Although vaccines have helped to control several of the most important viral pathogens, there is currently little prospect of an effective vaccine for either human immunodeficiency virus (HIV) or hepatitis C virus (HCV). These pathogens infect 
40 million and 170 million people worldwide, respectively, hastening the need for effective antiviral drugs.
Drug discovery and development efforts for HIV, based on advances in the understanding of the viral life cycle, have transformed what used to be a rapid and lethal infection into a chronic condition that can be controlled for many years through combination therapies with different classes of antiviral drugs — known as highly active antiretroviral therapy (HAART) (Box 1). Now, 22 years after the inhibitory effects of the first anti-HIV drug, azidothymidine (AZT)1 on the replication of HIV were described2 (then called HTLV-III (human T-cell lymphotropic virus type III)3 or LAV (lymphadenopathy-associated virus)4, 5), more than 20 anti-HIV drugs, belonging to 7 classes, have been approved6 (Supplementary information S1 (figure)).
Drug discovery and development for HCV has also progressed significantly in the past decade. At present, chronic HCV is treated with the combination of pegylated interferon (IFN) and ribavirin (Box 2). Pegylated IFN
-2a and ribavirin afforded an end-of-treatment response in 69% of the patients7, and a sustained viral response (SVR) was observed in 56% of the patients7. However, the question of whether SVR equates to viral clearance is controversial, as occult HCV infection after SVR has been observed in the liver and several types of lymphoid cells (peripheral blood mononuclear cells, T- and B-lymphocytes, T. Michalak, personal communication). Similar to HIV drug development, knowledge of the viral life cycle is providing new opportunities for therapeutic intervention, and the first drugs developed specifically to target HCV enzymes are showing promise in clinical settings.
Despite progress in the treatment of HIV and HCV, there is still considerable room for improvement and expansion of antiviral drugs. These should be: active against wild-type and mutant virus without allowing viral breakthrough; have high oral bioavailability and long elimination half-life, allowing once-daily oral treatment at low doses; have minimal adverse effects; and be easy to synthesize and formulate8. The molecular targets of existing antiviral drugs could be exploited in the design of novel, more potent and/or less toxic agents; new molecular targets within the HCV and HIV replicative cycle could be validated as sites of attack of new chemical entities; and judicious combinations of different drugs (Box 1) may provide more effective drug regimens with a reduced risk for drug resistance development. While building on the expertise and achievements obtained in the past, these approaches might lead to strategies that effectively control HIV and HCV in all infected individuals. This article considers the different stages in the life cycle of HIV and HCV for both established and emerging strategies in the design of antiviral drugs, and highlights the lessons learned in the past 25 years of antiviral drug discovery for the development of novel therapies.
Targets in the viral life cycle
HIV is a retrovirus that replicates through a proviral double-stranded DNA intermediate, whereas HCV is a (+)RNA virus that replicates through a (-)RNA intermediate. Nevertheless, parallels can be drawn between the life cycles of the two viruses with regard to potential targets for antiviral drugs. Fig. 1 shows a simplified flow chart of the life cycles of the two viruses, highlighting established and potential points for therapeutic intervention. Analogous steps in the two viral life cycles include viral entry, replication of the viral genome by polymerases and proteolytic processing of viral polyproteins. There are also steps that can be viewed as virus-specific — for example, HIV replication depends on the integration of the proviral DNA into the genome of infected cells, which is catalysed by HIV integrase. Furthermore, HIV and HCV differ in specific aspects of viral entry and assembly that may offer unique therapeutic opportunities. The following sections consider the latest progress in the development of drugs for HIV and HCV. These are grouped according to the stage of the viral life cycle they target and in approximate chronological order with regard to their development.
Figure 1 | Simplified flow charts of the life cycles of human immunodeficiency virus (HIV) and hepatitis C virus (HCV).
Established and potential targets for therapeutic intervention are highlighted, together with the class of antiviral drugs. Viral entry into the cells requires, for both HIV (a) and HCV (b), the interaction with a number of specific receptors and co-receptors, followed by virus–host cell fusion for HIV and endocytosis for HCV. Reverse transcriptase has a crucial role in the replicative cycle of HIV, as it is responsible for the transcription of the viral single-stranded (+)RNA genome to proviral double-stranded DNA, which is then integrated into the host cell genome with the help of the virally encoded integrase enzyme. Replication of the HCV genome depends on RNA replicase, which subsequently converts genomic (t)RNA into (-) and back to (+)RNA. The messenger (t)RNAs formed during the replicative cycle of HIV and HCV are translated to viral precursor proteins that are then cleaved by a virus-encoded protease into mature structural and functional proteins. For HIV, this maturation continues after the virus particles have been released (by budding) from the cells. CLDN1, claudin-1; GAG, glycosaminoglycans; LDLR, low-density lipoprotein receptor; NNRRI, non-nucleoside reverse replicase inhibitor; NNRTI, non-nucleoside reverse transcriptase inhibitor; NRRI, nucleoside reverse replicase inhibitor; NRTI, nucleoside reverse transcriptase inhibitor; NtRTI, nucleotide reverse transcriptase inhibitor; SR-BI, scavenger receptor class B type.
Viral genome replication inhibitors
Although HIV and HCV are different in their genomic structure, the pathogenesis and clinical symptoms they engender reveal related targets for therapeutic intervention, such as their polymerases (that is, RNA-dependent DNA polymerase or reverse transcriptase for HIV; RNA-dependent RNA polymerase or RNA replicase for HCV) and proteases (that is, aspartyl protease for HIV; serine protease for HCV). Many of the issues that had to be addressed during the development of HIV inhibitors, particularly those directed at the reverse transcriptase (nucleoside reverse transcriptase inhibitors (NRTIs) and non-nucleoside reverse transcriptase inhibitors (NNRTIs)) and the HIV protease, are likely to re-appear in the development of HCV inhibitors targeted at the RNA replicase (nucleoside and non-nucleoside reverse replicase inhibitors (NRRIs and NNRRIs)) and the HCV protease. These issues will probably concern pharmacokinetics, bioavailability and safety, and most importantly, drug resistance emergence, which could necessitate the combination of different drugs for HCV, as has proved mandatory for HIV.
HIV reverse transcriptase as a target for NRTIs and NtRTIs. The first anti-HIV drugs discovered (such as zidovudine) targeted the HIV reverse transcriptase, which offers two target sites for inhibitors: the catalytic substrate (dNTP) binding site, and an allosteric site, which is distinct from (yet closely located to) the substrate binding site9, 10 (Fig. 2 a,b).
