Meeting of the Minds on Mucosal Transmission

Researchers underscore the possible role of non-neutralizing antibodies in protection against HIV and discuss ways to optimize animal models to study mucosal transmission

By Andreas von Bubnoff

HIV is most often spread through sexual transmission and therefore primarily enters the body through mucosal surfaces at the genitals or the rectum. At a recent meeting on “New Insights into Mucosal Transmission of HIV/SIV and its Prevention by Vaccines and other Modalities,” which was held at the US National Institutes of Health June 3-4, researchers discussed efforts to better understand mucosal transmission of HIV in clinical trials, non-human primate models, and in vitro studies.

Now that there is the first evidence of vaccine-induced protection against HIV, many researchers are focused on trying to find an explanation for these results. RV144, an HIV vaccine efficacy trial conducted in Thailand involving more than 16,000 volunteers, showed that a canarypox vector-based candidate, ALVAC-HIV, administered in a prime-boost combination with an engineered HIV gp120 protein, AIDSVAX B/E, provided a modest 31% protection against HIV infection. While the candidate vaccine regimen used in RV144 does not seem to induce much of a neutralizing antibody response, it did induce antibodies capable of binding to gp120 in the majority of vaccinees. Some researchers believe it is possible that non-neutralizing, gp120-binding antibodies may have contributed to the modest protection, perhaps by inhibiting HIV at one or several points along its entry path through mucosal tissues.

At the meeting, Barton Haynes, a professor of medicine and immunology at Duke University Medical Center who chairs the scientific steering committee for RV144 analysis and follow-up studies, presented plans designed to test this hypothesis. In the absence of mucosal samples from RV144 vaccinees, researchers plan to conduct assays that measure whether antibodies taken from blood samples from RV144 vaccinees who remained HIV uninfected can inhibit HIV on its way into the body through the mucosa better than antibodies from vaccinees who were infected with HIV during the trial. “The assays are designed to look at every stage in the initial mucosal transmission event to make sure that we have assays that could pick up blocking at each one of those stages before traditional neutralization,” Haynes said.

For example, some assays will measure the ability of the vaccine-induced antibodies to slow the movement of virions through the mucus that covers the mucosal surfaces of the female genital tract. Other assays will measure whether these antibodies can inhibit transcytosis, which is one way HIV is thought to cross the epithelial layer of the mucosa. Inhibition of transcytosis can be tested in vitro using a cultured epithelial layer. Other assays will test for other antibody activities, including antibody-dependent cellular cytotoxicity (ADCC) and antibody-dependent cell-mediated virus inhibition (ADCVI; see Antibodies: Beyond Neutralization, IAVI Report, Jan.-Feb. 2010).

Before analyzing the most valuable samples from the 51 vaccinated volunteers in RV144 who subsequently became HIV infected, researchers from many different labs will test different assays in a pilot phase that will help them decide which are the best and most reliable assays to use. Haynes said the pilot phase has been delayed because of the political unrest in Thailand. “Hopefully it’s abating,” he said.

The assays that give the most solid and consistent results will then be used to compare, in a case-control study, the antibody activities of HIV-infected and uninfected vaccinees, Haynes said, adding that he expects to have first results in the fall. “Hopefully we will have something to say by the time of the vaccine meetings in October,” Haynes said. “That’s pushing it, but that’s the current timeline.”

Haynes said his lab will also test if antibodies in the serum of the RV144 vaccinees that specifically bind the Env proteins contained in AIDSVAX B/E, the protein boost used in RV144, can protect rhesus macaques in passive infusion experiments. To determine this, researchers will use specific probes with the exact Env variants used in AIDSVAX B/E to isolate memory B cells from the serum of RV144 vaccinees. They will then use the antibody genes in these cells to synthesize the antibodies, and then infuse the antibodies into rhesus macaques. The macaques will then be challenged with an R5-tropic simian immunodeficiency virus (SIV)/HIV hybrid known as SHIV that is recognized by the antibody, if one is available.

The role non-neutralizing antibody activities may play in protection against HIV was also discussed in other talks at the meeting. Marjorie Robert-Guroff, the chief of the immune biology of retroviral infection section of the vaccine branch at the National Cancer Institute (NCI), presented evidence that non-neutralizing antibodies do play a role in protection in rhesus macaque experiments. Guroff and colleagues recently reevaluated the serum from animals that had been primed twice with a replicating adenovirus serotype 5 vector expressing SIVmac239 gag and nefand HIV gp140 env, followed by two gp140 Env protein boosts. When the animals were challenged with SHIV89.6P that expressed the same genes contained in the vaccine regimen, they showed reduced acute and chronic viral load even though they did not have neutralizing antibodies until four weeks after challenge (1). When Guroff and colleagues explored whether non-neutralizing antibody activity could account for the reduction in viral load, they found that ADCC and ADCVI mediated by serum antibodies from these animals correlated with the observed reduction of acute and chronic viral load in these animals.

For the first time, they also showed that inhibition of transcytosis by antibodies in rectal secretions correlated with reduced chronic viral load, Guroff said (2). “The hypothesis [is] that non-neutralizing antibody contributed to the protection seen in the RV144 clinical trial,” said Guroff, who is one of the researchers involved in measuring ADCC activity in the Thai trial samples. “[The] data from our nonhuman primate studies support a role for non-neutralizing antibodies in protection.”

