Interactions of a Novel Inhibitor from an ExtremophilicBacillussp. with HIV-1 Protease

The active site cleft of the HIV-1 protease (PR) is bound by two identical conformationally mobile loops known as flaps, which are important for substrate binding and catalysis. The present article reports, for the first time, an HIV-1 PR inhibitor, ATBI, from an extremophilic Bacillus sp. The inhibitor is found to be a hydrophilic peptide with Mr of 1147, and an amino acid sequence of Ala-Gly-Lys-Lys-Asp-Asp-Asp-Asp-Pro-Pro-Glu. Sequence homology exhibited no similarity with the reported peptidic inhibitors of HIV-1 PR. Investigation of the kinetics of the enzyme-inhibitor interactions revealed that ATBI is a noncompetitive and tight binding inhibitor with the IC50 and K i values 18.0 and 17.8 nm, respectively. The binding of the inhibitor with the enzyme and the subsequent induction of the localized conformational changes in the flap region of the HIV-1 PR were monitored by exploiting the intrinsic fluorescence of the surface exposed Trp-42 residues, which are present at the proximity of the flaps. We have demonstrated by fluorescence and circular dichroism studies that ATBI binds in the active site of the HIV-1 PR and thereby leads to the inactivation of the enzyme. Based on our results, we propose that the inactivation is due to the reorganization of the flaps impairing its flexibility leading toward inaccessibility of the substrate to the active site of the enzyme.

HIV-1 1 protease (PR) has been classified as an aspartic protease that functions as a homodimer, based on its primary amino acid sequence, its inhibition by pepstatin, and its crystal structure (1)(2)(3). The retroviral protease is encoded in the viral pro gene for all retroviruses, including HIV-1 (4,5). During the replication cycle of HIV, gag and gag-pol gene products are translated as polyproteins. These proteins are subsequently processed by the virally encoded protease to yield structural proteins of the virus core, together with essential viral enzymes including the protease itself. The active site of HIV-1 PR is composed of the carboxylate side chains of two Asp residues, one from each subunit of the dimer (6 -10). The aspartic pro-teases are a large family of enzymes with diverse functional roles that also share a number of structural features (11)(12). One such feature is called the flap, which lies above the active site cleft. By sequence alignment, the conserved sequence domain of the flap region begins at position 47 of the HIV-1 PR sequence and extends through the Gly at position 52. These residues form a short stretch of ␤-sheet followed by a turn that ends with the conserved Gly at position 52 (13). The crystallographic structures of HIV-1 PR-inhibitor complexes demonstrated that the binding of a peptide analogue inhibitor or a peptide substrate involves numerous hydrogen-bonding interactions with the highly mobile flaps (13). Movement of these flaps apparently accompanies the binding of the peptide analogue or substrate, which binds in an extended ␤-sheet, such that hydrogen bonds are established between the complementary carbonyl oxygens and the amide protons of the peptide within the flaps. The presumed function of the flap-peptide interactions is to entrap and align the scissile peptide sequence in the HIV-1 PR active site (14).
HIV-1 PR has been an attractive target for the development of drugs against AIDS (15). The rational design of HIV-1 PR inhibitors may be considered under two broad categories based on (a) the substrate specificity and (b) the structural homology of HIV-1 PR dimer (16). Plethoras of synthetic inhibitory compounds targeting the active site of the HIV-1 PR have been reported (17,18). However, a lacuna of literature on biomolecules from microorganisms still exists. The present study deals with the isolation of an inhibitor, ATBI, of HIV-1 PR from an extremophilic Bacillus sp. and the evaluation of its kinetic parameters. Fluorescence spectroscopic studies revealed that ATBI binds in the active site of the HIV-1 PR and is the first report of a noncompetitive inhibitor from an extremophilic microorganism. It is well established that the Trp-42 is present adjacent to the flaps, and the flap regions of HIV-1 PR are the only dynamically flexible portions of the enzyme (19). We have investigated the conformational changes induced in the flap regions of the HIV-1 PR by monitoring the intrinsic fluorescence of the Trp residues and the effects on the secondary structure of the HIV-1 PR, by circular dichroism studies, upon binding of ATBI. We have also compared the results obtained with that of the substrate and active site-directed inhibitors of the HIV-1 PR. These results demonstrated that the enzyme inactivation is caused by the loss of the flexibility of the flaps restricting the entry and exit of the polypeptide substrate and products.

