Microbes behind the malodor: cross-feeding in bacterial vaginosis

By the end of their teens, many women will have learned—often the hard way—that the bacterial communities inhabiting the vagina are critical regulators of vaginal health. Disrupting the natural balance of these microbes can lead to a range of uncomfortable (and often stigmatized) conditions. Maybe it’s a yeast infection after too much time in a damp swimsuit, or your first urinary tract infection because no one told you to pee after sex.

In its optimal state, the vaginal microbiome is dominated by various species of Lactobacillus —the same type of bacteria that gives yogurt, kimchi, and miso their tang. These beneficial bacteria are often called commensals, and by filling the ecological niche, they help keep harmful microbes at bay—like choosing to sit next to a friendly face on the bus so no one else can take the seat.

But sometimes, despite our best efforts, the “bad” bacteria gain a foothold. This is the case in bacterial vaginosis (BV), a condition that affects more than a quarter of women in the U.S. and is marked by increased discharge and a distinct, unpleasant odor. BV isn’t the result of a single invading microbe—it emerges from a complex microbial power shift, where groups of bacteria interact not only with each other but also with host cells in ways that enable the metabolic changes responsible for its hallmark symptoms.

“Although BV has been studied for decades,” says Elliot Lee, a postdoctoral fellow in the Fredricks lab, “we still know relatively little about which species are performing particular functions, how they cause characteristic symptoms like malodor and thin vaginal discharge, and how they interact.”

Because there are so many players involved, studying this condition requires a broad approach that captures the microbial communities, the host cells, and the metabolic consequences. A new study published in The ISME Journal from Dr. David Fredricks group in collaboration with the Pacific Northwest National Laboratory has done exactly that, using untargeted metaproteomics to investigate samples from 9 healthy patients and 20 patients diagnosed with BV.

“We found evidence that Dialister micraerophilus is responsible for producing polyamine compounds like putrescine and cadaverine that contribute to malodor,” says Lee, “but this bacterium relies on precursor compounds made by other organisms such as Fannyhessea vaginae to make putrescine in high quantities.” Putrescine and cadaverine come by their names from the strong and distinctive odors they have.  In microbial communities like the vaginal microbiome, this phenomenon—where one species produces a compound that another species uses as a building block—is described as "cross-feeding". This kind of syntrophic relationship—where one microbe consumes another’s waste products—helps explain why BV isn’t caused by a single pathogen, but rather a disbalanced community working together in ways that harm the host.

The Fredricks lab also found that many BV-associated bacteria break down vaginal glycogen and ferment the resulting carbohydrates into small molecules like formic acid, acetate, succinate, and ethanol. A key enzyme in this process—pyruvate formate lyase (PFL)—was especially widespread among BV-associated bacteria but absent from beneficial Lactobacillus species. Lab experiments confirmed that a dozen BV-associated species could produce formic acid, whereas Lactobacillus species and some other bacteria could not. Interestingly, when formic acid production was blocked using an inhibitor called hypophosphite, the growth of formate-producing bacteria was significantly reduced, suggesting they depend on this pathway for survival.

“We also discovered that a yet-unnamed bacterium called DNF00809 isolated from BV patients consumes formic acid produced by other bacteria, buffering the environment and potentially protecting the community from being overwhelmed by acid,” says Lee.

Beyond reshaping the microbial community, BV also disrupts key metabolic processes and triggers noticeable shifts in host responses. On the microbial side, nitrogen metabolism appears to play an important role. The researchers identified extracellular proteases from Gardnerella and Prevotella, which likely contribute to tissue breakdown by degrading host proteins. While many BV-associated bacteria produced glutamate dehydrogenase—an enzyme involved in amino acid metabolism—Gardnerella lacked this enzyme entirely. Instead, it appears to rely on importing glutamate produced by other microbes, as evidenced by its expression of glutamate transporters and glutamine synthetase.

The team also identified new microbial culprits behind BV’s most infamous symptom: the distinct odor. They identified the enzyme betaine reductase (GrdH), which produces trimethylamine (TMA), in bacteria like Parvimonas micra and Finegoldia magna. Other pathways for TMA production were less common, reinforcing the idea that TMA is mainly generated from betaine by a small subset of BV-associated bacteria.

These microbial shifts are mirrored by changes in host protein expression. Using metaproteomics, the researchers found that proteins involved in maintaining epithelial structure were significantly less abundant in BV+ samples, suggesting that the protective lining of the vaginal epithelium may be compromised during infection. One unexpected finding was that transglutaminase 3 (TGase3), a protein involved in epithelial repair, was significantly more abundant in BV+ samples—an observation later validated in a separate patient cohort which suggests a host response aimed at repairing the damaged vaginal tissue.

These findings were made possible through a close collaboration between the Fredricks lab and scientists at the Pacific Northwest National Laboratory (PNNL), whose expertise helped drive the study’s metaproteomic depth. “PNNL has been on the cutting edge of proteomics work and has a national lab for using this technology to study microbes,” says Fredricks. “We collaborated with Brooke Kaiser, who was previously at UW in lab medicine and microbiology and is now a scientist at PNNL, along with several members of her team.” By pairing clinical samples with advanced analytical tools, the team was able to capture not just the identities of microbes present, but what they were doing—and how they were influencing both each other and their human host.

Fred Hutch/University of Washington/Seattle Children’s Cancer Consortium members Drs. David Fredricks and Daniel Raftery contributed to this research.

The spotlighted research was funded by the National Institutes of Health and the W.R. Wiley Environmental Molecular Science Laboratory.

Lee EM, Srinivasan S, Purvine SO, Fiedler TL, Leiser OP, Proll SC, Minot SS, Djukovic D, Raftery D, Johnston C, Fredricks DN, Deatherage Kaiser BL. 2025. Syntrophic bacterial and host–microbe interactions in bacterial vaginosis. The ESME Journal. https://doi.org/10.1093/ismejo/wraf055.