Figure 2 | Human immunodeficiency virus (HIV) reverse transcriptase.
a | Three-dimensional structure of HIV-1 reverse transcriptase, with the fingers, palm, thumb, connection and RNase H domains, belonging to the p66 subunit, and the p51 subunit, based on Ref. 9. b | HIV-1 reverse transcriptase complexed with the DNA template primer. The reverse transcriptase heterodimer consists of the p66 subunit (dark blue) and the p51 subunit (light blue). The two magnesium ions in the active site are shown as purple balls. The side-chains of active-site amino acids Y183, M184, D185, D186 and D10 are represented as green-coloured van der Waals spheres. Residues of the non-nucleoside reverse transcriptase inhibitor binding site (L100, K101, K103, V106, V108, V179, Y181, Y188, P225, F227, W229, L234, P236 and Y318) are represented as yellow-coloured van der Waals spheres10. Figure reproduced with permission from Ref. 10 © (2004) Elsevier Science Ltd.
NRTIs require three phosphorylation steps, catalysed by cell-derived kinases, to be converted to their active metabolites, the triphosphates of zidovudine or the other 2',3'-dideoxynucleosides11, 12. The active metabolites then act as competitive inhibitors or alternative substrates with respect to the normal substrates (either dATP, dGTP, dCTP or dTTP) and lead to the termination of chain elongation (Fig. 3a). So far, seven NRTIs (zidovudine, didanosine, zalcitabine, stavudine, lamivudine, abacavir and emtricitabine) have been approved by the US Food and Drug Administration (FDA) for the treatment of HIV.
Figure 3 | Nucleoside and non-nucleoside reverse transcriptase inhibitors.
a | Mechanism of action of nucleoside reverse transcriptase inhibitors (NRTIs), as exemplified for 2',3'-dideoxycytidine. Three phosphorylation steps convert the 2',3'-dideoxynucleoside to its 5'-triphosphate derivative; the latter will act as a chain terminator in the reverse transcriptase reaction owing to the absence of the 3'-OH group that is required for further chain elongation. b | Structural formulae of selected NRTIs. c | Structural formulae of selected non-nucleoside reverse transcriptase inhibitors (NNRTIs).
Nucleotide reverse transcriptase inhibitors (NtRTIs) have a similar mode of action to NRTIs, but the presence of the phosphonate group, which is analogous to a phosphate group, means that only two phosphorylation steps by cellular kinases are required for conversion to the active metabolite. NtRTIs are therefore able to bypass the nucleoside-kinase reaction, which can limit the activity of the dideoxynucleoside analogues against HIV. In addition, unlike phosphate, phosphonate can no longer be cleaved by the esterases that would normally convert nucleoside monophosphates back to their nucleoside form. At present, only one NtRTI, tenofovir, has been approved for the treatment of HIV.
Given the properties of existing drugs, for new NRTIs and NtRTIs, it will become increasingly difficult to comply with the demands for higher activity (potency), lower toxicity (side effects) and the more favourable resistance profile required for approval as antiviral drugs. Nevertheless, several NRTIs are currently in clinical development13: racivir ((
)FTC), apricitabine (AVX-754, formerly SPD-754), dexelvucitabine (
-D-Fd4C, formely Reverset), elvucitabine (-L-Fd4C), alovudine (MIV-310, FLT) and amdoxovir (DAPD) (Fig. 3b). Their mechanism of action can be considered to be analogous to that of zidovudine11, in that following successive phosphorylation to the 5'-mono-, 5'-di- or 5'-tri-phosphate, the ddNTP competes with the normal cellular substrate. Upon removal of the , 
-diphosphate, the remaining ddNMP is incorporated at the 3'-end of the DNA chain, leading to chain termination as it does not offer the necessary 3'-hydroxyl function for further DNA elongation (Fig. 3a).
Racivir corresponds to a 50:50 racemic mixture of the (-)--enantiomers and (+)--enantiomers of FTC14 and could therefore be considered as a combination of two different compounds, which should diminish the likelihood of resistance development. Racivir has proven safe and efficacious in a small Phase I trial in previously untreated (naive) patients15, and has now completed Phase II clinical trials.
Apricitabine appears to be one of the more promising new NRTIs; it showed a low propensity to select for resistance mutants in vitro and in vivo following 10-day monotherapy in antiretroviral-naive, HIV-infected patients16. The classical NRTI (or NNRTI) resistance mutations M184V, L74V (and K103N) are not expected to affect the efficiency of chain termination by apricitabine triphosphate. However, the K65R mutation (responsible for reduced sensitivity to tenofovir) may also reduce susceptibility to apricitabine through a reduction in the binding or incorporation of apricitabine triphosphate17. Apricitabine could be useful in treating patients who have failed on previous lamivudine- or emtricitabine-containing regimens18. However, apricitabine should not be combined with lamivudine or emtricitabine, as these might reduce antiviral activity19 and intracellular phosphorylation20 of apricitabine.
Amdoxovir proved safe and effective in a short-term (15 day) study in HIV-infected patients21, but did not add statistically significant antiretroviral activity at 24 weeks when added to enfuvirtide plus optimized background therapy (OBT) in advanced subjects with highly resistant virus22. Current attempts are now directed at the design of prodrugs of amdoxovir with alkyl or valyl substitutions at the N'-6 or C'-5 position, respectively23.
For alovudine, a recent study demonstrated that a 4-week course with the compound at 2 mg per day provided a modest but significant viral load reduction in patients harbouring viruses with a median of 4 thymidine-associated mutations (TAMSs)24. Whether alovudine will be of use in treating highly experienced patients remains to be determined25.
HIV non-nucleoside reverse transcriptase inhibitors. NNRTIs, which were discovered in 1990, interact with an allosteric binding site on HIV-1 reverse transcriptase that becomes exposed upon ligand binding26. NNRTIs are a key part of typical HAART regimes for treatment-naive patients (two NRTIs and one NNRTI), owing to their potency, favourable safety profile and ease of dosing. However, the relatively rapid emergence of resistance, resulting from mutations at amino-acid residues that surround the NNRTI binding site (in particular K103N and Y181C), is a serious limitation.
In addition to the three NNRTIs — nevirapine, delavirdine and efavirenz — that have been approved for the treatment of HIV, a few more are in clinical development13, including rilpivirine, etravirine and dapivirine (Fig. 3c). Their mechanism of action is similar to that of the approved NNRTIs in that they interact with a specific binding site of the reverse transcriptase, thereby blocking the enzyme's activity. The positioning of the NNRTI at its binding site is exemplified for etravirine (TMC 125) (Supplementary information S2 (figure)). It has recently been demonstrated that NNRTI binding to the polymerase domain of the reverse transcriptase interferes with RNase H activity, and that mutations in the NNRTI binding site (K103N, Y181C, Y188L and K103N/Y181C) reduce the potency of RNase H inhibition26.
