Evolutionary Rate and Genetic Heterogeneity of Human T-CellLymphotropic Virus Type II (HTLV-II) Using Isolates from EuropeanInjecting Drug UsersMarco Salemi, 1 Anne-Mieke Vandamme, 1 Chiara Gradozzi, 2 Kristel Van Laethem, 1 Ercole Cattaneo, 3Graham Taylor, 4 Claudio Casoli, 5 Patrick Goubau, 1 Jan Desmyter, 1 Umberto Bertazzoni 21 Rega Institute for Medical Research and University Hospitals, Katholieke Universiteit Leuven, Minderbroedersstraat 10, B-3000 Leuven,Belgium2 Istituto Biologia e Genetica, Universita ` di Verona, Strada Le Grasie 8, I-37 134 Verona, Italy3 IRCCS, Policlinico S. Matteo, I-27100 Pavia, Italy4 Imperial College School of Medicine at St. Mary’s, Norfolk Place, London W2 1PG, UK5 Istituto di Patologia Medica, Universita ` di Parma, Via Gramici 14, I-43100 Parma, ItalyAbstract. Seven new Italian and two new BritishHTLV-II isolates were obtained from injecting drug us-ers and the entire long terminal repeat (LTR) region wassequenced. Restriction analysis showed that all the Ital-ian isolates are of the IIb subtype, whereas the Britishisolates are of the IIa subtype. To understand whether thefurther differentiation of each two principal HTLV-IIsubtypes in several subgroups could be statistically sup-ported by phylogenetic analysis, the neighbor-joining,parsimony, and maximum likelihood methods were used.The separation between IIa and IIb is very well sup-ported by all three methods. At least two phylogeneticsubgroups exist within the HTLV-IIa and at least threewithin the HTLV-IIb subtype. In the present analysis, nostatistical support was obtained for additionalphylogroups. Two particular subgroups seem interestingbecause they include all European and North Americaninjecting drug user strains within the IIa and IIb sub-types, respectively. These data confirm that EuropeanHTLV-II infection among drug users is probably derivedfrom North America. They also suggest that though acertain differentiation by restriction analysis in differentsubgroups is possible, carefully interpreted phylogeneticanalyses remain necessary. Using the likelihood ratiotest, a molecular clock for the drug user strains was cali-brated. A fixation rate between 1.08 × 10 −4 and 2.7 ×10 −5 nucleotide substitutions per site per year was cal-culated for the IIa and IIb injecting drug user strains.This is the lowest fixation rate so far reported for RNAviruses, including for HIV, which typically range be-tween 10 −2 and 10 −4 .Key words: HTLV-II; IIa and IIb subtypes — Restric-tion analysis — Phylogenetic analysis — Maximum like-lihood — Neighbour-joining — Parsimony — Molecularclock — Fixation rateIntroductionHuman T-cell lymphotropic virus type I (HTLV-I) andtype II (HTLV-II) are closely related members of a groupof mammalian retroviruses and share many structuraland functional properties (Poiesz et al. 1993). HTLV-Ihas been etiologically connected to adult T-cell leukemia(ATL) and to HTLV-I-associated myelopathy-tropicalspastic paraparesis (HAM/TSP) (Gessain et al. 1985;Poiesz et al. 1980). No clear connection with any leuke-mia disease has yet been established for HTLV-II, al-though it was initially identified in a patient with atypicalhairy cell leukemia (Kalyanaraman et al. 1982; Rosen-blatt et al. 1986). Instead, recent studies seem to indicateCorrespondence to: M. Salemi; e-mail: marco.salemi@uz.kuleuven.ac.beJ Mol Evol (1998) 46:602–611© Springer-Verlag New York Inc. 1998
