Inhibitory effect of albumin-derived advanced glycosylation products on PMA-induced superoxide anion production by rat macrophages

Life .Sciem,

Vol. 60,

No. 25, pp. 227%2Xt9~ 1997

CoWright 0 1997Ekvier ScienceInc. Printed in the ELSEVIER

USA. All rights reserved 0024-m/97 317.00 t .cm

PII SOO24-3205(97)002S3-X

INHIBITORY EFFECT OF ALBUMIN-DERIVED ADVANCED GLYCOSYLATION PRODUCTS ON PMA-INDUCED SUPEROXIDE ANION PRODUCTION BY RAT MACROPHAGES Remedies Ramhz,

Francisco J. Bedoya, M. Dolores Chiara and Francisco Sobrino’

Departamento de Bioqufmica M&&x y Biologfa Molecular, Universidad de Sevilla, Avda Sihhez Piqjuan 4,41009 Sevilla, Spain (Received in final form March 17,1997)

Advanced glycosylation end products (AGE) are implicated in many of the complications of diabetes. In the same way, infectious diseases are frequently associatedwith this disease. An impaired respiratory burst in macrophages may be a cause of infectious complications in diabetic patients. To establish a possible mechanism of this altered cell function, we have analyzed the effect of AGEmodified proteins on PMA-dependent superoxide anion production (4.3 from normal rat peritoneal macrophages. We have used AGE-modified bovine serum albumin (AGE-BSA) prepared by incubation with glucose. AGE-BSA partially inhibits the phorbol ester-dependent superoxide production by macrophages in vitro. The specificity of this inhibitory effect is demonstrated by the fact that aminoguanidine, an inhibitor of the formation of AGE products, fully prevents the effect of AGE-BSA in vitro. Macrophages from diabetic rats shown an inhibition on PMA dependent-q- production. However, the treatment in vivo with aminoguanidine produced a cancelation of the inhibitory effect observed in the diabetic state. These data suggest that AGE-modified proteins could be implicated in the impairement of macrophage respiratory burst in diabetes. kky Wordr: albumin-derived advance glycosylation products, diabetes, macrophages, superoxide anion

Infectious diseases are frequently associated with diabetes. Phagocytic cells like polymorphonuclear leucocytes and macrophages play a critical role in host defense against infection. The generation of reactive oxygen intermediates is implicated in the antimicrobial activity of these cells. A diminished production of these oxygen species has been detected in neutrophils from diabetic patients (1) and in alveolar macrophages from diabetic rats (2). Similarly, glucose inhibits in vitro the respiratory burst in neutrophils isolated from healthy subjects (3). Glucose reacts non-enzymatically with the ammo groups of a wide range of proteins to form Schiff bases which, through multiple rearrangements, are transformed into complex irreversible protein adducts termed advanced glycosylation end products (AGE) (4,5). ‘To whom correspondence should be adressed.

2280

AGE-BSA on Maaophage Superoxide Anions

Vol. 60, No. 25,1997

Accumulation of AGE formed in vivo on long-lived protein increases with age and is accelerated in patients with diabetes (6,7). Once formed, AGE are chemically and biologically active and are believed to contribute to irreversible tissue damage by covalent protein-to protein cross-linking (8 10). A membrane-associated macrophage receptor has been identified that specifically recognixes proteins to which AGEs are bound. This macrophage receptor has been extensively characterized, and it was recently isolated (11-13). The interaction of AGEmodified proteins with the macrophage receptor induces the synthesis and release of cytokines such as TNF and IL-l (14) and growth factors (PDGF and IGF-1) (15,16), suggesting that the AGE receptor system plays an important role in normal tissue remodeling. So far, little is known about the mechanism responsible for the abnormal macrophage function in diabetes. In this paper, we have studied the effect of AGE-modified bovine serum albumin (AGEBSA), and the effect of diabetes on macrophage respiratory burst.