Perhaps the most promising investigational anti-HIV drug is rilpivirine (R278474, TMC278). When given at a dose of 25, 50, 100 or 150 mg once daily for 7 days to antiretroviral-naive HIV-infected subjects, rilpivirine was well tolerated and achieved a viral load decrease of more than 1.0 log10 copies per mL at all doses studied27. Phase III clinical trials have been initiated.
The prospects for the development of etravirine (TMC125), which is now under FDA review, also look favourable. In an open-label randomized clinical trial with two dosages of etravirine (400 or 800 mg twice daily), the compound led to a reduction of 1.04 or 1.18 log10 copies per mL after 24 weeks when added onto an optimized background28. In an unusual drug combination study where etravirine was combined with ritonavir-boosted darunavir (an HIV protease inhibitor), a mean reduction in HIV RNA of 2.7 log10 copies per mL was achieved after 24 weeks29. Furthermore, compared with the first-generation NNRTIs, efavirenz and nevirapine, HIV has a high genetic barrier to the development of resistance against etravirine30.
Dapivirine (TMC120) has been primarily pursued for its potential as a vaginal HIV-1 microbicide31. In a dose-ranging Phase I study in HIV-negative and HIV-positive female volunteers, a vaginal gel containing three different concentrations of TMC120 was well tolerated31. Long-term controlled release of TMC120 from silicone vaginal rings may be achievable32, and it has even been predicted that a protective vaginal concentration of TMC120 may be maintained for at least a year from a single ring device33.
HCV nucleoside RNA replicase inhibitors. The HCV RNA replicase (NS5B, a RNA-dependent RNA polymerase (RdRp)) replicates the HCV genome before viral assembly. Analogous to inhibitors of HIV reverse transcriptase, NS5B inhibitors can also be divided into two categories: nucleoside and non-nucleoside-based RNA replicase inhibitors (NRRIs and NNRRIs, respectively). NRRIs require sequential phosphorylation to their corresponding 5'-mono-, 5'-di- and 5'-triphosphates before entering into competition, at the RNA replicase level, with the natural substrate, CTP (Fig. 4). The NRRIs are assumed to interact as non-obligate chain terminators (unlike NRTIs and NtRTIs, NRRIs possess a 3'-hydroxyl function), and are therefore not classical chain terminators. Instead, they achieve chain termination by interfering with the subsequent elongation step through steric hindrance exerted by the 2'-C-methyl and/or 4'-C-azido groups. The fact that different ways of termination are effective for inhibiting HIV and HCV polymerases, respectively, is an empirical observation, and is related to the differences in DNA versus RNA replication.
Figure 4 | NS5B RNA replicase inhibitors: part 1.
a | Mechanism of action of nucleoside RNA replicase inhibitors (NRRIs) as exemplified for valopicitabine (2'-C-methylcytidine) and 4'-azidocytidine. Three phosphorylation steps convert the nucleoside analogues (that is, 2'-C-methylcytidine or 4'-azidocytidine) to their 5'-triphosphates, which then act as non-obligate chain terminators (in competition with the natural substrate CTP) of the HCV NS5B RNA-dependent RNA polymerase (RdRp) as they interfere with further elongation through steric hindrance. b | Structures of NRRIs.
The most important NRRIs (Fig. 4b) described to date are valopicitabine (the 3'-valine ester of 2'-C-methylcytidine)34, 2'-O-methylcytidine35, 2'-C-methyladenosine35, 36, 2'-C-methylguanosine37, 38, 7-deaza-2'-C-methyladenosine39, 7-deaza-7-fluoro-2'-C-methyladenosine40, 2'-deoxy-2'-fluoro-2'-C-methylcytidine (PSI-6,130)41, 4'-azidocytidine (R1479)42, oral prodrugs thereof (such as R1626)43 and novel 4'-azido-2'-deoxynucleoside analogues (such as RO-0622)44. The presence of a methyl (or fluorine) group in the 2'-C position or an azido group in the 4'-C position is crucial for the anti-HCV activity of NRRIs. It might well be worth exploring the anti-HCV activity of new nucleoside derivatives that contain both of these substitutions combined into the same molecule.
Valopicitabine (NM283) was the first NRRI that proceeded to clinical trials. Fifteen days of valopicitabine treatment at optimal dosing produced consistent HCV RNA reductions averaging more than 1.2 log10 in difficult-to-treat, predominantly nonresponder patients. The combination of valopicitabine with pegylated IFN shows antiviral synergy in preclinical studies, and expanded clinical testing of valopicitabine, alone and in combination with pegylated IFN, is underway45. However, this combination should not be extended to ribavirin, as ribavirin was shown to antagonize the anti-HCV activity of 2'-C-methylcytidine, the active component of valopicitabine46. Whether ribavirin also antagonizes the in vitro activity of other NRRIs such as PSI-6130, R1479 or RO-0622, remains to be explored.
R1479 (4'-azidocytidine) inhibits HCV replication in the subgenomic replicon system with a similar potency to 2'-C-methylcytidine42, and R1479 5'-triphosphate (R1479-TP) showed similar potency to the reference inhibitor 3'-dCTP (an obligate chain terminator). The S282T mutation in the NS5B RNA replicase, which has been shown to confer resistance to 2'-C-methylnucleosides, did not confer cross-resistance to R1479 (Ref. 42). In vitro studies mapped resistance to R1479 to amino-acid substitutions S96T and S96T/N142T of NS5B. These mutations, in turn, did not confer resistance to 2'-C-methylcytidine47, arguing for a combination therapy of R1479 with 2'-C-methylribonucleosides. Significantly higher concentrations of 2'-C-methyladenosine triphosphate were detected inside cells compared with 2'-O-methylcytidine triphosphate, which is consistent with the greater potency of 2'-C-methyladenosine over 2'-O-methylcytidine in the HCV replicon assay35. The relative inactivity of 2'-O-methylcytidine in inhibiting HCV replication could be ascribed to its poor intracellular conversion to the 5'-triphosphate; its activity could be restored by using a monophosphate prodrug36.
The 2'-C-methyl ribonucleosides 2'-C-methyladenosine and 2'-C-methylguanosine were identified as efficient chain-terminating inhibitors of HCV genome replication37, and the corresponding triphosphates were found to be potent inhibitors of the HCV NS5B-mediated RNA synthesis38. Characterization of drug-resistant HCV replicons defined a single S282T mutation within the active site of the viral RNA polymerase that conferred loss of sensitivity to 2'-C-methyl ribonucleosides in both replicon and isolated polymerase assays37.
