- Email: [email protected]
Effects of two new aldose reductase inhibitors, AL-1567 and AL-1576, in diabetic rats
Effects
of Two New Aldose
Brenda Walker
Reductase Inhibitors, in Diabetic Rats
Griffin, Loretta
G. McNatt,
Michael
AL-1567
L. Chandler,
and AL-1576,
and Billie M. York
Two new potent aldose reductase inhibitors, AL-l 567 (oL-spiro(2-fluoro-9H-fluoren-9,4’-imidazoiidine)-2’,5’-dione) and AL-l 576 ~spiro-(2.7-difluoro-9Kfluoren-9,4’-imidarolidine~2’,5’-dione), have been characterized with respect to in vitro activity toward rat lens and human placental aldose reductase and in vivo activity in uncontrolled, severely diabetic rats dosed acutely with the compounds. The IC, values for inhibition of rat lens aldose reductase are 2.7 x lo-’ mol/L for AL-1567 and 8.5 x 10-O mol/L for AL-1576; very similar IC, values were measured for each compound with the human placental enzyme. When the compounds were administered orally once per day to 3-week diabetic rats for a period of eight days, the ED, values for normalization of lens sorbitol levels were 0.60 mg/kg for AL-1567 and 0.05 mg/kg for AL-l 576, and for normalization of sciatic nerve sorbitol levels, 0.22 mg/kg for AL-1 567 and 0.04 mg/kg for AL-l 576. Compared with published data on other aldose reductase inhibitors evaluated in very similar diabetic rat models, both compounds have unusually high activity in lens, and AL-l 576 appears to be the most active such compound in both lens and sciatic nerve reported thusfar. The evidence linking increased sorbitol pathway activity to diabetic complications, such as cataract and neuropathy in animal models, suggests that aldose reductase inhibitors will be useful therapeutic agents in human diabetics. 8 1987 by Grune & Stratton, inc.
A
LDOSE REDUCTASE catalyzes the NADPH-dependent reduction of various aldehydes and aldoses to the corresponding alcohols and polyols.’ By sequential action of this enzyme and sorbitol dehydrogenase, an NAD+-requiring enzyme, glucose can be converted to fructose. Under hyperglycemic conditions, when the glucose concentration of insulin-unresponsive tissues, such as lens, rises in synchrony with plasma glucose concentration, the flux of glucose through the sorbitol pathway relative to glycolysis increases significantly.‘** Although the operation of this pathway under a glucose stress is well-documented, the immediate or long-term benefits of increased sorbitol pathway activity have eluded strict definition to date. However, because polyols such as sorbitol do not readily move across cell membranes, either by diffusion or transport, chronically elevated polyol levels can generate a severe osmotic stress within the responsive cell. In the case of lens, for example, the resultant swelling and membrane damage initiate a well-studied sequence of changes that progress eventually to nuclear opacity.3 The most definitive evidence for the critical role of aldose reductase activity and elevated polyol levels in sugar-induced cataractogenesis is the dramatic inhibition of this type of cataract by potent aldose reductase inhibitors administered to diabetic or galactosemic animals.3-6 Studies of several aldose reductase inhibitors in animal models of diabetes have suggested that the therapeutic effects of these compounds are not limited to lens, but can also occur in selected tissues, such as retina, kidney, and peripheral nerve, which contain aldose reductase and which are known to be at greatest risk for diabetic complications.31s.6Administration of potent aldose reductase inhibitors to diabetic or galactosemic animals can inhibit polyol accumulation in these tissues as well as prevent some of the From Alcon Laboratories. Inc. Fort Worth, TX Address reprint requests to Brenda W. Gri&. PhD, Alcon Laboratories, Inc, MS R2-43, 6201 S Freeway, PO Box 6600. Fort Worth, TX 76115. Q I987 by Grune & Stratton, Inc. 0026-0495/87/3605-0014$03.00/0
486
later-developing pathophysiologic and morphologic changes which mimic the human disease process.3-6 However, since sorbitol does not accumulate in other tissues of the diabetic rat to the levels observed for lens, the question of severity of the sorbitol-induced osmotic stress in those tissues has been raised. This issue is complicated by the cellular localization of aldose reductase within different tissues’ and by the fact that sorbitol accumulation depends not only on aldose reductase activity, but also on sorbitol dehydrogenase activity. It seems likely that a sustained increase of glucose metabolism through the sorbitol pathway may also create a metabolic stress, involving abnormal ratios of oxidized and reduced pyridine nucleotides, among other changes, that eventually compromise cellular integrity. Although our understanding of these phenomena is incomplete, the available evidence indicates that several aldose reductase inhibitors ameliorate biochemical, functional, and ultrastructural alterations in various tissues of animal models of diabetes.’ These results have provided a strong rationale for human clinical trials of this class of compounds.’ In this report, the in vitro and in vivo activities of two new potent aldose reductase inhibitors, AL- 1567 and AL-l 576, are described. The activities of these compounds are compared with published activities of other aldose reductase inhibitors that have been evaluated in very similar diabetic rat models. These data indicate that the more active compound AL-1576 is considerably more active in vivo than other aldose reductase inhibitors for which analogous data have been published. MATERIALS
AND METHODS
Aldose reductase activity was determined by a standard spectrophotometric assay’ in a solution containing 0.1 mol/L potassium phosphate buffer, pH 6.8, 10 mmol/L D,L-glyceraldehyde, 0.2 mmol/L NADPH, the enzyme, and, where present, an aldose reductase inhibitor. The inhibitor concentration inhibiting the rate by 50% (IC, value) was determined by linear regression analysis of the linear portion of plots of % inhibition v log [inhibitor concentration]. Since these very active compounds have very small ICso values, which are comparable to the actual concentration of aldose reduc-
Metabolism, Vol 36. No 5 (May), 1987: pp 486-490
487
NEW ALDOSE REDUCTASE INHIBITORS
tase in the assay mixture,
the rate of the uninhibited
reaction
0
was
always adjusted to a constant value, in order to make valid comparisons of I(& values among a series of inhibitors. Two sources of aldose reductase were employed: a soluble rat lens preparation described elsewhere* and a hemoglobin-free preparation from human placenta. The latter was prepared by homogenizing the tissue in 50 mmol/L triethanolamine buffer, pH 7.0, containing 1 mmol/L each mercaptoethanol and EDTA. After centrifugation, hemoglobin was removed from the supernatant by batch extraction with phosphocellulose and elution with the same buffer. This material, containing about 10 mg protein/ml, was diluted tenfold and used for evaluating the activities of aldose reductase inhibitors by the fluorescence assay described. A very sensitive fluorescence assay of aldose reductase activity was recently characterized*; it has been shown” that this assay has distinct sensitivity and specificity advantages compared to the standard spectrophotometric assay. With this method, it was possible to measure aldose reductase specifically in the crude human placental preparation because L-hexonate dehydrogenase has a much larger K, value for the oxidized pyridine nucleotide analog employed. Also, the fluorescence assay of aldose reductase permits a more accurate determination of the inhibition constants of very active compounds such as AL-I 576. The experimental conditions for the fluorescence assay were 0.05 mol/L potassium phosphate buffer, 12 mmol/L benzyl alcohol, 12.5 pmol/L oxidized 3-acetylpyridine adenine nucleotide phosphate, and enzyme in I mL total volume at 25 “C. The rate of formation of the fluorescent reduced pyridine nucleotide product was monitored with a Perkin-Elmer 650-10s or a Perkin-Elmer MPF-66 spectrofluorometer, set for 365 nm excitation and 480 nm emission, with IO nm slitwidths.’ Male Sprague-Dawley rats (SASCO, Omaha), weighing 230 to 300 g, were made diabetic by intravenous injection of 50 mg/kg streptozotocin (Sigma Chemical Co, St Louis), freshly dissolved in 0.05 mol/L citric acid buffer, pH 4.5, containing 0.5% NaCI; this solution was kept on ice and used within two hours. Within 1 week of induction of diabetes, plasma glucose levels of all rats were analyzed, after sampling the blood by tail venupuncture into heparinized capillary tubes. The glucose assay employed was a standard coupled enzymatic assay (Sigma), based on glucose oxidase and peroxidase. Injected rats with nonfasting plasma glucose values between 375 and 525 mg/dL were selected for the study, and randomly assigned to groups, so that the mean plasma glucose value of each diabetic group was the same within t 10%. Throughout the study, rats had unrestricted access to water and Purina Rodent Chow. (Richmond, IN). About 3 weeks after induction of diabetes, daily dosing of study animals was initiated and continued for a total of eight days. The compounds. in an aqueous vehicle, were administered orally by means of stainless steel feeding needles. Approximately one dose period after the last dose, animals were killed; lenses and samples of sciatic nerves were carefully dissected, weighed, and immediately frozen. Protein-free extracts of the tissues were prepared by stanusing ZnSO, and Ba(OH),. Analysis of sorbitol in dard procedures,” the neutralized extracts was performed according to published procedures, I’ in which sorbitol dehydrogenase catalyzes the stoichiometric reduction of NAD+ by sorbitol to the fluorescent product NADH. The reaction mixture contained I.0 mL of 0.1 mol/L glycine buffer, pH 9.5, 5 mmol/L EDTA, 10 mmol/L MgC&, 2 umts of sorbitol dehydrogenase, 0.8 mg NAD+, and 0.5 mL of the original or diluted tissue extract; sample blanks were identical except that sorbitol dehydrogenase was omitted. The increase in Ruorescence after a 20-minute incubation was measured with a PerkinElmer 650-10s fluorometer, set for 365 nm excitation and 455 nm emission wavelengths, and compared with a set of identically treated sorbitol standards. Sorbitol dehydrogenase and NAD’ were supplied by Sigma.
HN
-3
&3 NH
PA
0%
AL-1567: X = H Fig 1. Chemical structures of AL-l 567 and AL-l 576.
AL-1576: X = F
Statistical analysis of the sorbitol data was performed with a Digital Equipment Corporation PDPl l/44 computer and RS/l software (Bolt Beranek and Newman, Inc, Cambridge, MA). A parametric t-test comparison of the various group means was employed, with a 95% confidence level (P c .05) selected as the criterion for a significant difference between two group means. The chemical structures of AL-l 567 and AL- 1576 are shown in Fig I. RESULTS
Spectrophotometric assay of rat lens aldose reductase activity, with D,L-glyceraldehyde and NADPEI as substrates, established that AL- 1567 inhibition was uncompetitive with respect to the aldehyde substrate when the reduced pyridine nucleotide concentration was saturating (Fig 2). The IC,, values for inhibition of rat lens aldose reductase determined spectrophotometrically were 4.2 x 10-s mol/L for AL-1567 and 2.4 x 10-s mol/L for AL-1576; by the fluorescence
0.04 -10
1
I
0
10 i /
20
30
40
50
I GLYCERALOEBYDE , mMOLAR)
Fig 2. Kinetics of AL-1567 inhibition of D,L-glyceraldehyde reduction by NADPH catalyzed by rat lens aldose reductase. All reaction mixtures contained rat lens aldose reductase. 0.2 mmol/L NADPH, and variable concentrations of D.L-glyceraldehyde, in 0.1 mol/L potasssium phosphate buffer pH 6.6,25X, and the indicated concentration of AL-1567: (0). no inhibitor: (Cl), 1.5 x lo-* mol/L; (A). 3 x lo-‘mol/L: to), 6 x lo-* mol/L: (m). 6 x lo-* mol/L.
GRIFFIN
488
assay of aldose reductase activity,’ IC,, values for AL- 1567 and AL-1576 were 2.7 x IO-* mol/L and 8.5 x 10m9mol/L, respectively. With the fluorescence assay, very similar IC,, values were measured for inhibition of human placental aldose reductase: 2.5 x lo-’ mol/L for AL- 1567 and 1.8 x lo-” mol/L for AL-l 576. These data indicate that AL-1 576 is the more active inhibitor and that each compound has approximately the same absolute activity toward the rat and human enzymes. Concerning the uncompetitive inhibition pattern for AL1567 (Fig 2), the same pattern has been reported for some aldose reductase inhibitors,‘2,‘3 while mixed uncompetitivenoncompetitive patterns have been reported for others.‘4-‘6 These observations appear to relate less to chemical structures than to the inhibition constants of the inhibitors. Indeed all of the very active aldose reductase inhibitors now under development for human therapeutic use fall into the category of reversible “tight-binding” inhibitors,” ie, the inhibition constants are equal to or even smaller than the concentration of aldose reductase required in the spectrophotometric assay of enzymatic activity. As has been discussed at length,” kinetic patterns obtained under such conditions will depend on the manner in which the experiment is performed and are not diagnostic of the molecular mechanism of inhibition. Because the concentration of aldose reductase required in the fluorescence assay is decreased below the inhibition constants of most very active inhibitors, our studies indicate that accurate analyses of inhibition kinetics can be performed by this assay.8*9 For evaluating the aldose reductase inhibitor activity of these compounds in vivo, an acutely dosed diabetic rat model was chosen as the most appropriate system; very similar models have been employed by most laboratories studying It was considered essential to this class of compounds. ‘3~‘4~‘8~‘9 control as closely as possible the mean plasma glucose values of the various diabetic groups, for the following reasons: (1) the apparent K, value of aldose reductase for glucose in assays of enzyme activity in vitro has been shown to be very large, as large as 500 mmol/L,*' which greatly exceeds peak plasma glucose levels even in uncontrolled diabetic animals and (2) the glucose level in diabetic rat lens, an insulinunresponsive tissue, correlates with the plasma glucose level.18 Therefore, the plasma glucose concentration is a very important factor controlling sorbitol production in a given
