Reviews/Commentaries/Position Statements REVIEW ARTICLE

New Insights on Oxidative Stress and Diabetic Complications May Lead to a “Causal” Therapy

ANTONIO CERIELLO, MD Recent basic and clinical studies have uncovered new insights into the role of oxidative stress in diabetic complications, ABSTRACT—Evidence implicates hyperglycemia-derived oxygen free radicals as mediators of suggesting a different and innovative ap- diabetic complications. However, intervention studies with classic , such as proach to a possible “causal” antioxidant E, failed to demonstrate any beneficial effect. Recent studies demonstrate that a single hypergly- therapy. The aim of this review is to give cemia-induced process of overproduction of superoxide by the mitochondrial electron-transport an update on this topic. chain seems to be the first and key event in the activation of all other pathways involved in the pathogenesis of diabetic complications. These include increased polyol pathway flux, increased advanced glycosylation end product formation, activation of protein kinase C, and increased HYPERGLYCEMIA, hexosamine pathway flux. Superoxide overproduction is accompanied by increased nitric oxide OXIDATIVE STRESS, AND generation, due to an endothelial NOS and inducible NOS uncoupled state, a phenomenon ENDOTHELIAL favoring the formation of the strong oxidant peroxynitrite, which in turn damages DNA. DNA DYSFUNCTION — Vascular func- damage is an obligatory stimulus for the activation of the nuclear poly(ADP-ribose) tion in has been studied exten- polymerase. Poly(ADP-ribose) polymerase activation in turn depletes the intracellular concen- ϩ sively in both animal models and tration of its substrate NAD , slowing the rate of glycolysis, electron transport, and ATP forma- humans. Impaired endothelium-depen- tion, and produces an ADP-ribosylation of the GAPDH. These processes result in acute endothelial dent vasodilation has been a consistent dysfunction in diabetic blood vessels that, convincingly, also contributes to the development of diabetic complications. These new findings may explain why classic antioxidants, such as vita- finding in animal models of diabetes in- min E, which work by scavenging already-formed toxic oxidation products, have failed to show duced by alloxan or streptozotocin beneficial effects on diabetic complications and may suggest new and attractive “causal” antiox- (9,10). Similarly, studies in humans with idant therapy. New low–molecular mass compounds that act as SOD or catalase mimetics or type 1 or have found en- L-propionyl- and , which work as intracellular superoxide scavengers, im- dothelial dysfunction when compared proving mitochondrial function and reducing DNA damage, may be good candidates for such a with vascular function in nondiabetic strategy, and preliminary studies support this hypothesis. This “causal” therapy would also be subjects (11,12). associated with other promising tools such as LY 333531, PJ34, and FP15, which block the ␤ In vitro, the direct role of hyperglyce- protein kinase isoform, poly(ADP-ribose) polymerase, and peroxynitrite, respectively. While mia has been suggested by evidence that waiting for these focused tools, we may have other options: thiazolinediones, statins, ACE arteries isolated from normal animals, inhibitors, and angiotensin 1 inhibitors can reduce intracellular oxidative stress generation, and it has been suggested that many of their beneficial effects, even in diabetic patients, are due to this property. which are subsequently exposed to exog- enous hyperglycemia, also exhibit atten- Diabetes Care 26:1589–1596, 2003 uated endothelium-dependent relaxation (13). Consistently, in vivo studies have also demonstrated that hyperglycemia di- rectly induces, both in diabetic and nor- he relationship between diabetes It has been suggested that, in diabe- mal subjects, an endothelial dysfunction and premature vascular disease is tes, oxidative stress plays a key role in the (14,15). well established (1). Recent pro- pathogenesis of vascular complications, T The role of free radicals generation in spective studies indicate that long-term both microvascular and macrovascular producing the hyperglycemia-dependent glycemic control is an important predictor (6), and an early marker of such damage is endothelial dysfunction is suggested by not only of microvascular disease (2,3), but the development of an endothelial dys- studies showing that both in vitro (16,17) also of macrovascular complications (4). function (6,7). However, the role of oxi- and in vivo (18–21), the acute effects of Vascular endothelial cells are an im- dative stress in diabetes is questioned by hyperglycemia is counterbalanced by portant target of hyperglycemic damage, the results of intervention studies with an- antioxidants. but the mechanisms underlying this dam- tioxidants, which are elusive or unsuc- age are not fully understood (5). cessful (8). ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● Increased superoxide production in endothelial cells during From the Department of Pathology and Medicine, Experimental and Clinical, Chair of Internal Medicine, University of Udine, Udine, Italy. hyperglycemia: the unifying Address correspondence and reprint requests to Prof. Antonio Ceriello, Chair of Internal Medicine, University hypothesis for the development of of Udine, P.le S. Maria della Misericordia, 33100 Udine, Italy. E-mail: [email protected]. diabetic complications Received for publication 2 May 2002 and accepted in revised form 12 February 2003. Brownlee (22) recently pointed out the Abbreviations: AT1, angiotensin 1; NF, nuclear factor; NO, nitric oxide; NOS, nitric oxide synthase. A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion key role of superoxide production in en- factors for many substances. dothelial cells at the mitochondrial level © 2003 by the American Diabetes Association. during hyperglycemia in the pathogenesis

DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 1589 Causal antioxidant therapy for diabetic complications

genes related to vascular stress response (26). Figure 1 summarizes this finding.