Materials Bovine serum albumin (BSA), cytochrome c, Phorbol 1Zmyristate 13-acetate (PMA), streptoxotocin and xanthine were obtained from Sigma (Alcobendas, Spain). Peroxidase and xanthine oxidase were from Boehringer Mannheim (Barcelona, Spain). Aminoguanidinehydrochloride was provided by Fluke (Alcobendas, Spain). Luminol was obtained from Serva (Barcelona, Spain). Salts and other chemicals were from Merck (Barcelona, Spain).

Male Wistar rats weighing 170-180 g (7 week old) were used in these studies. Diabetes was induced by injection of streptoxotocin (60 mg STZ/kg body weight in citrate buffer, pH 4.5) via a lateral tail vein. Three days after the administration of STZ, the induction of diabetes was confirmed by measurement of urinary glucose. Diabetic animals were divided into two, groups, one of which was untreated for 6 weeks. The other group was treated with aminoguanidine-HCl 7.35 mmol/l given in the drinking water (17), which did not affect the water intake by the diabetic rats (approximately 250 ml/day). Both treated and untreated diabetic groups manifested similar levels of hyperglycemia (diabetic group: 33.9 f 1.1 mmol/l, aminoguanidine treated group: 35.4 f 1.4 mmolll vs control group: 9.7 f 2.9 mmoY1, n=lO). Preparation of albumin-derived advanced glycosylation end products (AGEBSA) Bovine serum albumin was dissolved at a concentration of 100 mg/ml in sodium phosphate buffer (0.5 mol/l, pH 7.2) alone, buffer plus D-glucose (200 mmol/l), or buffer plus D-glucose plus aminoguanidiie-HCl(200 mmol/l). Sodium axide (3 mmol/l) was added to prevent bacterial growth. All samples were incubated at 37OC for 0.5, 1.0, 1.5 and 2.0 weeks. The formation of AGE-BSA at different times of incubation was followed by measuring either the absorbance at 350 nm or the fluorescence at 370 nm excitation and 440 nm emission. Unreacted carbohydrate was removed before assay by extensive dialysis against PBS overnight. Test solutions were assayed without further dilution. Preparation of macrophages Macrophages were obtained from rat peritoneal exudate cells as described in (18). The viability of recovered cells was estimated from their ability to exclude Trypan blue and was always higher than 95 96.

Vol. 60, No. 25,1997

AGE-BSA on Macrophage Superoxide Anions

2281

Measurements of super&de anion production W- production was measured as described in (19). Briefly, macrophages (1-1.5x1@ cells/ml) were preincubated at 37OC in Krebs-Binger bicarbonate buffer containing 10 mmol/l glucose and the test solution for 15 min before the addition of cytochrome c (80 pmol/l) and PMA (100 nmol/l). Superoxide dismutase-inhibitable increase in absorbance at 550 nm was recorded continuously. The rate of 4’ generation was calculated by using a molar absorption coefficient of 21.1 mmol/ll cm-‘. a.- production in a cell free system was generated by a xanthin&amhine oxidase system consisting of 0.2 mmol/l xanthine, 2 mU xanthine oxidase in 20 mmol/l sodium carbonate pH 10. a- was detected by either cytochrome c reduction as described above or by chemiluminiscence following interaction with luminol(2OO pmol/l) in a reaction catalyxed by horseradish peroxidase (200 ccv). The effect of AGEBSA or BSA alone was analyzed by the addition of 200 or 500 ~1 (20-50 mg/ml) to the assay mixture. Chemiluminiscence was measured at 37OC in a luminometer (model LB 9500, Berthold). Protein determbtation The protein concentration was determined by the Lowry method (20) using BSA as standard.

