- Email: [email protected]
Interaction of Camel Lens ζ-Crystallin with Quinones: Portrait of a Substrate by Fluorescence Spectroscopy
Archives of Biochemistry and Biophysics Vol. 395, No. 2, November 15, pp. 185–190, 2001 doi:10.1006/abbi.2001.2538, available online at http://www.idealibrary.com on
Interaction of Camel Lens -Crystallin with Quinones: Portrait of a Substrate by Fluorescence Spectroscopy Mohammad D. Bazzi 1 Department of Biochemistry, College of Science, King Saud University, P.O. Box 2455, Riyadh 11451, Saudi Arabia
Received May 21, 2001, and in revised form July 24, 2001; published online October 26, 2001
Interaction of camel lens -crystallin, an NADPH: quinone oxidoreductase, with several quinone derivatives was examined by fluorescence spectroscopy and activity measurements. Fluorescence of -crystallin was quenched to different levels by the different quinones:juglone (5-OH, 1,4 naphthoquinone), 1,4 naphthoquinone (1,4-NQ), and 1,2 naphthoquinone (1,2-NQ) considerably quenched the fluorescence of -crystallin, where as the commonly used substrate, 9,10phenanthrenequinone (PQ) did not induce significant quenching. Activity measurements showed only PQ served as a substrate for camel lens -crystallin, while juglone, 1,4-NQ, and 1,2-NQ were inhibitors. Thus quinones that interacted with -crystallin directly inhibited the enzyme, whereas the substrate had very low affinity for the enzyme in the absence of NADPH. Another substrate, dichlorophenol indophenol (DCIP), conformed to the same pattern; DCIP did not quench the fluorescence of the enzyme significantly, but served as a substrate. This pattern is consistent with an ordered mechanism of catalysis with quinone being the second substrate. All three naphthoquinones were uncompetitive inhibitors with respect to NADPH and noncompetitive with respect to PQ. These kinetics are similar to those exhibited by cysteine- and/or lysinemodifying agents. Juglone, 1,4-NQ, and 1,2-NQ interacted with and quenched the fluorescence of camel lens ␣-crystallin, but to lesser extent than that of -crystallin. © 2001 Academic Press Key Words: camel lens; -crystallin; ␣-crystallin; quinones; fluorescence; inhibition; ordered mechanism.
-Crystallin is an NADPH:quinone oxidoreductase (EC 1.6.5.5) that catalyzes the reduction of some quinones through a one-electron transfer mechanism (1). 1 Address correspondence and reprint requests to author. E-mail: [email protected].
0003-9861/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.
-Crystallin was first characterized as a crystallin/enzyme in guinea pig lens (2), then in the lens of camel (3). The enzyme is distantly related to zinc-containing alcohol/polyol dehydrogenases, except that it shows specificity for NADPH and lacks the zinc binding domain (4, 5). -Crystallin may have a role in preventing or delaying cataract, at least in a strain of guinea pig that is prone to autosomal dominant hereditary cataract. -Crystallin from cataractous lenses of that strain was inactive, had a deletion of 34-residues at the NADPH-binding site, and failed to bind this nucleotide (4, 6). The physiological function (s) of -crystallin is unclear. Like other quinone reductases, -crystallin might function as a detoxification tool via two mechanisms. First, the enzyme exhibits high affinity for NADPH (7) and, thus, might serve as NADPH-binding protein to keep the lens environment in reduced form. Second, the oxidoreductase activity might also serve as a defense mechanism against quinones and other oxidizing agents (1). The presence of this enzyme in the lenses of some animals would appear to add certain advantages to these animals. However, the quinone oxidoreductase activity is supposed to proceed via a one-electron transfer mechanism (1), which under aerobic conditions produces oxygen radical, a highly harmful product. Guinea pig -crystallin is reported to have a broad range of substrates that includes juglone (5OH, 1,4 naphthoquinone), 1,4 naphthoquinone (1,4NQ), 1,2 naphthoquinone (1,2-NQ), 9,10-phenanthrenequinone (PQ), as well as dichlorophenol indophenol (DCIP), a nonquinone compound (1). In this study, direct interaction between -crystallin and several of its reported substrates was examined by fluorescence spectroscopy. The results showed that juglone, 1,4-NQ, and 1,2-NQ all interacted with and quenched the fluorescence of -crystallin significantly, while PQ and DCIP failed to interact with the enzyme directly in the absence of NADPH. Activity measure185
186
MOHAMMAD D. BAZZI
ments showed that only PQ and DCIP were substrates for camel lens -crystallin, whereas juglone, 1,4-NQ, and 1,2-NQ were strong inhibitors. Thus, substrates had low affinity for the enzyme in the absence of NADPH, while quinones that interacted with -crystallin directly did not serve as substrate. MATERIALS AND METHODS NADPH, PQ, juglone, 1,4-NQ, 1,2-NQ, and DICP were purchased from The Sigma Chemical Company (St. Louis, MO). Eyes of camels, all above 6 years old, were obtained from a local slaughterhouse. Stock solutions of the various quinones were prepared on the same day of the experiment. Concentration was determined by absorption using extinction coefficient of 2512 M ⫺1 cm ⫺1 at 340 nm, 2512 M ⫺1 cm ⫺1 at 330 nm, and 12590 M ⫺1 cm ⫺1 at 245 nm, 21000 M ⫺1 cm ⫺1 at 600 nm, and 31623 M ⫺1 cm ⫺1 at 252 nm for 1,2-NQ, 1,4-NQ, juglone, DCIP, and PQ, respectively. Protein purification. -Crystallin was purified from camel lens and assayed as described previously (8). ␣-Crystallin was prepared from the same camel lens homogenate used for the purification of -crystallin. The ␣L-crystallin peak was pooled, concentrated, and chromatographed on Suparose 12 HR gel filtration column (1 ⫻ 30 cm). The pooled peak consisted of ␣LA- and ␣LB-crystallin subunits and was used as such without further separation. Fluorescence measurements. Steady-state fluorescence measurements were conducted using 1.6 ml buffer containing 20 mM sodium phosphate, pH 7.8, 0.2 mM EDTA, and 1 M protein. Protein fluorescence was measured by exciting tryptophan (284 nm) and scanning the emission intensity (300 –500 nm). Titration of fluorescence quenching was carried out as a function of added ligand at a fixed protein concentration (1 M). Control experiments lacking the protein were run simultaneously with each assay. In all cases reported here, there were no time-dependent changes in the fluorescence of the samples; all fluorescence changes, if any, were virtually instantaneous with the addition of the