Alternate substrate inhibition of cholesterol esterase by thieno[2,3-d][1,3]oxazin-4-ones.

In a kinetic study, the interaction of bovine pancreatic cholesterol esterase (CEase) with fused 1,3-oxazin-4-ones and 1,3-thiazin-4-ones was investigated, and the compounds were characterized as alternate substrate inhibitors. Inhibition assays were performed in the presence of sodium taurocholate with p-nitrophenyl butyrate as chromogenic substrate. Strong active site-directed inhibition was detected for 2-diethylaminothieno[2,3-d][1,3]oxazin-4-ones with a cycloaliphatic chain at positions 5,6. The most potent inhibitors, compounds 3 and 4, exhibited K(i) values of 0.58 and 1.86 microm, respectively. An exchange of the ring oxygen by sulfur and the removal of the cycloaliphatic moiety as well as the replacement of the thiophene ring by benzene led to a loss of inhibitory potency. CEase has the capability to catalyze the hydrolysis of representatives of fused 1,3-oxazin-4-ones as well as the highly stable 1,3-thiazin-4-ones by using an acylation-deacylation mechanism. Hydrolyses were performed in the presence of a high enzyme concentration, and products were identified spectrophotometrically and by means of high performance liquid chromatography. The kinetic parameters V(max)I and V(max)I/K(m)(I) for the CEase-catalyzed turnover were determined. The compounds whose enzyme-catalyzed hydrolysis followed first-order kinetics (K(m)(I) > 25 microm) failed to inhibit CEase. On the other hand, inhibitors 3 (initial concentration of 25 microm) and 4 (20 microm) were hydrolyzed by CEase under steady-state conditions in the first phase of the reaction. Rate-limiting deacylation was demonstrated in nucleophilic competition experiments with ethanol as acyl acceptor wherein the conversion of compound 3 was accelerated up to an ethanol concentration of 1.5 m. The characterization of these compounds (i.e. 3 and 4) as alternate substrate inhibitors is not only based on the verification of the CEase-catalyzed hydrolysis. It also rests upon the concurrence of corresponding K(i) values obtained in the inhibition assay compared with separately determined K(m)(I) values of their enzyme-catalyzed consumption, as could be predicted from the kinetic model used in this study.

Bile salt-dependent lipase, also referred as cholesterol esterase (CEase) 1 (Sterol esterase, EC 3.1.1.13), is found in the pancreatic secretion of a wide range of species as well as in lactation of mammals. Pancreatic CEase once secreted into the duodenum and activated by primary bile salts catalyzes the hydrolysis of a broad spectrum of substrates including cholesteryl esters, triacylglycerides, phospholipids, and esters of lipid-soluble vitamins (1,2). Most probably, the role of CEase extends beyond that of simply hydrolyzing dietary lipids. Circulating CEase may function as a cholesterol transfer protein (3) and may have a deleterious effect in atherosclerosis processes, because it has been reported to convert the larger and less atherogenic low density lipoprotein to the smaller and more atherogenic low density lipoprotein subspecies (4). However, the role of plasmatic CEase in atherogenesis and the relationship of the enzyme to various pathological conditions are not clearly established so far (2). CEase belongs to the ␣/␤-hydrolase fold family of proteins whose members, mostly serine esterases, share secondary and tertiary structural features (5,6). Ester hydrolysis is catalyzed by the operation of a catalytic triad (Ser-194, His-435, and Asp-320 in the case of CEase numbered for the rat enzyme). This triad is stereochemically convergent with catalytic triads of serine proteases, and like serine proteases, serine esterases of the ␣/␤-hydrolase fold family use an acylation-deacylation mechanism. Both the acylation and deacylation stages transit tetrahedral intermediates that are stabilized by a tripartite oxanion hole (5)(6)(7).