EXPERIMENTAL PROCEDURES
Bacterial Strains and Growth Conditions-The extremophilic Bacillus sp. was grown on a liquid medium containing soyameal (2%) and other nutrients at 50°C for 48 h as described (20) (the medium was adjusted to pH 10 by the addition of sterile 10% sodium carbonate). The Escherichia coli strain harboring the recombinant plasmid containing * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
‡ Supported by the Council of Scientific and Industrial Research, Government of India.
Purification and Biochemical Characterization of ATBI-Extracellular culture filtrate (1000 ml) of the extremophilic Bacillus sp. was treated with activated charcoal (65 g) and incubated at 4°C overnight. The colorless filtrate was subjected to membrane filtration through amicon-UM10 (molecular weight cut-off 10,000) and subsequently through amicon-UM2 (molecular weight cut-off 2000). The resulting inhibitor sample was concentrated by lyophilization (50 ml). The residual concentrate was further purified by reverse phase high performance liquid chromatography (rp-HPLC). The concentrated inhibitor sample (100 l) was loaded onto a prepacked UltroPac column (Lichrosorb RP-18, LKB), which was preequilibrated with 10% acetonitrile (CH 3 CN) and 0.1% trifluoroacetate. The fractions were eluted on a linear gradient of 0 -50% CH 3 CN with H 2 O containing 0.01% trifluoroacetate at a flow rate of 0.5 ml/min and monitored at a wavelength of 210 nm. The eluate was evaporated and lyophilized. The residual matter was dissolved in distilled H 2 O and assayed for the anti-HIV-1 PR activity. The active fractions were rechromatographed on rp-HPLC under similar experimental conditions as described above. The active peak was finally purified by rp-HPLC using the Lichrosorb RP18 column.
The amino acid sequence of the purified peptide was analyzed with a protein sequencer (Applied Biosystems model 476A), and the sequence homology was done manually after retrieving the peptide sequences from the data bank. Molecular mass of the purified ATBI was determined on a VG Biotek Platform-II quadrupole electrospray mass spectrometer using CH 3 CN-H 2 O (1:1) as mobile phase. The isoelectric point of the inhibitor was determined as described (21).
Enzyme Purification, Assay, and Kinetic Analysis-The recombinant HIV-1 PR harbored in Escherichia coli was expressed by temperature induction after the onset of the log phase of bacterial growth and purified by ammonium sulfate precipitation, dialysis, and gel filtration chromatography as described (22). The HIV-1 PR activity was assayed using the synthetic substrate Lys-Ala-Arg-Val-Nle-p-nitro-Phe-Glu-Ala-Nle-amide (23)(24). The HIV-1 PR was incubated at 37°C at different concentrations of the substrate in a reaction mixture containing 100 mM NaCl, 5 mM ␤-mercaptoethanol, 5 mM EDTA, and 50 mM sodium acetate buffer, pH 5.6. After 15 min, the reaction was stopped by the addition of an equal volume of 5% trichloroacetate and followed by a 30-min incubation at 28°C. The cleavage products were analyzed by rp-HPLC and by decrease in optical absorbance at 300 nm. The inhibition constant K i was determined as described by Dixon (25) and by Lineweaver-Burk's equation. In Dixon's method, proteolytic activity of the recombinant HIV-1 PR was measured at two different concentrations of substrate as a function of inhibitor concentration. The kinetic constants were determined by incubating the HIV-1 PR in the absence and presence of ATBI with increasing concentrations of the substrate for 15 min at 37°C. The inhibition was analyzed by the double reciprocal plot. The kinetics of the HIV-1 PR inhibition was analyzed by a model for tight binding inhibition (26). Kinetic determinations of enzyme interaction with the inhibitor in the absence of substrate were determined at short intervals by assaying the residual protease activity. In these experiments, the residual enzymatic activity was measured after the HIV-1 PR and the inhibitor were mixed and the samples were subsampled at increasing time intervals and assayed with the substrate. In all of the experiments, the inhibition of the HIV-1 PR was too rapid to measure under first order conditions. Rates of the HIV-1 PR were therefore determined in all cases by second order association rate kinetics. The association rate constants were calculated according to the integrated second order rate equation (27), where [E] is the enzyme concentration, [I] is the inhibitor concentration, and [EI] is the concentration of enzyme-inhibitor complex. The residual values (free enzyme) at 60 s were subtracted from that of the total enzyme, and this gives the concentration of the enzyme-inhibitor complex. The dissociation rate constants (kЈ or ␣) were determined from the formula ␣ ϭ ␤[I] by plotting the slope of the rate of inhibition (␤) or association rate constant (kЉ) in each reaction versus time multiplied by the inhibitor concentration. The slopes of the values were fitted by linear regression. Substrate protection studies were carried out by incubating the HIV-1 PR with different concentrations of substrate and then assaying the proteolytic activity at increased concentration of the inhibitor.
