点击显示 收起
Rega Institute for Medical Research, Katholieke Universiteit Leuven, Leuven, Belgium (J.A., E.D.C., J.B.)
Department of Pharmacology, University of Alcale? Alcale?de Henares, Spain (F.R.-B., F.G.)
Department of Experimental Medicine, University of Rome "Tor Vergata", Rome, Italy (F.C.-S., C.-F.P.)
Instituto de Que猰ica Meedica, Consejo Superior de Investigaciones Cienteicas, Madrid, Spain (A.S.-F., S.V., M.-J.C.)
Abstract
The coumarins represent a unique class of non-nucleoside reverse transcriptase inhibitors (NNRTIs) that were isolated from tropical plants. (+)-Calanolide A, the most potent compound of this class, selects for the T139I resistance mutation in HIV-1 reverse transcriptase (RT). Seven RTs mutated at amino acid position 139 (Ala, Lys, Tyr, Asp, Ile, Ser, and Gln) were constructed by site-directed mutagenesis. The mutant T139Q enzyme retained full catalytic activity compared with wild-type RT, whereas the mutant T139I, T139S, and T139A RTs retained only 85 to 50% of the activity. Mutant T139K, T139D, and T139Y RTs had seriously impaired catalytic activities. The mutations in the T139I and T139D RTs were shown to destabilize the RT heterodimer. (+)-Calanolide A lost inhibitory activity (up to 20-fold) against the mutant T139Y, T139Q, T139K, and T139I enzymes. All of the mutant enzymes retained marked susceptibility toward the other NNRTIs, including nevirapine, delavirdine, efavirenz, thiocarboxanilide UC-781, quinoxaline GW867420X, TSAO [[2',5'-bis-O-(tert-butyldimethylsilyl)--D-ribofuranosyl]-3'-spiro-5''-(4''-amino-1'',2''-oxathiole-2'',2''-dioxide)] derivatives, and the nucleoside inhibitor, ddGTP. The fact that the T139I RT 1) proved to be resistant to (+)-calanolide A, 2) represents a catalytically efficient enzyme, and 3) requires only a single transition point mutation (ACAATA) in codon 139 seems to explain why mutant T139I RT virus strains, but not virus strains containing other amino acid changes at this position, predominantly emerge in cell cultures under (+)-calanolide A pressure.
Polycyclic coumarins, originally isolated as natural products from several plants of the genus Calophyllum, have been demonstrated to be active against HIV-1 (for a review, see Yu et al., 2003). (+)-Calanolide A, the most potent compound of this class, has been evaluated in antiviral activity studies against non-nucleoside reverse transcriptase inhibitor (NNRTI)-resistant HIV-1 strains and related mutated RTs (Buckheit et al., 1999; Quan et al., 1999). Detailed enzyme kinetic studies on RT inhibition by (+)-calanolide A have also been performed. Unlike NNRTIs, which noncompetitively inhibit RT with respect to the substrate and template/primer, (+)-calanolide A is at least partly competitive compared with dNTP binding (Currens et al., 1996b). These findings suggest that (+)-calanolide A most probably interacts with RT in a manner that is mechanistically different from that of other NNRTIs described previously. In this respect, (+)-calanolide A may represent a unique class of HIV-1-specific NNRTIs. Despite these kinetic differences with NNRTIs, (+)-calanolide A, similar to most NNRTIs, is inactive against HIV-2 strains or other (retro)viruses (Kashman et al., 1992; Currens et al., 1996a).
One of the major problems associated with the NNRTIs is the rapid emergence of drug resistant virus strains (Vandamme et al., 1998; Balzarini, 1999, 2004; De Clercq, 1999). Indeed, HIV-1 resistance to regular NNRTIs is primarily associated with mutations of amino acids that line the lipophilic NNRTI-specific binding pocket (Balzarini, 1999). (+)-Calanolide A selects in cell culture for the rather unusual T139I mutation in the HIV-1 RT (Buckheit et al., 1995). The mutant T139I HIV-1 strains are resistant to (+)-calanolide A but retain marked sensitivity to many other NNRTIs as well as several nucleoside RT inhibitors (NRTIs) (Buckheit et al., 1999). The Thr139 amino acid is part of the so-called 7-8 loop, which comprises a six-amino acid motif denoted as SINNET. Whereas this loop is exposed to the solvent in the p66 subunit, the equivalent loop in p51 is snugly lodged into a cleft on the surface of the p66 subunit (Kohlstaedt et al., 1992). In fact, this loop, which is close to both the putative entrance to the NNRTI-binding pocket and the active site (Fig. 1), is essential for the catalytic function of the p66 subunit because it is required to form a stable heterodimeric enzyme (Pandey et al., 2001, 2002). It is also worth mentioning that resistance to the TSAO [[2',5'-bis-O-(tert-butyldimethylsilyl)--D-ribofuranosyl]-3'-spiro-5''-(4''-amino-1'',2''-oxathiole-2'',2''-dioxide)] class of NNRTIs is achieved through mutation of Glu138 to lysine in the p51 7-8 loop (Balzarini et al., 1993a, 1993b). Because of the close proximity of Thr139 and Glu138 in HIV-1 RT, it is plausible that both (+)-calanolide A and TSAO derivatives share a similar site and/or mode of interaction with HIV-1 RT. Extending this analogy, (+)-calanolide A resistance must arise from the T139I mutation taking place in the p51 subunit rather than in the p66 subunit. This is in sharp contrast to the vast majority of mutations in the NNRTI-binding pocket conferring resistance to other NNRTIs, which are the result of substitutions occurring in the p66 subunit rather than in the p51 subunit. Recent reports on the mechanism of TSAO resistance and structural modeling have suggested an influence of TSAO on RT dimerization, which places this compound in a unique position among the NNRTIs (Sluis-Cremer et al., 2000; Rodre猤uez-Barrios et al., 2001).
