Dimerization of human immunodeficiency virus type 1 reverse transcriptase as an antiviral target.

Emergence of drug resistant strains of human immunodeficiency virus type 1 (HIV-1) is a major hindrance in the long-term treatment of HIV-1 infected individuals. Alternative strategies, including those directed to structural elements of viral targets, are needed to combat the growing acquired immune deficiency syndrome (AIDS) pandemic. The HIV-1 reverse transcriptase (RT) dimer interface, critical for dimer stability and catalytic function, is a novel target for designing new anti-HIV-1 drugs. Several existing RT inhibitors are known to impair polymerase function by destabilizing RT dimer stability and can serve as useful leads in this direction. Conversely, studies have shown that potent nonnucleoside reverse transcriptase inhibitors (NNRTIs) can enhance RT subunit interaction, which may contribute in part to the inhibitory effect of these drugs. Interface peptides are reported to suppress enzyme activity by interfering with active RT heterodimer formation. This review focuses on small molecule and peptide inhibitors that interfere with the formation of the active RT heterodimer and also discusses regions in the RT that are critical for RT dimerization that can be considered as potential targets for chemotherapeutic intervention.

Relationship between enzyme activity and dimeric structure of recombinant HIV-1 reverse transcriptase.

The multifunctional enzyme human immunodeficiency virus type 1 (HIV-1) reverse transcriptase (RT) is a heterodimer composed of a 66-kDa (p66) subunit and a p66-derived 51-kDa (p51) subunit. p66/p51 HIV-1 RT contains 1 functional DNA polymerase and 1 ribonuclease H (RNase H) active site, which both reside in the p66 subunit at spatially distinct regions. In this study, we have investigated the relationship between the heterodimeric structure of HIV-1 RT and its enzymatic properties by introducing mutations at RT codon W401 that inhibit the formation of p66/p51 heterodimers. We demonstrate a striking correlation between abrogation of both HIV-1 RT dimerization and DNA polymerase activity. In contrast, the p66 monomers exhibited only moderately slowed catalytic rates of DNA polymerase-dependent and DNA polymerase-independent RNase H cleavage activity compared with the wild-type (WT) enzyme. Furthermore, no major changes in the unique cleavage patterns were observed between the WT and mutant enzymes for the different substrates used in the RNase H cleavage assays. Based on these results, and on our current understanding of HIV-1 RT structure, we propose that the p66 monomer can adopt an open tertiary conformation that is similar to that observed for the subunit in the heterodimeric enzyme. We also propose that the formation of intersubunit interactions in HIV-1 RT regulates the establishment of a functional DNA polymerase active site.

Molecular mechanisms of HIV-1 resistance to nucleoside reverse transcriptase inhibitors (NRTIs).

Nucleoside reverse transcriptase inhibitors (NRTIs), such as 3'-azido-3'-deoxythymidine, 2',3'-dideoxyinosine and 2',3'-dideoxy-3'-thiacytidine, are effective inhibitors of human immunodeficiency type 1 (HIV-1) replication. NRTIs are deoxynucleoside triphosphate analogs, but lack a free 3'-hydroxyl group. Once NRTIs are incorporated into the nascent viral DNA, in reactions catalyzed by HIV-1 reverse transcriptase (RT), further viral DNA synthesis is effectively terminated. NRTIs should therefore represent the ideal antiviral agent. Unfortunately, HIV-1 inevitably develops resistance to these inhibitors, and this resistance correlates with mutations in RT. To date, three phenotypic mechanisms have been identified or proposed to account for HIV-1 RT resistance to NRTIs. These mechanisms include alterations of RT discrimination between NRTIs and the analogous dNTP (direct effects on NRTI binding and/or incorporation), alterations in RT-template/primer interactions, which may influence subsequent NRTI incorporation, and enhanced removal of the chain-terminating residue from the 3' end of the primer. These different resistance phenotypes seem to correlate with different sets of mutations in RT. This review discusses the relationship between HIV-1 drug resistance genotype and phenotype, in relation to our current knowledge of HIV-1 RT structure.

