- Email: [email protected]
Glycation of Hepatocyte Cytosolic Proteins in Streptozotocin-Induced Diabetic Rats
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
229, 952–958 (1996)
1907
Glycation of Hepatocyte Cytosolic Proteins in Streptozotocin-Induced Diabetic Rats A. Gugliucci1 and M-F. Allard Department of Anatomy, University of Montreal, CP 6128, Succ Centre Ville, H3C 3J7, Montreal, Quebec, Canada Received November 13, 1996 A role for glycation in diabetic pathology appears beyond doubt and one of the present trends is to focus the poorly explored field of intracellular glycation. In this work we studied the pattern of early glycation in hepatocyte cytosolic proteins from streptozotocin-induced diabetic rats (nÅ14) compared to control animals (nÅ8). Glycated proteins were present in the cytosol of control rats and increased three-fold after one month of diabetes, while glycated Hb and glycated plasma proteins rose two- and three-fold, respectively. A good correlation (rÅ0.82, põ0.001) was found between glycated cytosolic proteins and glycated plasma proteins, suggesting the latter could provide an indirect indication of intracellular glycation. Using PBA affinity chromatography followed by SDS-PAGE we detected 7 major glycated bands in cytosols from control animals which increased dramatically in diabetic rats. Moreover, other glycated proteins, which were undetectable in control animals, became prominent, and more than 15 major bands can thus be resolved. No major differences in the patterns can be seen after 1, 5, or 12 months of diabetes, suggesting that early glycation in cytosolic proteins reaches an equilibrium in a short period of one to two weeks (further supported by the tight correlation with glycated plasma proteins). Through comparison of the patterns obtained with an antiglucytollysine antibody on Western blots with those of silver stained gels from the PBA eluates we present evidence that intracellular glycation is mediated by glucose but mainly by other sugars. q 1996 Academic Press
Structural and functional modifications of proteins as a result of the glycation reaction have been extensively studied for the past twenty years (1). Most efforts concentrated on glucose as the main glycating agent, and considering the low rate of the glycation reaction by this sugar (weeks) attention was naturally drawn to circulating or long-lived proteins, either extracellular or from non-insulin dependent tissues (2, 3, 4, 5). However, recent reports from Brownlee’s group demonstrate intracellular glycation through incubation of endothelial cells in 30 mmol/l glucose for only one week. These cell culture experiments indicate therefore, that even advanced glycation, though formerly believed to occur in the long term on extracellular matrix, can occur intracellularly in short-lived proteins and even modify biological activities (1, 2, 6). Our own recent work demonstrated that histones from the liver of diabetic rats showed glycation levels three-fold higher than those of their age-matched controls (7). Based on this evidence we decided to tackle the issue of glycation of short-lived proteins, which is the case for most cytosolic proteins. In this work we studied hepatocyte cytosolic proteins since, given the metabolic characteristics of these cells, they offer a vulnerable target for glycation by glucose and other sugars. For that purpose, an experimental diabetes was induced by streptozotocin on rats which were maintained hyperglycemic for different periods. We then studied the pattern of early glycation in hepatocyte cytosolic proteins from thses animals as compared to 1
Corresponding author. Fax: (1514) 343 2459. E-mail: [email protected]. Abbreviations: ECL, enhanced chemiluminiscence; PBA, phenylboronate; SDS-PAGE, sodium dodecyl sulfate polyacrylamide electrophoresis. 952 0006-291X/96 $18.00 Copyright q 1996 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 5868
/
6914$$$861
12-06-96 10:29:27
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
age-matched control animals and correlated these findings with parameters of extracellular glycation and glycemic control. MATERIALS AND METHODS Apparatus. Ultracentrifugation was carried out in a 50 Ti rotor, in a Beckman XL Ultracentrifuge (Bioanalytical System Group, Mississauga, Ont, Canada); other centrifugations were performed in a Biofuge 22 R refrigerated centrifuge (Baxter Canlab, Mississauga, Ont, Canada). Spectrophotometric measurements were made in a Shimadzu UV 160 U Double Beam Recorder Spectrophotometer (RPI Instruments Inc., Montreal, Quebec, Canada). Proteins were measured according to Bradford (Bio Rad, Mississauga Ont, Canada). Experimental diabetes model. Diabetes was induced in male Sprague-Dawley (nÅ18) rats weighing 100-120 g by a single i.p. injection of 70 mg/kg body weight of streptozotocin in citrate buffer. Control animals (nÅ10) were shaminjected with citrate buffer only. Monitoring i) Diabetic state was confirmed and monitored by glycemia (weekly), glycosuria (daily) (Ames, Ontario, Canada), glycated hemoglobin and glycated plasma proteins , measured by phenylboronate (PBA) affinity chromatography on GlycoTest II mini columns from Pierce (Pierce, Chromatographic Specialties Inc, Brockville, Ont, Canada) (8, 9). Rats were killed after one month (nÅ14), 5 months (nÅ2) and 12 (nÅ2) months and hepatocytes were prepared as reported before. (10). Cytosol fractions were obtained essentially as previously described (11, 12). Glycated proteins in hepatocyte cytosol. For the analysis of early glycated protein profiles, hepatic cytosol fractions (500 mg protein) were first loaded onto PBA columns. Only proteins bearing early glycation or Amadori adducts are thus recognized and retained (13). After elution the whole glycated fraction was concentrated by ultrafiltration on Centricon 3, precipitated on cold acetone, loaded onto 15% acrylamide gels and processed as described below. Glycated hemoglobin and plasma proteins. Blood was drawn at sacrifice from inferior vena cava, collected on heparin and centrifuged at 800 1 g, at 4 7C for 15 min. Glycated proteins were determined on separated plasma by PBA chromatography; glycated Hb was measured on hemolysates by the same technique (13). SDS-PAGE. Electrophoresis was run according to Laemmli (14) on 15% acrylamide gels under reducing conditions for cytosolic fractions or on 7.5% gels (non reducing conditions) for plasma fractions. Equipment employed was Mini Gel from BioRad (BioRad Laboratories, Mississauga, Ont, Canada). After electrophoresis; bands were revealed by Silver staining (15, 16). Immunochemical detection of glucose-derived early glycation products. Upon reduction with NaBH4 glucose-derived early glycation products become glucitollysine. In parallel to the aforementioned experiments, 40 mg of total cytosolic fractions were electrophoresed and Western blotted (17). Early glucose-derived glycation adducts, were detected with a monoclonal anti-glucitollysine antibody (18, 19). Immunoblots were revealed by enhanced chemiluminiscence (ECL) from Boeringher Mannheim using anti-mouse peroxidase labelled second antibody (Amersham, Ont, Canada). Statistical analysis. Unless otherwise stated, data are expressed as mean { SD. Comparisons between data were performed by the Student’s t test (two-tailed) for unpaired samples. Correlations between different parameters were calculated by linear regression. Data were processed on StatWorks from Cricket Software, Philadelphia.
