New inhibitors of aldose reductase: Anti-oximes of aromatic aldehydes

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS

Vol. 286, No. 2, May 1, pp. 596-603,199l

New Inhibitors of Aldose Reductase: Anti-Oximes of Aromatic Aldehydes

Chun Shen and David S. Sigmanl Department of Biological Chemistry School of Medicine and Department of Chemistry and Biochemistry, Molecular Biology Institute, University of California, Los Angeles, California 90024

Received August 21, 1990, and in revised form December 19, 1990

Aldose reductase is an NADPH-dependent enzyme which catalyzes the reduction of glucose to sorbitol. Spe- cific potent inhibitors of aldose reductase are of potential pharmacological use because elevated levels of sorbitol produced by this enzyme in lens, peripheral nerve, retina, and renal glomeruli may be responsible for the patho- genesis associated with chronic diabetes. These inhibitors could also serve as probes of the mechanism of action of aldose reductase. anti-Oximes of aromatic aldehydes (e.g., benzaldoxime and 4-fluorobenzaldoxime) have proved to be effective inhibitors of aldose reductase ri- valing pharmacological agents currently used to inhibit this enzyme in viva. The kinetic patterns of inhibition in which benzyl alcohol is used as the oxidizable substrate suggest that the inhibition is due to the formation of a stable ternary complex composed of aldose reductase, NADP+, and the anti-oxime. Analogous ternary complexes are formed at the active site of horse liver alcohol dehydrogenase which is also inhibited by anti-oximes of efficient substrates. 0 isw Academic PPZSS, IIIC.

Aldose reductase is an NADPH-dependent enzyme that is widely distributed in mammalian tissues (l-4). Al- though the enzyme is principally responsible for the con- version of glucose to sorbitol in vivo, aldose reductase catalyzes the reversible oxidation/reduction of a variety of simple alcohols/aldehydes including ethanol/acetal- dehyde, n-butanolln-butyraldehyde, isobutanol/isobu- tyraldehyde, and benzyl alcohol/benzaldehyde (5). This broad substrate specificity is similar to that of vertebrate liver NAD+-dependent alcohol dehydrogenase, an enzyme which has been thoroughly studied both mechanistically and structurally (6).

r To whom correspondence should be addressed at the Molecular Bi- ology Institute.

596

Aldose reductase, by increasing tissue levels of sorbitol, may play a central role in the development of a variety of complications in chronic diabetics (7). As a result, sub- stantial effort has been expended in the development of specific inhibitors of aldose reductase (8). Several effective inhibitors have been discovered, but their mechanisms of action have not been convincingly characterized (2, 8).

Because of the similarities of the substrate specificity of aldose reductase and liver alcohol dehydrogenase, we have investigated whether analogs of potent inhibitors of alcohol dehydrogenase might serve as effective inhibitors of aldose reductase. Since we had previously demonstrated that anti-oximes of specific aldehyde substrates, not their syn-isomers, were powerful inhibitors (8&s, 10m7 to lo-’ M) of alcohol dehydrogenase (9), we wondered if comparable derivatives would be potent inhibitors of aldose reductase. In this communication, we report that anti-oximes of aromatic aldehydes are potent inhibitors of aldose reductase, whereas the syn-analogs are not potent inhibitors.

EXPERIMENTAL PROCEDURES

Materials and Methods Chemicals and reagents were obtained from the following sources. (i)

Sigma: 3-acetylpyridine adenine dinucleotide phosphate (oxidized form, 98% purity), reduced nicotinamide adenine dinucleotide phosphate, 2- hydroxybenzaldoxime, 2-nitrobenzaldoxime, 4-dimethylaminobenzal- dehyde, acetaldoxime, propionaldoxime, butyraldoxime, acetone oxime, pyrazole; (ii) Aldrich: syn-benzaldoxime, syn-4-fluorobenzaldoxime, 4- nitrohenzaldoxime, cyclopropanone oxime, cyclopentanone oxime, cy- clohexanone oxime, 1,2-cyclohexanedione dioxime, indazole, rt-nitroindazole, 5nitroindazole; (iii) Lancaster Synthesis: benzamide, cinnamaldoxime, 4-nitrocinnamaldehyde, isobutyraldoxime, diphenyl ketooxime; (iv) Kodak: isobutyric acid; (v) Fisher Scientific: hydroxyl- amine, EDTA. Tolrestat was a gift from Wyeth-Ayerst; sorbinil from Pfizer; and glucon-(1,5)-hydroximolactone was a gift from Professor A. T. Vasella.

