Bovine lens aldose reductase: Tight binding of the pyridine coenzyme

ARCHIVES

OF BIOCHEMISTRY

AND

BIOPHYSICS

Vol. 283, No. 2, December, pp. 512418,

1990

Bovine Lens Aldose Reductase: Tight Binding of the Pyridine Coenzyme’ Antonella Del Corso,* Daniela Barsacchi,* Maggiorana Giannessi,* Maria Grazia Tozzi,* Marcella Camici,*s2 Julien L. Houben,+ Maurizio Zandomeneghi,* and Umberto Mura*p3 *Dipartimento di Fisiologia e Biochimica, Universitci di Pisa, Via S. Maria 55, 56100 Pisa-Italy; tIstituto di Chimica Quuntistica ed Energetica Molecolare, C.N.R; and SDipartimento di Chimica e Chimica Industriale, Universita’ di Piss

Received April 20, 1990, and in revised form August 8, 1990

Analysis by HPLC of the protein-free supernatant obtained after denaturation of aldose reductase shows that the native form of the enzyme (ARb) contains a tightly bound NADP+, which is absent in the oxidatively modified form (ARa). The absorption, fluorescence, and circular dichroism spectra of ARb and ARa are consistent with the presence of the cofactor only in the native form of aldose reductase. On the other hand, the modified enzyme, in appropriate thiol reducing conditions, can tightly bind NADP+. This indicates a potential reversibility of the modification of aldose reductase, at least in terms of retention of the cofactor. 0 1990 Academic Press,

Inc.

Aldose reductase (alditol:NADP+ oxidoreductase, EC l.l.l.Zl), the promoter of the polyol pathway, catalyzes the NADPH-dependent reduction of a variety of aldoses and aldehydes to the corresponding alcohols. The enzyme is responsible for the formation of sorbitol, whose accumulation represents one of the major contributions to diabetic complications (l-3). Bovine lens aldose reductase has been purified and characterized for its catalytic properties by several authors (4-9), and its amino acid sequence has been determined i This work was supported in part by National Institutes of Health Grant ROl-EY07832 and by Italian C.N.R., Progetto Finalizzato “Invecchiamento”. We thank Prof. P. L. Ipata for his constant encouragement and interest in the work. We are indebted to Prof. F. Testi, Institute of Veterinary General Pathology and Pathological Anatomy, University of Bologna, for his valuable cooperation. Thanks are also due to Mr. G.L. Dall’Olio, BECA of Prunaro di Budrio, for the kind supply of bovine lenses, and to Dr. T. Mazza (BECA) for his cooperation. * Present address: Istituto di Produzione Animale, Universita di Udine (Italy). 3 To whom correspondence should be addressed. 512

(10). We have recently reported that this enzyme, in the presence of oxygen radical generating systems, undergoes a process of modification (11,12). The native (ARb)4 and the modified (ARa) enzyme forms exhibited a different sensitivity to inhibition by Sorbinil, ARa being completely insensitive to the inhibitor (12). The enzyme modification was accompanied by a marked change in stereospecificity toward the two enantiomers of glyceraldehyde (13). In this paper, spectroscopic and chemical lines of evidence are presented that indicate that the native form of bovine lens aldose reductase is characterized by the presence of tightly bound pyridine coenzyme, which is retained during the purification steps, while no evidence for the presence of bound NADP+ is observed for the modified form of the enzyme. A tight binding of pyridine coenzymes has been reported for NADH-dependent sheep liver aldehyde dehydrogenase (14), as well as for bovine liver and human erythrocyte catalase (15, 16), where the pyridine coenzyme, which is not involved in the catalysis, appears to protect the enzyme against inactivation by its own substrate Hz02. The tight binding of the pyridine coenzyme to bovine lens aldose reductase, which has never been reported so far for other aldose reductases, appears to be related to the redox state of cysteinyl residues of the enzyme molecule.