In recent years, much attention has focused on the inhibition of CEase as a potential target particularly for the development of hypocholesterolemic agents. Sharing a common mechanism for substrate hydrolysis, CEase and serine proteases might be expected to be inhibited by the same classes of mechanismbased inhibitors that indeed have been demonstrated for boronic and borinic acids, aryl haloketones, aryl phosphates and phosphonates, and carbamates (8). Aryl and cholesteryl carbamates comprise the most studied class of CEase inhibitors (8 -11). A detailed characterization of transient inhibition by aryl carbamates was reported by Feaster et al. (9). Inhibition occurs because of rapid carbamylation of the active site serine followed by slow decarbamylation. Biphasic time courses reflect the progressive loss of enzyme activity in a nonlinear phase and a subsequent steady-state phase of the reaction. From their kinetic mechanism, such inhibitors are best described as pseudo-substrate inhibitors or as alternate substrate inhibitors (8,12).
The incorporation of a scissile CO-O or CO-N bond into a ring system has frequently been used in the design of mechanism-based inhibitors of serine proteases, e.g. of leukocyte elastase (13)(14)(15)(16)(17)(18). This concept has found less attention for the development of CEase inhibitors. However, 6-chloro-2-pyrones, representatives of a known class of mechanism-based inhibitors of serine proteases, have recently been described as potent CEase inhibitors (19). Although having the potential to act as alternate substrates or suicide inhibitors, competitive inhibition of CEase was postulated.
In the kinetic study presented in this paper, fused 1,3-oxa-* 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. zin-4-ones and 1,3-thiazin-4-ones were investigated as inhibitors of CEase. The acylation-deacylation mechanism by which 3,1-benzoxazin-4-ones and analogous thieno [1,3]oxazin-4-ones interact with various serine proteases is well documented (18, 20 -24). Their potency to inhibit serine proteases derives from the ratio of the acylation and deacylation rates. So far, these alternate substrate inhibitors of serine proteases have not been investigated as inhibitors of CEase. The aim of this study was (i) characterization of fused 1,3-oxazin-4-ones and 1,3-thiazin-4-ones as inhibitors of CEase, (ii) investigation of a CEasecatalyzed turnover of these heterocyclic compounds, and (iii) examination to what extent inhibitory potency can be concluded from the kinetic parameters of the enzyme-catalyzed conversion.

EXPERIMENTAL PROCEDURES
Materials and Instruments-CEase from bovine pancreas (41 units/ mg), sodium taurocholate (TC), and p-nitrophenyl butyrate (pNPB) were obtained from Sigma (Steinheim, Germany). Compounds 1-6 and 8 were prepared as described elsewhere (22,23,25). The synthesis and structural elucidation of compound 7 was done using general methods and instruments as reported previously (22). Spectrophotometric assays were done on a Perkin-Elmer Lambda 16 UV-visible spectrophotometer with a cell holder equipped with a constant temperature water bath. Incubation experiments were performed using an Eppendorf thermomixer comfort. Analytical HPLC was performed on a Bischoff chromatograph 2200 with an UV detector Lambda 1000. A 5-m Phenomenex Jupiter 250 ϫ 4.6-mm column was used at a flow rate of 0.5 ml/min.
Kinetic Parameters of the Non-enzymatic Hydrolysis-Alkaline hydrolysis of the compounds was followed spectrophotometrically at a fixed wavelength (compound 1 ϭ 348 nm; compounds 2-4 ϭ 335 nm; and compounds 5 and 6 ϭ 385 nm) by monitoring the disappearance of the compounds at 25°C in 50 mM CAPS, pH 9.5, for compound 1 and pH 11.25 for all other compounds, respectively. Stock solutions of the compounds were prepared in Me 2 SO. The final inhibitor concentration was 20 M, and the final Me 2 SO concentration was 5%. The reactions were monitored for at least two half-lives. For compounds 5 and 6 (5 M each), a final Me 2 SO concentration of 15% was used, and reactions were followed for 50 h. Curves were analyzed as first-order reactions.