Another possible mechanism that could help explain the protection in RV144 is that antibodies might inhibit HIV movement in vaginal mucus. Tom Hope, a professor of cell and molecular biology at Northwestern University, presented evidence for this. He isolated cervicovaginal mucus from women and measured HIV movement with and without anti-HIV antibodies present. He found that both non-neutralizing Env-binding antibodies and broadly neutralizing antibodies such as b12 slowed HIV movement in the mucus. HIV movement was also slower in mucus from HIV-infected women, suggesting HIV-specific antibodies were present in their mucus. “We think this could be another mechanism through which antibodies binding to virus could influence transmission,” Hope said.


Fighting Viruses with Bacteria  

One topic covered at the recent meeting on “New Insights into Mucosal Transmission of HIV/SIV and its Prevention by Vaccines and other Modalities,” which was held at the US National Institutes of Health from June 3-4, was the development of novel approaches to prevent mucosal HIV transmission in humans. Laurel Lagenaur, a senior scientist at the California-based company Osel Inc., presented an update of efforts to develop a live vaginal protein-based microbicide by introducing the gene for Cyanovirin into Lactobacilli, bacteria that normally live in the vaginal mucosa (see Mucosal Vaccines: Insights from Different Fields, IAVI Report, Nov.-Dec. 2008).

Cyanovirin is a protein that binds to HIV gp120 with a very high affinity. Lagenaur showed that in rhesus macaques, the Lactobacilli expressed Cyanovirin protein for up to six weeks after the animals had been inoculated with the transgenic bacteria.

To see if the transgenic bacteria could protect from challenge, the researchers inoculated macaques with the bacteria every week. Each weekly inoculation was followed 24 hours later by a low-dose vaginal challenge with a simian immunodeficiency virus (SIV)/HIV hybrid, known as SHIV. More weekly inoculation/challenge cycles were necessary to infect the animals with the Cyanovirin-expressing Lactobacilli than control animals. According to Lagenaur, this translated into a 62% reduction of the SHIV transmission rate in the animals with the Cyanovirin-expressing Lactobacilli. “This is the first successful demonstration of a live microbicide,” said Lagenaur. She thinks it might also be interesting to express some of the recently isolated broadly neutralizing antibodies, such as PG16 or VRC01 (see Raft of Results Energizes Researchers, IAVI Report, Sep.-Oct. 2009), in the Lactobacilli in future experiments. —AvB



Refining animal models

Other work related to mucosal transmission involves fine-tuning animal models. In humans, most productive clinical HIV infections can be traced back to a single transmitted founder virus (see HIV Transmission: The Genetic Bottleneck, IAVI Report, Nov.-Dec. 2008). And, ideally, researchers would want to be able to reproduce these same transmission dynamics in nonhuman primate models so that they could mimic the human situation as closely as possible.

Brandon Keele, a senior scientist at NCI and the Science Applications International Corporation in Frederick, Maryland (SAIC-Frederick/NCI), analyzed the number and sequence of SIV variants that establish productive infection after infecting rhesus macaques intra-rectally with SIVmac251 stocks grown by different labs. He found that diluting the same 251 challenge stock led to a reduced number of transmitted founder viruses, eventually resulting in animals that were infected by a single SIV variant.

Guroff has used a repeat low-dose rectal challenge with a 1:500 dilution of an SIVmac251 stock that was grown in the lab of Ronald Desrosiers at the New England Primate Research Center at Harvard Medical School and provided by the Division of AIDS at NIAID. In this infection model, up to nine challenges were needed to infect all naive macaques, and Keele found that eight of nine animals became infected by a single viral variant (the ninth was infected with two variants). “We think that’s a pretty reasonable approach to recapitulate the single variant infection we find in humans,” Keele said.

Genoveffa Franchini, chief of the animal models and retroviral vaccine section at NCI, said she plans to use Guroff’s intra-rectal challenge regimen in her efforts to develop an animal model that could recapitulate the results of RV144, when the vaccine regimen is tested in rhesus macaques. “[So far], we have used a dose that gives us a lot of variants,” Franchini said. “Now we want to use a dose that gives us one variant.”

Meanwhile, Keele is also studying other transmission routes. He finds intra-vaginal infection to be more variable than intra-rectal infection in terms of the number of transmitted founder viruses after infection with the same dose. In collaboration with Jake Estes, a senior scientist at SAIC-Frederick/NCI, Keele is now studying the location and sequence of individual SIV-infected cells in rhesus macaques that were infected vaginally to better characterize where transmitted founder viruses first take hold in the animals. He showed that it is possible to infect an animal vaginally and then use a laser to isolate SIV-infected cells from histological sections of the cervix of the infected animal. The integrated proviral SIV from these infected cells can then be sequenced. Along with Chris Miller, a professor at the School of Veterinary Medicine at the University of California in Davis, Keele has also started to analyze the number of transmitted founder viruses in male rhesus macaques whose penises were exposed to SIVmac251.

1. Virology 374, 322, 2008
2.J. Virol. 84, 7161, 2010