Inhibition of 1,4-β-d-Xylan Xylanohydrolase by the Specific Aspartic Protease Inhibitor Pepstatin

This is the first report that describes the inhibition mechanism of xylanase from Thermomonospora sp. by pepstatin A, a specific inhibitor toward aspartic proteases. The kinetic analysis revealed competitive inhibition of xylanase by pepstatin A with an IC50 value 3.6 ± 0.5 μm. The progress curves were time-depended, consistent with a two-step slow tight binding inhibition. The inhibition followed a rapid equilibrium step to form a reversible enzyme-inhibitor complex (EI), which isomerizes to the second enzyme-inhibitor complex (EI*), which dissociated at a very slow rate. The rate constants determined for the isomerization of EI to EI* and the dissociation of EI* were 15 ± 1 × 10-5 and 3.0 ± 1 × 10-8 s-1, respectively. The Ki value for the formation of EI complex was 1.5 ± 0.5 μm, whereas the overall inhibition constant Ki* was 28.0 ± 1 nm. The conformational changes induced in Xyl I by pepstatin A were monitored by fluorescence spectroscopy, and the rate constants derived were in agreement with the kinetic data. Thus, the conformational alterations were correlated to the isomerization of EI to EI*. Pepstatin A binds to the active site of the enzyme and disturbs the native interaction between the histidine and lysine, as demonstrated by the abolished isoindole fluorescence of o-phthalaldehyde-labeled xylanase. Our results revealed that the inactivation of xylanase is due to the interference in the electronic microenvironment and disruption of the hydrogen-bonding network between the essential histidine and other residues involved in catalysis, and a model depicting the probable interaction between pepstatin A with xylanase has been proposed.

In recent years considerable research efforts have been expended in the design and synthesis of glycosidase inhibitors not only to understand about the active site structures and mechanisms of these interesting enzymes but also in generating new therapeutic agents. Specific inhibitors of glycosidases have proved valuable in a number of applications ranging from mechanistic studies (1,2) to possible therapeutic uses such as viral infectivity through interference with normal glycosylation of coat proteins (3), against cancer, bacterial infections, and as insecticides (4). Glycosidase inhibitors as therapeutics has been highlighted in recent years with the successful marketing of the ␣-amylase inhibitor acarbose for the control of blood glucose levels and of the sialidase inhibitor relenza as a treatment for influenza. Acarbose is one of the new classes of oral ␣-glucosidase inhibitors used in the treatment type II diabetes. It is a linear tetrasaccharide of ␣-D-glucose residues modified at the non-reducing end. The non-reducing end group saccharide residue has a nitrogen replacing the interglycosidic oxygen atom and a double bond between C-4 and C-5 positions (5). Relenza (4-guanidino-2,4-dideoxy-2,3-dehydro-N-acetylneuraminic acid; Zanamivir) is a potent and highly specific sialidase inhibitor with inhibitory activity in vivo against both influenza A and B viruses (6).
Xylanases (1,4-␤-D-xylan xylanohydrolase) are glycosidases that catalyze the hydrolytic cleavage of ␤-1,4-linked polymers of D-xylose (7). They have raised enormous interest in the past decade in view of their application in clarification of juices and wines, conversion of renewable biomass into liquid fuels, and in development of environmentally sound prebleaching processes in the paper and pulp industry (8). Although extensive studies have been carried out on the industrial applications of xylanases, there is a paucity of reports on their molecular enzymology and clinical implications. Recently, glycosidases have been studied with a clinical perspective of locating enzyme-allergens (9). Some of these enzymes including xylanases and cellulases have been found to cause occupational and non-occupational allergies, such as respiratory and irritant contact dermatitis. Therefore, from the biomedical point of view inhibitors of this class of enzymes will have tremendous importance in near future. In addition, inhibition of cellulolytic and hemicellulolytic enzymes have potential applications to prevent the degradation of wood and cloth by the action of the hydrolytic enzymes present in the gut of the termites.