Results Effect of AGEBSA on cytochrome c reduction Albumin-derived advanced glycosylation end products (AGE-BSA) directly reduced cytochrome c (Fig. 1). AGBBSA (20 mg/ml) formed at different times of incubation of BSA with glucose (under the conditions described in Methods) was incubated with cytochrome c in the absence of other stimuli or cells. A positive correlation between cytochrome c reduction and the amount of AGE-BSA present in the assay was observed. Under the assay conditions routinely used, 12% of cytochrome c was reduced directly by the AGEBSA generated following 2 weeks of incubation of BSA with glucose. However, the effect of unmodified BSA (bovine serum albumin incubated in the absence of glucose) on cytochrome c reduction was negligible. Effect of AGEBSA on 4’ production by rat macrophages We next studied the effect of AGE-BSA on respiratory burst of peritoneal macrophages stimulated with PMA. Since the velocity of cytochrome c reduction by w depends on the concentration of exogenous oxidized cytochrome c, a dose-dependency study of cytochrome c was performed. It was observed that the velocity of cytochrome c reduction by Q- reached a plateau at cytochrome c levels above 70 pmol/l (data not shown). Thus, it is clear that the concentration of oxidized cytochrome c that remained after AGE-BSA incubation (about 70 nmol/ml) coincides with optimal level, of cytochrome c to detect Q‘ in the assay. Since AGE-BSA reduces directly cytochrome c, the production of w by macrophages could be overstimated. Hence appropiate controls were routinely run when asses&g w generation by macrophages, and the rate of cytochrome c reduced by AGE-BSA in the absence of PMA was subs&acted from the overall rate of cytochrome c reduced by PMA-stimulated macrophages in the presence of AGE-BSA. As shown in Fig. 2, AGE-BSA inhibited PI&A-stimulated w- production in peritoneal macrophages and the inhibition was associated with the amount of AGE-BSA formed during the incubation of BSA with glucose. The inhibition of w production by advanced glycosylation products reached a maximal value (68.5% of control value) with 20 mg/ml AGE-BSA formed after 2 weeks of incubation of BSA with glucose. The rate of AGE-BSA formation (measured by fluorescence emission at 440 nm) followed a biphasic pattern with a slow phase during the first week and a rapid, linear phase during the second week of incubation. The presence of

2282

AGE-BSA on MacrophageSuperoxideAnions

Vol. 60, No. 25,1997

aminoguanidine during the incubation of ISA with glucose inhibited the formation of albuminderived advanced glycosylation products and prevented the inhibitory effect of AGE-BSA on Qproduction by rats macrophages.

0

0.5 1.0 1.5 Incubation Time (weeks)

2.0

Fig. 1. Reduction of cytochrome c by AGE-BSA. Unmodified BSA (0) or AGEBSA (0) formed at different times of incubation were incubated at a concentration of 20 mg/ml with cytochrome c in the absence of PMA and cells for 45 min. Values are means f SEM for two independent experiments performed in duplicate.

In order to discard a non-specific effect of AGE-BSA on PMA (e.g. non-specifically binding to phorbol ester) two experiments were performed. First, cells were activated with PMA (100 nmol/l) and AGE-BSA was added after 10 min of incubation. The respiratory burst triggered with PMA alone produced 63.9 nmolOi710 min per mg of protein. The subsequent addition of AGE-BSA produced a clear inhibition in the rate of Q.- production in the next 10 min (15.2 nmol @‘/lOmin per mg of protein. Second, cells were incubated with different doses of PMA. At 100 nmol/l and 200 nmol/l PMA, the inhibitory rate of AGE-BSA was similar (about 30%). At 20 nmol/l PMA a higher degree of inhibition was found (about 45-50 %). These data suggest that at the concentrations of PMA routinely used (100 nmoYl), a specific inhibitory effect of AGE-BSA on PMA-induced Oi- production predominates over the non-specific binding of AGE-BSA to PMA, which could have some relevance at low concentrations of PMA (e.g. 20 nmol/l). To analyze the potential scavenging effect of AGE-BSA, experiments in a cell free system were performed. When Oz.-is generated by xanthinekanthine oxidase system, unmodified BSA (50 mg/ml) inhibited Oz.-production by 13 % (3.08 f 0.10 nmol cytochrome c reduced /min vs 3.54 f 0.15 nmol cytochrome c reduced/min in control conditions, n=2). AGE-BSA (50 mg/ml) inhibited @‘ production by 44% (2.00 f 0.11 nmol cytochrome c reduced/min vs 3.54 f 0.15 nmol cytochrome c reduced /min in control conditions, n=2). When O;- was detected by chemiluminiscence, unmodified BSA (20 mglml) inhibited Oz.- induced light production by 88 96.