SUPEROXIDE, NITRIC OXIDE, PEROXYNITRITE, AND NITROTYROSINE FORMATION — Even increased su- peroxide generation in hyperglycemia is a key event in activating the other pathways involved in the pathogenesis of diabetic complications; it represents only a first step in the production of the endothelial dysfunction in diabetes. Nitric oxide (NO) production plays a central role in modulating endothelial function (27). NO is generated from the of L-arginine by the enzyme nitric oxide synthase (NOS), of which there are three isoforms: the constitutive Figure 1—In endothelial cells, can pass freely through the cell membrane in an insulin- types, brain (bNOS) and endothelial independent manner via GLUT1. Intracellular hyperglycemia induces overproduction of super- (eNOS), and the inducible type, iNOS oxide at the mitochondrial level. Overproduction of superoxide is the first and key event in the (28). iNOS is induced de novo by various activation of all other pathways involved in the pathogenesis of diabetic complications, such as stimuli, including hyperglycemia (29), polyol pathway flux, increased advanced glycosylation end product (AGE) formation, activation ␬ while the mitochondrial-generated super- of protein kinase C (PKC) and NF- B, and increased hexosamine pathway flux. oxide can inhibit eNOS, although enough NO is still produced (30). of diabetic complications. This new in- oxidoreductase), one of the four inner The superoxide anion may quench sight is consistent with the four pathways membrane–associated complexes central NO, thereby reducing the efficacy of a suggested to be involved in the develop- to oxidative . The data in potent endothelium-derived vasodila- ment of diabetic complications (increased the papers (23) indicate that, at least in tor system that participates in the general polyol pathway flux, increased advanced the cell culture, endothelium in an envi- of the vasculature (31), and glycosylation end product formation, ac- ronment mimicking physiological hyper- evidence suggests that during hyperglyce- tivation of protein kinase C, and increased glycemia cannot control its appetite for mia, reduced NO availability exists (14). hexosamine pathway flux) and with a uni- glucose. Accelerated flux of glucose The activation of protein kinase C, fying hypothesis regarding the effects of through glycolysis and feeding of pyru- due to superoxide overproduction, in- hyperglycemia on cellular dysfunction vate (thus formed) to the tricarboxylic duces a de novo synthesis of the enzyme (23,24). The authors used endothelial acid cycle overloads mitochondria, caus- NAD(P)H oxidase, which significantly cells subjected to physiologically relevant ing excessive generation of free radicals. contributes to produce more superoxide glucose concentrations as a model system Although oxygen free radicals have been anions (32). for analyzing the vascular response to hy- shown to have a physiological role in sig- Hyperglycemia also favors, through perglycemia because the non-insulin- nal transduction, their sustained genera- the activation of NF-␬B, an increased ex- dependent glucose transporter GLUT1 tion at the levels shown in endothelial pression of both NAD(P)H and iNOS facilitates diffusion of high levels of glu- cells exposed to high glucose can be ex- (33). Overexpression of iNOS is accom- cose into the endothelium. In the pres- pected to have substantial effects on cel- panied by increased generation of NO ence of increased glucose, endothelial lular properties. Each of the pathways (33). generation of reactive oxygen species, implicated in secondary complications of Superoxide overproduction, when particularly superoxide anion, was shown diabetes has been shown to arise by a sin- accompanied by