It has been suggested that at the level of the RNA polymerization reaction, the 2'-C-methyl entity sterically blocks the next incoming ribonucleotide 5'-triphosphate (NTP)37. Additional modifications, such as the 7-deaza modification, may further disrupt the alignment of the 3'-OH for nucleophilic attack on the -phosphate of this incoming NTP, so as to more efficiently terminate chain elongation39.
HCV non-nucleoside RNA replicase inhibitors. In contrast to NNRTIs, which all target the same single binding pocket of the HIV reverse transcriptase, and NRRIs, which interact with the HCV NS5B catalytic site (Fig. 5a,b), NNRRIs can interact at a number of different allosteric sites34. Consequently, many structurally different NNRRIs (Fig. 5c) have been identified: DKA compound 30 (Refs 48,49), a benzimidazole 5-carboxamide derivative50, an indole-N-acetamide derivative51, 52, a benzothiadiazine derivative53, 54, a phenylalanine derivative55, a thiophene 2-carboxylic acid derivative56, 57, a dihydropyrone derivative (J. Tatlock, personal communication), the tetrahydropyranoindolyl acetic acid derivative HCV-371 (Ref. 58), a series of 5-hydroxy-3(2H)-pyridazinones59, 60 and the benzofuran derivative HCV-796 (Ref. 61).
Figure 5 | NS5B RNA replicase inhibitors: part 2.
a | X-ray crystallographic structure of the NS5B RNA replicase. The thumb, palm and finger domains are coloured in blue, green and red, respectively. The two catalytic aspartates (residues 318 and 319) are shown in stick and ball style in the active site. Residues that are responsible for resistance to the various inhibitors are coloured according to the resistance pattern: magenta (benzimidazoles/indoles), light blue (thiophenes), brown (benzothiadiazines), yellow (dihydroxypyrimidines) and orange (2'-C-Me nucleosides)63. Figure reproduced with permission from Ref. 63 © (2005) International Medical Press. b | Binding sites for non-nucleoside RNA replicase inhibitors (NNRRIs) at the NS5B polymerase. Whereas there is only one binding site (S282) for the nucleoside reverse replicase inhibitors (NRRIs), there are at least four binding sites for the NNRRIs (NNI sites 1, 2, 3 and 4 for benzimidazole, thiophene carboxylic acid, benzothiadiazine and benzofuran, respectively)147, 148, 149 (and A. Y. M. Howe, personal communication). Yellow spheres represent amino acid residues at the NNRRI binding sites; turquoise dots represent Mg2+ ions at the catalytic aspartates. Figure is courtesy of Inge Vliegen and Weidong Zhong. c | Structures of NNRRIs.
The X-ray crystallographic structure of HCV NS5B54, 62, 63 allowed the mapping of the binding sites for the NRRIs (2'-C-methyl nucleosides) and NNRRIs (that is, benzimidazoles/indoles, thiophenes, benzothiadiazines and dihydroxypyrimidines) (Fig. 5a,b). As shown for at least one NNRRI, the interaction between this inhibitor and NS5B occurs without dramatic changes to the structure of the protein62.
Before the HCV NS5B RdRp was actively pursued as the target enzyme for development of both the NRRI and NNRRI type of HCV inhibitors, the NS5B of the related pestivirus BVDV (bovine viral diarrhoea virus) was formally validated as a target for antiviral drug discovery and development64 (Supplementary information S3 (figure)). Following the identification of the lead compound VP 32947 (Ref. 64), we described the compound BPIP as an NRRI for pestiviruses such as BVDV65. The latter compound was modelled near position F224 in the RdRp (NS5B) of the BVDV65. Introduction of a fluorine atom in position 2 of the 2-phenyl substituent of the lead antipestivirus compound I (5-[(4-bromophenyl)methyl]-2-phenyl-5H-imidazo[4,5-c]pyridine) resulted in an analogue with selective activity against HCV in the subgenomic replicon system66. Further pursuing this line of research, we have recently identified newly substituted imidazopyridines as potent inhibitors of HCV replication, targeting the viral RNA replicase67.
Viral protease inhibitors
The life cycles of both HIV and HCV include a step in which newly expressed viral precursor polyproteins are proteolytically cleaved by viral proteases into smaller, mature functional or structural viral proteins. The HIV (aspartyl) protease cleaves the group-specific antigen (gag) and gag–polymerase (pol) precursor proteins to structural capsid proteins (matrix antigen (MA p17), capsid antigen (CA p24) and nucleocapsid (NC p7, NC p1)) and functional proteins (protease (p11/p11), reverse transcriptase (p66/p51) and integrase (p32)). The HCV (serine) protease NS3/NS4A cleaves the polyprotein at the peptide linkages between the non-structural proteins NS3, NS4A, NS4B, NS5A and NS5B, the latter corresponding to the RNA replicase. The HIV and HCV proteases are therefore attractive antiviral drug targets.
HIV protease inhibitors. HIV protease inhibitors were a key component of the first HAART regimes, although the adverse metabolic side effects, such as lipodystrophy, and dosing issues of the first-generation protease inhibitors caused them to be overtaken by NNRTIs for treatment-naive patients. However, the newest protease inhibitors, atazanavir, tipranavir and darunavir have been developed with the aim of overcoming some of these issues, and/or to have activity against viruses resistant to other protease inhibitors. All of the currently available HIV protease inhibitors, except for tipranavir, which is based on the coumarin lactone scaffold, can be considered as peptidomimetics (Fig. 6a). They are built upon a hydroxyethylene group, which mimics the peptide linkage(s) in the polyprotein precursor gag–pol, which is cleaved by the HIV protease into the mature capsid proteins and enzymes. The protease inhibitors mimic the structure of its regular substrate, and compete with its binding, thus blocking protease activity. The binding of the protease inhibitor to its active site within the HIV protease dimer is shown for darunavir (TMC114, Prezista) in Fig. 6b.
Figure 6 | Human immunodeficiency virus (Hiv) protease inhibitors.
a | Except for tipranavir, all other HIV protease inhibitors can be considered as peptidomimetic in that they contain the non-scissile hydroxyethylene [-CH2-CHOH-] bond instead of the readily hydrolysable peptide [-CO-NH-] bond. b | HIV protease structure with darunavir (TMC114) in the active site150. Figure reproduced with permission from Ref. 150 © (2006) Elsevier Science Ltd.
Tipranavir, which is not a peptidomimetic analogue, could possess a drug-resistance profile that is different from that of the other HIV protease inhibitors, and might have a different potency. Of note, in two studies conducted in treatment-experienced patients, ritonavir-boosted tipranavir demonstrated superior antiviral activity at week 24, compared with other ritonavir-boosted protease inhibitors68, 69.
