Selective inhibition of HIV-1 reverse transcriptase-associated ribonuc
http://www.100md.com
《核酸研究医学期刊》
Resistance Mechanisms Laboratory, HIV Drug Resistance Program, National Cancer Institute at Frederick Frederick, MD 21702, USA 1Protein Biophysics Resource, Division of Bioengineering and Physical Sciences, National Institutes of Health Bethesda, MD 20892, USA 2Laboratory of Molecular Genetics, National Institute of Child Health and Human Development Bethesda, MD 20892, USA 3Molecular Targets Development Program, National Cancer Institute at Frederick Frederick, MD 21702, USA 4SAIC-Frederick, Frederick MD 21702, USA 5Division of Infectious Diseases, School of Medicine, University of Pittsburgh Pittsburgh, PA 15213, USA
*To whom correspondence should be addressed. Tel: +1 301 846 5256; Fax: +1 301 846 6013; Email: slegrice@ncifcrf.gov
ABSTRACT
High-throughput screening of a National Cancer Institute library of pure natural products identified the hydroxylated tropolone derivatives ?-thujaplicinol (2,7-dihydroxy-4-1(methylethyl)-2,4,6-cycloheptatrien-1-one) and manicol (1,2,3,4-tetrahydro-5-7-dihydroxy-9-methyl-2-(1-methylethenyl)-6H-benzocyclohepten-6-one) as potent and selective inhibitors of the ribonuclease H (RNase H) activity of human immunodeficiency virus-type 1 reverse transcriptase (HIV-1 RT). ?-Thujaplicinol inhibited HIV-1 RNase H in vitro with an IC50 of 0.2 μM, while the IC50 for Escherichia coli and human RNases H was 50 μM and 5.7 μM, respectively. In contrast, the related tropolone analog ?-thujaplicin (2-hydroxy-4-(methylethyl)-2,4,6-cycloheptatrien-1-one), which lacks the 7-OH group of the heptatriene ring, was inactive, while manicol, which possesses a 7-OH group, inhibited HIV-1 and E.coli RNases H with IC50 = 1.5 μM and 40 μM, respectively. Such a result highlights the importance of the 2,7-dihydroxy function of these tropolone analogs, possibly through a role in metal chelation at the RNase H active site. Inhibition of HIV-2 RT-associated RNase H indirectly indicates that these compounds do not occupy the nonnucleoside inhibitor-binding pocket in the vicinity of the DNA polymerase domain. Both ?-thujaplicinol and manicol failed to inhibit DNA-dependent DNA polymerase activity of HIV-1 RT at a concentration of 50 μM, suggesting that they are specific for the C-terminal RNase H domain, while surface plasmon resonance studies indicated that the inhibition was not due to intercalation of the analog into the nucleic acid substrate. Finally, we have demonstrated synergy between ?-thujaplicinol and calanolide A, a nonnucleoside inhibitor of HIV-1 RT, raising the possibility that both enzymatic activities of HIV-1 RT can be simultaneously targeted.
INTRODUCTION
Reverse transcriptase (RT)-associated ribonuclease H (RNase H) activity is responsible for both non-specifically degrading the RNA strand of the RNA/DNA replication intermediate as well as specifically removing the minus (–) and plus (+) strand RNA primers from nascent DNA (1). The absolute requirement for RNase H activity for human immunodeficiency virus (HIV) replication (2,3) suggests that this might be an attractive target for the development of antiviral agents to complement DNA polymerase-based HIV-1 RT inhibitors currently in clinical use . In this respect, recent reports have documented several promising candidates effective at low micromolar concentrations, including hydrazones (5–7), tetragalloylglucopyranose (8), diketo acids (9) and N-hydroxyimides (10). Although it remains to be established that their mode of inhibition is through direct binding to the RNase H catalytic center, both diketo acids and N-hydroxyimides have been shown to inhibit an enzymatically active peptide derived from the RNase H domain of HIV-1 RT (9,11). Thus, while antiviral activity of these select RNase H antagonists is yet to be demonstrated, sufficient evidence has accumulated to justify further screening for inhibitors of HIV-1 and HIV-2 RNase H. Moreover, although Klumpp et al. (10) have recently evaluated N-hydroxyimides for the inhibition of retroviral and bacterial RNases H, none of the more potent HIV-1 RNase H inhibitors reported to date has addressed selectivity with respect to the human counterpart. This issue is of particular importance, since recent data have indicated that disruption of the RNase H1 allele in mice confers a lethal embryonic defect (12). To take this issue into account, we recently described two simple, robust fluorescence-based methodologies which could be applied in parallel to retroviral, bacterial and human RNases H using the same assay platform, together with their use for high-throughput screening (13,14). The availability of purified human RNase H1 has also allowed us to investigate the specificity of inhibitors uncovered from our current screening efforts.
Here, we describe two tropolone (2-hydroxy-2,4,6-cycloheptatrien-1-one) derivatives with a 7-OH substitution, identified from high-throughput screening of National Cancer Institute libraries of pure natural products, which inhibit retroviral, bacterial and human RNases H. Inhibition of HIV-2 RNase H activity indirectly indicates that these compounds do not interact with the nonnucleoside-binding site located near the DNA polymerase catalytic center of this enzyme. The most potent of the hydroxylated tropolone analogs, ?-thujaplicinol, derived from the heartwood of several cupressaceous plants (e.g. Thuja plicata, Thuja occidentalis and Chamaecyparis obtusaI) (15–17), is active at sub-micromolar concentrations against the RNases H of HIV-1 and HIV-2 RT, while the DNA polymerase activity of either enzyme is unaffected at an inhibitor concentration of 50 μM. The same compound is 30- to 250-fold less active against the human and Escherichia coli RNases H, respectively, demonstrating that selective inhibition of the retroviral enzyme can be achieved. Finally, we demonstrate here that ?-thujaplicinol acts synergistically with calanolide A, a nonnucleoside inhibitor of HIV-1 RT (18,19), opening the possibility of simultaneously targeting the DNA polymerase and RNase H functions of HIV-1 and HIV-2 RT.
A number of reports have demonstrated that tropolone derivatives elicit a variety of biological effects, including anti-tumor (20), insecticidal (21), antifungal (22,23) and antimicrobial (24) activity, while their metal chelates have been shown to inhibit human influenza virus-induced apoptosis (25). Wakabayashi et al. (26) have investigated cytotoxic activity of tropolone derivatives against a variety of human oral tumor cell lines, and also indicated that they fail to protect MT-4 cells from HIV-1 infection.
MATERIALS AND METHODS
Tropolone derivatives
The structures of -thujaplicin , ?-thujaplicin , -thujaplicin , ?-thujaplicinol , manicol and nootkatin are provided in Figure 1. All compounds were members of a National Cancer Institute library of purified, natural compounds, and their identity and purity were confirmed by 1H-NMR. Each compound was prepared as a 20 mM stock solution in 100% dimethyl sulfoxide (DMSO) and diluted into the appropriate reaction buffer immediately prior to analysis.
Figure 1 Structures of tropolone and its derivatives.
Enzymes and enzyme assays
Recombinant, wild-type p66/p51 HIV-1 RT was expressed and purified as described previously (27). HIV-2 RT was prepared using the same E.coli expression system (27). E.coli RNase HI and recombinant human RNase H were prepared as described previously (28,29). The strategy for high-throughput screening and confirmation of RNase H activity by capillary electrophoresis has recently been described by Parniak et al. (13) and Chan et al. (14), respectively. For HIV-1 RT, conditions for cleavage of the HIV-1 PPT RNA primer extended at its 3' terminus by five deoxynucleotides have been described by Kvaratskhelia et al. (30). DNA-dependent DNA polymerase activity was analyzed on a 71 nt DNA to which a labeled 36 nt DNA primer was hybridized as described by Dash et al. (31). For qualitative analysis of PPT cleavage and DNA-dependent DNA polymerase activity, all tropolone derivatives were evaluated at a single concentration of 50 μM.
Surface plasmon resonance analysis
All experiments were performed with a Biacore 2000 optical biosensor at 25°C. CM5 Sensor chips with a carboxymethylated dextran matrix, EDC, NHS and P20 surfactant were obtained from Biacore AB, (Uppsala, Sweden), and streptavidin was from Pierce (Rockford, IL). Streptavidin was immobilized on the sensor chip surface to the level of 4000–4500 RU in each flow cell. Biotinylated nucleic acid (12mer RNA–DNA hybrid in the form of a hairpin within which a thymidine residue of a connection tetraloop was biotinylated) was injected at a concentration of 50 nM in HBS buffer (10 mM HEPES, 150 mM NaCl, 1 mM EDTA, 0.005% P20, pH 7.4) onto streptavidin-modified chip surface at a flow rate of 5 μl/min until the desired level (1000–1500 RU) of immobilization was achieved. Blocking of unoccupied streptavidin with biotin and washing the chip with regeneration buffer (2 M NaCl) was performed as described previously (28). Experiments on inhibitor or marker binding to nucleic acid were carried out in Tris-buffered saline/DMSO running buffer (10 mM Tris–HCl, 50 mM NaCl, 1 mM EDTA, 0.005% P20, 5% DMSO, pH 8.0) using a flow rate of 50 μl/min. Typically, 50 μl of a marker or inhibitor (each separately or in mixtures of 2 or 3 simultaneously) diluted in the same buffer in the concentration range of 50 nM to 10 μM was successively injected using the ‘kinject’ command defining a dissociation time of 2 min. After each round of injection, the system was intensively washed using running buffer. All experiments were performed in duplicate. Data transformation of the primary sensograms and overlay plots were prepared using the BIAevaluation 3.0 software (Biacore AB). The response from the reference flow cell that does not contain nucleic acid was subtracted from that of the experimental flow cell.