The three-dimensional structures of family 10 xylanases (10,11) have revealed the extended substrate binding cleft in which the surface residues are linked by an extensive hydrogen bond network. The clefts form deep grooves, consistent with their endo-mode action, and comprise a series of subsites, each one capable of binding a xylose moiety (12). The active site of xylanase contains two essential catalytic groups, one playing the role of acid/base and the other functioning as a nucleophile (13). These two groups have been identified as carboxyl groups, and a covalent intermediate is formed that undergoes hydrolysis to afford hemiacetal with net retention of anomeric stereochemistry (14). The transition states leading to and from the covalent intermediate have substantial oxacarbonium ion character, as indicated by kinetic isotope effects and by the effects of electron-withdrawing substituents on the sugar ring upon reaction rate (15). Analysis of active site amino acids that play an important role in substrate binding and in catalysis has been greatly facilitated by solving the crystal structure of family 10 xylanases covalently linked to mechanism-based cello-biosyl (2-deoxy-2-fluorocellobiosyl) and xylobiosyl (2-deoxy-2fluoroxylobiosyl) inhibitors (15). To gain further insight into the details of the hydrolytic mechanism of glycosidases, specific inhibitors are necessary that can act as mechanistic and structural probes. A diverse array of extremely potent, basic nitrogen-containing inhibitors has been developed over the years, and they have been found to be of great utility in the study of the glycosidase mechanism (7,16). However, there have been very few reports of naturally occurring inhibitors of xylanases. Pepstatin is a naturally occurring low molecular weight potent inhibitor specific for aspartic proteases (17). The unusual potency of pepstatin toward aspartic proteases has been widely exploited as a research tool for unraveling the mechanism of this group of enzymes (18), its biological functions (19), and in affinity chromatography (20,21). It also has been tested as a therapeutic agent for experimental control of gastric ulcer (17,22), carrageenin edema (17), hypertension (23,24), and infectious diseases like human immunodeficiency virus (25). To our knowledge there are no reports of pepstatin inhibition on xylanolytic enzymes.
In this paper we present slow-tight binding inhibition of the thermostable xylanase (Xyl I) 1 from a Thermomonospora sp. by a specific aspartic protease inhibitor, pepstatin A. The steadystate kinetics revealed a two-step inhibition mechanism, and the conformational modes observed during the binding of inhibitor to the enzyme were conveniently monitored by fluorescence analysis. The mechanism of inactivation of Xyl I by pepstatin A was delineated by monitoring the isoindole fluorescence of the o-phthalaldehyde (OPTA)-labeled enzyme and a model for the probable interactions have been proposed.

EXPERIMENTAL PROCEDURES
Materials-Oat spelled xylan, dinitrosalicylic acid, pepstatin A, and o-phthalaldehyde (OPTA) were obtained from Sigma. Sephadex A-50 and Sephacryl S-200 were obtained from Amersham Biosciences . All other chemicals used were of analytical grade.
Microorganisms Growth Condition and Purification of Xyl I-Thermomonospora sp. producing Xyl I is an alkalothermophilic actinomycete having optimum growth at pH 9 and 50°C. It was isolated from self-heating compost from the Barabanki district of Uttar Pradesh, India (26). The Thermomonospora sp. was grown at 50°C for 96 h for the production of Xyl I. The enzyme was purified to homogeneity from the extracellular culture filtrate by fractional ammonium sulfate precipitation (35-55%), DEAE-Sephadex ion exchange chromatography, and Sephacryl S-200 gel filtration chromatography (27).
Xylanase Assay and Inhibition Kinetics-Xylanase assay was carried out in phosphate buffer, 0.05 M, pH 6.0, by mixing a specified concentration of the enzyme with 0.5 ml of oat spelled xylan (10 mg/ml) in a reaction mixture of 1 ml and incubating at 50°C for 30 min. The reducing sugar released was determined by the dinitrosalicylic acid method (28). One unit of xylanase activity was defined as the amount of enzyme that produced 1 mol of xylose equivalent per min using oatspelled xylan as the substrate under assay conditions. Protein concentration was determined according to the method of Bradford (29) using bovine serum albumin as the standard.
For initial kinetic analysis the kinetic parameters for the substrate hydrolysis were determined by measuring the initial rate of enzymatic activity. The inhibition constant (K i ) was determined by Dixon method (30) and also by the Lineweaver-Burk equation. The K m value was also calculated from the double-reciprocal equation by fitting the data into the computer software Microcal Origin. For the Lineweaver-Burk analysis Xyl I (2 M) was incubated with pepstatin A at (1 M) and (2 M) and assayed at increased concentrations of xylan (1-10 mg/ml) at 50°C for 30 min. The reciprocals of substrate hydrolysis (1/v) for each inhibitor concentration were plotted against the reciprocals of the substrate concentrations and the K i was determined by fitting the resulting data. In Dixon's method (30), xylanolytic activity of Xyl I (2 M) was measured in the presence of 5 and 10 mg/ml xylan at concentrations of pepstatin A ranging from 0 to 7 M at 50°C for 30 min. The reciprocals of substrate hydrolysis (1/v) were plotted against the inhibitor concentration, and the K i was determined by fitting the data using Microcal Origin.