RESULTS AND DISCUSSION
In order to study the effects of diabetes on intracellular glycation we induced diabetes in male Sprague-Dawley rats (100-120g) by a single injection of streptozotocin. All the animals showed hyperglycemia and glycosuria 48 h after the injection. Animals were monitored daily for glycosuria which was always above 52 mmol/l. As in this work we focus on early glycation of cytosolic proteins for which a rapid turnover is expected, we killed most of our animals after one month of diabetes. At that time, glycemia was 23.6 { 1.1 mmol/l for diabetic animals (nÅ14) vs 5.6 { 0.6 mmol/l for controls (nÅ8), põ 0.0001. One month after the injections diabetic animals (nÅ14) weighed 225 { 15 g vs 362 { 14 g for controls (nÅ8), põ 0.0001. Glycated hemoglobin, glycated plasma proteins and glycated hepatocyte cytosol proteins were quantitated in control and diabetic animals by using the same technique of affinity chromatography, as depicted in Figure 1. Glycated hemoglobin (Fig1a) and glycated plasma proteins (Fig1b), are employed as markers of average glycemic control , one month and more for Hb and two weeks and less for plasma proteins (8, 9). Glycated hepatocyte cytosol proteins in diabetic rats and age-matched controls is depicted in (Fig1c). Glycated proteins are present in the cytosol of control rats, while one month of experimental diabetes induced a three-fold rise in their concentration. As glucose transport is mostly insulin independent in hepatocytes, which bear Glut II and Glut I glucose transporters (20, 21), this important increase in protein 953
AID
BBRC 5868
/
6914$$$861
12-06-96 10:29:27
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Glycated proteins in control and diabetic rats. Glycated hemoglobin (a), plasma proteins (b), and hepatocyte cytosol proteins (c) were quantitated by phenylboronate affinity chromatography in control (nÅ8) and streptozotocininduced diabetic rats kept hyperglycemic for one month (nÅ14). Glycated hemoglobin and glycated plasma proteins, markers of average glycemic control, were respectively increased two- and three-fold in diabetic animals. Glycated proteins in hepatocyte cytosol showed rised three-fold in diabetic rats compared to controls. 954
AID
BBRC 5868
/
6914$$5868
12-06-96 10:29:27
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. Correlation between glycation of hepatocyte cytosol proteins and markers of glycemic control. Glycated hemoglobin, plasma proteins, and hepatocyte cytosol proteins were quantitated by phenylboronate affinity chromatography in streptozotocin-induced diabetic rats kept hyperglycemic for one month (nÅ14). A good significant correlation between intracellular early glycation and a short term marker of extracellular glycation, glycated plasma proteins is apparent in (a). When correlation is performed against a long term marker of glycemic control, glycated hemoglobin, this correlation is poorer (b).
glycation in cytosol should be interpreted mainly as a deficiency in insulin effects on sugar intermediate metabolism. In fact, it is expected that lack of appropriate insulin secretion leading to decreased rates of glycogenesis and increased glycogenolysis produce shifts in steady state concentrations of free glucose and on metabolites derived from key irreversible reactions, such as glucose-6-phosphate and other intermediates (22, 23). On the other hand, the increase in glycation of cytosolic hepatocyte proteins we found is in agreement with the qualitative data from another report showing increased glycation of crude total liver homogenate proteins in diabetic rats detected by immunochemistry (24). As expected, glycated Hb -a marker of average glycemic control- was two-fold higher in diabetic rats (Fig1a), values which compare well with those usually encountered in poorly controlled diabetic patients (8, 25, 26). Glycated plasma proteins, averaging glycemic control for a shorter period (one-two weeks) showed in turn a threefold increase in concentration in diabetic vs control animals (Fig1b). In order to establish whether a link between extracellular and intracellular glycation is apparent we performed correlations between glycated plasma proteins and glycated cytosolic proteins. Figure 2a shows a good significant correlation between these two parameters. When correlation is performed against a long term marker of glycemic control, glycated hemoglobin, this correlation is poorer (Fig 2b). These results are consistent with the fact that both cytosolic and plasma proteins have short half-lives, while Hb is longerlived. The good correlation found also suggests that glycated plasma proteins measurement 955
AID
BBRC 5868
/
6914$$$861
12-06-96 10:29:27
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 3. Electrophoretic profile on SDS-PAGE of glycated plasma proteins in control and diabetic rats. Plasma proteins (250 mg) were loaded onto a PBA column and the whole glycated fractions were concentrated by ultrafiltration on Centricon 3, precipitated on cold acetone, and loaded onto 7.5% acrylamide gels under non reducing conditions. Coloration was performed by silver staining. 1-4 diabetic rats (one month of hyperglycemia); Cont, control animal. Diabetic rats display a dramatic increase in glycated plasma proteins, namely for albumin (Ab), IgG, and high molecular weight proteins (HMW) which comprise mainly lipoproteins, IgM, and alpha -2-macroglobulin.