IC50 and Ki determinations
Dose–response plots were prepared by measuring RNase H activity on a fluorescent RNA/DNA hybrid as described previously (13) while varying the inhibitor concentration by 2-fold dilutions from 100 μM to 0.19 nM. Curve fitting was performed using Sigma Plot. In order to determine the Ki for ?-thujaplicinol, the substrate concentration was varied from 200 to 1.56 nM while varying the inhibitor concentration from 1.56 to 0.024 nM at each substrate concentration. Hydrolysis was followed by fluorescence measurements at 40 s intervals over the 30 min reaction time using a Molecular Devices Spectramax Gemini EM fluorescence spectrometer. Slopes of the reaction curves were determined from the linear portion of the plots using SoftMAX Pro version 4.3 supplied with the instrument. Curve fitting and prediction of the mode of inhibition was performed using the Enzyme Kinetics module of Sigma Plot version 8 (SPSS, Chicago, IL).
Synergy between RNase H and DNA polymerase inhibitors
In order to determine whether the binding site for ?-thujaplicinol overlapped that of nonnucleoside RT inhibitors, RNase H activity was measured in the presence of varying concentrations of tropolone inhibitor and varying concentrations of the nonnucleoside calanolide A (18) according to the method of Yonetani and Theorell (32) and Cook et al. (33). Data were processed using the Calcusyn software package (BioSoft, Cambridge, UK).
RESULTS
Tropolone derivatives specifically inhibit HIV-1 RNase H function
The structures of the tropolone derivatives used in the present study are provided in Figure 1. Since HIV-1 RT is a bifunctional enzyme, inhibition of RNase H activity might reflect direct binding to the RNase H domain or, alternatively, an allosteric effect through binding to the nonnucleoside-binding pocket close to the DNA polymerase domain, which has been shown previously to modulate RNase H function (34). Preliminary experiments were, therefore, designed to investigate the specificity of each analog for the RNase H activity of HIV-1 RT. In Figure 2A, RNase H activity was evaluated on a 5'labeled RNA–DNA chimera hybridized to its DNA complement, which mimics processing of the HIV-1 PPT primer from nascent DNA, following initiation of second or (+) strand synthesis. Using this substrate, we have previously demonstrated RNase H cleavage at the PPT RNA-U3 DNA junction (defined as –1), and additionally at positions –2 and –3 (30). For these initial experiments, a single analog concentration of 50 μM was employed, which resulted in almost complete inhibition of PPT processing by both ?-thujaplicinol (lane 4) and manicol (lane 7). In Figure 2B, the same derivatives were re-evaluated for their ability to inhibit DNA-dependent DNA polymerase activity on a duplex containing a 5'labeled 36 nt DNA primer hybridized to a 71 nt DNA template. At a single concentration of 50 μM, negligible inhibition of DNA polymerase activity was observed for all analogs. The combined data of Figure 2 thus indicate that inhibition of HIV-1 RNase H activity is not an consequence of analog binding to the DNA polymerase catalytic center and exerting an indirect effect, as has been reported for certain diketo acids and phenylhydrazones by Shaw-Reid et al. (9).
Figure 2 (A) Inhibition of HIV-1 RT-catalyzed polypurine tract processing by tropolone and its derivatives. Left, schematic representation of the RNase H assay. Right, RNase H activity in the presence of 50 μM tropolone derivative. Lane C, no inhibitor; lane 1, no inhibitor, +DMSO; lane 2, ?-thujaplicin; lane 3, -thujaplicin; lane 4, ?-thujaplicininol; lane 5, -thujaplicin; lane 6, nootkatin; lane 7, manicol; and lane 8, tropolone. PPT hydrolysis products have been indicated. (B) DNA-dependent DNA polymerase activity of HIV-1 RT in the presence of tropolone and hydroxylated derivatives. DNA polymerase activity was determined at a single inhibitor concentration of 50 μM. Lane notations are as in (A).
Selectivity of inhibition
Clearly, a desirable feature of an inhibitor of retroviral RT-associated RNase H activity would be selectivity with respect to its human counterpart. To address this issue, ?-thujaplicinol and manicol were examined for their inhibitory effect on the RTs of HIV-1 and HIV-2 as well as human RNase H. Figure 3A indicates that ?-thujaplicinol inhibited HIV-1 RT with an IC50 of 0.21 ± 0.03 μM. For HIV-2 RT, the IC50 for the same analog was 3-fold higher (Figure 3C, 0.77 ± 0.08 μM). Such a result might reflect subtle differences in the architecture of RNase H domains of these two enzymes, which has been suggested from biochemical analysis of the purified RTs (35). More importantly, ?-thujaplicinol inhibited human RNase H with an IC50 of 5.7 ± 0.7 μM (Figure 3E), indicating 30-fold selectivity for the HIV-1 enzyme. Although not shown here, an IC50 value of 50 μM was determined for E.coli RNase H, indicating that this enzyme was 250-fold less sensitive to ?-thujaplicinol inhibition. In Figure 3B, D and F, inhibition of retroviral and human RNases H by manicol was compared. While this analog was slightly less potent against HIV-1 RNase H (IC50 = 0.60 ± 0.09 μM), 6-fold enhanced selectivity over the human enzyme was achieved (IC50 = 3.5 ± 0.1 μM).
Figure 3 Selectivity of RNase H inhibition. Dose–response curves for RNase I inhibition by ?-thujaplicinol (A, C and E) and manicol (B, D and F) are presented. (A and B) HIV-1 RT; (C and D), HIV-2 RT; (E and F), human RNase H. IC50 determinations are the results of triplicate assays.
IC50 values for tropolone and its derivatives are presented in Table 1. Interestingly, ?-thujaplicin, which differs from ?-thujaplicinol in that it lacks the hydroxyl function at position 7 of the heptatriene ring, was completely inactive against all enzymes tested, despite reports that it possesses metal chelating properties (36). Relocation of the hydroxyl function on the heptatriene ring produced a different effect, i.e. while -thujaplicin failed to inhibit retroviral RNases H, -thujaplicin was weakly active, with an IC50 value of 50 and 33 μM for the HIV-1 and HIV-2 enzymes, respectively.
Table 1 Inhibition of retroviral, bacterial and human RNases H by hydroxylated tropolone derivatives
?-Thujaplicinol does not inhibit RNase H activity through intercalation
Although DNA polymerase activity was unaffected, one possible mode of inhibition we had to consider was the intercalation of the planar tropolone derivatives into the RNA/DNA hybrid, effectively making the substrate unavailable for enzyme binding. In order to eliminate this possibility, inhibitor binding to an immobilized RNA/DNA hybrid was investigated via surface plasmon resonance, which we have successfully used to investigate the binding properties of HIV-1 RT derivatives (37). As a control, binding of ethidium bromide to the immobilized hybrid was analyzed, the results of which are depicted in the sensorgram of Figure 4A. From this data, we calculated that 4–6 molecules of ethidium bromide intercalated into the hybrid with a Kd of 1.5 μM. In contrast, data of Figure 4B, which presents the combined sensorgrams obtained when the RNA/DNA hybrid was incubated with -thujaplicin, ?-thujaplicin and ?-thujaplicinol, clearly indicate that they do not bind the substrate over a concentration range of 50 nM–10 mM.
Figure 4 Analysis of inhibitor intercalation by surface plasmon resonance. In (A), binding of ethidium bromide to an immobilized 12mer RNA/DNA hybrid was performed as a control. Sensorgrams a–f illustrate the response to successive injection of the intercalator at concentrations of 10, 5, 1.0, 0.5, 0.25 and 0.1 μM, respectively. (B) Combined sensorgrams following injection of ?-thujaplicin, -thujaplicin and ?-thujaplicinol over the same concentration range as in (A).
Mode of inhibition
?-Thujaplicin and manicol were further investigated in order to gain insight into their mechanism of inhibition of HIV-1 RT/RNase H. From the Michaelis–Menten plot of Figure 5, a Ki of 0.20 ± 0.07 μM was determined for ?-thujaplicinol, and a separate study indicated a Ki of 1.0 ± 0.4 μM for manicol (data not shown). The data presented in Figure 5 show a change in Vmax without affecting the Km, which is suggestive of mixed inhibition. Non-linear fitting of the data confirms that this is best fit to mixed-type inhibition compared with competitive or uncompetitive modes. With this type of inhibition, the antagonist can bind either the free enzyme or the enzyme–substrate complex. This suggests that the tropolone derivatives could occupy (or create) a pocket near the RNase H active site, since access to that site would be impaired by the RNA/DNA duplex substrate.