For the progress curve analysis assays were carried out in a reaction mixture of 1 ml containing enzyme, substrate, and inhibitor at various concentrations. The reaction mixture contained Xyl I (50 nM) in sodium phosphate buffer, 0.05 M, pH 6.0, and varying concentrations of pepstatin A (0.3-3 M) and xylan (10 mg/ml). The reaction was initiated by the addition of Xyl I at 50°C, and the release of products was monitored at different time intervals by estimating the reducing sugar at 540 nm. In each slow binding inhibition experiment, five to six assays were performed with appropriate blanks. For the kinetic analysis and rate constant determinations, the assays were carried out in triplicate, and the average value was considered throughout. Further details of the experiments are given in the respective figure legends.
Evaluation of Kinetic Parameters-Initial rate studies for reversible; competitive inhibition was analyzed according to Equation 1, where K m is the Michaelis constant, V max is the maximal catalytic rate at saturating substrate concentration S, K i ϭ (k 4 /k 3 ) is the dissociation constant for the first reversible enzyme-inhibitor complex, and I is the inhibitor concentration (31). The progress curves for the interactions between pepstatin A and Xyl I were analyzed using Equation 2 (32,33), where [P] is the product concentration at any time t, v 0 and v s are the initial and final steady-state rates, respectively, and k is the apparent first-order rate constant for the establishment of the final steady-state equilibrium. As a prerequisite for tight binding inhibitors, corrections have been made for the reduction in the inhibitor concentration that occurs on formation of the enzyme inhibitor (EI) complex. This is because in the case of tight binding inhibition, the concentration of EI is not negligible in comparison to the inhibitor concentration, and the free inhibitor concentration is not equal to the added concentration of the inhibitor. The corrections of the variation of the steady-state velocity with the inhibitor concentrations were made according to Equation 3 and 4 as described by Morrison and Walsh (34), where K i Ј ϭ K i * (1 ϩ S/K m ), k 7 rate constant for the product formation, I t , and E t stands for total inhibitor and enzyme concentration, respectively. The relationship between the rate constant of enzymatic reaction k, and the kinetic constants for the association and dissociation of the enzyme and inhibitor was determined as per Equation 5.
The progress curves were analyzed by Equations 2 and 5 using nonlinear least-square parameter minimization to determine the best-fit values with the corrections for the tight binding inhibition. The overall inhibition constant is determined as given by Equation 6.
For the time-dependent inhibition, there exists a time range in the progress curves wherein formation of EI* is small. Within this time range, it is possible to directly measure the effect of the inhibitor on v 0 , i.e. to measure K i directly. Values for K i were obtained from Dixon analysis at a constant substrate concentration as described in Equation 7.
The rate constant k 6 for the dissociation of the second enzymeinhibitor complex was measured directly from the time-dependent inhibition. Concentrated Xyl I and pepstatin A were incubated in a reaction mixture to reach equilibrium followed by large dilutions in assay mixtures containing near-saturating substrate. Xyl I (2 mM) was preincubated with equimolar concentrations of pepstatin A for 120 min in sodium phosphate buffer, 0.05 M, pH 6.0. 5 l of the preincubated sample was removed and diluted 5000-fold in the same buffer and assayed at 50°C using xylan at (150 mg/ml) at different time intervals.
Fluorescence 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 -400 nm at 25°C. The slit widths on both the excitation and emission were set at 5 nm, and the spectra were obtained at 100 nm/min.
For inhibitor binding studies, Xyl I (2 M) was dissolved in sodium phosphate buffer, 0.05 M, pH 6.0. Titration of the enzyme with pepstatin A was performed by the addition of different concentrations of the inhibitor to a fixed concentration of enzyme solution. For each inhibitor concentration on the titration curve a new enzyme solution was used. All the data on the titration curve were corrected for dilutions, and the graphs were smoothened. The magnitude of the rapid fluorescence decrease (F 0 Ϫ F) occurring at each Pepstatin A concentration was computer-fitted to Equation 8 to determine the calculated value of K i and ⌬F max (35).
The first order rate constants for the slow loss of fluorescence k obs , at each inhibitor concentration [I] were computer fitted to the Equation 9 (31) for the determination of k 5 under the assumption that for a tight binding inhibitor, k 6 can be considered negligible at the onset of the slow loss of fluorescence.