increased NO genera- to be enhanced (23). Several pathways gle unifying mechanism (22). A central tion, favors the formation of the strong can be considered as likely candidates for contribution of the works of Brownlee et oxidant peroxynitrite (34), which in turn oxygen free radical formation in cells. al. (23) is to demonstrate that suppression avidly oxidizes , an These include NAD(P)H oxidase, the mi- of intracellular free radicals, using low iNOS cofactor, to dihydrobiopterin (35). tochondrial respiratory chain, xantine molecular inhibitors or by expression of Under conditions of tetrahydrobiopterin oxidase, the arachidonic cascade (lipoxy- the antioxidant enzyme - deficiency, iNOS is in an uncoupled state, genase and cycloxygenase), and microso- superoxide dismutase, prevents each of which means that electrons flowing from mal (25). Brownlee et al. (23) these events (i.e., glucose-induced forma- the iNOS reductase domain to the oxy- have determined that the source of free tion of oxidants is a proximal step in cell genase domain are diverted to molecular radicals in endothelial cells incubated in perturbation). oxygen rather than to L-arginine, resulting high glucose is the transport of glycolysis- Furthermore, activation of nuclear in production of superoxide rather than derived pyruvate in mitochondria at the factor (NF)-␬B by this mechanism ties hy- NO (36,37). Exposure to peroxynitrite level of complex II (succinate:ubiquinone perglycemia to the expression of multiple during hyperglycemia also produces an

1590 DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 Ceriello

and ATP formation, and produces the ADP-rybosylation of GAPDH (51). This process results in acute endothelial dys- function in diabetic blood vessels (48). The possibility of normalizing diabetic Ϫ Figure 2—O2 reacting with NO endothelial dysfunction using a specificof produces peroxynitrite (ONOOϪ) poly(ADP-ribose) polymerase inhibitor, (see Fig. 3 for details). The radi- PJ34, supports this evidence (52,53). cal scavenging antioxidant vita- min E is active only against certain A NEW THERAPEUTIC components of oxidative stress, APPROACH: THE particularly peroxidation products, leaving other conse- “CAUSAL” ANTIOXIDANT quences of O Ϫ untouched. THERAPY — As previously stated, 2 convincing evidence is now available about the possible role of oxidative stress uncoupling state of eNOS, presumably The toxic action of nitrotyrosine, and in the development of diabetic complica- via a depletion of the enzyme, favor- therefore perhaps of peroxynitrite, is sup- tions (6). However, clinical trials with an- ing superoxide overproduction (38). ported by study evidence. Increased apo- tioxidants, in particular with , Consistently, in hyperglycemic condi- ptosis of myocytes, endothelial cells, and have failed to demonstrate any beneficial tions, an overproduction of both superox- fibroblasts in heart biopsies from diabetic effect (8). ide and NO has been reported, with a patients (46), in hearts from streptozoto- On this matter, it has recently been threefold increase in superoxide genera- cin-induced diabetic rats (47), and in suggested that antioxidant therapy with tion (37). As previously reported, the si- working hearts from rats during hyper- vitamin E or other antioxidants is limited multaneous overgeneration of NO and glycemia (44) is selectively associated to scavenging already-formed oxidants superoxide favors the production of a with levels of nitrotyrosine found in those and may therefore be considered a more toxic reaction product, the peroxynitrite cells. “symptomatic” rather than a causal treat- anion (34). The peroxynitrite anion is cy- ment for vascular oxidative stress (54). totoxic because it oxidizes sulfydryl