In this study, we constructed seven different recombinant RT enzymes bearing a mutation at position 139 of RT and determined their catalytic activity as well as their resistance profiles against a variety of NNRTIs, including (+)-calanolide A and TSAO. The data obtained provide a rationale for the finding that it is the T139I mutation and not any other substitution at position 139 of HIV-1 RT that emerges under (+)-calanolide A pressure in cell culture.
Materials and Methods
Compounds. TSAO derivatives of N3-methylthymine (m3T) and thymine were synthesized as described previously (Balzarini et al., 1992). Nevirapine (BI-RG-587, dipyridodiazepinone) was obtained from Boehringer Ingleheim USA (Ridgefield, CT). Delavirdine [bis-(heteroaryl)piperazine, U-90152], and efavirenz (DMP-266) were provided by Dr. R. Kirch (at that time at Aventis, Frankfurt, Germany) and Dr. J.-P. Kleim (currently at GlaxoSmithKline, Stevenage, UK). The thiocarboxanilide derivative UC-781 was obtained from W. G. Brouwer (Crompton, Ltd., Guelph, Ontario, Canada). The quinoxaline GW420867X was provided by Dr. J.-P. Kleim. 2',3'-dideoxy-GTP (ddGTP) was obtained from Sigma-Aldrich (St. Louis, MO). (+)-Calanolide A was delivered by Sarawak MediChem Pharmaceuticals Inc. (Sarawak, Malaysia).
Site-Directed Mutagenesis of HIV-1 Reverse Transcriptase. Mutant RT enzymes containing the T139A, T139Q, T139Y, T139K, T139I, T139S, or T139D mutations in both p66 and p51 subunits were derived from the RT sequence cloned in pKRT2His (D'Aquila and Summers, 1989). Site-directed mutagenesis was performed using the QuikChange site-directed mutagenesis kit (Stratagene, Cambridge, UK) as described previously (Pelemans et al., 1998). Two synthetic oligonucleotide primers (Sigma-Aldrich, St. Louis, MO) contained the desired mutation at amino acid position 139 of HIV-1 RT. The presence of the desired mutation was confirmed by sequencing of the RT gene on an ABI Prism 3100 sequencer (Applied Biosystems, Foster City, CA) using the ABI Prism Big Dye Terminator Cycle Sequencing Ready reaction kit (Applied Biosystems). For all mutants, the mutation was introduced in both p66 and p51 subunits. For only the mutant T139I and T139D RT, the mutation was introduced solely in p66, solely in p51, or in both p66 and p51 subunits.
Construction of Plasmids Expressing Mutant and Wild-Type Recombinant HIV-1 RT. Recombinant HIV-1 RT were expressed from a two-plasmid coexpression system as described previously (Jonckheere et al., 1996). The p66 subunit of RT was expressed from pACYC66His and the p51 subunit from pKRT51. To construct wild-type and 139-mutated pACYC66His, wild-type and 139-mutated pKRT2His were digested with EcoRI and FspI, and the RT-containing fragments were ligated into pACYC184 digested with EcoRI and ScaI. To construct wild-type and 139-mutated pKRT51, wild-type and 139-mutated pKRT2His were digested with NcoI and KpnI, and the RT-containing fragment was ligated into pKRT51 digested with NcoI and KpnI.
Expression and Purification of Wild-Type and Mutant HIV-1 RT. Expression of recombinant HIV-1 RT was performed as described previously (Auwerx et al., 2004). Luria broth (800 ml) containing 100 e蘥/ml ampicillin and 10 e蘥/ml tetracycline was inoculated with an overnight culture of Escherichia coli JM109 transformed with both plasmids of the coexpression system and started at an OD600 of 0.1. The culture was grown at 37°C and induced with a 1 mM final concentration of isopropyl 1-thio--D-galactopyranoside for expression of RT, and after centrifugation, the pellet was stored at -20°C. Later, the bacterial cell pellet was resuspended in 15 ml of lysis buffer (50 mM sodium phosphate buffer, 5 mM -mercaptoethanol, 0.9% glucose, 100 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 10 e蘥/ml pepstatin, 10 e蘥/ml leupeptin, and 10% glycerol) and passed through an SLM Aminco French Pressure Cell Press (Beun de Ronde, La Abcoude, The Netherlands). The lysate was centrifuged for 20 min at 17,000g.
The purification of RT was performed as described previously (Auwerx et al., 2004). In brief, the supernatant of the lysed bacterial cell culture was incubated with Ni-NTA resin (QIAGEN, Valencia, CA). After sedimentation of the Ni-NTA resin with the bound (His)6-tagged proteins, a column was formed and washed twice with a sodium phosphate buffer containing 10 mM imidazole. The RT then was eluted from the column with a sodium phosphate buffer containing 125 mM imidazole. The imidazole-containing buffer was exchanged with a Tris-HCl buffer, and the eluate was concentrated to 2 ml using Ultrafree-15 centrifugal filtration devices (Millipore Corporation, Billerica, MA). The (His)6-tagged RT was further purified to approximately 98% purity over a HiTrap Heparin column (Amersham Biosciences UK, Ltd., Little Chalfont, Buckinghamshire, UK). All of the fractions containing heterodimer RT were pooled and stored in a 50% glycerol buffer at -20°C. Protein concentrations in these stock solutions were determined using the Bio-Rad protein assay (Bio-Rad, Hemel Hempstead, UK) with bovine serum albumin as standard.