CEase Inhibition Assay in the Presence of a Chromogenic Substrate-CEase inhibition was assayed spectrophometrically (9, 26) at 25°C. Assay buffer was 100 mM sodium phosphate, 100 mM NaCl, pH 7.0. A stock solution of CEase (2.44 mg/ml) was prepared with 100 mM sodium phosphate buffer, pH 7.0, freshly diluted with the same buffer and kept at 0°C. TC (12 mM) was dissolved in assay buffer and kept at 0°C. Stock solutions of pNPB and of the inhibitors were prepared in acetonitrile. Into a cuvette containing 430 l of assay buffer, 500 l of TC solution, 40 l of acetonitrile, 10 l of pNPB solution, and 10 l of an inhibitor solution were added and thoroughly mixed. After incubation for 2 min at 25°C, the reaction was initiated by adding 10 l of the enzyme solution. It was mixed again, and the reaction was monitored 1 min after the addition of the enzyme. Concentrations were as follows: 200 M pNPB, 6 mM TC, 19.5 ng/ml CEase, 6% acetonitrile, and different inhibitor concentrations. Uninhibited enzyme activity was determined by adding acetonitrile instead of the inhibitor solution. The determination of the Michaelis-Menten constant and maximum velocity for the substrate pNPB was done at eight different pNPB concentrations to result in values of K m S ϭ 110 Ϯ 11 M and V max S ϭ 5.09 Ϯ 0.15 M/min, respectively. A molar extinction coefficient, ⑀ ϭ 7.67 mM Ϫ1 cm Ϫ1 , for p-nitrophenol at pH 7.0 was used. The rates of CEasecatalyzed pNPB hydrolysis were corrected by those of the non-enzymatic hydrolysis of pNPB as determined by using 10 l of 100 mM sodium phosphate buffer, pH 7.0, instead of the enzyme solution.
Progress curves were monitored at 405 nm over 5 min, fitted, and analyzed as described below. Similarly, the inhibition of CEase by compound 3 was studied in the presence of different substrate concentrations.
CEase-catalyzed Turnover of the Heterocyclic Compounds-The enzymatic conversion of the compounds was followed spectrophotometrically at 25°C. Into a cuvette containing assay buffer, 500 l of TC solution, 50 l of acetonitrile, and 10 l of an inhibitor solution were added, thoroughly mixed, and incubated for 2 min at 25°C. The reaction was initiated by adding a volume (9 -12 l) of an enzyme solution (122 g/ml). This volume was adjusted by the determination of a 1:6250 dilution that converts pNPB (200 M) with a rate of 1.8 M/min. The entire volume was 1 ml containing the following concentrations: 6 mM TC, CEase (adjusted activity), 6% acetonitrile, and 10 -25 M inhibitor. The reactions were analyzed by monitoring UV-visible spectra in fixed time intervals or by following the time course at a fixed wavelength. For the latter experiments, the absorption maxima of the compounds were used. The molar extinction coefficient for each analyzed compound at the corresponding wavelength was determined separately in triplicate experiments. Control experiments to prove the stability of the compounds were performed by adding 100 mM sodium phosphate buffer, pH 7.0, instead of the enzyme solution.
Similarly, the influence of different ethanol concentrations on the kinetics of the CEase-catalyzed turnover of compounds 2 (measured at 331 nm) and 3 (measured at 349 nm) was determined. The entire volume was 1 ml containing the following concentrations: 6 mM TC, 12.2 g/ml CEase, 6% acetonitrile, 50 M compounds 2 or 3, and different ethanol concentrations.
HPLC Measurements-The CEase-catalyzed turnover of selected compounds was followed by HPLC. The mobile phase was a 1:1 mixture of acetonitrile and 50 mM phosphate buffer (Na 2 HPO 4 /KH 2 PO 4 ), pH 5.1. A flow rate of 0.5 ml/min was used, and absorption was monitored at 225 nm. A mixture of compound 4, TC, and assay buffer was incubated at 25°C for 10 min. The reaction was initiated by adding a volume (9 -12 l) of an enzyme solution (122 g/ml). This volume was adjusted by the determination of a 1:6250 dilution that converts pNPB (200 M) with a rate of 1.8 M/min. The entire volume was 1 ml containing the following concentrations: 6 mM TC, CEase (adjusted activity), 6% acetonitrile, and 20 M compound 4. The mixture was incubated for 30 h at 25°C, and 20-l aliquots were injected 5 min after enzyme addition and then in intervals of 6 h. The linearity of the peak areas versus concentration of compound 4 (range 1-24 M) was controlled in separate experiments. The reaction of CEase with compound 6 (10 M) was similarly investigated. The following retention times were observed: compound 4 ϭ 64 min; compound 6 ϭ 104 min; compound 8 ϭ 10.7 min.

