Mapping the Microbiome: Vaginal Bacteria and HIV Risk

The list of things that increase a person's risk for HIV acquisition are pretty well known: poverty, lack of access to health care, condomless sex with someone who doesn't know they have HIV. But vaginal bacteria?

Some bacteria might also increase risk, at least according to evidence presented at last month's International AIDS Conference (AIDS 2016) in Durban, South Africa. In two separate presentations by two different teams, researchers showed that half a dozen bacteria that are more prevalent among African women might increase HIV susceptibility. But how to remove those bacteria is more complicated, and researchers don't yet have a good solution.

"Some women are going to acquire HIV irrespective of their vaginal microbiomes," said R. Scott McClelland, M.D., M.P.H., associate director of the University of Washington's Center for AIDS Research, who presented one of the studies at AIDS 2016. "But I really think that the vaginal microbiota does drive the vaginal immune milieu, and I think it does influence susceptibility to HIV."

A 20-Year-Old Hypothesis

The vaginal microbiome is a "good news/bad news" kind of situation when it comes to HIV.

On the one hand, researchers have known for years that women with a microbiome dominated by one specific bacterial type, Lactobacillus, had lower rates of HIV. (Details about that are available here.) They've also known that women with HIV-protective microbiomes tend to be white and Asian.

On the other hand, women without Lactobacillusdominant microbiomes -- women with a great diversity of bacterial types, a condition doctors generally call bacterial vaginosis (BV) -- tend to be African American, Latinx or live in sub-Saharan Africa. In other words, BV is more common among people and in areas with the highest HIV rates.

But it's not just geography. BV is itself strongly associated with HIV acquisition, presumably because it increases inflammation in the vagina, which attracts more immune cells to the vaginal surface -- cells that HIV then hijacks. A meta-analysis published in AIDS in 2008 found that women with BV were 60% more likely to have HIV.

This has been a problem for researchers because diverse microbiomes are difficult to dislodge.

"[BV] has been mysterious," said Richard Cone, Ph.D., a biophysicist who has run a lab at Johns Hopkins University for years, looking for ways to harness the vaginal microbiome to protect women from HIV. "We don't know what starts it. We don't know how to help women get over it. It's recurrent."

Researchers have been looking at the vagina for about two decades, trying to decipher whether it's just BV -- that is, lack of Lactobacillus -- that increases women's HIV risk, or whether it's something more specific. So, when McClelland and the Center for the AIDS Programme of Research in South Africa (CAPRISA) discussed their findings at AIDS 2016 in July, their presentations were 20 years in the making.

CAPRISA Findings

While there were two presentations at AIDS 2016, only one got spotlight status. AIDS 2016 convened a special symposium, chaired by U.S. Ambassador Deborah Birx and National Institute of Allergy and Infectious Disease Director Anthony Fauci, M.D., to present data from CAPRISA.

That study found that, while women have a variety of bacteria in their microbiomes, one bacteria, Prevotella bivia, was associated with a 13-fold increased risk of HIV. But the results represented a small sample -- 11% of 119 women who had participated in CAPRISA's trials of an HIV-prevention gel. And, because it was a symposium, the methodologies weren't available for conference attendees to analyze. A question-and-answer session was planned for the end of the symposium, which presented data on three interlinked studies on women's HIV risk, but time constraints made that impossible, said Brent L. Williams, Ph.D., an assistant professor of clinical pathology and cell biology at Columbia University.

Williams conducted the bacterial sequencing and analysis for the CAPRISA study. In an email, he explained that the analysis was based on broad-range 16S rRNA sequencing, which can analyze the abundance of a specific bacteria in a woman's microbiome, relative to the other bacteria in the vagina. That is, the study didn't just check that the bacteria existed at all; it also ascertained the relative amount of P. bivia in the microbiome compared with other bacteria present.

That was followed by shotgun metagenomic sequencing in a subset of samples to double check the 16S sequencing's findings, as well as a quantitative PCR using primers specific to P. bivia, to ascertain the absolute number of P. bivia copies in each woman's microbiome.

"Both the 16S rRNA sequencing and quantitative PCR with an assay specific for Prevotella bivia revealed the strong association with inflammation and HIV acquisition," Williams stated. "This layered approach, using a combination of methodologies, while laborious, provides us as scientists with statistical rigor required to fully validate the findings."