reagents. Binding curves were generated by calculating the area of the protein emission (between 300 and 450 nm) as a function of added ligand. Quenching was expressed as a percentage change of the original fluorescence: Q ⫽ (F o – F i) * 100/F o, where F o and F i represent the protein fluorescence in the absence and the presence of added ligand, respectively. In the cases where the added material absorbs light at 284 nm (the excitation wavelength), the fluorescence spectra were corrected for inner filter effect as follows (9, 10): F corr ⫽ F obs * 10 A/2, where F corr is the corrected fluorescence, A is the absorption of the added ligand at 284 nm, and F obs is the observed fluorescence. In most of the cases reported here, the inner filter contribution was minimal and did not exceed 3%. Correction factors of larger magnitude were necessary only at high concentrations of PQ and DCIP. All fluorescence measurements were conducted at 22°C on Jasco FP750 spectrofluorometer attached to a PC. Data collection and manipulation were carried out using the manufacturer’s software. Enzyme Assay. The NADPH: quinone oxidoreductase activity of -crystallin was determined at 22°C in a phosphate buffer containing 100 M NADPH, 25 M substrate, and 0.5 g of enzyme in a final volume of 1.0 ml. The reaction was initiated by the addition of the substrate and the decrease in absorbance at 340 nm was monitored using spectrophotometer (Ultrospec 2000 Pharmacia-LKB). All quinone derivatives used in this study were examined as potential substrates for -crystallin. Blanks lacking either enzyme or substrate were run routinely. Activity was expressed as micromoles of NADPH oxidized per minute per mg of protein. Inhibition kinetics. Quinone derivatives that did not serve as substrates were examined as potential inhibitors of -crystallin using PQ as a substrate. First, inhibition was determined as a function
of added ligand in the presence of 100 M NADPH and 25 M PQ. Kinetics of inhibition were determined by double reciprocal plots of the initial velocities as a function of NADPH or PQ at a fixed concentration of inhibitor; -crystallin activity was measured as a function of NADPH (0 –100 M) at constant PQ (25 M) and in the presence of various concentrations of juglone, 1,2-NQ, or 1,4-NQ; the activity was also measured as a function of PQ (0 –25 M) at constant NADPH (100 M) and in the presence of the above mentioned quinone derivatives. All measurements were carried out in duplicates. The inhibition constant (K i) was calculated by plotting the reciprocals of apparent K m or V max versus inhibitor. Other methods. Protein was determined by the method of Bradford (11), using bovine serum albumin as a standard or by absorption using the extinction coefficient of 1.34 for 0.1% -crystallin at 280 nm (8). Binding parameters were determined by nonlinear least square fitting of the data to the general rate equation: Y i ⫽ Y max*X i/(K i ⫹ X i), where Y i is observed quenching at concentration X i; Y max and K i were the maximum quenching and the binding constant, respectively. All data manipulations were performed using Kaleidagraph software.
RESULTS
Figure 1 shows the fluorescence spectra of camel lens -crystallin in the presence of various concentrations of juglone. The protein had a multicomponent fluorescence spectrum with emission maximum at about 313 nm (7). Addition of juglone quenched the protein fluorescence in a specific and dose-dependent manner (Fig. 1). -Crystallin–juglone interaction quenched but did not alter the shape or the position of the protein fluorescence. This suggested that the interaction consisted of simple binding event, probably without major conformational changes in the enzyme. Quantitative analysis of fluorescence quenching is shown in Fig. 1B. Juglone quenched about 80% of the protein fluorescence with an apparent K d of about 3 M. Interaction of -crystallin with the various quinone derivatives was examined in a similar manner (Fig. 2). Based on the magnitude of fluorescence quenching, the tested quinones can be divided into two groups. The first group comprised juglone, 1,4-NQ, and 1,2-NQ, all considerably quenched the protein fluorescence. There were minor variations among these quinones. For example, 1,4-NQ had the highest affinity for the enzyme, but quenched the protein fluorescence the least. Juglone and 1,2-NQ both quenched the fluorescence to nearly the same magnitude, but juglone had the higher affinity. Nevertheless, interaction of -crystallin with any of these quinones was of moderate affinity (k dⱕ 10 M) and produced strong fluorescence quenching (ⱖ70%). The second group consisted of PQ and DCIP. PQ marginally quenched the fluorescence of -crystallin, even at very high concentrations. DCIP, a nonquinone substrate also quenched the fluorescence only at high concentrations (Fig. 2). It should be noted that DCIP, and to lesser extent PQ, absorb light at 284 nm (the excitation wavelength) and, thus, the magnitude of fluorescence quenching at high concentrations is uncertain due to artifacts of the innerfilter effect. In
INTERACTION OF -CRYSTALLIN WITH QUINONES
187
small variations among them, all these quinones were efficient inhibitors (ⱖ80%) at moderate concentrations (IC50 ranged between 3 and 11 M). Inhibition kinetics was investigated to determine whether these quinones interacted with the same binding of the substrate. -Crystallin activity was measured as a function of NADPH at constant PQ and inhibitor (Fig. 4A) or as a function PQ at constant NADPH and inhibitor (Fig. 4B). 1,2 NQ, juglone, and 1,4NQ all inhibited -crystallin uncompetitively with respect to NADPH with Ki of about 4, 9, and 15 M, respectively (Fig. 4A, inset). All three quinones also inhibited the enzyme noncompetitively with respect to PQ with Ki of about 7, 22, and 30 M, in the same order (Fig. 4B, inset). These results indicated that these inhibitors interacted with a site other than that of PQ or NADPH. The ability of the above mentioned naphthoquinones to interact with ␣-crystallin was also examined by fluorescence measurements. Juglone, 1,4-NQ, and 1,2-NQ interacted with ␣-crystallin and quenched its fluorescence in a dose-dependent manner (Fig. 5). Like that with -crystallin, interactions of ␣-crystallin with these quinones appeared to represent simple binding events. Quantitatively, affinity of the interaction as well as the magnitude of fluorescence quenching were higher with -crystallin than with ␣-crystallin for each of the tested quinones (Fig. 5).
