CLINICAL SIGNIFICANCE OF THE CELL DEATH INDUCED BY PR
Discordance
The initiation of highly active antiretroviral therapy (HAART) in HIV-infected patients most commonly results in a decrease in plasma HIV viral load (a virologic response) with a corresponding increase in CD4 T cell count and CD4 percentage (an immunologic response).
Within 2 years of the introduction of protease inhibitor (PI)-based antiretroviral therapy, discordant responses between the immunologic and virologic effects of antiretroviral therapy were reported.36 Discordant responses can involve a virologic reponse with immunologic failure (inhibition of viral replication and decrease in HIV plasma viral load, with no increase in CD4 count or percentage). Conversely, virologic failure with immunologic response (ongoing viral replication accompanied by lack of CD4 T cell loss or even CD4 T cell increase) may also occur.Soon after the initial recognition of discordant virologic responses, improved clinical outcomes were observed in patients with virologic failure but immunologic response to antiretroviral therapy. One retrospective analysis of a cohort of patients with an earlier nucleoside reverse transcriptase
inhibitor (NRTI) treatment and CD4 counts virologic response.39 These and other studies reporting sustained increases in CD4 T cells and significant clinical benefit despite this ongoing detectable viral replication37-42 generated important data challenging the view that the beneficial immunological effect of PI-based antiretroviral therapy was solely due to the suppression of ongoing viral replication.
Understanding the mechanisms behind these discordant responses to PI-based ART may provide important insight into T cell biology and the causes of CD4 cell loss during HIV infection. A variety of reasons behind discordant responses to ART has been proposed, including impaired replicative (and possibly pathogenic) fitness of a resistant virus; direct cytotoxic effects of HIV protease; viral escape, leading to autoimmunization and enhanced cytotoxic T cell function; and possibly other unidentified mechanisms.
An emerging body of literature indicates that a drug-resistant virus (especially a PI-resistant virus) has impaired replicative potential compared with a wild-type virus.43-49 Such studies use laboratory strains of HIV with point mutations in PR or clinical isolates containing PR mutations and compare replicative capacity between the mutant and wild-type virusus, either in simple replication assays or in competition experiments. More recently, these experiments have been performed in severe combined immunodeficient (SCID)-hu Thy/LIV mice, activated peripheral blood mononuclear cells, and cultures of thymic organ specimens from humans. In the overwhelming majority of cases, an impaired replicative fitness of mutant virus was shown.50 It is, therefore, reasonable to suspect that such mutations, which result in diminished replicative capacity, might also be associated with impaired pathogenic capacity, which manifests clinically as decreased propensity for CD4 T cell depletion.
Protease inhibitor resistance
HIV protease is a homodimer of two identical 99 amino acid chains with an active site at positions 25 to 27 in each chain. This enzyme cleaves the Gag (p55) and Gag-Pol (p160) polyproteins to yield the HIV core proteins p6, p7, p17, and p24 and the enzymes reverse transcriptase, integrase, and protease.51 The critical role of HIV protease in the replication cycle of HIV via the production of mature virions was recognized as a target for antiretroviral drug therapy. In 1995, HIV protease inhibitors, the third class of antiretroviral agents for treatment of HIV infection, were introduced.52 PIs exert their therapeutic effect by selectively binding to and inhibiting the catalytic activity of HIV protease, thereby preventing cleavage of the large Gag and Gag-Pol polyprotein precursors. Shortly after introduction of PIs into routine use for the care of HIV-infected individuals, antiretroviral combination therapy that included PIs was recognized as having a profound ability to alter the progressive course of HIV infection.
Clinical research analyzing the survival of patients treated with these drugs confirmed these observations.53 In the 9 years after their introduction, PIs have remained potent agents against HIV and, when used in combination with other drug classes, form the cornerstone of current ART against HIV.Similar to the previously recognized phenomenon of viral resistance to NRTIs and non-NRTIs (NNRTIs) development of resistance to PIs was soon recognized to occur commonly and to adversely affect the susceptibility of HIV to PI therapy.54 High levels of ongoing viral replication and turnover in the steady state of HIV infection, as well as the highly error-prone mechanism of HIV reverse transcriptase that lacks a proofreading mechanism, combine to produce frequent mutations in the population of newly produced virions. In addition, recombination of genetic material among HIV strains increases the genetic diversity of the HIV population within a host.55
Currently available combination ART does not fully suppress viral replication in all patients. Ongoing viral replication during periods of drug exposure may be associated with drug concentration sufficient to exert selective pressure on the population of HIV quasispecies present. In this environment, viral quasispecies with resistance to antiretrovirals emerge.56 Extensive cross-resistance among PIs also commonly occurs and has been associated with a genotypic basis for class resistance to available PIs.57 Ultimately, an HIV mutational variant is selected that exhibits a survival advantage over wild-type virus, at least when that virus remains in the continued presence of the antiretroviral selective process. In addition to altering substrate specificities, these resistance-associated mutations typically confer altered enzyme kinetics due to their location in or around the active site of the HIV PR homodimer.