Fluorescence and Circular Dichroism Analysis-Fluorescence measurements were performed on a PerkinElmer Life Sciences LS50 luminescence spectrometer connected to a Julabo F20 water bath. Protein fluorescence was excited at 295 nm, and the emission was recorded from 300 to 500 nm at 25°C. The slit widths on both the excitation and emission were set at 5 nm, and the spectra were obtained at 500 nm/min. Fluorescence data were corrected by running control samples of buffer and smoothened. For binding studies, the HIV-1 PR (25 g/ml) was dissolved in 50 mM sodium acetate buffer (pH 5.6) containing 100 mM NaCl, 5 mM EDTA, 5 mM ␤-mercaptoethanol. Titration of the enzyme with ATBI was performed by the addition of different concentrations of the inhibitor to the enzyme solution. For each inhibitor concentration on the titration curve, a new enzyme solution was used. All of the data on the titration curve were corrected for dilutions. Further, the emission spectra of the HIV-1 PR were recorded in the presence of the substrate and the active site-based inhibitors N-acetyl-Leu-Val-Phe-Al (where Al is aldehyde) and pepstatin at 25°C. Accessibility calculations and visualization of Trp residues were performed by Insight-II (28) from the crystallographic structure of the HIV-1 PR as described (19). 2 CD spectra were recorded in a Jasco-J715 spectropolarimeter at ambient temperature using a cell of 1-mm path length. Replicate scans were obtained at 0.1-nm resolution, 0.1-nm bandwidth, and a scan speed of 50 nm/min. Spectra were averages of six scans with the base line subtracted spanning from 280 to 200 nm in 0.1-nm increments. The CD spectrum of the HIV-1 PR (25 g/ml) was recorded in 50 mM sodium phosphate buffer (pH 5.6) containing 100 mM NaCl, 5 mM EDTA, 5 mM ␤-mercaptoethanol in the absence/presence of substrate (40 M) or ATBI (20 nM). Secondary structure content of the HIV-1 PR, the HIV-1 PR-substrate complex, and the HIV-1 PR-ATBI complex was calculated using the algorithm of the K2d program (30,31).

Purification and Biochemical Characterization of ATBI-
The extracellular culture filtrate of the extremophilic Bacillus sp. was subjected to activated charcoal treatment and ultrafiltration to remove the high molecular weight impurities. The concentrated inhibitor sample was further purified by rp-HPLC. The anti-HIV-1 PR activity was associated with the peak A (Fig. 1a), and other eluted peaks showed no inhibitory activity. Homogeneity of the active fractions was indicated by the single peak as analyzed on rp-HPLC (Fig. 1b). Further, the purified ATBI showed a single band on an analytical isoelectric focusing gel unit with a pI of 10.0. The amino acid sequence of the purified inhibitor determined by a protein sequencer was Ala-Gly-Lys-Lys-Asp-Asp-Asp-Asp-Pro-Pro-Glu and was distinctly different from the sequence of the other reported inhibitors of HIV-1 PR (32)(33)(34)(35). The predominance of the charged amino acid residues in the inhibitor sequence indicated its hydrophilic nature. The molecular mass of ATBI as determined from electrospray mass spectrometry was 1147 Da (Fig. 2a).