Aldose reductase. Bovine lenses were purchased from Pel-Freez (P.O. Box 68, Rogers, AR 72756) and stored frozen until needed for preparation of partially purified aldose reductases (5). Enzyme preparation was carried out in a 4°C cold room. Bovine lens (40 g) was homogenized in 3

ooo3-9&X/91 $3.00 Copyright 0 1991 by Academic Press, Inc.

All rights of reproduction in any form reserved.

MECHANISM-BASED INHIBITORS OF BOVINE ALDOSE REDUCTASE 597

A

I anfi-bentaldoxime

Time (min)

an/i-Fluorobenxoldoxime B

9s-ET-- Time (min)

FIG. 1. Reverse phase HPLC separations of benzaldoxime and 4- fluorobenzaldoxime. (A) Benzaldoxime. The bottom chromatogram of commercial syn-benzaldoxime contains 95% syn-isomer, retention time 22.8 min, and 5% anti-isomer, retention time 26.9 min. The upper chromatogram is of HPLC-purified anti-benzaldoxime, retention time 26.9 min. (B) 4-Fluorobenzaldoxime. The bottom chromatogram of commercial syn-4-fluorobenzaldoxime contains 78% syn-isomer, retention time 15.4 min, and 20% anti-isomer, retention time 18.1 min. The upper chromatogram is of HPLC-purified anti-4-fluorobenzaldoxime, retention time 18.1 min.

vol of cold distilled water in a glass homogenizer. The homogenized material was centrifuged at 10,OOOg for 15 min at 4”C, insoluble material was discarded, and saturated ammonium sulfate was added to the supernatant to 40% saturation. After standing for 15 min with occasional stirring to allow for precipitation, the solution was centrifuged at 10,OOOg for 15 min at 4°C. Ammonium sulfate was added to the supernatant to 50% saturation. After standing for 15 min with occasional stirring, it was centrifuged once again. The supernatant was then brought to 75% saturation by further ammonium sulfate addition. Precipitate was re- moved by centrifugation. All the precipitates and supernatants were assayed for aldose reductase activity, which was highest in the precipitate of the 50% ammonium sulfate fraction. This fraction was used in subsequent studies.

Horse liver alcohol dehydrogenase and sorbitol dehydrogenase. Horse liver alcohol dehydrogenase was purchased from Sigma. The crystalline suspension was centrifuged and dissolved in 0.05 M phosphate buffer, pH 7.0, and dialyzed against six lOOO-ml portions of 0.05 M phosphate buffer at pH 7.0 at 4°C for 72 h. The dialyzed enzyme solution was stored at 4°C (10). Sorbitol dehydrogenase (sheep liver) was purchased from Sigma, and dissolved in water (1.65 units/ml) for use without any further treatment.

HPLC. The HPLC system used consisted of two Waters M-45 solvent delivery systems, a Lambda-Max 481 LC spectrophotometer, a Waters automated gradient controller, a Shimadzu C-R3A recorder, and a Vydac C-18 semipreparative column. Solvents were all filtered and degassed before use.

NMR. The proton NMR analysis was performed on a Bruker AM 360/Wb equipped with a data system (Aspect 3000 with 768 kbyte mem- ory) and a plotter (Hewlett Packard Model 7550A). Five-millimeter sample tubes were used, predried samples were dissolved in deuterated chloroform or methanol (1.0 mg sample in 0.5 ml solvent) with TMS as reference.

Enzyme Assay

Aldose reductase. The enzyme activity was measured using the fluorescence assay based on the reduction of 3-acetyl-NADP+ by benzyl alcohol (11). The fluorescence measurement was performed with a Aminco-Bowman spectrophotofluorometer with 365-nm excitation, and emission wavelength of 465 nm (NADPH) or 480 nm (3-acetyl-NADPH). All enzymatic reactions were carried out at 25°C.