Vol. 60, No. 25,197

AGE-BSA on MacrophageSupexoxideAnions

2283

AGE-BSA (20 mg/ml) induced inhibition of light production was 9596, 30 seconds after starting the reaction; and no differences in the percentage of inhibition of light production were observed between BSA and AGE-BSA, 3 mm after starting the reaction. Hence, the magnitude of Q.scavenging by AGE-BSA and BSA is bigger when Oz.- is detected with luminol.

0

1.5 1.0 0.5 Incubation Time (weeks)

2.0

Fig. 2. Relationship between the formation of AGE-BSA and the 4.‘ production by rat macrophages. Formation of AGE-BSA in the abscence (A) or presence (A) of 200 mmol/l aminoguanidine. PMA-induced @- formation by macrophages incubated for 45 min with 20 mg/ml of AGE-BSA formed in the absence (0) or presence (0) of aminoguanidine. In control conditions, the production of 9.‘ was 334.2 f 17.8 nmoV45 min per mg of protein. This value was normalized to 100%. Data are means f SEM for two separate experiments performed in duplicate.

In order to analyze a dose-dependent effect of AGE-BSA on Q.- production in peritoneal macrophages, differents amounts of AGE-BSA were taken from the same preparation (obtained after 2 weeks of incubation of BSA with glucose). As shown in Fig. 3, AGE-BSA inhibited CJ.production in a dose-dependent manner. The maximal inhibitory effect (34 46of the Oz.-produced by control cells) was observed at 50 mg/ml AGE-BSA. Unmodified BSA also inhibited the 4.‘ production but to a lesser extent (about 75% at 50 mg/ml). The time course of the effect of two concentrations of AGE-BSA (formed after 2 weeks of incubation of the albumin with glucose) on @.- production by rat macrophages is shown in Fig. 4. An inhibition of Q.- production in the presence of 20 mg/ml and 50 mg/ml AGE-BSA was observed 10 mm after PMA stimulation. The inhibitory effect of AGE-BSA on Q.- production is also dependent of the concentration of AGE-BSA present in the incubation medium. Experiments performed in the presence of unmodified BSA showed that BSA alone was also able to inhibit the Oz.‘production by rat macrophages in a dose dependent fashion. However, the extent of inhibition was clearly lesser than the inhibition observed in the presence of AGE-BSA.

AGE-BSA

2284

on Macrophage Superoxide Anions

Vol. 60, No. 2.5,1!397

lot 1 ‘ij s E

75

.

‘i; s

50

0” 3 Y -0 8 a !N 0

25

0

‘I 0

10

20 AGE-BSA

30

40

50

(mg/ml)

Fig. 3. Dose-dependency of the effect of AGE-MA on Q- production by rat macrophages. Cells were incubated with 100 nmol/l PMA and increasing amounts of unmodified BSA (0) or AGE-BSA (0) for 60 min. In control conditions, the production of %- was 178.9 f 8.5 nmoV60 min per mg of protein. This value was normalixed to 100%. Values shown are means f SEM from duplicates and are representative of 2 experiments.

