Phylogenetic analysis of 48 papillomavirus types and 28 subtypes and variants: A showcase for the molecular evolution of DNA viruses

Abstract

Papillomaviruses are attractive models for studying the molecular evolution of DNA viruses because of the large number of isolates that exhibit genomic diversity and host species and tissue specificity. To examine their relationship, we selected two amino acid sequences, one of 52 residues within the early gene E1 and the other of 44 residues within the late gene L1, which allowed insertion- and deletion-free alignment of all accessible papillomavirus sequences. We constructed phylogenetic trees from the amino acid and corresponding nucleotide sequences from 28 published and 20 newly determined animal and human papillomavirus (HPV) genomic sequences by using distance matrix, maximum-likelihood, and parsimony methods. The trees agreed in all important topological aspects. One major branch with two clearly separated clusters contained 11 HPV types associated with epidermodysplasia verruciformis. A second major branch had all the papillomaviruses involved in genital neoplasia and, in distant relationship, the cutaneous papillomaviruses HPV type 2a (HPV-2a), HPV-3, and HPV-10 as well as the "butcher's" papillomavirus HPV-7 and two simian papillomaviruses. Four artiodactyl (even-toed hoofed mammal) papillomaviruses, the cottontail rabbit papillomavirus, and avian (chaffinch) papillomavirus type 1 formed a third major branch. Last, four papillomaviruses exhibited little affinity to any of these three branches; these were the cutaneous types HPV-1a, HPV-4, and HPV-41 and B-group bovine papillomavirus type 4. The phylogeny suggests that some branches of papillomavirus evolution are restricted to particular target tissues and that a general process of long-term papillomavirus-host coevolution has occurred. This latter hypothesis is still conjectural because of bias in the current data base for human types and the paucity of animal papillomavirus sequences. The comparison of evolutionary distances for the most closely related types with those of 28 subtypes and variants of HPV-2, HPV-5, HPV-6, HPV-16, and HPV-18 supports the type as a natural taxonomic unit, with subtypes and variants being expressions of minor intratype genomic diversity similar to that found in the natural populations of all biological species. An exception to this seems to be HPV-2c, which has an evolutionary distance from HPV-2a of the intertype magnitude and may eventually have to be regarded as a distinct type. We describe an experimental approach that estimates the taxonomic and phylogenetic positions of newly identified papillomaviruses without viral isolation and complete genomic sequencing. Finally, the paper also explores concepts for a natural, i.e., phylogenetically based, papillomavirus taxonomy; demonstrates the rooting of an HPV variant cluster against its closest type relative; and presents a hypothetical combined phylogeny of the papillomaviruses with other DNA tumor viruses.