Williams said he'd been working on it for three years.

University of Washington Findings

But CAPRISA's wasn't the only vaginal bacteria analysis presented at AIDS 2016. McClelland presented a poster of his team's findings -- which mapped the Prevotella genus, but not P. bivia.

Instead, the University of Washington found five other bacteria associated with statistically significant increases of HIV incidence in their cohort of people with vaginas (the studies were all conducted on non-trans women). Those bacteria were Eggerthella species type 1, Gemella asaccharolytica, Leptotrichia/Sneathia, Megasphaera, and Mycoplasma hominis.

The results were based on vaginal samples from 349 women, 87 of whom were living with HIV. Like the CAPRISA group, the University of Washington employed broad range 16S rRNA sequencing. In addition, the team analyzed both the relative abundance of a bacteria in the microbiome -- that is, what percentage of the total community was occupied by a particular strain -- as well as the absolute number of a particular bacteria present. Then the team did a quantitative PCR probe to zero in on the five bacteria with the highest association with HIV.

Differentiating between relative and absolute bacterial copies is important because the sheer volume of bacteria in a microbiome varies by woman, too, said McClelland.

"We saw [in a recent JAIDS paper] a 100-fold difference in overall vaginal bacteria concentration," he said. "A relative abundance of 1% in one woman might be 100 times more of that bacteria than in another woman, who has the exact same relative abundance."

And that concentration may be important to HIV risk, said McClelland -- though he said it's too soon yet to tell. University of Washington's data showed an association between abundance of bacteria and HIV susceptibility: Gemella asaccharolytica was most associated with HIV at the highest concentrations, but not at lower concentrations. Megasphaera, meanwhile, was associated with a three-fold increase of HIV -- but only at the smallest concentrations. Eggerthella species 1 had its highest association with HIV when present in mid-range quantities, but the association dropped off at either higher or lower concentrations.

The challenge now is to delve more deeply into the bacterial species and do more research to confirm and replicate the findings.

"What I'd love to know is if a high concentration of Gemella, for instance, is associated with having a low concentration of Megasphaera," said McClelland. "This might suggest that a particular bacterial community with both of these species in particular concentrations is associated with higher risk for HIV."

The findings were gratifying for McClelland, who said that the study was a risk. It could have turned out that any BV-associated species would increase HIV risk. And then we'd just be back at what researchers knew in 1998.

"I would have been disappointed had we simply ended up in the same old place --that there aren't individual taxa driving this, but that it's just all of them," he said. "I was delighted -- I would not say surprised, but delighted -- that, in fact, the data emerged that does support the hypothesis I first wrote into a grant nine years ago, that there may be key bacterial species associated with HIV acquisition."

Making Progress for Women

The lack of agreement between the two studies about which bacteria increase risk didn't trouble Douglas Kwon, M.D., Ph.D., who heads a research team on the vaginal microbiome and inflammation at Harvard Medical School. He called the disparities an example of the scientific method at work. All the results will need to be replicated around the world, anyway.

But he did think the results, as varied as they are, represented progress.

"For a while, we linked specific sexually transmitted infections [STIs] to increased HIV acquisition risk, and then more recently to baseline genital inflammation," Kwon said. "From the research presented, the emerging idea is that the vaginal microbiome increases HIV risk -- not just STIs, not just specific infections, but actually the community of microorganisms that live in the genital tract."

If that turns out to be true, he said, "that means attempts to manipulate the microbiome would be useful for reducing HIV risk in women in sub-Saharan Africa."

(Read more about ways researcher are attempting to manipulate the microbiome to boost women's natural HIV prevention here.)

But that's a big "if." Even McClelland said there's so much more to learn about this part of the body that we have to go back to the beginning, to basic science, to figure out what these results mean.

"These are observational studies, and so we have to ask the question: What are the other possible explanations for these associations?" McClelland said. "For example, in young women becoming sexually active, could these [bacteria] simply be a marker for unprotected sex, or unprotected sex with an uncircumcised, HIV-positive male partner? HIV may change the man's microbiome, too. Bacteria that are under the foreskin are part of the penile microbiome. From there, they may be passed into the vaginal microbiome of their partners."

As researchers move forward, he said, "it's important to be circumspect."