could provide an indirect indication of intracellular glycation. A former study concerning whole kidney and liver plasma membrane glycation had shown similar evidence (27). Glycation was already present in plasma proteins from control animals and increased in diabetic rats as shown in Figure 3 which depicts the electrophoretic profile on SDS-PAGE. Main glycated proteins were albumin (Ab), IgG and a dramatic increase for proteins over 200 kDa is apparent in agreement with data from human diabetes (26, 28). These include lipoproteins, IgM and alpha -2-macroglobulin which show little or no glycation in control animals (26). Figure 4 shows electrophoretic profiles of glycated hepatocyte cytosol proteins in diabetic rats and age-matched controls. Figure 4A displays typical profiles of fractions obtained from
FIG. 4. Electrophoretic profile on SDS-PAGE of glycated hepatocyte cytosol proteins in control and diabetic rats. Hepatocyte cytosol proteins (500 mg) were loaded onto a PBA column and the whole glycated fractions were concentrated by ultrafiltration on Centricon 3, precipitated on cold acetone, and loaded onto 15% acrylamide gels and run under reducing conditions. (A) 1-6 diabetic rats (one month of hyperglycemia); Cont, typical control animal. Coloration was performed by silver staining; 7 typical Western blotting of cytosolic proteins (40 mg) from a diabetic rat (number 6), detection was performed with an antiglucytollysine antibody and ECL. (B) Effect of duration of hyperglycemia in glycated protein profile. Cont, control animal (6 months); 1, diabetic rat (five months of hyperglycemia); 2, diabetic rat (twelve months of hyperglycemia). 956
AID
BBRC 5868
/
6914$$$861
12-06-96 10:29:27
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
rats sacrifized after one month of diabetes as well as from their age-matched controls. Figure 4B shows the patterns for rats sacrifized after 5 months (4B.1) or 12 months (4B.2) of diabetes. Using PBA affinity chromatography followed by SDS-PAGE we detect 7 major glycated bands in cytosols from control animals (at approximately 52, 46, 34, 28, 27, 22 and 16 kDa, respectively). This pattern does not change with the age of the animal since it appears essentially the same for either one month (Figure 4A) or one year control animals (Figure 4B). When diabetic animals are considered, a dramatic increase in the glycated bands already present in controls is observed. Moreover, glycated proteins which are undetectable or present in only trace amounts in control animals, become prominent, and more than 15 major bands can thus be resolved. Similar profiles were obtained in all cases (nÅ14) while the quantitative and correlation data presented above were in agreement with the profiles obtained. In order to better illustrate this issue, we show typical runs from 6 different diabetic animals. Rats 2-4 display the heaviest glycation and were precisely those animals with poorer glycemic control as judged by plasma protein glycation, weight loss, and the presence of cataracts (a major metabolic complication in which glycation and the polyol pathway play a key role), even after only one month of hyperglycemia. In order to ascertain whether further modifications of the pattern of early glycation of proteins occurred with the duration of diabetes we studied animals killed after 5 months (Fig 4, B1) or 12 months (Fig 4, B2) of diabetes. No further important modifications in the major bands patterns (no greater than the inter-individual variation found in profiles from one month diabetic rats) can be seen. These data are consistent with the fact that early glycation in cytosolic proteins reaches an equilibrium in a short period of one-two weeks an issue which is further supported by the correlation we found with glycated plasma proteins. This steady-state which is equally apparent at 1, 5 or 12 months of diabetes is due to the intrinsic turnover of cytosolic proteins which prevents the accumulation of glycated sites beyond the level reached in the first weeks. Most of early glycation products (whether derived from glucose, glucose-6-phosphate, fructose) share a common coplanar cis-diol moiety which permits to isolate and quantify them by affinity chromatography (8). In order to gain further insight as to which proteins were effectively glycated by glucose and which by other sugars, in parallel with the affinity chromatography experiments, we specifically detected glucose adducts with a monoclonal antiglucytollysine antibody. No signal could be found for control animals (data not shown), showing that in the euglycemic state, glycation of cytosolic proteins by glucose is under the limit of detection of ECL. Conversely, a strong signal was obtained for diabetic animals as seen in Figure 4A.7 where we depict a typical blot from a diabetic rat. When one compares the patterns obtained with the antiglucytollysine antibody (Figure 4A.7) on Western blots with those of silver stained gels from the PBA eluates (Figure 4A.1-6) several differences can be substantiated. Some of the bands are present in both cases (in particular those over 40 kDa). Many