Mutational analysis of Lys65 of HIV-1 reverse transcriptase.

Amino acid Lys(65) is part of the highly flexible beta3-beta4 loop in the fingers domain of the 66 kDa subunit of human immunodeficiency virus type 1 (HIV-1) reverse transcriptase (RT). Recent crystal data show that the epsilon-amino group of Lys(65) interacts with the gamma-phosphate of the bound deoxynucleoside triphosphate ('dNTP') substrate [Huang, Chopra, Verdine and Harrison (1998) Science 282, 1669-1675]. In order to biochemically define the function of RT Lys(65), we have used site-specific mutagenesis to generate RT with a variety of substitutions at this position, including K65E, K65Q, K65A and K65R. Kinetic analyses demonstrate that if Lys(65) in RT is substituted with an amino acid other than arginine the enzyme exhibits dramatic decreases in the binding affinity (K(m)) for all dNTP substrates, in RT catalytic efficiency (k(cat)/K(m)) and in the mutant enzyme's ability to carry out pyrophosphorolysis, the reverse reaction of DNA synthesis. The pH optimum for the DNA polymerase activity of K65E RT was 6.5, compared to 7.5 for the wild-type enzyme, and 8.0 for the K65R, K65A and K65Q mutants. Molecular modelling studies show that mutations of Lys(65) do not affect the geometry of the loop's alpha-carbon backbone, but rather lead to changes in positioning of the side chains of residues Lys(70) and Arg(72). In particular, Glu in K65E can form a salt bridge with Arg(72), leading to the diminution of the latter residue's interaction with the alpha-phosphate of the dNTP residue. This alteration in dNTP-binding may explain the large pH-dependent changes in both dNTP-binding and catalytic efficiency noted with the enzyme. Furthermore, the K65A, K65Q and K65E mutant enzymes are 100-fold less sensitive to all dideoxynucleoside triphosphate ('ddNTP') inhibitors, whereas the K65R mutation results in a selective 10-fold decrease in binding of ddCTP and ddATP only. This implies that mutations at position 65 in HIV-1 RT influence the nucleotide-binding specificity of the enzyme.

Mechanism by which phosphonoformic acid resistance mutations restore 3'-azido-3'-deoxythymidine (AZT) sensitivity to AZT-resistant HIV-1 reverse transcriptase.

The development of phosphonoformic acid (PFA) resistance against a background of 3'-azido-3'-deoxythymidine (AZT) resistance in human immunodeficiency virus type 1 (HIV-1) restores viral sensitivity to AZT. High level AZT resistance requires multiple mutations (D67N/K70R/T215F/K219Q). In order to characterize the mechanism of PFA resistance-mediated resensitization to AZT, the A114S mutation associated with PFA resistance was introduced into the reverse transcriptase (RT) of both wild type and drug-resistant virus. We previously showed that pyrophosphorolytic removal of chain-terminating AZT is the primary mechanism of the AZT resistance phenotype (Arion, D., Kaushik, N., McCormick, S., Borkow, G., and Parniak, M. A. (1998) Biochemistry 37, 15908-15917). Introduction of A114S into the AZT resistance background significantly diminishes both the enhanced pyrophosphorolytic activity and the DNA synthesis processivity associated with the AZT-resistant RT. The A114S mutation also alters the nucleotide-dependent phosphorolysis activity associated with AZT resistance. The presence of the A114S mutation therefore severely impairs the mutant enzyme's ability to excise chain-terminating AZT. The decrease in phosphorolytic activity of RT conferred by the PFA resistance A114S mutation resensitizes AZT-resistant HIV-1 to AZT by allowing the latter to again function as a chain terminator of viral DNA synthesis. These data further underscore the importance of phosphorolytic removal of chain-terminating AZT as the primary mechanism of HIV-1 AZT resistance.