Reverse Transcriptase Assay. For determination of the 50% inhibitory concentration (IC50) of the test compounds against HIV-1 RT, the RNA-dependent DNA polymerase assay was performed as follows: 50 e蘬 of the reaction mixture contained 50 mM Tris-HCl, pH 7.8, 5 mM dithiothreitol, 300 e glutathione, 500 e EDTA, 150 mM KCl, 5 mM MgCl2, 1.25 e蘥 of bovine serum albumin, a fixed concentration of the labeled substrate [8-3H]dGTP (1.6 e, 1 e藽i; specific activity, 12.6 Ci/mmol; Amersham Biosciences), a fixed concentration of the template/primer poly(rC)·oligo(dG)12-18 (0.1 mM; Amersham Biosciences), 0.06% Triton X-100, 5 e蘬 of inhibitor solution containing various concentrations (10-fold dilutions) of the compounds, and 5 e蘬 of the RT preparations that correspond to 8, 11, 6, 61, 575, 6, 9, and 137 ng of enzyme (protein) for wild type and the T139A, T139Q, T139Y, T139K, T139I, T139S, and T139D mutant RTs, respectively. The reaction was initiated by the addition of the enzyme, and the reaction mixtures were incubated at 37°C for 30 min, at which time 200 e蘬 of yeast RNA (2 mg/ml) and 1 ml of trichloroacetic acid (5%, v/v) in 20 mM Na4P2O7 were added. The solutions were incubated on ice for at least 30 min, after which the acid-insoluble material was filtered over GF/C glass-fiber filters (Whatman, Maidstone, UK) and washed with 5% trichloroacetic acid in H2O and ethanol. The filters were then analyzed for radioactivity in a liquid scintillation counter (Canberra Industries, Meriden, CT). The IC50 for each test compound was determined as the compound concentration that inhibited HIV-1 RT activity by 50%.
Steady-state kinetic assays were also performed as described previously (Balzarini et al., 1992), except that the reaction mixtures were incubated for 30 min instead of 60 min during the assays with variable substrate (dGTP) or template/primer [poly(rC)·oligo-(dG)12-18] concentrations. Under these experimental conditions, the catalytic reaction of the different enzymes proceeded linearly and proportionally with time. The Km and Vmax (kcat) values for poly(rC)·oligo(dG)12-18 and dGTP were determined in the presence of fixed concentrations of [8-3H]dGTP (specific radioactivity, 14.1 Ci/mmol [1.25 e(1 e藽i)]) and poly(rC)·oligo(dG)12-18 (0.1 mM), respectively. The Km and Vmax (kcat) values were derived from the double reciprocal Lineweaver-Burk plots of the variable substrate (dGTP) or template/primer [poly(rC)·oligo(dG)12-18] concentrations versus the velocities of dGTP incorporation at each substrate or template/primer concentration.
Stability of Wild-Type and Mutant Heterodimer HIV-1 RTs in the Presence of Urea. Denaturation curves were plotted by preincubation of RT with different concentrations of urea ranging from 0.0625 up to 2.5 M for 10 min at 37°C in a 50-e蘬 reaction buffer containing 50 mM Tris-HCl, pH 7.8, 0.06% Triton X-100, 5 mM dithiothreitol, 150 mM, 0.3 mM glutathione, 1.25 mg/ml bovine serum albumin, 0.5 mM EDTA, 5 mM MgCl2, and 1.4 mM poly(rC)·oligo(dG)12-18 (Amersham Biosciences). The polymerase reaction was initiated by adding [8-3H]dGTP (0.1 mM, 1 mCi/ml) (Amersham Biosciences) substrate. After incubating for 10 min at 37°C, the reactions were terminated by the addition of 1 ml of 5% ice-cold trichloroacetic acid in 200 mM Na4P2O7 and 200 e蘬 of yeast RNA (2 mg/ml, pH 8.0). Reaction products were incubated on ice for 30 min and precipitated on a Whatman GF/C filter. The filters were washed with 20 ml of 5% trichloroacetic acid and dried with 2 ml of ethanol. The amount of incorporated radioactive substrate was analyzed in a TR-2500 liquid scintillation counter (PerkinElmer Life and Analytical Sciences, Boston, MA) by adding 4 ml of HiSafe2 (PerkinElmer Life and Analytical Sciences). Polymerase activity was determined as the amount of nucleotide incorporated at each urea concentration relative to the amount of nucleotide incorporation in the absence of denaturant. The percentage polymerase activity was plotted versus the urea concentration, and the data were fitted to a curve using the program SigmaPlot Version 8.0 (SPSS Inc., Chicago, IL) to determine the concentration of urea at the midpoint of the denaturation curve.
Molecular Modeling and Structure Visualization. The protein environment around Thr139 in HIV-1 RT was visualized and pictured using the PyMOL molecular graphics program (DeLano, 2004), and the X-ray coordinates of a covalently trapped catalytic complex between HIV-1 RT and a DNA template:primer were deposited in the Protein Data Bank (http://www.rcsb.org/PDB/) with code 1RTD (Huang et al., 1998).
Results
RNA-Dependent DNA Polymerase Activities of Wild-Type and Mutant T139X HIV-1 RTs. To assess the influence of changes at the amino acid residue Thr139 on the catalytic activity of HIV-1 RT, seven recombinant RTs were constructed by site-directed mutagenesis: T139A, T139Q, T139Y, T139K, T139I, T139S, and T139D. In this way, the different types of amino acid side chains were represented: an aliphatic side chain in alanine, an aromatic side chain in tyrosine, a protonated amino group in the positively charged lysine, a carboxylate group in the negatively charged aspartic acid, and a polar uncharged group in serine and glutamine. The T139I mutation was also introduced, because it consistently appears in cell culture under (+)-calanolide A-selective pressure. The mutations were introduced in both p66 and p51 subunits of the RT heterodimer, and all of the mutant recombinant RTs were purified to 98% homogeneity through Ni-NTA- and heparin-containing affinity columns.
Using poly(rC)·oligo(dG)12-18 as the template/primer and [8-3H]dGTP as the radiolabeled substrate, the RNA-dependent DNA polymerase activity was fully retained in the T139Q mutant and was reduced by only 15% in the mutant T139I RT compared with wild type (Fig. 2). The polymerase activity was reduced by approximately 50% in the T139S and T139A RT mutants and was seriously impaired in the other mutants containing the negatively charged Asp139, the positively charged Lys139, and the aromatic amino acid mutation Tyr139.