1. Benzyl alcohol as substrate: A typical assay mixture of aldose reductase using benzyl alcohol as substrate contained 10 pM 3-acetyl- NADP+, 0.05 M potassium phosphate at pH 8.25, and 5 mM benzyl alcohol as well as enzyme. The reaction mixture was incubated for 2 min and then the assay was initiated by the addition of benzyl alcohol. The increase in fluorescence of 3-acetyl-NADPH generated in the reaction was followed continuously. K,,, of aldose reductase for benzyl alcohol was 0.5 mM.

2. D-Glucose as substrate: The assay mixtures for the reduction of D-glucose typically contained 2.5 PM NADPH, 0.05 M potassium phosphate buffer at pH 7.0, enzyme solution, and 10 mM D-glucose in a 3- ml cuvette. Reduction of D-glucose was followed continuously by the decrease of fluorescence of NADPH. The nonenzymatic decay of NADPH was subtracted by adding to a reference cuvette all components except glucose. K,,, of aldose reductase for glucose was 6 mM.

Alcohol dehydrogenase. The activity of alcohol dehydrogenase was determined by measuring the fluorescence increase of NADH generated in a reaction mixture of 1 mM NAD+, 0.05 M phosphate buffer at pH 7.0, alcohol dehydrogenase, and inhibitor. Assay mixtures were incubated at room temperature for 10 min prior to initiation by the addition of ethanol (3 mM).

Sorbitol dehydrogenase. The activity of sorbitol dehydrogenase was determined by measuring the fluorescence increase of NADH generated in a reaction mixture containing 0.5 mM NAD*, 0.05 M phosphate buffer at pH 8.25, sorbitol dehydrogenase, and inhibitor (12). Solutions were incubated at room temperature for IO min prior to initiation by sorbitol (0.5 mM).

Preparation of Oximes anti-Benzaldoxime and anti+luorobenzaldoxime. anti-Benzaldox-

ime and anti-4-fluorobenzaldoxime were isolated from commercial syn- benzaldoxime and syn-4+luorobenzaldoxime, respectively, using a Vydac C-18 semipreparative column eluting with acetonitrile/water. Fractions of anti-benzaldoxime and anti-4-fluorobenzaldoxime were then concen- trated by a rotatory evaporator. Their concentrations were determined by absorption spectroscopy. The anti-aromatic oximes were kept frozen until needed. The syn-forms of the respective aromatic oximes were prepared in the same way as the anti-oximes.

598 SHEN AND SIGMAN

A

h

(a)

Assignments.

a. 7.35 PPM b. 7.57 PPM

~ c. 8.07 PPM

I I I I I I I I I I I I I

0.2 0.0 7.0 7.6 7.4 7.2 PPM

C

! Assignments : a. 7.10 PPM b. 7.62 PPM c. 0.07 PPM

I I,,,,,,,,, *I,

0.2 0.0 7.0 7.6 7.4 7.2 zo PPM

i..., llil 0.2 0.0 7.0 7.6

PPY

D

(4

Assignments: a. 7.40 ppm 6. 7.97 ppm c. 7.3Oppm

(Cl

‘., Assignments : a. 7.15 PPM 6. 8.08 PPM c. 7.30 PPM

,181 1 1 113 1~1~ 1 a 10 1 0.2 0.0 7.0 7.6 7.4 7.2 7.0 6.0 6.6

PPM

FIG. 2. 360 MHz proton NMR spectra of benzaldoxime and 4-fluorobensaldoxime. (A) The proton spectrum of commercial syn-benzaldoxime in deuterated methanol. (B) The proton spectrum of HPLC-purified anti-bensaldoxime in deuterated methanol. (C) The proton spectrum of commercial syn-4-fluorobensaldoxime in deuterated methanol. (D) The proton spectrum of HPLC-purified anti-4-fkorobensaldoxime in deuterated methanol.

4-Dimethylamino-benzaldonime and 4-nitrocinnamaldonime. These oximes were prepared by standard methods. A single spot for each oxime was detected under 254-nm uv light and with the CuC& spot test for oximes (13).