Figure 5 Michaelis–Menten plot for the determination of the Ki for ?-thujaplicinol for HIV-1 RT-associated RNase H. A Ki of 0.20 ± 0.07 μM was determined from non-linear curve fitting. Inhibitor concentrations were 0.024 μM (open circle), 0.049 μM (filled triangle), 0.098 μM (open triangle), 0.20 μM (filled square), 0.39 μM (open square), 0.78 μM (filled diamond), 1.56 μM (open diamond), RNase H activity in the presence of 5% DMSO (filled circle).
?-Thujaplicinol and the nonnucleoside calanolide A occupy different sites on HIV-1 RT
The multifunctional nature of HIV-1 RT opens the possibility that antiviral agents could be separately targeted to the DNA polymerase and RNase H catalytic centers. In order to assess this, we investigated whether ?-thujaplicinol would act in combination with calanolide A (18). This coumarin derivative, isolated from the tropical rainforest tree Calophyllum lanigerum, has been shown to inhibit HIV-1 RT activity by a mechanism similar to other nonnucleoside-based drugs (19). Figure 6 presents a Yonetani–Theorell plot of RNase H inhibition in the presence of both calanolide A and ?-thujaplicinol. The data indicate that the inhibitors bind to two independent sites since the plot shows converging lines (32). In addition, HIV-2 is naturally resistant to NNRTIs; thus, inhibition of HIV-2 RNase H is additional, albeit indirect, evidence that ?-thujaplicinol binds to an independent site. Independent binding sites for calanolide A and ?-thujaplicinol were further supported by inhibitor-combination assays using the methods described by Chou and Talalay (38,39). The combination index (CI) derived for mixed inhibitors can be used to describe additivity (CI = 1), synergy (CI < 1) and antagonism (CI > 1). The inhibitor combination of ?-thujaplicinol/calanolide A was found to be 0.50, i.e. the inhibition by a ?-thujaplicinol/calanolide A combination is more potent than predicted by adding the effect of the individual inhibitors. This synergistic effect would not be observed with inhibitors competing for the same or closely positioned binding sites. In a control experiment, a CI of 1.8 was determined for a ?-thujaplicinol/manicol combination, indicating antagonism. This would be predicted, since the two tropolone derivatives most likely compete for the same binding site.
Figure 6 Yonetani–Theorell plot for the inhibition of HIV-1 RT in the presence of the NNRTI calanolide A and ?-thujaplicinol. The inverse of the rate of RNase H cleavage was plotted as a function of ?-thujaplicinol concentration at calanolide A concentrations of 12.5 μM (open square), 0.78 μM (filled square), 0.39 μM (open circle) and DMSO (filled circle). The convergent best-fit lines indicate mutually exclusive binding sites for calanolide A and ?-thujaplicinol.
DISCUSSION
In this communication, we demonstrate that the natural products ?-thujaplicinol and manicol, derived from the heartwood of several cupressaceous trees, inhibit the RNase H activity of HIV-1 and HIV-2 RT at sub-micromolar concentrations without affecting DNA polymerase function of either enzyme. Moreover, the most potent inhibitor, ?-thujaplicinol, is 30-fold more active on HIV-1 RT/RNase H than on human RNase H1. In light of the requirement for RNase H1 during embryonal development in mice (12), selectivity of the retroviral enzyme over the human counterpart should be a major goal during antiviral drug development, but to date had not been addressed. Finally, our data indicate that ?-thujaplicinol and the nonnucleoside inhibitor calanolide A (18,19) occupy different sites on HIV-1 RT, raising the possibility that the DNA polymerase and RNase H domains of HIV-1 RT might be simultaneously targeted by antiviral agents.
Although the mechanistic basis for retroviral RNase H inhibition observed here remains to be established, the data of Figure 1 and Table 1 provide some insights. Since the ?-diketone moiety of thujaplicins has been demonstrated capable of chelating metals, such as Cu++ and Zn++ (36), chelating and/or altering the co-ordination geometry of the divalent metal in the RNase H catalytic center essential for catalysis would seem a plausible mechanism of action. However, we observe here that neither ?-thujaplicin nor -thujaplicin is active against the retroviral enzymes, while -thujaplicin is only modestly active (Table 1). ?-thujaplicinol introduces a second hydroxyl function at position 7 of the heptatriene ring, which substantially improves potency. At the same time, although manicol introduces a bulky substituent onto the heptatriene ring, it provides almost the same potency against HIV-1 RNase H as ?-thujaplicinol. While addition of the 7-OH function enhances the specificity for HIV-1 and HIV-2 RT/RNase H, it has little to no effect on potency against the bacterial and human enzymes. Poor inhibition of E.coli RNase H and the consistent but modest inhibition of human RNase H by all tropolones tested suggests that if metal chelation is responsible for inhibition, the 7-OH function is critical for stabilizing an interaction, which is specific to the active-site geometry of the retroviral enzymes. Interestingly, a two-metal catalyzed catalytic mechanism has been proposed for HIV-1 RNase H, based on crystallographic data with the isolated Mn++-doped domain showing two divalent cations at the active site (40). Although speculative, the potency we observe for ?-thujaplicinol may reflect its ability to form a more stable complex with the divalent cation at both metal-binding sites. If an interaction with the divalent cation at the RNase H catalytic center is the basis for inhibition by ?-thujaplicinol and manicol, it is interesting to note that these two compounds do not inhibit DNA polymerase function, despite the fact that the equivalent divalent cation is coordinated in this catalytic center by a triad of carboxyate residues (Asp110, Asp185 and Asp186). One possible explanation might be that the divalent cation is more tightly bound at the DNA polymerase catalytic center and thus more difficult to access. However, titration calorimetry studies suggest the contrary, namely that the divalent cation in the RNase H catalytic center is more tightly coordinated (41). Based on these observations, differences in the overall architecture of the metal-binding site might exclude these inhibitors from the polymerase catalytic center. Further studies of the interaction of ?-thujaplicinol and manicol with HIV-1 and HIV-2 RT have the potential to clarify their mechanism of action. While inactive in cellular models of HIV-1 infection, these two compounds nonetheless will serve as useful probes of HIV-1 and HIV-2 RNase H enzymology. For example, recent studies have suggested that inhibition of HIV-1 RNase H activity might increase the rate at which chain terminators of DNA synthesis are excised, thus leading to enhanced nucleoside resistance (V. Pathak, personal communication). The availability of inhibitors selective for the RNase H activity of HIV-1 RT will allow this hypothesis to be directly evaluated in vitro.
ACKNOWLEDGEMENTS
We would like to thank the Drug Synthesis and Chemistry Branch of the National Cancer Institute for the pure natural product library and for providing the tropolone derivatives. Funding to pay the Open Access publication charges for this article was provided by the National Cancer Institute, National Institutes of Health, Department of Health and Human Services.
REFERENCES
Klarmann, G.J., Hawkins, M.E., Le Grice, S.F. (2002) Uncovering the complexities of retroviral ribonuclease H reveals its potential as a therapeutic target AIDS Rev., 4, 183–194 .
Schatz, O., Cromme, F., Naas, T., Lindemann, D., Gruninger-Leitch, F., Mous, J., Le Grice, S.F.J. (1990) Inactivation of the RNaseH domain of HIV-1 reverse transcriptase blocks viral infectivity In Papas, T. (Ed.). Oncogenesis and AIDS, Houston, TX Portfolio Publishing Company pp. 55–68 .
Tisdale, M., Schulze, T., Larder, B.A., Moelling, K. (1991) Mutations within the RNase H domain of human immunodeficiency virus type 1 reverse transcriptase abolish virus infectivity J. Gen. Virol., 72, 59–66 .
Parniak, M.A. and Sluis-Cremer, N. (2000) Inhibitors of HIV-1 reverse transcriptase Adv. Pharmacol., 49, 67–109 .
Borkow, G., Fletcher, R.S., Barnard, J., Arion, D., Motakis, D., Dmitrienko, G.I., Parniak, M.A. (1997) Inhibition of the ribonuclease H and DNA polymerase activities of HIV-1 reverse transcriptase by N-(4-tert-butylbenzoyl)-2-hydroxy-1-naphthaldehyde hydrazone Biochemistry, 36, 3179–3185 .
Gabbara, S., Davis, W.R., Hupe, L., Hupe, D., Peliska, J.A. (1999) Inhibitors of DNA strand transfer reactions catalyzed by HIV-1 reverse transcriptase Biochemistry, 38, 13070–13076 .
Davis, W.R., Tomsho, J., Nikam, S., Cook, E.M., Somand, D., Peliska, J.A. (2000) Inhibition of HIV-1 reverse transcriptase-catalyzed DNA strand transfer reactions by 4-chlorophenylhydrazone of mesoxalic acid Biochemistry, 39, 14279–14291 .
Min, B.S., Nakamura, N., Miyashiro, H., Kim, Y.H., Hattori, M. (2000) Inhibition of human immunodeficiency virus type 1 reverse transcriptase and ribonuclease H activities by constituents of Juglans mandshurica Chem. Pharm. Bull. (Tokyo), 48, 194–200 .