others (under 40 kDa) are completely absent in Western blots but are conspicuous in PBA eluates. These differences can then be ascribed to those proteins modified by other sugars than glucose. As glucose concentration is low in cytosol, great attention is drawn at present to glycolytic intermediates as putative candidates for intracellular glycation agents. Glucose-6-phosphate, for instance, is indeed several-fold more potent than glucose (29, 30) and has been shown to alter DNA mutation rate in prokaryotic strains which accumulate this intermediate (29,31). Moreover, the sorbitol pathway generates fructose which is also more reactive than glucose (30, 32), providing a link between the polyol pathway and glycation. In this regard, our data, although indirect, are consistent with and support an important role of sugars other than glucose in intracellular glycation. Particularly interesting in this regard is the main band seen in control and diabetic animals, at about 16 kDa. Its strong glycation and apparent molecular mass could be interpreted a priori as indications of it being hemoglobin chains. However, its complete absence in Western blots strongly militates against this and rather indicates it should be a 957
AID
BBRC 5868
/
6914$$$861
12-06-96 10:29:27
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
different, heavily glycated polypeptide which reacted with sugars different from glucose. It is quite possible that some of the glycated proteins we have found in hepatocyte cytosol are enzymes precisely involved in glucose metabolism such as glucose-6-phosphate dehydrogenase, glucokinase or aldehyde reductase, and for which in vitro glycation has been previously shown to produce deleterious effects (33, 34, 35). In conclusion, this study provides evidence for a rapid, steady-state early glycation of hepatocyte cytosol proteins during experimental diabetes. Our data suggest this intracellular glycation is not only mediated by glucose but by other sugars as well. Cytosol protein glycation correlates tightly with plasma protein glycation. Further studies are needed to ascertain the nature of the glycated polypeptides and the eventual functional effect of glycation. An effort to identify the main glycated proteins by 2D electrophoresis and sequencing is currently under way in our laboratory. ACKNOWLEDGMENTS This work was supported by the Association Diabe`te Que´bec (ADQ). M-FA. was a fellow from ADQ. The monoclonal antiglucytollysine antibody G8C11 was kindly provided by the Scripps Research Institute (La Jolla, CA).
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35.
Brownlee, M. (1995) Annu. Rev. Med. 46, 223–234. Brownlee, M. (1994) Diabetes 43, 836–841. Smith, M., Monnier, V., et al. (1995) Neuroreport 6, 1595. Monnier, V. (1989) in Proceedings of the NIH Conference on the Maillard Reaction in Aging, Diabetes and Nutrition (Baynes, J. W., and Monnier, V. M., Eds.), pp. 1–22, Liss, New York. Monnier, V., Sell, D., et al. (1992) Diabetes 41, 36–41. Giardino, I., Edelstein, D., and Brownlee, M. (1994) J. Clin. Invest. 94, 110–117. Gugliucci, A., et al. (1995) Biochem. Biophys. Res. Commun. 212, 56–62. Armbruster, D. A. (1987) Clin. Chem. 33, 2157–2163. Benjamin, R., and Sacks, D. (1994) Clin. Chem. 40, 683–687. Ose, L., Roken, I., et al. (1981) Scand. J. Clin. Lab. Invest. 41, 63–73. Fleischer, S., and Kervina, M. (1974) Methods Enzymol. 31, 6–40. Jarret, L. (1974) Methods Enzymol. 31, 60–74. Shaklai, N., Garlick, R. L., and Bunn, H. F. (1984) J. Biol. Chem. 259, 3812–3817. Laemmli, U. K. (1970) Nature 227, 680–685. Merril, C. R., Goldman, D., and Van Keuren, M. L. (1984) Meth. Enzymol. 104, 441–446. Rabilloud, T., Vuillard, L., et al. (1994) Cell. Mol. Biol. 40, 57–75. Bjerrum, O. J., and Schafer-Nielsen, C. (1986) in Analytical Electrophoresis (Dun, M. J., Ed.), pp. 315, Verlag Chemie, Weinheim. Curtiss, L. K., and Witztum, J. L. (1985) Diabetes 34, 452–461. Curtiss, L. K., Pepe, M. G., and Witztum, J. L. (1986) Methods Enzymol. 128, 354–364. Baldwin, S. (1993) Biochim. Biophys. Acta 1154, 17–49. Kahn, B. (1992) J. Clin. Invest. 89, 1367–1374. Lorenzi, M. (1992) Diabetes/Metab. Rev. 8, 85–103. Brownlee, M., and Cerami, A. (1981) Annu. Rev. Biochem. 50, 385–432. Myint, T., Hoshi, S., et al. (1995) Biochimica et Biophysica Acta 1272, 73–79. Koenig, R. J., Peterson, C. M., et al. (1976) N. Engl. J. Med. 295, 417–420. Menini, T., Gugliucci, A., et al. (1993) Ann. Biol. Clin. 51, 887–891. Cefalu, W., Wang, Z., et al. (1991) Diabetes 40, 902–907. Gugliucci, A., Menini, T., and Stahl, A. J. C. (1994) Glycosylation and Disease 1, 177–184. Bucala, R., Model, P., et al. (1985) Proc. Natl. Acad. Sci. USA 82, 8439–8442. Gugliucci, A. (1994) Biochem. Biophys. Res. Commun. 203, 588–593. Bucala, R., Lee, A. T., et al. (1993) Proc. Natl. Acad. Sci. USA 90, 2666–2670. McPherson, J., Shilton, B., and Walton, D. (1988) Biochemistry 27, 1901–1907. Ganea, E., and Harding, J. J. (1994) Biochem. Soc. Transac. 22, 445S. Zahner, D., Ramirez, R., and Malaisse, W. J. (1990) Diabetes Research 14, 109–115. Takahashi, M., Lu, Y. B., et al. (1995) Biochemistry 34, 1433–1438.