In a variety of different T cell lines, viral replication of virions containing mutant protease (assessed using indirect immunofluorescence assay) was impaired tenfold in the mutant virus compared to that in the wild-type virus; processing of the pr55 Gag polyprotein precursor and p24 production were similarly impaired.
Of particular interest was that the ability of mutant protease to kill MT4 cells in vitro was also impaired.58 Other groups subsequently demonstrated similar impairment in HIV PR activity associated with resistance mutations using techniques such as dual infection competition assays,47’49 pr55 Gag polyprotein processing, single-cell infectivity assays, viral RNA and p24 production rates, and Gag cleavage kinetics. These altered outcomes (compared with wild-type virus) were observed using patient viral isolates, in vitro-selected protease mutations, site-directed protease mutations, and recombinant HIV-1 vectors using laboratory strains of HIV deleted for HIV PR or reverse transcriptase, into which patient-derived protease or reverse transcriptase is inserted.43,46’47’49’59It now seems that at least three classes of mutations may result from subinhibitory HIV PI selective pressure: primary HIV resistance mutations that impair viral replicative potential in vitro, secondary protease resistance mutations that may compensate for the impairment associated with primary mutations, and Gag-Pol cleavage site mutations. Examples include the primary resistance mutations D30N and L90M, both of which significantly impact viral replicative fitness compared with wild-type virus, whereas the secondary mutation, L63P, does not independently impact viral fitness.60 However, the addition of L63P to either of the primary mutations compensates for their impaired fitness — significantly in the case of the L90M isolate and slightly for the D30N isolate.46 Alternately, defects within Gag-Pol processing may also impact replicative capacity, presumably by altering the efficacy of viral polyprotein processing, and virion packaging and progeny release. For example, paired pretherapy and postresistance viral isolates assayed for infectivity and Gag-Pol precursor cleavage demonstrated that impaired Gag-Pol cleavage (assessed by the accumulation of cleavage intermediates) is associated with impaired viral infectivity.45 Impaired Gag-Pol processing can be caused by HIV PR or Gag-Pol cleavage site mutations.44,61
Although the majority of PR mutations impair protease activity, viral infectivity, packaging, and polyprotein processing, some mutations may actually enhance fitness.
For example, viral evolution in one patient receiving protease inhibitor monotherapy demonstrated mutant viruses that initially had reduced protease activity and, consequently, reduced viral replicative capacity. In vivo replication of these viral isolates in the presence of drug selected for viral populations with more protease mutations eventually acquired enhanced replicative potential compared with the wild-type viral strain.62 Insertions of up to six amino acids were also associated with enhanced viral replicative fitness compared with wild-type controls.63 Mutational improvements in viral replicative capacity may also enhance viral susceptibility to some therapeutic PIs. For example, the Gag polyproteinprocessing mutations at positions 449 and 353 are associated with improved viral fitness, despite impaired sensitivity to antiretroviral agent amprenavir. Similarly, the N88 mutation in HIV PR is associated with impaired viral fitness and hypersusceptibility to amprenavir.64Overall, the interrelationships between viral replication, viral drug sensitivity, and Gag-Pol processing are complex and remain poorly defined. It is likely that each of these measures confers different information regarding the clinical impact of such mutations. At a molecular level, because protease inhibitor-resistant isolates have impaired catalytic activity, other viral functions will also likely be impaired. For example, the impaired ability of HIV PR to cleave the Gag-Pol polyprotein precursor may result in impaired production of reverse transcriptase (RT) and, therefore, reduced level of RT activity in infected cell cultures.65,66 Moreover, such mutants are likely to have an impaired ability to cleave host cell proteins.
One example concerns the PR active site mutation (T26S), which has approximately five- to tenfold lower activity on HIV-1 polyprotein processing.67-68 Of interest is that viruses containing this mutation also have impaired CD4 T cell killing potential. They did not induce toxicity upon bacterial expression, and their cleavage of cytoskeletal proteins in vitro was significantly impaired in infected T cell lines.67 Another PR mutant (V32I/A71V) grows only slightly more slowly than wild-type virus on peripheral blood mononuclear cells (PBMCs)43; however, it cannot recognize and cleave several cellular proteins, including vimentin, focal adhesion plaque kinase (FAK), fimbrin, and integrins.27 Altogether, these observations demonstrate that substrates of host cell origin can have different affinities for HIV PR than does the HIV polyprotein, supporting the hypothesis that mutations within HIV PR can, and do, impact the ability of HIV to kill target T cells.