The widespread use of HIV-1 protease inhibitors (PI) in multidrug combinations has been a major factor in the current success of antiretroviral therapy. Unfortunately, therapy failure remains common and is frequently attributed to the emergence of drug-resistant HIV-1 strains [1–3]. A feature of resistance to the PI has been the observation that HIV-1 highly resistant to one drug in this class is often cross-resistant to other approved PI [4–6]. For example, strains highly resistant to indinavir (IDV) are commonly cross-resistant to ritonavir (RTV), nelfinavir (NFV) and saquinavir (SQV) [7,8]. Currently, substantial numbers of HIV-1-infected individuals receiving antiretroviral therapy may harbor virus broadly cross-resistant to PI [9,10; for review see 11]. Consequently, there is an urgent clinical need to develop new PI which are able to inhibit these broadly resistant HIV-1 variants.
Tipranavir (TPV) is a novel non-peptidic HIV-1 protease inhibitor, currently under clinical investigation [12,13]. Unlike the approved PI that are modeled on peptidic backbones, TPV was developed from a non-peptidic coumarin template , the antiprotease activity of which was discovered by high-throughput screening . Like other PI, TPV is a potent inhibitor of the HIV-1 protease activity in vitro (Ki 0.008 nmol/l) and in cell culture, with an average 50% inhibitory dose (IC50) below 0.1 μmol/l . Furthermore, phase II clinical trials have shown that TPV is safe and well tolerated in HIV-infected patients [17,18].
It was anticipated from the non-peptidic nature of TPV and the crystallographic analysis of TPV-binding interactions with the protease that this inhibitor might have activity against HIV-1 strains that have become resistant to the current peptidomimetic PI. Indeed, preliminary data have suggested that TPV can block the replication of a number of viruses resistant to the current range of PI . In this study, TPV was screened against a substantial panel of highly PI-resistant clinical isolates in order to determine the spectrum of activity of this inhibitor. Extensive genotypic analysis was performed to identify the major amino acid substitutions in HIV-1 protease associated with TPV sensitivity and resistance.
Materials and methods
Samples obtained from patients were submitted for routine assessment of drug susceptibility in order to identify samples resistant to PI. Overall, 127/134 of the samples had been taken from different individuals (multiple samples from the same individuals were from different sampling times). Although antiretroviral therapy histories were unavailable, the phenotypic resistance patterns and genotypes suggested that the patients had received extensive PI therapy. Subsequently, the in vitro susceptibility to IDV, RTV, NFV and SQV was determined and a group of 105 samples was selected with tenfold or greater increases in IC50 (relative to a wild-type control virus) to at least three of these four inhibitors (as this study was performed prior to the approval of amprenavir, this PI was not included in the analysis). A further group of 29 recombinant viruses had tenfold or greater resistance individually to RTV, NFV or SQV. The susceptibility of all 134 variants was then re-tested to the four PI simultaneously with TPV.
Viral RNA was extracted from 200 μl patient plasma using the QIAamp Viral RNA Extraction Kit (Qiagen, Hilden, Germany), according to the manufacturer's instructions. cDNA encompassing part of pol was produced using Expand reverse transcriptase (RT) (Boehringer Mannheim, Mannheim, Germany) as described previously . A 2.2 kb fragment encoding the protease and RT regions was then amplified by nested polymerase chain reaction (PCR) using PCR primers and conditions as described . This genetic material was subsequently used in both phenotyping and genotyping experiments.
Phenotypic susceptibility testing
MT-4 cells  were co-transfected with pol PCR fragments and a HIV-1 molecular clone with deletion of gag– protease–RT, as described [20,22]. This resulted in viable recombinant viruses containing part of gag, plus protease and RT (up to codon 400) from the donor PCR fragment. Phenotypic susceptibility to PI was determined using an MT-4 cell assay the AntivirogramTM[20,22]. Resistance values were derived by dividing the mean IC50 for a patient's recombinant virus by the mean IC50 for wild-type control virus (strain HXB2-D).
The PCR products obtained from patient plasma samples were genotyped by dideoxynucleotide-based sequence analysis. Samples were sequenced using the Big Dye terminator kit (Applied Biosystems, Foster City, California, USA) and resolved on an ABI 377 DNA sequencer .