RESULTS

Purification and NMR Spectra of Oximes

The HPLC chromatograms of commercial syn-benzaldoxime show that this product contains 95% syn-isomer with retention time 22.8 min and 5% anti-isomer with retention time 26.9 min (Fig. 1A). After rechromatogra- phy, isolation of pure anti- and syn-isomers of benzaldoxime is possible (see Experimental Procedures). The bottom chromatogram shows the separation of anti- and syn-isomers of the commercial sample while the upper chromatogram shows the separation of the HPLC-purified anti-isomer of benzaldoxime. The HPLC chromatograms of commercial syn-4-fluorobenzaldoxime show that this product contains 78% syn-isomer with retention time 15.4 min and 20% anti-isomer with retention time 18.1 min (Fig. 1B). The proton NMR spectra of anti- and syn-isomers of benzaldoxime and 4-fluorobenzaldoxime in deuterated methanol are presented in Fig. 2.

Assaying Aldose Reductase with Benzyl Alcohol and 3-Acetyl-NADPf

Lens aldose reductase catalyzes the reversible oxidation/reduction of a number of alcohols/aldehydes using NADP+/NADPH as a coenzyme. The oxidation of benzyl alcohol by the oxidized form of 3-acetylpyridine adenine dinucleotide phosphate (3-acetyl-NADP+) was used to screen the inhibitory potency of oxime inhibitors because it provides a sensitive and reliable assay of aldose reductase activity (11). In addition to the low Michaelis constant for benzyl alcohol, an additional reason for screening the inhibitory action of oximes with this assay is that oxime inhibitors of alcohol dehydrogenase form stable ternary complexes composed of the enzyme, oxidized coenzyme, and oxime (9). By assaying the oxidation of an alcohol, oxidized coenzyme is available in excess to drive the formation of the anticipated inhibitory complex.

Benzaldoximes. Table I records the inhibitory eth- ciency of benzaldoxime and closely related compounds in the benzyl alcohol/3-acetyl-NADP+ reaction of aldose reductase. For comparison, the inhibitory potencies of tolrestat and sorbinil were determined under comparable


TABLE I

Inhibition by Aromatic Derivatives of Aldose Reductase

% Inhibition

Inhibitor Concentration

(M)

0.5 mM

benzyl alcohol

5mM benzyl alcohol

anti-Benzaldoxime (HPLC purified)

syn-Benzaldoxime 5 x lo-@ 64 27

(HPLC purified) Benzoic acid Benzamide anti-4-Fluorobenzaldoxime

5 x 1o-5 50 13 1 x 1o-4 0 0 1 x 1om4 0 0

(HPLC purified) 5 x lo-* 51 14 Tolrestat 5 x 1om9 41 43 Sorbinil 5 x 10-r 48 42

Note. Aldose reductase activity was measured fluorimetrically by the reduction of 3-acetyl-NADP+ with benzyl alcohol. The assay mixture containing 10 pM 3-acetyl-NADP+, enzyme, 0.05 M phosphate buffer (pH 8.25), and inhibitor was incubated for 2 min at 25’C. The reaction was initiated by the addition of benzyl alcohol (0.5 mM and 5 mM, respectively).

assay conditions. anti-Benzaldoxime is lOOO-fold more efficient than the syn-isomer as an inhibitor. anti-Fluoro- benzaldoxime is equivalent to anti-benzaldoxime. Both anti-geometric isomers are more effective inhibitors than sorbinil but less effective than tolrestat, two compounds which have been used to decrease sorbitol accumulation in vivo. Since benzoic acid and benzamide are not inhibitors of aldose reductase, the inhibitory action of benzaldoxime is attributed to the oxime functionality rather than the benzene ring.

Aliphatic oximes. Aliphatic oximes are ineffective inhibitors of aldose reductase (Table II). The syn- and anti- isomers of these molecules have not been resolved prior to assay because preparations typically contain roughly 50% of each isomer. Since the inhibition constant of the anti-isomer of n-butyraldoxime is approximately 9 nM for liver alcohol dehydrogenase, the ineffectiveness of these aliphatic oximes as inhibitors of aldose reductase is sur- prising given the enzyme’s efficient reduction of the parent aldehyde. Equally unexpected is the inability of pyrazole, also a potent inhibitor of alcohol dehydrogenase, to block aldose reductase (Table II) (14). The pyrazole derivatives indazole (I), 4-nitroindazole (II), and 5nitroindazole (III), which might be expected to share the specificity determinants of benzaldoxime, are also ineffective inhibitors.