To assess the role of the T139D and T139I mutations when separately located in the p66 and the p51 subunits of the RT heterodimer, four additional mutant RTs were constructed in which the T139D or T139I mutation was introduced solely in either the p66 or the p51 subunit of the heterodimeric RT enzyme. Whereas the mutant RT enzyme in which T139D was solely present in the p51 subunit had a catalytic activity that was 21 ± 3% wild-type enzyme, the heterodimeric enzyme at which T139D was solely present in p66 had a catalytic activity of 91 ± 3% wild-type enzyme. For the T139I mutation solely present in the p51 subunit, the catalytic activity was 59 ± 1% wild type, whereas the presence of this mutation in the p66 alone was 89 ± 3% wild type. Thus, the exclusive presence of the T139D or T139I mutation in the p51 subunit of the RT heterodimer had a much more deleterious effect on the catalytic activity of the mutant enzyme than when these mutations were solely present in the p66 subunit of the RT heterodimer.
Inhibitory Activities of NNRTIs and ddGTP against Wild-Type and Mutant T139X HIV-1 RTs. The mutant T139Q/T139I/T139S/T139A/T139K/T139D/T139Y RT enzymes were evaluated for their sensitivity to the inhibitory activity of a variety of NNRTIs and the NRTI ddGTP (Table 1). Among all of the NNRTIs evaluated, (+)-calanolide A showed the most pronounced loss of inhibitory potential against the mutated enzymes. Indeed, the mutant T139K (20-fold), T139I (8-fold), T139Y (6-fold), and T139Q (6-fold) HIV-1 RT enzymes displayed marked resistance toward (+)-calanolide A compared with wild-type enzyme.
Template/primer, 1.4 mM poly(rC) ·oligo(dG)12-18; substrate, 1.6 e [8-3H]dGTP. Data are the means of two to three independent experiments ± S.D. The amino acid mutations are concomitantly present in both p66 and p 51 subunits of the HIV-1 RT heterodimer.
The RT enzyme bearing the T139K mutation was 4-fold less susceptible to the inhibitory activity of the TSAO derivatives of thymine and m3T. In contrast, most mutant enzymes gained significant sensitivity toward the second-generation NNRTIs, such as thiocarboxanilide UC-781 and efavirenz. This was most noticeable for the mutant T139A and T139D RTs (increases of up to 5- to 10-fold and 3- to 4-fold in sensitivity for UC-781 and efavirenz, respectively). The greater susceptibility of most mutant Thr139 enzymes to UC-781 and efavirenz was not a general property of second-generation NNRTIs, because the quinoxaline GW867420X kept a virtually similar inhibitory potential against each of the mutant RTs. As already observed for UC-781 and efavirenz, the mutated T139D RT enzyme was 5-fold more sensitive toward the inhibitory activity of the first-generation NNRTI nevirapine (Table 1). We were surprised that ddGTP showed a markedly decreased inhibitory activity (7-fold) against several mutant RTs, in particular, T139Y and T139D RT.
When (+)-calanolide A was evaluated for its inhibitory activity against the HIV-1 RT heterodimers that contained the T139I or T139D mutation solely in either the p66 or the p51 subunit, marked resistance of mutant T139I RT toward (+)-calanolide A was only evident when the T139I mutation was solely introduced in the p51 subunit (Fig. 3). Thus, the resistance against (+)-calanolide A is clearly originating from the amino acid mutation in the p51 subunit and not in the p66 subunit of the heterodimer.
Kinetic Analysis of Wild-Type and Mutant T139I HIV-1 RTs. Kinetic analysis of the wild-type and mutant T139I RTs was performed with the substrate dGTP or the template/primer poly(rC)·oligo(dG) as variables. The kinetic parameters are summarized in Table 2. When dGTP or the template/primer poly(rC)·oligo(dG) was used as the variable substrate, no marked differences in Km were noted between wild-type and mutant T139I enzyme. The catalytic efficiency of the mutant enzyme (kcat/Km) was very comparable between the mutant T139I and wild-type enzyme, indicating that the T139I mutation has no marked influence on positioning of the template/primer or the substrate in an optimal position to allow efficient catalysis.
The data are means (± S.D.) of at least two to three independent experiments.
Effects of Urea on Wild-Type and Mutant T139I and T139D HIV-1 RT Activity. Wild-type and mutant T139I and T139D RTs were exposed to a variety of urea concentrations, and their catalytic activity was determined (Fig. 4). For the wild-type enzyme, the polymerase activity gradually decreased in the presence of increasing concentrations of urea. A urea concentration as high as 0.5 M decreased the catalytic activity of wild-type RT by 20% (residual activity 80%), whereas 2.0 M urea decreased its catalytic activity to less than 10%. Half of the wild-type RT catalytic activity was retained at 0.80 M urea (i.e., the urea-IC50). When the mutant T139I and T139D RT enzymes were exposed to similar concentrations of urea, the enzymes showed increased sensitivity toward the denaturing effect of urea compared with wild-type RT. Indeed, whereas the urea-IC50 shifted from 0.80 to 0.55 M for the mutant T139I RT, the urea IC50 was further decreased to 0.45 M for the mutant T139D RT enzyme (Fig. 4). A similar increased sensitivity to urea was observed for mutant T139D RT when the T139D mutation was solely introduced in the p51 subunit (urea-IC50 = 0.60 M), whereas the sensitivity to urea was not increased (urea-IC50 = 0.95 M) when the T139D mutation was solely introduced in the p66 subunit (data not shown). There was a close correlation between the catalytic activity of the mutant T139D RT with mutation in both subunits, p66 solely and p51 solely, and the urea concentration required to decrease RT activity by 50%. The r-value of the regression line was as high as 0.995.