The results of the 105 highly cross-resistant panel are shown in Fig. 1 and the susceptibilities of the 29 individually resistant variants (defined as having a greater than tenfold increase in IC50 value to a single PI) are shown in Table 1. The majority of broadly PI cross-resistant isolates were TPV sensitive (95/105; 90%); 8/105 (8%) had a four- to tenfold increase in TPV IC50 and only 2/105 (2%) had greater than a tenfold increase in TPV IC50 value. The mean increases in IC50 (SE in parentheses) were as follows: IDV, 44-fold (2.8); RTV, 87-fold (6.0); NFV, 45-fold (2.1); SQV, 46-fold (2.8); and TPV, 2-fold (0.23). All of the 29 variants that were individually resistant to RTV, NFV, or SQV remained fully sensitive to TPV or were even hypersensitive relative to the wild-type control (Table 1). Therefore, our data established that TPV was extremely effective at inhibiting a substantial range of PI-resistant clinical strains.
Genotypic analysis of all 134 samples revealed complex patterns of multiple mutations in the protease coding regions. Figure 2a illustrates the relative frequencies of recognized primary and secondary mutations in the 105 highly PI-resistant samples. The average number of documented PI resistance mutations per sample in this group was 6.1. Substitutions at the following 19 HIV-1 protease amino acid residues were considered to be associated with PI resistance: 10, 20, 24, 30, 32, 33, 36, 46, 47, 48, 50, 54, 71, 73, 77, 82, 84, 88, and 90 (major or ‘primary’ mutations are shown in bold type). As anticipated from previous studies [10,11], there was a predominance of major or ‘primary’ mutations at codons 82, 84, 90 and secondary mutations at codons 10, 36, 46, 54, 71 and 77. The frequencies of PI mutations seen in the RTV-, NFV- or SQV-resistant strains are also shown in Fig. 2. This analysis revealed, as expected, marked differences in PI mutation patterns between the different groups. For example, there was a predominance of characteristic codon 30 and 88 mutations in the group resistant only to NFV and a predominance of characteristic codon 48 and 90 mutations in the group resistant only to SQV. This genotypic analysis demonstrated that the 134 samples studied contained a wide spectrum of different PI resistant genotypes, the vast majority of which remained TPV susceptible.
In an attempt to define those mutations responsible for TPV resistance and mutational patterns that did not confer resistance, two subsets of samples from the 105 highly resistant group were analysed. Since there were only two samples with greater than tenfold increase in IC50 values, all samples with a greater than fourfold increase in TPV IC50 were grouped together (n = 10); the mean increases in IC50 (SE in parentheses) were: TPV, 8-fold (1.1); IDV, 62-fold (6.4); RTV, 140-fold (18.7); NFV, 54-fold (5.0); and SQV, 57-fold (8.3). The mutational patterns were also examined in 19 samples with > 2.5-fold increased TPV sensitivity (`hypersensitivity') relative to the wild-type control. The mean fold decrease in IC50 (SE in parentheses) for TPV was 3 (0.02); mean increases in IC50 for the other PI were: IDV, 35-fold (6.3); RTV, 57-fold (10.4); NFV, 37-fold (6.2); and SQV, 51-fold (7.0). The PI mutation frequencies in these two subsets are illustrated in Fig. 3a. The average number of documented PI resistance mutations per sample was 6.8 in the TPV-resistant group and 6.3 in the ‘hypersensitive’ group (although there were numerous additional polymorphisms in both groups relative to the laboratory reference strain HXB2-D). It was striking that the TPV-hypersensitive group had a high frequency of G48V and V82A mutations (Fig. 3a). By contrast, the TPV-resistant group had a high frequency of the relatively rare mutation V82T, plus I84V (Fig. 3b). Examination of individual samples with reduced TPV susceptibility revealed two clusters of mutation patterns, either 82T with 84V or 84V with 90M (both with numerous secondary mutations) (data not shown). However, we also identified TPV-sensitive samples that had combinations of mutations including 84V and 90M. Therefore, the precise combinations of PI resistance mutations that dictate resistance to TPV remain to be elucidated. It is possible that novel mutation patterns could confer TPV resistance since the samples with reduced TPV susceptibility contained an average of 9.2 polymorphisms in the protease, in addition to recognized sites of PI resistance.