Derivatives of cinnamaldoxime and benzaldoxime. 4 Dimethylamino-cinnamaldoxime is an efficient inhibitor of horse liver alcohol dehydrogenase (Ki, 5 nM) and forms a stable ternary complex with enzyme and NAD+ (9).

Spectroscopic studies have suggested that the anti-oxime of this derivative is much more stable than the syn-isomer. 4-Dimethylamino-cinnamaldoxime, 4-nitrocinnamaldoxime, and the unsubstituted cinnamaldoxime have proven to be effective inhibitors of aldose reductase as well (Table III). The derivatives of benzaldoxime such as 2-hydroxybenzaldoxime, 2-nitrobenzaldoxime, 4-nitrobenzaldoxime, and 4-dimethylaminobenzaldoxime have also proven to be potent inhibitors of aldose reductase. However, since the geometric isomers could not be resolved prior to assay, the inhibition constants could not be accurately determined.

Oximes of aldoses. The oximes of D-glucose, D-fructose, and glucon-( 15) -hydroximolactone were examined as inhibitors of aldose reductase (Table IV). These derivatives are not inhibitors despite their close relationships with the two substrates of the enzyme. The oximes of D- glucose and D-fructose are acyclic but the hydroximolactone is not. Simple unfunctionalized nonaromatic derivatives cyclic systems are not sufficient for inhibition as evidenced by the minimal inhibition of cyclopropanecarboxaldoxime, cyclopentanecarboxaldoxime, cyclohexanecarboxaldoxime, and 1,2-cyclohexanedione dioxime (Table IV).

Inhibition patterns of anti-benzaldoxime. The mechanism of inhibition of alcohol dehydrogenase by anti-oximes involves the formation of a stable ternary complex composed of enzyme, NAD+, and anti-oxime. This hypothesis is supported by the competitive inhibition of the oxime relative to the oxidizable substrate, ethanol. To test for an analogous mode of inhibition of aldose reductase, the inhibition pattern of anti-benzaldoxime was determined in the direction of benzyl alcohol oxidation by

TABLE II

Inhibition by Aliphatic Oximes and Pyrazole Derivatives of Aldose Reductase


(M) % Inhibition

Acetaldoxime Propionaldoxime Butyraldoxime Isobutyraldoxime Isobutyric acid Acetone Oxime Indazole 4-Nitroindazole 5Nitroindazole EDTA Pyrazole l,lO-Phenanthroline

5 x 1om4 7 5 x lo-” 13 5 x 10-a 93 5 x 1o-5 81 1 x 1o-3 0 5 x 10-b 40

2.5 X 1Om5 0 1 x 10-e 0 1 x 10-e 36 5 x 1om2 0 1 x 10-a 0 1 x 10-a 29

Note. Aldose reductase activity was measured fluorimetrically by the reduction of 3-acetyl-NADP+ with benzyl alcohol. The assay mixture containing 10 HAM 3-acetyl-NADP+, enzyme, 0.05 M phosphate buffer (pH 8.25), and inhibitor was incubated for 2 min at 25’C. The reaction was initiated by the addition of benzyl alcohol (0.5 mM).

600 SHEN AND SIGMAN

H I / XD 02N \

I /

varying the alcohol concentration at a fixed 3-acetyl- NADP+ concentration. Although a slight intercept (i.e., displacement on the l/u axis) is apparent in the double reciprocal analysis, the slope effect predominates (Fig. 3A). The increased intercept on the l/v axis can be attributed to the binding of the inhibitor to the enzyme-3- acetyl-NADPH complex (15). In forming this complex, the oxime adds subsequent to the binding of benzyl alcohol and therefore can never be blocked from binding even with an infinite substrate concentration (15). When the kinetic data is collected by varying the 3-acetyl-NADP+ concentration at fixed benzyl alcohol, parallel lines, di- agnostic of uncompetitive inhibition, are observed (Fig. 3B). These kinetic patterns are consistent with the formation of a stable ternary complex composed of enzyme, 3-acetyl-NADP+, and oxime.

