Nevirapine-selected mutations Y181I/C of HIV-1 reverse transcriptase confer cross-resistance to stavudine

Stavudine administration did not increase the frequency of Y181I/C reverse transcriptase (RT) mutations in non-nucleoside reverse transcriptase inhibitor (NNRTI)-treated patients. However, recombinant Y181C HIV-1 showed reduced stavudine susceptibility with respect to both recombinant wild-type and K103N HIV-1 strains. In addition, recombinant Y181I RT enzyme showed reduced susceptibility to stavudine with respect to both wild-type and K103N RT. A previously unnoticed role of Y181I/C RT changes selected by nevirapine or other NNRTI in determining stavudine resistance is documented.

HIV-1 resistance to reverse transcriptase (RT) inhibitors is mediated by mutations acquired under the selective pressure of nucleoside reverse transcriptase inhibitors (NRTI) and non-nucleoside reverse transcriptase inhibitors (NNRTI). Combination treatments including NRTI and NNRTI often select multidrug-resistant HIV-1 strains carrying multiple RT mutations [¹]. In addition, mutations conferring simultaneous cross-resistance to multiple NRTI have been reported [²]. The mechanisms of resistance to stavudine are still elusive. Although associated with V75T substitution [³], stavudine resistance appears more frequently to be related to complex patterns of RT mutations induced by NRTI [²].

In the present report, the role of RT mutations Y181I/C, both conferring high resistance to nevirapine [⁴], in determining increased resistance to stavudine is documented.

In the period January 1999 to December 2001, HIV RT sequencing data were prospectively collected from 816 patients failing antiretroviral therapy (ART). Among these patients, 243 (29.8%) failed NNRTI-containing ART regimens. Overall, 66 out of 243 NNRTI-treated patients (27.2%) harboured HIV-1 with RT Y181I/C mutations, whereas 177 out of 243 (72.8%) did not. No mutation at RT codon 181 was found in patients not exposed to NNRTI. Stavudine was administered in combination with NNRTI in 48 out of 66 patients (72.7%) showing the RT Y181I/C mutations and in 122 out of 177 patients (68.9%) without RT mutations at codon 181 (odds ratio 1.49, 95% CI 0.89-1.23, Fisher's exact test P > 0.05). In more detail, 101 out of 243 NNRTI-treated patients (41.5%) failed ART regimens including nevirapine. Among these, 43 out of 101 patients (42.6%) harboured RT Y181I/C mutant HIV-1. Again, the proportion of patients receiving stavudine in combination with nevirapine was not significantly different between patients with (31/43, 72.1%) or without (34/58, 58.6%) mutations at RT codon 181 (Fisher's exact test P > 0.05).

The effect of NNRTI-induced RT mutations on stavudine resistance was analysed by evaluating in parallel the drug susceptibility of recombinant HIV-1 carrying either wild-type RT or RT mutated at codons 181 or 103. In detail, HIV-1 RT sequences were selected from one treatment-naive patient showing only amino acid polymorphisms not known to be associated with drug resistance (M16T, R83K, K101R, K166R, and R211K) and two NNRTI-treated patients showing as the only RT drug-resistance-associated mutation Y181C (plus polymorphisms K101R, K122P, I135T, K166R, and D177E) or K103N (plus polymorphisms V35T, V60I, R83K, D121H, I178L, Q207E, R211K, V245E, A272P, and K277R) changes, respectively. Briefly, RT genes were amplified by polymerase chain reaction and cloned in the pHXB2ΔRT2-261 plasmid (carrying the HIV-1 strain XB2 backbone deleted of the RT gene) as previously described [⁵]. Plasmid DNA was then transfected into HeLa CD4 cells to reconstitute infectious HIV-1. The drug susceptibility of recombinant HIV-1 strains was determined as reported.

In parallel, the level of resistance to RT inhibitors was predicted using a mathematical algorithm for the interpretation of RT mutations (Stanford software system; http://hiv-4.stanford.edu/cgi-test/hivtest-web.pl).