The inhibitory effect of 20 mg/ml AGE-BSA on 4; production is suppressed when the formation of advanced glycosylation products is blocked by the presence of 200 mmol/l aminoguanidine (AG) (Fig. 5). The Oz.-production by rat peritoneal macrophages incubated with 20 mg/ml AGE-BSA-AG (BSA incubated with glucose and aminoguanidme during 2 weeks) was similar to the amount of @- produced by control cells after 45 min of incubation. Under these experimental conditions the amount of cytochrome c reduced by AGE-BSA-AG in absence of PMA is nearly negligible and similar to reduction induced by unmodified BSA. Production of 0,’ by macrophages of diabetic rats In order to analyze if the above described effects of AGE-BSA on macrophage function could have some physiological relevance, experiments in STZ-treated rats were performed. Fig. 6 illustrates that macrophages from the diabetic rats presented a reduced capacity to respond to PMA. At 30 min of incubation an inhibition of 20-2546 on Q- production was found. When the synthesis of AGE-products is blocked by in vivo administration of aminoguanidine to diabetic rats, the capacity of macrophages to respond to PMA was recovered. These data suggest that the AGE BSA effect observed in vitro, also present a relevance in experiments in vivo.

Vol. 60, No. 25,197

AGE-BSA on Macrophage Superoxide Anions

w

2

s 250 e a

;

200

%

T-d150

.

2 .g 100 J s a g 50 !N 0

0

:

0

10

30 20 Time (min)

40

Fig. 4. Time course- of the effect of AGE-BSA on a- production by rat macrophages. Cells were incubated with 100 nmoY1 PMA in the presence of the following concentrations of unmodified BSA or AGE-BSA: no additions (control) (O), 20 mg/ml unmodified BSA (A), 20 mg/ml AGE-BSA (A), 50 mg/ml BSA (Cl) and 50 mg/ml AGE-BSA (U). Data are means f SEM from duplicates and are representative of 3-4 experiments. ‘2‘ ._ s 2 ;

300

;r 225

0

10

20

30

40

Time (min)

Fig. 5. Prevention by aminoguanidine (AG) of the inhibitory effect of AGE-BSA on Oz.-production by rat macrophages. Cells were incubated with 100 nmol/l PMA and the following additions: no additions (control)( 0), 20 mg/ml unmodified BSA (A), 20 mg/ml AGE-BSA (A) and 20 mg/ml AGE-BSA-AG (0). Values are means f SEM from duplicates and are representative of 2-3 experiments.

2285

22%

AGE-MA on Maaophage Superoxide Anions

10

20 Time

30

Vol. 60, No. 2S,l!I97

40

(min)

Fig. 6. Time-course of w- production by peritoneal macrophages from control (0), diabetic (A) and aminoguanidine-treated diabetic (A) ratsin response to PMA. Values are means f SBM from 4 replicates and are representative of 3-4 experiments. Values labelled with * are significantly different from control group (pCO.05).

Discussion It is well accepted that superoxide anion production can be taken as an index of oxidative metabolism in phagocytic cells (21). Although an impaired oxidative burst is observed in leukocytes of diabetic patients (1) and in alveolar macrophages of diabetic rats (2), the cause of these altered cell functions is not fully understood. Since protein glycosylation is a process most often implicated in diabetes complications (22,23), it is likely that some products of glycosylation are implicated in the impairment of superoxide anion production. To test this hypothesis we analyzed PMA-dependent superoxide anion production from normal rat peritoneal macrophages in the presence of AGE-modified albumin. Proteins undergo a series of non-enzymatic reactions with glucose over time to form advanced glycosylation end products (4,5). We have confirmed the scavenging effect of BSA on 4; reported previously (24). In addition, we observed that AGE-BSA also scavenges Oz.- by xanthine/xanthine oxidase system. We also found that AGE-BSA inhibits Or’- production by macrophages. We then propose that AGE-BSA inhibits the NADPH-oxidase activity since the extent of inhibition of G; formation is greater in a cell system than in the xanthine/xanthine oxidase system. Moreover, the differences between AGE and AGE-BSA dissapeared in a luminol-xanthine/xanthine oxidase system when incubation times over 3 min were taken. In an opposite way, the differences between BSA and AGE-BSA in a cell system were greater at longer incubation times (over 20 min) than at short times. These data suggest that AGE-BSA specifically inhibits the release of PMA-dependent Oz.by incubated macrophages in vitro. Present results are particularly remarkable by the following