Assaying Aldose Reductase with D-Glucose and NADPH

Since D-glucose is the biologically significant substrate for aldose reductase in the polyol pathway, inhibitors discovered using the benzyl alcohol-dependent assay were also evaluated by studying the reduction of glucose by NADPH. This assay is less sensitive than that which uti- lizes 3-acetyl-NADP+ and benzyl alcohol. An additional

TABLE III

Inhibition by Oximes of Substituted Cinnamaldehydes and Benzaldehydes of Aldose Reductase


(M) % Inhibition

2-Hydroxybenzaldoxime 2-Nitrobenzaldoxime I-Nitrobenzaldoxime 4-Dimethylaminobenzaldoxime

(synthesized) Cinnamaldoxime 4-Nitrocinnamaldoxime

(synthesized) anti-4-Dimethylaminocinna-

maldoxime (synthesized)

5 x 10-T 11 5 x 10-b 68 5 x 10-s 75

5 x 10-T 47 5 x 10-s 57

5 x 10-T 33

5 x 10-s 38

Note. Aldose reductase activity was assayed by the method described in Table II.

complexity arises because free NADP+, as a product in the initial velocity measurements, is not present. As a result, the level of inhibition will depend on the steady- state concentration of E-NADP+ in the reaction scheme, and direct comparisons of the inhibition constants determined by the two assays are not possible. The inhibitory potencies of anti-benzaldoxime, tolrestat, and sorbinil are compared in Table V. As anticipated, the anti- benzaldoxime is a more efficient inhibitor in the oxidation of benzyl alcohol than in the reduction of glucose. The magnitude of the inhibition constant is more assay dependent for anti-benzaldoxime than either tolrestat or sorbinil.

Inhibitory Specificity of Benzaldorimes

If an inhibitor of aldose reductase is to be useful for in viva applications, it must be specific for that enzyme; therefore, the inhibitory potencies of anti-benzaldoxime for aldose reductase, sorbitol dehydrogenase, and alcohol dehydrogenase were compared (Table VI). Sorbitol dehydrogenase shows no sensitivity to the inhibitor, and alcohol dehydrogenase is 20-fold less effectively inhibited than aldose reductase.

DISCUSSION

The Michaelis constant for glucose in the aldose reductase-catalyzed reaction is 6 mM, approximately twice

TABLE IV

Inhibition by Oximes of Cyclic Aldehydes and Sugars of Aidose Reductase


(M) % Inhibition

Cyclopropanecarboxaldoxime 5 x 1o-6 13 Cyclopentanecarboxaldoxime 5 x w4 27 Cyclohexanecarboxaldoxime 5 x lo-’ 33 1,2-Cyclohexanedione dioxime 5 x lo-’ 83 Diphenyl ketoxime 5 x 10-s 83 D-Glucose oxime 1 x 1o-3 0 D-Fructose oxime 1 x 10-a 0 Glucon-(1,5)-hydroximolactone 5 x lo-’ 0

Note. Adlose reductase activity was assayed by the method described in Table II.


-I -3 -2 -I 0 I 2 3 4 5 -4 -3 -2 -I 0 I 2 3

I/ mM Iknz.-OH l/uM acctyl NADP+

FIG. 3. Inhibition patterns of aldose reductase with anti-benzaldoxime. (A) Competitive inhibition of inhibitor (anti-benzaldoxime) with substrate (benzyl alcohol). Aldose reductase was assayed with various concentrations of benzyl alcohol, and a fixed concentration of 3-acetyl-NADP+ (10 HIM) in 0.05 M phosphate buffer (pH 8.25) at 25°C in the absence or presence of inhibitor. No inhibitor (m); 0.33 MM anti-benzaldoxime (+). (B) Noncompetitive inhibition of inhibitor (anti-benzaldoxime) with coenzyme (3-acetyl-NADP+). Aldose reductase was assayed with fixed concentration of benzyl alcohol (5 mM) and various concentrations of 3-acetyl-NADP+ in 0.05 M phosphate buffer (pH 8.25) at 25°C in the absence or presence of inhibitor. No inhibitor (Kl); 0.33 pM anti-benzaldoxime (+).