Time course of the protein fluorescence after the addition of inhibitor was measured for 300 s with excitation and emission wavelengths fixed at 295 and 340 nm, respectively, with data acquisition at 0.1-s intervals. Corrections for the inner filter effect were performed as described by Equation 10 (36), where F c and F stand for the corrected and measured fluorescence intensities, respectively, and A ex and A em are the absorbances of the solution at the excitation and emission wavelengths, respectively. Background buffer spectra were subtracted to remove the contribution from Raman scattering. Effect of Pepstatin A on the Isoindole Fluorescence of OPTA-labeled Xyl I-Fresh OPTA solution was prepared in methanol for each experiment. The modification was carried out by incubating Xyl I (2 M) in 1 ml of 0.05 M sodium phosphate buffer, pH 6, with 50 M OPTA at 25°C. Methanol had no effect on the activity of the enzyme and was always less than 2% (v/v). The formation of Xyl I-isoindole derivative was followed spectrofluorometrically by monitoring the increase in fluorescence with the excitation wavelength fixed at 330 nm. To monitor the effect of pepstatin A on the isoindole fluorescence of Xyl I, the enzyme was preincubated with pepstatin A (2 M) for 20 min, and then OPTA was added, and the formation of isoindole derivative was monitored as described above.

RESULTS
Kinetic Analysis of the Inhibition of Xyl I-The aspartic protease inhibitor (pepstatin A) was well documented for its inhibitory activity toward pepsin, renin, cathepsin D, and human immunodeficiency virus 1 protease (37,38,39,25). The bifunctional nature of pepstatin A was established by its potency toward Xyl I, the xylanase purified from the Thermomonospora sp. Xyl I, a member of family 10 xylanase, is highly thermostable with half-lives of 86, 30, and 15 min at 80, 90, and 100°C, respectively, and is stable in an expansive pH range of 5-10 with more than 75% residual activity. Initial kinetic assessments revealed that pepstatin A is a competitive inhibitor of Xyl I with an IC 50 value of 3.6 Ϯ 0.5 M (Fig. 1). In the absence of pepstatin A, the steady-state rate of xylanolytic activity of Xyl I was reached rapidly, whereas in its presence a time-dependent decrease in the rate as a function of the inhibitor concentration was observed. Examination of the progress curves revealed a time range where the initial rate of reaction did not deviate from linearity (Fig. 2), and the conversion of EI to EI* was minimal. For a low concentration of pepstatin A this time range was 10 min, within which classical competitive inhibition experiments was used to determine the K i values (Equation 5). The value of the inhibition rate constant K i , associated with the formation of the reversible enzyme-inhibitor complex (EI) determined from the fits of data to the reciprocal equation, was 1.5 Ϯ 0.5 M (Fig. 3A), which was corroborated by the Dixon method (Fig. 3B). The apparent rate constant k, derived from the progress curves when plotted versus the inhibitor concentration, followed a hyperbolic function (Fig. 4), revealing that a fast equilibrium precedes the formation of the final slow dissociating enzyme-inhibitor complex (EI*), indicating a two-step, slow-tight inhibition mechanism (Scheme I). Indeed, the data could be fitted to Equation 5 by non-linear regression analysis, which yielded the best estimate of the overall inhibition constant K i * of 28.0 Ϯ 1 nM.
In an alternative method, the rate constant k 6 , for the conversion of EI* to EI, was determined by preincubating high concentrations of enzyme and inhibitor for sufficient time to allow the system to reach equilibrium. Dilution of the enzymeinhibitor complex into a relatively large volume of assay mixture containing saturating substrate concentration causes dissociation of the enzyme-inhibitor complex and, thus, regeneration of enzymatic activity. Under these conditions, v 0 and the effective inhibitor concentration can be considered approximately equal to 0, and the rate of activity regeneration will provide the k 6 value. After preincubating Xyl I with pepstatin A, the enzyme inhibitor mixture was diluted 5000-fold into the assay mixture containing the substrate at 50 K m . By least-squares minimization of Equation 2 to the data for recovery of enzymatic activity, the determined k 6 value was 3.0 Ϯ 1 ϫ 10 Ϫ8 s Ϫ1 (Fig. 5), which clearly indicated a very slow dissociation of EI*. The final steady-state rate v S was determined from the control that was preincubated without the inhibitor. The value of the rate constant k 5 , associated with the isomerization of EI to EI*, was 15 Ϯ 1 ϫ 10 Ϫ5 s Ϫ1 , as obtained from fits of Equation 3 to the onset of inhibition data using the experimentally determined values of K i and k 6 ( Table I). The overall inhibition constant K i * is a function of k 6 /(k 5 ϩ k 6 ) and is equal to the product of K i and this function. The k 6 value indicated a slower rate of dissociation of EI* complex and the half-life t1 ⁄2 , for the reactivation of EI* as determined from k 6 values was 64 Ϯ 2 ϫ 10 2 h, suggesting higher binding affinity of pepstatin A toward Xyl I. Where E stands for free enzyme, I is free inhibitor, EI is a rapidly forming preequilibrium complex, and EI* is the final enzyme-inhibitor complex. E may undergo inter-conversion into another form, E*, which binds to the inhibitor by a fast step, where k cf and k Ϫcf stand for the rate constants for forward and backward reaction, respectively, for the conversion of the enzyme.