Kinetics of Inactivation of the Recombinant HIV-1 PR by ATBI-ATBI was found to inhibit the purified recombinant HIV-1 PR with an IC 50 value (50% inhibitory concentration) of 18 nM (Fig. 3a). The inhibition of the HIV-1 PR followed a sigmoidal pattern with increasing concentrations of the inhibitor. However, the secondary plot (the slope of inhibition graph versus inhibitor concentration) was not linear, suggesting that the application of Michaelis-Menten inhibition kinetics was not appropriate in this study. The inhibition constant K i , determined by the classical double reciprocal plot and also by Dixon plot was 17.8 nM (Fig. 3b), which is almost equal to the IC 50 value of the inhibitor. The Lineweaver-Burk's reciprocal plot (Fig. 3c) showed that ATBI was a noncompetitive inhibitor of the HIV-1 PR. For the inhibition kinetic studies, the HIV-1 PR activity was monitored in the presence of various concentrations of inhibitor and substrate as a function of time. A very rapid inhibition of the HIV-1 PR was observed, which necessitated measuring all of the kinetic parameters at second order association conditions. The ␣-values obtained for ATBI were reasonably constant (data not shown), and the average calculated value of ␣ in the presence of substrate is 8.25 Ϯ 0.50 ϫ 10 Ϫ4 s Ϫ1 . The association rate constants ␤ in the presence and absence of substrate were calculated from the plot of the HIV-1 PR inhibition versus time. The values of ␤ (data not shown) were not affected by the presence of the substrate, indicating that the presence of substrate had no implication on the interaction of the inhibitor and enzyme. The reciprocal of the values of ␤ were plotted as a function of the substrate concentration (Fig. 4a). The plotted line did not fit a linear plot but was a good fit for a rectangular hyperbola, revealing noncompetitive inhibition. The mechanism of inhibition of the HIV-1 PR was further deciphered by the plot of ⌬␤ versus substrate concentration (Fig. 4b). Noncompetitive inhibition was represented by the straight line. This plot is analogues to the diagnostic plot of slope versus substrate concentration (36).
Fluoremetric Analysis of Enzyme Inhibitor Interactions-The localized conformational changes induced in the HIV-1 PR due to the interaction with ATBI were investigated by fluorescence spectroscopic studies. The sequence data indicated the presence of four Trp residues A-6, A-42, B-6, and B-42, two on each monomer of the HIV-1 PR (19). The visualization and accessibility calculations of these Trp residues revealed that they are present on the surface of the enzyme and thus are excellent probes to monitor the changes in the tertiary structure due to ligand binding. Therefore, the conformational changes induced in the HIV-1 PR upon binding of ATBI were monitored by exploiting the intrinsic fluorescence by excitation of the -* transition in the Trp residues. The fluorescence emission spectra of the HIV-1 PR exhibited an emission maxima ( max ) at ϳ342 nm as a result of the radiative decay of the -* transition from the Trp residues, confirming the hydrophilic nature of the Trp environment. The titration of the native enzyme with increasing concentrations of ATBI resulted in a concentrationdependent quenching of the tryptophanyl fluorescence (Fig. 5). However, the max of the fluorescence profile indicated no blue or red shift, revealing that the ligand binding caused reduction in the intrinsic protein fluorescence. A progressive quenching in the fluorescence of the HIV-1 PR at 342 nm was observed concomitant to the binding of substrate (Lys-Ala-Arg-Val-Nlep-nitro-Phe-Glu-Ala-Nle-amide). Further, to throw light upon the mechanism of inactivation of the HIV-1 PR by ATBI, we have analyzed the interaction of two representative competitive inhibitors, N-acetyl-Leu-Val-Phe-Al (where Al is aldehyde) (37) and pepstatin (2) by steady-state intrinsic fluorescence measurements. The binding of the competitive inhibitors led to the decrease in the quantum yield of the tryptophanyl fluorescence as indicated by the quenching of the emission spectra of the HIV-1 PR. The comparative analysis of the intensity changes in the fluorescence spectra of the HIV-1 PR upon binding of the substrate or the known active site-based inhibitors was found to be similar to that of ATBI, suggesting that ATBI binds in the active site of the enzyme.
