Research Article

Selection bias at the heterosexual HIV-1 transmission bottleneck

Science  11 Jul 2014:
Vol. 345, Issue 6193, pp.
DOI: 10.1126/science.1254031

You are currently viewing the abstract.

View Full Text

Structured Abstract


Heterosexual HIV-1 transmission is an inefficient process with rates reported at <1% per unprotected sexual exposure. When transmission occurs, systemic infection is typically established by a single genetic variant, taken from the swarm of genetically distinct viruses circulating in the donor. Whether that founder virus represents a chance event or was systematically favored is unclear. Our work has tested a central hypothesis that founder virus selection is biased toward certain genetic characteristics.

Embedded Image

Fitter viruses (red) are favored more in woman-to-man (bottom curve) than in man-to-woman (top curve) transmission. The probability that a majority donor amino acid variant is transmitted is a function of relative fitness, here estimated by the frequency of the variant in the Zambian population. Even residues common in the population are less likely to be transmitted to healthy men than to women, indicative of higher selection bias in woman-to-man transmission.


If HIV-1 transmission involves selection for viruses with certain favorable characteristics, then such advantages should emerge as statistical biases when viewed across many viral loci in many transmitting partners. We therefore identified 137 Zambian heterosexual transmission pairs, for whom plasma samples were available for both the donor and recipient partner soon after transmission, and compared the viral sequences obtained from each partner to identify features that predicted whether the majority amino acid observed at any particular position in the donor was transmitted. We focused attention on two features: viral genetic characteristics that correlate with viral fitness, and clinical factors that influence transmission. Statistical modeling indicates that the former will be favored for transmission, while the latter will nullify this relative advantage.


We observed a highly significant selection bias that favors the transmission of amino acids associated with increased fitness. These features included the frequency of the amino acid in the study cohort, the relative advantage of the amino acid with respect to the stability of the protein, and features related to immune escape and compensation. This selection bias was reduced in couples with high risk of transmission. In particular, significantly less selection bias was observed in men with genital inflammation and in women (regardless of inflammation status), compared to healthy men, suggesting a more permissive environment in the female than male genital tract. Consistent with this observation, viruses transmitted to women were characterized by lower predicted fitness than those in men. The presence of amino acids favored during transmission predicted which individual virus within a donor was transmitted to their partner, while chronically infected individuals with viral populations characterized by a predominance of these amino acids were more likely to transmit to their partners.


These data highlight the clear selection biases that benefit fitter viruses during transmission in the context of a stochastic process. That such biases exist, and are tempered by certain risk factors, suggests that transmission is frequently characterized by many abortive transmission events in which some target cells are nonproductively infected. Moreover, for efficient transmission, some changes that favored survival in the transmitting partner are frequently discarded, resulting in overall slower evolution of HIV-1 in the population. Paradoxically, by increasing the selection bias at the transmission bottleneck, reduction of susceptibility may increase the expected fitness of breakthrough viruses that establish infection and may therefore worsen the prognosis for the newly infected partner. Conversely, preventive or therapeutic approaches that weaken the virus may reduce overall transmission rates via a mechanism that is independent from the quantity of circulating virus, and may therefore provide long-term benefits to the recipient if transmission does occur.

HIV needs to be fit to transmit

Although you might not think it, it's hard to catch HIV. Less than 1% of unprotected sexual exposures result in infection. What then leads to transmission? Carlson et al. determined the amino acid sequence of viruses infecting 137 Zambian heterosexual couples in which one partner infected the other (see the Perspective by Joseph and Swanstrom). The authors then used statistical modeling and found that transmitted viruses are typically the most evolutionarily fit. That is, compared to other viral variants in the infected person, the transmitted virus most closely matches the most common viral sequence found in the Zambian population.

Science, this issue 10.1126/science.1254031; see also p. 136


Heterosexual transmission of HIV-1 typically results in one genetic variant establishing systemic infection. We compared, for 137 linked transmission pairs, the amino acid sequences encoded by non-envelope genes of viruses in both partners and demonstrate a selection bias for transmission of residues that are predicted to confer increased in vivo fitness on viruses in the newly infected, immunologically naïve recipient. Although tempered by transmission risk factors, such as donor viral load, genital inflammation, and recipient gender, this selection bias provides an overall transmission advantage for viral quasispecies that are dominated by viruses with high in vivo fitness. Thus, preventative or therapeutic approaches that even marginally reduce viral fitness may lower the overall transmission rates and offer long-term benefits even upon successful transmission.

  • These authors contributed equally to this work.

  • Present address: Department of Pharmaceutics, Washington National Primate Research Center, University of Washington, Seattle, WA 98121, USA.

  • § Present address: Google Inc., Venice, CA 90291, USA.

  • Present address: Department of Mathematics and Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, MA 02139, USA.

View Full Text