Scheme I describes two alternative models for the time-dependent inhibition. The mechanism in Scheme Ia, where the binding of the inhibitor to the enzyme is slow and tight but occurs in a single step, is eliminated based on the data of Table  I. Scheme Ic represents the inhibition model where the inhibitor binds only to the free enzyme because the inhibitor has a measurable effect on the initial rates before the onset of slowtight binding inhibition, which has slowly adopted the transition-state configuration, can also be eliminated by the observed rates of onset of inhibition. Our foregoing results for the inactivation of Xyl I were, therefore, consistent with the slow-tight binding mechanism as described in Scheme Ib.
Effect of Inhibitor Binding on the Fluorescence of Xyl I-The kinetic analysis revealed a two-step inhibition mechanism, where the EI complex isomerizes to a tightly bound, slow dissociating EI* complex. This isomerization is a consequence of the conformational changes induced in Xyl I due to the binding of pepstatin A. The tryptophanyl fluorescence of Xyl I exhibited an emission maxima ( max ) at ϳ340 nm as a result of the radioactive decay of the -* transition from the Trp residues (Fig. 6). The binding of pepstatin A resulted in a concentrationdependent quenching of the fluorescence with saturation reaching at/above 6 M pepstatin A (inset of Fig. 6). The absence of blue or red shift in max negated any drastic gross conformational changes in the three-dimension structure of the enzyme due to inhibitor binding. The subtle conformational changes induced during the isomerization of EI to EI* were monitored by analyzing the tryptophanyl fluorescence of the complexes as a function of time. Binding of pepstatin A resulted an exponential decay of the fluorescence intensity as indicated by a sharp decrease in the quantum yield of fluorescence followed by a slower decline to a stable value (Fig. 7). Furthermore, titration of pepstatin A against Xyl I revealed that the magnitude of the initial rapid fluorescence loss (F 0 Ϫ F) increased in a saturation-type manner (Fig. 8), which corrobo-rated the two-step slow tight binding inhibition of Xyl I by pepstatin A. From the data in Fig. 8, the magnitude of the rapid fluorescence decrease at a specific pepstatin A concentration was found to be close to the total fluorescence quenching observed in Fig. 6, indicating that the EI and EI* complexes have the same intrinsic fluorescence. The value of K i determined by fitting the data for the magnitude of the rapid fluorescence decrease (F 0 Ϫ F) was 1.2 Ϯ 0.5 M, and the k 5 value determined from the data derived from the slow decrease in fluorescence was 14 Ϯ 1 ϫ 10 Ϫ5 s Ϫ1 . These rate constants are in good agreement with that obtained from the kinetic analysis; therefore, the initial rapid fluorescence decrease can be correlated to the formation of the reversible complex EI, whereas the slow, time-dependent decrease reflected the accumulation of the tight bound slow dissociating complex EI*.