958
AID
BBRC 5868
/
6914$$$861
12-06-96 10:29:27
bbrca
229, 952–958 (1996)
1907
Glycation of Hepatocyte Cytosolic Proteins in Streptozotocin-Induced Diabetic Rats A. Gugliucci1 and M-F. Allard Department of Anatomy, University of Montreal, CP 6128, Succ Centre Ville, H3C 3J7, Montreal, Quebec, Canada Received November 13, 1996 A role for glycation in diabetic pathology appears beyond doubt and one of the present trends is to focus the poorly explored field of intracellular glycation. In this work we studied the pattern of early glycation in hepatocyte cytosolic proteins from streptozotocin-induced diabetic rats (nÅ14) compared to control animals (nÅ8). Glycated proteins were present in the cytosol of control rats and increased three-fold after one month of diabetes, while glycated Hb and glycated plasma proteins rose two- and three-fold, respectively. A good correlation (rÅ0.82, põ0.001) was found between glycated cytosolic proteins and glycated plasma proteins, suggesting the latter could provide an indirect indication of intracellular glycation. Using PBA affinity chromatography followed by SDS-PAGE we detected 7 major glycated bands in cytosols from control animals which increased dramatically in diabetic rats. Moreover, other glycated proteins, which were undetectable in control animals, became prominent, and more than 15 major bands can thus be resolved. No major differences in the patterns can be seen after 1, 5, or 12 months of diabetes, suggesting that early glycation in cytosolic proteins reaches an equilibrium in a short period of one to two weeks (further supported by the tight correlation with glycated plasma proteins). Through comparison of the patterns obtained with an antiglucytollysine antibody on Western blots with those of silver stained gels from the PBA eluates we present evidence that intracellular glycation is mediated by glucose but mainly by other sugars. q 1996 Academic Press
Structural and functional modifications of proteins as a result of the glycation reaction have been extensively studied for the past twenty years (1). Most efforts concentrated on glucose as the main glycating agent, and considering the low rate of the glycation reaction by this sugar (weeks) attention was naturally drawn to circulating or long-lived proteins, either extracellular or from non-insulin dependent tissues (2, 3, 4, 5). However, recent reports from Brownlee’s group demonstrate intracellular glycation through incubation of endothelial cells in 30 mmol/l glucose for only one week. These cell culture experiments indicate therefore, that even advanced glycation, though formerly believed to occur in the long term on extracellular matrix, can occur intracellularly in short-lived proteins and even modify biological activities (1, 2, 6). Our own recent work demonstrated that histones from the liver of diabetic rats showed glycation levels three-fold higher than those of their age-matched controls (7). Based on this evidence we decided to tackle the issue of glycation of short-lived proteins, which is the case for most cytosolic proteins. In this work we studied hepatocyte cytosolic proteins since, given the metabolic characteristics of these cells, they offer a vulnerable target for glycation by glucose and other sugars. For that purpose, an experimental diabetes was induced by streptozotocin on rats which were maintained hyperglycemic for different periods. We then studied the pattern of early glycation in hepatocyte cytosolic proteins from thses animals as compared to 1
Corresponding author. Fax: (1514) 343 2459. E-mail: [email protected]. Abbreviations: ECL, enhanced chemiluminiscence; PBA, phenylboronate; SDS-PAGE, sodium dodecyl sulfate polyacrylamide electrophoresis. 952 0006-291X/96 $18.00 Copyright q 1996 by Academic Press All rights of reproduction in any form reserved.
AID
BBRC 5868
/
6914$$$861
12-06-96 10:29:27
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
age-matched control animals and correlated these findings with parameters of extracellular glycation and glycemic control. MATERIALS AND METHODS Apparatus. Ultracentrifugation was carried out in a 50 Ti rotor, in a Beckman XL Ultracentrifuge (Bioanalytical System Group, Mississauga, Ont, Canada); other centrifugations were performed in a Biofuge 22 R refrigerated centrifuge (Baxter Canlab, Mississauga, Ont, Canada). Spectrophotometric measurements were made in a Shimadzu UV 160 U Double Beam Recorder Spectrophotometer (RPI Instruments Inc., Montreal, Quebec, Canada). Proteins were measured according to Bradford (Bio Rad, Mississauga Ont, Canada). Experimental diabetes model. Diabetes was induced in male Sprague-Dawley (nÅ18) rats weighing 100-120 g by a single i.p. injection of 70 mg/kg body weight of streptozotocin in citrate buffer. Control animals (nÅ10) were shaminjected with citrate buffer only. Monitoring i) Diabetic state was confirmed and monitored by glycemia (weekly), glycosuria (daily) (Ames, Ontario, Canada), glycated hemoglobin and glycated plasma proteins , measured by phenylboronate (PBA) affinity chromatography on GlycoTest II mini columns from Pierce (Pierce, Chromatographic Specialties Inc, Brockville, Ont, Canada) (8, 9). Rats were killed after one month (nÅ14), 5 months (nÅ2) and 12 (nÅ2) months and hepatocytes were prepared as reported before. (10). Cytosol fractions were obtained essentially as previously described (11, 12). Glycated proteins in hepatocyte cytosol. For the analysis of early glycated protein profiles, hepatic cytosol fractions (500 mg protein) were first loaded onto PBA columns. Only proteins bearing early glycation or Amadori adducts are thus recognized and retained (13). After elution the whole glycated fraction was concentrated by ultrafiltration on Centricon 3, precipitated on cold acetone, loaded onto 15% acrylamide gels and processed as described below. Glycated hemoglobin and plasma proteins. Blood was drawn at sacrifice from inferior vena cava, collected on heparin and centrifuged at 800 1 g, at 4 7C for 15 min. Glycated proteins were determined on separated plasma by PBA chromatography; glycated Hb was measured on hemolysates by the same technique (13). SDS-PAGE. Electrophoresis was run according to Laemmli (14) on 15% acrylamide gels under reducing conditions for cytosolic fractions or on 7.5% gels (non reducing conditions) for plasma fractions. Equipment employed was Mini Gel from BioRad (BioRad Laboratories, Mississauga, Ont, Canada). After electrophoresis; bands were revealed by Silver staining (15, 16). Immunochemical detection of glucose-derived early glycation products. Upon reduction with NaBH4 glucose-derived early glycation products become glucitollysine. In parallel to the aforementioned experiments, 40 mg of total cytosolic fractions were electrophoresed and Western blotted (17). Early glucose-derived glycation adducts, were detected with a monoclonal anti-glucitollysine antibody (18, 19). Immunoblots were revealed by enhanced chemiluminiscence (ECL) from Boeringher Mannheim using anti-mouse peroxidase labelled second antibody (Amersham, Ont, Canada). Statistical analysis. Unless otherwise stated, data are expressed as mean { SD. Comparisons between data were performed by the Student’s t test (two-tailed) for unpaired samples. Correlations between different parameters were calculated by linear regression. Data were processed on StatWorks from Cricket Software, Philadelphia.
