Search

In a recent report, Balestrieri et al. (1) asserted that the effect of lamivudine (3TC) against human T-cell leukemia virus type 1 (HTLV-1) at therapeutic concentrations is seen mainly on cell proliferation, with little or no direct antiviral effect. These findings are supported by biochemical studies that have shown a high level of 3TC resistance by the reverse transcriptase (RT) of HTLV-1 (3). We and others have examined the efficacy of 3TC in HTLV-1-infected patients with HTLV-1-associated myelopathy-tropical spastic paraparesis (6, 8). In most of these patients, we saw a reduction in HTLV-1 proviral DNA load following several weeks of therapy. However, in almost all cases this decline in HTLV-1 viremia was transient and most patients regained baseline HTLV-1 load titers after a while. Despite experiencing virological failure, only silent mutations and few amino acid changes in the HTLV-1 RT could be recognized after extensive analyses, which hardly could explain high levels of 3TC resistance. Furthermore, none of the HTLV-1-positive patients examined so far have developed genetic changes during therapy with 3TC in the conserved LPQG (Q151 M) or YMDD (M184V) motifs, which are associated with resistance to this drug in other lentiviruses (7). Taken together, these data suggest that HTLV-1 RT presents primary resistance to 3TC, and codons other than 151 or 184 could be implicated. In human immunodeficiency virus type 1 (HIV-1), a novel set of mutations in the RT, E44D and V118I, has been reported to produce 3TC resistance in the absence of the M184V mutation (5). In order to examine the possible role of these mutations in HTLV-1, we compared the amino acid sequence of the pol gene of HTLV-1 (GenBank {"type":"entrez-nucleotide","attrs":{"text":"D13784","term_id":"221866","term_text":"D13784"}}D13784) and subtype B HIV-1 (HXB2; GenBank {"type":"entrez-nucleotide","attrs":{"text":"K03455","term_id":"1906382","term_text":"K03455"}}K03455). Alignment was performed with the CLUSTAL X program. Whereas the aspartic acid at residue 44 was not found in the HTLV-1 RT, 118I was recognized as the consensus amino acid in HTLV-1. In vitro studies of HIV-1 have shown that an isoleucine at position 118 decreases the rate of 3TC incorporation, and this reduction does not depend on the presence of classical AZT mutations (M. Girouard, K. Diallo, B. Marchand, S. McCormick, M. A. Wainberg, and M. Gotte, XI Int. HIV Drug Resist. Workshop, Seville, Spain, abstr. 26, 2002). Resistance to 3TC develops when less of the drug monophosphate gets incorporated into the nascent cDNA due to steric conformation changes in the RT. In contrast, resistance to AZT and many other nucleoside analogs develops mainly as result of pyrophosphorolysis, by which there is an increased excision and removal of the incorporated chain terminator nucleotides (2). The V118I mutation produces an alteration of the structure-dependent binding in the HIV-1 RT which translates into lower affinity for 3TC. Although HIV-1 and HTLV-1 share only ∼20% of amino acid sequence identity, their RT enzymes are quite similar. In fact, RTs from almost all lentiviruses share similar tertiary structures in palm and finger domains, in spite of high divergence in their amino acid residues (4). Taking into account all these considerations, the presence of an isoleucine at codon 118 could be in part responsible for a natural 3TC resistance by HTLV-1. Biochemical and site-direct mutagenesis studies should be conducted to confirm our hypothesis. In the meantime, for the benefit of patients with HTLV-1-associated myelopathy-tropical spastic paraparesis, antiretroviral therapy should no longer include 3TC.