Shaw-Reid, C.A., Munshi, V., Graham, P., Wolfe, A., Witmer, M., Danzeisen, R., Olsen, D.B., Carroll, S.S., Embrey, M., Wai, J.S., et al. (2003) Inhibition of HIV-1 ribonuclease H by a novel diketo acid, 4--2,4-dioxobutanoic acid J. Biol. Chem., 278, 2777–2780 .
Klumpp, K., Hang, J.Q., Rajendran, S., Yang, Y., Derosier, A., Wong Kai In, P., Overton, H., Parkes, K.E., Cammack, N., Martin, J.A. (2003) Two-metal ion mechanism of RNA cleavage by HIV RNase H and mechanism-based design of selective HIV RNase H inhibitors Nucleic Acids Res., 31, 6852–6859 .
Hannoush, R.N., Carriero, S., Min, K.L., Damha, M.J. (2004) Selective inhibition of HIV-1 reverse transcriptase (HIV-1 RT) RNase H by small RNA hairpins and dumbbells Chembiochem, 5, 527–533 .
Cerritelli, S.M., Frolova, E.G., Feng, C., Grinberg, A., Love, P.E., Crouch, R.J. (2003) Failure to produce mitochondrial DNA results in embryonic lethality in Rnaseh1 null mice Mol. Cell, 11, 807–815 .
Parniak, M.A., Min, K.L., Budihas, S.R., Le Grice, S.F., Beutler, J.A. (2003) A fluorescence-based high-throughput screening assay for inhibitors of human immunodeficiency virus-1 reverse transcriptase-associated ribonuclease H activity Anal. Biochem., 322, 33–39 .
Chan, K.C., Budihas, S.R., Le Grice, S.F., Parniak, M.A., Crouch, R.J., Gaidamakov, S.A., Isaaq, H.J., Wamiru, A., McMahon, J.B., Beutler, J.A. (2004) A capillary electrophoretic assay for ribonuclease H activity Anal. Biochem., 331, 296–302 .
Gardner, J.A.F., Barton, G.M., MacLean, H. (1957) Occurrence of 2,7-dihydroxy-4-isopropyl-2,4,6-cycloheptatrien-1-one (7-hydroxy-4-isopropyltropolone) in western red cedar (Thuja plicata Donn.) Can J. Chem., 35, 1039–1048 .
Zavarin, E., Smith, L.V., Bicho, J.G. (1967) Tropolones of Cupressaceae—III Phytochemistry, 6, 1387–1394 .
Polonsky, J., Beloeil, J.-C., Prange, T., Pascard, C., Lacquemin, H., Donnelly, D.M.X., Kenny, P. (1983) Manicol: a sesquiterpenoid hydroxytropolone from Dulacia guianensis. A revised structure (X-ray analysis) Tetrahedron, 39, 2647–2655 .
Kashman, Y., Gustafson, K.R., Fuller, R.W., Cardellina, J.H., II, McMahon, J.B., Currens, M.J., Buckheit, R.W., Jr, Hughes, S.H., Cragg, G.M., Boyd, M.R. (1992) The calanolides, a novel HIV-inhibitory class of coumarin derivatives from the tropical rainforest tree, Calophyllum lanigerum J. Med. Chem., 35, 2735–2743 .
Boyer, P.L., Currens, M.J., McMahon, J.B., Boyd, M.R., Hughes, S.H. (1993) Analysis of nonnucleoside drug-resistant variants of human immunodeficiency virus type 1 reverse transcriptase J. Virol., 67, 2412–2420 .
Sugawara, K., Ohbayashi, M., Shimizu, K., Hatori, M., Kamei, H., Konishi, M., Oki, T., Kawaguchi, H. (1988) BMY-28438 (3,7-dihydroxytropolone), a new antitumor antibiotic active against B16 melanoma. I. Production, isolation, structure and biological activity J. Antibiot. (Tokyo), 41, 862–868 .
Morita, Y., Matsumura, E., Okabe, T., Shibata, M., Sugiura, M., Ohe, T., Tsujibo, H., Ishida, N., Inamori, Y. (2003) Biological activity of tropolone Biol. Pharm. Bull., 26, 1487–1490 .
Inamori, Y., Sakagami, Y., Morita, Y., Shibata, M., Sugiura, M., Kumeda, Y., Okabe, T., Tsujibo, H., Ishida, N. (2000) Antifungal activity of Hinokitiol-related compounds on wood-rotting fungi and their insecticidal activities Biol. Pharm. Bull., 23, 995–997 .
Morita, Y., Matsumura, E., Tsujibo, H., Yasuda, M., Okabe, T., Sakagami, Y., Kumeda, Y., Ishida, N., Inamor, Y. (2002) Biological activity of 4-acetyltropolone, the minor component of Thujopsis dolabrata SIeb. et Zucc. hondai Mak Biol Pharm Bull, 25, 981–985 .
Johnston, W.H., Karchesy, J.J., Constantine, G.H., Craig, A.M. (2001) Antimicrobial activity of some Pacific Northwest woods against anaerobic bacteria and yeast Phytother. Res., 15, 586–588 .
Miyamoto, D., Kusagaya, Y., Endo, N., Sometani, A., Takeo, S., Suzuki, T., Arima, Y., Nakajima, K., Suzuki, Y. (1998) Thujaplicin-copper chelates inhibit replication of human influenza viruses Antiviral Res., 39, 89–100 .
Wakabayashi, H., Yokoyama, K., Hashiba, K., Hashimoto, K., Kikuchi, H., Nishikawa, H., Kurihara, T., Satoh, K., Shioda, S., Muto, S., et al. (2003) Cytotoxic activity of tropolones against human oral tumor cell lines Anticancer Res., 23, 4757–4763 .
Miller, J.T., Rausch, J.W., Le Grice, S.F.J. (2001) Evaluation of retroviral ribonuclease H activity Methods Mol. Biol., 160, 335–354 .
Ma, W.P., Hamilton, S.E., Stowell, J.G., Byrn, S.R., Davisson, V.J. (1994) Sequence specific cleavage of messenger RNA by a modified ribonuclease H Bioorg. Med. Chem., 2, 169–179 .
Pileur, F., Andreola, M.L., Dausse, E., Michel, J., Moreau, S., Yamada, H., Gaidamakov, S.A., Crouch, R.J., Toulme, J.J., Cazenave, C. (2003) Selective inhibitory DNA aptamers of the human RNase H1 Nucleic Acids Res., 31, 5776–5788 .
Kvaratskhelia, M., Budihas, S.R., Le Grice, S.F. (2002) Pre-existing distortions in nucleic acid structure aid polypurine tract selection by HIV-1 reverse transcriptase J. Biol. Chem., 277, 16689–16696 .
Dash, C., Rausch, J.W., Le Grice, S.F. (2004) Using pyrrolo-deoxycytosine to probe RNA/DNA hybrids containing the human immunodeficiency virus type-1 3' polypurine tract Nucleic Acids Res., 32, 1539–1547 .
Yonetani, T. (1982) The Yonetani–Theorell graphical method for examining overlapping subsites of enzyme active centers Methods Enzymol., 87, 500–509 .
Cook, G.A., Mynatt, R.L., Kashfi, K. (1994) Yonetani–Theorell analysis of hepatic carnitine palmitoyltransferase-I inhibition indicates two distinct inhibitory binding sites J. Biol. Chem., 269, 8803–8807 .
Palaniappan, C., Fay, P.J., Bambara, R.A. (1995) Nevirapine alters the cleavage specificity of ribonuclease H of human immunodeficiency virus 1 reverse transcriptase J. Biol. Chem., 270, 4861–4869 .
Fan, N., Rank, K.B., Poppe, S.M., Tarpley, W.G., Sharma, S.K. (1996) Characterization of the p68/p58 heterodimer of human immunodeficiency virus type 2 reverse transcriptase Biochemistry, 35, 1911–1917 .
Endo, M., Mizutani, T., Matsumura, M., Moriyasu, M., Ichimaru, M., Kato, A., Hashimoto, Y. (1988) High-performance liquid chromatographic determination of hinokitiol in cosmetics by the formation of difluoroborane compounds J. Chromatogr., 455, 430–433 .
Gorshkova, II, Rausch, J.W., Le Grice, S.F., Crouch, R.J. (2001) HIV-1 reverse transcriptase interaction with model RNA-DNA duplexes Anal. Biochem., 291, 198–206 .
Chou, T.C. and Talalay, P. (1981) Generalized equations for the analysis of inhibitions of Michaelis–Menten and higher-order kinetic systems with two or more mutually exclusive and nonexclusive inhibitors Eur. J. Biochem., 115, 207–216 .
Chou, T.C. and Talalay, P. (1984) Quantitative analysis of dose-effect relationships: the combined effects of multiple drugs or enzyme inhibitors Adv. Enzyme Regul., 22, 27–55 .
Davies, J.F., II, Hostomska, Z., Hostomsky, Z., Jordan, S.R., Matthews, D.A. (1991) Crystal structure of the ribonuclease H domain of HIV-1 reverse transcriptase Science, 252, 88–95 .