Inhibition of CEase by Compounds 1-6-A series of fused
1,3-oxazin-4-ones and analogous 1,3-thiazin-4-ones was evaluated as inhibitors of bovine pancreatic cholesterol esterase. Detailed investigations were performed with 2-diethylamino derivatives 1-6 whose structures are shown in Table I. The reactions were followed over the period of 1-6 min after initiation by the addition of the enzyme. Progress curves were characterized by a linear steady-state turnover of the substrate, and values of a linear regression were fitted to an equation of competitive inhibition to obtain K i values (Table I). Strong inhibition was detected for thieno [2,3-d] [1,3]oxazin-4ones 3 and 4 with K i values Ͻ 2 M. As an example, the analysis of the inhibition kinetics of the thieno[2,3-d] [1,3]oxazin-4-one 3 is illustrated in Fig. 1.
It was determined next whether the inhibitors are active site-directed. This was done exemplary for the most potent inhibitor 3. The Hanes-Woolf plot [S]/v versus substrate concentration is shown in Fig. 2. The relevant relationship is given in Equation 1 , and K m S are the concentration and the Michaelis-Menten parameters of pNPB, respectively, and v is the relative steady-state velocity (v ϭ v s /v 0 ϫ 100%).
[I] is the concentration of inhibitor 3, and K i1 and K i2 are inhibition constants to estimate the type of inhibition. Linear regression gave values for slopes and vertical intercepts that were replotted against [I] to calculate K i1 and K i2 , respectively. A formally competitive inhibition could be deduced, and a value of K i1 ϭ 0.65 M was obtained in agreement with the result from the inhibition assay for compound 3 (K i ϭ 0.58 M) (Table I).
Product Identification of the CEase-catalyzed Turnover of Compounds 1-6-The possibility that fused 1,3-oxazinones and 1,3-thiazinones are substrates of CEase was checked for compounds 1-6. The solutions of each compound were incubated with CEase at 25°C in the presence of TC. The enzyme concentration was ϳ60-fold higher compared with the inhibition experiments in the presence of pNPB. Reactions were monitored spectrophometrically or by means of HPLC. Compound 3 indeed underwent an enzymatic turnover to form the corresponding thiophenecarboxylic acid 7 (see structure in Fig.  3). This could be concluded, because the final UV spectra obtained for the reaction with CEase (Fig. 5A) were identical with that of the reference compound 7 synthesized. Accordingly, the final spectra of the CEase-catalyzed conversion of compound 4 (data not shown) revealed the thiophenecarboxylic acid 8 (see structure in Fig. 3) as the product. The formation of compound 8 as the only product of the CEase-catalyzed transformation of compound 4 was additionally demonstrated by HPLC analysis using the synthesized thiophenecarboxylic acid 8 as reference. Enzyme-catalyzed hydrolysis of compounds 1 and 2 (Fig. 4A) and 4 was also examined spectrophotometrically. To a minor extent, compound 6 was converted by CEase. This reaction was monitored by HPLC. The acylation-deacylation mechanism as concluded from these results is shown exemplary for compounds 3 and 4 in Fig. 3.
Kinetic Parameters of the CEase-catalyzed Turnover of Compounds 1-6-The enzymatic transformation of the compounds was investigated with an equal CEase activity adjusted toward pNPB prior to all experiments. Kinetic parameters are given in Table I. The first-order rate constants obtained for compounds 1, 2, and 5 refer to V max I /K m I (27). V max I and K m I are the Michaelis-Menten parameters for the heterocyclic compounds as substrates of CEase. Michaelis-Menten constants could be ascertained being K m I Ͼ 25 M for 1, 2, and 5. The kinetic analysis for the CEase-catalyzed hydrolysis of compound 2 is illustrated in Fig. 4.
