Malnutrition seems like an intuitive problem: you don’t eat enough food, so your health suffers. But it’s not that simple. One mysterious type of malnutrition known as kwashiorkor—characterised by leaky blood vessels, puffy limbs, distended stomachs, and fragile skin—often affects children who eat just as much as their healthy neighbours. And even if these kids get to munch on protein-rich food, some don’t recover.
A team of scientists, led by Jeff Gordon at the Washington University School of Medicine, has shown that children with kwashiorkor harbour 11 species of gut bacteria that, together with their poor diets, conspire to damage their guts.
These results suggest that this particular type of malnutrition isn’t just caused by the absence of food, but also by the presence of the wrong microbes.
The team first started studying kwashiorkor in Malawi a few years ago, after noticing that some children developed the condition while their identical twins did not. Why the difference? The twins had the same genes. They ate the same food. They lived in the same village. But their gut microbes were very different.
These microbial communities change over time, much like plants colonising a burnt forest or a new island. Lichens and mosses come first, before giving way to shrubs, then trees. Likewise, in the gut, milk-eating microbes give way those that digest plant matter. Waves of species succeed and replace each other, until they settle on a stable, mature, and more diverse community. In a normal gut, this takes about three years. But in the kwashiorkor kids, the changing communities had stagnated, leaving them with immature microbes for their age.
When Gordon’s team transplanted these immature communities into mice with no microbes of their own, the rodents lost weight—but only if they also ate the equivalent of a poor Malawian diet. The combination of poor food and immature microbes triggered the symptoms of kwashiorkor.
But which microbes are important? Is it the entire community, with its hundreds or thousands of species? Or does the problem lie with a smaller cabal? To find out, Gordon relied on an antibody called IgA. Immune cells release this substance into the gut, where it piles onto microbes to create immobilising coats. Around half the bacteria in our gut are restrained in this way. By looking for these targeted species, “you can use the immune system to mine the microbiota,” says Gordon.
His postdocs Andrew Kau and Joseph Planer began by transplanting microbes from a pair of 21-month-old Malawian twins—one with kwashiorkor and one without—into germ-free mice. They then used a technique called BugFACS to pull out any bacteria that were restrained by IgA. They effectively used the antibody as a fishing rod to hook microbes that draw the immune system’s attention.
In the mice colonised by the ‘kwashiorkor’ microbes, IgA pulled out large numbers of Enterobacteriaceae (prounounced En-ter-oh-back-tee-ree-ay-see-ay). The team then loaded this IgA-targeted set of microbes into more germ-free mice. The rodents fared badly. Half of them died within five days. When the team looked at their guts under a microscope, they saw carnage.
A normal gut has tightly packed cells to prevent microbes from slipping through, and dense forests of tall pillars for absorbing nutrients. In these guts, the cells were pulling away from each other, and the pillars were shrunken and shredded. Imagine a fence with wide gaps between rotting slats. “The [lining] was really just torn apart,” says Gordon. “It was pervasive and dramatic.”
Kau and Planer isolated several of the bacteria within this lethal community and identified a set of 11 species that collectively destroy the gut. These included three Enterobacteriaceae, and several common gut inhabitants like Bacteroides fragilis and Bacteroides thetaiotamicron. Individually, these microbes did very little. Collectively, they led to shredded guts and severe weight loss. “It’s not just one actor,” says Gordon. “It’s the concentred effort of several organisms.”
And as before, the team showed that this cabal only harmed mice that ate a Malawian diet. If the rodents ate more nutritious meals, the microbes were benign. As the team showed in their earlier work, it’s the combination of diet and microbes that makes for poor health.
The team used the same techniques to show that healthy twins, who don’t get kwashiorkor, have guts that are rich in two particular bacteria. The first of these, Akkermansia muciniphila, can protect mice from being obese; it seems that it protects them from malnutrition too. The second one, Clostridium scindens, is part of a group that stops the immune system from overreacting. It was recently shown to single-handedly block infections by its deadlier cousin—Clostridium difficile, a bug that causes severe diarrhoea. These two microbes were enough to defend mice from the more destructive ones.
Having done all these experiments in mice, the team then returned to humans. They used their BugFACS technique on 19 more pairs of Malawian twins, to pull out the IgA-targeted microbes in their guts. And they found the same patterns: more Enterobacteriaceae meant a greater risk of kwashiorkor.
“This is a major advance in the field,” says Charlotte Kaetzel from the University of Kentucky, who studies IgA. “Of course, Jeff Gordon’s lab brings the most state-of-the-art methods to this type of study.” It’s important, she says, that the team combined experiments in germ-free mice, where microbes can be precisely controlled, with direct analyses of the stools of healthy and undernourished children.
This kind of approach is a staple of Gordon’s group. It lends weight to their claims that the microbes are actually causing kwashiorkor, rather than just going along for the ride, and that the patterns in mice are relevant to humans.
The team are now trying to understand how the 11 microbes that they identified damage the gut, and how C.scindens and A.muciniphila thwart them. They also want to know if the same patterns apply to malnourished people from other parts of the world, with different genes, diets, and cultural practices. In the long-term, they hope to develop ways of analysing a child’s microbes (perhaps, using BugFACS) to diagnose their risk of malnutrition before symptoms show, or even to develop probiotics containing bacteria that can forestall these diseases in places where food is scarce.
Reference: Kau, Planer, Liu, Rao, Yatsunenko, Trehan, Manary, Liu, Stappenbeck, Maleta, Ashorn, Dewey, Houpt, Hsieh & Gordon. 2015. Functional characterization of IgA-targeted bacterial taxa from undernourished Malawian children that produce diet-dependent enteropathy. Science Translational Medicine. http://dx.doi.org/10.1126/scitranslmed.aaa4877
PS: I’ve written two pieces today about the microbiome. In this one, Akkermansia protected mice from malnutrition caused by other microbes and a poor diet. In the other, Akkermansia was associated with inflammatory disease, in mice that ate a diet rich in food additives. In other rodent studies, it stops mice from getting fat, but is more common in cases of bowel cancer. All of this illustrates a point I’ve made before: any one microbe can have very different effects in different contexts and circumstances. There is no universally “good” bacterium, no universally “healthy” microbiome.