Previous studies have provided clues as to why TPV might be less affected by the mutations described above that give rise to resistance to peptidomimetic PI. Specifically, peptidomimetic inhibitors bind to the protease via an extensive hydrogen bonding network, resulting in the main chain of the inhibitor being extended and rigidly constrained in the binding pocket . Hydrogen bonds are highly directional, allowing for little flexibility. In contrast, TPV binds to the protease with far fewer key hydrogen bonds and is more dependent on hydrophobic interactions . The notion of flexibility of the inhibitor to adjust to amino acid changes in the active site is supported by the fact that a range of different hydrophobic groups at C-6 and C-3 positions are well tolerated. In addition, X-ray derived structures of different protease–inhibitor complexes revealed subtle differences in the conformation of different ligands in the active site . These observations provided an interesting rationale for testing the susceptibility of TPV against a large range of PI-resistant clinical isolates.
In view of the IC50 value (about 0.1 μmol/l) that is typically seen with TPV against wild-type HIV-1 (somewhat higher than the US Food and Drug Administration approved PI), the potential clinical efficacy of this inhibitor was of obvious interest. To address this issue, a recent phase II study in antiretroviral drug-naive subjects involved treatment with TPV with or without low-dose RTV . Median decreases of HIV-1 plasma RNA of 0.8–1.6 log10 copies/ml were achieved with median TPV trough concentrations of 0.76–67 μmol/l . These data clearly demonstrated that TPV has clinical activity against susceptible HIV-1 strains.
In summary we have used an in vitro culture approach to define the antiviral activity of TPV against a wide spectrum of clinically relevant, highly PI-resistant strains of HIV-1. This extensive assessment demonstrated a substantial and surprising lack of cross-resistance. Mutational patterns could be identified that differentiated TPV-resistant strains from those hypersensitive to the inhibitor. However, extensive site-directed mutagenesis and in vitro drug-selection studies will be required to define the precise genetic nature of TPV resistance, which is likely to be relatively complex. Since TPV has already shown significant clinical activity in antiretroviral drug-naive patients [17,18,24], it will be important to establish if the promising in vitro activity reported here translates into activity in patients who harbor PI-resistant strains. Such studies in patients with a history of therapy with other PI are currently in progress. Finally, pre-defining an individual's resistance status should facilitate the selection of potentially active partner drugs that can be used in combination with TPV to protect its activity and optimize viral suppression.
The authors would like to acknowledge the Virco Belgium phenotyping team and the Virco UK genotyping team, as well as the Pharmacia & Upjohn HIV-1 protease inhibitor team for their many contributions to this work.
1. D'Aquila RT, Johnson VA, Welles SL. et al. Zidovudine resistance and HIV-1 disease progression during antiretroviral therapy.
Ann Intern Med 1995, 122: 401 –408.
2. Kuritzkes DR. Clinical significance of drug resistance in HIV-1 infection.
AIDS 1996, 10: S27 –S31.
3. Poppe SM, Slade DE, Chong KT. et al. Antiviral activity of the dihydropyrone PNU-140690, a new nonpeptidic human immunodeficiency virus protease inhibitor.
Antimicrob Agents Chemother 1997, 41: 1058 –1063.
4. Condra JH, Schleif WA, Blahy OM. et al. In vivo emergence of HIV-1 variants resistant to multiple protease inhibitors.
Nature 1995, 374: 569 –571.
5. Gulnik SV, Suvorov LI, Liu B. et al. Kinetic characterization and cross-resistance patterns of HIV-1 protease mutants selected under drug pressure.
Biochemistry 1995, 34: 9282 –9287.
6. Tisdale M, Myers RE, Maschera B, Parry NR, Oliver NM, Blair ED. Cross-resistance analysis of human immunodeficiency virus type 1 variants individually selected for resistance to five different protease inhibitors.
Antimicrob Agents Chemother 1995, 39: 1704 –1710.
7. Condra JH, Holder DJ, Schleif WA. et al. Genetic correlates of in vivo viral resistance to indinavir, a human immunodeficiency virus type 1 protease inhibitor.
J Virol 1996, 70: 8270 –8276.