FIG. 1. Interaction of juglone with camel lens -crystallin. (A) The fluorescence emission of 1 M protein in the presence of 0, 0.2, 0.4, 0.8, 1.6, 3.2, 6.4, and 12.8 M juglone (from top to bottom, in order). (B) Fluorescence quenching as function of juglone. The area under each curve above was numerically integrated, and fluorescence quenching was calculated as described under Materials and Methods. The panel shows the average and standard deviation of three independent measurements. The excitation wavelength was 284 nm.
any event, fluorescence quenching by either PQ or DCIP was much lower than that of juglone, 1,4-NQ, or 1,2-NQ. The ability of the various quinones to serve as substrate for -crystallin was examined by activity measurements (Fig. 3). The NADPH: quinone oxidoreductase activity of camel lens -crystallin was evident only with PQ or DCIP as a substrate, with PQ being the superior one. However, there was no oxidoreductase activity with either juglone, 1,4-NQ, or 1,2-NQ as a substrate, even at 100-folds the enzyme concentration used with PQ or DCIP. Instead, juglone, 1,4-NQ, and 1,2-NQ inhibited the NADPH:PQ oxidoreductase activity of -crystallin (Fig. 3B). 1,2-NQ was the most potent inhibitor followed by juglone and 1,4-NQ. Despite the
FIG. 2. Quinone–-crystallin interaction. The fluorescence emission of 1 M -crystallin was measured as a function of juglone (E), 1,2-NQ (䊐), 1,4-NQ (F), DCIP (), or PQ (‚). The results with juglone were the same as those described in the legend of Fig. 1 and were added here for purpose of illustration. Fluorescence quenching was calculated as described in the legend of Fig. 1.
188
MOHAMMAD D. BAZZI
could interact with, and possibly modulate ␣- and -crystallins. For multisubstrate enzymes, fluorescence measurements may provide a direct approach to measure the binding of the individual substrates. Often, protein– ligand interaction induces changes in the protein fluorescence, which can be used to quantitate the binding parameters (12). In the activity measurements, -crystallin must interact with both NADPH and the quinone substrate for the catalysis to proceed and, thus, binding parameters of individual substrates are obtained
FIG. 3. -Crystallin activity with the various quinones. (A) The NADPH oxidoreductase activity of -crystallin measured as a function of substrate. The activity was determined at 22°C in a phosphate buffer containing 100 M NADPH, the indicated concentration of quinone, and 0.5 g of enzyme in a final volume of 1.0 ml. Only PQ (F, left hand scale) and DCIP (E, right hand scale) produced measurable activity. There was no measurable activity with juglone, 1,2-NQ, or 1,4-NQ, even with 50 g enzyme in the assay. (B) The inhibitory effect of 1,2-NQ (E), 1,4-NQ (䊐), and juglone (F). These measurements were conducted at 22°C in a phosphate buffer containing 100 M NADPH, 25 M PQ, the indicated concentration of quinone, and 0.5 g of enzyme in a final volume of 1.0 ml.
DISCUSSION
-Crystallin is generally described as an NADPH: quinone oxidoreductase that catalyzes the reduction of various quinones. The enzyme has been reported to have a broad range of substrates that includes juglone, PQ, several naphthoquinones, and DCIP, a nonquinone compound (1). This study examined the binding of these quinones to camel lens -crystallin by fluorescence spectroscopy. The results revealed that different quinones had different affinities for -crystallin and probably different modes of interaction as well. Comparison between activity and fluorescence measurements suggested that only certain quinones could function as substrate for -crystallin, while other quinones
FIG. 4. Kinetics of inhibition. (A) Double reciprocal plots of the initial velocities determined as a function of NADPH. The panel shows results obtained with 0 (E), 4 (䊐), 8 (F), and 12 M juglone (■). 1,2 NQ and 1,4 NQ exhibited similar results and were omitted for brevity. The inset shows the intercepts obtained for 1,2 NQ (X), juglone (Œ), or 1,4 NQ (‚) plotted against its concentration. (B) Double reciprocal plots of the initial velocities measured as a function of PQ. The panel shows results obtained with juglone (symbols are the same as above), and again, results with 1,2 NQ and 1,4 NQ were omitted for brevity. The inset shows the slopes obtained for 1,2 NQ (X), juglone (Œ), or 1,4 NQ (‚) plotted as function of quinone. All measurements were carried out in duplicates.
INTERACTION OF -CRYSTALLIN WITH QUINONES
FIG. 5. Interaction of quinones with ␣-crystallin. The figure compares fluorescence quenching of ␣-crystallin (F) versus -crystallin (E) by juglone (top panel), 1,2-NQ (Middle panel), or 1,4-NQ (bottom panel). In all cases, fluorescence emission of 1 M protein was measured as a function of the added ligand. Fluorescence quenching was calculated as described in the legend of Fig. 1.