Phenotypic drug susceptibility testing showed an increase in the stavudine ID₅₀ value of the recombinant HIV-1 carrying the mutated RT 181 codon (4.8 ± 1.0 μM) with respect to recombinant HIV-1 strains carrying either wild-type RT (0.6 ± 0.4 μM) or RT mutated at codon 103 (1.4 ± 0.03 μM) (Fig. 1a). No significant increase in resistance to thymidine analogue zidovudine was observed for the recombinant 181 and 103 RT mutants (ID₅₀, 0.07 ± 0.07 and 0.07 ± 0.06 μM, respectively) with respect to the recombinant wild-type RT strain (0.03 ± 0.04 μM). Finally, as expected, both recombinant 181 and 103 RT mutants showed high ID₅₀ levels for nevirapine (ID₅₀ > 24 μM) with respect to the recombinant wild-type RT strain (0.1 ± 0.07 μM).

Fig. 1

As expected, the Stanford software interpretation system scored the three recombinant strains as stavudine susceptible.

To confirm that NNRTI-induced mutations at codon 181 could impair stavudine processing by HIV-1 RT, recombinant wild-type and mutated HIV-1 RT enzymes were expressed, purified and assayed as reported [⁶]. The co-expression vectors pUC12N/p66(His)/p51 with the wild-type or the mutant (Y181I and K103N) forms of HIV-1 RT p66 were kindly provided by Dr S.H. Hughes.

The susceptibility of recombinant HIV-1 RT enzymes carrying NNRTI resistance-associated mutations to inhibition by the dideoxynucleoside triphosphate analogues 2′,3′-dideoxythimidine triphosphate (ddTTP), zalcitabine triphosphate (ddCTP), zidovudine triphosphate (ZDVTP) and stavudine triphosphate (d4TTP) (NRTI active forms) was tested in comparison with the wild-type enzyme. As reflected by the increase in the corresponding K_m values, mutations at codon 181 significantly reduced the affinity of the enzyme for the dideoxynucleoside triphosphate substrate (K_m, 11 ± 0.5 μM) with respect to wild-type enzyme (K_m, 4.2 ± 0.5 μM). In contrast, a mutation at codon 103 did not decrease the binding affinity (K_m, 2 ± 0.2 μM). The increase in K_i (competitive inhibition constant) values for ddCTP, ddTTP and d4TTP of RT enzyme with a mutation at codon 181 (0.4 ± 0.05, 0.07 ± 0.01 and 0.09 ± 0.005 μM), with respect to wild-type enzyme (0.2 ± 0.03, 0.02 ± 0.001, and 0.02 ± 0.005 μM), showed that the mutation at codon 181 was also associated with a 2, 3.5 and 4.5-fold decrease in susceptibility to the relevant drug (Fig. 1b). In contrast, no significant differences between the wild-type RT enzyme (Ki, 0.007 ± 0.001 μM) and the enzyme mutated at codon 181 (Ki, 0.01 ± 0.002 μM) or codon 103 (Ki, 0.006 ± 0.001 μM) were found for ZDVTP inhibition (Fig. 1b).

Mutations Y181I/C are involved in high-level NNRTI resistance, especially to nevirapine and delavirdine [⁴]. We previously showed that the 181 mutation reduced the affinity of RT for the deoxyribonucleotide triphosphate substrate, and was also involved in the stereochemical discrimination between the d- and l- enantiomers of deoxyribonucleotide triphosphate. Here, we show that mutations at codon 181 are selected by NNRTI, but also determine stavudine resistance. In fact, RT 181 mutations were detected in patients treated with NNRTI only. In addition, no increase in the frequency of RT 181 mutations was observed in patients treated with NNRTI in combination with stavudine. Therefore, stavudine does not appear to contribute to the selection of RT Y181I/C mutations. On the other hand, once acquired, the Y181I/C mutations are responsible for an eightfold reduction in HIV stavudine susceptibility, with respect to HIV harbouring either wild-type RT or RT carrying other NNRTI resistance-associated mutations. In addition, the re recombinant enzyme carrying an NNRTI resistance-associated mutation at codon 181 is four to fivefold less susceptible to inhibition by d4TTP.

These findings have an immediate clinical impact. In fact, rescue ART regimens after the failure of NNRTI-including treatments should avoid the inclusion of stvudine when in the presence of RT Y181I/C changes. Caution in administering ART regimens including a combination of stavudine with nevirapine is thus advisable. In addition, but no less important, software-based interpretation systems (improperly defined as 'virtual phenotypes') must be utilized with caution because they are unable to evaluate unusual drug-resistance profiles correctly.