Nature vs. Nurture in the gut microbiota

It’s fun to watch how scientific truth changes with time. It shows that we can learn. But it’s amusing for me to compare what I told my first students with what I told my last students. When I started teaching, we didn’t know much about the microbes that lived in our gut—there were a lot of them, but they were generally regarded as being neither particularly harmful nor helpful, but just along for the ride. In the last decade, we have learned that these cells (and there are about ten of them for every one of our cells) are essential for providing nutrients, regulating metabolism, regulating the immune system, regulating the development and maintenance of the gut, and much more.

Although everyone has more or less similar general types of organisms in their gut, the balance of types varies considerably. Comparing the intestinal microbiota of two individuals is similar to comparing the yellow pages of Los Angeles and Fairbanks—they’ll both have the same classifications, but one will have a lot more acting coaches and the other will have more hunting guides. While the origin of the differences in yellow pages is largely due to environment, it’s more difficult to pin down the origin of the differences in our gut microbiota. Does diet have an effect on community composition? Does your genetics determine what microbes live in your gut?

A group from Washington University in St. Louis looked at an extremely simplified system to examine the influence of diet on gut microbes. Rather than studying humans, they used “gnotobiotic” mice—these are animals delivered by C-section, and raised in an absolutely sterile environment. These animals are pretty sickly (since the microbes that normally aid in development and digestion are absent), but researchers can deliberately add specific, known species of microbes to their food. This way, they know exactly what microbes are in there (gnotobiotic = “known life”), and by comparison with germ-free animals, they can know the effect of specific types of microbes.

To simplify their experiments, the researchers infected these mice with a community of 10 different types of microbes, chosen to represent the most general characteristics of the microbes in the human gut: one good at digesting starches, one good at fermenting amino acids, etc. They then gave these mice extremely simplified diets: mostly corn oil, mostly casein, mostly sucrose, or mostly cornstarch. After two weeks, the researchers collected the mouse poop, and analyzed the composition of the microbial community therein. This was repeated, until each mouse had enjoyed each diet.

After they had amassed a lot of mouse poop, they were able to develop a pretty good model of how the structure of the microbial community in the mouse gut was shaped by the mouse’s food intake. The model is actually relatively simple, and quite predictive: such a percent of protein in the diet leads to such a percent of Clostridium cells, and so on. The researchers were able to test their predictions by feeding the mice slightly more complex food (baby food, actually—pureed turkey, pureed carrots, etc). Sure enough, by knowing what went into the mouse, they were able to roughly predict the microbial community structure.

So, are we what we eat? Sorta. This was an extremely simplified system; thankfully our diets are more than corn oil, corn starch, casein, sucrose, and mashed peas, and we have thousands of different microbes in our guts. However, despite its simplicity, this study suggests that we can alter our gut microbiota at will. This can be medically important. For instance, the amino acid fermenting bacteria associated with the high protein diet are also associated with a variety of inflammatory syndromes; it would be nice to reduce their relative numbers.

However, another study reminds us that our family background is also a significant factor in determining our personal microbial community structure, and our health. Using mice, a group of researchers from Nebraska found a correlation between certain regions of DNA and certain types of bacteria in the gut.

Genetecists use the term “Quantitative Trait” to discuss something that can be measured (quantified) and inherited (trait). So, height is a quantitative trait. Once you have a quantitative trait, you can look for genes that influence it. In the case of height, there are dozens of such genes, each of which makes a small contribution to overall height. Each gene occurs in a specific region, or locus, in the DNA, so a geneticist could refer to a "Quantitative Trait Locus" (QTL) associated with height. It's useful to note that a QTL is not a gene; it's a region of DNA, potentially carrying many genes, associated with a particular trait.

The researchers decided to treat the composition of the gut microbiota as a quantitative trait. They did this by making a mathematical model that combined the relative abundances of 64 different microbes into one measurable variable; you could measure the relative numbers of these 64 microbes in the poop of any mouse, and compare it with any other mouse. Given the existence of so many different laboratory strains of mice, this made for a lot of data to analyze.

The big question, then, was whether there were any genetic loci that corresponded with variation in this quantitative trait. There were. At least 13 different places in the mouse genome corresponded with definite variations in their gut microbiota. If two mice had different versions of the same gene in one of these places, then they would likely have different ratios of different microbes in their guts.

So what genes determine what microbes we have? The answers are still a bit vague; we know that certain regions of the DNA are associated with these traits, but each region has several genes. Nonetheless, there are some cases where there are candidate genes that really stand out. For instance, there’s one QTL associated with increased numbers of two specific types of bacteria; this QTL contains genes specifically involved in controlling immune responses in the gut.

This study is highly suggestive, but not conclusive. It establishes a correlation, but the results were all retrospective; no predictions were tested. For example, it would be great to see a cross of two different strains of mice predicted to favor certain different bacteria. Would it be possible to accurately predict what bacteria the offspring would favor? Even without this bit of proof, this study represents a buttload of work, and a very promising entrée into a challenging question. Like the study with gnotobiotic mice, this study also has relevance to health and medicine, as it may explain the observed linkage between certain QTLs and diseases such as Crohn’s disease.

This research also raises a big question in the symbiosis between mammals and our gut microbiota. The authors of this study use the link between certain immune system genes and certain microbes to support the argument that the immune system evolved for the purpose to maintaining favorable microbes in the gut. However, I’m a microbiologist, and I look through the other end of the telescope. To me, this is further evidence that certain microbes have been exerting selection pressure upon us, in order to make a better environment for themselves.

Either way, your doctor’s advice for good health and longevity remains much the same: eat right, and choose good parents.

Andrew K. Benson, Scott A. Kelly, Ryan Legge, Fangrui Ma, Soo Jen Low, Jaehyoung Kim, Min Zhang, Phaik Lyn Oh, Derrick Nehrenberg, Kunjie Hua, Stephen D. Kachman, Etsuko N. Moriyama, Jens Walter, Daniel A. Peterson, and Daniel Pomp (2010). Individuality in gut microbiota composition is a complex polygenic trait shaped by multiple environmental and host genetic factors. Proceedings Natl. Acad. Sci. USA 107: 18933-18938.

Jeremiah Faith, Nathan P. Mc Nulty, Federico E. Rey, Jeffrey I. Gordon (2011). Predicting a Human Gut Microbiota’s Response to Diet in Gnotobiotic Mice. Science 333: 101-104.