Effect of Pepstatin A on the Isoindole Fluorescence of Xyl I by OPTA-In our earlier report we have investigated the role of essential histidine and lysine residues in the active site of the Xyl I and shown that binding of the chemoaffinity label OPTA to these residues of the active site resulted in the formation of an isoindole derivative (40,41). The active site of Xyl I is composed of the catalytic carboxylic groups and the histidine residue, which play a crucial role in catalysis. To investigate the binding of pepstatin A to the active site and changes in the native intermolecular interactions, we have monitored the changes in the interaction of the lysine and histidine due to pepstatin A binding and their influence on the isoindole fluorescence of Xyl I (Fig. 9). The unbound enzyme did not show fluorescence when excited at 338 nm; however, incubation of OPTA with Xyl I resulted in an increase in the fluorescence with a max at 417 nm due to the formation of the isoindole derivative. The pepstatin A preincubated Xyl I failed to react with OPTA as revealed by the total loss of isoindole fluorescence, which not only confirmed the binding of pepstatin A to the active site of Xyl I but also further revealed that the binding of pepstatin A resulted in the formation of a new set of hydrogen bonding and other nonionic interactions. These altered weak interactions cause disruption of the native hydrogen-bonding network of the histidine and lysine residues, which are essential for the formation of isoindole derivative. DISCUSSION In recent years a number of naturally occurring reversible glycosidase inhibitors such as nojirimycin, castanospermine, and swainsonine have been reported (1). Nojirimycin and castanospermine are naturally occurring polyhydroxylated alkaloids known as aza-sugars. Nojirimycin is produced by Bacilli and Streptomyces sp. and shows inhibitory activity against plant and fungal glucosidases (42). Castanospermine was isolated from the seeds of Castanospermum australe (43) and was found to be a powerful inhibitor of ␣ and ␤ glucosidases (44). Swainsonine is an indolizidine alkaloid was isolated from the plant Swainsona canescens. Swainsonine is a specific and potent inhibitor of ␣-mannosidase (45). Another class of inhibitors is the covalent, irreversible type, typically affinity labels. These are generally synthetic analogues of sugars containing reactive groups such as epoxides, isothiocyanates, and ␣-halocarbonyls (1,46). There are reports of mechanism-based inhibitors such as conduritol epoxides (47), the quinone methide-generating glycosides (48), and the glycosylmethyl triazenes (49). These are more selective inhibitors whose efficacy depends upon binding and subsequent enzymatic action to generate a reactive species. Although a plethora of synthetic inhibitors has been reported, there is a lacuna of naturally occurring low molecular weight inhibitors of glycosidases from microorganisms. It is noteworthy to point out that several extremophiles are known to produce highly thermostable xylanases and cellulases; how- ever, these organisms have not been studied extensively for their potential exploitation toward isolation of inhibitors of important enzymes. Pepsin and Xyl 1 catalyze hydrolytic cleavage of two different substrates made of peptide bond and glycosidic bond, respectively. Pepsin, a model enzyme for aspartic protease, did show any activity on xylan, a sugar polymer made of glycosidic bond. The Xyl 1 did not show any activity against protein substrates, such as hemoglobin and casein, made of peptide bond (data are not shown), indicating these enzymes are absolutely specific toward their natural substrates. Here we present the first report of an aspartic protease specific inhibitor, pepstatin A, exhibiting slow-tight binding inhibition against xylanase from Thermomonospora sp. There are several reports on the inhibition of pepstatin against aspartic proteases (17,25,37,38,39), but this is the first study that explores the inhibition mechanism of this specific aspartic protease inhibitor against a xylanase. The inhibitor showed exceptionally high potency against Xyl I, the thermostable xylanase from a Thermomonospora sp., and its 1:1 molar ratio of interaction with the enzyme indicated its "tight binding" nature. The two-step inhibition mechanism was corroborated by the equilibrium binding studies of the enzyme and inhibitor and the correlation of the kinetic data with the conformational changes induced in the enzyme-inhibitor complexes. In the presence of competitive inhibitors, a number of enzymatic reactions do not respond immediately but display a slowonset of inhibition, which is referred as slow binding inhibition (50 -54). The establishment of the equilibria between enzyme, inhibitor, and enzyme-inhibitor complexes in slow binding inhibition occurs slowly on the steady-state time scale (54 -58). Enzyme-catalyzed reactions, where the concentrations of the enzyme and inhibitor are comparable and the equilibria are set up rapidly are referred as tight binding inhibition. Kinetically the slow binding inhibition can be illustrated by three mechanisms (Scheme I). When an inhibitor has a low K i value and the concentration of I varies in the region of K i , both k 3 I and k 4 values would be low. Thus, a simple second-order interaction between enzyme and inhibitor and low rates of association and dissociation would lead to slow binding inhibition. Alternatively, a two-step model depicts the rapid formation of an initial collisional complex EI, which slowly isomerizes to form a tightly bound slow dissociating complex EI*. Slow binding inhibition can also arise due to an initial slow