tissue; to eliminate the possibility of introducing bias into the data by variable rates of sorbitol production among the diabetic groups, the mean plasma glucose levels of these groups were the same within f 10% (Table 1). Table 1 also summarizes the mean sorbitol contents of lens and sciatic nerve, in absolute amounts and as percentages of the diabetic control group, for the drug-treated and control groups. For both aldose reductase inhibitors, a characteristic dependence of sorbitol level on dose was observed in each of the tissues. It should be noted that of the four doses of AL-l 576 included in the study, only the lowest dose did not decrease the sciatic nerve sorbitol level of treated diabetics to normal, or below in the case of the 0.50 mg/kg/d dose. From log dose-response plots of these data, ED,, values were estimated: for sciatic nerve sorbitol inhibition, 0.22 mg/kg/d for AL-1567 and 0.04 mg/kg/d for AL-1576; for lens sorbitol inhibition, 0.60 mg/kg/d for AL-l 567 and 0.05 mg/kg/d for AL- 1576. The data demonstrate that AL- 1576 is measurably more active than AL-1567 in both tissues of diabetic rats, qualitatively consistent with their in vitro aldose reductase inhibitor activities. In other studies, it has been shown that chronic dosing of these compounds at much higher dose levels to normal rats produces small but measurable inhibition of the very low normal sorbitol levels of lens and sciatic nerve. The sorbitol values for normal rat lens and sciatic nerve measured in this study (Table 1) agree with the published value of 0.15 nmol/mg for lens,18 and from 0.07 to 0.34 nmol/mg for sciatic nerve.‘8’2’We have confirmed (data not on the time course of sorbishown) previous observations **z*~ to1 accumulation in diabetic rat lens; the polyol attains its maximal value about 3 weeks after streptozotocin injection, and decreases slowly thereafter. Reported values of sorbitol content in diabetic rat tissues at comparable times after induction of diabetes are 22.518 to 50.0 nmol/mg23 for lens and 1.924 to 3.5 nmol/mg25 for sciatic nerve. In various studies, sorbitol has been measured by either gas chromatographic techniques23,24 or by the enzymatic fluorescence assay employed in this study’8.25;where the two methods have been directly compared,** excellent agreement has been found. Since the enzymatic method” has several distinct advantages, including higher sensitivity, it has become the assay of choice for studies of aldose reductase inhibitor activity in diabetic animal models.‘31’8*‘9
Table 1. Mean Sorbitol Values in Lenses and Nerves of Drug-Treated, DCX3e Drug
he/kg/d)
Plasma Glucose (m&L)
Lens Sorbitol”
(nmol/mal
Normal, and Diabetic Control Groups
Lens Sorbitol
Nerve Sorbitol*
Nerve Sorbitol
% Diabetic Control
(nmollmg)
I% Diabetic Control)
AL-1567
0.20
446
f 58
39.0
+ 10.8
97.5
1.01
+ 0.18
AL-1567
0.50
437
t 36
21.0
k 3.2
52.5
0.73
+ 0.13
37.6
AL-1567
1.50
436
+ 53
5.85
+ 1.30
14.6
0.46
? 0.10
23.7
AL-1 576
0.032
443
* 47
22.6
t 5.3
56.5
1.24
k 0.18
63.9
AL-l
576
0.0s
448
+ 47
17.9
* 5.4
44.8
0.33
* 0.17
17.0
AL-l
576
0.20
451
k 46
7.98
+ 4.0
20.0
0.21
+ 0.06
10.8
0.50
3.4
0.08
f 0.02
1.94
k 0.61
100.0
0.25
k 0.07
12.9
AL-1576
443
ir 40
1.35
k 0.24
None (controls)
-
448
+ 55
40.0
+ 17.0
Normals
-
136
k 16
0.40
* 0.11
Values
are given as mean f SD (n = 6).
*Sorbitol units, nmol/mg tissue wet weight.
ET AL
100.0 1.0
52.1
4.1
489
NEW ALDOSE REDUCTASE INHIBITORS
DISCUSSION
diabetic rats. Oral dosing of AL-1576 (0.50 mg/ kg/d) or AL-1567 (1.5 mg/kg/d) for 15 weeks completely prevented appearance of cataract, evaluated by opthalmoscopic examination; in this same time period, nuclear opacity developed in the diabetic control (untreated) group.6 Both compounds have also been shown to be potent anticataract agents in a more accelerated and severe sugar cataract model, weanling rats maintained on a 30% galactose diet.26 It should be noted that other aldose reductase inhibitors for which information has been published appear to be measurably less active in lens than in sciatic nerve, as determined from EDSo values for normalization of sorbitol levels. However, AL-1576 is uniquely active in lens of diabetic rats; the low EDSovalue, 0.05 mg/kg/d, is comparable to the activity in sciatic nerve. Aldose reductase activity, linked to elevated intracellular sorbitol levels, has been implicated in diabetic complications of other tissues of diabetic animal models based on: (1) the presence of the enzyme in the tissue, (2) elevation of tissue sorbitol levels well above normal values in diabetic animals, and (3) the ability of several aldose reductase inhibitors to normalize sorbitol, as well as improve other measures of tissue function. Uncontrolled diabetic rats develop neuronal dysfunction, which closely resembles that of diabetes of long duration in humans.27S28It has been shown that both AL1567 and AL-1576,6 as well as other aldose reductase inhibitors,‘4,‘9 can significantly improve the nerve conduction velocity deficiency of chronically diabetic rats. The present study confirms other reports that the early increase in sorbitol content of sciatic nerve of diabetic rats can be subsequently normalized by short-term systemic administration of potent aldose reductase inhibitors. Thus, the data support the idea that elevation of nerve sorbitol is an early biochemical event in the later-developing neuropathy of insulin-uncontrolled, severely diabetic rats, and by inference, in human diabetics. In summary, the data presented in this report on the activities of two new aldose reductase inhibitors, AL-1567 and AL-1576, indicate that in vivo, AL-1576 is the most potent member of this class of compound thus far described, with unusually high activity in diabetic rat lens. Moreover, severely
The activities of two very potent aldose reductase inhibitors, AL-1567 and AL-1576, have been described. From recently published data, it is possible to compare the in vivo activities of these compounds with other aldose reductase inhibitors in various stages of development and human clinical testing. Although there are some variations in study protocols for diabetic rat models used in various labs, the following conclusions can be drawn. Tolrestat,i3 ONO2235,19 and statill appear to have very similar activities in sciatic nerve of acutely dosed diabetic rats, with ED,, values for normalization of sorbitol ranging between 5 and 10 mg/kg/d. The lens activity of ONO-2235 is not known, but statil was reported to inhibit cataract in diabetic rats at 25 mg/kg/d.14 The ED,, value of tolrestat in diabetic rat lens was recently reported as 10 to 20 mg/kg/d.13 However, it appears that the model in that study” was somewhat less severe than ours, since (1) drug treatment was initiated either 11 days prior to, or on the day of, induction of diabetes by streptozotocin and (2) animals were killed after only 10 additional days of diabetes/drug treatment, when tissue sorbitol values have not yet attained their maximal values.22,23 The diabetic rat model used in our study is very similar to that in which sorbinil has been reported to have ED,, values of approximately 1.5 and 0.80 mg/kg/d for normalization of lens and sciatic nerve sorbitol levels, respectively.” In this model, AL- 1567 and AL-I 576 are measurably more active than sorbinil; based on relative EDSo values, AL-1576 is estimated to be about 20 times more active than sorbinil in sciatic nerve and about 30 times more active in lens. Comparisons of molar in vivo activities of these compounds support these conclusions. Although sorbinil has the lowest molecular weight (236), the molecular weight of AL-1576 (286) is greater by only 20%. which is comparable to the experimental error in the ED,, values. ‘The documented anticataract activities of potent aldose reductase inhibitors in galactosemic and diabetic animals support the osmotic stress hypothesis of sugar cataractogenesis.’ whereby hyperosmolality produced by chronically elevated intracellular polyol levels initiates tissue swelling and subsequent biochemical alterations that progress to lens opacity. That inhibition of the polyol pathway does, in fact, prevent cataracts has been demonstrated by chronic administration of both AL-I 567 and AL-1576 to uncontrolled,
these data and other evidence cited provide additional support for a critical role of increased polyol pathway activity in
the later-developing pathophysiologic alterations of various tissues of animal models of human diabetes.