the normal fasting blood glucose level. This high K, re- The strongest evidence for the specificity of the inter- sults in increased sorbitol production at the elevated glu- action between aromatic oximes and aldose reductase is cose concentrations characteristic of hyperglycemia, es- the different inhibition properties of the syn- and anti- pecially in tissues that are not insulin dependent in geometric isomers of benzaldoxime. The syn-isomer is a glucose transport such as kidney, lens, and retina. Sorbitol poor inhibitor while the anti-isomer is a strong inhibitor. accumulation, by causing osmotic imbalance in these tis- Liver alcohol dehydrogenase is also inhibited by anti-ali- sues, may be responsible for many complications of phatic oximes prepared from specific aldehyde substrates. chronic diabetes. Specific potent inhibitors of aldose re- In the case of alcohol dehydrogenase, the inhibition can ductase are of interest because of their potential phar- be attributed to the formation of a stable ternary complex macological uses. In addition, if their mode of action is composed of enzyme, NAD+, and anti-oxime. The tight known, these inhibitors would provide a new approach binding in the ternary complex is a reflection of three for investigating the detailed mechanism of the action of specific interactions: the structure of the parent aldehyde, aldose reductase and comparing it to other dehydrogen- the coordination of the nitrogen to zinc ion bound at the ases. Recent studies have revealed that human placental active site (17), and the nucleophilic addition of oxime aldose reductase shows sequence homology not only to oxygen to the nicotinamide ring (9) (Fig. 4A). Presumably, human liver aldehyde reductase, but also to 2,5-diketo- the stereochemistry of the active site prohibits the syn- gluconic acid reductase from corynebacterium, frog p- isomer to participate in the latter two interactions si- crystallin, and bovine lung prostaglandin F synthase (16). multaneously.

TABLE V

Comparison of Inhibition of Aldose Reductase Using Different Substrates

Substrate Coenzyme KW2

bM) Inhibitor Concentration

(M) % Inhibition

D-Glucose NADPH

Benzyl 3-acetyl- Alcohol NADP+

6

0.5

anti-Benzaldoxime 1 x 10-s 14 Tolrestat 2 x 10-s 80 Sorbinil 1 x 10-s 20 anti-Benzaldoxime 5 x 10-s 64 Tolrestat 5 x 10-9 47 Sorbinil 5 x 10-T 48

Note. Aldose reductase activity was measured Auorimetrically by the decrease of NADPH using D-glucose as substrate. The assay mixture containing 2.5 pM NADPH, 0.05 M phosphate buffer at pH 7.0, enzyme, and inhibitor was incubated for 10 min at 25°C. The reaction was initiated by the addition of D-ghCOSe. Aldose reductase activity using benzyl alcohol as substrate was assayed by the method described in Table I.

602 SHEN AND SIGMAN

TABLE VI

Inhibitory Specificity of Benzaldoximes on Enzymes

Aldose reductase

Concentration (PM) % Inhibition

Alcohol dehydrogenase


anti-Benzaldoxime 0.05 64 0.82 50 syn-Benzaldoxime 50 50 28.5 50 Tolrestat 0.005 48 20 24 Sorbinil 0.5 48 200 0

Sorbitol dehydrogenase


500 9 N/A N/A

100 48 1000 0

Note. Aldose reductase activity was assayed by the method described in Table I. Alcohol dehydrogenase activity was assayed by measuring the fluorescence increase of NADH generated in a reaction mixture of 1 mM NAD+, 0.05 M phosphate buffer (pH 7.0), enzyme, and inhibitor. Assay mixture was incubated for 10 min at 25’C prior to initiation by the addition of ethanol. Sorbitol dehydrogenase activity was assayed by measuring the fluorescence increase in NADH generated in a reaction of 0.5 mM NAD+, 0.05 M phosphate buffer (pH 8.25), enzyme, and inhibitor. Assay mixture was incubated for 10 min at 25°C prior to initiation by the addition of sorbitol. N/A, not available.

At present, the kinetic data on the inhibition of aldose reductase by the anti-benzaldoxime is consistent with the formation of a stable ternary complex; however, the specific interactions at the active site which contribute to this stability are not known. No metal ion has been demonstrated at the active site, nor has the enzyme proven to be sensitive to l,lO-phenanthroline, an effective inhibitor of horse liver alcohol dehydrogenase (10). The Lewis

A COENZYME

B COENZYME

FIG. 4. (A) Model for the inactive ternary complex formed with anti- oxime, NAD+, and horse liver alcohol dehydrogenase. (B) Postulated model of ternary complex formed with anti-oxime, NADP+, and aldose reductase. Lewis acid catalyst indicated.

acid function, provided by the metal ion at the active site of alcohol dehydrogenase, may be accomplished in aldose reductase by a general acid catalyst (Fig. 4B). Purified enzyme will be required for spectroscopic experiments in order to determine if the nucleophilic addition of the oxime oxygen to the nicotinamide ring is a source of stability of the ternary complex.

