and dozens of other debilitating diseases.

“Because we couldn’t know who they are and what they’re doing, we focused on pathogenesis,” explains Margaret McFall-Ngai, a professor in the Department of Medical Microbiology & Immunology at the UW–Madison School of Medicine and Public Health. Once we knew that bacteria existed and developed the germ theory, modern medicine grew by leaps and bounds. “Pathogenesis has had such a profound effect on human history,” she notes.

Except that’s not how the world normally works. McFall-Ngai studies mutualism, where microbes and their host organisms scratch each other’s backs. For the last 25 years, she and colleague Ned Ruby have been untangling the elegant relationship between the Hawaiian bobtail squid and its luminescent bacterial symbiont Vibrio fischeri. She argues that these collaborative relationships are far more important than we realize—that instead of viewing the world through the framework of disease, biology needs to be understood through the prism of beneficial microbes.

“I think we are in a revolution,” McFall-Ngai says. She recently was lead author—Ruby was one of 25 others—of a major PNAS review. They argued that new technologies have “revealed a bacterial world astonishing in its ubiquity and diversity” and that the resulting relationships in symbiosis and in larger ecosystems are “fundamentally altering” our biological understanding. “All biologists will be challenged to broaden their appreciation of these interactions and to include investigations of the relationships between and among bacteria and their animal partners as we seek a better understanding of the natural world,” the authors state.

Model species like yeast, mice and fruit flies are common workhorses for scientific discovery. By exhaustively breaking down and manipulating these organisms in the lab, scientists have been able to decode a huge array of biological puzzles.

Studying symbiosis—two organisms intertwined, often in a mutually beneficial relationship—adds a layer of complexity to these models. Since the 1960s scientists have been working with Steinernema, a large family of very small worms—nematodes—and their symbiont, Xenorhabdus bacteria. When Heidi Goodrich-Blair joined the CALS bacteriology faculty in 1997, she had already begun unraveling how the worms and their symbionts communicate on a molecular level.

Steinernema are a popular organic control for greenhouse pests. Just a couple hundred microns in length—about 250 million fit in a cup—they prey on insects in their larval stage, entering their target through natural body openings. Creepy, but it’s their bacterial symbionts that do the killing. Xenorhabdus live in the intestinal tract of Steinernema—the worm protects them from ultraviolet radiation. Xenorhabdus infect the insect host when they’re excreted and cause a raging infection that kills the insect, setting up a perfect incubator that can produce more than a million Steinernema offspring.

So how can this worm and its virulent symbiont help us understand human microbiota? The two biggest inquiries are which bacteria are present, and how they contribute to our health. The Steinernema/Xenorhabdus relationships don’t give us direct answers, but they help us refine the questions.

Current techniques for analyzing gut bacteria decode DNA and its related compounds. Genomics identifies all of the genes present, but a lot of DNA is not used regularly, if at all. It’s like drawing conclusions about your diet by analyzing your cookbook library. Genomics shows everything you could possibly create—but not what you actually make. “All we’re doing is sequencing the potential,” says Goodrich-Blair.

Transcriptomics, on the other hand, decode what genes are active by recording the RNA messages from the genes actually in use. But with so many species present in the gut, it’s not possible to link microbes to genes.

These tools provide a lot of information, but how do you make sense of it? “That’s where model systems come in,” says Goodrich-Blair. “We have the ability to tease cause and effect out of our systems. We can inactivate specific genes in specific microbes, and then we can ask, ‘What impact does that have?’”

With a good model, nature has done some of that genetic engineering already. It’s often reported how close human genes are to those of evolutionary cousins like chimpanzees (98 percent identical). Bacteria have fewer genes overall but are far more diverse. In E. coli, for example, nearly 40 percent of the gene content is variable within that species. To make things even more confusing, bacteria can even transmit DNA horizontally, across species barriers.

“It’s very difficult to define what a species is in bacteriology,” explains Kristen Murfin, a fifth-year grad student in Goodrich-Blair’s lab. Her work with animal-associated microbes focuses on variations in strains, a level below species.

Murfin is testing how important strains are by examining a group of closely related Steinernema worms. In nature each subspecies has its very own Xenorhabdus bovienii strain for a symbiont. In the lab she can cross worms with various bacterial strains to see if the worms’ fitness—their ability to find and infect prey, and how much (if any) offspring are produced—suffers.

“Strains are so different,” explains Murfin. “Arguably, who the microbes are is important. But what they can do is maybe more important because if you have two strains of the same species that can do two different things metabolically, they are going to have very different impacts on the host.”

If strain matters in these very simple models, the implication is that we probably need to be looking at a finer scale in the human gut. “You would not be able to distinguish the difference between these strains using the technology that we currently have and at the level we’re looking in the human gut,” says Goodrich-Blair. “If some strains of bacteria are better for us than others, then it matters which strain we have in us. So we will have to dig down to a deeper level of bacterial identity than we have been.”

