The global spread of hepatitis C virus 1a and 1b: a phylodynamic and phylogeographic analysis

Abstract

Background: Hepatitis C virus (HCV) is estimated to affect 130-180 million people worldwide. Although its origin is unknown, patterns of viral diversity suggest that HCV genotype 1 probably originated from West Africa. Previous attempts to estimate the spatiotemporal parameters of the virus, both globally and regionally, have suggested that epidemic HCV transmission began in 1900 and grew steadily until the late 1980s. However, epidemiological data suggest that the expansion of HCV may have occurred after the Second World War. The aim of our study was to elucidate the timescale and route of the global spread of HCV.

Methods and findings: We show that the rarely sequenced HCV region (E2P7NS2) is more informative for molecular epidemiology studies than the more commonly used NS5B region. We applied phylodynamic methods to a substantial set of new E2P7NS2 and NS5B sequences, together with all available global HCV sequences with information in both of these genomic regions, in order to estimate the timescale and nature of the global expansion of the most prevalent HCV subtypes, 1a and 1b. We showed that transmission of subtypes 1a and 1b "exploded" between 1940 and 1980, with the spread of 1b preceding that of 1a by at least 16 y (95% confidence interval 15-17). Phylogeographic analysis of all available NS5B sequences suggests that HCV subtypes 1a and 1b disseminated from the developed world to the developing countries.

Conclusions: The evolutionary rate of HCV appears faster than previously suggested. The global spread of HCV coincided with the widespread use of transfused blood and blood products and with the expansion of intravenous drug use but slowed prior to the wide implementation of anti-HCV screening. Differences in the transmission routes associated with subtypes 1a and 1b provide an explanation of the relatively earlier expansion of 1b. Our data show that the most plausible route of the HCV dispersal was from developed countries to the developing world. Please see later in the article for the Editors' Summary.

Figure 1. Similarity plot of the genotype 3 full-length reference sequence versus reference sequences for other genotypes along with the HCV genome.

Similar plots are produced for all the genotypes. The shaded regions, E2P7NS2 and NS5B, are clearly the most divergent ones, and were selected for PCR-sequencing.

Figure 2. Phylogenetic trees of the isolates used in the population dynamics (yellow circles) along with all the available NS5B sequences (tips without circles).

Figure 3. Global population dynamics of the hepatitis C virus genotypes 1a and 1b based on relaxed-clock analysis of NS5B.

The tMRCA estimated from E2P7NS2 was used to provide a gamma-distributed prior for the tMRCA of strains also available for NS5B.

Figure 4. Phylogeographic trees of all the partial NS5B sequences available for 1a and 1b subtypes.

The red, green, and black lines indicate events attributed to the sequences sampled from the US, other developed, and developing countries, respectively.