Vol. 60, No. 25,1997

AGE-BSA on Macrophage Superoxide Anions

2287

facts. First, there is a clear positive relationship between the amount of AGE-BSA added to macrophage suspension and the inhibition of PMAdependent superoxide anion production. Second, this inhibition seems to be specific for AGE products, since the presence of aminoguanidine during the formation of AGE-BSA fully prevented the effect of AGE-BSA. Aminoguanidine, a nucleophilic hydraxine compound, inhibits the formation of AGE products both in vitro and in vivo (25,26) and also attenuates the development of several diabetes-induced vascular, neural and collagen changes (25, 27-29). Further experiments analyzed whether macrophages from diabetic rats presented a similar response to PMA. It was observed a clear inhibition on PMA dependent Q- production as consequence of diabetes state. This inhibitory effect was suppressed by treatment of diabetic rats with aminoguanidine. Since aminoguanidine also inhibits aldose reductase (30) and NO formation (31), the possibility that these mechanisms are involved in macrophage dysfunction in diabetic rats remains to be substantiatied. It has recently been demonstrated that macrophages have specific membrane receptors that recognizes the AGE moiety and mediates the uptake and degradation of AGE proteins (11-13). This is of particular significance to give a physiological relevance to the present data. It may be hypothetized that the impaired respiratory burst in macrophages observed during the diabetic state could be caused by the binding to macrophages of AGE-modified proteins, which are increased in this disease. The possibility that AGE-modified proteins were the main factors to explain the impaired respiratory burst during diabetes has been previously suggested on the basis that high levels of glucose at short-term also inhibit Oz.-production by PMN (3). It has recently been proposed that AGE-modified proteins stimulate the secretion of macrophages-derived cytokines such as IL-l and TNF (14). Since both cytokines are able to activate or prime phagocytic cells in order to stimulate the respiratory burst (32,33) discrepancies with the data presented here are to be considered. The great difference in the concentrations of AGE-BSA used either to stimulate the cytolcine production (about 250 pg/ml) or to inhibit the respiratory burst (our data) would explain the different response of macrophages in both processes. Thus, a dual physiological role of AGE-BSA on macrophage function could be suggested depending on its concentration (i.e. stimulation of cytolcines release at low concentrations and inhibition of superoxide anion production at higher amounts). We respiratory complicates cytochrome

also are describing a non-specific effect of AGE-BSA on the measurement of burst in this paper. That is, a direct reduction of oxidized cytchrome c, which the analysis of Q.- production. Thus, adequate control samples (AGE-BSA and alone) should also be routinely included.

In conclusion, although the altered macrophage function in the diabetic state may be caused by multiple mechanisms, the present results suggest that AGE-modified proteins might play an important role in the impairment of respiratory burst. Thus, we have shown an inhibition in the Oz.- production in rat peritoneal macrophages induced by AGE-modified albumin, which may contribute to the diminished bactericidal capacity observed in the diabetic state.

Acknowledgments This study was supported by FISSS (Grants No. 9411484 and 94/1489), and DGICYT (Grant No. 96/0205).

22&l

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.

