HIV-infected vaginal cells do not transmit HIV if plasma viral load is undetectable, researchers find

A group of researchers have cleared up an important question about HIV transmission, in experiments on mice. Although HIV-infected CD4 cells persist in the vagina even on antiretroviral therapy (ART) that fully suppresses free HIV in the blood and body fluids, these cells are not anything like numerous enough to pose any transmission threat.  

Two other studies that looked at genital changes after HIV infection have been published recently

In one, a study of the mucosal immune response in the vagina in HIV-positive women has found that some women develop broadly-neutralising antibodies, and an antibody profile similar to one already key to vaccine efficacy.

In the other, a study of the sperm of HIV-positive men has found that, although like HIV-negative men’s sperm, it contains proteins that enhance HIV infection, it also contains competing proteins that mitigate that effect.

HIV-infected vaginal cells in women on antiretroviral therapy

One of the issues that HIV prevention studies have to deal with is that it is easier to study HIV immune responses in the blood than in the actual sites of most HIV infection, the epithelial mucosa in the genitals. As a result our knowledge of how the mucosal immune system responds both to infection and to drug treatment is still quite deficient; this impacts on our ability to forecast the success of biomedical prevention methods like oral pre-exposure prophylaxis (PrEP), microbicides and vaccines.   

In women, oral PrEP has appeared somewhat less effective than in men and although most of this appears to be due to poor adherence, there remain doubts that lower drug levels in the vagina, and the fact that it appears more common for HIV viral shedding to persist in the vagina under fully-suppressive ART than in the sperm, may both contribute to lower efficacy. It has been theorised that in situations where free HIV virus has been suppressed, intracellular HIV remains and that direct cell-to-cell transmission could pose a residual risk.

There have been human studies that suggest vaginal shedding of HIV during fully-suppressive ART is unlikely ever to reach infectious levels, but to investigate the issue on a cellular level, researchers from the University of North Carolina, USA and Aarhus University in Denmark investigated free and cell-associated virus in female mice that were given human immune-system transplants so they could respond to HIV, infected and then given ART.

The mice responded to HIV therapy very much like humans though with somewhat higher off-treatment viral load (about one million copies/ml in blood and 100,000 copies/ml in rectal fluid and 50,000 copies/ml in vaginal fluid). Their absolute CD4 cell count, after an initial ‘spike’ following infection, did not then ‘crash’ but the CD4 percentage – the proportion of T-lymphocytes that were CD4 calls, fell gradually in the blood but much more profoundly in rectal fluid, saliva and, especially, vaginal fluid. This was due to a vast overproliferation of CD8 cells in response to HIV infection. This supports many previous observations that the immune disorder of HIV infection is initially characterised by the vast over-stimulation HIV imposes on the immune system. The researchers were able to determine that these cells were capable of direct cell-to-cell contact and of passing on infection.

When mice were given ART, free virus in vaginal fluids fell rapidly and was undetectable (below 50 copies/ml) within as little as two weeks after starting. The CD4 percentage rapidly renormalised too. However while HIV viral RNA within cells did fall, it did not become undetectable and fell to a lesser extent in cervico-vaginal secretions than in blood. In blood plasma the intracellular HIV RNA count was about 80,000 copies per million cells before ART and about 30 copies per million cells after. In cervico-vaginal secretions the post-ART viral load was similar, but the pre-ART viral load was only 6500 copies per million cells and this therefore represents a smaller proportional decrease. And in the female reproductive tract overall, there were about 80 copies/million cells of cell-associated HIV RNA.

Would this be enough to pose an infection threat? The researchers first used a cell assay where vaginal cells in the lab dish were exposed to CD3 cells from blood or vaginal fluid from treated mice. One week after ART initiation, there were still nearly as many HIV-infected cells capable of infection in ART-treated mice as in ones not on ART. However by two weeks after initiation, no cellular infections using either cells from the blood or from vaginal fluid were seen.

The researchers then exposed HIV-negative female mice vaginally to two different doses of T-lymphocyte (CD3) cells, which include CD4 and CD8 cells. They gave them two different doses of infected dells – 5000 individual cells and 10,000. None of the five mice given 5,000 cells became infected but two out of five given 10,000 cells did. Five thousand infected cells is well in excess of the number of HIV-infected cells seen in credible doses of either blood or vaginal fluid.

“Our analysis shows that the residual levels of HIV-RNA cells present in mice receiving ART were too low to transmit HIV in vitro”, the researchers conclude.

Broadly neutralising antibodies in vaginal fluids

Two other studies documented how the mucosal and seminal immune responses develop in response to HIV infection, and in both cases researchers found useful changes in response to HIV that could possibly be amplified by a vaccine or mimicked by a drug.

In the first, researchers studied immune responses in 13 women who became infected with HIV during Caprisa 002, a study in South Africa designed to establish HIV incidence rates and suggest the right sites and population for the subsequent Caprisa 004 microbicide study.

The researchers found that, within only a year or two of infection, seven out of the 13 women (54%) developed so-called broadly neutralising antibodies in their vaginal fluids that could prevent infection by a couple of different viruses. One promising finding was that the antibodies that women developed vaginally were of the ‘immunoglobulin G’ (IgG) type, which has been shown to be protective against HIV infection, rather than the IgA type, which actually enhances it. This is the opposite of what had been expected as IgA is more usually associated with mucosal immune responses.

Secondly, two of the women (15%) developed a specific subtype called IgG3, which has been shown to be the crucial protective factor in the RV144 vaccine study – the only efficacy study that has yet produced a positive result.

In the women these broadly neutralising antibodies are not induced at potent enough levels to have any effect on their own HIV infection but the results suggest that a systemically-administered vaccine can produce a protective response within the vaginal mucosa, and there might be no need for topical ones.

Infection enhancement and inhibition in semen

In a third study, researchers from the French national research institute INSERM compared the degree to which semen from HIV-negative and HIV-positive men enhanced the infectivity of HIV.

It has been known since 2007 that semen naturally enhances HIV infection. Semen contains proteins called SEVIs (Semen-derived Enhancers of Virus Infection), sticky strands that HIV clings to so that is gets delivered efficiently to the dendritic cells that ferry the virus away to the CD4 cells it, in the main, infects. However all the studies that established this cultured cells with HIV virus in the presence of semen samples from HIV-negative men when, obviously, it is semen from HIV-positive men that infects people.  

The French researchers co-cultured CD4 cells with HIV in lab dishes with no semen, semen from HIV-negative men, and semen from HIV-positive men.  

They confirmed that semen is a powerful enhancer of infection. Semen from negative men increased the number of cells infected by anything from 70 to 200%. Semen from HIV positive men did so too, but only by 50% compared to no semen.

The researchers investigated if there were any factors in the semen’s ability to enhance infection and found only one – the cytokine (immune-signalling protein) RANTES, which has already been used as the progenitor of a class of anti-HIV drugs, the CCR5 entry inhibitors (notably maraviroc (Celsentri)).

Clearly there is not enough RANTES present in HIV-positive men’s semen to counteract the effect of SEVI proteins, let alone prevent infection. But understanding how the body’s immune response to HIV generates RANTES protein in semen could lead to the artificial generation of more potent RANTES analogues by genetic engineering or a vaccine.