PEROXYNITRITE, DNA Figure 2 helps explain this concept. groups in proteins, initiates lipid peroxi- DAMAGE, AND According to the evidence discussed dation, and nitrates amino acids such as ENDOTHELIAL in this article, it is suggested that inter- tyrosine, which affects many signal trans- DYSFUNCTION: THE ROLE rupting the overproduction of superoxide duction pathways (34). The production OF POLY(ADP-RIBOSE) by the mitochondrial electron-transport of peroxynitrite can be indirectly inferred POLYMERASE ACTIVATION — chain would normalize the pathways by the presence of nitrotyrosine (39). Peroxynitrite is a potent initiator of DNA involved in the development of diabet- The possibility that diabetes is associ- single-strand breakage, which is an oblig- ic complications. It might, however, be ated with increased nitrotyrosine forma- atory stimulus for the activation of the difficult to accomplish this using con- tion is supported by the recent detection nuclear enzyme poly(ADP-ribose) poly- ventional antioxidants, because these of increased nitrotyrosine plasma levels in merase (48). As described above, when scavenge reactive oxygen species in a ste- type 2 diabetic patients (40). Several endothelial cells respond to high glucose, chiometric manner. New low–molecular pieces of evidence support a direct role of reactive nitrogen and oxygen species gen- mass compounds that act as SOD or cata- hyperglycemia in favoring this phenome- eration occurs (49). These reactive spe- lase mimetics have the theoretical advan- non. Nitrotyrosine formation is detected cies trigger DNA single-strand breakage, tage of scavenging reactive oxygen species in the artery wall of monkeys during hy- which induces a rapid activation of poly- continuously by acting as catalysts with perglycemia (41) and also in the plasma of (ADP-ribose) polymerase (49). The role efficiencies approaching those of native healthy subjects during a hyperglycemic of hyperglycemia and related oxidative enzymes (54). Such compounds normal- clamp (42) and in diabetic patients during stress in producing DNA damage is sup- ize endothelial dysfunction in streptozo- an increase of postprandial hyperglyce- ported by the recent findings that in- tocin-induced diabetic rats (55) and mia (43). Hyperglycemia is also accompa- creased amounts of 8-hydroxyguanine improve diabetes-induced decreases in nied by nitrotyrosine deposition in a and 8-hydroxydeoxyguanosine (markers endoneurial blood flow and motor nerve perfused working heart from rats, and it is of oxidative damage to DNA) can be conduction velocity (56). Another inter- reasonably related to unbalanced produc- found in both the plasma and tissues of esting compound is L-propionyl-carni- tion of NO and superoxide through iNOS streptozotocin diabetic rats (50). These tine. This substance has been shown to act overexpression (44). Nitrotyrosine for- concentrations are correlated and can be as an intracellular superoxide scavenger, mation is followed by the development of reduced by the control of hyperglycemia improving mitochondrial function and an endothelial dysfunction in both and by the use of the antioxidants probu- reducing DNA damage (57–60). These healthy subjects (41) and in coronaries of col and vitamin E (50). properties have been shown to have ben- perfused hearts (44), and this effect is not Poly(ADP-ribose) polymerase activa- eficial effects on diabetic heart function, surprising because it has been shown that tion in turn depletes the intracellular con- peripheral nerve function, and vascular ϩ nitrotyrosine can also be directly harmful centration of its substrate NAD , slowing blood flow in experimental diabetes to endothelial cells (45). the rate of glycolysis, electron transport, (58,60,61).

DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 1591 Causal antioxidant therapy for diabetic complications

In the last years, another substance has received much attention: lipoic acid (62). It may have a unique self-regene- rating capacity as a mitochondrial antiox- idant (62), and the possibility that it restores endothelial dysfunction in both animal models of diabetes and in diabetic patients has been reported (63,64). Other promising tools are LY 333531, PJ34, and FP15, which block the protein kinase ␤ isoform, poly(ADP-ribose) poly- merase, and peroxynitrite, respectively. Not surprisingly, they have been shown to ameliorate the endothelial dysfunction induced by hyperglycemia (52,53,65, 66). LY 333531 has been demonstrated to reduce oxidative stress generation in the retina: this is consistent with the evidence that protein kinase C activation may in- crease superoxide generation through NAD(P)H (67). PJ34 is not an antioxidant and does not directly interfere with the reactivity of peroxynitrite (48); however, Ϫ Ϫ poly(ADP-ribose) polymerase / mice show that after coronary occlusion and reperfusion, nitrotyrosine staining is markedly reduced by PJ34 (68). There is evidence that poly(ADP-ribose) polymer- ase inhibition can suppress the expression of iNOS (69,70) and that poly(ADP- ribose) polymerase can regulate the func- tion of mitochondria in oxidatively challenged cells: poly(ADP-ribose) poly- merase inhibitors or poly(ADP-ribose) polymerase deficiency can suppress mito- Figure 3—Superoxide overproduction reduces eNOS activity, but through NF-␬B and protein chondrial permeability transition and mi- kinase C (PKC) activates NAD(P)H and increases iNOS expression: the final effect is an increased tochondrial oxidant generation (71). NO generation. This condition favors the formation of the strong oxidant peroxynitrite, which in Because mitochondria represent a princi- turn produces, in iNOS and eNOS, an uncoupled state, resulting in the production of superoxide pal source of reactive oxidants in endo- rather than NO, and damages DNA. DNA damage is an obligatory stimulus for the activation of the nuclear enzyme poly(ADP-ribose) polymerase (PARP). Poly(ADP-ribose) polymerase acti- thelial cells placed in high glucose (22) ϩ may be that poly(ADP-ribose) polymerase vation in turn depletes the intracellular concentration of its substrate NAD , slowing the rate of inhibition suppresses nitrotyrosine glycolysis, electron transport, and ATP formation, and produces an ADP-ribosylation of the generation preserving mitochondrial GAPDH. This process results in acute endothelial dysfunction in diabetic blood vessels, which contributes to the development of diabetic complications. NF-␬B activation also induces a proin- integrity. flammatory conditions and adhesion molecule overexpression. All these alterations produce the FP15 is a potent peroxynitrite decom- final picture of diabetic complications. SOD or catalase mimetics, L-propionyl-carnitine and lipoic position catalyst (66). It inhibits tyrosine acid, reducing mitochondrial overproduction of superoxide, and reducing DNA damage may be nitration and reduces the toxicity of per- good candidates for causal intracellular antioxidants. This causal therapy would also be associ- oxynitrite for ␤-cells and vascular endo- ated with LY 333531, PJ34, and FP15, which block protein kinase ␤ isoform, poly(ADP-ribose) thelium during the development of polymerase, and peroxynitrite, respectively. Other current options may be thiazolinediones, st- diabetes in rats (66). Therefore, in the atins, ACE inhibitors, and ATI inhibitors, which also have, at different levels, intracellular causal near future, a causal antioxidant therapy antioxidant activity. AGE, advanced glycosylation end product. may include SOD and catalase mimetics, L-propionyl-carnitine, lipoic acid, protein cused tools, we may already have other inhibiting effect on iNOS and signifi- kinase C␤ and poly(ADP-ribose) poly- possibilities. It is now evident that thiazo- cantly reduce peroxynitrite generation merase inhibitors, and peroxynitrite cata- linediones (72), statins (73), ACE inhibi- (72). An animal study supports the effi- lysts. This combination would aim to tors (74), and ATI inhibitors (75,76) have cacy of thiazolidinediones in reducing the block the noxious cascade activated by a strong intracellular antioxidant activity, harmful effect of peroxynitrite (77). In- hyperglycemia through the overproduc- and it has been suggested that many of deed, statins increase NO tion of superoxide and NO. their beneficial ancillary effects are due to and decrease superoxide production However, while waiting for these fo- this property. Thiazolinediones have an (78), probably interfering with NAD(P)H