AGE-ISA

on

Macrophage Superoxide Anions

Vol. 60, No. 25,1997

S.V. SHAH, J.D. WALLIN, and S.D. EILEN, J. Clin. Endocrinol. Metab. s 402-409 (1983). V. MOH-, and J. LATIFPOUR, J. Appl. Physiol. 68 2384-2390 (1990). C.P. NIELSON, and D.A. HINDSON, Dii 3 1031-1035 (1989). F.G. NJOROGE, and V.M. MONNIER, Prog. Clin. Biol. Res. m 85-197 (1989). R. BUCALA, and A. CERAMI, Adv. Pharmacol. a l-34 (1992). V.M. MONNIER, R.R. KOHN, and A. CERAMI, Proc. Natl. Acad. Sci. U.S.A. 81.583587 (1984). V.M. MONNIER, J. Gerentol. G B105-Blll (1990). M. BROWNLEE, S. PONGOR, and A. CERAMI, J. Exp. Med. 15_8 1739-1744 (1983). M. BROWNLEE, H. VLASSARA, and A. CERAMI, Diabetes a 938-941 (1985). M. BROWNLEE, A. CERAMJ, and H. VLASSARA, N. Engl. J. Med. 318 1315-1321 (1988). H. VLASSARA, M. BROWNLEE, and A. CERAMI, Proc. Natl. Acad. Sci. U.S.A. 82 5588-5592 (1985). H. VLASSARA, M. BROWNLEE, and A. CERAMI, J. Exp. Med. w 1301-1309 (1986). S. RADOFF, A. CERAMI, and H. VLASSARA, Diabetes 2 1510-1518 (1990). H. VLASSARA, M. BROWNLEE, K.R. MANOGUE, C.A. DINARELLO, and A. PASAGIAN, Science m 1546-1548 (1988). M. KIRSTEIN, J. BRETT, S. RADOFF, S. OGAWA, and H. VLASSARA, Proc. Natl. Acad. Sci. U.S.A. fl9010-9014 (1990). M. KIRSTEIN, C. ASTON, R. HINTZ, and H. VLASSARA, J. Clin. Invest. % 439-446 (1992). N.E. CAMERON, M.A. COTTER, K. DINES, and A. LOVE, Diabetologia 3 946-950 (1992). D.E. MOSIER, Methods Enzymol. m 294-297 (1984). B.M. BABIOR, R.S. KIPNES, and J.T. CURNE’ITE, J. Clin. Invest. a 741-744 (1973). O.H. LOWRY, N.J. ROSEBROUGH, A.L. FARR, and R.J. RANDALL, J. Biol. Chem. m 265-275 (1951). B.M. BABIOR, Blood a 959-966 (1984). M. BROWNLEE, H. VLASSARA, and A. CERAMI, Ann. Intern. Med. m527-537 (1984). M. BROWNLEE, and A. CERAMI, Annu. Rev. Biochem. jQ 385-432 (1981). M.E. HOLT, M.T.E. RYALL, and A.K. CAMPBELL, B. J. Exp. Path. 65 231-241 (1984). M. BROWNLEE, H. VLASSARA, T. KOONEY, P. ULRICH, and A. CERAMI, Science m 1629-1632 (1986). K. NICHOLLS, and T.E. MANDEL, Lab. Invest. @ 486-493 (1989). J.R. WILLIAMSON, K. CHANG, Y. IDO, E. OSTROW, W. ALLISON, J. HARLOW, and R.G. TILTON, Diabetes Metab. 16 369-370 (1990). T. SOULIS-LIPAROTA, M. COOPER, D. PAPAZOGLOU, B. CLARKE, and G. JERUMS, Diabetes @ 1328-1334 (1991). M. KIHARA, J.D. SCHMELZER, J.F. PODUSLO, G.L. CURRAN, K.K. MCKANDER, and P.A. LOW, Proc. Natl. Acad. Sci. U.S.A. SH 6107-6111 (1991). K. KUMARI, S. UMAR, V. BANSAL, and M.K. SAHIB, Diabetes a 1079-1084 (1991). R.G. TILTON, K. CHANG, K.S. HASAN, S.R. SMITH, J.M. PETRASH, T.P. MISKO, W.M.MOORE, M.G. CURRIE, J.A. CORBERTI’, M.L. MCDANIEL, and J.R. WILLIAMSON, Diabetes Q 221-232(1993).

Vol. 60, No. 25,1997

AGE-MA on Macrophage Superoxide Anions

T. MATSUBARA, and M. ZIFF, J. Immunol. m 32953298 (1986). S.H. PINCUS, E.A. WHlTCOMB, and C.A. DINARELLO, J. Immunol. m (1986).

2289

3509-3514