Cowan, J.A., Ohyama, T., Howard, K., Rausch, J.W., Cowan, S.M., Le Grice, S.F. (2000) Metal-ion stoichiometry of the HIV-1 RT ribonuclease H domain: evidence for two mutually exclusive sites leads to new mechanistic insights on metal-mediated hydrolysis in nucleic acid biochemistry J. Biol. Inorg. Chem., 5, 67–74 .(Scott R. Budihas, Inna Gorshkova1, Serge)
*To whom correspondence should be addressed. Tel: +1 301 846 5256; Fax: +1 301 846 6013; Email: slegrice@ncifcrf.gov
ABSTRACT
High-throughput screening of a National Cancer Institute library of pure natural products identified the hydroxylated tropolone derivatives ?-thujaplicinol (2,7-dihydroxy-4-1(methylethyl)-2,4,6-cycloheptatrien-1-one) and manicol (1,2,3,4-tetrahydro-5-7-dihydroxy-9-methyl-2-(1-methylethenyl)-6H-benzocyclohepten-6-one) as potent and selective inhibitors of the ribonuclease H (RNase H) activity of human immunodeficiency virus-type 1 reverse transcriptase (HIV-1 RT). ?-Thujaplicinol inhibited HIV-1 RNase H in vitro with an IC50 of 0.2 μM, while the IC50 for Escherichia coli and human RNases H was 50 μM and 5.7 μM, respectively. In contrast, the related tropolone analog ?-thujaplicin (2-hydroxy-4-(methylethyl)-2,4,6-cycloheptatrien-1-one), which lacks the 7-OH group of the heptatriene ring, was inactive, while manicol, which possesses a 7-OH group, inhibited HIV-1 and E.coli RNases H with IC50 = 1.5 μM and 40 μM, respectively. Such a result highlights the importance of the 2,7-dihydroxy function of these tropolone analogs, possibly through a role in metal chelation at the RNase H active site. Inhibition of HIV-2 RT-associated RNase H indirectly indicates that these compounds do not occupy the nonnucleoside inhibitor-binding pocket in the vicinity of the DNA polymerase domain. Both ?-thujaplicinol and manicol failed to inhibit DNA-dependent DNA polymerase activity of HIV-1 RT at a concentration of 50 μM, suggesting that they are specific for the C-terminal RNase H domain, while surface plasmon resonance studies indicated that the inhibition was not due to intercalation of the analog into the nucleic acid substrate. Finally, we have demonstrated synergy between ?-thujaplicinol and calanolide A, a nonnucleoside inhibitor of HIV-1 RT, raising the possibility that both enzymatic activities of HIV-1 RT can be simultaneously targeted.
INTRODUCTION
Reverse transcriptase (RT)-associated ribonuclease H (RNase H) activity is responsible for both non-specifically degrading the RNA strand of the RNA/DNA replication intermediate as well as specifically removing the minus (–) and plus (+) strand RNA primers from nascent DNA (1). The absolute requirement for RNase H activity for human immunodeficiency virus (HIV) replication (2,3) suggests that this might be an attractive target for the development of antiviral agents to complement DNA polymerase-based HIV-1 RT inhibitors currently in clinical use . In this respect, recent reports have documented several promising candidates effective at low micromolar concentrations, including hydrazones (5–7), tetragalloylglucopyranose (8), diketo acids (9) and N-hydroxyimides (10). Although it remains to be established that their mode of inhibition is through direct binding to the RNase H catalytic center, both diketo acids and N-hydroxyimides have been shown to inhibit an enzymatically active peptide derived from the RNase H domain of HIV-1 RT (9,11). Thus, while antiviral activity of these select RNase H antagonists is yet to be demonstrated, sufficient evidence has accumulated to justify further screening for inhibitors of HIV-1 and HIV-2 RNase H. Moreover, although Klumpp et al. (10) have recently evaluated N-hydroxyimides for the inhibition of retroviral and bacterial RNases H, none of the more potent HIV-1 RNase H inhibitors reported to date has addressed selectivity with respect to the human counterpart. This issue is of particular importance, since recent data have indicated that disruption of the RNase H1 allele in mice confers a lethal embryonic defect (12). To take this issue into account, we recently described two simple, robust fluorescence-based methodologies which could be applied in parallel to retroviral, bacterial and human RNases H using the same assay platform, together with their use for high-throughput screening (13,14). The availability of purified human RNase H1 has also allowed us to investigate the specificity of inhibitors uncovered from our current screening efforts.
Here, we describe two tropolone (2-hydroxy-2,4,6-cycloheptatrien-1-one) derivatives with a 7-OH substitution, identified from high-throughput screening of National Cancer Institute libraries of pure natural products, which inhibit retroviral, bacterial and human RNases H. Inhibition of HIV-2 RNase H activity indirectly indicates that these compounds do not interact with the nonnucleoside-binding site located near the DNA polymerase catalytic center of this enzyme. The most potent of the hydroxylated tropolone analogs, ?-thujaplicinol, derived from the heartwood of several cupressaceous plants (e.g. Thuja plicata, Thuja occidentalis and Chamaecyparis obtusaI) (15–17), is active at sub-micromolar concentrations against the RNases H of HIV-1 and HIV-2 RT, while the DNA polymerase activity of either enzyme is unaffected at an inhibitor concentration of 50 μM. The same compound is 30- to 250-fold less active against the human and Escherichia coli RNases H, respectively, demonstrating that selective inhibition of the retroviral enzyme can be achieved. Finally, we demonstrate here that ?-thujaplicinol acts synergistically with calanolide A, a nonnucleoside inhibitor of HIV-1 RT (18,19), opening the possibility of simultaneously targeting the DNA polymerase and RNase H functions of HIV-1 and HIV-2 RT.
A number of reports have demonstrated that tropolone derivatives elicit a variety of biological effects, including anti-tumor (20), insecticidal (21), antifungal (22,23) and antimicrobial (24) activity, while their metal chelates have been shown to inhibit human influenza virus-induced apoptosis (25). Wakabayashi et al. (26) have investigated cytotoxic activity of tropolone derivatives against a variety of human oral tumor cell lines, and also indicated that they fail to protect MT-4 cells from HIV-1 infection.
MATERIALS AND METHODS
Tropolone derivatives
The structures of -thujaplicin , ?-thujaplicin , -thujaplicin , ?-thujaplicinol , manicol and nootkatin are provided in Figure 1. All compounds were members of a National Cancer Institute library of purified, natural compounds, and their identity and purity were confirmed by 1H-NMR. Each compound was prepared as a 20 mM stock solution in 100% dimethyl sulfoxide (DMSO) and diluted into the appropriate reaction buffer immediately prior to analysis.
Figure 1 Structures of tropolone and its derivatives.
Enzymes and enzyme assays
Recombinant, wild-type p66/p51 HIV-1 RT was expressed and purified as described previously (27). HIV-2 RT was prepared using the same E.coli expression system (27). E.coli RNase HI and recombinant human RNase H were prepared as described previously (28,29). The strategy for high-throughput screening and confirmation of RNase H activity by capillary electrophoresis has recently been described by Parniak et al. (13) and Chan et al. (14), respectively. For HIV-1 RT, conditions for cleavage of the HIV-1 PPT RNA primer extended at its 3' terminus by five deoxynucleotides have been described by Kvaratskhelia et al. (30). DNA-dependent DNA polymerase activity was analyzed on a 71 nt DNA to which a labeled 36 nt DNA primer was hybridized as described by Dash et al. (31). For qualitative analysis of PPT cleavage and DNA-dependent DNA polymerase activity, all tropolone derivatives were evaluated at a single concentration of 50 μM.
Surface plasmon resonance analysis
All experiments were performed with a Biacore 2000 optical biosensor at 25°C. CM5 Sensor chips with a carboxymethylated dextran matrix, EDC, NHS and P20 surfactant were obtained from Biacore AB, (Uppsala, Sweden), and streptavidin was from Pierce (Rockford, IL). Streptavidin was immobilized on the sensor chip surface to the level of 4000–4500 RU in each flow cell. Biotinylated nucleic acid (12mer RNA–DNA hybrid in the form of a hairpin within which a thymidine residue of a connection tetraloop was biotinylated) was injected at a concentration of 50 nM in HBS buffer (10 mM HEPES, 150 mM NaCl, 1 mM EDTA, 0.005% P20, pH 7.4) onto streptavidin-modified chip surface at a flow rate of 5 μl/min until the desired level (1000–1500 RU) of immobilization was achieved. Blocking of unoccupied streptavidin with biotin and washing the chip with regeneration buffer (2 M NaCl) was performed as described previously (28). Experiments on inhibitor or marker binding to nucleic acid were carried out in Tris-buffered saline/DMSO running buffer (10 mM Tris–HCl, 50 mM NaCl, 1 mM EDTA, 0.005% P20, 5% DMSO, pH 8.0) using a flow rate of 50 μl/min. Typically, 50 μl of a marker or inhibitor (each separately or in mixtures of 2 or 3 simultaneously) diluted in the same buffer in the concentration range of 50 nM to 10 μM was successively injected using the ‘kinject’ command defining a dissociation time of 2 min. After each round of injection, the system was intensively washed using running buffer. All experiments were performed in duplicate. Data transformation of the primary sensograms and overlay plots were prepared using the BIAevaluation 3.0 software (Biacore AB). The response from the reference flow cell that does not contain nucleic acid was subtracted from that of the experimental flow cell.
