Genetic analysis reveals the complex structure of HIV-1 transmission within defined risk groups

Abstract

We explored the epidemic history of HIV-1 subtype B in the United Kingdom by using statistical methods that infer the population history of pathogens from sampled gene sequence data. Phylogenetic analysis of HIV-1 pol gene sequences from Britain showed at least six large transmission chains, indicating a genetically variable, but epidemiologically homogeneous, epidemic among men having sex with men. Through coalescent-based analysis, we showed that these chains arose through separate introductions of subtype B strains into the United Kingdom in the early to mid-1980s. After an initial period of exponential growth, the rate of spread generally slowed in the early 1990s, which is more likely to correlate with behavior change than with reduced infectiousness resulting from highly active antiretroviral therapy. Our results provide insights into the complexity of HIV-1 epidemics that must be considered when developing HIV monitoring and prevention initiatives.

Fig. 1.

Schematic representation of the phylogeny generated from 3,429 U.K. and worldwide HIV-1 subtype B pol sequences. Filled circles represent sequences from the U.K., and open squares represent non-U.K. sequences. Three branching patterns were distinguished: non-U.K. transmission clusters (a), sporadic U.K. infections (b), and U.K. transmission clusters (c). Transmission clusters are clades of sequences from a particular location that descend from a common ancestor, indicating a successful spread of the virus in that location. U.K. transmission clusters are defined as those clades that include at least 25 sequences, 90% or more of which are of U.K. origin.

Fig. 2.

Phylogenetic trees of the six U.K. transmission clusters and their corresponding estimated epidemic histories (all shown on the same time scale). The trees represent the ancestral relationships of sequences belonging to each cluster. (a) Cluster 1. (b) Cluster 2. (c) Cluster 3. (d) Cluster 4. (e) Cluster 5. (f) Cluster 6. The demographic histories were estimated by Bayesian MCMC inference with a model of logistic growth (see text for details) and show change in the effective number of infections through time (time scale in calendar years). The blue line shows the median estimate of the effective number of infections, whereas the black lines show the 95% confidence limits of the estimate.

Fig. 3.

Schematic representation of the logistic model of population growth. According to this model, the number of infections population grows exponentially at rate r from time t_a (time of the most recent common ancestor of the sampled sequences). The growth rate slows as time moves toward the present, such that Ne represents the effective number of infections at the present. Ne can be thought of as the number of infections contributing to new infections rather than the total number of prevalent infections within the cluster.