MATERIALS

AND

METHODS

Materials. NADP+, NADPH, D,L-glyceraldehyde and all other biochemicals were obtained from Sigma Chemical Co. Ammonium sulfate and all inorganic chemicals were of reagent grade from BDH. Calf eyes

4 Abbreviations used: ARb, native form of bovine lens aldose reductase; ARa, modified form of bovine lens aldose reductase; DTT, dithiothreitol; SDS, sodium dodecyl sulfate. 0003-9&x/90 $3.00 Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.

NADP-BOUND

ALDOSE

REDUCTASE

were obtained from freshly slaughtered animals at the local slaughterhouse and the lenses were removed and kept frozen until needed. Preparation of the native (ARb) and modified (ARa) forms of a&se redwtase. In order to purify ARb, frozen lenses, after incision of the capsule, were suspended in 50 mM sodium phosphate, pH 6.8, containing 5 mM 2-mercaptoethanol and 2 mM dithiothreitol (1 g tissue/3.5 ml), and stirred in an ice-cold bath for 1 h. The suspension was then centrifuged at 40,OOOgat 4°C for 40 min and the supernatant was processed as previously described (12), with the only exception being that the Orange Matrex purification step preceded the Sephadex G-75 chromatographic column (2.6 X 80 cm). In order to prepare the modified form of aldose reductase, the purified ARb was incubated (0.03 mg/ml) in 85 mM sodium phosphate, pH 6.8, containing 5 mM P-mercaptoethanol, 0.42 M ammonium sulfate, 0.1 mM FeSO, and 0.3 mM EDTA (11). The modification process was monitored by following the activation of aldose reductase at different times, using D,L-glyceraldehyde as substrate. After activation of the enzyme, usually obtained in 2 h of incubation, ARa was dialyzed by Amicon YM5 membrane, against 10 mM sodium phosphate pH 7 (standard buffer). Both enzyme preparations, which were electrophoretically homogeneous, were concentrated by Amicon YM5 membranes and stored at 4’C under sterile conditions. The assay for the enzyme activity Enzyme actiuity measurement. as previously described was performed with 4.7 mM D,L-glyceraldehyde (ll), by following the decrease in absorbance at 340 nm, which parallels the NADPH oxidation. One unit of enzyme activity is the amount of the enzyme which catalyzes the oxidation of 1 Fmol NADPH/min. HPLC analysis. NADP+ in the protein-free supernatant obtained after denaturation of the native and modified forms of aldose reductase was detected by a Beckman Model 332 HPLC system on a reverse-phase Supelco LC-18-DB column (25 cm X 4.6 mm). The elution was performed essentially according to Stocchi et al. (17). Buffer A was 0.1 M potassium phosphate, pH 6, and Buffer B was 0.1 M potassium phosphate, pH 6, containing 10% methanol. The chromatographic conditions were the following: 9 min 100% Buffer A, 6 min up to 25% Buffer B, 2.5 min up to 90% Buffer B, 2 min up to 100% Buffer B, 6 min 100% Buffer B. The flow rate was 1 ml/min and detection was performed at 254 nm. The samples for the HPLC analysis were prepared starting from purified ARa and ARb solutions (0.6 ml containing 1 mg protein). Each sample was acidified by addition of 20 ~1 of 1 N HCl, kept at 100°C for 9 min, and then centrifuged. The supernatant was filtered through UltrafreePF PTGC filters (Millipore, 10,000 MW cutoff), and the pH was brought to about 7 by addition of 5 ~1 1 M NaOH. A 0.5.ml aliquot was then applied to the HPLC column and analyzed. Standard NADP+, exhibiting a retention time of 18 min, when subjected to a treatment similar to that described for the enzyme samples, gives rise to two distinct peaks with retention times of 4.4 and 18 min (see Fig. 1, curve c). Spectroscopic analyses. Absorption spectra were performed on a Beckman DU-7 spectrophotometer, with lo-mm pathlength cuvettes. Fluorescence emission spectra were recorded on a ISSGreg 200 spectrofluorimeter wit.h excitation and emission resolution of 15 and 8 nm, respectively. Concentrations of the enzyme samples were chosen to obtain an absorbance of approximately 0.1 at the excitation wavelength of 295 nm. Quartz cells of lo-mm pathlength were used. No oxygen degassing was attempted. Circular dichroism spectra were obtained on a Jasco J40AS spectropolarimeter with a cylindrical lo-mm pathlength cuvette, kept at 10°C. A spectral band width of 2 nm was used. Other methods. Protein was determined by the Coomassie blue binding assay (18), using bovine serum albumin as standard. The purity of both enzyme forms was assessed by SDS-polyacrylamide gel electrophoresis in reducing conditions as previously described (12).