Secondary Structural Analysis of Enzyme Substrate-Inhibitor Complexes-To evaluate the effects of the inhibitor on the secondary structure of the enzyme, we have analyzed the CD spectra of the HIV-1 PR-ATBI complex. The secondary structure contents of the HIV-1 PR as determined from the crystallographic data were 4.04% ␣-helix, 47.47% ␤-sheet, and 48.49% of aperiodic conformation (19). 2 The estimated secondary structure contents from the CD analysis were 5% ␣-helix, 48% ␤-sheet, and 47% aperiodic structure, which are in total agreement with the crystallographic data. The circular dichroism spectrum of the HIV-1 PR-ATBI complex showed a pronounced shift in the negative band at 220 nm of the native enzyme to 225 nm (Fig. 6). This shift reveals a subtle change in the secondary structure of the enzyme upon ligand binding. To elucidate the changes in the secondary structure of the enzymeinhibitor complex, we have compared it with that of the HIV-1 PR-substrate complex. Interestingly, the HIV-1 PR-ATBI and HIV-1 PR-substrate complexes exhibited a similar pattern of negative ellipticity in the far-UV region, suggesting that the inhibitor causes similar structural changes and was distinctly different from that of the unliganded enzyme. DISCUSSION The present paper describes a spectrofluorometric approach toward investigating the localized conformational changes induced in the HIV-1 PR upon binding of the noncompetitive inhibitor ATBI. We have shown by analyzing the kinetic parameters of the interactions of the HIV-1 PR and the inhibitor that Michaelis-Menten kinetics cannot be applied for this inhibition study. The failure of substrate protection against HIV-1 PR inhibition by ATBI and the nondissociative nature of the HIV-1 PR-ATBI complex with multiple dilutions and washings led us to apply tight binding inhibition kinetics. The short time observed for the inhibition mandated performance of the kinetics under second-order rate conditions. Observed ␣ and ␤ values for ATBI were independent of the substrate concentration and relatively constant, implying that the binding of the inhibitor was not influenced by the binding of the substrate. However, a typical rectangular hyperbola resulted in a reciprocal plot of 1/␤ versus [S]. We have concluded, by using a diagnostic   FIG. 1. rp-HPLC purification of ATBI. Shown is the elution profile of ATBI obtained by rp-HPLC on a Lichrosorb RP-18 column preequilibrated with 10% acetonitrile and 0.1% trifluoroacetate. a, 100 l of the lyophilized ATBI sample was loaded on a linear gradient of 0 -50% acetonitrile with water containing 0.01% trifluoroacetate for 15 min at a flow rate of 0.5 ml/min and monitored at 210 nm. The fractions containing the peaks A, B, and C having retention times of 2.553, 3.268, and 3.853 min, respectively, were collected manually and assayed for the anti-HIV-1 PR activity. b, 10 l of the fractions containing the peak A (associated with the anti-HIV-1 PR activity) was reloaded onto the rp-HPLC system under similar experimental conditions. The peak detected showed a retention time of 2.560 min.
plot of ⌬␤ versus [S], that the inactivation of the HIV-1 PR by ATBI was noncompetitive.