Human immunodeficiency virus type 1 reverse transcriptase dimer destabilization by 1-[Spiro[4"-amino-2",2" -dioxo-1",2" -oxathiole-5",3'-[2', 5'-bis-O-(tert-butyldimethylsilyl)-beta-D-ribofuranosyl]]]-3-ethylthy mine.

The nonnucleoside inhibitor binding pocket is a well-defined region in the p66 palm domain of the human immunodeficiency virus type-1 reverse transcriptase (HIV-1 RT). This binding pocket opens toward the interface of the p66/p51 heterodimer and we have investigated whether ligand binding at or near this site induces structural changes that have an impact on the dimeric structure of HIV-1 RT. 1-[2',5'-bis-O-(tert-butyldimethylsilyl]-3'-spiro-5' '-(4' '-amino-1' ',2' '-oxathiole-2' ',2' '-dioxide)-3-ethylthymine (TSAOe(3)T) was found to destabilize the subunit interactions of both the p66/p51 heterodimer and p66/p66 homodimer enzymes. The Gibbs free energy of dimer dissociation (DeltaG(D)(H)2(O)) is decreased with increasing concentrations of TSAOe(3)T, resulting in a loss in dimer stability of 4.0 and 3.2 kcal/mol for the p66/p51 and p66/p66 HIV-1 RT enzymes, respectively. This loss of energy is not sufficient to induce the dissociation of the subunits in the absence of denaturant. This destabilizing effect seems to be unique for TSAOe(3)T, since neither the tight-binding inhibitor UC781 nor nevirapine showed any effects on the stability of HIV-1 RT dimers. TSAOe(3)T was unable to destabilize the subunit interactions of the E138K mutant enzyme, which exhibits significant resistance to TSAOe(3)T inhibition. Molecular modeling of TSAOm(3)T into the nonnucleoside inhibitor binding pocket of wild-type RT suggests that it makes significant interactions with the p51 subunit of the enzyme, a feature that has not been observed with other types of nonnucleoside inhibitors. The observed destabilization of the dimeric HIV-1 RT may result from structural/conformational perturbations at the reverse transcriptase subunit interface.

Aflatoxin B1 and sulphobromophthalein binding to the dimeric human glutathione S-transferase A1-1: a fluorescence spectroscopic analysis.

The binding interactions between dimeric human class alpha glutathione S-transferase A1-1 (GST A1-1) and aflatoxin B1 or sulphobromophthalein (BSP) were characterised. Aflatoxin B1 binds to GST A1-1 with a stoichiometry of 1.1 mol/mol of dimeric enzyme. The binding interaction, which can be described by a hyperbolic saturation isotherm (Kd = 8+/-2 microM), does not induce major structural changes in the enzyme, nor does it inhibit enzymatic activity. The average distance between the single tryptophan residue (Trp20) of GST A1-1 and protein-bound aflatoxin B1 was calculated to be 22.7 A by means of fluorescence resonance energy transfer. The aflatoxin-binding region, according to this calculated distance, was determined to be located in the dimer interface cleft near the crystallographic two-fold axis. Hill-plot analyses suggest that a positive co-operative interaction exists between BSP and the dimeric GST A1-1 (h = 1.6+/-0.1; K' = 14+/-0.6 microM). The binding of BSP induces a conformational change in the enzyme which is accompanied by a decrease in the molecular flexibility and in the solvent-accessible properties of the enzyme's Trp20 residue. Site-directed mutagenesis of Trp20 (Trp20-->Phe) confirms that this residue is situated in the binding environment and although it is not essential for BSP binding, it is involved in the interaction. Furthermore, the structural change associated with BSP binding alters the hyperbolic character of the glutathione saturation curve. This indicates that there may also be a cooperative interaction between glutathione and BSP or that BSP binding induces asymmetric functioning of the two enzyme subunits so that they become unequal in catalytic activity.

Equilibrium and kinetic unfolding properties of dimeric human glutathione transferase A1-1.