RESULTS

HPLC analysis. HPLC analysis of the supernatant obtained after boiling the acid-treated preparations of the native and modified forms of aldose reductase is consistent

a 1 FIG. 1. HPLC denatured native a: elution profile elution profile for and Methods.

20

10

elution

time

(min)

analysis of the protein-free supernatant of the and modified forms of aldose reductase. Curve for ARa. Curve b: elution profile for ARb. Curve c: standard NADP+. For further details see Materials

with the presence of the pyridine coenzyme only in the native form (Fig. lb). In fact, the elution profile obtained from the ARb preparation was superimposable with that of a standard solution of NADP+ subjected to the same treatment of the enzyme sample (Fig. lc). On the contrary, no absorbing material was observed for the supernatant of the modified form of aldose reductase (Fig. la). Absorption spectra. The absorption spectra of purified ARa and ARb were measured in the wavelength region 200-400 nm, and the effect of DTT and NADP+ tested. The spectral region 300-400 nm is shown in Fig. 2. The addition of DTT to the native purified aldose reductase determined the appearance of a well-defined shoulder at 340 nm (curve 5 versus curve 3) while the thiol compound did not affect the absorption spectrum of the modified form (curve 1). The presence of near stoichiometric amounts of NADP+ in the ARa solution in the absence of the thiol compound slightly increased the absorption spectrum of the protein (curve 2), while the simultaneous presence of NADP+ and DTT (curve 4) determined a significant hyperchromic effect, giving rise to an absorption spectrum similar to that of ARb plus DTT. When ARa was treated with NADP+ and then DTT, extensive dialysis against standard buffer did not remove the cofactor. In fact, the absorption spectrum of the di-

514

DEL

CORSO

0.15

$0.10 5 n : i 0.05

0 320 340 wavelength

360 380 (nm)

ET AL.

ARa was observed when DTT was added to the protein solutions (Fig. 4B). Indeed, the spectrum of the modified NADP+-supplemented aldose reductase exhibited a progressive increase of dichroism activity at 330 nm upon addition of increasing amounts of the thiol reagent. In this regard, it is noteworthy that NADP+ up to 30 pM, alone or in the presence of DTT, did not show any detectable dichroic signal until 290 nm, while a positive couplet centered at 262 (maximum of positive and negative bands at 270 and 255 nm, respectively) was observed. At higher concentrations of NADP+, the positive band of this couplet influenced the spectral region 300-350 nm by increasing the elongation values. Despite the interaction of ARa with NADP+, this enzyme form, unlike the native aldose reductase, did not efficiently retain the cofactor. In fact, after extensive dialysis against standard buffer of the NADP+-saturated enzyme, a dichroic signal at 330 nm accounting for only 30% of the signal of the undialyzed sample was observed. The addition of NADP+ to this enzyme sample gave rise to the expected elongation

FIG. 2. Absorption spectra of the purified native and modified forms of aldose reductase. The enzyme solutions contained 1 mg protein/ml. Curve 1: ARa in standard buffer, either alone or in the presence of 2 mM DTT. Curve 2: ARa supplemented with 30 PM NADP+. Curve 3: ARb in standard buffer. Curve 4: ARa supplemented with 30 pM NADP+ and 2 mM DTT. Curve 5: ARb supplemented with 2 mM DTT. All spectra were reported assuming zero absorbance at 400 nm.

