Science Spotlight

HIV superinfection may inform vaccine design

From the Overbaugh Lab, Human Biology Division

June 19, 2017

It is paramount that scientists make ethical choices, especially when it comes to medical research. This has resulted in a timeless question for researchers – how do we understand human biology and disease without harming humans? The answer to this question has taken countless forms. The basics of human genetics were characterized in fruit flies and stickleback fish, and work in yeast uncovered many features of the cell cycle, while many disease treatments were identified using human cells grown in dishes and mice. Over the decades scientists have demonstrated creativity and ingenuity to answer biological questions; however, for some studies surrogate systems are exceedingly difficult. In the case of HIV it only infects humans; it does not even infecte primates such as macaques. Much of the viral biology has been characterized safely in cell culture models, but these systems cannot recreate an organism’s immune response. This is one of the many reasons developing an HIV vaccine is exceedingly difficult. Researchers in Dr. Julie Overbaugh’s Laboratory (Human Biology Division) are leveraging real world situations to understand immune response to HIV infection. A recent study led by Dr. Keshet Ronen analyzed patient samples from a cohort of HIV infected Kenyan women. These women were infected with a single HIV variant, but were at high risk of exposure to other HIV variants from different sexual partners. In a recent eBioMedicine publication, these scientists analyzed whether the initial infection mobilized components of the immune system to protect against infection with a secondary strain (superinfection).

Dr. Ronen and colleagues utilized samples from a previous study population called the Mombasa Cohort. In the late 1990s 146 HIV-1 infected women were enrolled in this substudy, which is part of a larger 25-year study of women at high risk of infection by HIV. This was prior to antiretroviral therapy and thus all plasma and vaginal swab samples collected were therapy naïve. Within this group there were 21 cases of superinfection, where two distinct viral genomes were sequenced. These 21 cases are the largest population of superinfection analyzed to date as previous studies were composed of 3-12 cases. Dr. Ronen had previously found that superinfection occurred at a lower rater than initial infection suggesting a protective feature. For this study she and colleagues, including co-first author, Adam Dingens, compared the 21 superinfected individuals with 63 matched singly infected individuals from the same cohort looking for differences in the immune response between women who acquired superinfection and those who remained singly infected despite ongoing risk behavior. The importance of this analysis being that if a unique immune feature prevents superinfection then a potent vaccine should elicit a similar immune response.

This study analyzed humoral immune response in superinfected individuals; this facet of human immunity relies on antibodies that are secreted by B cells and circulate in extracellular fluid. To develop HIV immunity these circulating antibodies would latch onto the virus either preventing it from infecting new cells, and/or recruiting other immune cells to destroy the viral particles. While this is only one aspect of the immune system it is key to preventative medicine, “Humoral immunity is the primary correlate of protection in the majority of licensed vaccines, though cellular immunity is often generated in addition” said Dr. Ronen.

Individuals infected with a single HIV strain (blue) are less likely to be infected with a second HIV variant (red). Four different measures of antibody-based humoral immune response showed no difference between individuals who did not acquire a second infection and those who did, suggesting an alternative feature of the immune system provides protection against superinfection.

Figure provided by Dr. Keshet Ronen

The first aspect of immune response measured was the potency of neutralizing antibodies to prevent viral infection in both superinfected and singly infected individuals. Researchers measured the ability of a patient plasma sample (and thus antibodies) to inhibit the replication of four different HIV variants. Measuring activity against multiple viruses is important because the HIV genome is constantly mutating and often allowing the virus to evade the immune system. When superinfected cases were compared to singly infected ones, there was no association between superinfection status and either the potency or the breadth of the response to different variants. Another feature of the humoral immune response is for antibodies to bind infected cells allowing other immune components to destroy them. Antibody-dependent cellular cytotoxicity (ADCC) is measured by mixing HIV infected cells with patient antibody samples and peripheral blood mononuclear cells from HIV negative donors. Cell death was measured by the loss of a cytoplasmic dye that only exits the cell if the membrane has been punctured. Dr. Ronen and colleagues tested ADCC in both plasma and cervical samples, and as before found no association between superinfection and ADCC.

It is possible the superinfected and singly infected samples behaved similarly in these cell culture models, but in a full immune context would react differently. Thus researchers measured the most basic antibody activity – binding to a target protein. HIV membrane buds from host cells, thus its major distinguishing factor is the presence of its envelope protein, encoded by the env gene. Previous work with vaccine development and superinfected individuals has indicated that antibodies targeting the V1V2 region of the gp140 envelope protein are associated with protection from infection. However, in this study the single and superinfected samples behaved similarly, both in how many samples demonstrated any binding, and in the relative binding affinities.

While epidemiological studies demonstrated a protection from superinfection it appears this is not due to humoral immune activities. The Overbaugh Lab is working to understand which other immune mechanisms could provide this protection, still with vaccine development in mind, “Currently we are developing studies in collaboration with Catherine Blish at Stanford University to evaluate whether NK cell phenotype and innate activity is associated with superinfection acquisition. We are also interested in the impact of superinfection on immune responses. Our group has previously found that superinfected women in the Mombasa Cohort have broader and more potent neutralizing antibody activity than singly infected women and we are working on isolating antibodies from these women to understand how such breadth and potency develops” said Dr. Ronen.


Ronen K, Dingens AS, Graham SM, Jaoko W, Mandaliya K, McClelland RS, Overbaugh J. 2017. Comprehensive Characterization of Humoral Correlates of Human Immunodeficiency Virus 1 Superinfection Acquisition in High-risk Kenyan Women. EBioMedicine, 18, 216-224.


Funding for this research was provided by the National Institutes of Health.