interconversion of the enzyme E into another form, E*, which binds to the inhibitor by a fast step. Understanding the basis of the isomerization of EI to EI* could lead to the design of inhibitors that allow titration of the lifetime of the EI*. In case of slow-tight binding inhibition, the inhibitor will inhibit the enzyme competitively at the onset of the reaction; however, at increasing concentrations of inhibitor the rate of substrate hydrolysis will decrease hyperbolically as a function of time. In tight binding inhibition, corrections have to be made for the reduction in the inhibitor concentration that occurs on formation of the EI complex, since the concentration of EI is not negligible in comparison to the inhibitor concentration and the free inhibitor concentration is not equal to the added concentration of the inhibitor. The kinetic analysis of the xylanase inhibition in this paper provides a unique opportunity for the quantitative determination of these rates and affinities, which can be extended to other slow-tight binding inhibition reactions. The formation of EI complex between Xyl I and pepstatin A was too rapid to be measured at steady-state kinetics and was likely to be near diffusion control. However, the isomerization of EI to the second tightly bound enzyme-inhibitor complex, E*, was too slow and relatively independent of the stability of the EI or the ability of the inhibitor to stabilize the EI*. The k 6 values revealed very slow dissociation of the inhibitor from the EI*, indicating a highly stable, non-dissociative nature of the second complex. Therefore, for slow-tight binding inhibition the major variable is k 6 , the first-order rate constant associated with the conversion of EI* to EI, and the apparent inhibitor constant K i * depends on the ability of the inhibitor to stabilize the EI*. The half-life as derived from the k 6 value indicated a longer half-life of the EI*, which is an essential parameter for an inhibitor to have biomedical applications.
The characteristic feature of slow binding inhibition is the induction of conformational changes in the enzyme-inhibitor Xyl I (2 mM) was pre-incubated without (•) or with (f) equimolar concentrations of pepstatin A for 120 min in sodium phosphate buffer, 0.05 M, pH 6.0. At the specified times indicated by the points, 5 l of the preincubated sample was removed, diluted 5000-fold in the same buffer, and assayed for the xylanolytic activity using xylan (150 mg/ml). The rate constant associated with the regeneration of activity (k 6 ) was determined by estimating the reducing sugar as described under "Experimental Procedures."

TABLE I Inhibition constants of pepstatin A against Xyl I
Values of rate constants for Xyl I inhibition by pepstatin A were calculated from Scheme I at 50°C in sodium phosphate buffer, 0.05 M, pH 6.0, using oat spelt xylan as the substrate. IC 50 is from the inhibition profile. K i was determined from the steady-state time range for the competitive inhibition. k 6 is calculated from the regeneration assay, K i * and k 5 were determined from the equations as described under "Experimental Procedures." Inhibition constants Values IC 50 3.6 Ϯ 0.5 ϫ 10 Ϫ6 M K i 1.5 Ϯ 0.5 ϫ 10 Ϫ6 M K i * 28.0 Ϯ 1 ϫ 10 Ϫ9 M k 5 15 Ϯ 1 ϫ 10 Ϫ5 s Ϫ1 k 6 3.0 Ϯ 1 ϫ 10 Ϫ8 s Ϫ1 t1 ⁄2 64 Ϯ 2 ϫ 10 2 h SCHEME 1

Slow-Tight Binding Inhibition of Xylanase by Pepstatin
complex, resulting in the clamping down of the enzyme to the inhibitor, thus, the formation of a stable enzyme-inhibitor complex. The two-step inhibition mechanism of Xyl I by pepstatin A was reflected in the quenching pattern of the fluorescence of the enzyme-inhibitor complexes. The rate constants derived from the fluorescence analysis of the complexes corroborated the values derived from the kinetic analysis. Therefore, we propose that the initial rapid fluorescence loss reflected the formation of the reversible complex EI, whereas the subse- quent slower decrease was correlated to the accumulation of the tightly bound complex EI*. Any major alteration in the three-dimensional structure of Xyl I due to the binding of pepstatin A can be ruled out, since there was no shift in the tryptophanyl fluorescence of the complexes. Any disturbance in the environment of tryptophan residues may reflect in an alternation in emission wavelength, quantum yield, and susceptibility to quenching (59). Energy transfer to an acceptor molecule having an overlapping absorption spectrum can also contribute to fluorescence quenching (60). However, because the inhibitor has no absorption in the region of 290 -450 nm, the fluorescence quenching due to the energy transfer between the inhibitor and the tryptophan residues of Xyl I has been neglected. The other possibility is the presence of multiple sites, where binding at one induced rapid fluorescence change and at a second site caused the slow fluorescence decrease. This was verified by titrating a fixed concentration of Xyl I with increasing concentrations of pepstatin A. The xylanolytic activity decreased linearly with increasing concentrations of pepstatin A, yielding a stoichiometry close to 1:1 (also revealed by fluorescence) expected for the slow-tight binding inhibition, therefore, inconsistent with the presence of multiple high affinity sites. From the physical explanation for the quenching process it was apparent that the inhibitor induced fluorescence quenching followed by the formation of both the complexes. The agreement of the rate constants concomitant with the fluorescence changes observed during the time-dependent inhibition led us to correlate the localized conformational changes in the enzyme-inhibitor complex to the isomerization of the EI to EI*.
