Using molecular techniques that detect all known types of microbes and borrowing statistical techniques from field ecology and population genetics, Paul Eckburg, MD, a postdoctoral scholar in infectious diseases and geographic medicine, conducted the most extensive study to date surveying the inhabitants of the lower digestive tract. In the three healthy subjects he studied, he found 395 unique bacterial species.
"The intestinal flora is critical to human physiology and a wide spectrum of disease, but the first step in studying this ecosystem is to figure out who is there and how the community census varies in time and space," said Eckburg, the lead author of the study published in this week's online edition of Science Express. "But even with this large sequencing project, which produced orders of magnitude more sequence data than had been generated in the past, we are not completely there yet. This is just the tip of the iceberg."
Eckburg works with David Relman, MD, associate professor of medicine (infectious diseases and geographic medicine) and of microbiology and immunology. Relman's lab at the Veteran Affairs Palo Alto Heath Care System specializes in microbial pathogen discovery and human microbial ecology, and in appreciating the roles played by microbes in human health and disease.
The paper by Eckburg and Relman is intended as the first of several studies looking at how microbial communities vary according to host, diet, geography, disease and other variables.
To distinguish individual bacteria among the hundreds of types in the samples, Eckburg took advantage of a technique that compares the genetic sequence of a molecule shared by bacteria and archaea. The molecule, 16S rRNA, plays a role in the translation of the genetic code and thus is critical to the organisms. However, small variations in the 16S rRNA gene sequences allow scientists to detect distinct bacteria.
For each of the three research subjects, the researchers analyzed samples from six different anatomical sites inside the large intestine as well as a stool sample.
Eckburg and his team determined more than 13,000 sequences of 16S rRNA. Most of what they identified included the usual suspects in the intestinal flora, but there were surprises. Nearly two-thirds of the bacteria they identified were novel, meaning that they had no genetic close neighbors in the existing databases that store sequence information about all known species.
"We thought we would find new ones, but it was a bit surprising to see such a large percentage that had remained unidentified," said Eckburg. "Despite this large effort, we are still not approaching the point of complete coverage of the intestinal community in any one individual - much less the complete coverage of all human intestinal communities." In contrast, the investigators found very limited diversity within the archaea, microbes that look a lot like bacteria but are genetically and biochemically as different from bacteria as bacteria are from humans.
The team also used the samples to examine how microbial communities differ among locations within the intestine as well as among different people.
"We were surprised at the degree to which an individual determines the particular picture of microbial diversity that we see," said Relman. "Variation between individuals had been suspected, but our data proved to be a dramatic confirmation of this belief."
An interesting corollary of this finding is that the ability to create a comprehensive census of the microbes living inside a person could have implications for human forensics. Unlike people's genetic makeup, their endogenous microbial communities may indicate where they have been and what activities they have pursued in terms of travel, diet, antibiotic use and the like.
Relman emphasized that the study looked at only a limited number of individuals and examined only one genetic feature of the microbes, so there are many studies that must and will follow.
It took the team about a year to carry out this analysis of the microbial flora from three people. Although the process can now be done faster, Eckburg and Relman said they plan to complement this current approach with the use of microarray technology. The latter will allow them to analyze samples much more quickly. They have been working with biochemistry professor Patrick Brown, MD, PhD, and graduate student Chana Palmer to develop a gene chip that can perform a microbial census on complex communities and samples.