indirectly. A previous study indicated that NADPH interacted with and strongly quenched the fluorescence of -crystallin (7). The present study showed that the second substrate, PQ or DCIP, did not quench the enzyme’s fluorescence significantly, and by inference, probably did not interact with the enzyme directly. A major difference between the fluorescence measurements and activity measurement was the presence of NADPH in the latter. Thus, the enzyme had very low affinity for the second substrate (quinone) in the absence of the first (NADPH). This is consistent with the ordered Bi–Bi mechanism of catalysis, which is followed by most dehydrogenases (13, 14), as well as any other mechanism stipulating NADPH– enzyme binding as the first step of catalysis. Fluorescence measurements indicated that juglone, 1,4-NQ, and 1,2-NQ all interacted directly with and
189
quenched the fluorescence of camel lens -crystallin (Fig. 2). The same quinones also inhibited the NADPH: PQ oxidoreductase activity of the enzyme (Fig. 3). Thus, quinones that interacted with the enzyme directly were inhibitors, while substrates failed to interact with the enzyme in the absence of NADPH. The ability of juglone, 1,4-NQ, or 1,2-NQ to inhibit -crystallin, while unexpected, was not surprising; juglone and/or other naphthoquinones can inhibit a variety of enzymes (15–20); also, there are conflicting reports as to whether juglone a substrate or an inhibitor of -crystallin (1, 21). Nevertheless, fluorescence measurements indicated that juglone, 1,4-NQ and 1,2-NQ interacted with -crystallin in a different manner from that exhibited by PQ and DCIP. Kinetics of inhibitions by juglone, 1,4-NQ or 1,2-NQ were unexpected. Competitive inhibition with respect to PQ was expected, since they are, like the substrate, quinone derivatives. Instead, they all inhibited the enzyme noncompetitively with respect to PQ, and uncompetitively with respect to NADPH. Thus, interaction site (s) of these quinones with -crystallin was probably different from that of NADPH or PQ. The nature of their binding site on -crystallin remained unknown; however, several studies have proposed that naphthoquinones could interact with cysteine (s) of several enzymes (19, 22, 23). The present study showed that juglone, 1,4-NQ, and 1,2-NQ interacted with ␣-crystallin, the major structural protein of the lens (Fig. 5). These results pointed out two important considerations. First, these quinones exhibited similar affinities toward ␣- versus -crystallin, and thus, probably do not selectively target -crystallin in the lens. Second, these quinones may have specific and non-specific modes of interaction with proteins; their ability to interact with - and ␣-crystallins, as well as a variety of other proteins (24 –28) argues for their ability to interact with proteins via a general mode of interaction; it is unlikely that all these proteins had unique binding sites for these quinones. Again a general mode of interaction involving cysteine or other amino acids (e.g., lysine), as cited above, seems plausible in many cases. ACKNOWLEDGMENTS Many thanks to Drs. Nayyar Rabbani and Ali Duhaiman for their help in this work.
REFERENCES 1. Rao, P. V., Krishna, C. M., and Zigler, J. S., Jr. (1992) J. Biol. Chem. 267, 96 –102. 2. Huang, Q. L., Russell, P., Stone, S. H., and Zigler, J. S., Jr. (1987) Curr. Eye Res. 6, 725–732. 3. Garland, D., Rao, P. V., Del Corso, A., Mura, U., and Zigler, J. S., Jr. (1991) Arch. Biochem. Biophys. 285, 134 –136.
190
MOHAMMAD D. BAZZI
4. Borras, T., Jornvall, H., Rodokanaki, A., Gonzalez, P., Rodriguez, I., and Hernandez-Calzadilla, C. (1990) Exp. Eye Res. 50, 729 – 735. 5. Edwards, K. J., Barton, J. D., Rossjohn, J., Thorn, J. M., Taylor, G. L., and Ollis, D. L. (1996) Arch. Biochem. Biophys. 328, 173–183. 6. Rao, P. V., and Zigler, J. S., Jr. (1992) Exp. Eye Res. 54, 627– 630. 7. Bazzi, M. D., Rabbani, N., and Duhaiman, A. S. (2001) Biochim. Biophys. Acta 1544, 283–288. 8. Duhaiman, A. S., Rabbani, N., AlJafari, A. A., and Alhomida, A. S. (1995) Biochem. Biophys. Res. Commun. 215, 632– 640. 9. Stevens, A., and Augusteyn, R. C. (1997) Eur. J. Biochem. 243, 792–797. 10. Brands, L., and Witholt, B. (1967) Methods Enzymol. 11, 776 – 856. 11. Bradford, M. M. (1976) Anal. Biochem. 72, 248 –254. 12. Ward, L. D. (1985) Methods Enzymol. 117, 400 – 414. 13. Trivic, S., and Leskovac, V. (1994) Biochem. Mol. Biol. Int. 32, 399 – 407. 14. Ganzhorn, A. J., and Plapp, B. V. (1988) J. Biol. Chem. 263, 5446 –5454. 15. Imamura, Y., Koga, T., Uriu, Y., Otagiri, M., Satoh, K., and Hara, A. (2000) Biol. Pharm. Bull. 23, 155–158. 16. Thiboldeaux, R. L., Lindroth, R. L., and Tracy, J. W. (1998) Comp. Biochem. Physiol. C Pharmacol. Toxicol. Endocrinol. 120, 481– 487.
17. Sugie, S., Okamoto, K., Rahman, K. M., Tanaka, T., Kawai, K., Yamahara, J., and Mori, H. (1998) Cancer Lett. 127, 177–183. 18. Kamei, H., Koide, T., Kojima, T., Hashimoto, Y., and Hasegawa, M. (1998) Cancer Biotherm. Radiopharmacol. 13, 185–188. 19. Hennig, L., Christner, C., Kipping, M., Schelbert, B., Rucknagel, K. P., Grabley, S., Kullertz, G., and Fischer, G. (1998) Biochemistry 37, 5953–5960. 20. Duroux, L., Delmotte, F. M., Lancelin, J. M., Keravis, G., and Jay-Allemand, C. (1998) Biochem. J. 333, 275–283. 21. Duhaiman, A. S. (1996) Biochem. Biophys. Res. Commun. 218, 648 – 652. 22. Floreani, M., Forlin, A., Bellin, S., and Carpenedo, F. (1995) Biochem. Mol. Biol. Int. 37, 757–763. 23. Vos, R. M., Van Ommen, B., Hoekstein, M. S., De Goede, J. H., and Van Bladeren, P. J. (1989) Chem. Biol. Interact. 71, 381–392. 24. Chao, S. H., Greenleaf, A. L., and Price, D. H. (2001) Nucleic Acids Res. 29, 767–773. 25. Haubrich, D. R., and Wang, P. F. (1976) Biochem. Pharmacol. 25, 669 – 672. 26. Inouye, K., Saito, A., Orita, M., Tonomura, B., Imaishi, H., and Ohkawa, H. (2000) J. Biochem. (Tokyo) 127, 1041–1046. 27. Rao, P., and Zigler, J. S. (1992) Biochim. Biophys. Acta 1116, 75– 81. 28. Varga, Z., Bene, L., Pieri, C., Damjanovich, S., and Gaspar, R. (1996) Biochem. Biophys. Res. Commun. 218, 828 – 832.