The determination of kinetic parameters of the enzymatic hydrolysis of compound 3 to form 7 is depicted in Fig. 5. The reaction approximated a zero-order kinetics until the conver-  , where v i is the initial velocity of the transformation, k is the first-order rate constant, and [I 0 ] is the initial concentration of the heterocyclic compound. Progress curves of the conversion of 3 and 4 were analyzed as described in the text. Alkaline hydrolysis was followed spectrophotometrically at a fixed wavelength and analyzed as first-order reaction. NI, no inhibition, which refers to a rate of Ն94% of the reaction in the absence of the inhibitor; ND, not determined. Enzyme-catalyzed hydrolysis of compound 4 to form 8 deviated from zero-order kinetics already in the initial part of the reaction and had to be analyzed by the methodology described above for compound 3. The disappearance of 4 was followed continuously at 346 nm to obtain kinetic parameters as well as by HPLC in 6-h intervals. The determination of the respective concentrations of compound 4 by HPLC (data not shown) was completely in agreement with the spectrophotometrical measurement. The very slow conversion of compound 6 by CEase was determined by HPLC. After an incubation over 24 h, 96% of the unchanged compound still was detected.

,3-oxazin-4-ones and 1,3-thiazin-4-ones
Nucleophilic Competition-The effect of ethanol on the CEase-catalyzed transformation of compounds 2 and 3 was investigated by monitoring the reactions at the absorption maxima of both compounds. The conversion of compound 2 (Fig. 6A) was significantly inhibited by ethanol. Transformations were analyzed as first-order reactions. Initial velocities were plotted as relative values against ethanol concentration (Fig. 6B) to allow for the determination of an apparent inhibition constant of ethanol (K i Ј ϭ 530 M) by non-linear regression using an equation of competitive inhibition.
A different feature was found in the case of compound 3. Transformations followed zero-order kinetics (Fig. 7A), and an increase of maximum velocities up to an ethanol concentration of 1.5 M was observed. A plot of relative maximum velocities versus ethanol concentration is shown in Fig. 7B. Thus, an inhibitory effect of ethanol as it could be concluded from the above mentioned experiment is more than compensated. This result indicated a nucleophilic activation by ethanol to affect deacylation as the rate-determining step that V max I monitors.

DISCUSSION
The aim of the current study was the kinetic characterization of the interaction of pancreatic cholesterol esterase with fused 1,3-oxazin-4-ones and 1,3-thiazin-4-ones. Based on preliminary studies (data not shown), we intended to focus on 2-diethylamino derivatives. Benzoxazinone 1 is a poor inhibitor of CEase with a K i value of 53 M ( Table I)  ment of the benzene unit by a (substituted) thiophene ring (compounds 2-4) resulted in a strong enzyme inhibition by thieno [1,3]oxazinones with cycloaliphatic moieties (compounds 3 and 4). An exchange of the ring oxygen by sulfur (compound 3 versus 6) led to a loss of inhibition.
Active site-directed inhibition as shown for compound 3 (Fig.  2) is a prerequisite for alternate substrate inhibition. Clear evidence for this type of inhibition was provided by the findings that the inhibitors themselves undergo an enzymatic hydrolysis as shown exemplary in Fig. 3. A similar mode of interaction of fused 1,3-oxazinones and 1,3-thiazinones has been reported for the inhibition of several serine proteases including chymotrypsin, leukocyte elastase, and chymase (18, 20 -24). Typically, a time-dependent inhibition was observed, and the rates of acylation and deacylation, respectively, were available by using slow binding kinetics (28).