The equilibrium and kinetic unfolding properties of homodimeric class alpha glutathione transferase (hGST A1-1) were characterized. Urea-induced equilibrium unfolding data were consistent with a folded dimer/unfolded monomer transition. Unfolding kinetics were investigated, using stopped-flow fluorescence, as a function of denaturant concentration (3.5-8.9 M urea) and temperature (10-40 degrees C). The unfolding pathway, monitored by tryptophan fluorescence, was biphasic with a fast unfolding event (millisecond time range with enhanced fluorescence properties) and a slow unfolding event (seconds to minutes time range with quenched fluorescence properties). Both events occurred simultaneously from 3.5 M urea. Each phase displayed single-exponential behavior, consistent with two unimolecular reactions. Urea-dependence studies and thermodynamic activation parameters (transition-state theory) suggest that the transition state for each phase is well-structured and is closely related to native protein in terms of solvent exposure. The apparent activation Gibbs free energy change in the absence of denaturant, DeltaG (H2O), indicates that the slow unfolding event represents the transition state for the overall unfolding pathway. The rate and urea independence of each phase on the initial condition exclude the possibility of a preexisting equilibrium between various native forms in the pretransition baseline. The unfolding pathways monitored by energy transfer to or direct excitation of AEDANS covalently linked to Cys111 in hGST A1-1 were monophasic with urea and temperature properties similar to those observed for the slow unfolding event (described above). A sequential unfolding kinetic mechanism involving the partial dissociation of the two structurally distinct domains per subunit followed by complete domain and subunit unfolding is proposed.

Conformational stability of pGEX-expressed Schistosoma japonicum glutathione S-transferase: a detoxification enzyme and fusion-protein affinity tag.

A glutathione S-transferase (Sj26GST) from Schistosoma japonicum, which functions in the parasite's Phase II detoxification pathway, is expressed by the Pharmacia pGEX-2T plasmid and is used widely as a fusion-protein affinity tag. It contains all 217 residues of Sj26GST and an additional 9-residue peptide linker with a thrombin cleavage site at its C-terminus. Size-exclusion HPLC (SEC-HPLC) and SDS-PAGE studies indicate that purification of the homodimeric protein under nonreducing conditions results in the reversible formation of significant amounts of 160-kDa and larger aggregates without a loss in catalytic activity. The basis for oxidative aggregation can be ascribed to the high degree of exposure of the four cysteine residues per subunit. The conformational stability of the dimeric protein was studied by urea- and temperature-induced unfolding techniques. Fluorescence-spectroscopy, SEC-HPLC, urea- and temperature-gradient gel electrophoresis, differential scanning microcalorimetry, and enzyme activity were employed to monitor structural and functional changes. The unfolding data indicate the absence of thermodynamically stable intermediates and that the unfolding/refolding transition is a two-state process involving folded native dimer and unfolded monomer. The stability of the protein was found to be dependent on its concentration, with a delta G degree (H2O) = 26.0 +/- 1.7 kcal/mol. The strong relationship observed between the m-value and the size of the protein indicates that the amount of protein surface area exposed to solvent upon unfolding is the major structural determinant for the dependence of the protein's free energy of unfolding on urea concentration. Thermograms obtained by differential scanning microcalorimetry also fitted a two-state unfolding transition model with values of delta Cp = 7,440 J/mol per K, delta H = 950.4 kJ/mol, and delta S = 1,484 J/mol.

Class-pi glutathione S-transferase is unable to regain its native conformation after oxidative inactivation by hydrogen peroxide.