Interaction of Camel Lens -Crystallin with Quinones: Portrait of a Substrate by Fluorescence Spectroscopy Mohammad D. Bazzi 1 Department of Biochemistry, College of Science, King Saud University, P.O. Box 2455, Riyadh 11451, Saudi Arabia
Received May 21, 2001, and in revised form July 24, 2001; published online October 26, 2001
Interaction of camel lens -crystallin, an NADPH: quinone oxidoreductase, with several quinone derivatives was examined by fluorescence spectroscopy and activity measurements. Fluorescence of -crystallin was quenched to different levels by the different quinones:juglone (5-OH, 1,4 naphthoquinone), 1,4 naphthoquinone (1,4-NQ), and 1,2 naphthoquinone (1,2-NQ) considerably quenched the fluorescence of -crystallin, where as the commonly used substrate, 9,10phenanthrenequinone (PQ) did not induce significant quenching. Activity measurements showed only PQ served as a substrate for camel lens -crystallin, while juglone, 1,4-NQ, and 1,2-NQ were inhibitors. Thus quinones that interacted with -crystallin directly inhibited the enzyme, whereas the substrate had very low affinity for the enzyme in the absence of NADPH. Another substrate, dichlorophenol indophenol (DCIP), conformed to the same pattern; DCIP did not quench the fluorescence of the enzyme significantly, but served as a substrate. This pattern is consistent with an ordered mechanism of catalysis with quinone being the second substrate. All three naphthoquinones were uncompetitive inhibitors with respect to NADPH and noncompetitive with respect to PQ. These kinetics are similar to those exhibited by cysteine- and/or lysinemodifying agents. Juglone, 1,4-NQ, and 1,2-NQ interacted with and quenched the fluorescence of camel lens ␣-crystallin, but to lesser extent than that of -crystallin. © 2001 Academic Press Key Words: camel lens; -crystallin; ␣-crystallin; quinones; fluorescence; inhibition; ordered mechanism.
-Crystallin is an NADPH:quinone oxidoreductase (EC 1.6.5.5) that catalyzes the reduction of some quinones through a one-electron transfer mechanism (1). 1 Address correspondence and reprint requests to author. E-mail: [email protected].
0003-9861/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.
-Crystallin was first characterized as a crystallin/enzyme in guinea pig lens (2), then in the lens of camel (3). The enzyme is distantly related to zinc-containing alcohol/polyol dehydrogenases, except that it shows specificity for NADPH and lacks the zinc binding domain (4, 5). -Crystallin may have a role in preventing or delaying cataract, at least in a strain of guinea pig that is prone to autosomal dominant hereditary cataract. -Crystallin from cataractous lenses of that strain was inactive, had a deletion of 34-residues at the NADPH-binding site, and failed to bind this nucleotide (4, 6). The physiological function (s) of -crystallin is unclear. Like other quinone reductases, -crystallin might function as a detoxification tool via two mechanisms. First, the enzyme exhibits high affinity for NADPH (7) and, thus, might serve as NADPH-binding protein to keep the lens environment in reduced form. Second, the oxidoreductase activity might also serve as a defense mechanism against quinones and other oxidizing agents (1). The presence of this enzyme in the lenses of some animals would appear to add certain advantages to these animals. However, the quinone oxidoreductase activity is supposed to proceed via a one-electron transfer mechanism (1), which under aerobic conditions produces oxygen radical, a highly harmful product. Guinea pig -crystallin is reported to have a broad range of substrates that includes juglone (5OH, 1,4 naphthoquinone), 1,4 naphthoquinone (1,4NQ), 1,2 naphthoquinone (1,2-NQ), 9,10-phenanthrenequinone (PQ), as well as dichlorophenol indophenol (DCIP), a nonquinone compound (1). In this study, direct interaction between -crystallin and several of its reported substrates was examined by fluorescence spectroscopy. The results showed that juglone, 1,4-NQ, and 1,2-NQ all interacted with and quenched the fluorescence of -crystallin significantly, while PQ and DCIP failed to interact with the enzyme directly in the absence of NADPH. Activity measure185
186
MOHAMMAD D. BAZZI
ments showed that only PQ and DCIP were substrates for camel lens -crystallin, whereas juglone, 1,4-NQ, and 1,2-NQ were strong inhibitors. Thus, substrates had low affinity for the enzyme in the absence of NADPH, while quinones that interacted with -crystallin directly did not serve as substrate. MATERIALS AND METHODS NADPH, PQ, juglone, 1,4-NQ, 1,2-NQ, and DICP were purchased from The Sigma Chemical Company (St. Louis, MO). Eyes of camels, all above 6 years old, were obtained from a local slaughterhouse. Stock solutions of the various quinones were prepared on the same day of the experiment. Concentration was determined by absorption using extinction coefficient of 2512 M ⫺1 cm ⫺1 at 340 nm, 2512 M ⫺1 cm ⫺1 at 330 nm, and 12590 M ⫺1 cm ⫺1 at 245 nm, 21000 M ⫺1 cm ⫺1 at 600 nm, and 31623 M ⫺1 cm ⫺1 at 252 nm for 1,2-NQ, 1,4-NQ, juglone, DCIP, and PQ, respectively. Protein purification. -Crystallin was purified from camel lens and assayed as described previously (8). ␣-Crystallin was prepared from the same camel lens homogenate used for the purification of -crystallin. The ␣L-crystallin peak was pooled, concentrated, and chromatographed on Suparose 12 HR gel filtration column (1 ⫻ 30 cm). The pooled peak consisted of ␣LA- and ␣LB-crystallin subunits and was used as such without further separation. Fluorescence measurements. Steady-state fluorescence measurements were conducted using 1.6 ml buffer containing 20 mM sodium phosphate, pH 7.8, 0.2 mM EDTA, and 1 M protein. Protein fluorescence was measured by exciting tryptophan (284 nm) and scanning the emission intensity (300 –500 nm). Titration of fluorescence quenching was carried out as a function of added ligand at a fixed protein concentration (1 M). Control experiments lacking the protein were run simultaneously with each assay. In all cases reported here, there were no time-dependent changes in the fluorescence of the samples; all fluorescence changes, if any, were virtually instantaneous with the addition of the reagents. Binding curves were generated by calculating the area of the protein emission (between 300 and 450 nm) as a function of added ligand. Quenching was