Discussion
(+)-Calanolide A possesses antiviral properties characteristically ascribed to NNRTIs (i.e., selectivity for HIV-1, but not HIV-2 strains, and rapid selection of drug-resistant virus strains containing NNRTI-characteristic mutations in RT). It is noteworthy that (+)-calanolide A exhibits a 10-fold enhanced activity against certain drug-resistant viruses that bear the most prevalent NNRTI resistance mutations such as the Y181C mutation (Buckheit et al., 1999). (+)-Calanolide A-resistant virus strains may carry, besides T139I, also L100I, Y188H, L187F, and N348K mutations in the RT (Currens et al., 1996b; Buckheit et al., 1999). Among these mutations, the single amino acid substitution T139I proved to be of major importance for the resistance against (+)-calanolide A and the effect of this mutation on drug resistance is engendered from its presence in the p51 subunit (Boyer et al., 1994). This p51 subunit dependence was also proven by the fact that resistance toward (+)-calanolide A was found in the mutant RT with a wild-type p66 and a mutant T139I p51 subunit and not vice versa. However, the occurrence of additional amino acid changes besides the T139I mutation in the RT of the selected virus strains must explain the pronounced resistance to (+)-calanolide A, because the single amino acid T139I mutation confers a relatively low level of resistance to this drug (up to 20-fold).
To the best of our knowledge, no detailed site-directed mutagenesis studies of the RT enzyme at this amino acid position have ever been performed to assess the impact of the different mutations on (+)-calanolide A sensitivity, RT catalytic activity, and the structural dynamics of HIV-1 RT. Moreover, the proximity of Thr139 to Glu138, the amino acid that is important for binding of TSAO derivatives to HIV-1 RT (Camarasa et al., 2004), also suggests that the binding sites for (+)-calanolide A and TSAO in HIV-1 RT could be near each other or even overlapping. Glu138 is one of the amino acid residues located near the putative entrance to the well defined NNRTI-binding pocket (Esnouf et al., 1997). Although (+)-calanolide A is structurally quite distinct from TSAO, it is not unfeasible that the predominant interaction of (+)-calanolide A with HIV-1 RT occurs with the 7-8 loop located in the p51 RT subunit of RT. However, attempts to dock this drug in this region failed to provide a unique and distinctive binding mode (data not shown). This 7-8 loop is important for heterodimerization of the RT enzyme as shown earlier for the TSAO interaction with Glu138 of the 7-8 loop (Pandey et al., 2001, 2002; Rodre猤uez-Barrios et al., 2001). As found for TSAO-derivatives, perhaps (+)-calanolide A can also enhance RT heterodimer dissociation, a feature that has previously not been observed for other types of NNRTIs (Sluis-Cremer et al., 2000).
The T139I mutation that appears under (+)-calanolide A selection is indeed located in the 7-8 loop of p51 (Fig. 1) that is of crucial importance for efficient dimerization of both subunits. In agreement with the experimental data, molecular modeling of TSAO-m3T binding to this loop in wild-type RT suggests that the observed destabilization of the heterodimeric RT may result from structural and conformational perturbations at the RT subunit interface (Sluis-Cremer et al., 2000; Rodre猤uez-Barrios et al., 2001). Likewise, mutations at the 139 position of the HIV-1 RT may destabilize the p66/p51 heterodimer in a similar way, thus diminishing the catalytic activity of the enzyme as described previously for some amino acid mutations at position 138 (Pelemans et al., 2001) and recently also at positions 136 (Balzarini et al., 2005) and 137 (Auwerx et al., 2005) of HIV-1 RT. Indeed, a seriously compromised RT activity was observed for several amino acid mutations at position 139, especially for the lysine and aspartic acid residues, which also resulted in a higher susceptibility of the mutated RT to the inactivating (denaturation) action of urea (Fig. 4). Therefore, it may be hypothesized that certain mutations at position 139, such as Lys or Asp, compromise the optimal conformation at the p66/p51 heterodimer interface, resulting in a decreased catalytic activity and easier separation of both subunits in the presence of urea. Low concentrations of urea probably have already a marked influence on RT activity, because it can destabilize the heterodimer before full dissociation into monomers occurs. Sluis-Cremer et al. (2000) reported a higher urea-IC50 for wild-type RT (2.3 M) than that found in our study. This could be explained by the fact that the reaction buffer and purification methods of the RT are quite different in our methods; it also cannot be excluded that different salt concentrations have an influence on the urea-IC50 as well. The different nature of the particular RT gene construct may also account for the observed differences.
Although Thr139 is rather conserved in wild-type HIV-1 strains, Ceccherini-Silberstein et al. (2005) found that 3.7% of drug-treated HIV-1-infected patients were bearing mutations at position 139. These amino acid mutations seemed to be Ala, Arg, Lys, Met, Ser, Val, Ile, and Pro. The absence of the aspartic acid and the tyrosine mutations at position 139 in NNRTI-treated patients is in agreement with our site-directed mutagenesis observations, because viruses with highly impaired enzymes, such as the mutant T139D RT (2.5% wild-type activity) and the mutant T139Y RT (0.6% wild-type activity), would probably not be viable in cell culture and/or patients. Because none of the patients was treated with (+)-calanolide A, it seems that other drugs may in some cases also select for mutations at Thr139 either to decrease the sensitivity of the virus for the particular drug or to compensate for a potential decreased replication capacity of other mutations in RT. There was clearly a drug pressure on the virus, because in drug-naive individuals, six different mutations were spontaneously found in 10 of 457 isolates (i.e., Val1, Met1, Ala1, Arg2, Ile2, and Pro3), whereas in drug-treated individuals nine different mutations were found in 72 of 1556 isolates (i.e., Lys19, Arg18, Ala13, Gln9, Met6, Ile3, Ser2, Val1, and Pro1).