REFERENCES
1. Hayman S, Kinoshita JH: Isolation and properties of lens aldose reductase. J Biol Chem 240:877-882, 1965 2. Gabbay KH: The sorbitol pathway and the complications of diabetes. N Engl J Med 288:831-836, 1973 3. Kador PF, Robison WG Jr, Kinoshita JH: The pharmacology of aldose reductase inhibitors. Annu Rev Pharmacol Toxicol25:691714,1985 4. Dvornik K, Simard-Duquesne N, Krami M. et al: Polyol accumulation in galactosemic and diabetic rats: Control by an aldose reductase inhibitor. Science 182:1146-l 148, 1973
5. Chandler ML, Shannon WA, DeSantis L: Prevention of retinal capillary basement membrane thickening in diabetic rats by aldose
reductase inhibitors. Invest Ophthalmol Vis Sci 25:159. 1984 (suppi) 6. Griffin BW, Chandler ML, DeSantis L: Prevention of diabetic cataract and neuropathy in rats by two new aldose reductase inhibitors. Invest Ophthalmol Vis Sci 25: 136, 1984 (suppl) 7. Young RJ, Ewing DJ, Clarke BF: A controlled trial of Sorbinil, an aldose reductase inhibitor, in chronic painful diabetic neuropathy. Diabetes 32:938-942, 1983 8. Griffin BW, McNatt LG: Characterization of the reduction of 3-acetylpyridine adenine dinucleotide phosphate by benzyl alcohol catalyzed by aldose reductase. Arch Biochem Biophys 246:75-81, 1986 9. Griffin BW, McNatt LG. York BM: Characterization of
490
aldose reductase activities from human and animal sources by a sensitive fluorescence assay, in Weiner H (ed): Third International Workshop on Enzymology of Carbonyl Metabolism: Aldehyde Dehydrogenase, Aldo-Keto Reductases and Alcohol Dehydrogenase. New York, Liss (in press) 10. Sherman WA, Stewart MA: Identification of sorbitol in mammalian nerve. Biochem Biophys Res Commun 22:492-497, 1966 11. Malone JI, Knox G, Benford S, et al: Red cell sorbitol. An indicator of diabetic control. Diabetes 29:861-864, 1980 12. Peterson MJ, Sarges R, Aldinger CE, et al: CP-45, 634: A novel aldose reductase inhibitor that inhibits polyol pathway activity in diabetic and galactosemic rats. Metabolism 28: 456-461, 1979 (suppl 1) 13. Simard-Duquesne N, Greselin E, Dubuc J, et al: The effects of a new aldose reductase inhibitor (Tolrestat) in galactosemic and diabetic rats. Metabolism 34: 885-892, 1985 14. Stribling D, Mirrlees DJ, Harrison HE, et al: Properties of ICI 128,436, a novel aldose reductase inhibitor, and its effects on diabetic complications in the rat. Metabolism 34:336-344, 1985 15. Terashima H, Hama K, Yamamoto R, et al: Effects of a new aldose reductase inhibitor on various tissues in vitro. J Pharmacol Exp Ther 229:226-230,1984 16. Kador PF, Goosey JD, Sharpless NE, et al: Stereospecific inhibition of aldose reductase. Eur J Med Chem 16:293-298, 1981 17. Williams JW, Morrison JF: The kinetics of reversible tightbinding inhibition. Methods Enzymol69:437-467, 1979 18. Malone JI, Leavengood H, Peterson MJ, et al: Red blood cell sorbitol as an indicator of polyol pathway activity. Inhibition by Sorbinil in insulin-dependent diabetic subjects. Diabetes 33:45-49, 1984 19. Kikkawa R, Hatanaka I, Yasuda H, et al: Effect of a new
GRIFFIN ET AL
aldose reductase inhibitor, (E)-3-carboxymethyl-5-[(2E)-Methyl3-phenylpropenylidenelrhodanine (ONO-2235) on peripheral nerve disorders in streptozotocin-diabetic rats. Diabetologia 24:290-292, 1983 20. Gabbay KH, Cathcart ES: Purification and immunologic identification of aldose reductases: Diabetes 23:460-468, 1974 21. Yue DK, Hanwell MA, Satchel1 PM, et al: The effect of aldose reductase inhibition on motor nerve condition velocity in diabetic rats: Diabetes 3 1:789-794, 1982 22. Hutton JC, Schofield PJ, Williams JF, et al: Sorbitol metabolism in the retina: Accumulation of pathway intermediates in streptozotccin induced diabetes in the rat. Aust J Exp Biol Med Sci 52:361-373, 1974 23. Kuriyama H, Sasaki K, Fukuda M: Studies on diabetic cataract in rats induced by streptozotocin II. Biochemical examinations of rat lenses in relation to cataract stages. Ophthalmic Res 15:191-197, 1983 24. Poulsom R, Mirrlees DJ, Earl DCN, et al: The effects of an aldose reductose inhibitor upon the sorbitol pathway, fructosel-phosphate, and lactate in the retina and nerve of streptozotocindiabetic rats. Exp Eye Res 36:751-760, 1983 25. Stewart MA, Sherman WR, Anthony S: Free sugars in alloxan diabetic rat nerve. Biochem Biophys Res Commun 22:488491,1966 26. Chandler ML, Boltralik J, York B, et al: Prevention of cataracts in diabetic rats by aldose reductase inhibitors. Invest Ophthalmol Vis Sci 22: 156, 1982 (suppl) 27. Gabbay KH: Role of Sorbitol pathway in neuropathy, in Advances in Metabolic Disorders (suppl2). Orlando, FL, Academic, 1973, pp 417-424 28. Clements RS: Diabetic neuropathy: New concepts of its etiology. Diabetes 28:604-611, 1979
of Two New Aldose
Brenda Walker
Reductase Inhibitors, in Diabetic Rats
Griffin, Loretta
G. McNatt,
Michael
AL-1567
L. Chandler,
and AL-1576,
and Billie M. York
Two new potent aldose reductase inhibitors, AL-l 567 (oL-spiro(2-fluoro-9H-fluoren-9,4’-imidazoiidine)-2’,5’-dione) and AL-l 576 ~spiro-(2.7-difluoro-9Kfluoren-9,4’-imidarolidine~2’,5’-dione), have been characterized with respect to in vitro activity toward rat lens and human placental aldose reductase and in vivo activity in uncontrolled, severely diabetic rats dosed acutely with the compounds. The IC, values for inhibition of rat lens aldose reductase are 2.7 x lo-’ mol/L for AL-1567 and 8.5 x 10-O mol/L for AL-1576; very similar IC, values were measured for each compound with the human placental enzyme. When the compounds were administered orally once per day to 3-week diabetic rats for a period of eight days, the ED, values for normalization of lens sorbitol levels were 0.60 mg/kg for AL-1567 and 0.05 mg/kg for AL-l 576, and for normalization of sciatic nerve sorbitol levels, 0.22 mg/kg for AL-1 567 and 0.04 mg/kg for AL-l 576. Compared with published data on other aldose reductase inhibitors evaluated in very similar diabetic rat models, both compounds have unusually high activity in lens, and AL-l 576 appears to be the most active such compound in both lens and sciatic nerve reported thusfar. The evidence linking increased sorbitol pathway activity to diabetic complications, such as cataract and neuropathy in animal models, suggests that aldose reductase inhibitors will be useful therapeutic agents in human diabetics. 8 1987 by Grune & Stratton, inc.