Brecanavir (GW640385), another HIV protease inhibitor, was reported to show potent antiviral activity when boosted by ritonavir in HIV-1-infected patients harbouring both protease inhibitor-sensitive and protease inhibitor-resistant virus, following 24 weeks of treatment70. This compound was reported to be over 100-fold more potent than previously marketed protease inhibitors and even 10-fold more potent than the most recently marketed inhibitor, darunavir. In addition, it has been shown to be safe and well-tolerated upon both single-dose71 and repeat72 administration. However, hopes for brecanavir were recently dashed when the company responsible for its development announced "insurmountable issues regarding formulation"73.
HCV serine protease (NS3/4A) inhibitors. The conformation of HCV NS3 serine protease interacting with the NS4A co-factor peptide, which is important for its proteolytic activity, has been the most intensively pursued anti-HCV target. However, its shallow active site has made the design of inhibitors more challenging than for HIV protease inhibitors. Foremost among the HCV serine protease (NS3/4A) inhibitors (Fig. 7a), which also inhibit HCV replication, are ciluprevir (BILN 2061)74, telaprevir (VX-950)75, boceprevir (SCH503034)76, 77, 78 and SCH446211 (also referred to as SCH6)79, 80.
Figure 7 | NS3 hepatitis C virus (HCV) protease inhibitors.
a | Structures of HCV serine NS3 protease inhibitors. b | NS3 protease inhibitor (SCH503034) complexed with its target enzyme. Left panel shows Connolly surface for the NS3 protease. SCH503034 is rendered in CPK format (Corey–Pauling–Koltun space-filling model): gold, carbon; red, oxygen; and blue, nitrogen. Right panel is a close up of the NS3:SCH503034 complex showing side chains that are perturbed upon binding of SCH503034. Side chains of the complex are coloured as follows: carbon, green; nitrogen, blue; and oxygen, red. Side chains of the apoenzyme structure are shown in purple77. Figure reproduced with permission from Ref. 77 © (2006) American Society for Microbiology.
The antiviral efficacy of ciluprevir in HCV-infected patients, as monitored by plasma virus load reductions (upon a treatment period of no longer than 2 days), has been clearly demonstrated74. This efficacy was further confirmed with doses as low as 25 mg ciluprevir administered twice a day in patients infected with HCV genotype 1a or 1b81. However, HCV genotypes 2 and 3 exhibited a markedly reduced sensitivity to ciluprevir82. Replacement of five residues of HCV NS3 at positions 78, 79, 80, 122 and 132 accounted for most of the reduced sensitivity of genotype 2b, whereas replacement of residue 168 alone could account for the reduced sensitivity of genotype 3a82. Substitution of amino acids in HCV NS3 genotype 1, based on genotype 2 and 3, has revealed residues that affect the binding of ciluprevir. Substitution of residues 78–80, together with 122 and 132, accounted for most of the reduced sensitivity of genotype 2. The most critical position affecting inhibitor binding to genotype 3 protease was 168. Substitution of residues at positions 168, 123 and 132 fully accounted for the reduced sensitivity of genotype 3 (Ref. 83).
The most advanced HCV protease inhibitor is telaprevir. Exposure of cells to telaprevir resulted in a more than 4 log10 reduction in the HCV RNA levels after a 2-week incubation of replicon cells, with no rebound of viral RNA observed after withdrawal of the inhibitor. In several animal species, telaprevir exhibited a favourable pharmacokinetic profile with high exposure in the liver. Also, in a recently developed HCV protease mouse model, telaprevir showed excellent inhibition of HCV NS3/4A protease activity in the liver. Therefore, the overall preclinical profile of telaprevir supports its candidacy as a novel oral therapy against hepatitis C84.
The combination of telaprevir and IFN was additive to moderately synergistic in reducing HCV RNA in replicon cells with no significant increase in cytotoxicity. The benefit of the combination was sustained over time: a 4 log10 reduction in HCV RNA levels was achieved following a 9-day incubation with telaprevir and IFN at lower concentrations than when either telaprevir or IFN was used alone. The combination of telaprevir and IFN also suppressed the emergence of in vitro resistance mutations against telaprevir in replicon cells85. In fact, combination of telaprevir with pegylated IFN would seem mandatory to avoid the development of resistance to telaprevir86.
The major ciluprevir-resistant mutations at N168 were fully susceptible to telaprevir, and the dominant resistant mutation against telaprevir at A156 (A156S) remained sensitive to ciluprevir. Modelling analysis suggests that there are different mechanisms of resistance to telaprevir and ciluprevir75. The cross-resistance towards ciluprevir and telaprevir conferred by substitution of A156 with either valine or threonine was confirmed by characterization of the purified enzymes and reconstituted replicon cells containing the single amino-acid substitution A156V or A156T. Both of these cross-resistance mutations displayed significantly diminished fitness (or replication capacity) in a transient replicon cell system87. Besides A156V and A156T, the R155K mutation has also been reported to confer low-level resistance to telaprivir (greater than 25-fold) while retaining sensitivity to IFN and showing reduced replication capacity88.
HCV RNA replicons that are resistant to the novel ketoamide inhibitor of the NS3/4A protease, SCH6 (originally SCH446211), have also been reported79. Resistant replicon RNAs were generated by G418 selection in the presence of SCH6 in a dose-dependent fashion; the emergence of resistance was reduced at higher SCH6 concentrations. Sequencing demonstrated remarkable consistency in the mutations conferring SCH6 resistance in genotype 1b replicons derived from two different strains of hepatitis C virus, A156T/A156V and R109K. The novel mutation R109K had not been reported previously to cause resistance to NS3/4A inhibitors; it conferred moderate resistance only to SCH6. Structural analysis indicated that this reflects unique interactions of SCH6 with P'-side residues in the protease active site.
By contrast, A156T conferred high-level resistance to SCH6 and the related ketoamide SCH503034 (boceprevir) (Fig. 7b), as well as ciluprevir and telaprevir. Unlike R109K, which had minimal impact on NS3/4A enzymatic function, A156T significantly reduced NS3/4A catalytic efficiency, polyprotein processing and replicon fitness. However, three separate second-site mutations, P89L, Q86R, and G162R, were capable of partially reversing A156T-associated defects in polyprotein processing and/or replicon fitness, without significantly reducing resistance to the protease inhibitor79. The A156T mutation is undoubtedly a key mutation in the resistance of HCV to NS3/4A inhibitors; together with the R155Q, D168A and D168V mutations, the A156T mutation may provide further insights into the mechanism of HCV drug escape, as well as into the rational design of novel protease inhibitors89.