A mutation at position 139 to either Gln, Ile, Lys, or Tyr would result in a 6- to 20-fold resistance to this compound at the enzymatic level, whereas the influence of the mutations Gln, Ile, Lys, or Tyr on the resistance/sensitivity of RT to other NNRTIs or NRTIs are very minor (Table 1). Therefore, these amino acid mutations can be theoretically expected to appear under (+)-calanolide A pressure. However, a Thr-to-Ile conversion can emerge by a single transition point mutation (ACAATA), whereas double transversion point mutations (ACACAA) are required for the Thr-to-Gln conversion, and even triple transversion point mutations (ACATAT) are needed for the Thr to Tyr conversion. Therefore, a T139I RT mutation would be more likely to appear than the T139Q or T139Y mutations in HIV-1 RT. Moreover, the (+)-calanolide A-resistant T139K RT (which can also arise from a single point mutation) and mutant T139Y RT have a much lower catalytic activity than the mutant T139I RT enzyme. Thus, our site-directed mutagenesis and kinetic analysis of mutated T139X RTs makes it obvious why the T139I mutation must preferentially occur under (+)-calanolide A pressure in cell culture.
In a previous study using the Y2H RT dimerization assay, (+)-calanolide A had no detectable effect on RT dimerization (Tachedjian et al., 2001; Tachedjian and Goff, 2003). However, failure of an effect of (+)-calanolide A on RT dimerization can be explained by lack of entry of this compound in the yeast cells. Because the kinetic studies with (+)-calanolide A may suggest more than one site of interaction with the HIV-1 RT (Currens et al., 1996b), it would be very interesting to evaluate (+)-calanolide A in other p66-p51 RT dimerization assays. Given the fact that it has been shown that TSAO derivatives affect RT subunit dimerization, the appearance of the T139I resistance mutation under (+)-calanolide A pressure, which is near the TSAO-binding amino acid, Glu138, supports this view. In addition, the fact that the mutant T139D and T139I RT enzymes have increased sensitivity toward the denaturation activity of urea is also supportive of a role for Thr139 in the stabilization of the HIV-1 RT heterodimer. In addition, the lower catalytic efficiency that was observed for the mutant T139I enzyme can also be explained by ensuing structural/conformational differences at the dimerization interface that critically affect proper positioning of the primer/template and/or the orientation of the incoming substrate molecules (i.e., dGTP). Additional support to this hypothesis is provided by the markedly decreased inhibitory activity of the NRTI ddGTP against some of the T139X mutants (Table 1).
The observed hypersensitivity (up to 20-fold) toward the thiocarboxanilide UC-781 (and efavirenz) for the majority of HIV-1 RT mutants, with the exception of T139I, is intriguing. In particular, hypersensitivity of RT to UC-781 as a result of mutations at amino acid position 139 in RT is rather puzzling because UC-781 makes direct contacts with Lys101, Val106, Tyr181, and Phe227 of p66 but not with the p51 subunit (Balzarini et al., 1998; Ren et al., 2004). One possible explanation is that the definite shape that the NNRTI-binding pocket adopts when this inhibitor is lodged into it can be achieved more easily in the presence of these types of mutations.
The protein stretch ranging from Ile135 to Thr139 makes up the tip of the so-called 7-8 loop that is present in both p66 and p51 subunits. An important difference, however, is that in p66, this loop is exposed to the solvent, whereas in p51 it lies close to the dimerization interface (Fig. 1). The importance of the structural support imparted by the 7-8 loop is demonstrated by the severe impairment of the polymerase function of the heterodimeric RT enzyme upon deletion or alanine substitution of amino acids 136-139 (Pandey et al., 2001) and, more specifically, by the dramatic changes in activity taking place upon mutation of Thr139 (Fig. 2). In the absence of direct experimental structural evidence, we suggest that the presence of Lys, Asp, or Tyr at this position is incompatible with the required loop conformation that is necessary for tight interaction with the p66 subunit and essential for catalytic activity. On the other hand, this conformation should virtually remain intact in the case of Thr139Glu and Thr139Ile substitutions and would be perturbed only slightly in the case of Thr139Ala or Thr139Ser mutations.
In conclusion, among the Thr139 RT mutants investigated, T139I RT has the highest resistance profile against (+)-calanolide A. It also retains a marked catalytic activity (85% wild type), which may result in a preferential replication of the mutated T139I RT virus in the presence of (+)-calanolide A, compared with the other mutant RT viruses. In addition, the single transition point mutation ACAATA that is necessary to convert the wild-type Thr139 into the mutant T139I can occur more readily than double- or triple-transversion base changes that code for Gly (102% wild-type activity) or Tyr (0.6% wild-type activity) respectively. Thus, these kinetic properties of the mutant RTs (a pronounced catalytic activity of the T139I RT enzyme and a marked degree of resistance of T139I RT against (+)-calanolide A and the requirement of only one (transition) point mutation in the 139 codon to afford an amino acid substitution may probably explain why the T139I mutation in HIV-1 RT consistently and predominantly appears in cell culture under (+)-calanolide A pressure. Based on the location of the 139-mutation in the crucial 7-8 loop of the p51 subunit in the p66/p51 dimer interface and the more pronounced denaturation effect of urea against mutant T139I and T139D versus wild-type RT enzymes, it can be well possible that (+)-calanolide A affects dimerization of the HIV-1 RT heterodimer and thus represents the prototype compound of a new class of RT dimerization inhibitors.
Acknowledgements
We thank Kris Uyttersprot and Kristien Minner for excellent technical assistance and Christiane Callebaut for fine editorial help.
References
Auwerx J, Esnouf R, De Clercq E, and Balzarini J (2004) Susceptibility of feline immunodeficiency virus/human immunodeficiency virus type 1 reverse transcriptase chimeras to non-nucleoside RT inhibitors. Mol Pharmacol 65: 244-251.
Auwerx J, Van Nieuwenhove J, Rodre猤uez-Barrios F, de Castro S, Velequez S, Ceccherini-Silberstein F, De Clercq E, Camarasa MJ, Perno CF, Gago F, et al. (2005) The N137 and P140 amino acids in the p51 and the P95 amino acid in the p66 subunit of human immunodeficiency virus type 1 (HIV-1) reverse transcriptase are instrumental to maintain catalytic activity and to design new classes of anti-HIV-1 drugs. FEBS Lett 579: 2294-2300.