expressed as a percentage change of the original fluorescence: Q ⫽ (F o – F i) * 100/F o, where F o and F i represent the protein fluorescence in the absence and the presence of added ligand, respectively. In the cases where the added material absorbs light at 284 nm (the excitation wavelength), the fluorescence spectra were corrected for inner filter effect as follows (9, 10): F corr ⫽ F obs * 10 A/2, where F corr is the corrected fluorescence, A is the absorption of the added ligand at 284 nm, and F obs is the observed fluorescence. In most of the cases reported here, the inner filter contribution was minimal and did not exceed 3%. Correction factors of larger magnitude were necessary only at high concentrations of PQ and DCIP. All fluorescence measurements were conducted at 22°C on Jasco FP750 spectrofluorometer attached to a PC. Data collection and manipulation were carried out using the manufacturer’s software. Enzyme Assay. The NADPH: quinone oxidoreductase activity of -crystallin was determined at 22°C in a phosphate buffer containing 100 M NADPH, 25 M substrate, and 0.5 g of enzyme in a final volume of 1.0 ml. The reaction was initiated by the addition of the substrate and the decrease in absorbance at 340 nm was monitored using spectrophotometer (Ultrospec 2000 Pharmacia-LKB). All quinone derivatives used in this study were examined as potential substrates for -crystallin. Blanks lacking either enzyme or substrate were run routinely. Activity was expressed as micromoles of NADPH oxidized per minute per mg of protein. Inhibition kinetics. Quinone derivatives that did not serve as substrates were examined as potential inhibitors of -crystallin using PQ as a substrate. First, inhibition was determined as a function
of added ligand in the presence of 100 M NADPH and 25 M PQ. Kinetics of inhibition were determined by double reciprocal plots of the initial velocities as a function of NADPH or PQ at a fixed concentration of inhibitor; -crystallin activity was measured as a function of NADPH (0 –100 M) at constant PQ (25 M) and in the presence of various concentrations of juglone, 1,2-NQ, or 1,4-NQ; the activity was also measured as a function of PQ (0 –25 M) at constant NADPH (100 M) and in the presence of the above mentioned quinone derivatives. All measurements were carried out in duplicates. The inhibition constant (K i) was calculated by plotting the reciprocals of apparent K m or V max versus inhibitor. Other methods. Protein was determined by the method of Bradford (11), using bovine serum albumin as a standard or by absorption using the extinction coefficient of 1.34 for 0.1% -crystallin at 280 nm (8). Binding parameters were determined by nonlinear least square fitting of the data to the general rate equation: Y i ⫽ Y max*X i/(K i ⫹ X i), where Y i is observed quenching at concentration X i; Y max and K i were the maximum quenching and the binding constant, respectively. All data manipulations were performed using Kaleidagraph software.
RESULTS
Figure 1 shows the fluorescence spectra of camel lens -crystallin in the presence of various concentrations of juglone. The protein had a multicomponent fluorescence spectrum with emission maximum at about 313 nm (7). Addition of juglone quenched the protein fluorescence in a specific and dose-dependent manner (Fig. 1). -Crystallin–juglone interaction quenched but did not alter the shape or the position of the protein fluorescence. This suggested that the interaction consisted of simple binding event, probably without major conformational changes in the enzyme. Quantitative analysis of fluorescence quenching is shown in Fig. 1B. Juglone quenched about 80% of the protein fluorescence with an apparent K d of about 3 M. Interaction of -crystallin with the various quinone derivatives was examined in a similar manner (Fig. 2). Based on the magnitude of fluorescence quenching, the tested quinones can be divided into two groups. The first group comprised juglone, 1,4-NQ, and 1,2-NQ, all considerably quenched the protein fluorescence. There were minor variations among these quinones. For example, 1,4-NQ had the highest affinity for the enzyme, but quenched the protein fluorescence the least. Juglone and 1,2-NQ both quenched the fluorescence to nearly the same magnitude, but juglone had the higher affinity. Nevertheless, interaction of -crystallin with any of these quinones was of moderate affinity (k dⱕ 10 M) and produced strong fluorescence quenching (ⱖ70%). The second group consisted of PQ and DCIP. PQ marginally quenched the fluorescence of -crystallin, even at very high concentrations. DCIP, a nonquinone substrate also quenched the fluorescence only at high concentrations (Fig. 2). It should be noted that DCIP, and to lesser extent PQ, absorb light at 284 nm (the excitation wavelength) and, thus, the magnitude of fluorescence quenching at high concentrations is uncertain due to artifacts of the innerfilter effect. In
INTERACTION OF -CRYSTALLIN WITH QUINONES
187
small variations among them, all these quinones were efficient inhibitors (ⱖ80%) at moderate concentrations (IC50 ranged between 3 and 11 M). Inhibition kinetics was investigated to determine whether these quinones interacted with the same binding of the substrate. -Crystallin activity was measured as a function of NADPH at constant PQ and inhibitor (Fig. 4A) or as a function PQ at constant NADPH and inhibitor (Fig. 4B). 1,2 NQ, juglone, and 1,4NQ all inhibited -crystallin uncompetitively with respect to NADPH with Ki of about 4, 9, and 15 M, respectively (Fig. 4A, inset). All three quinones also inhibited the enzyme noncompetitively with respect to PQ with Ki of about 7, 22, and 30 M, in the same order (Fig. 4B, inset). These results indicated that these inhibitors interacted with a site other than that of PQ or NADPH. The ability of the above mentioned naphthoquinones to interact with ␣-crystallin was also examined by fluorescence measurements. Juglone, 1,4-NQ, and 1,2-NQ interacted with ␣-crystallin and quenched its fluorescence in a dose-dependent manner (Fig. 5). Like that with -crystallin, interactions of ␣-crystallin with these quinones appeared to represent simple binding events. Quantitatively, affinity of the interaction as well as the magnitude of fluorescence quenching were higher with -crystallin than with ␣-crystallin for each of the tested quinones (Fig. 5).