Conformational integrity of the active site of an enzyme is essential for its catalysis, and investigations on the molecular orientation of the functional groups of the active site as well as their microenvironment are areas of growing scientific interest. Chemo-affinity labeling is a powerful technique to assign the binding sites of ligand-macromolecule complexes, which combines some of the advantages of both the photoactivated and electrophilic affinity labeling (61). OPTA is a bifunctional, fluorescent chemoaffinity label, which until recently was known to have absolute specificity for amino and thiol groups (62) for the formation of an isoindole derivative. However, application of OPTA as a probe to ascertain the conformational flexibility and polarity of the active site of Xyl I by the formation of a fluorescent isoindole derivative with the lysine and histidine residue has been reported in our laboratory (40). OPTA contains two aldehyde groups, one of which reacts with the primary amine of lysine while the second group reacts with the secondary amine of the imidazole ring of histidine, resulting in the formation of the isoindole derivative. Our foregoing results revealed that, when Xyl I was preincubated with pepstatin A, OPTA failed to form the isoindole derivative, as reflected by the loss of fluorescence. The inability of OPTA to form the isoindole derivative with the pepstatin A-bound Xyl I could be attributed to the interaction of the inhibitor with either lysine or histidine or both the residues, thereby changing the native molecular interactions of these residues. The chemical structure of pepstatin is isovaleryl-L-valyl-L-valyl-statyl-L-alanyl-statin, which contains two residues of unusual amino acid statin ((3S,4S)-4amino-3-hydroxy-6-methyl-heptanoic acid). The statin is the major structural component responsible for the pepstatin inhibition against the aspartic protease. Aspartic proteases consist of two carboxyl groups at the active site, one of which has to be protonated and the other deprotonated for the enzyme to be active. Aspartic protease undergoes a general acid-base catalysis that may be called a "push-pull" mechanism. The nucleophilic attack is achieved by two simultaneous proton transfers, one from a water molecule to the diad of the two carboxyl groups and a second one from the diad to the carbonyl oxygen of the substrate with the concurrent CO-NH bond cleavage. Pepstatin is an analog of the transition state of peptic catalysis that tightly binds to active site of the enzyme (37). In the case of xylanase the active site contains two essential catalytic groups, one playing the role of acid/base and the other functioning as a nucleophile. We propose that pepstatin makes several interactions with the active site residues of Xyl I both through hydrogen bonds and non-bonded interactions (Fig.  10A). The hydroxyl group of statin residues of the inhibitor form hydrogen bonding with the amines of Lys and Gln. The model also depicts the hydrogen bonding between the carbonyl oxygen of Ala of the inhibitor with the secondary amine of the His of Xyl I. The catalytic site of xylanases consists of two carboxyl groups and an essential lytic water molecule and follows a general acid-base catalytic mechanism (63). Fig. 10B demonstrates the probable interaction of the essential His and Lys residues of Xyl I with the hydroxyl and carbonyl functional groups of the inhibitor. Based on the existing experimental evidences, we further propose that the other residues of the inhibitor could form many intermolecular hydrogen bonds and other weak interactions with the residues in or near the active site of Xyl I. We also visualize that the tight binding nature of pepstatin A in conjunction with the multiple nonbonded interactions may be sufficient to interfere in the native weak interactions between the carboxyl groups, the lytic water molecule, and the essential histidine residue of the active site, leading toward the inactivation of Xyl I.
Based on our observations, we conclude that the inhibition of Xyl I by pepstatin A followed slow-tight binding inhibition mechanism and the induced conformational changes are conveniently monitored by fluorescence spectroscopy. Chemo-affinity labeling of the enzyme active site has demonstrated that the inactivation of xylanase is due to the interference in the electronic microenvironment and disruption of the hydrogenbonding network between the essential histidine and other residues involved in catalysis. However, the crystal structure of the enzyme-inhibitor complex will aid to understand the mechanism of inactivation of Xyl I in depth and will further shed light on the molecular interactions between the enzyme and inhibitor.