Deciphering the crystal structure of the HIV PR and its inhibitor complexes has gained immense interest among the crystallographers in the last decade. From the available crystallographic data, it is deduced that binding of substrate or peptide-analogue inhibitors in the substrate-binding site of HIV-1 PR induces conformational changes in the flaps (3,14,38). The apparent function of these flaps is to force the peptide substrate into a ␤-sheet in the active site and to correctly position its scissile bond between the two catalytic aspartyl residues. The flaps accomplish this by the establishment of a series of hydrogen-bonding interactions between amide nitrogens and carbonoyl oxygens of the peptide substrate. Our interpretation for the changes observed in the secondary structure of the HIV-1 PR due to the binding of the substrate to the active site can be correlated to the inward movement of the flaps. It is significant to note that the secondary structure of the HIV-1 PR undergoes similar pattern of changes upon binding of the substrate or inhibitor. Thus, we have attributed the observed secondary structure changes in the HIV-1 PR-ATBI complex to the inward movement of the flaps of the HIV-1 PR. The noncompetitive nature of the inhibitor may be addressed due to the better binding affinity of the inhibitor to the active site than the substrate. This, however, does not exclude the possibility of the differential binding pockets for the inhibitor and the substrate in the active site of the enzyme.
The tryptophanyl fluorescence appears to be uniquely sensitive to shielding by a variety of ligands because of the propensity of the excited indole nucleus to emit energy in the excited state. The Trp residues (A-42 and B-42) of the HIV-1 PR are present next to the Lys-43, the first residue of the flap region, FIG. 2. Chemical properties of ATBI. a, the purified ATBI was analyzed for the determination of the molecular mass using an acetonitrile/ water (1:1) system as the mobile phase on a quadrupole electrospray mass spectrometer. b, schematic representation of the chemical structure of ATBI.
which extends from Lys-43 to Arg-57 (Fig. 7) (19). There have been reports of introducing a Trp residue, which would act as a highly specific reporter to monitor the structural changes in the flap regions by substrate-inhibitor binding (14). However, the site-directed mutagenesis studies of the HIV-1 PR have revealed that the enzymatic activity is extremely sensitive to mutations in the flap regions (39). The inhibitors that bind to the active site also bind to the inner face of the flaps of the HIV-1 PR. The binding of the inhibitor-substrate and the subsequent movement of the flaps may have influence on the intrinsic fluorescence of the Trp-42 residues. Based on the above assumption, we have exploited these two Trp residues of the HIV-1 PR to investigate the localized conformational changes induced upon substrate or inhibitor binding. Our fore-

FIG. 3. Binding of ATBI to the HIV-1 PR and inhibition kinetics
analyses. a, the proteolytic activity of the purified HIV-1 PR was determined in the presence of increasing concentrations of ATBI. The percentage inhibition of the HIV-1 PR activity was calculated from the residual enzymatic activity. The sigmoidal curve indicates the best fit for the percentage inhibition data obtained, and the IC 50 value was calculated from the graph. b, enzymatic activity of the HIV-1 PR (25 g/ml) was estimated using the substrate Lys-Ala-Arg-Val-Nle-p-nitro-Phe-Glu-Ala-Nle-amide (40 M (OE) or 80 M (q)) at different concentrations of ATBI. Reciprocals of the reaction velocity were plotted versus the inhibitor concentration. The straight lines indicated the best fit of the data obtained. The inhibition constant K i was calculated from the point of the intersection of the plots. c, the HIV-1 PR (25 g/ml) was incubated, without (f) or with the inhibitor at 10 nM (q) and 20 nM (OE) and assayed at increasing concentrations of the substrate. The reciprocals of the rate of the substrate hydrolysis for each inhibitor concentration were plotted against the reciprocals of the substrate concentrations. K i was determined from the formula as per the noncompetitive type of inhibition. going results have suggested that ATBI binds in the active site of the enzyme and is a unique example where the conformational changes in the flaps were investigated by monitoring the radiative decay of the -* transition from the Trp residues without mutating any of the residues in the flaps. The fluorescence quenching of the HIV-1 PR by ATBI revealed that the binding of the inhibitor reduces the quantum yield of the Trp emission. These results were further corroborated by the quenching studies of the HIV-1 PR in the presence of the substrate and the known competitive inhibitors. The inhibition of the HIV-1 PR by N-acetyl-Leu-Val-Phe-Al (where Al is aldehyde) and pepstatin is well documented. The quenching of the tryptophanyl fluorescence upon binding of the substrate or the active site inhibitors can be very well explained by the shielding effect of the Trp residues due to the inward movement of the flaps. The comparison of the emission spectra of the HIV-1 PR upon binding of ATBI with that of the substrate or the active site inhibitors led us to conclude that ATBI binds to the active site of the enzyme and induces the inward movement of the flaps, thereby reducing the radiative decay of the intrinsic Trp fluorescence. The concentration-dependent quenching of Trp fluorescence showed that max did not undergo any red or blue shift, wherein the quenching of fluorescence was consid-erably high. These findings indicated that the polarity of the Trp environment was negligibly altered after binding of the inhibitor, suggesting minimal conformational changes in the tertiary structure of the HIV-1 PR.