FIG. 1. Interaction of juglone with camel lens -crystallin. (A) The fluorescence emission of 1 M protein in the presence of 0, 0.2, 0.4, 0.8, 1.6, 3.2, 6.4, and 12.8 M juglone (from top to bottom, in order). (B) Fluorescence quenching as function of juglone. The area under each curve above was numerically integrated, and fluorescence quenching was calculated as described under Materials and Methods. The panel shows the average and standard deviation of three independent measurements. The excitation wavelength was 284 nm.
any event, fluorescence quenching by either PQ or DCIP was much lower than that of juglone, 1,4-NQ, or 1,2-NQ. The ability of the various quinones to serve as substrate for -crystallin was examined by activity measurements (Fig. 3). The NADPH: quinone oxidoreductase activity of camel lens -crystallin was evident only with PQ or DCIP as a substrate, with PQ being the superior one. However, there was no oxidoreductase activity with either juglone, 1,4-NQ, or 1,2-NQ as a substrate, even at 100-folds the enzyme concentration used with PQ or DCIP. Instead, juglone, 1,4-NQ, and 1,2-NQ inhibited the NADPH:PQ oxidoreductase activity of -crystallin (Fig. 3B). 1,2-NQ was the most potent inhibitor followed by juglone and 1,4-NQ. Despite the
FIG. 2. Quinone–-crystallin interaction. The fluorescence emission of 1 M -crystallin was measured as a function of juglone (E), 1,2-NQ (䊐), 1,4-NQ (F), DCIP (), or PQ (‚). The results with juglone were the same as those described in the legend of Fig. 1 and were added here for purpose of illustration. Fluorescence quenching was calculated as described in the legend of Fig. 1.
188
MOHAMMAD D. BAZZI
could interact with, and possibly modulate ␣- and -crystallins. For multisubstrate enzymes, fluorescence measurements may provide a direct approach to measure the binding of the individual substrates. Often, protein– ligand interaction induces changes in the protein fluorescence, which can be used to quantitate the binding parameters (12). In the activity measurements, -crystallin must interact with both NADPH and the quinone substrate for the catalysis to proceed and, thus, binding parameters of individual substrates are obtained
FIG. 3. -Crystallin activity with the various quinones. (A) The NADPH oxidoreductase activity of -crystallin measured as a function of substrate. The activity was determined at 22°C in a phosphate buffer containing 100 M NADPH, the indicated concentration of quinone, and 0.5 g of enzyme in a final volume of 1.0 ml. Only PQ (F, left hand scale) and DCIP (E, right hand scale) produced measurable activity. There was no measurable activity with juglone, 1,2-NQ, or 1,4-NQ, even with 50 g enzyme in the assay. (B) The inhibitory effect of 1,2-NQ (E), 1,4-NQ (䊐), and juglone (F). These measurements were conducted at 22°C in a phosphate buffer containing 100 M NADPH, 25 M PQ, the indicated concentration of quinone, and 0.5 g of enzyme in a final volume of 1.0 ml.
DISCUSSION
-Crystallin is generally described as an NADPH: quinone oxidoreductase that catalyzes the reduction of various quinones. The enzyme has been reported to have a broad range of substrates that includes juglone, PQ, several naphthoquinones, and DCIP, a nonquinone compound (1). This study examined the binding of these quinones to camel lens -crystallin by fluorescence spectroscopy. The results revealed that different quinones had different affinities for -crystallin and probably different modes of interaction as well. Comparison between activity and fluorescence measurements suggested that only certain quinones could function as substrate for -crystallin, while other quinones
FIG. 4. Kinetics of inhibition. (A) Double reciprocal plots of the initial velocities determined as a function of NADPH. The panel shows results obtained with 0 (E), 4 (䊐), 8 (F), and 12 M juglone (■). 1,2 NQ and 1,4 NQ exhibited similar results and were omitted for brevity. The inset shows the intercepts obtained for 1,2 NQ (X), juglone (Œ), or 1,4 NQ (‚) plotted against its concentration. (B) Double reciprocal plots of the initial velocities measured as a function of PQ. The panel shows results obtained with juglone (symbols are the same as above), and again, results with 1,2 NQ and 1,4 NQ were omitted for brevity. The inset shows the slopes obtained for 1,2 NQ (X), juglone (Œ), or 1,4 NQ (‚) plotted as function of quinone. All measurements were carried out in duplicates.
INTERACTION OF -CRYSTALLIN WITH QUINONES
FIG. 5. Interaction of quinones with ␣-crystallin. The figure compares fluorescence quenching of ␣-crystallin (F) versus -crystallin (E) by juglone (top panel), 1,2-NQ (Middle panel), or 1,4-NQ (bottom panel). In all cases, fluorescence emission of 1 M protein was measured as a function of the added ligand. Fluorescence quenching was calculated as described in the legend of Fig. 1.