1592 DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 Ceriello activity (79), and improve endothelial discovery and to the evaluation of new thelium-dependent and independent va- dysfunction (73), even in diabetic pa- antioxidant molecules, such as SOD and sodilation in patients with type 2 (non- tients (80,81). This hypothesis is sup- catalase mimetics, that hopefully may in- insulin dependent) diabetes mellitus. ported by the evidence that the action of hibit, at an early stage, the mechanism Diabetologia 35:771–776, 1992 statins is independent of the cholesterol- leading to diabetic complications. While 13. Bohlen HG, Lash JM: Topical hyperglyce- mia rapidly suppresses EDRF-mediated lowering effect (73,80,81) and that stud- waiting for these new and specific com- vasodilatation of normal rat arterioles. ies show that (at least) simvastatin pounds, it is reasonable to suggest that Am J Physiol 265:H219–H225, 1993 ameliorates diabetic retinopathy (82) and already-available substances, such as thia- 14. Giugliano D, Marfella R, Coppola L, Ver- reduces as primary zolinediones, statins, ACE inhibitors, and razzo G, Acampora R, Giunta R, Nappo F, prevention in diabetic patients (83). ATI blockers, should also be used because Lucarelli C, D’Onofrio F: Vascular effects Moreover, statins reduce nitrotyrosine they are effective causal antioxidants. of acute hyperglycemia in humans are formation in diabetic patients (81) and at- reversed by L-arginine: evidence for re- tenuate angiotensin II–induced free radi- duced availability of nitric oxide during cal production (84). This effect is of References hyperglycemia. Circulation 95:1783–1790, particular interest because it is now well 1. Kannel WB, McGee DL: Diabetes and car- 1997 diovascular diseases: the Framingham 15. Kawano H, Motoyama T, Hirashima O, recognized that angiotensin II and acti- Study. JAMA 241:2035–2038, 1979 Hirai N, Miyao Y, Sakamoto T, Kugiyama vated angiotensin 1 (AT-1) receptors pro- 2. UK Prospective Diabetes Study Group: K, Ogawa H, Yasue H: Hyperglycemia duce intracellular oxidative stress Intensive blood-glucose control with sul- rapidly suppresses flow-mediated endo- (85,86), and evidence shows that hyper- phonylureas or insulin compared with thelium-dependent vasodilation of bra- glycemia is able to directly modulate cel- conventional treatment and risk of com- chial artery. J Am Coll Cardiol 34:146– lular angiotensin production (87). plications in patients with type 2 diabetes 154, 1999 Therefore, it is not surprising that the pos- (UKPDS 33). Lancet 352:837–853, 1998 16. Tesfamariam B, Cohen RA: Free radicals sible use of both ACE inhibitors and ATI 3. The Diabetes Control and Complications mediate endothelial cell dysfunction receptor blockers should be of particular Trials (DCCT) Research Group: The effect caused by elevated glucose. Am J Physiol interest in the prevention of hyperglyce- of intensive treatment of diabetes on the 263:H321–H326, 1992 development and progression of long- 17. Diederich D, Skopec J, Diederich A, Dai F: mia-derived oxidative stress. term complications in insulin-dependent Endothelial dysfunction in mesenteric re- Important trials have shown that diabetes mellitus. N Engl J Med 329:977– sistance arteries of diabetic rats: role of these compounds are particularly effi- 986, 1993 free radicals. Am J Physiol 266:H1153– cacious in preventing diabetic compli- 4. Laakso M: Hyperglycemia and cardiovas- H1161, 1994 cations such as retinopathy (88), cular disease in type 2 diabetes. Diabetes 18. Marfella R, Verrazzo G, Acampora R, La nephropathy (89–93), and cardiovascu- 48:937–942, 1999 Marca C, Giunta R, Lucarelli C, Paolisso lar disease (91,93). ACE inhibitors and 5. De Vriese AS, Verbeuren TJ, Van de G, Ceriello A, Giugliano D: ATI blockers have been demonstrated to Voorde J, Lameire NH, Vanhoutte PM: reverses systemic hemodynamic changes act as causal antioxidants (85,86), and it Endothelial dysfunction in diabetes. Br J by acute hyperglycemia in healthy sub- has recently been suggested that this Pharmacol 130:963–974, 2000 jects. Am J