8. Schapiro JM, Winters MA, Lawrence J, Merigan TC. Clinical cross-resistance between the HIV-1 protease inhibitors saquinavir and indinavir and correlations with genotypic mutations.
AIDS 1999, 13: 359 –365.
9. Schinazi RF, Larder BA, Mellors JW. Mutations associated with HIV-1 drug resistance.
Int Antiviral News 1997, 5: 129 –142.
10. Hertogs K, Kemp S, Bloor S. et al. Patterns of cross-resistance among protease inhibitors in over 1500 clinical HIV-1 isolates.
:Comparison of genotypic and phenotypic resistance profiles.
Antiviral Ther 1998, 3: 49 –50.
11. Boden D, Markowitz M. Resistance to human immunodeficiency virus type 1 protease inhibitors.
Antimicrob Agents Chemother 1998, 42: 2775 –2783.
12. Turner SR, Strohbach JW, Tommasi RA. et al. Tipranavir (PNU-140690): a potent, orally bioavailable nonpeptidic HIV protease inhibitor of the 5,6-dihydro-4-hydroxy-2-pyrone sulfonamide class.
J Med Chem 1998, 41: 3467 –3476.
13. Thaisrivongs S, Strohbach JW. Structure-based discovery of tipranavir disodium (PNU-140690E): a potent, orally bioavailable, nonpeptidic HIV protease inhibitor.
Biopolym Pept Sci 1999, 51: 51 –58.
14. Kashman Y, Gustafson KR, Fuller RW. et al. The calanolides, a novel HIV-inhibitory class of coumarin derivatives from the tropical rainforest treeCalophyllum lanigerum.
J Med Chem 1992, 35: 2735 –2743.
15. Thaisrivongs S, Tomich PK, Watenpaugh KD. et al. Structure-based design of HIV protease inhibitors: 4-hydroxycoumarins and 4-hydroxy-2-pyrones as non-peptidic inhibitors.
J Med Chem 1994, 37: 3200 –3204.
16. Thaisrivongs S, Skulnick HI, Turner SR. et al. Structure-based design of HIV protease inhibitors: sulfonamide-containing 5,6-dihydro-4-hydroxy-2-pyrones as non-peptidic inhibitors.
J Med Chem 1996, 39: 4349 –4353.
17. Wang Y, Tutton CM, Borin MT, et al
. The safety, tolerance, pharmacokinetics, and efficacy of PNU-140690, a new non-peptidic HIV protease inhibitor, in a phase I/II study.XII international Conference on AIDS
. Geneva, June 1998 [abstract 41176].
18. Wang Y, Freimuth WW, Daenzer CL. et al. Safety and efficacy of PNU-140690, a new non-peptidic HIV protease inhibitor, and HIV genotypic changes in patients in a Phase II study.
Antiviral Ther 1998, 3: 5. 5.
19. Gubler U, Hoffmann BJ. A simple and very efficient method for generating cDNA libraries.
Gene 1983, 25: 263 –269.
20. Hertogs K, de Bethune MP, Miller V. et al. A rapid method for simultaneous detection of phenotypic resistance to inhibitors of protease and reverse transcriptase in recombinant human immunodeficiency virus type 1 isolates from patients treated with antiretroviral drugs.
Antimicrob Agents Chemother 1998, 42: 269 –276.
21. Harada S, Koyanagi Y, Yamamoto N. Infection of HTLV-III/LAV in HTLV-I-carrying cells MT-2 and MT-4 and application in a plaque assay.
Science 1985, 229: 563 –566.
22. Pauwels R, Hertogs K, Kemp S. et al. Comprehensive HIV drug resistance monitoring using rapid, high-throughput phenotypic and genotypic assays with correlative data analysis.
Antiviral Ther 1998, 3: 35 –36
23. Larder BA, Kohli A, Kellam P. et al. Quantitative detection of HIV-1 drug resistance mutations by automated DNA sequencing.
Nature 1993, 365: 671 –673.
24. Wang Y, Daenzer C, Wood R, et al
. The safety, efficacy, and viral dynamics analysis of tipranavir, a new-generation protease inhibitor, in a phase II study in antiretroviral-naive HIV-1 infected patients.7th Conference on Retroviruses and Opportunistic Infections.
San Francisco, February 2000 [abstract 673].