alyzed sample, which was superimposable with curve 2 of Fig. 2, became similar to curve 4 of Fig. 2 after addition of DTT alone (data not shown). No effect on the absorption spectrum was observed when NADP+ was added to ARb in the presence of DTT. Circular dichroism spectra. Circular dichroism spectra of purified native and oxidatively modified forms of bovine lens aldose reductase exhibited marked differences only in the wavelength region 250-340 nm (Fig. 3). The addition of DTT did not affect the dichroic spectrum of the modified form of aldose reductase, but generated a remarkable change in the spectrum of the native aldose reductase (Fig. 3). In particular, a large negative dichroic band with a maximal elongation at 330 nm was observed. The addition of increasing amounts of NADP+ to ARa (Fig. 4A) gave rise to a dramatic change in the dichroic spectrum of this protein, which became similar to the ARb spectrum of Fig. 3A. Note that all spectra in Fig. 4 are reported without correction for progressive dilution since the overall spectra are due to two components whose concentration ratio varies during the experiment. While the elongation of the peak at 305 nm, generated by the addition of increasing amounts of NADP+, reached a saturation value (see inset Fig. 4A), the elongation of the two peaks at lower wavelength (283 and 272 nm) appeared to progressively increase with the amount of the added pyridine coenzyme (Fig. 4A). A further similarity between the spectra of ARb (Fig. 3A) and NADP+-supplemented

250

280 310 340 wavelength (nm)

370

FIG. 3. Circular dichroic spectra of the two forms of aldose reductase. A and B refer to CD spectra obtained with 0.8 mg/ml of ARb and ARa, respectively. (-) Aldose reductase in standard buffer; ( * * a) aldose reductase in standard buffer supplemented with 2 mM DTT; (- - -) baseline of standard buffer, either alone or supplemented with 2 mM DTT. AC at 330 nm for ARh in the presence of DTT, evaluated on the basis of a M, = 34,000, was 11.6 M-km-‘.

NADP-BOUND

ALDOSE

515

REDUCTASE

1

dithiothreitol

CmM)

A t

280

I

310

I

I

340 370 wavelength

280 (nm)

310

340

370

FIG. 4. Effect of NADP+ and dithiothreitol on the circular dichroism of the modified form of aldose reductase. A, CD spectra of ARa before (curve 1) and after sequential addition of 5, 3, 3, 5 and 90 al (curve 2 to curve 6) of 1 mM NADP+ to 0.97 ml of 0.8 mg/ml enzyme in standard at 305 nm as a function of buffer. At at 305 nm for NADP+-saturated ARa (curve 6) was 4.4 M-km- I. The inset shows the increase of CD, I, the indicated amount of added NADP+. Values are corrected for dilution and, at higher concentrations, for free NADP+ contribution. B, 1 ml of NADP+-saturated ARa (same as curve 6, dotted line) was supplemented with sequential addition of 50 mM DTT up to 3.5 mM final concentration. The inset shows the increase of CD at 330 nm as a function of the indicated amount of DTT. Values are corrected for dilution.

for a complete saturation. On the other hand, a complete retention of the cofactor by ARa was observed when the dialysis against standard buffer was performed using the enzyme saturated with NADP+ in the presence of DTT. Fluorescence emission spectra. The difference between the fluorescence spectra of ARa and ARb using an excitation wavelength of 295 nm is reported in Fig. 5, where the ratio of the emission intensity between the two enzyme forms is also shown. A quenching effect on the fluorescence emission of ARa was observed upon addition of NADP+ (Fig. 6). The presence of DTT further reduced the fluorescence emission of NADP+-supplemented ARa, while the effect of the thiol compound on the ARa spectrum in the absence of the cofactor was negligible (data not shown). The NADP+-dependent quenching of tryptophan emission is reported in the inset of Fig. 6. DISCUSSION

