NEW YORK (GenomeWeb News) — For the microbiome, there may be many versions of normal and healthy, according to a new metagenomic analysis by a pair of researchers at the University of Michigan.
As reported in Nature today, Michigan's Tao Ding and Patrick Schloss examined the microbial composition of various body sites of 300 healthy people from the Human Microbiome Project. Each site contained a number of microbial community types. For instance, community type A, one of the four distinct community types identified in stool samples, was marked by high levels of Bacteroides.
The makeup of the microbiomes at the different sites was also related. For example, the community of the oral microbiome could predict the gut microbiome community type. The makeup of each person's microbiome, the researchers said, also appeared to be influenced by the person's life history, such as having been breastfed.
"Understanding the diversity of community types and the mechanisms that result in an individual having a particular type or changing types will allow us to use their community type to assess disease risk and to personalize their medical care," Schloss, an associate professor of microbiology and immunology, said in a statement.
To study the bacterial community composition across the human body, Schloss and Ding turned to data from the HMP Consortium. This dataset included 16S rRNA gene sequence data and clinical metadata from 300 healthy people sampled at 15 body sites or 18 body sites, depending on if they were a man or woman. All 300 individuals were sampled twice and 100 of them were sampled three times.
Using a Dirichlet multinomial mixture model approach, the duo assigned the samples to community types. Each body site, they found, had between two and 18 different community types.
The types, they noted, weren't only marked by the most abundant bacterial populations they contained, but by complex patterns of co-occurring bacterial populations. In the gut, Schloss and Ding found that community type A had high Bacteroides levels, but lacked Prevotella and Ruminococcaceae, while community type C also lacked Prevotella, though had a lower relative abundance of Bacteroides.
The community type that a person had at one site could predict what community type they had at another. Schloss and Ding noted such correlations among sites in the mouth, vagina, elbow creases, and behind the ears. Stool samples, they said, showed a high correlation with samples from the oral cavity, particularly saliva samples.
For instance, people with stool community type D, which is marked by high Prevotella levels, were 2.1-times more likely to have saliva communities types A and C, which also have high Prevotella levels as compared to saliva communities type B and D.
"These results are intriguing because they suggest that although the oral and stool communities share little taxonomic resemblance, oral bacterial populations seed the gut, and those populations experience the ecological environment of the gut to give rise to consistent community types by the time they reach the stool," Schloss and Ding wrote.
By drawing on the repeated sampling data, the researchers further found that the microbiomes at some sites appeared to be more stable than others. The most stable microbiomes, by their analysis, are the stool and vaginal microbiomes, and the least stable is the supragingival plaque of the mouth. Among the stool communities, the researchers reported that stool community type D was the most stable, followed by types A, C, and B.
Based on the metadata that was collected by surveying participants, Schloss and Ding found that certain demographic and life history events could also be correlated to different microbial community types at different sites.
For instance, whether the person had been breastfed as an infant was associated with their stool community type — people who had been breastfed were 2.4 times more likely to have stool community type A and those who were not were 2.2 times more likely to have stool community type D.
Gender also influenced community types at a number of sites. Men were three times more likely than women to have stool community type D.
Additionally, whether a woman had a baccalaureate degree was also correlated with community type at the vaginal introitus and at the posterior formix, though the researchers noted that this association was likely due to a combination of related factors including socioeconomic class, sexual behavior, and race or ethnicity.
Still, the researchers said that because such a range of microbial community was seen among healthy women, they suspect that there is no one normal vaginal microbiome, and likely no one healthy microbiome at other sites as well.
"What our data shows is that just because a person's microbiome is different doesn't make it unhealthy," Schloss added. "It demonstrates there's more to learn about the factors that cause one's microbiome to change."
Further, knowing why communities change could help researchers understand related diseases and disease risk and the effect that treatments like probiotics, antibiotics, and fecal transplants may have.