IC50 and Ki determinations
Dose–response plots were prepared by measuring RNase H activity on a fluorescent RNA/DNA hybrid as described previously (13) while varying the inhibitor concentration by 2-fold dilutions from 100 μM to 0.19 nM. Curve fitting was performed using Sigma Plot. In order to determine the Ki for ?-thujaplicinol, the substrate concentration was varied from 200 to 1.56 nM while varying the inhibitor concentration from 1.56 to 0.024 nM at each substrate concentration. Hydrolysis was followed by fluorescence measurements at 40 s intervals over the 30 min reaction time using a Molecular Devices Spectramax Gemini EM fluorescence spectrometer. Slopes of the reaction curves were determined from the linear portion of the plots using SoftMAX Pro version 4.3 supplied with the instrument. Curve fitting and prediction of the mode of inhibition was performed using the Enzyme Kinetics module of Sigma Plot version 8 (SPSS, Chicago, IL).
Synergy between RNase H and DNA polymerase inhibitors
In order to determine whether the binding site for ?-thujaplicinol overlapped that of nonnucleoside RT inhibitors, RNase H activity was measured in the presence of varying concentrations of tropolone inhibitor and varying concentrations of the nonnucleoside calanolide A (18) according to the method of Yonetani and Theorell (32) and Cook et al. (33). Data were processed using the Calcusyn software package (BioSoft, Cambridge, UK).
RESULTS
Tropolone derivatives specifically inhibit HIV-1 RNase H function
The structures of the tropolone derivatives used in the present study are provided in Figure 1. Since HIV-1 RT is a bifunctional enzyme, inhibition of RNase H activity might reflect direct binding to the RNase H domain or, alternatively, an allosteric effect through binding to the nonnucleoside-binding pocket close to the DNA polymerase domain, which has been shown previously to modulate RNase H function (34). Preliminary experiments were, therefore, designed to investigate the specificity of each analog for the RNase H activity of HIV-1 RT. In Figure 2A, RNase H activity was evaluated on a 5'labeled RNA–DNA chimera hybridized to its DNA complement, which mimics processing of the HIV-1 PPT primer from nascent DNA, following initiation of second or (+) strand synthesis. Using this substrate, we have previously demonstrated RNase H cleavage at the PPT RNA-U3 DNA junction (defined as –1), and additionally at positions –2 and –3 (30). For these initial experiments, a single analog concentration of 50 μM was employed, which resulted in almost complete inhibition of PPT processing by both ?-thujaplicinol (lane 4) and manicol (lane 7). In Figure 2B, the same derivatives were re-evaluated for their ability to inhibit DNA-dependent DNA polymerase activity on a duplex containing a 5'labeled 36 nt DNA primer hybridized to a 71 nt DNA template. At a single concentration of 50 μM, negligible inhibition of DNA polymerase activity was observed for all analogs. The combined data of Figure 2 thus indicate that inhibition of HIV-1 RNase H activity is not an consequence of analog binding to the DNA polymerase catalytic center and exerting an indirect effect, as has been reported for certain diketo acids and phenylhydrazones by Shaw-Reid et al. (9).
Figure 2 (A) Inhibition of HIV-1 RT-catalyzed polypurine tract processing by tropolone and its derivatives. Left, schematic representation of the RNase H assay. Right, RNase H activity in the presence of 50 μM tropolone derivative. Lane C, no inhibitor; lane 1, no inhibitor, +DMSO; lane 2, ?-thujaplicin; lane 3, -thujaplicin; lane 4, ?-thujaplicininol; lane 5, -thujaplicin; lane 6, nootkatin; lane 7, manicol; and lane 8, tropolone. PPT hydrolysis products have been indicated. (B) DNA-dependent DNA polymerase activity of HIV-1 RT in the presence of tropolone and hydroxylated derivatives. DNA polymerase activity was determined at a single inhibitor concentration of 50 μM. Lane notations are as in (A).
Selectivity of inhibition
Clearly, a desirable feature of an inhibitor of retroviral RT-associated RNase H activity would be selectivity with respect to its human counterpart. To address this issue, ?-thujaplicinol and manicol were examined for their inhibitory effect on the RTs of HIV-1 and HIV-2 as well as human RNase H. Figure 3A indicates that ?-thujaplicinol inhibited HIV-1 RT with an IC50 of 0.21 ± 0.03 μM. For HIV-2 RT, the IC50 for the same analog was 3-fold higher (Figure 3C, 0.77 ± 0.08 μM). Such a result might reflect subtle differences in the architecture of RNase H domains of these two enzymes, which has been suggested from biochemical analysis of the purified RTs (35). More importantly, ?-thujaplicinol inhibited human RNase H with an IC50 of 5.7 ± 0.7 μM (Figure 3E), indicating 30-fold selectivity for the HIV-1 enzyme. Although not shown here, an IC50 value of 50 μM was determined for E.coli RNase H, indicating that this enzyme was 250-fold less sensitive to ?-thujaplicinol inhibition. In Figure 3B, D and F, inhibition of retroviral and human RNases H by manicol was compared. While this analog was slightly less potent against HIV-1 RNase H (IC50 = 0.60 ± 0.09 μM), 6-fold enhanced selectivity over the human enzyme was achieved (IC50 = 3.5 ± 0.1 μM).
Figure 3 Selectivity of RNase H inhibition. Dose–response curves for RNase I inhibition by ?-thujaplicinol (A, C and E) and manicol (B, D and F) are presented. (A and B) HIV-1 RT; (C and D), HIV-2 RT; (E and F), human RNase H. IC50 determinations are the results of triplicate assays.
IC50 values for tropolone and its derivatives are presented in Table 1. Interestingly, ?-thujaplicin, which differs from ?-thujaplicinol in that it lacks the hydroxyl function at position 7 of the heptatriene ring, was completely inactive against all enzymes tested, despite reports that it possesses metal chelating properties (36). Relocation of the hydroxyl function on the heptatriene ring produced a different effect, i.e. while -thujaplicin failed to inhibit retroviral RNases H, -thujaplicin was weakly active, with an IC50 value of 50 and 33 μM for the HIV-1 and HIV-2 enzymes, respectively.
Table 1 Inhibition of retroviral, bacterial and human RNases H by hydroxylated tropolone derivatives
?-Thujaplicinol does not inhibit RNase H activity through intercalation
Although DNA polymerase activity was unaffected, one possible mode of inhibition we had to consider was the intercalation of the planar tropolone derivatives into the RNA/DNA hybrid, effectively making the substrate unavailable for enzyme binding. In order to eliminate this possibility, inhibitor binding to an immobilized RNA/DNA hybrid was investigated via surface plasmon resonance, which we have successfully used to investigate the binding properties of HIV-1 RT derivatives (37). As a control, binding of ethidium bromide to the immobilized hybrid was analyzed, the results of which are depicted in the sensorgram of Figure 4A. From this data, we calculated that 4–6 molecules of ethidium bromide intercalated into the hybrid with a Kd of 1.5 μM. In contrast, data of Figure 4B, which presents the combined sensorgrams obtained when the RNA/DNA hybrid was incubated with -thujaplicin, ?-thujaplicin and ?-thujaplicinol, clearly indicate that they do not bind the substrate over a concentration range of 50 nM–10 mM.
Figure 4 Analysis of inhibitor intercalation by surface plasmon resonance. In (A), binding of ethidium bromide to an immobilized 12mer RNA/DNA hybrid was performed as a control. Sensorgrams a–f illustrate the response to successive injection of the intercalator at concentrations of 10, 5, 1.0, 0.5, 0.25 and 0.1 μM, respectively. (B) Combined sensorgrams following injection of ?-thujaplicin, -thujaplicin and ?-thujaplicinol over the same concentration range as in (A).
Mode of inhibition
?-Thujaplicin and manicol were further investigated in order to gain insight into their mechanism of inhibition of HIV-1 RT/RNase H. From the Michaelis–Menten plot of Figure 5, a Ki of 0.20 ± 0.07 μM was determined for ?-thujaplicinol, and a separate study indicated a Ki of 1.0 ± 0.4 μM for manicol (data not shown). The data presented in Figure 5 show a change in Vmax without affecting the Km, which is suggestive of mixed inhibition. Non-linear fitting of the data confirms that this is best fit to mixed-type inhibition compared with competitive or uncompetitive modes. With this type of inhibition, the antagonist can bind either the free enzyme or the enzyme–substrate complex. This suggests that the tropolone derivatives could occupy (or create) a pocket near the RNase H active site, since access to that site would be impaired by the RNA/DNA duplex substrate.
Figure 5 Michaelis–Menten plot for the determination of the Ki for ?-thujaplicinol for HIV-1 RT-associated RNase H. A Ki of 0.20 ± 0.07 μM was determined from non-linear curve fitting. Inhibitor concentrations were 0.024 μM (open circle), 0.049 μM (filled triangle), 0.098 μM (open triangle), 0.20 μM (filled square), 0.39 μM (open square), 0.78 μM (filled diamond), 1.56 μM (open diamond), RNase H activity in the presence of 5% DMSO (filled circle).