Viral entry inhibitors
Cell entry inhibitors that interact with either fusion or co-receptor binding of the virus with the host cell have been described for HIV, however as HCV entry inhibitors cannot be detected by using the replicon systems, similar cell entry inhibitors for HCV remain to be identified. HIV entry into CD4+ cells requires successive non-specific interactions with cell surface heparan sulphates, followed by specific binding to the CD4 receptor and the CXCR4 or CCR5 co-receptor, before the viral envelope fuses with the plasma cell membrane and the viral nucleocapsid enters the cells90, 91 (Figs 8,9). The binding of HCV to the cell surface and its entry into the cells may involve the low-density lipoprotein receptor (LDLR), glycosaminoglycans (GAG), scavenger receptor class B type (SR-BI, also known as SCARB1), the tetraspanin protein CD81 and claudin-1 (CLDN1), before internalization occurs through clathrin-mediated endocytosis93.
Figure 8 | Inhibiting human immunodeficiency virus (HIV) fusion.
When HIV infects a CD4+ T cell (a), the viral glycoprotein gp120 first interacts with the CD4 receptor, then with the CCR5 or CXCR4 co-receptor, upon which the viral gp41 will bring the viral envelope in contact with the host cell membrane (b). The gp41 glycoprotein contains four major functional domains: starting from the N terminus towards the C terminus these are the fusion peptide, the heptad repeat 1 (HR1), the heptad repeat 2 (HR2) and the transmembrane domain that anchors gp41 into the viral lipid bilayer. Enfuvirtide is homologous to part of the HR2 region. When the N terminal fusion peptide of gp41 is inserted into the host cell membrane, the three HR2 domains of the gp41 trimer loop back in a triple hairpin and 'zip' themselves into three highly conserved hydrophobic grooves on the outer face of the HR1 trimeric bundle to form a six-helix bundle that pulls the outer membranes of the virus and the cell into close physical proximity, thus enabling the two membranes to fuse13. This process depends on an interaction of the heptad repeat HR2 with HR1. By being homologous to the HR2 domain, enfuvirtide blocks this interaction90.
Figure 9 | Human immunodeficency virus (HIV) co-receptor antagonists.
When the HIV glycoprotein gp120 binds to CD4 (a), it induces a conformational change in gp120 that exposes the co-receptor binding site (b); this is a complex domain comprising the V3 loop and specific amino-acid residues in CD4, collectively termed the 'bridging sheet'. Exposure of the co-receptor binding site permits binding of gp120 to the co-receptor (c). Co-receptor antagonists inhibit this step by binding to the co-receptor and changing its shape so that gp120 cannot recognize it. Co-receptor binding induces conformational changes in gp41 and insertion of the fusion peptide into the host cell membane (d), ultimately resulting in fusion of the viral envelope with the host cell membrane91. (e) Structural formulae of selected CCR5 antagonists.
HIV fusion inhibition. The first viral entry inhibitor to be approved, enfuvirtide (DP-178, T-20), is a peptide corresponding to the amino-acid residues 643–678 of the HIV-1 envelope precursor glycoprotein gp160. This precursor glycoprotein is cleaved by cellular serine proteases to yield a noncovalent gp120–gp41 heterodimer13. The gp41 glycoprotein is responsible for the fusion of the viral envelope with the cell membrane, a process that is inhibited by enfuvirtide90 (Fig. 8).
Due to its peptidic nature, and unlike all other anti-HIV drugs, enfuvirtide has to be subcutaneously injected twice a day93. It is generally used in combination with HAART regimens, and conveys an incremental benefit if added to highly active three-drug regimens94. The virological and immunological response obtained at an early time point (week 12) during enfuvirtide-based therapy is predictive of the responses at weeks 24, 48 and 96 in the T-20 versus optimized regimen only (TORO) trials95. This may be highly beneficial in treatment-experienced patients with limited options.
However, HIV has a low genetic barrier to resistance to enfuvirtide96, with HR1 mutations at codon 36 (G36E, G36D or G36S) and 38 (V38A, V38G or V38M) being the most commonly detected resistance mutations at week 2, and mutations at codon 40 (Q40H) and codon 43 (N43D) being more prevalent at week 4. This low genetic barrier to resistance underscores the importance of combining enfuvirtide with other antiretrovirals96. However, depending on the envelope genetic context, mutations in the HR1 domain may not always lead to phenotypic resistance97; furthermore, these HR1 mutations do not affect susceptibility of HIV-1 to other classes of viral entry inhibitors, including co-receptor (CCR5 and CXCR4) inhibitors98. On the contrary, the activity of enfuvirtide against strains of HIV-1 that use CCR5 as co-receptor may be enhanced by compounds such as rapamycin, which have been shown to downregulate CCR5 expression99, 100.
HIV co-receptor antagonists. The human chemokine receptors CCR5 and CXCR4 are attractive targets for small-molecule antagonists that inhibit HIV-1 infection. In the overall process of viral entry, the point of attack of such antagonists occurs after the HIV envelope gp120 has interacted with its primary receptor CD4, but before the HIV envelope gp41 inserts the fusion peptide into the host cell membrane (Fig. 9).
HIV-1 strains can be categorized by co-receptor tropism; that is, their ability to use CCR5 or CXCR4, or both for entry into cells. CCR5 has received more attention as a potential target, because a natural deletion (
32) in the CCR5 gene confers natural resistance to infection by CCR5-tropic strains, which are the most frequently sexually transmitted strains71. CXCR4 has received less attention, although a switch from CCR5 to CXCR4 tropism may occur spontaneously in approximately 50% of HIV-infected patients and has been associated with, but is not required for, disease progression91. The most advanced CXCR4 antagonist, AMD3100 (Mozobil) has not been further pursued as a therapeutic modality for HIV infections, but as a highly selective CXCR4 antagonist it has proved to be particularly promising for the mobilization of haematopoietic stem cells in patients with haematologic malignancies such as non-Hodgkin's lymphoma (NHL)101.
Three CCR5 antagonists have been evaluated in clinical trials — maraviroc, vicriviroc and aplaviroc (Fig. 9e). However, the development of aplaviroc was halted in 2005, following the observation of elevated liver enzymes and total bilirubin in one of the HIV-1-infected patients receiving the compound91.
Maraviroc is highly selective for CCR5, as confirmed against a wide range of receptors (including the hERG ion channel), indicating potential for an excellent clinical safety profile102. It was FDA-approved in August 2007 (Selzentry), and a 10-day monotherapy with maraviroc at twice daily doses greater than, or equal to, 100 mg achieved a mean reduction of approximately 1.6 log10 viral RNA copies per mL at a median of 10–15 days, thus providing proof-of-concept that CCR5 antagonism is a viable antiretroviral therapeutic approach103. CXCR4-tropic variants are obviously not sensitive to maraviroc, and such variants may emerge when the CCR5-tropic virus is suppressed104. However, there is no evidence for a co-receptor switch from CCR5 to CXCR4 under selection pressure from maraviroc104, and virus strains that have become resistant to maraviroc still use inhibitor-bound receptor for entry105. Although complex, this pathway to resistance presents less of a barrier to escape from maraviroc than the switch to CXCR4 tropism.