In this inhibition study, CEase-catalyzed pNPB hydrolysis gave straight lines of the reaction progress, indicating that a steady-state was already reached at the beginning of the measurements. A relief from steady-state attributed to enzymecatalyzed turnover of the inhibitor did not occur within 6 min. The kinetic model of the inhibition assay is part ([Nu] ϭ 0) of FIG. 4. Kinetics of the CEase-catalyzed hydrolysis of compound 2. Reaction was performed in 100 mM sodium phosphate, 100 mM NaCl, pH 7.0, with 6 mM TC and 6% acetonitrile. Initial concentration of compound 2 was 25 M. A, depletion of compound 2 is illustrated by monitoring UV-visible spectra in 4-min intervals. The arrow indicates the initial spectrum. B, hydrolysis was followed at 331 nm and analyzed as first-order reaction. A first-order rate constant of 0.085 Ϯ 0.00002 min Ϫ1 that corresponds to V max I /K m I was obtained by nonlinear regression. Control reaction in the absence of CEase is shown to demonstrate the stability of compound 2.
The rates of the enzyme-catalyzed inhibitor consumption, v I , were determined in the absence of both pNPB ([S] ϭ 0) and ethanol ([Nu] ϭ 0). The kinetic model is part of Fig. 8 On the basis of these expressions, the kinetic results of 2-diethylamino-substituted compounds will be discussed. Benzoxazinone 1 is a comparable good substrate for CEase, whereas the replacement of the benzene unit by thiophene leads to a 6-fold decrease in V max I /K m I (compound 1 versus 2, Table I). The reaction of both compounds with CEase followed first-order kinetics, and thus neither EI nor E-I did accumulate within the time course studied. The influence of the benzenethiophene replacement on the affinity of the compound toward the active site is assumed to be rather small. Different rates simply reflect a reduced chemical reactivity of the thiophenederived compound 2. An enhanced electron density at the thiophene C-atoms results in a decreased carbonyl activity in the case of compound 2 (22,23). The second-order rate constant k OH Ϫ of the alkaline hydrolysis can be used to estimate intrinsic reactivity toward nucleophiles and is two orders of magnitude lower for compound 2 compared with 1.
The replacement of the ring oxygen of compound 2 by sulfur results in a further enhanced chemical stability and decreased carbonyl reactivity in the case of compound 5, which shows a 10-fold lower k OH Ϫ value (compound 2 versus 5, Table I). This effect is attributed to a stronger resonance stabilization for the thiolactone-containing heterocycles as discussed elsewhere for Reactions were performed in 100 mM sodium phosphate, 100 mM NaCl, pH 7.0, with 6 mM TC and 6% acetonitrile. Initial concentration of compound 3 was 50 M. A, depletion of compound 3 was followed at 349 nm in the absence of ethanol and at eight different ethanol concentrations (0.25-2 M). Progress curves were analyzed as zero-order reactions, and rates were determined by linear regression. Control reactions in the absence of CEase were performed for each ethanol concentration. The control for [EtOH] ϭ 1.5 M is shown. B, the relative rates obtained from fits to the data shown in A (rate in the absence of ethanol ϭ 100% activity) were plotted against the concentrations of ethanol. experiment in the absence of a chromogenic substrate is sufficient to determine K i . We are currently investigating this methodology for alternate substrate inhibitors of other serine esterases.
Our experimental data are in agreement with the classification of alternate substrate inhibitors as stable analogs that have the potential for conversion to products during a normal course of catalysis with a rate of one or more steps that has become extremely slow (12). The separately determined kinetic parameters for the consumption of the alternate substrate inhibitor are operative in the inhibition assay system but where consumption might practically not occur. For example, inhibitor 3 was assayed at concentrations 2-10 M over 6 min, and for this period of time a product formation of less (because of substrate competition) than 9 nM could be estimated by considering V max I and the enzyme concentrations used. In summary, we have analyzed alternate substrate inhibitors of CEase based on derivations of a kinetic system including a one-substrate, two-products, two-step irreversible reaction for substrate hydrolysis and a one-substrate, one-product, one-step irreversible reaction for the hydrolysis of the alternate substrate inhibitors. Strong inhibition was achieved by introducing a cycloaliphatic moiety into the thieno[2,3d] [1,3]oxazin-4-one skeleton, which results in a remarkable decrease of the deacylation rate. These derivatives (i.e. compounds 3 and 4) show high chemical stability and act as true alternate substrate inhibitors of pancreatic cholesterol esterase.