?-Thujaplicinol and the nonnucleoside calanolide A occupy different sites on HIV-1 RT
The multifunctional nature of HIV-1 RT opens the possibility that antiviral agents could be separately targeted to the DNA polymerase and RNase H catalytic centers. In order to assess this, we investigated whether ?-thujaplicinol would act in combination with calanolide A (18). This coumarin derivative, isolated from the tropical rainforest tree Calophyllum lanigerum, has been shown to inhibit HIV-1 RT activity by a mechanism similar to other nonnucleoside-based drugs (19). Figure 6 presents a Yonetani–Theorell plot of RNase H inhibition in the presence of both calanolide A and ?-thujaplicinol. The data indicate that the inhibitors bind to two independent sites since the plot shows converging lines (32). In addition, HIV-2 is naturally resistant to NNRTIs; thus, inhibition of HIV-2 RNase H is additional, albeit indirect, evidence that ?-thujaplicinol binds to an independent site. Independent binding sites for calanolide A and ?-thujaplicinol were further supported by inhibitor-combination assays using the methods described by Chou and Talalay (38,39). The combination index (CI) derived for mixed inhibitors can be used to describe additivity (CI = 1), synergy (CI < 1) and antagonism (CI > 1). The inhibitor combination of ?-thujaplicinol/calanolide A was found to be 0.50, i.e. the inhibition by a ?-thujaplicinol/calanolide A combination is more potent than predicted by adding the effect of the individual inhibitors. This synergistic effect would not be observed with inhibitors competing for the same or closely positioned binding sites. In a control experiment, a CI of 1.8 was determined for a ?-thujaplicinol/manicol combination, indicating antagonism. This would be predicted, since the two tropolone derivatives most likely compete for the same binding site.
Figure 6 Yonetani–Theorell plot for the inhibition of HIV-1 RT in the presence of the NNRTI calanolide A and ?-thujaplicinol. The inverse of the rate of RNase H cleavage was plotted as a function of ?-thujaplicinol concentration at calanolide A concentrations of 12.5 μM (open square), 0.78 μM (filled square), 0.39 μM (open circle) and DMSO (filled circle). The convergent best-fit lines indicate mutually exclusive binding sites for calanolide A and ?-thujaplicinol.
DISCUSSION
In this communication, we demonstrate that the natural products ?-thujaplicinol and manicol, derived from the heartwood of several cupressaceous trees, inhibit the RNase H activity of HIV-1 and HIV-2 RT at sub-micromolar concentrations without affecting DNA polymerase function of either enzyme. Moreover, the most potent inhibitor, ?-thujaplicinol, is 30-fold more active on HIV-1 RT/RNase H than on human RNase H1. In light of the requirement for RNase H1 during embryonal development in mice (12), selectivity of the retroviral enzyme over the human counterpart should be a major goal during antiviral drug development, but to date had not been addressed. Finally, our data indicate that ?-thujaplicinol and the nonnucleoside inhibitor calanolide A (18,19) occupy different sites on HIV-1 RT, raising the possibility that the DNA polymerase and RNase H domains of HIV-1 RT might be simultaneously targeted by antiviral agents.
Although the mechanistic basis for retroviral RNase H inhibition observed here remains to be established, the data of Figure 1 and Table 1 provide some insights. Since the ?-diketone moiety of thujaplicins has been demonstrated capable of chelating metals, such as Cu++ and Zn++ (36), chelating and/or altering the co-ordination geometry of the divalent metal in the RNase H catalytic center essential for catalysis would seem a plausible mechanism of action. However, we observe here that neither ?-thujaplicin nor -thujaplicin is active against the retroviral enzymes, while -thujaplicin is only modestly active (Table 1). ?-thujaplicinol introduces a second hydroxyl function at position 7 of the heptatriene ring, which substantially improves potency. At the same time, although manicol introduces a bulky substituent onto the heptatriene ring, it provides almost the same potency against HIV-1 RNase H as ?-thujaplicinol. While addition of the 7-OH function enhances the specificity for HIV-1 and HIV-2 RT/RNase H, it has little to no effect on potency against the bacterial and human enzymes. Poor inhibition of E.coli RNase H and the consistent but modest inhibition of human RNase H by all tropolones tested suggests that if metal chelation is responsible for inhibition, the 7-OH function is critical for stabilizing an interaction, which is specific to the active-site geometry of the retroviral enzymes. Interestingly, a two-metal catalyzed catalytic mechanism has been proposed for HIV-1 RNase H, based on crystallographic data with the isolated Mn++-doped domain showing two divalent cations at the active site (40). Although speculative, the potency we observe for ?-thujaplicinol may reflect its ability to form a more stable complex with the divalent cation at both metal-binding sites. If an interaction with the divalent cation at the RNase H catalytic center is the basis for inhibition by ?-thujaplicinol and manicol, it is interesting to note that these two compounds do not inhibit DNA polymerase function, despite the fact that the equivalent divalent cation is coordinated in this catalytic center by a triad of carboxyate residues (Asp110, Asp185 and Asp186). One possible explanation might be that the divalent cation is more tightly bound at the DNA polymerase catalytic center and thus more difficult to access. However, titration calorimetry studies suggest the contrary, namely that the divalent cation in the RNase H catalytic center is more tightly coordinated (41). Based on these observations, differences in the overall architecture of the metal-binding site might exclude these inhibitors from the polymerase catalytic center. Further studies of the interaction of ?-thujaplicinol and manicol with HIV-1 and HIV-2 RT have the potential to clarify their mechanism of action. While inactive in cellular models of HIV-1 infection, these two compounds nonetheless will serve as useful probes of HIV-1 and HIV-2 RNase H enzymology. For example, recent studies have suggested that inhibition of HIV-1 RNase H activity might increase the rate at which chain terminators of DNA synthesis are excised, thus leading to enhanced nucleoside resistance (V. Pathak, personal communication). The availability of inhibitors selective for the RNase H activity of HIV-1 RT will allow this hypothesis to be directly evaluated in vitro.
ACKNOWLEDGEMENTS
We would like to thank the Drug Synthesis and Chemistry Branch of the National Cancer Institute for the pure natural product library and for providing the tropolone derivatives. Funding to pay the Open Access publication charges for this article was provided by the National Cancer Institute, National Institutes of Health, Department of Health and Human Services.
REFERENCES
Klarmann, G.J., Hawkins, M.E., Le Grice, S.F. (2002) Uncovering the complexities of retroviral ribonuclease H reveals its potential as a therapeutic target AIDS Rev., 4, 183–194 .
Schatz, O., Cromme, F., Naas, T., Lindemann, D., Gruninger-Leitch, F., Mous, J., Le Grice, S.F.J. (1990) Inactivation of the RNaseH domain of HIV-1 reverse transcriptase blocks viral infectivity In Papas, T. (Ed.). Oncogenesis and AIDS, Houston, TX Portfolio Publishing Company pp. 55–68 .
Tisdale, M., Schulze, T., Larder, B.A., Moelling, K. (1991) Mutations within the RNase H domain of human immunodeficiency virus type 1 reverse transcriptase abolish virus infectivity J. Gen. Virol., 72, 59–66 .
Parniak, M.A. and Sluis-Cremer, N. (2000) Inhibitors of HIV-1 reverse transcriptase Adv. Pharmacol., 49, 67–109 .
Borkow, G., Fletcher, R.S., Barnard, J., Arion, D., Motakis, D., Dmitrienko, G.I., Parniak, M.A. (1997) Inhibition of the ribonuclease H and DNA polymerase activities of HIV-1 reverse transcriptase by N-(4-tert-butylbenzoyl)-2-hydroxy-1-naphthaldehyde hydrazone Biochemistry, 36, 3179–3185 .
Gabbara, S., Davis, W.R., Hupe, L., Hupe, D., Peliska, J.A. (1999) Inhibitors of DNA strand transfer reactions catalyzed by HIV-1 reverse transcriptase Biochemistry, 38, 13070–13076 .
Davis, W.R., Tomsho, J., Nikam, S., Cook, E.M., Somand, D., Peliska, J.A. (2000) Inhibition of HIV-1 reverse transcriptase-catalyzed DNA strand transfer reactions by 4-chlorophenylhydrazone of mesoxalic acid Biochemistry, 39, 14279–14291 .
Min, B.S., Nakamura, N., Miyashiro, H., Kim, Y.H., Hattori, M. (2000) Inhibition of human immunodeficiency virus type 1 reverse transcriptase and ribonuclease H activities by constituents of Juglans mandshurica Chem. Pharm. Bull. (Tokyo), 48, 194–200 .
Shaw-Reid, C.A., Munshi, V., Graham, P., Wolfe, A., Witmer, M., Danzeisen, R., Olsen, D.B., Carroll, S.S., Embrey, M., Wai, J.S., et al. (2003) Inhibition of HIV-1 ribonuclease H by a novel diketo acid, 4--2,4-dioxobutanoic acid J. Biol. Chem., 278, 2777–2780 .