RESULTS AND DISCUSSION
In order to study the effects of diabetes on intracellular glycation we induced diabetes in male Sprague-Dawley rats (100-120g) by a single injection of streptozotocin. All the animals showed hyperglycemia and glycosuria 48 h after the injection. Animals were monitored daily for glycosuria which was always above 52 mmol/l. As in this work we focus on early glycation of cytosolic proteins for which a rapid turnover is expected, we killed most of our animals after one month of diabetes. At that time, glycemia was 23.6 { 1.1 mmol/l for diabetic animals (nÅ14) vs 5.6 { 0.6 mmol/l for controls (nÅ8), põ 0.0001. One month after the injections diabetic animals (nÅ14) weighed 225 { 15 g vs 362 { 14 g for controls (nÅ8), põ 0.0001. Glycated hemoglobin, glycated plasma proteins and glycated hepatocyte cytosol proteins were quantitated in control and diabetic animals by using the same technique of affinity chromatography, as depicted in Figure 1. Glycated hemoglobin (Fig1a) and glycated plasma proteins (Fig1b), are employed as markers of average glycemic control , one month and more for Hb and two weeks and less for plasma proteins (8, 9). Glycated hepatocyte cytosol proteins in diabetic rats and age-matched controls is depicted in (Fig1c). Glycated proteins are present in the cytosol of control rats, while one month of experimental diabetes induced a three-fold rise in their concentration. As glucose transport is mostly insulin independent in hepatocytes, which bear Glut II and Glut I glucose transporters (20, 21), this important increase in protein 953
AID
BBRC 5868
/
6914$$$861
12-06-96 10:29:27
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Glycated proteins in control and diabetic rats. Glycated hemoglobin (a), plasma proteins (b), and hepatocyte cytosol proteins (c) were quantitated by phenylboronate affinity chromatography in control (nÅ8) and streptozotocininduced diabetic rats kept hyperglycemic for one month (nÅ14). Glycated hemoglobin and glycated plasma proteins, markers of average glycemic control, were respectively increased two- and three-fold in diabetic animals. Glycated proteins in hepatocyte cytosol showed rised three-fold in diabetic rats compared to controls. 954
AID
BBRC 5868
/
6914$$5868
12-06-96 10:29:27
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. Correlation between glycation of hepatocyte cytosol proteins and markers of glycemic control. Glycated hemoglobin, plasma proteins, and hepatocyte cytosol proteins were quantitated by phenylboronate affinity chromatography in streptozotocin-induced diabetic rats kept hyperglycemic for one month (nÅ14). A good significant correlation between intracellular early glycation and a short term marker of extracellular glycation, glycated plasma proteins is apparent in (a). When correlation is performed against a long term marker of glycemic control, glycated hemoglobin, this correlation is poorer (b).
glycation in cytosol should be interpreted mainly as a deficiency in insulin effects on sugar intermediate metabolism. In fact, it is expected that lack of appropriate insulin secretion leading to decreased rates of glycogenesis and increased glycogenolysis produce shifts in steady state concentrations of free glucose and on metabolites derived from key irreversible reactions, such as glucose-6-phosphate and other intermediates (22, 23). On the other hand, the increase in glycation of cytosolic hepatocyte proteins we found is in agreement with the qualitative data from another report showing increased glycation of crude total liver homogenate proteins in diabetic rats detected by immunochemistry (24). As expected, glycated Hb -a marker of average glycemic control- was two-fold higher in diabetic rats (Fig1a), values which compare well with those usually encountered in poorly controlled diabetic patients (8, 25, 26). Glycated plasma proteins, averaging glycemic control for a shorter period (one-two weeks) showed in turn a threefold increase in concentration in diabetic vs control animals (Fig1b). In order to establish whether a link between extracellular and intracellular glycation is apparent we performed correlations between glycated plasma proteins and glycated cytosolic proteins. Figure 2a shows a good significant correlation between these two parameters. When correlation is performed against a long term marker of glycemic control, glycated hemoglobin, this correlation is poorer (Fig 2b). These results are consistent with the fact that both cytosolic and plasma proteins have short half-lives, while Hb is longerlived. The good correlation found also suggests that glycated plasma proteins measurement 955
AID
BBRC 5868
/
6914$$$861
12-06-96 10:29:27
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 3. Electrophoretic profile on SDS-PAGE of glycated plasma proteins in control and diabetic rats. Plasma proteins (250 mg) were loaded onto a PBA column and the whole glycated fractions were concentrated by ultrafiltration on Centricon 3, precipitated on cold acetone, and loaded onto 7.5% acrylamide gels under non reducing conditions. Coloration was performed by silver staining. 1-4 diabetic rats (one month of hyperglycemia); Cont, control animal. Diabetic rats display a dramatic increase in glycated plasma proteins, namely for albumin (Ab), IgG, and high molecular weight proteins (HMW) which comprise mainly lipoproteins, IgM, and alpha -2-macroglobulin.