More potent inhibitors of aldose reductase than the anti-benzaldoxime have been identified through screening programs. However, the mode of action of the most effective inhibitors that have been identified (e.g., sorbinil, tolrestat, and Ono-2235) (17) are unknown. Nevertheless, judging from the many nucleophiles which can form addition compounds at the active sites of NAD+ and NADP+-dependent dehydrogenases, these inhibitors may be acting via a method analogous to that of the anti-oximes. Once the mechanism of inhibition of aldose reductase by anti-benzaldoxime has been identified, it may provide a useful clue for investigating this hypothesis. In this context, it is of interest to note that Ono-2235 is a cinnamaldehyde derivative (18), and oximes of cinnamaldehyde and its derivatives have proven to be effective inhibitors of aldose reductase.

ACKNOWLEDGMENTS

This research has been supported by a United States Public Health Service grant (HD 21437). We thank Professor Robert Delange of the Department of Biological Chemistry for allowing us to use the Aminco- Bowman spectrophotofluorometer, and Professor A. T. Vasella of the University of Zurich for providing glucon-(1,5)-hydroximolactone.

REFERENCES

1. Cohen, M. P. (1986) The Polyol Paradigm and Complications of Diabetes, Springer-Verlag, New York.

2. Dvornik, D. (1987) Aldose Reductase Inhibition, McGraw-Hill, New York.

3. Humber, L. G. (1987) Prog. Med. Chem. 24,299-341.


4.

5.

6.

7.

8.

9.

10.

Larson, E. R., Lipinski, C. A., and Sarges, R. (1988) Med. Res. Reu. 8.159-181.

Hayman, S., and Kinoshita, J. H. (1965) J. Biol. Chem. 240,877- 882.

Branden, C. I., Jornvall, H., Eklund, H., and Furguren, D. (1975) The Enzymes (Boyer, P. D., Ed.), Vol. 11, pp. 1033189, Academic Press, San Diego. Dvornik, D., Simard-Duquesne, N., Krami, M., Sestanj, K., Gabbay, K. H., Kinoshita, J. H., Varma, S. D., and Merola, L. D. (1973) Science 182,1146-1148.

Butera, J., Bagli, J., Doubleday, W., Humber, L., Treasuyrwala, A., Loughney, D., Sestanj, K., Millen, J., and Sredy, J. (1989) J. Med. Chem. 32,757-765. Sigman, D. S., Frolich, M., and Anderson, R. E. (1982) Eur. J. Biochem. 126,523-529. Sigman, D. S. (1967) J. Biol. Chem. 242, 3815-3824.

11.

12.

13.

14.

15.

16.

17.

18.

Griffin, B. W., and McNatt, L. G. (1986) Arch. Biochem. Biophys. 246,75-81.

Jeffery, J., and Jornvall, H. (1988) Adu. Enzymol. 61,47-106.

Hranisavljevic-Jakovljevic, M., Pejkovic-Tadic, I., and Stojiljkovic, A. (1963) J. Chromatography 12, 70-73.

Yonetani, T. (1963) Biochem. Z. 338, 300-316.

Sigman, D. S., and Glazer, A. N. (1972) J. Biol. Chem. 247, 334- 341.

Bohren, K. M., Bullock, B., Wermuth, B., and Gabbay, K. H. (1989) J. Biol. Chem. 264,9547-9551.

Dunn, M. F., Bielman, J. F., and Branlant, G. (1975) Biochemistry 14,3176-3X32.

Kikkawa, R., Hatanaka, I., Yasuda, H., Kobayashi, N., Shigeta, Y., Terashima, H., Morimura, T., and Tsuboshima, M. (1983) Diabe- tologia 24, 290-292.

Documents

New inhibitors of aldose reductase: Anti-oximes of aromatic aldehydes