The molecular clock of HIV-1 unveiled through analysis of a known transmission history

Detailed knowledge about the rate and mode of the genetic variation is vital for understanding how HIV-1 induces disease and develops resistance as well as for studies on the molecular epidemiology and origin of the virus. To unveil the molecular clock of HIV-1 we analyzed a unique set of viruses from a known transmission history with separation times between samples of up to 25 years. The env V3 and p17gag regions of the genome were sequenced, and genetic distances were estimated by using the true tree and a nucleotide substitution model based on a general reversible Markov process with a gamma distribution to account for differences in substitution rates among sites. Linear regression analysis showed that separation times were significantly correlated with synonymous as well as nonsynonymous nucleotide distances in both V3 and p17, giving strong support for the existence of a molecular clock. The estimated rate of nucleotide substitution was 6.7 +/- 2.1 x 10(-3) substitutions/site per year in V3 and 2.7 +/- 0.5 x 10(-3) in p17. Importantly, the regression analyses showed that there was a significant genetic distance at zero divergence times. This pretransmission interval exists because the ramifications in the phylogenetic trees do not correspond to time of transmission, but rather to the coalescence time of the most recent common ancestor of the viruses carried by the transmitter and the recipient. Simulation experiments showed that neither the V3 nor the p17 clocks were overdispersed, which indicates that the introduction of nucleotide substitutions can be described adequately by a simple stochastic Poisson process.