A
LDOSE REDUCTASE catalyzes the NADPH-dependent reduction of various aldehydes and aldoses to the corresponding alcohols and polyols.’ By sequential action of this enzyme and sorbitol dehydrogenase, an NAD+-requiring enzyme, glucose can be converted to fructose. Under hyperglycemic conditions, when the glucose concentration of insulin-unresponsive tissues, such as lens, rises in synchrony with plasma glucose concentration, the flux of glucose through the sorbitol pathway relative to glycolysis increases significantly.‘** Although the operation of this pathway under a glucose stress is well-documented, the immediate or long-term benefits of increased sorbitol pathway activity have eluded strict definition to date. However, because polyols such as sorbitol do not readily move across cell membranes, either by diffusion or transport, chronically elevated polyol levels can generate a severe osmotic stress within the responsive cell. In the case of lens, for example, the resultant swelling and membrane damage initiate a well-studied sequence of changes that progress eventually to nuclear opacity.3 The most definitive evidence for the critical role of aldose reductase activity and elevated polyol levels in sugar-induced cataractogenesis is the dramatic inhibition of this type of cataract by potent aldose reductase inhibitors administered to diabetic or galactosemic animals.3-6 Studies of several aldose reductase inhibitors in animal models of diabetes have suggested that the therapeutic effects of these compounds are not limited to lens, but can also occur in selected tissues, such as retina, kidney, and peripheral nerve, which contain aldose reductase and which are known to be at greatest risk for diabetic complications.31s.6Administration of potent aldose reductase inhibitors to diabetic or galactosemic animals can inhibit polyol accumulation in these tissues as well as prevent some of the From Alcon Laboratories. Inc. Fort Worth, TX Address reprint requests to Brenda W. Gri&. PhD, Alcon Laboratories, Inc, MS R2-43, 6201 S Freeway, PO Box 6600. Fort Worth, TX 76115. Q I987 by Grune & Stratton, Inc. 0026-0495/87/3605-0014$03.00/0
486
later-developing pathophysiologic and morphologic changes which mimic the human disease process.3-6 However, since sorbitol does not accumulate in other tissues of the diabetic rat to the levels observed for lens, the question of severity of the sorbitol-induced osmotic stress in those tissues has been raised. This issue is complicated by the cellular localization of aldose reductase within different tissues’ and by the fact that sorbitol accumulation depends not only on aldose reductase activity, but also on sorbitol dehydrogenase activity. It seems likely that a sustained increase of glucose metabolism through the sorbitol pathway may also create a metabolic stress, involving abnormal ratios of oxidized and reduced pyridine nucleotides, among other changes, that eventually compromise cellular integrity. Although our understanding of these phenomena is incomplete, the available evidence indicates that several aldose reductase inhibitors ameliorate biochemical, functional, and ultrastructural alterations in various tissues of animal models of diabetes.’ These results have provided a strong rationale for human clinical trials of this class of compounds.’ In this report, the in vitro and in vivo activities of two new potent aldose reductase inhibitors, AL- 1567 and AL-l 576, are described. The activities of these compounds are compared with published activities of other aldose reductase inhibitors that have been evaluated in very similar diabetic rat models. These data indicate that the more active compound AL-1576 is considerably more active in vivo than other aldose reductase inhibitors for which analogous data have been published. MATERIALS
AND METHODS
Aldose reductase activity was determined by a standard spectrophotometric assay’ in a solution containing 0.1 mol/L potassium phosphate buffer, pH 6.8, 10 mmol/L D,L-glyceraldehyde, 0.2 mmol/L NADPH, the enzyme, and, where present, an aldose reductase inhibitor. The inhibitor concentration inhibiting the rate by 50% (IC, value) was determined by linear regression analysis of the linear portion of plots of % inhibition v log [inhibitor concentration]. Since these very active compounds have very small ICso values, which are comparable to the actual concentration of aldose reduc-
Metabolism, Vol 36. No 5 (May), 1987: pp 486-490
487
NEW ALDOSE REDUCTASE INHIBITORS
tase in the assay mixture,
the rate of the uninhibited
reaction
0
was
always adjusted to a constant value, in order to make valid comparisons of I(& values among a series of inhibitors. Two sources of aldose reductase were employed: a soluble rat lens preparation described elsewhere* and a hemoglobin-free preparation from human placenta. The latter was prepared by homogenizing the tissue in 50 mmol/L triethanolamine buffer, pH 7.0, containing 1 mmol/L each mercaptoethanol and EDTA. After centrifugation, hemoglobin was removed from the supernatant by batch extraction with phosphocellulose and elution with the same buffer. This material, containing about 10 mg protein/ml, was diluted tenfold and used for evaluating the activities of aldose reductase inhibitors by the fluorescence assay described. A very sensitive fluorescence assay of aldose reductase activity was recently characterized*; it has been shown” that this assay has distinct sensitivity and specificity advantages compared to the standard spectrophotometric assay. With this method, it was possible to measure aldose reductase specifically in the crude human placental preparation because L-hexonate dehydrogenase has a much larger K, value for the oxidized pyridine nucleotide analog employed. Also, the fluorescence assay of aldose reductase permits a more accurate determination of the inhibition constants of very active compounds such as AL-I 576. The experimental conditions for the fluorescence assay were 0.05 mol/L potassium phosphate buffer, 12 mmol/L benzyl alcohol, 12.5 pmol/L oxidized 3-acetylpyridine adenine nucleotide phosphate, and enzyme in I mL total volume at 25 “C. The rate of formation of the fluorescent reduced pyridine nucleotide product was monitored with a Perkin-Elmer 650-10s or a Perkin-Elmer MPF-66 spectrofluorometer, set for 365 nm excitation and 480 nm emission, with IO nm slitwidths.’ Male Sprague-Dawley rats (SASCO, Omaha), weighing 230 to 300 g, were made diabetic by intravenous injection of 50 mg/kg streptozotocin (Sigma Chemical Co, St Louis), freshly dissolved in 0.05 mol/L citric acid buffer, pH 4.5, containing 0.5% NaCI; this solution was kept on ice and used within two hours. Within 1 week of induction of diabetes, plasma glucose levels of all rats were analyzed, after sampling the blood by tail venupuncture into heparinized capillary tubes. The glucose assay employed was a standard coupled enzymatic assay (Sigma), based on glucose oxidase and peroxidase. Injected rats with nonfasting plasma glucose values between 375 and 525 mg/dL were selected for the study, and randomly assigned to groups, so that the mean plasma glucose value of each diabetic group was the same within t 10%. Throughout the study, rats had unrestricted access to water and Purina Rodent Chow. (Richmond, IN). About 3 weeks after induction of diabetes, daily dosing of study animals was initiated and continued for a total of eight days. The compounds. in an aqueous vehicle, were administered orally by means of stainless steel feeding needles. Approximately one dose period after the last dose, animals were killed; lenses and samples of sciatic nerves were carefully dissected, weighed, and immediately frozen. Protein-free extracts of the tissues were prepared by stanusing ZnSO, and Ba(OH),. Analysis of sorbitol in dard procedures,” the neutralized extracts was performed according to published procedures, I’ in which sorbitol dehydrogenase catalyzes the stoichiometric reduction of NAD+ by sorbitol to the fluorescent product NADH. The reaction mixture contained I.0 mL of 0.1 mol/L glycine buffer, pH 9.5, 5 mmol/L EDTA, 10 mmol/L MgC&, 2 umts of sorbitol dehydrogenase, 0.8 mg NAD+, and 0.5 mL of the original or diluted tissue extract; sample blanks were identical except that sorbitol dehydrogenase was omitted. The increase in Ruorescence after a 20-minute incubation was measured with a PerkinElmer 650-10s fluorometer, set for 365 nm excitation and 455 nm emission wavelengths, and compared with a set of identically treated sorbitol standards. Sorbitol dehydrogenase and NAD’ were supplied by Sigma.
HN
-3
&3 NH
PA
0%
AL-1567: X = H Fig 1. Chemical structures of AL-l 567 and AL-l 576.
AL-1576: X = F
Statistical analysis of the sorbitol data was performed with a Digital Equipment Corporation PDPl l/44 computer and RS/l software (Bolt Beranek and Newman, Inc, Cambridge, MA). A parametric t-test comparison of the various group means was employed, with a 95% confidence level (P c .05) selected as the criterion for a significant difference between two group means. The chemical structures of AL-l 567 and AL- 1576 are shown in Fig I. RESULTS
Spectrophotometric assay of rat lens aldose reductase activity, with D,L-glyceraldehyde and NADPEI as substrates, established that AL- 1567 inhibition was uncompetitive with respect to the aldehyde substrate when the reduced pyridine nucleotide concentration was saturating (Fig 2). The IC,, values for inhibition of rat lens aldose reductase determined spectrophotometrically were 4.2 x 10-s mol/L for AL-1567 and 2.4 x 10-s mol/L for AL-1576; by the fluorescence
0.04 -10
1
I
0
10 i /
20
30
40
50
I GLYCERALOEBYDE , mMOLAR)
Fig 2. Kinetics of AL-1567 inhibition of D,L-glyceraldehyde reduction by NADPH catalyzed by rat lens aldose reductase. All reaction mixtures contained rat lens aldose reductase. 0.2 mmol/L NADPH, and variable concentrations of D.L-glyceraldehyde, in 0.1 mol/L potasssium phosphate buffer pH 6.6,25X, and the indicated concentration of AL-1567: (0). no inhibitor: (Cl), 1.5 x lo-* mol/L; (A). 3 x lo-‘mol/L: to), 6 x lo-* mol/L: (m). 6 x lo-* mol/L.
GRIFFIN
488
assay of aldose reductase activity,’ IC,, values for AL- 1567 and AL-1576 were 2.7 x IO-* mol/L and 8.5 x 10m9mol/L, respectively. With the fluorescence assay, very similar IC,, values were measured for inhibition of human placental aldose reductase: 2.5 x lo-’ mol/L for AL- 1567 and 1.8 x lo-” mol/L for AL-l 576. These data indicate that AL-1 576 is the more active inhibitor and that each compound has approximately the same absolute activity toward the rat and human enzymes. Concerning the uncompetitive inhibition pattern for AL1567 (Fig 2), the same pattern has been reported for some aldose reductase inhibitors,‘2,‘3 while mixed uncompetitivenoncompetitive patterns have been reported for others.‘4-‘6 These observations appear to relate less to chemical structures than to the inhibition constants of the inhibitors. Indeed all of the very active aldose reductase inhibitors now under development for human therapeutic use fall into the category of reversible “tight-binding” inhibitors,” ie, the inhibition constants are equal to or even smaller than the concentration of aldose reductase required in the spectrophotometric assay of enzymatic activity. As has been discussed at length,” kinetic patterns obtained under such conditions will depend on the manner in which the experiment is performed and are not diagnostic of the molecular mechanism of inhibition. Because the concentration of aldose reductase required in the fluorescence assay is decreased below the inhibition constants of most very active inhibitors, our studies indicate that accurate analyses of inhibition kinetics can be performed by this assay.8*9 For evaluating the aldose reductase inhibitor activity of these compounds in vivo, an acutely dosed diabetic rat model was chosen as the most appropriate system; very similar models have been employed by most laboratories studying It was considered essential to this class of compounds. ‘3~‘4~‘8~‘9 control as closely as possible the mean plasma glucose values of the various diabetic groups, for the following reasons: (1) the apparent K, value of aldose reductase for glucose in assays of enzyme activity in vitro has been shown to be very large, as large as 500 mmol/L,*' which greatly exceeds peak plasma glucose levels even in uncontrolled diabetic animals and (2) the glucose level in diabetic rat lens, an insulinunresponsive tissue, correlates with the plasma glucose level.18 Therefore, the plasma glucose concentration is a very important factor controlling sorbitol production in a given
tissue; to eliminate the possibility of introducing bias into the data by variable rates of sorbitol production among the diabetic groups, the mean plasma glucose levels of these groups were the same within f 10% (Table 1). Table 1 also summarizes the mean sorbitol contents of lens and sciatic nerve, in absolute amounts and as percentages of the diabetic control group, for the drug-treated and control groups. For both aldose reductase inhibitors, a characteristic dependence of sorbitol level on dose was observed in each of the tissues. It should be noted that of the four doses of AL-l 576 included in the study, only the lowest dose did not decrease the sciatic nerve sorbitol level of treated diabetics to normal, or below in the case of the 0.50 mg/kg/d dose. From log dose-response plots of these data, ED,, values were estimated: for sciatic nerve sorbitol inhibition, 0.22 mg/kg/d for AL-1567 and 0.04 mg/kg/d for AL-1576; for lens sorbitol inhibition, 0.60 mg/kg/d for AL-l 567 and 0.05 mg/kg/d for AL- 1576. The data demonstrate that AL- 1576 is measurably more active than AL-1567 in both tissues of diabetic rats, qualitatively consistent with their in vitro aldose reductase inhibitor activities. In other studies, it has been shown that chronic dosing of these compounds at much higher dose levels to normal rats produces small but measurable inhibition of the very low normal sorbitol levels of lens and sciatic nerve. The sorbitol values for normal rat lens and sciatic nerve measured in this study (Table 1) agree with the published value of 0.15 nmol/mg for lens,18 and from 0.07 to 0.34 nmol/mg for sciatic nerve.