Microbiome to the rescue

SAN FRANCISCO—One of the most common bacterial organisms found in the vagina has helped scientists at the University of California, San Francisco (USCF) reveal the extent to which the human microbe could serve as a rich source for antibiotic compounds. The researchers found that Lactobacillus gasseri, a common species in the microbe community of the vagina, produces a compound, lactocillin, which closely resembles an antibiotic compound that the pharmaceutical company Novartis is currently testing in a Phase 2 clinical trial. The discovery, along with other key finding of UCSF’s study, suggests there could be many other small molecules produced by microbes in or on the human body that possess therapeutic potential.

Dr. Mohamed Donia, who drafted the new study as a postdoctoral fellow at UCSF, tells DDNews that the ability of organisms like Lactobacillus gasseri to produce antibiotic compounds turns the typical drug development process on its head. “We’re used to developing drugs by discovering natural products or using synthetic chemistry, and then spending years modifying these compounds to achieve the best biological activity and the least toxicity,” says Donia, who is currently an assistant professor of molecular biology at Princeton University. “This particular compound has bypassed that whole process, and there could be many molecules in other parts of the human microbiota with the same potential.”

The UCSF study, which was published in Cell in September, found that lactocillan serves to kill common vaginal bacterial pathogens while sparing other bacteria known to dwell harmlessly in the vagina. “This bacteria actually produced the right drug in the right place, with the exact biological activity that is needed, and probably at the time,” says Donia.

While many medicines are derived from microbes and plants, few efforts have been made to use bacteria within the human body as a source for new drug molecules. But recent efforts to better understand the human mirobiome has created new opportunities to explore this therapeutic potential. Scientists have made significant progress toward mapping the bacterial ecosystems found in the gut, skin, nasal passages, mouth, vagina and other parts of the human body through research funded by the NIH’s Human Microbiome Project.

However, much less is known about the small molecules that govern interactions between microbes and their human hosts. A primary purpose of the UCSF study was to identify the biosynthetic gene clusters (BGCs) that contain the genetic blueprints for creating such molecules. “Small molecules are the common language that almost every living cell can understand, so we wanted to interrogate what small molecules are being produced by these bacteria and find out how they interact with their host and interact with other bacteria,” says Donia.

Researchers created a machine-learning algorithm that enabled a computer to identify known genes that produce small molecules with potential as drugs. When they used this algorithm to systematically analyze genes in the human microbiome, they identified 3,000 BGCs at different body sites. “We were surprised to find so many BGCs producing every small molecule type, and we were also surprised to find that so many of them were very common within a population of healthy humans,” says Donia. One of the compounds that researchers identified and then studied in greater detail was the molecule produced within the vaginal microbe community, lactocillin, which belongs to a class of antibiotics called thiopeptides.

The approach taken by the UCSF researchers differed in significant ways from that of other studies carried out under the umbrella of the Human Microbiome Project. “Most studies have been focused on continuing to sequence and document just which microbes are part of the microbiome,” says Donia. “We took a functional approach and tried to find out what these microbes are actually doing.”

The most common method of identifying bacteria residing in humans involves genus-level analysis, but UCSF researchers found that this method is not detailed enough to predict which drug-like molecules bacteria will produce. Individual species, and different strains within each species, produce different molecules. “We need to learn what these molecules are and what they are doing,” said Dr. Michael Fischbach, an assistant professor of bioengineering at UCSF and senior author of the study. “This could represent a pool of molecules with many tantalizing candidates for drug therapy.”