could provide an indirect indication of intracellular glycation. A former study concerning whole kidney and liver plasma membrane glycation had shown similar evidence (27). Glycation was already present in plasma proteins from control animals and increased in diabetic rats as shown in Figure 3 which depicts the electrophoretic profile on SDS-PAGE. Main glycated proteins were albumin (Ab), IgG and a dramatic increase for proteins over 200 kDa is apparent in agreement with data from human diabetes (26, 28). These include lipoproteins, IgM and alpha -2-macroglobulin which show little or no glycation in control animals (26). Figure 4 shows electrophoretic profiles of glycated hepatocyte cytosol proteins in diabetic rats and age-matched controls. Figure 4A displays typical profiles of fractions obtained from
FIG. 4. Electrophoretic profile on SDS-PAGE of glycated hepatocyte cytosol proteins in control and diabetic rats. Hepatocyte cytosol proteins (500 mg) were loaded onto a PBA column and the whole glycated fractions were concentrated by ultrafiltration on Centricon 3, precipitated on cold acetone, and loaded onto 15% acrylamide gels and run under reducing conditions. (A) 1-6 diabetic rats (one month of hyperglycemia); Cont, typical control animal. Coloration was performed by silver staining; 7 typical Western blotting of cytosolic proteins (40 mg) from a diabetic rat (number 6), detection was performed with an antiglucytollysine antibody and ECL. (B) Effect of duration of hyperglycemia in glycated protein profile. Cont, control animal (6 months); 1, diabetic rat (five months of hyperglycemia); 2, diabetic rat (twelve months of hyperglycemia). 956
AID
BBRC 5868
/
6914$$$861
12-06-96 10:29:27
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
rats sacrifized after one month of diabetes as well as from their age-matched controls. Figure 4B shows the patterns for rats sacrifized after 5 months (4B.1) or 12 months (4B.2) of diabetes. Using PBA affinity chromatography followed by SDS-PAGE we detect 7 major glycated bands in cytosols from control animals (at approximately 52, 46, 34, 28, 27, 22 and 16 kDa, respectively). This pattern does not change with the age of the animal since it appears essentially the same for either one month (Figure 4A) or one year control animals (Figure 4B). When diabetic animals are considered, a dramatic increase in the glycated bands already present in controls is observed. Moreover, glycated proteins which are undetectable or present in only trace amounts in control animals, become prominent, and more than 15 major bands can thus be resolved. Similar profiles were obtained in all cases (nÅ14) while the quantitative and correlation data presented above were in agreement with the profiles obtained. In order to better illustrate this issue, we show typical runs from 6 different diabetic animals. Rats 2-4 display the heaviest glycation and were precisely those animals with poorer glycemic control as judged by plasma protein glycation, weight loss, and the presence of cataracts (a major metabolic complication in which glycation and the polyol pathway play a key role), even after only one month of hyperglycemia. In order to ascertain whether further modifications of the pattern of early glycation of proteins occurred with the duration of diabetes we studied animals killed after 5 months (Fig 4, B1) or 12 months (Fig 4, B2) of diabetes. No further important modifications in the major bands patterns (no greater than the inter-individual variation found in profiles from one month diabetic rats) can be seen. These data are consistent with the fact that early glycation in cytosolic proteins reaches an equilibrium in a short period of one-two weeks an issue which is further supported by the correlation we found with glycated plasma proteins. This steady-state which is equally apparent at 1, 5 or 12 months of diabetes is due to the intrinsic turnover of cytosolic proteins which prevents the accumulation of glycated sites beyond the level reached in the first weeks. Most of early glycation products (whether derived from glucose, glucose-6-phosphate, fructose) share a common coplanar cis-diol moiety which permits to isolate and quantify them by affinity chromatography (8). In order to gain further insight as to which proteins were effectively glycated by glucose and which by other sugars, in parallel with the affinity chromatography experiments, we specifically detected glucose adducts with a monoclonal antiglucytollysine antibody. No signal could be found for control animals (data not shown), showing that in the euglycemic state, glycation of cytosolic proteins by glucose is under the limit of detection of ECL. Conversely, a strong signal was obtained for diabetic animals as seen in Figure 4A.7 where we depict a typical blot from a diabetic rat. When one compares the patterns obtained with the antiglucytollysine antibody (Figure 4A.7) on Western blots with those of silver stained gels from the PBA eluates (Figure 4A.1-6) several differences can be substantiated. Some of the bands are present in both cases (in particular those over 40 kDa). Many others (under 40 kDa) are completely