‘8’2’We have confirmed (data not on the time course of sorbishown) previous observations **z*~ to1 accumulation in diabetic rat lens; the polyol attains its maximal value about 3 weeks after streptozotocin injection, and decreases slowly thereafter. Reported values of sorbitol content in diabetic rat tissues at comparable times after induction of diabetes are 22.518 to 50.0 nmol/mg23 for lens and 1.924 to 3.5 nmol/mg25 for sciatic nerve. In various studies, sorbitol has been measured by either gas chromatographic techniques23,24 or by the enzymatic fluorescence assay employed in this study’8.25;where the two methods have been directly compared,** excellent agreement has been found. Since the enzymatic method” has several distinct advantages, including higher sensitivity, it has become the assay of choice for studies of aldose reductase inhibitor activity in diabetic animal models.‘31’8*‘9
Table 1. Mean Sorbitol Values in Lenses and Nerves of Drug-Treated, DCX3e Drug
he/kg/d)
Plasma Glucose (m&L)
Lens Sorbitol”
(nmol/mal
Normal, and Diabetic Control Groups
Lens Sorbitol
Nerve Sorbitol*
Nerve Sorbitol
% Diabetic Control
(nmollmg)
I% Diabetic Control)
AL-1567
0.20
446
f 58
39.0
+ 10.8
97.5
1.01
+ 0.18
AL-1567
0.50
437
t 36
21.0
k 3.2
52.5
0.73
+ 0.13
37.6
AL-1567
1.50
436
+ 53
5.85
+ 1.30
14.6
0.46
? 0.10
23.7
AL-1 576
0.032
443
* 47
22.6
t 5.3
56.5
1.24
k 0.18
63.9
AL-l
576
0.0s
448
+ 47
17.9
* 5.4
44.8
0.33
* 0.17
17.0
AL-l
576
0.20
451
k 46
7.98
+ 4.0
20.0
0.21
+ 0.06
10.8
0.50
3.4
0.08
f 0.02
1.94
k 0.61
100.0
0.25
k 0.07
12.9
AL-1576
443
ir 40
1.35
k 0.24
None (controls)
-
448
+ 55
40.0
+ 17.0
Normals
-
136
k 16
0.40
* 0.11
Values
are given as mean f SD (n = 6).
*Sorbitol units, nmol/mg tissue wet weight.
ET AL
100.0 1.0
52.1
4.1
489
NEW ALDOSE REDUCTASE INHIBITORS
DISCUSSION
diabetic rats. Oral dosing of AL-1576 (0.50 mg/ kg/d) or AL-1567 (1.5 mg/kg/d) for 15 weeks completely prevented appearance of cataract, evaluated by opthalmoscopic examination; in this same time period, nuclear opacity developed in the diabetic control (untreated) group.6 Both compounds have also been shown to be potent anticataract agents in a more accelerated and severe sugar cataract model, weanling rats maintained on a 30% galactose diet.26 It should be noted that other aldose reductase inhibitors for which information has been published appear to be measurably less active in lens than in sciatic nerve, as determined from EDSo values for normalization of sorbitol levels. However, AL-1576 is uniquely active in lens of diabetic rats; the low EDSovalue, 0.05 mg/kg/d, is comparable to the activity in sciatic nerve. Aldose reductase activity, linked to elevated intracellular sorbitol levels, has been implicated in diabetic complications of other tissues of diabetic animal models based on: (1) the presence of the enzyme in the tissue, (2) elevation of tissue sorbitol levels well above normal values in diabetic animals, and (3) the ability of several aldose reductase inhibitors to normalize sorbitol, as well as improve other measures of tissue function. Uncontrolled diabetic rats develop neuronal dysfunction, which closely resembles that of diabetes of long duration in humans.27S28It has been shown that both AL1567 and AL-1576,6 as well as other aldose reductase inhibitors,‘4,‘9 can significantly improve the nerve conduction velocity deficiency of chronically diabetic rats. The present study confirms other reports that the early increase in sorbitol content of sciatic nerve of diabetic rats can be subsequently normalized by short-term systemic administration of potent aldose reductase inhibitors. Thus, the data support the idea that elevation of nerve sorbitol is an early biochemical event in the later-developing neuropathy of insulin-uncontrolled, severely diabetic rats, and by inference, in human diabetics. In summary, the data presented in this report on the activities of two new aldose reductase inhibitors, AL-1567 and AL-1576, indicate that in vivo, AL-1576 is the most potent member of this class of compound thus far described, with unusually high activity in diabetic rat lens. Moreover, severely
The activities of two very potent aldose reductase inhibitors, AL-1567 and AL-1576, have been described. From recently published data, it is possible to compare the in vivo activities of these compounds with other aldose reductase inhibitors in various stages of development and human clinical testing. Although there are some variations in study protocols for diabetic rat models used in various labs, the following conclusions can be drawn. Tolrestat,i3 ONO2235,19 and statill appear to have very similar activities in sciatic nerve of acutely dosed diabetic rats, with ED,, values for normalization of sorbitol ranging between 5 and 10 mg/kg/d. The lens activity of ONO-2235 is not known, but statil was reported to inhibit cataract in diabetic rats at 25 mg/kg/d.14 The ED,, value of tolrestat in diabetic rat lens was recently reported as 10 to 20 mg/kg/d.13 However, it appears that the model in that study” was somewhat less severe than ours, since (1) drug treatment was initiated either 11 days prior to, or on the day of, induction of diabetes by streptozotocin and (2) animals were killed after only 10 additional days of diabetes/drug treatment, when tissue sorbitol values have not yet attained their maximal values.22,23 The diabetic rat model used in our study is very similar to that in which sorbinil has been reported to have ED,, values of approximately 1.5 and 0.80 mg/kg/d for normalization of lens and sciatic nerve sorbitol levels, respectively.” In this model, AL- 1567 and AL-I 576 are measurably more active than sorbinil; based on relative EDSo values, AL-1576 is estimated to be about 20 times more active than sorbinil in sciatic nerve and about 30 times more active in lens. Comparisons of molar in vivo activities of these compounds support these conclusions. Although sorbinil has the lowest molecular weight (236), the molecular weight of AL-1576 (286) is greater by only 20%. which is comparable to the experimental error in the ED,, values. ‘The documented anticataract activities of potent aldose reductase inhibitors in galactosemic and diabetic animals support the osmotic stress hypothesis of sugar cataractogenesis.’ whereby hyperosmolality produced by chronically elevated intracellular polyol levels initiates tissue swelling and subsequent biochemical alterations that progress to lens opacity. That inhibition of the polyol pathway does, in fact, prevent cataracts has been demonstrated by chronic administration of both AL-I 567 and AL-1576 to uncontrolled,
these data and other evidence cited provide additional support for a critical role of increased polyol pathway activity in
the later-developing pathophysiologic alterations of various tissues of animal models of human diabetes.