Like maraviroc, vicriviroc binds specifically to the CCR5 receptor and prevents infection of target cells by CCR5-tropic HIV-1 strains106. During 14-day monotherapy with vicriviroc at doses of 25 or 50 mg twice daily, reductions of 1.0–1.5 log10 HIV RNA copies per mL were achieved107. In another study108 in HIV-1-infected treatment-experienced patients, the reduction in viral load was sustained throughout a 24-week treatment period with doses as low as 10 or 15 mg. Malignancies occurred in a number of patients10; however, the relationship of vicriviroc to malignancy has remained uncertain. As has been noted for maraviroc, vicriviroc-resistant viruses may enter target cells by using the inhibitor-bound form of CCR5109.
Infection with dual (CCR5 and CXCR4)-tropic or mixed HIV-1 populations that use both CCR5 and CXCR4 is common among highly-treatment experienced patients, whereas infection with virus that uses CXCR4 alone is uncommon110. HIV-1 that uses CXCR4 exclusively (X4 virus) is not sensitive to CCR5 inhibitors, but how the dual-tropic or mixed X4/R5 HIV-1 populations respond to these compounds in vivo remains to be further explored. Ultimately, combinations of CCR5 inhibitors and CXCR4 inhibitors will be needed to cope with the dual-tropic or mixed X4/R5 HIV-1 populations.
Also worth exploring are combinations of the small-molecule CCR5 inhibitors (that is, maraviroc and vicriviroc) with monoclonal antibodies against CCR5, as potent synergy was observed between small-molecule CCR5 antagonists and some of the monoclonal antibodies that recognize different epitopes on CCR5111, 112. These findings suggest that monoclonal antibodies combined with small-molecule CCR5 inhibitors could be a promising new strategy for HIV-1 therapy.
Other HIV and HCV targets
The viral life cycles of HIV and HCV offer several other attractive targets for intervention strategies. For HIV, raltegravir (Isentress), which was FDA approved in October 2007, is the first member of the new drug class of viral integrase inhibitors, and bevirimat, a first-in-class HIV maturation inhibitor, is currently undergoing Phase IIb clinical trials. Further investigational strategies include cyclophilin inhibitors for both HIV and HCV, and internal ribosome entry site (IRES) and ion channel inhibitors for HCV.
HIV integrase inhibitors. Fifteen years after the search for HIV-1 integrase inhibitors started, these compounds are becoming a therapeutic reality. MK-0518 (raltegravir) and GS-9137 (elvitegravir) (Fig. 10a) are foremost among the various integrase inhibitors that have been described (S-1360, GSK-364735, L-870810, L-870812, MK-0518, GS-9137, L-900564, GS-992, and BMS-707035), and raltegravir was recently approved by the FDA. Both compounds inhibit the strand transfer reaction in the integration process, a crucial step in the stable maintenance of the viral genome, as well as efficient viral gene expression and replication. Unlike integrase inhibitors such as L-870810 and its derivative MK-0518, which can be viewed as diketo acids (bioisosteres), the core structure of elvitegravir is derived from the quinolone antibiotics113. It was therefore gratifying to note that in a newly developed cellular assay, elvitegravir could be validated as a genuine HIV integrase inhibitor114. Amino-acid residues associated with the emergence of reduced susceptibility to elvitegravir have been mapped on the HIV-1 integrase monomeric structure (HIV-1 integrase is assumed to be biologically active as a tetramer) (Fig. 10b).
Figure 10 | Human immunodeficiency virus (HIV) integrase inhibitors.
a | Structural formulae of raltegravir (MK-0518) and elvitegravir (GS-9137, JTK-303). b | HIV-1 integrase monomer structure and amino-acid residues associated with the emergence of reduced susceptibility to GS-9137 in vitro. Integrase monomer crystal structure, as based on Ref. 151; residues associated with the emergence of reduced susceptibility to elvitegravir, as defined in Ref.152. The catalytic core domain is shown in green, the C terminal domain in cyan and the DDE catalytic triad in white. Orange and yellow residues denote resistance selections.
The most relevant property of raltegravir is its impressive potency; a 10-day monotherapy study resulted in a 2 log10 reduction in viral load115. Another Phase II study in which 203 naive patients were randomized to receive lamivudine and tenofovir plus efavirenz or raltegravir showed faster viral suppression with raltegravir (at week 4, 60–80% of patients achieved a viral load of less than 50 copies per mL versus 25% with efavirenz) with a lower rate of adverse events116. In another multicentre, triple-blind, dose-ranging randomized study, raltegravir added onto an optimized background regimen at three doses (either 200 mg, 400 mg or 600 mg given orally twice daily), was found to provide better viral suppression than placebo at all doses117.
Initial clinical trials with elvitegravir have indicated that monotherapy for 10 days with twice-daily doses of 400 or 800 mg, or once daily dosing of 50 mg plus ritonavir, can achieve HIV-1 RNA reductions up to 2 log10 copies per mL in treatment-naive and treatment-experienced patients, with an adverse event profile that is similar to placebo118. No clinically relevant drug–drug interactions were noted when ritonavir-boosted elvitegravir was combined with emtricitabine/tenofovir disoproxil fumarate119. Further studies have corroborated the initial findings that elvitegravir (125 mg) can achieve more than 2 log10 reductions in viral load after 24 weeks when combined with other active drugs120.
HIV maturation inhibition. Bevirimat (3-O-(3', 3'-dimethylsuccinyl)-betulinic acid; PA-457; Fig. 11a) has been described as a maturation inhibitor121, as it disrupts a late step in HIV-1 gag processing that involves the conversion of the capsid precursor p25 (CA-SP1) to mature capsid protein p24 (CA) by inhibiting the CA-SP1 cleavage (Fig. 11b)122. Virions generated in the presence of bevirimat exhibit aberrant capsid morphology and are no longer infectious. Resistance mutations to bevirimat have been found at the p25 to p24 cleavage site: three at or near the C-terminus of CA (CA-H226Y, CA-L231F and CA-L231M) and three at the first and third residues of SP1 (SP1-AIV, SP1-A3T and SP1-A3V)122.
Figure 11 | Human immunodeficiency virus (HIV) maturation inhibition.
a | Structural formula of bevirimat (PA-457). b | Bevirimat (PA-457) resistance mutations. The precursor Pr55Gag protein is represented at the top, with the matrix (MA), capsid (CA), nucleocapsid (NC) and p6 domains and the spacer peptides (SP1 and SP2) indicated. The alignment shows each of the six potential PA-457 resistance mutations that have been identified122.