The native and the modified forms of bovine lens aldose reductase are both dependent on NADPH. Therefore, the

binding of the cofactor is an obvious requirement of the catalytic action of both enzyme forms. However, independently of the catalytic properties of the enzyme, our results reveal that the native, but not the oxidatively modified aldose reductase, has tightly bound pyridine coenzyme. On the other hand, the modified form of aldose reductase can interact with and, in appropriate thiol reducing conditions, tightly bind the cofactor, thus suggesting a potential reversibility of the modification, at least in terms of retention of NADP+. The presence of the pyridine coenzyme in the native aldose reductase is demonstrated by the HPLC analysis of the supernatant obtained after boiling the acid-treated preparations of the purified native and modified forms of the enzyme. In fact, the acid thermal denaturation released NADP+ and its degradation products only from the native form of aldose reductase (Fig. 1). The two enzyme forms used in this study were electrophoretically homogeneous and were subjected to a highly efficient gel

DEL

CORSO

ARa

280

330 wavelength

380 (nm)

430

FIG. 5. Fluorescence emission spectra of the native and modified forms of aldose reductase. The enzyme solutions contained 0.25 mg protein/ ml. Dashed line refers to the intensity ratio between the two spectra. For further details, see Materials and Methods.

filtration chromatography before the analysis: this implies that the pyridine coenzyme is tightly bound to the native aldose reductase, while it is lost after modification of the enzyme. Circular dichroism spectroscopy appears to be a very sensitive technique to monitor the cofactor-enzyme binding (Figs. 3 and 4). The 305-nm dichroic band of the native purified aldose reductase appears to be associated with the interaction between the protein and the coenzyme. Such a signal, which is not present in dichroic spectra of ARa, can be readily obtained when NADP+ is added to this enzyme form. As shown in Fig. 4A, this signal appears proportional to the bound cofactor, reaching a saturation value, which depends on the amount of protein (data not shown). Two additional peaks at lower wavelength, namely 283 and 272 nm, are still related to the protein-cofactor complex. The apparent blue shift of these bands on addition of oversaturating amounts of NADP+ (Fig. 4A) can be associated with the increasing contribution of the positive CD couplet of the free NADP+. A useful tool which enhances the signal of NADP+ bound to the protein is the thiol reagent DTT. The presence of this compound with the native enzyme generates a wide intense band with a maximum at approximately 330 nm (Fig. 3A). While no effect of DTT is observed with ARa in the absence of NADP+ (Fig. 3B), a thiol-dependent increase of the 330-nm dichroic band is observed using ARa saturated with NADP+ (Fig. 4B). Since the purification of the native form of aldose reductase included an

ET AL.

affinity chromatography step, in which the enzyme was eluted by NADPH (12), one could wonder whether the cofactor tight binding is an artifact of the purification procedure. Indeed, when the affinity chromatography step was circumvented by alternative purification procedures, the enzyme still showed the dichroic spectrum typical of the enzyme-cofactor complex. This indicates that aldose reductase retains the cofactor originally present in the lens. The dramatic difference in fluorescence emission spectra of ARa and ARb using an excitation wavelength of 295 nm (Fig. 5) is in line with the well-known quenching effect of pyridine coenzymes on tryptophan fluorescence (19-22). This interpretation is supported by the quenching effect of exogenously added NADP+ on the ARa fluorescence emission spectrum (Fig. 6). A further decrease in the fluorescence emission intensity is observed when DTT is added to the NADP+-supplemented ARa (Fig. 6). Since DTT quenches the fluorescence emission only when the cofactor is present, it is conceivable that its action is exerted through the protein-bound NADP+. The blue shift of the ARb fluorescence spectrum compared to the ARa spectrum, as judged by the ratio of the fluorescence emission spectra of the two enzyme forms (Fig. 5), is also observed after addition of NADP+ and DTT to the modified form of aldose reductase.

a

280

330 wavelength

380 (nm)

430

FIG. 6. Effect of NADP+ and DTT on the fluorescence emission spectra of the modified form of aldose reductase. The enzyme solution contained 0.25 mg protein/ml. ARa in standard buffer alone (curve a), or supplemented either with 15 pM NADP+ (curve b) or with 15 PM NADP+ and 2 mM DTT (curve c). The inset shows the quenching effect on fluorescence at 330 nm of the indicated amount of NADP+ added to a solution of ARa containing 2 mM DTT.

