|Home | About | Journals | Submit | Contact Us | Français|
9-R-2-Phosphonomethoxypropyl adenine (PMPA) is an acyclic nucleoside phosphonate analog that has demonstrated efficacy against human immunodeficiency virus (HIV). We recently described the synthesis, metabolism, and biological activities of bis(isopropyloxymethylcarbonyl)PMPA [bis(poc)PMPA] as an orally bioavailable prodrug for PMPA. Among a large panel of drug-resistant HIV type 1 variants, only the K65R virus was resistant to PMPA. K65R virus also showed reduced susceptibility to bis(poc)PMPA, although the prodrug could still inhibit these viruses at submicromolar, nontoxic concentrations. Among a panel of seven primary clinical isolates from patients with diverse treatment histories, only one isolate showed reduced susceptibility to PMPA and was found to carry three mutations (M41L, T69N, R73K) in its reverse transcriptase catalytic domain.
9-R-2-Phosphonomethoxypropyl adenine (PMPA), a member of the acyclic nucleoside phosphonate family of antiviral agents, is a potent inhibitor of both hepadnaviruses and retroviruses including the human immunodeficiency virus (HIV) (for a review, see reference 2). PMPA is highly effective for the pre- or postexposure prophylaxis of simian immunodeficiency virus (SIV) infection in rhesus monkeys and prevents the establishment of SIV if it is administered 48 h before or 4 or 24 h after virus inoculation (35). PMPA is also effective against chronic SIVmne infection in cynomolgus monkeys (36). A single daily dose of PMPA administered subcutaneously reduces plasma SIV RNA levels by >99% within 2 weeks of treatment. PMPA is currently undergoing phase I/II clinical trials for the treatment of HIV infection in AIDS patients. Preliminary results have shown that PMPA given intravenously appears to be safe and is well tolerated, and it caused a 1.1 log reduction of the HIV RNA levels after the administration of only eight doses (7). PMPA is poorly absorbed after oral administration, presumably due to the negative charges on the phosphonyl group. We recently described a bis(isopropyloxymethylcarbonyl) ester of PMPA, designated bis(poc)PMPA, which functions as an orally available prodrug of PMPA (3). Phase I/II clinical trials with bis(poc)PMPA are ongoing. Current treatment guidelines recommend the use of two or more nucleoside inhibitors plus a protease inhibitor (4). Understanding of the cross-resistance patterns of different drugs within a class is essential for determining suitable drug combinations. Toward this end, we have investigated the susceptibility of a panel of drug-resistant HIV isolates with various phenotypes for their susceptibility to PMPA and bis(poc)PMPA.
The synthesis of PMPA and bis(poc)PMPA has been described before (1). PMPA and bis(poc)PMPA were generously supplied by Norbert Bischofberger (Gilead Sciences, Foster City, Calif.). The antiviral and cytotoxic activities of these compounds in human peripheral blood mononuclear cells (PBMCs) or the T-lymphocytic MT-2 cell line were determined by HIV type 1 (HIV-1) p24 yield reduction assays as described previously (29, 30). PMPA was a potent inhibitor of HIV-1 replication, with effective antiviral concentrations of ~0.6 and ~0.2 μM in MT-2 cells and PBMCs, respectively (Table (Table1).1). PMPA was also relatively nontoxic to cells, with 50% inhibitory concentrations (IC50s) of ~1,200 to 1,250 μM, and it had a high therapeutic index. Bis(poc)PMPA was ~30- to 90-fold more potent than PMPA in both its antiviral and cytotoxic effects, with little change in its therapeutic index. Thus, bis(poc)PMPA appears to be an effective prodrug for the efficient intracellular delivery of PMPA.
Table Table22 summarizes the sources, phenotypes, and genotypes of the mutant viruses used in this study. The genotypic information on the isolates is from the literature, while the phenotypes of the mutants were confirmed prior to antiviral assays with PMPA and bis(poc)PMPA. Interestingly, all isolates with the exception of the K65R virus were susceptible to PMPA and bis(poc)PMPA. The K65R virus was about eightfold less susceptible to PMPA than the wild-type Hxb2 molecular clone. Likewise, bis(poc)PMPA was about ninefold less active against the K65R virus, although the virus was still inhibited by submicromolar concentrations of the drug that were well below the drug’s cytotoxic concentration.
With the increasing use of protease inhibitors, HIV variants with resistance to one or more protease inhibitors are emerging. Although protease inhibitors and PMPA target different viral gene products, protease and reverse transcriptase (RT) are both processed from a common precursor polyprotein. However, neither a saquinavir-resistant HIV variant (16), nor an HIV variant resistant to multiple protease inhibitors (6) showed any cross-resistance to PMPA or bis(poc)PMPA (Table (Table22).
K65R RT shows diminished chain termination in the presence of ddCTP (13). We therefore investigated whether K65R RT is resistant to inhibition by PMPApp, the biologically active intracellular derivative of PMPA and bis(poc)PMPA. Virion-associated RT activity and its inhibition by PMPApp were determined by using MS-2 phage RNA as a heteropolymeric template and a synthetic oligonucleotide (5′-CGT TAG CCA CTC CGA AGT GCG T-3′) complementary to residues 3326 to 3347 of MS-2 RNA (8) as the primer. The reactions were initiated by the addition of HIV-1IIIB, and K65R virus preparations concentrated 100-fold by ultracentrifugation. RT reactions were carried out for 90 min at 37°C in the presence of 32P-dATP, unlabelled dCTP, dGTP, and dTTP, and various concentrations of PMPApp. At the end of the incubation period, samples were spotted onto DE-81 filters, washed three times in 2× SSC (1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate), rinsed in 95% ethanol, and air dried prior to scintillation counting. As shown in Fig. Fig.1,1, PMPApp was 10-fold less effective in inhibiting RT from K65R virus (IC50, ~1 μM) than it was in inhibiting RT from wild-type IIIB virus (IC50, ~0.1 μM) or purified recombinant RT from HIV-1 BH-10 (IC50, ~0.1 μM) (data not shown). The concordance between the reduced susceptibility of virion-associated RT activity to inhibition by PMPApp in vitro and the phenotypic resistance of the K65R virus in cell culture suggests that screening for enzymatic resistance may provide a surrogate approach to the rapid detection and characterization of at least a subset of PMPA-resistant mutants.