indirectly. A previous study indicated that NADPH interacted with and strongly quenched the fluorescence of -crystallin (7). The present study showed that the second substrate, PQ or DCIP, did not quench the enzyme’s fluorescence significantly, and by inference, probably did not interact with the enzyme directly. A major difference between the fluorescence measurements and activity measurement was the presence of NADPH in the latter. Thus, the enzyme had very low affinity for the second substrate (quinone) in the absence of the first (NADPH). This is consistent with the ordered Bi–Bi mechanism of catalysis, which is followed by most dehydrogenases (13, 14), as well as any other mechanism stipulating NADPH– enzyme binding as the first step of catalysis. Fluorescence measurements indicated that juglone, 1,4-NQ, and 1,2-NQ all interacted directly with and
189
quenched the fluorescence of camel lens -crystallin (Fig. 2). The same quinones also inhibited the NADPH: PQ oxidoreductase activity of the enzyme (Fig. 3). Thus, quinones that interacted with the enzyme directly were inhibitors, while substrates failed to interact with the enzyme in the absence of NADPH. The ability of juglone, 1,4-NQ, or 1,2-NQ to inhibit -crystallin, while unexpected, was not surprising; juglone and/or other naphthoquinones can inhibit a variety of enzymes (15–20); also, there are conflicting reports as to whether juglone a substrate or an inhibitor of -crystallin (1, 21). Nevertheless, fluorescence measurements indicated that juglone, 1,4-NQ and 1,2-NQ interacted with -crystallin in a different manner from that exhibited by PQ and DCIP. Kinetics of inhibitions by juglone, 1,4-NQ or 1,2-NQ were unexpected. Competitive inhibition with respect to PQ was expected, since they are, like the substrate, quinone derivatives. Instead, they all inhibited the enzyme noncompetitively with respect to PQ, and uncompetitively with respect to NADPH. Thus, interaction site (s) of these quinones with -crystallin was probably different from that of NADPH or PQ. The nature of their binding site on -crystallin remained unknown; however, several studies have proposed that naphthoquinones could interact with cysteine (s) of several enzymes (19, 22, 23). The present study showed that juglone, 1,4-NQ, and 1,2-NQ interacted with ␣-crystallin, the major structural protein of the lens (Fig. 5). These results pointed out two important considerations. First, these quinones exhibited similar affinities toward ␣- versus -crystallin, and thus, probably do not selectively target -crystallin in the lens. Second, these quinones may have specific and non-specific modes of interaction with proteins; their ability to interact with - and ␣-crystallins, as well as a variety of other proteins (24 –28) argues for their ability to interact with proteins via a general mode of interaction; it is unlikely that all these proteins had unique binding sites for these quinones. Again a general mode of interaction involving cysteine or other amino acids (e.g., lysine), as cited above, seems plausible in many cases. ACKNOWLEDGMENTS Many thanks to Drs. Nayyar Rabbani and Ali Duhaiman for their help in this work.
REFERENCES 1. Rao, P. V., Krishna, C. M., and Zigler, J. S., Jr. (1992) J. Biol. Chem. 267, 96 –102. 2. Huang, Q. L., Russell, P., Stone, S. H., and Zigler, J. S., Jr. (1987) Curr. Eye Res. 6, 725–732. 3. Garland, D., Rao, P. V., Del Corso, A., Mura, U., and Zigler, J. S., Jr. (1991) Arch. Biochem. Biophys. 285, 134 –136.
190
MOHAMMAD D. BAZZI
4. Borras, T., Jornvall, H., Rodokanaki, A., Gonzalez, P., Rodriguez, I., and Hernandez-Calzadilla, C. (1990) Exp. Eye Res. 50, 729 – 735. 5. Edwards, K. J., Barton, J. D., Rossjohn, J., Thorn, J. M., Taylor, G. L., and Ollis, D. L. (1996) Arch. Biochem. Biophys. 328, 173–183. 6. Rao, P. V., and Zigler, J. S., Jr. (1992) Exp. Eye Res. 54, 627– 630. 7. Bazzi, M. D., Rabbani, N., and Duhaiman, A. S. (2001) Biochim. Biophys. Acta 1544, 283–288. 8. Duhaiman, A. S., Rabbani, N., AlJafari, A. A., and Alhomida, A. S. (1995) Biochem. Biophys. Res. Commun. 215, 632– 640. 9. Stevens, A., and Augusteyn, R. C. (1997) Eur. J. Biochem. 243, 792–797. 10. Brands, L., and Witholt, B. (1967) Methods Enzymol. 11, 776 – 856. 11. Bradford, M. M. (1976) Anal. Biochem. 72, 248 –254. 12. Ward, L. D. (1985) Methods Enzymol. 117, 400 – 414. 13. Trivic, S., and Leskovac, V. (1994) Biochem. Mol. Biol. Int. 32, 399 – 407. 14. Ganzhorn, A. J., and Plapp, B. V. (1988) J. Biol. Chem. 263, 5446 –5454. 15. Imamura, Y., Koga, T., Uriu, Y., Otagiri, M., Satoh, K., and Hara, A. (2000) Biol. Pharm. Bull. 23, 155–158. 16. Thiboldeaux, R. L., Lindroth, R. L., and Tracy, J. W. (1998) Comp. Biochem. Physiol. C Pharmacol. Toxicol. Endocrinol. 120, 481– 487.
17. Sugie, S., Okamoto, K., Rahman, K. M., Tanaka, T., Kawai, K., Yamahara, J., and Mori, H. (1998) Cancer Lett. 127, 177–183. 18. Kamei, H., Koide, T., Kojima, T., Hashimoto, Y., and Hasegawa, M. (1998) Cancer Biotherm. Radiopharmacol. 13, 185–188. 19. Hennig, L., Christner, C., Kipping, M., Schelbert, B., Rucknagel, K. P., Grabley, S., Kullertz, G., and Fischer, G. (1998) Biochemistry 37, 5953–5960. 20. Duroux, L., Delmotte, F. M., Lancelin, J. M., Keravis, G., and Jay-Allemand, C. (1998) Biochem. J. 333, 275–283. 21. Duhaiman, A. S. (1996) Biochem. Biophys. Res. Commun. 218, 648 – 652. 22. Floreani, M., Forlin, A., Bellin, S., and Carpenedo, F. (1995) Biochem. Mol. Biol. Int. 37, 757–763. 23. Vos, R. M., Van Ommen, B., Hoekstein, M. S., De Goede, J. H., and Van Bladeren, P. J. (1989) Chem. Biol. Interact. 71, 381–392. 24. Chao, S. H., Greenleaf, A. L., and Price, D. H. (2001) Nucleic Acids Res. 29, 767–773. 25. Haubrich, D. R., and Wang, P. F. (1976) Biochem. Pharmacol. 25, 669 – 672. 26. Inouye, K., Saito, A., Orita, M., Tonomura, B., Imaishi, H., and Ohkawa, H. (2000) J. Biochem. (Tokyo) 127, 1041–1046. 27. Rao, P., and Zigler, J. S. (1992) Biochim. Biophys. Acta 1116, 75– 81. 28. Varga, Z., Bene, L., Pieri, C., Damjanovich, S., and Gaspar, R. (1996) Biochem. Biophys. Res. Commun. 218, 828 – 832.