When choosing a model organism, one could do worse than the Hawaiian bobtail squid—Euprymna scolopes. (Imagine punctuating the grad school grind with occasional collection trips to Paiko Lagoon on Oahu.) But that’s not what led microbiologist Ned Ruby and invertebrate zoologist Margaret McFall-Ngai to the squid in the first place. What makes these mollusks special is their light-emitting bacterial ally, Vibrio fischeri. The squid live in the shallows and spend their days buried in the sand. At night they hunt—and are hunted. Under the bright tropical night, the squid would cast a faint shadow on the ocean floor. The squid use the bacteria in their light organs to erase their shadow, so predators can’t triangulate their position.

This unique biology makes the squid-vibrio relationship incredibly useful. In a lot of animal-microbe associations, animals are born with their symbionts. For example, if you want a mouse without microbes, you have to deliver it via caesarean section. With the squid, you just have to let it hatch in water without V. fischeri. Because V. fischeri glow, you don’t have to kill the squid to find out if it’s been colonized—another challenge with many symbiosis models. And because V. fischeri provide light instead of the more common nutritional assistance, you can deprive the squid of its symbiont in the lab without affecting its health or ability to survive.

But perhaps the biggest benefit is that V. fischeri live in direct contact with the squid’s epithelial cells, similar to the cells that the human body presents to the microbial world. Humans have 10 organ systems, and eight have epithelial and mucosal surfaces that interact with the external environment and maintain communities of beneficial bacteria. The squid-vibrio system offered “a way to understand how bacteria talk to animal tissue,” explains McFall-Ngai.

A young squid has a juvenile light organ that filters V. fischeri from the vast array of species available at sea—the first communication between the symbionts. Once the squid has captured its V. fischeri, the filtering organ isn’t needed anymore, and within four days it’s gone. This development is triggered principally by exposure to two compounds excreted by the V. fischeri, lipopolysaccharide and peptidoglycan, that are commonly associated with bacterial pathogenesis. In addition, the light of the symbiont itself participates in triggering these changes.

In the lingo of pathogenic microbiologists, these compounds are pathogen-associated molecular patterns (PAMPs). But in a paper in Science, McFall-Ngai argues that they would more accurately be called microbial-associated molecular patterns (MAMPs). These substances may have been discovered while unraveling a few virulent pathogenic processes, but in fact many species of bacteria in your gut create—and communicate with—the same substances.

It’s an attempt to wrest scientific lingo from the pathogenic worldview. Sometimes these bacterial products are benign. Sometimes they’re necessary in the gut, but they’re bad actors in the bloodstream. “I see them as a language. It’s not just what you say, it’s how you say it,” explains Elizabeth Heath-Heckman, a senior graduate student in the McFall-Ngai lab. “It’s a difference between talking to someone in a normal tone and yelling or swearing at them. It can be the same word, but it’s a completely different context.”

Getting back to the squid, the discovery that MAMPs trigger the loss of parts of the juvenile light organ has important implications. It’s been known for a long time that the mammalian gut and its associated immune tissue requires interaction with gram-negative bacteria for proper development. “Nobody could ever figure out why,” says McFall-Ngai. This suggests that the cellular language that animals use to communicate with bacteria is deeply embedded in our genetic code. “Of course it’s doing a different thing, but in animals as divergent as mice and squid it’s the same simple molecules,” she says. “We know that animals have been associating with bacteria since the beginning of their evolution,” she explains. “It allows you to look at the experiments that nature has done through evolution to try to gain insight into how these things work at a basic level.”

Recently McFall-Ngai’s lab has linked microbial symbionts to another hot field, circadian rhythms. V. fischeri produce the same blue light that plants and animals use to tell that it’s daytime. Circadian circuitry controls far more than bedtime; its malfunction could underlie a wide variety of problems with immunity, metabolism and mental health.

Heath-Heckman wanted to know if the V. fischeri bacteria helped the squid tell time. Using genetically modified V. fischeri to turn off light production, she deduced that light alone couldn’t keep the squid on a circadian cycle: it also needed circadian feedback from the bacteria to tell time.

Her paper, published in April, was the first to link microbiota and circadian rhythms. Just two months later, another group reported that in mice treated with antibiotics, the circadian rhythms in the gut subsided—suggesting that, like the squid, the bacteria and their chemical language might be necessary for maintaining those rhythms.

“If it only happens in the squid, then it’s only cool in one dimension,” says Heath-Heckman. But if microbes are major players in timekeeping in general, the microbiologists have just opened a new frontier in circadian biology. “This may be a shared trait among animals and their bacterial symbionts. That’s part of the power of our system,” she concludes. “It really can tell us things about how bacteria associate with animals.”

Jim Steele’s lab primarily works on Lactobacillus casei, and its diversity is a testament to the power of just one microbial species. One of Steele’s projects focuses on cheese flavor; another is tweaking strains to enhance ethanol production for use as a biofuel; and yet another works on the utility of this species as a probiotic.