The oxidation of protein sulphydryls to disulphides and their reduction back to sulphydryls is an early cellular response to oxidative stress. Hydrogen-peroxide-mediated oxidation of class-pi glutathione S-transferase results in the formation of disulphide bonds, which inhibits its catalytic function yet allows it to retain its non-substrate-ligand-binding properties. The overall hydrodynamic volume of the oxidised class-pi glutathione S-transferase type-1 homodimer (GSTP1-1) is decreased, and its tertiary-structural and secondary-structural elements are changed with respect to the native protein. Structural differences appear to be prominent in domain 1 of oxidised GSTP1-1, in that the exposure of both tryptophan residues is increased, while the electric potential about one of them is altered. Treatment of the oxidised protein with dithiothreitol or glutathione restores its enzymatic capabilities, albeit with lower specific activities for 1-chloro-2,4-dinitrobenzene and ethacrynic acid. The hydrophobic binding site (H-site) for electrophilic substrates is negatively affected in that the K(m) and catalytic-efficiency values are diminished significantly with respect to those values obtained for the native protein. The dithiothreitol-treated oxidised GSTP1-1 is able to regain its overall hydrodynamic volume; however, both its secondary-structural and tertiary-structural elements remain modified with respect to the native protein, as do both tryptophanyl environments. Furthermore, the oxidised glutathione S-transferase and dithiothreitol-treated oxidised glutathione S-transferase are less thermostable (tM = 55.5 degrees C and 56.3 degrees C, respectively) than the native enzyme (tM = 59 degrees C). These results indicate that the class-pi glutathione S-transferase is unable to regain its native conformation after oxidative inactivation.

Determination of a binding site for a non-substrate ligand in mammalian cytosolic glutathione S-transferases by means of fluorescence-resonance energy transfer.

To determine the location of the non-substrate-ligand-binding region in mammalian glutathione S-transferases, fluorescence-resonance energy transfer was used to calculate distances between tryptophan residues and protein-bound 8-anilinonaphthalene 1-sulphonate (an anionic ligand) in the human class-alpha glutathione S-transferase, and in a human Trp28-->Phe mutant class-pi glutathione S-transferase. Distance values of 2.21 nm and 1.82 nm were calculated for the class-alpha and class-pi enzymes, respectively. Since glutathione S-transferases bind one non-substrate ligand/protein dimer, the ligand-binding region, according to the calculated distances, is found to be located in the dimer interface near the twofold axis. This region is the same as that in which the parasitic helminth Schistosoma japonicum glutathione S-transferase binds praziquantel, a non-substrate drug used to treat schistosomiasis [McTigue, M. A., Williams, D. R. & Tainer, J. A. (1995) J. Mol. Biol. 246, 21-27]. Since the overall folding topology is conserved and certain features at the dimer interface are similar throughout the superfamily, it is reasonable to expect that all cytosolic glutathione S-transferases bind non-substrate ligands in the amphipathic groove at the dimer interface.

Conformational stability of Cys45-alkylated and hydrogen peroxide-oxidised glutathione S-transferase.

A highly reactive cysteine residue in class pi glutathione S-transferases enhances their susceptibility to chemical alkylation and oxidative stress. Alkylation of the reactive Cys45 in the porcine class pi enzyme (pGSTP1-1) with either N-iodoacetyl-N'-(5-sulpho-1-naphthyl)ethylenediamine or iodoacetamide results in a loss of enzyme activity and glutathione-binding function. Similarly, oxidation of pGSTP1-1 with hydrogen peroxide (H2O2) also results in a loss of catalytic and glutathione-binding function, but these effects are reversed by the addition of 5 mM glutathione or dithiothreitol. Analysis by SDS-PAGE of the H2O2-oxidised enzyme indicates oxidation-induced formation of disulphide bonds involving Cys45. Equilibrium-unfolding studies with guanidinium chloride indicate that the unfolding of Cys45-alkylated and H2O2-oxidised pGSTP1-1 can be described by a two-state model in which the predominant thermodynamically stable species are the folded dimer and unfolded monomer. Unfolding transition curves suggest that the introduction of a large and bulky AEDANS at Cys45 does not affect the unfolding pathway for pGSTP1-1. H2O2-oxidised pGSTP1-1, on the other hand, appears to follow a different unfolding pathway. This appears not to be a result of the introduction of disulphide bonds since the reduction of these bonds in the oxidised protein with dithiothreitol does not affect the unfolding transition. Furthermore, the conformational stability of the oxidised protein is significantly diminished (delta G(H2O) = 11.6 kcal/mol) when compared with unmodified and AEDANS-alkylated enzyme (delta G(H2O) = 22.5 kcal/mol).