REFERENCES
1. Hayman S, Kinoshita JH: Isolation and properties of lens aldose reductase. J Biol Chem 240:877-882, 1965 2. Gabbay KH: The sorbitol pathway and the complications of diabetes. N Engl J Med 288:831-836, 1973 3. Kador PF, Robison WG Jr, Kinoshita JH: The pharmacology of aldose reductase inhibitors. Annu Rev Pharmacol Toxicol25:691714,1985 4. Dvornik K, Simard-Duquesne N, Krami M. et al: Polyol accumulation in galactosemic and diabetic rats: Control by an aldose reductase inhibitor. Science 182:1146-l 148, 1973
5. Chandler ML, Shannon WA, DeSantis L: Prevention of retinal capillary basement membrane thickening in diabetic rats by aldose
reductase inhibitors. Invest Ophthalmol Vis Sci 25:159. 1984 (suppi) 6. Griffin BW, Chandler ML, DeSantis L: Prevention of diabetic cataract and neuropathy in rats by two new aldose reductase inhibitors. Invest Ophthalmol Vis Sci 25: 136, 1984 (suppl) 7. Young RJ, Ewing DJ, Clarke BF: A controlled trial of Sorbinil, an aldose reductase inhibitor, in chronic painful diabetic neuropathy. Diabetes 32:938-942, 1983 8. Griffin BW, McNatt LG: Characterization of the reduction of 3-acetylpyridine adenine dinucleotide phosphate by benzyl alcohol catalyzed by aldose reductase. Arch Biochem Biophys 246:75-81, 1986 9. Griffin BW, McNatt LG. York BM: Characterization of
490
aldose reductase activities from human and animal sources by a sensitive fluorescence assay, in Weiner H (ed): Third International Workshop on Enzymology of Carbonyl Metabolism: Aldehyde Dehydrogenase, Aldo-Keto Reductases and Alcohol Dehydrogenase. New York, Liss (in press) 10. Sherman WA, Stewart MA: Identification of sorbitol in mammalian nerve. Biochem Biophys Res Commun 22:492-497, 1966 11. Malone JI, Knox G, Benford S, et al: Red cell sorbitol. An indicator of diabetic control. Diabetes 29:861-864, 1980 12. Peterson MJ, Sarges R, Aldinger CE, et al: CP-45, 634: A novel aldose reductase inhibitor that inhibits polyol pathway activity in diabetic and galactosemic rats. Metabolism 28: 456-461, 1979 (suppl 1) 13. Simard-Duquesne N, Greselin E, Dubuc J, et al: The effects of a new aldose reductase inhibitor (Tolrestat) in galactosemic and diabetic rats. Metabolism 34: 885-892, 1985 14. Stribling D, Mirrlees DJ, Harrison HE, et al: Properties of ICI 128,436, a novel aldose reductase inhibitor, and its effects on diabetic complications in the rat. Metabolism 34:336-344, 1985 15. Terashima H, Hama K, Yamamoto R, et al: Effects of a new aldose reductase inhibitor on various tissues in vitro. J Pharmacol Exp Ther 229:226-230,1984 16. Kador PF, Goosey JD, Sharpless NE, et al: Stereospecific inhibition of aldose reductase. Eur J Med Chem 16:293-298, 1981 17. Williams JW, Morrison JF: The kinetics of reversible tightbinding inhibition. Methods Enzymol69:437-467, 1979 18. Malone JI, Leavengood H, Peterson MJ, et al: Red blood cell sorbitol as an indicator of polyol pathway activity. Inhibition by Sorbinil in insulin-dependent diabetic subjects. Diabetes 33:45-49, 1984 19. Kikkawa R, Hatanaka I, Yasuda H, et al: Effect of a new
GRIFFIN ET AL
aldose reductase inhibitor, (E)-3-carboxymethyl-5-[(2E)-Methyl3-phenylpropenylidenelrhodanine (ONO-2235) on peripheral nerve disorders in streptozotocin-diabetic rats. Diabetologia 24:290-292, 1983 20. Gabbay KH, Cathcart ES: Purification and immunologic identification of aldose reductases: Diabetes 23:460-468, 1974 21. Yue DK, Hanwell MA, Satchel1 PM, et al: The effect of aldose reductase inhibition on motor nerve condition velocity in diabetic rats: Diabetes 3 1:789-794, 1982 22. Hutton JC, Schofield PJ, Williams JF, et al: Sorbitol metabolism in the retina: Accumulation of pathway intermediates in streptozotccin induced diabetes in the rat. Aust J Exp Biol Med Sci 52:361-373, 1974 23. Kuriyama H, Sasaki K, Fukuda M: Studies on diabetic cataract in rats induced by streptozotocin II. Biochemical examinations of rat lenses in relation to cataract stages. Ophthalmic Res 15:191-197, 1983 24. Poulsom R, Mirrlees DJ, Earl DCN, et al: The effects of an aldose reductose inhibitor upon the sorbitol pathway, fructosel-phosphate, and lactate in the retina and nerve of streptozotocindiabetic rats. Exp Eye Res 36:751-760, 1983 25. Stewart MA, Sherman WR, Anthony S: Free sugars in alloxan diabetic rat nerve. Biochem Biophys Res Commun 22:488491,1966 26. Chandler ML, Boltralik J, York B, et al: Prevention of cataracts in diabetic rats by aldose reductase inhibitors. Invest Ophthalmol Vis Sci 22: 156, 1982 (suppl) 27. Gabbay KH: Role of Sorbitol pathway in neuropathy, in Advances in Metabolic Disorders (suppl2). Orlando, FL, Academic, 1973, pp 417-424 28. Clements RS: Diabetic neuropathy: New concepts of its etiology. Diabetes 28:604-611, 1979