Recently, Steele’s graduate student Travis De Wolfe got his first real data, and he can hardly sit still. Now entering his second year in the lab, he is helping to build a model of the entire human gut, using both mice and pigs. Steele’s goal is to figure out how to use probiotics to treat Clostridium difficile, a nasty human diarrheal illness that is associated with antibiotic usage. De Wolfe eagerly pulls up a chart breaking down the microbes in two mice: one treated with 1 million colony-forming units (a measurement of living bacteria) of Lactobacillus casei strain 32G and one treated with 100 million of the same organism.

Two interesting things appeared in the data. One was that, despite the massive infusion of Lactobacillus casei, none were detected in the cecum, the region midway through the gastrointestinal tracts that De Wolfe sampled. The other was that the higher dose of probiotic significantly reduced another bacterium, Lachnospiraceae. That’s the same order as C. difficile, their ultimate target. Both findings suggest complex ecological relationships.

“It’s pretty astounding for preliminary data,” De Wolfe says, then cautions that it means nothing by itself; they’re just trying to get their methods down. “We need to pull it apart.”

A probiotic is defined as a live microorganism which, when consumed in an adequate amount, confers a health benefit on the host. It’s implied in the definition that the microbe matters, and that the dose matters. But there’s a caveat, says Steele: “Weirdly, we’re this deep into the research … and no one’s actually proven those two very basic tenets of probiotic therapy.”

That there are benefits of taking probiotics is clear. Steele describes a famous study that looked at preschoolers in China. One group received a single strain of probiotics, another received two different strains, and the third received placebos. Over six months, the groups that got probiotics missed a lot less school and had fewer symptoms of upper respiratory infection. And the kids who received two strains did the best, though ultimately the difference between receiving one or two probiotic strains wasn’t statistically significant.

The benefit of the probiotics was clear, but how did it work? Was the children’s immune system on alert because of exposure to a bolus of probiotics? Did the probiotics trigger a change in the microbial ecosystem in the ileum that in turn resulted in upregulating the immune system? “We don’t know the mechanism,” states Steele. “If scientists are going to take probiotics and dietary interventions to the next level, we’ve got to understand the underlying mechanisms. If you don’t understand the mechanism then you simply can’t optimize and control the health outcome.”

A large number of people already take probiotics and follow special diets, and scientists are trying to capture some insights from that real-life experiment. Steele needs a middle ground between this complexity and the stripped-down systems like the squid-vibrio.

Pigs are favorite model systems because they so closely mimic human systems. Steele has already fed pigs a humanized diet and found that he can get reasonably close to re-creating the human GI tract. “I’m willing to have a smaller n and to pay more to utilize a model that more closely mimics humans for some experiments,” he says (n referring to the number of experimental subjects). But with yards of intestinal tract, pigs are still unwieldy. He’s hoping to use mice in parallel with pigs to shorten the experimental cycle.

There is vast experience in translating mice to humans, but Steele is still uncomfortable with the trade-offs. For example, mice have a foregut colonized by Lactobacillus—an organ with an obvious role in defining the gastrointestinal microbiota, but missing from humans altogether. But by extrapolating between mice and pigs he hopes to make the translation to human health. “There is just no other model system that allows us to economically get the n,” says Steele. “We have to utilize model systems where we can have greater control, run a larger number of samples—and mice give you that opportunity.”

On his office door, Steenbock Professor of Microbiological Sciences Ned Ruby has posted a New York Times article about a do-it-yourself fecal transplant. The writer’s friend had ulcerative colitis, and early research suggested that implanting microbes from a healthy gut might help defeat this difficult disease. But new guidelines from the federal Food and Drug Administration had curtailed the willingness of physicians to perform the experimental procedure.

Instead the procedure was guided, long-distance, by a physician. The overall logic about why it might work was sound. But exactly why it worked could remain a mystery for years. In a nutshell, that’s the promise and peril of microbes in human health.

Ruby has worked with McFall-Ngai on the squid-vibrio system for the last 25 years, and the accelerating proliferation of ideas and evidence around microbes is exciting. “The field is opening up beautifully,” he says. “It’s like you’re just coming over the horizon and you’re beginning to see the tops of a town. You can’t see the whole town yet, but you’re beginning to see the tops of the buildings. It’s pretty clear there is going to be a town down there.”

For all the excitement, Steele thinks the data on the health impacts of probiotics is probably going to get more confusing before it becomes clearer. The answers will come from complex mathematical analysis, crowd-sourced epidemiology, happy accidents and dogged insight. But without the model systems being built at UW and beyond, we’d never unlock the box.

“The importance of studying model systems cannot be overstated, in my opinion,” concludes Goodrich-Blair. “You can’t study one thing and get a paradigm. You have to study a whole bunch of different things to get the paradigm.”

“It is a huge black box,” Steele notes. “The tools we have today are not going to be the last tools that we employ to look at this system.”

Email this to someoneShare on LinkedInShare on RedditShare on TumblrTweet about this on TwitterShare on Facebook

Leave a Reply

Your email address will not be published. Required fields are marked *