Balzarini J (1999) Suppression of resistance to drugs targeted to human immunodeficiency virus reverse transcriptase by combination therapy. Biochem Pharmacol 58: 1-27.
Balzarini J (2004) Current status of the non-nucleoside reverse transcriptase inhibitors of human immunodeficiency virus type 1. Curr Top Med Chem 4: 921-944.
Balzarini J, Auwerx J, Rodre猤uez-Barrios F, Chedad A, Ceccherini-Silberstein A, Garce猘-Aparicio C, Velequez S, De Clercq E, Perno CF, Camarasa MJ, et al. (2005) The amino acid Asn136 in HIV-1 reverse transcriptase (RT) maintains efficient association of both RT subunits and enables the rational design of novel RT inhibitors. Mol Pharmacol 68: 49-60.
Balzarini J, Karlsson A, Vandamme A-M, Peereez-Peereez M-J, Zhang H, Vrang L, berg B, Backbro K, Unge T, San-Feelix A, et al. (1993a) Human immunodeficiency virus type 1 (HIV-1) strains selected for resistance against the HIV-1-specific [2',5'-bis-O-(tert-butyldimethylsilyl)-3'-spiro-5''-(4''-amino-1'',2''-oxathiole-2'',2''-dioxide)]--D-pentofuranosyl (TSAO) nucleoside analogues retain sensitivity to HIV-1-specific nonnucleoside inhibitors. Proc Natl Acad Sc USA 90: 6952-6956.
Balzarini J, Pelemans H, Esnouf R, and De Clercq E (1998) A novel mutation (F227L) arises in the reverse transcriptase of human immunodeficiency virus type 1 on dose-escalating treatment of HIV type 1-infected cell cultures with the nonnucleoside reverse transcriptase inhibitor thiocarboxanilide UC-781. AIDS Res Hum Retroviruses 14: 255-260.
Balzarini J, Peereez-Peereez MJ, San-Felix A, Camarasa MJ, Bathurst IC, Barr PJ, and De Clercq E (1992) Kinetics of inhibition of human immunodeficiency virus type 1 (HIV-1) reverse transcriptase by the novel HIV-1-specific nucleoside analogue [2',5'-bis-O-(tert-butyldimethylsilyl)--D-ribofuranosyl]-3'-spiro-5''-(4''-amino-1'',2''-oxathiole-2'',2''-dioxide)thymine (TSAO-T). J Biol Chem 267: 11831-11838.
Balzarini J, Velequez S, San-Feelix A, Karlsson A, Peereez-Peereez MJ, Camarasa MJ, and De Clercq E (1993b) Human immunodeficiency virus type 1-specific [2',5'-bis-O-(tert-butyldimethylsilyl)--D-ribofuranosyl]-3'-spiro-5''-(4''-amino-1'',2''-oxathiole-2'',2''-dioxide)-purine analogues show a resistance spectrum that is different from that of the human immunodeficiency virus type 1-specific non-nucleoside analogues. Mol Pharmacol 43: 109-114.
Boyer PL, Ding J, Arnold E, and Hughes SH (1994) Subunit specificity of mutations that confer resistance to nonnucleoside inhibitors in human immunodeficiency virus type 1 reverse transcriptase. Antimicrob Agents Chemother 38: 1909-1914.
Buckheit RW Jr, Fliakas-Boltz V, Decker WD, Roberson JL, Stup TL, Pyle CA, White EL, McMahon JB, Currens MJ, Boyd MR, et al. (1995) Comparative anti-HIV evaluation of diverse HIV-1-specific reverse transcriptase inhibitor-resistant virus isolates demonstrates the existence of distinct phenotypic subgroups. Antiviral Res 26: 117-132.
Buckheit RW Jr, White EL, Fliakas-Boltz V, Russell J, Stup TL, Kinjerski TL, Osterling MC, Weigand A, and Bader JP (1999) Unique anti-human immunodeficiency virus activities of the nonnucleoside reverse transcriptase inhibitors calanolide A, costatolide and dihydrocostatolide. Antimicrob Agents Chemother 43: 1827-1834.
Camarasa MJ, San-Feelix A, Velequez S, Peereez-Peereez MJ, Gago F, and Balzarini J (2004) TSAO compounds: the comprehensive story of a unique family of HIV-1 specific inhibitors of reverse transcriptase. Curr Top Med Chem 4: 945-963.
Ceccherini-Silberstein F, Gago F, Santoro M, Gori C, Svicher V, Rodre猤uez-Barrios F, D'Arrigo R, Ciccozzi M, Bertoli A, d'Arminio Monforte A, et al. (2005) High sequence conservation of HIV-1 reverse transcriptase under drug pressure despite a continuous appearance of mutations. J Virol, in press.
Currens MJ, Gulakowski RJ, Mariner JM, Moran RA, Buckheit RW Jr, Gustafson KR, McMahon JB, and Boyd MR (1996a) Antiviral activity and mechanism of action of calanolide A against the human immunodeficiency virus type-1. J Pharmacol Exp Ther 279: 645-651.
Currens MJ, Mariner JM, McMahon JB, Boyd MR, Shulman N, and Winters M (1996b) Kinetic analysis of inhibition of human immunodeficiency virus type-1 reverse transcriptase by calanolide A. J Pharmacol Exp Ther 279: 652-661.
D'Aquila RT and Summers WC (1989) HIV-1 reverse transcriptase/ribonuclease H: high level expression in Escherichia coli from a plasmid constructed using the polymerase chain reaction. J Acquir Immune Defic Syndr 2: 579-587.
De Clercq E (1999) Perspectives of non-nucleoside reverse transcriptase inhibitors (NNRTIs) in the therapy of HIV-1 infection. Il Farmaco 54: 26-45.
DeLano WL (2004) The PyMOL Molecular Graphics System. DeLano Scientific LLC, San Carlos, CA.