NADP-BOUND

ALDOSE

Even though the fluorescence quenching of the bound cofactor is likely to be different for the five tryptophan residues present in bovine lens aldose reductase (lo), it is difficult to interpret the observed blue shift exclusively in terms of specific tryptophan quenching. In fact, the shift toward the red region of the fluorescence emission spectrum of ARa, an enzyme form likely structurally different from ARb, might be caused by an exposure to a more polar environment of some tryptophan residues of this enzyme form with possible changes in fluorescence quantum yield. Although more experimental lines of evidence are required to define the mechanism of interaction of DTT with the native and modified forms of aldose reductase, it is clear that the dichroic band at 330 nm and the increase in absorption in the region 300-360 nm are strictly dependent on the simultaneous presence of NADP+ and DTT. In this respect, one could speculate that the effect of DTT on the spectral properties of aldose reductase could result from an interaction of the thiol compound with the protein-bound NADP+ either directly or mediated through a reduction of specific cysteine residue(s) present on the enzyme molecule. Dithiothreitol acting as a disulfide reducing agent could generate and keep in a reduced state a specific cysteine residue present on the protein molecule which can interact with the bound cofactor as a charge transfer agent. In this regard, it has been recently reported that reactive sulfbydryls in human placenta aldose reductase may be located at or near the NADPH binding site (23). The electron transfer from thiol ions to the nicotinamide ring of NADP+ could well account for the generation of the 330-nm dichroic band. This hypothesis is consistent with the results of the absorption spectra which show a shoulder in the region 300360 nm of ARb and NADP+-supplemented ARa, both in the presence of DTT (Fig. 2). Similar considerations apply also for a direct interaction between thiol group(s) of DTT and NADP+ bound to the enzyme. It is noteworthy that DTT is a sulfur analog of a reaction product of the enzyme (i.e., threitol) and could, therefore, exert a site-specific directed effect. If this is true, however, since 2 mM threitol is ineffective in generating the 330-nm dichroic band, the presence of SH groups appears to be relevant. In order to explain the observed effect of DTT on dichroic and absorption spectra, the possibility of a DTTdependent reduction of protein-bound NADP+ cannot be ruled out. Indeed, the observed increase in absorption at 340 nm upon addition of DTT to ARb or NADP+-supplemented ARa (Fig. 2), which is proportional to the amount of the enzyme, implies that if any DTT-dependent NADPH formation occurs, it is restricted to the proteinbound cofactor (i.e., only one catalytic cycle). A NADP’ reduction, however, is not consistent with the results obtained by CD spectra of ARa supplemented with the reduced form of t.he cofactor. In fact, either in the absence or in the presence of DTT, a wide dichroic signal in the

REDUCTASE

517

region 330-400 nm with a maximum at 350 nm (results not shown) is generated, which is significantly shifted with respect to the dichroic band obtained by the addition of NADP+ plus DTT to ARa. Finally, it is possible that the oxidized form of DTT, which exhibits a broad absorption peak in the 260- to 330-nm spectral region, with a maximum at 285 nm, might interact with NADP+ bound to the protein and account for the observed 330-nm dichroic band. However, since oxidized DTT, added to ARb or NADP+-supplemented ARa does not generate any dichroic signal at 330 nm, this possibility should be ruled out, unless the interaction between protein-bound NADP+ and oxidized DTT occurs only when the oxidation of the thiol compound is realized in situ. In any case, an interpretation of the absorption spectra exclusively in terms of the contribution of the oxidized form of DTT is possible only if postulating an increased extinction of the bound oxidized DTT as compared to the free compound (24). Our results show the inability of ARa to segregate the cofactor by a tight binding. In fact, if ARa is supplemented with an excess of NADP+ and then dialyzed, at least 70% of the enzyme is recovered free of the cofactor, thus indicating that the lack of NADPf in the modified aldose reductase is combined with a structural change of this protein. It is also evident, both from dichroism and from absorption spectroscopy analyses, that the treatment of ARa with DTT makes this enzyme form capable of a tight binding with the cofactor. Therefore, it appears that the efficiency of the binding between NADP+ and aldose reductase is strongly affected by the reduced state of cysteine residue(s) in the enzyme molecule.