Klumpp, K., Hang, J.Q., Rajendran, S., Yang, Y., Derosier, A., Wong Kai In, P., Overton, H., Parkes, K.E., Cammack, N., Martin, J.A. (2003) Two-metal ion mechanism of RNA cleavage by HIV RNase H and mechanism-based design of selective HIV RNase H inhibitors Nucleic Acids Res., 31, 6852–6859 .
Hannoush, R.N., Carriero, S., Min, K.L., Damha, M.J. (2004) Selective inhibition of HIV-1 reverse transcriptase (HIV-1 RT) RNase H by small RNA hairpins and dumbbells Chembiochem, 5, 527–533 .
Cerritelli, S.M., Frolova, E.G., Feng, C., Grinberg, A., Love, P.E., Crouch, R.J. (2003) Failure to produce mitochondrial DNA results in embryonic lethality in Rnaseh1 null mice Mol. Cell, 11, 807–815 .
Parniak, M.A., Min, K.L., Budihas, S.R., Le Grice, S.F., Beutler, J.A. (2003) A fluorescence-based high-throughput screening assay for inhibitors of human immunodeficiency virus-1 reverse transcriptase-associated ribonuclease H activity Anal. Biochem., 322, 33–39 .
Chan, K.C., Budihas, S.R., Le Grice, S.F., Parniak, M.A., Crouch, R.J., Gaidamakov, S.A., Isaaq, H.J., Wamiru, A., McMahon, J.B., Beutler, J.A. (2004) A capillary electrophoretic assay for ribonuclease H activity Anal. Biochem., 331, 296–302 .
Gardner, J.A.F., Barton, G.M., MacLean, H. (1957) Occurrence of 2,7-dihydroxy-4-isopropyl-2,4,6-cycloheptatrien-1-one (7-hydroxy-4-isopropyltropolone) in western red cedar (Thuja plicata Donn.) Can J. Chem., 35, 1039–1048 .
Zavarin, E., Smith, L.V., Bicho, J.G. (1967) Tropolones of Cupressaceae—III Phytochemistry, 6, 1387–1394 .
Polonsky, J., Beloeil, J.-C., Prange, T., Pascard, C., Lacquemin, H., Donnelly, D.M.X., Kenny, P. (1983) Manicol: a sesquiterpenoid hydroxytropolone from Dulacia guianensis. A revised structure (X-ray analysis) Tetrahedron, 39, 2647–2655 .
Kashman, Y., Gustafson, K.R., Fuller, R.W., Cardellina, J.H., II, McMahon, J.B., Currens, M.J., Buckheit, R.W., Jr, Hughes, S.H., Cragg, G.M., Boyd, M.R. (1992) The calanolides, a novel HIV-inhibitory class of coumarin derivatives from the tropical rainforest tree, Calophyllum lanigerum J. Med. Chem., 35, 2735–2743 .
Boyer, P.L., Currens, M.J., McMahon, J.B., Boyd, M.R., Hughes, S.H. (1993) Analysis of nonnucleoside drug-resistant variants of human immunodeficiency virus type 1 reverse transcriptase J. Virol., 67, 2412–2420 .
Sugawara, K., Ohbayashi, M., Shimizu, K., Hatori, M., Kamei, H., Konishi, M., Oki, T., Kawaguchi, H. (1988) BMY-28438 (3,7-dihydroxytropolone), a new antitumor antibiotic active against B16 melanoma. I. Production, isolation, structure and biological activity J. Antibiot. (Tokyo), 41, 862–868 .
Morita, Y., Matsumura, E., Okabe, T., Shibata, M., Sugiura, M., Ohe, T., Tsujibo, H., Ishida, N., Inamori, Y. (2003) Biological activity of tropolone Biol. Pharm. Bull., 26, 1487–1490 .
Inamori, Y., Sakagami, Y., Morita, Y., Shibata, M., Sugiura, M., Kumeda, Y., Okabe, T., Tsujibo, H., Ishida, N. (2000) Antifungal activity of Hinokitiol-related compounds on wood-rotting fungi and their insecticidal activities Biol. Pharm. Bull., 23, 995–997 .
Morita, Y., Matsumura, E., Tsujibo, H., Yasuda, M., Okabe, T., Sakagami, Y., Kumeda, Y., Ishida, N., Inamor, Y. (2002) Biological activity of 4-acetyltropolone, the minor component of Thujopsis dolabrata SIeb. et Zucc. hondai Mak Biol Pharm Bull, 25, 981–985 .
Johnston, W.H., Karchesy, J.J., Constantine, G.H., Craig, A.M. (2001) Antimicrobial activity of some Pacific Northwest woods against anaerobic bacteria and yeast Phytother. Res., 15, 586–588 .
Miyamoto, D., Kusagaya, Y., Endo, N., Sometani, A., Takeo, S., Suzuki, T., Arima, Y., Nakajima, K., Suzuki, Y. (1998) Thujaplicin-copper chelates inhibit replication of human influenza viruses Antiviral Res., 39, 89–100 .
Wakabayashi, H., Yokoyama, K., Hashiba, K., Hashimoto, K., Kikuchi, H., Nishikawa, H., Kurihara, T., Satoh, K., Shioda, S., Muto, S., et al. (2003) Cytotoxic activity of tropolones against human oral tumor cell lines Anticancer Res., 23, 4757–4763 .
Miller, J.T., Rausch, J.W., Le Grice, S.F.J. (2001) Evaluation of retroviral ribonuclease H activity Methods Mol. Biol., 160, 335–354 .
Ma, W.P., Hamilton, S.E., Stowell, J.G., Byrn, S.R., Davisson, V.J. (1994) Sequence specific cleavage of messenger RNA by a modified ribonuclease H Bioorg. Med. Chem., 2, 169–179 .
Pileur, F., Andreola, M.L., Dausse, E., Michel, J., Moreau, S., Yamada, H., Gaidamakov, S.A., Crouch, R.J., Toulme, J.J., Cazenave, C. (2003) Selective inhibitory DNA aptamers of the human RNase H1 Nucleic Acids Res., 31, 5776–5788 .
Kvaratskhelia, M., Budihas, S.R., Le Grice, S.F. (2002) Pre-existing distortions in nucleic acid structure aid polypurine tract selection by HIV-1 reverse transcriptase J. Biol. Chem., 277, 16689–16696 .
Dash, C., Rausch, J.W., Le Grice, S.F. (2004) Using pyrrolo-deoxycytosine to probe RNA/DNA hybrids containing the human immunodeficiency virus type-1 3' polypurine tract Nucleic Acids Res., 32, 1539–1547 .
Yonetani, T. (1982) The Yonetani–Theorell graphical method for examining overlapping subsites of enzyme active centers Methods Enzymol., 87, 500–509 .
Cook, G.A., Mynatt, R.L., Kashfi, K. (1994) Yonetani–Theorell analysis of hepatic carnitine palmitoyltransferase-I inhibition indicates two distinct inhibitory binding sites J. Biol. Chem., 269, 8803–8807 .
Palaniappan, C., Fay, P.J., Bambara, R.A. (1995) Nevirapine alters the cleavage specificity of ribonuclease H of human immunodeficiency virus 1 reverse transcriptase J. Biol. Chem., 270, 4861–4869 .
Fan, N., Rank, K.B., Poppe, S.M., Tarpley, W.G., Sharma, S.K. (1996) Characterization of the p68/p58 heterodimer of human immunodeficiency virus type 2 reverse transcriptase Biochemistry, 35, 1911–1917 .
Endo, M., Mizutani, T., Matsumura, M., Moriyasu, M., Ichimaru, M., Kato, A., Hashimoto, Y. (1988) High-performance liquid chromatographic determination of hinokitiol in cosmetics by the formation of difluoroborane compounds J. Chromatogr., 455, 430–433 .
Gorshkova, II, Rausch, J.W., Le Grice, S.F., Crouch, R.J. (2001) HIV-1 reverse transcriptase interaction with model RNA-DNA duplexes Anal. Biochem., 291, 198–206 .
Chou, T.C. and Talalay, P. (1981) Generalized equations for the analysis of inhibitions of Michaelis–Menten and higher-order kinetic systems with two or more mutually exclusive and nonexclusive inhibitors Eur. J. Biochem., 115, 207–216 .
Chou, T.C. and Talalay, P. (1984) Quantitative analysis of dose-effect relationships: the combined effects of multiple drugs or enzyme inhibitors Adv. Enzyme Regul., 22, 27–55 .
Davies, J.F., II, Hostomska, Z., Hostomsky, Z., Jordan, S.R., Matthews, D.A. (1991) Crystal structure of the ribonuclease H domain of HIV-1 reverse transcriptase Science, 252, 88–95 .
Cowan, J.A., Ohyama, T., Howard, K., Rausch, J.W., Cowan, S.M., Le Grice, S.F. (2000) Metal-ion stoichiometry of the HIV-1 RT ribonuclease H domain: evidence for two mutually exclusive sites leads to new mechanistic insights on metal-mediated hydrolysis in nucleic acid biochemistry J. Biol. Inorg. Chem., 5, 67–74 .(Scott R. Budihas, Inna Gorshkova1, Serge)