The majority of the inhibitors of HIV-1 PR are hydrophobic in nature. There is a scarcity of hydrophilic peptidic inhibitors of HIV-1 PR, which is significant for the bioavailability of the drug. Inhibition of the HIV-1 PR by the hydrophilic peptides derived from the transframe region of gag-pol has been reported (40). The transframe octapeptide Phe-Leu-Arg-Glu-Asp-Leu-Ala-Phe, the N terminus of the transframe octapeptide, and its analogues are competitive inhibitors of the mature HIV-1 PR. The tripeptides Glu-Asp-Leu and Glu-Asp-Phe derived from the transframe octapeptide were found to be most potent. The x-ray crystallographic studies showed the interactions of Glu at P2 and Leu at P1 of Glu-Asp-Leu with residues of the active site of the HIV-1 PR. As a hydrophilic inhibitor containing residues Asp, Ala, and Glu, ATBI may have a similar mode of interaction with the residues in the active site of the HIV-1 PR to that of the transframe octapeptide, but it is not feasible to understand the interactions at the atomic level at present. However, the crystal structure will aid in understanding the mechanism of inactivation of the HIV-1 PR by ATBI. With the existing experimental evidence, we visualize that the charged side chains of the amino acids, the amide nitrogens, and the carbonoyl oxygen groups of ATBI could form many intermolecular hydrogen bonds and other weak interactions (van der Waals, ionic, etc.) with the ␤-sheet of the flaps and with the other residues present in or near the active site. Further, we propose that the tight binding and noncompetitive nature of ATBI in conjunction with the multiple nonbonded interactions may be sufficient to cause the loss of the dynamic flexibility of the flaps, which is crucial for the substrate binding and catalysis of the enzyme. A schematic diagram representing the proposed mechanism is depicted in Fig. 7. The noncompetitive nature of the ATBI indicated that the inhibitor-complexed form of the HIV-1 PR loses its binding ability to the substrate, since the flaps can no more open up for the substrate to be aligned in the active site of the HIV-1 PR, which subsequently results in the inactivation of the enzyme. These observations are at variance with the binding of the substrate to the enzyme in the absence of ATBI, where the flexibility of the flaps can be regained after catalysis.
Inhibitors directed toward the active site of the HIV-1 PR are well documented (41,42). Despite their loss in potency due to the spontaneous mutations occurring in the active site (29,43) leading toward the drug resistance behavior of the virus, there is a paucity of literature on noncompetitive inhibitors. A constant search for the new class of HIV-1 PR inhibitors with high potency is a frontier area of biomedical research. The side chains of the peptidic inhibitor ATBI, which are capable of forming many nonbonded interactions (hydrogen bonding, van der Waals interactions, etc.) with the enzyme, might result in superior resistance characteristics in comparison with the competitive inhibitors. ATBI, as proposed, could interact with the backbone of the ␤-sheet of the flaps of the HIV-1 PR, thus eliminating the probability of drug resistance by a single mutation. ATBI, by virtue of its unique sequence and noncompet- (shown above the arrows), which is important for the substrate binding and catalysis. The binding of the inhibitor (as indicated by the solid block) in the active site induces inward movement of the flaps (as indicated by the arrows). Further, we propose that the noncompetitive nature of ATBI, along with its multiple nonbonded interactions with the flaps, is responsible for the loss of the dynamic flexibility of the flaps, resulting in the inactivation of the HIV-1 PR. The structure of the HIV-1 PR is as described in PDB ID.1AID.