REFERENCES 1. Kinoshita, J. H., Fukushi, S., Kador, P., and Merola, L. 0. (1979) Metabolism 28(Suppl. l), 462-469. 2. Kador, P. F., Robinson, W. G., Jr., and Kinoshita, J. H. (1984) Annu. Rev. Pharmacol. Toxicol. 25,691-714. 3. Dvornik, D., Simard-Duquesne, N., Krami, M., Sestanj, K., Gabbay, K., Kinoshita, J. H., Varma, S. D., and Merola, L. 0. (1973) Science 182,1146-1148. 4. Hayman, S., and Kinoshita, J. H. (1965) J. Biol. Cfzern. 240, 877882. 5. Sheaff, C. M., and Doughty, C. C. (1976) J. Biol. Chem. 251,26962702. 6. Halder, A. B., and Crabbe, M. J. C. (1984) Biochem. J. 219,33-39. I. Inagaki, K., Miwa, I., and Okuda, J. (1982) Arch. Biochm. Biophys. 216,337-344. 8. Conrad, S. M., and Doughty, C. C. (1982) Biochin. Biophys. Acta 708,348-357. 9. Doughty, C. C., and Conrad, S. M. (1982) Biochin. Biophys. Acta 708,358-364. 10. Early, S. L., Shade, S. Z., Williams, T. R., and Doughty, C. C. (1987) International Workshop on Aldose Reductase Inhibitors, Honolulu, Hawaii (Abstract (X.1). 11. Del Corso, A., Camici, M., and Mura, U. (1987) Biochem. Biophys. Res.Commun. 148,369-375.

518

DEL

12. Del Corso, A., Barsacchi, D., Camici, M., Garland, U. (1989) Arch. Biochem. Biophys. 270, 604-610.

CORSO ET AL.

D., and Mura,

13. Del Corso, A., Barsacchi, D., Giannessi, M., Tozzi, M. G., Camici, M., and Mura, U. (1989) J. Biol. Chem. 264, 17,653-17,655. 14. Hart, G. J., and Dickinson,

F. M. (1983) Biochem. J. 211,363-371.

15. Kirkman, H. N., Galiano, Chem. 262,660-666.

S., and Gaetani,

16. Kirkman, H. N., and Gaetani, USA 81.4343-4347.

G. F. (1987) J. Biol.

G. F. (1984) Proc. Natl. Acad. Sci.

17. Stocchi, V., Cucchiarini, L., Magnani, M., Chiarantini, L., Palma, P., and Crescentini, G. (1985) And. Biockm. 146,118-124.

18. Bradford,

M. M. (1976) Anal. Biochem. 72, 248-254.

19. Abdallah, M. A., Biellmann, J.-F., Wiget, P., Joppich-Kuhn, Luisi, P. L. (1978) J. Biochem. 89.397-405. 20. Eftink,

M. R. (1986) Biochemistry

21. Luisi, P. L., and Favilla, 22. Theorell,

R., and

25,6620-6624.

R. (1970) Eur. J. Biochem. 17, 91-94.

H., and Tatemoto, K. (1971) Arch. Biochem. Biophys. 142,

68-82. 23. Liu, S., Bhatnagar, A., Das, B., and Srivastava, Biochem. Biophys. 275,112-121. 24. Cleland, W. W. (1964) Biochemistry

3, 480-485.

S. K. (1989) Arch.