Esnouf RM, Ren J, Hopkins AL, Ross CK, Jones EY, Stammers DK, and Stuart DI (1997) Unique features in the structure of the complex between HIV-1 reverse transcriptase and the bis(heteroaryl)piperazine (BHAP) U-90152 explain resistance mutations for this nonnucleoside inhibitor. Proc Natl Acad Sci USA 94: 3984-3989.
Huang H, Chopra R, Verdine GL, and Harrison SC (1998) Structure of a covalently trapped catalytic complex of HIV-1 reverse transcriptase: implications for drug resistance. Science (Wash DC) 282: 1669-1675.
Jonckheere H, De Vreese K, Debyser Z, Vandekerckhove J, Balzarini J, Desmyter J, De Clercq E, and Annee J (1996) A two plasmid co-expression system in Escherichia coli for the production of virion-like reverse transcriptase of the human immunodeficiency virus type 1. J Virol Methods 61: 113-125.
Kashman Y, Gustafson KR, Fuller RW, Cardellina JH II, McMahon JB, Currens MJ, Buckheit RW Jr, Hughes SH, Cragg GM, and Boyd MR (1992) The calanolides, a novel HIV-inhibitory class of coumarin derivatives from the tropical rainforest tree, Calophyllum lanigerum. J Med Chem 35: 2735-2743.
Kohlstaedt LA, Wang J, Friedman JM, Rice PA, and Steitz TA (1992) Crystal structure at 3.5 A resolution of HIV-1 reverse transcriptase complexed with an inhibitor. Science (Wash DC) 256: 1783-1790.
Pandey PK, Kaushik N, Singh K, Sharma B, Upadhyay AK, Kumar S, Harris D and Pandey VN (2002) Insertion of a small peptide of six amino acids into the beta7-beta8 loop of the p51 subunit of HIV-1 reverse transcriptase perturbs the heterodimer and affects its activities. BMC Biochem 3: 18.
Pandey PK, Kaushik N, Talele TT, Yadav PN, and Pandey VN (2001) The beta7-beta8 loop of the p51 subunit in the heterodimeric (p66/p51) human immunodeficiency virus type 1 reverse transcriptase is essential for the catalytic function of the p66 subunit. Biochemistry 40: 9505-9512.
Pelemans H, Aertsen A, Van Laethem K, Vandamme A-M, De Clercq E, Peereez-Peereez M-J, San-Feelix A, Velequez S, Camarasa M-J, and Balzarini J (2001) Site-directed mutagenesis of human immunodeficiency virus type 1 reverse transcriptase at amino acid position 138. Virology 280: 97-106.
Pelemans H, Esnouf RM, Jonckheere H, De Clercq E, and Balzarini J (1998) Mutational analysis of Tyr-318 within the non-nucleoside reverse transcriptase inhibitor binding pocket of human immunodeficiency virus type I reverse transcriptase. J Biol Chem 273: 34234-34239.
Peerez-Peerez M-J, San-Feelix A, Balzarini J, De Clercq E, and Camarasa M-J (1992) TSAO analogues. Stereospecific synthesis and anti-HIV-1 activity of 1-[2',5'-bis-O-(tert-butyldimethylsilyl)-beta-D-ribofuranosyl]-3'-spiro-5''-(4''-amino-1'',2''-oxathiole 2'',2''-dioxide) pyrimidine and pyrimidine-modified nucleosides. J Med Chem 35: 2988-2995.
Quan Y, Motakis D, Buckheit R Jr, Xu ZQ, Flavin MT, Parniak MA, and Wainberg MA (1999) Sensitivity and resistance to (+)-calanolide A of wild-type and mutated forms of HIV-1 reverse transcriptase. Antivir Ther 4: 203-209.
Ren J, Nichols CE, Chamberlain PP, Weaver KL, Short SA, and Stammers DK (2004) Crystal structures of HIV-1 reverse transcriptases mutated at codons 100, 106 and 108 and mechanisms of resistance to non-nucleoside inhibitors. J Mol Biol 336: 569-578.
Rodre猤uez-Barrios F, Peerez C, Lobaton E, Velequez S, Chamorro C, San-Feelix A, Peerez-Peerez M-J, Camarasa M-J, Pelemans H, Balzarini J, and Gago F (2001) Identification of a putative binding site for [2',5'-bis-O-(tert-butyldimethylsilyl)-beta-D-ribofuranosyl]-3'-spiro-5''-(4''-amino-1'',2''-oxathiole-2'',2''-dioxide)thymine (TSAO) derivatives at the p51-p55 interface of HIV-1 reverse transcriptase. J Med Chem 44: 1853-1865.
Sluis-Cremer N, Arion D, and Parniak MA (2002) Destabilization of the HIV-1 reverse transcriptase dimer upon interaction with N-acyl hydrazone inhibitors. Mol Pharmacol 62: 398-405.
Sluis-Cremer N, Dmitrienko GI, Balzarini J, Camarasa M-J, and Parniak MA (2000) 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-ethylthymine. Biochemistry 39: 1427-1433.
Tachedjian G and Goff SP (2003) The effect of NNRTIs on HIV reverse transcriptase dimerization. Curr Opin Investig Drugs 4: 966-973.
Tachedjian G, Orlova M, Sarafianos SG, Arnold E, and Goff SP (2001) Nonnucleoside reverse transcriptase inhibitors are chemical enhancers of dimerization of the HIV type 1 reverse transcriptase. Proc Natl Acad Sci USA 98: 7188-7193.
Vandamme A-M, Van Laethem K, Van Vaerenbergh K, and De Clercq E (1998) Anti-HIV virus combination therapy and resistance management. Int Antiviral News 6: 182-187.
Yu D, Suzuki M, Xie L, Morris-Natschke SL, and Lee KH (2003) Recent progress in the development of coumarin derivatives as potent anti-HIV agents. Med Res Rev 23: 322-345.