absent in Western blots but are conspicuous in PBA eluates. These differences can then be ascribed to those proteins modified by other sugars than glucose. As glucose concentration is low in cytosol, great attention is drawn at present to glycolytic intermediates as putative candidates for intracellular glycation agents. Glucose-6-phosphate, for instance, is indeed several-fold more potent than glucose (29, 30) and has been shown to alter DNA mutation rate in prokaryotic strains which accumulate this intermediate (29,31). Moreover, the sorbitol pathway generates fructose which is also more reactive than glucose (30, 32), providing a link between the polyol pathway and glycation. In this regard, our data, although indirect, are consistent with and support an important role of sugars other than glucose in intracellular glycation. Particularly interesting in this regard is the main band seen in control and diabetic animals, at about 16 kDa. Its strong glycation and apparent molecular mass could be interpreted a priori as indications of it being hemoglobin chains. However, its complete absence in Western blots strongly militates against this and rather indicates it should be a 957
AID
BBRC 5868
/
6914$$$861
12-06-96 10:29:27
bbrca
Vol. 229, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
different, heavily glycated polypeptide which reacted with sugars different from glucose. It is quite possible that some of the glycated proteins we have found in hepatocyte cytosol are enzymes precisely involved in glucose metabolism such as glucose-6-phosphate dehydrogenase, glucokinase or aldehyde reductase, and for which in vitro glycation has been previously shown to produce deleterious effects (33, 34, 35). In conclusion, this study provides evidence for a rapid, steady-state early glycation of hepatocyte cytosol proteins during experimental diabetes. Our data suggest this intracellular glycation is not only mediated by glucose but by other sugars as well. Cytosol protein glycation correlates tightly with plasma protein glycation. Further studies are needed to ascertain the nature of the glycated polypeptides and the eventual functional effect of glycation. An effort to identify the main glycated proteins by 2D electrophoresis and sequencing is currently under way in our laboratory. ACKNOWLEDGMENTS This work was supported by the Association Diabe`te Que´bec (ADQ). M-FA. was a fellow from ADQ. The monoclonal antiglucytollysine antibody G8C11 was kindly provided by the Scripps Research Institute (La Jolla, CA).
REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35.
Brownlee, M. (1995) Annu. Rev. Med. 46, 223–234. Brownlee, M. (1994) Diabetes 43, 836–841. Smith, M., Monnier, V., et al. (1995) Neuroreport 6, 1595. Monnier, V. (1989) in Proceedings of the NIH Conference on the Maillard Reaction in Aging, Diabetes and Nutrition (Baynes, J. W., and Monnier, V. M., Eds.), pp. 1–22, Liss, New York. Monnier, V., Sell, D., et al. (1992) Diabetes 41, 36–41. Giardino, I., Edelstein, D., and Brownlee, M. (1994) J. Clin. Invest. 94, 110–117. Gugliucci, A., et al. (1995) Biochem. Biophys. Res. Commun. 212, 56–62. Armbruster, D. A. (1987) Clin. Chem. 33, 2157–2163. Benjamin, R., and Sacks, D. (1994) Clin. Chem. 40, 683–687. Ose, L., Roken, I., et al. (1981) Scand. J. Clin. Lab. Invest. 41, 63–73. Fleischer, S., and Kervina, M. (1974) Methods Enzymol. 31, 6–40. Jarret, L. (1974) Methods Enzymol. 31, 60–74. Shaklai, N., Garlick, R. L., and Bunn, H. F. (1984) J. Biol. Chem. 259, 3812–3817. Laemmli, U. K. (1970) Nature 227, 680–685. Merril, C. R., Goldman, D., and Van Keuren, M. L. (1984) Meth. Enzymol. 104, 441–446. Rabilloud, T., Vuillard, L., et al. (1994) Cell. Mol. Biol. 40, 57–75. Bjerrum, O. J., and Schafer-Nielsen, C. (1986) in Analytical Electrophoresis (Dun, M. J., Ed.), pp. 315, Verlag Chemie, Weinheim. Curtiss, L. K., and Witztum, J. L. (1985) Diabetes 34, 452–461. Curtiss, L. K., Pepe, M. G., and Witztum, J. L. (1986) Methods Enzymol. 128, 354–364. Baldwin, S. (1993) Biochim. Biophys. Acta 1154, 17–49. Kahn, B. (1992) J. Clin. Invest. 89, 1367–1374. Lorenzi, M. (1992) Diabetes/Metab. Rev. 8, 85–103. Brownlee, M., and Cerami, A. (1981) Annu. Rev. Biochem. 50, 385–432. Myint, T., Hoshi, S., et al. (1995) Biochimica et Biophysica Acta 1272, 73–79. Koenig, R. J., Peterson, C. M., et al. (1976) N. Engl. J. Med. 295, 417–420. Menini, T., Gugliucci, A., et al. (1993) Ann. Biol. Clin. 51, 887–891. Cefalu, W., Wang, Z., et al. (1991) Diabetes 40, 902–907. Gugliucci, A., Menini, T., and Stahl, A. J. C. (1994) Glycosylation and Disease 1, 177–184. Bucala, R., Model, P., et al. (1985) Proc. Natl. Acad. Sci. USA 82, 8439–8442. Gugliucci, A. (1994) Biochem. Biophys. Res. Commun. 203, 588–593. Bucala, R., Lee, A. T., et al. (1993) Proc. Natl. Acad. Sci. USA 90, 2666–2670. McPherson, J., Shilton, B., and Walton, D. (1988) Biochemistry 27, 1901–1907. Ganea, E., and Harding, J. J. (1994) Biochem. Soc. Transac. 22, 445S. Zahner, D., Ramirez, R., and Malaisse, W. J. (1990) Diabetes Research 14, 109–115. Takahashi, M., Lu, Y. B., et al. (1995) Biochemistry 34, 1433–1438.
958
AID
BBRC 5868
/
6914$$$861
12-06-96 10:29:27
bbrca