We also investigated the antiviral efficacy of PMPA against primary isolates from treatment-naive patients or from patients experiencing treatment failure (Table (Table3).3). The treatment histories of these subjects are summarized in Table Table3.3. Six of the seven isolates studied were highly susceptible to PMPA. One isolate from a patient previously treated with zidovudine (AZT) and ddI showed reduced sensitivity to PMPA and bis(poc)PMPA. The catalytic domain of RT from this isolate was amplified by PCR and was cloned into plasmid vectors and analyzed by automated dye-terminator sequencing. One of the three clones sequenced contained nonconservative mutations that, compared to wild-type RT sequences, resulted in amino acid changes at codons 41 (M41V), 69 (T69N), and 73 (R73K) (28a). The M41V mutation has previously been associated with AZT resistance (17), while a mutation at T69 (T69N) leads to ddI and ddC resistance (9). Mutagenesis studies are in progress to determine whether one or more of these mutations are associated with the reduced sensitivity to PMPA. These results nevertheless suggest that primary clinical isolates are generally sensitive to PMPA, although occasional variants with reduced sensitivity may be encountered in the clinic.
The K65R mutation determines HIV resistance to several other antiviral agents including ddC, ddI, lamivudine (3TC), 9-(2-phosphonylmethoxyethyl)adenine (PMEA), and 1592U89, as well as dioxalane guanosine and other dioxalane derivatives (12, 14, 26, 34, 40). In a recent study, K65R mutations were found to appear in RT of SIV from PMPA-treated newborn macaques, and the K65R SIV showed reduced susceptibility to PMPA (38), although the animals remained disease free for more than 13 months. Recently, Cherrington et al. (5) described the in vitro selection of HIV-1 K65R variants with reduced sensitivity to PMPA. Hence, the selection of the K65R mutation by the use of PMPA is a potential cause of concern. Interestingly, K65R mutants remain susceptible to AZT, and the K65R mutation results in the reversal of AZT resistance, at least in the D67N, K70R, T215Y, and K219Q background. Thus, it may be possible to suppress the emergence of the K65R mutation by combination therapy with AZT. The K65R mutation has been reported among isolates from a small proportion (~25%) of patients on monotherapy with ddI (39), but it is rarely seen among isolates from patients receiving ddI and AZT (15). Likewise, even though ddC can select for the K65R mutation in vitro, K65R mutants are not detected in patients on combination therapy with ddC and AZT (32). Together, these results support the inclusion of AZT in therapeutic regimens involving PMPA (or other drugs affected by the K65R mutation).
HIV-1 variants that carry the Q151M mutation and that are resistant to AZT, ddI, ddC, ddG, and stavudine (d4T; but not 3TC) have been isolated from patients receiving combination therapy with AZT and ddI or AZT and ddC. Additional mutations arising at codons 62, 75, 77, and 116 enhance the degree of antiviral resistance in isolates carrying the Q151M mutation (27–28, 37). It is interesting that Q151M mutants remain sensitive to PMPA and bis(poc)PMPA, suggesting that these compounds may be particularly useful in the treatment of patients harboring HIV isolates with the Q151 mutation.
Finally, bis(poc)PMPA appears to be much more potent than PMPA and achieves much greater intracellular concentrations of PMPA, PMPAp (PMPA monophosphate), and PMPApp (11, 21, 24). Bis(poc)PMPA is rapidly hydrolyzed by esterases within cells and plasma. After oral administration of bis(poc)PMPA, only PMPA, but not bis(poc)PMPA, is detected in the plasma. Thus, it is unclear whether high intracellular PMPApp levels could be achieved in tissues and the circulation following oral bis(poc)PMPA administration. Since the K65R virus could be inhibited by submicromolar concentrations of bis(poc)PMPA that were well below the levels resulting in toxicity, the use of systemic formulations of bis(poc)PMPA or other methods for improving the intracellular delivery of these compounds may provide a novel approach to limiting the problem of drug resistance.
This work was supported in part by PHS grants RO1 AI27652 and UO1-AI32908, by Cancer Center (CORE) grant P30 CA21765 from the National Institutes of Health, and by the American Lebanese Syrian Associated Charities. We are grateful to the following various contributors to the NIH AIDS Research and Reference Reagent Program (Rockville, Md.) for supplying the indicated reagents: Douglas Richman, G910-6, G762, 1391-1, 1391-4, and N119; Brendan Larder and S. Kemp, RTMF, RTMC, RTMDR-1, and HIV-174V; Emilio Emini, HIV-1IIIB A17; John Mellors and Raymond Schinazzi, HIV-1LAI-M184V; and Donald Dubin and Joseph Fitzgibbon, pJF4a. We thank H. Mitsuya (National Institutes of Health, Bethesda, Md.) for supplying pSUM-9, pSUM8, and pSUM12, and Mark Wainberg (McGill University, Montreal, Quebec, Canada) for providing the HIV-1 K65R molecular clone.