Bevirimat has so far only been evaluated for its safety and pharmacokinetics in healthy volunteers123, 124, and repeated dosing appeared to be well tolerated123. It is well absorbed after oral administration and its half-life is unexpectedly long (60–80 hours)124, which may facilitate infrequent (that is, twice weekly) dosing. Bevirimat (Fig. 11a) can be glucuronidated, thus facilitating its elimination by the liver, and two monoglucuronides and one diglucuronide have been isolated125. Berivamat has undergone Phase IIa clinical trials with once-daily oral dosing of the compound for 10 days. A Phase IIb clinical trial is currently taking place (www.panacos.com).
HIV and HCV cyclophilin inhibitors. More than a decade ago, cyclophilin A (Cyp A) was recognized as a target for the anti-HIV activity of cyclosporin A and non-immunosuppressive analogues thereof126. Cyp A was found to bind specifically to the HIV-1 gag polyprotein127 and was also reported as a functional regulator of the HCV NS5B RNA replicase128. According to the anti-HCV strategies proposed by Tan et al.129, cyclophilin inhibitors fit into the category of membrane-association intervention. Cyclophilins are peptidyl-prolyl cis-trans isomerases (PPIases) involved in protein folding and function. Cyclophilin B (CypB, also known as PPIB) was recently shown to function as a stimulatory regulator of NS5B128. In fact, CypB enhanced the ability of NS5B to bind to its template, and thus facilitated RNA replication130. This was abolished in the presence of cyclosporin A, suggesting that it could be considered an attractive drug candidate for the treatment of HCV. However, the non-immunosuppressant derivative of cyclosporin A, Debio-025 (Supplementary information S4 (figure)), may serve as a better option in this regard, as it combines a non-immunosuppressive effect with strong inhibitory activity on HCV replication131, 132. Furthermore, Debio-025 also inhibits HIV replication in cell culture (R.G. Ptak, personal communication).
Other investigational strategies for HCV. Initiation of the HCV genomic mRNA translation process may be blocked by inhibitors of the internal ribosomal entry site (IRES), and both the transcription and translation of the viral mRNA may be considered as targets for ribozymes, antisense oligomers133, small interfering (si)RNAs134 and short hairpin (sh)RNAs135. The HCV p7 protein, which forms ion channels in lipid membranes, could be inhibited by long-alkyl-chain iminosugar derivatives that possess antiviral activity against the HCV surrogate bovine viral diarrhea virus (BVDV), so p7 might also be a potential target for anti-HCV therapy136. Glucosidase inhibitors could affect viral morphogenesis, and hence viral infectiousness137, and compounds like arsenic trioxide might inhibit HCV replication by an as yet unknown, probably multipronged, mechanism of action138.
Conclusions
Now, 20 years after the first HIV drug, AZT, was approved, the number of FDA-approved HIV drugs has increased to 24. For HCV, which was identified 6 years after HIV, drug development is now at a similar, if not more, rapid pace than anti-HIV drug development has been. With the advent of the HIV integrase inhibitors (raltegravir and elvitegravir), the CCR5 inhibitors (maraviroc and vicriviroc) and the new NNRTI etravirine that could be used for patients as rescue therapy, Hatano and Deeks139 speculated that 2007 may be comparable to the landmark events of 1996, when the near miraculous effects of HIV combination therapy were first observed. Stephenson140 wrote that researchers are now buoyed by novel HIV drugs, referring to the fact that raltegravir and maraviroc plus OBT demonstrated significantly greater activity compared with placebo added onto OBT in antiretroviral-experienced patients. This was further echoed by the Swedish recommendations for antiretroviral treatment of HIV infection 2007 (Ref. 141).
Is this hype or is the hope for multidrug-resistant HIV-infected patients justified? Although the integrase inhibitors and the CCR5 antagonists interact with novel targets, and the second-generation NNRTIs are active against viruses that are resistant to the first generation, caution should be exercised. For CCR5 antagonists, there is the liability that those viruses that use the CXCR4 co-receptor to enter the cells are not sensitive, and thus may outgrow in the presence of CCR5 inhibitors. For the integrase inhibitors, there is the possibility of rapid drug-resistant strain emergence, as has been witnessed with reverse transcriptase inhibitors and, in particular, with the first-generation NNRTIs. Regarding the novel NNRTIs etravirine and rilpivirine, their efficiency in the clinic against drug-resistant (for example, Y181C or newly emerging) mutants still needs to be substantiated.
For all new HIV drugs, it will remain a research priority to attempt to synthesize new congeners with improved sensitivity/resistance profiles, as exemplified for the NNRTIs142, 143. This is equally true for the integrase inhibitors, where the development of novel inhibitors could be concomitant with the recognition of the clinical impact of current integrase inhibitors (raltegravir and elvitegravir) and the importance of resistance mutations, which are inevitably going to arise against HIV integrase inhibitors.
Future anti-HIV chemotherapy will continue to be based on the judicious choice of drug combinations, which result in higher potency with less toxicity and lower risk for drug resistance development. As different compounds emerge as potential drugs for the treatment of HIV infection, unexpected (and unfavourable) drug–drug interactions might be observed (for example, between tipranavir/ritonavir and enfuvirtide144), and some classes of anti-HIV agents (that is, protease inhibitors) could fall out of favour due to the risk of myocardial infarction145. Similarly, protease inhibitors could be discouraged from entering multiple-drug combinations with, for example, NNRTIs as a regime for treatment-naive patients with HIV146. As a rule, future drug combination schedules should be guided by the successes (efficacy and safety) of what has been achieved with drug combinations in the past, and complemented by the new drugs, once their efficacy has been demonstrated.
The potential opportunities for the future treatment of (chronic) HCV infection are remarkably similar to the combination regimens currently used for the therapy of HIV infection. The treatment of HCV infection, which is currently based on the combination of pegylated IFN with ribavirin, may in the future evolve to include nucleoside analogues that may be referred to as NRRIs, NNRRIs and HCV protease inhibitors. Potential drug candidates have already been identified for each of these three classes. It is likely that, as for the HAART for AIDS, we may soon witness the 'HAART' ('highly active anti-RNA viral therapy') for hepatitis C.
The ultimate goal in the therapy of any virus infection is the elimination of the virus from the organism. For HIV, which hides away in its proviral DNA form, the eventual eradication of the virus may not be achievable. However, as HCV is an RNA virus that does not replicate through a DNA intermediate, the prospect of a true cure seems much more realistic, and the combined use of the compounds that have been developed for the 'HAART' of HCV should help to accomplish this goal.