Physiol 268:E1167–E1173, 6. Giugliano D, Ceriello A, Paolisso G: Oxi- 1995 property may account for their beneficial dative stress and diabetic vascular compli- 19. Ting HH, Timimi FK, Boles KS, Creager effect on diabetic complications (94). cations. Diabetes Care 19:257–267, 1996 SJ, Ganz P, Creager MA: This hypothesis is strongly supported by 7. Cai H, Harrison DG: Endothelial dysfunc- improves endothelium-dependent vaso- the recent evidence that treatment with tion in cardiovascular disease: the role of dilation in patients with non-insulin-de- either ACE inhibitors or AT-1 blockers oxidant stress. Circ Res 87:840–844, pendent diabetes mellitus. J Clin Invest 97: prevents nitrotyrosine deposition in the 2000 22–28, 1996 kidney and the development of protein- 8. Marchioli R, Schweiger C, Levantesi G, 20. Ceriello A, Motz E, Giugliano D: Vitamin uria in diabetic rats (95). A summary of Gavazzi L, Valagussa F: Antioxidant vita- C and hypertension. Lancet 355:1271– the various options proposed as causal mins and prevention of cardiovascular 1272, 2000 antioxidant therapy is reported in Fig. 3. disease: epidemiological and clinical trial 21. Beckman JA, Goldfine AB, Gordon MB, data. 36:S53–S63, 2001 Creager MA: Ascorbate restores endothe- In conclusion, a large body of evi- 9. Meraji S, Jayakody L, Senaratne PJ, Thom- lium-dependent vasodilation impaired by dence suggests a possible role of oxidative son ABR, Kappagoda T: Endothelium-de- acute hyperglycemia in humans. Circula- stress in the pathogenesis of diabetic com- pendent relaxation in aorta of BB rat. tion 103:1618–1623, 2001 plications. This raises the concept that an Diabetes 36:978–981, 1987 22. Brownlee M: Biochemistry and molecular antioxidant therapy may be of great inter- 10. Mayhan WG: Impairment of endotheli- cell biology of diabetic complications. Na- est in these patients. It has recently been um-dependent dilatation of cerebral ar- ture 414:813–820, 2001 suggested that diabetic subjects with terioles during diabetes mellitus. Am J 23. Nishikawa T, Edelstein D, Du X-L, Yam- complications may have a defective cellu- Physiol 256:H621–H625, 1989 agishi S, Matsumura T, Kaneda Y, Yorek lar antioxidant response against the oxi- 11. Johnstone MT, Creager SJ, Scales KM, M, Beebe D, Oates P, Hammes HP, Gi- dative stress generated by hyperglycemia, Cusco JA, Lee BK, Creager MA: Impaired ardino I, Brownlee M: Normalizing mito- endothelium-dependent vasodilation in chondrial superoxide production blocks which can predispose the patient to organ patients with insulin-dependent diabetes three pathways of hyperglycaemic dam- damage (96). New insights into the mech- mellitus. Circulation 88:2510–2516, 1993 age. Nature 404:787–790, 2000 anisms leading to the generation of oxida- 12. McVeigh GE, Brennan GM, Johnston GD, 24. Du XL, Edelstein D, Rossetti L, Fantus IG, tive stress in diabetes are now available. McDermott BJ, McGrath LT, Henry WR, Goldberg H, Ziyadeh F, Wu J, Brownlee Presumably these findings have led to the Andrews JW, Hayes JR: Impaired endo- M: Hyperglycemia-induced mitochon-

DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 1593 Causal antioxidant therapy for diabetic complications

drial superoxide overproduction activates 2002 8-hydroxydeoxyguanosine in streptozo- the exosamine pathway and induces plas- 39. Ischiropoulos H: Biological tyrosine nitra- tocin-induced diabetic rats. Diabetes 50: minogen activator inhibitor-1 expression tion: a pathophysiological function of ni- 2837–2841, 2001 by increasing Sp1 glycosylation. Proc Natl tric oxide and reactive oxygen species. 51. Kamoshima W, Kitamura Y, Nomura Y, Acad SciUSA97:12222–12226, 2002 Arch Biochem Biophys 356:1–11, 1998 Taniguchi T: Possible involvement of 25. Cross A, Jones O: Enzymic mechanisms of 40. Ceriello A, Mercuri F, Quagliaro L, Assa- ADP-ribosylation of particular enzymes in superoxide production. Biochim Biophys loni R, Motz E, Tonutti L, Taboga C: De- cell death induced by nitric oxide-donors Acta 1057:281–198, 1991 tection of nitrotyrosine in the diabetic in human neuroblastoma cells. Neuro- 26. Collins T: Endothelial nuclear factor kB plasma: evidence of oxidative stress. Dia- chem Int 30:305–311, 1997 and the initiation of the atherosclerotic le- betologia 44:834–838, 2001 52. Garcia Soriano F, Pacher P, Mabley JG, sion. Lab Invest 68:499–508, 1993 41. Pennathur S, Wagner JD, Leeuwenburgh Liaudet L, Szabo E, Szabo C: Rapid rever- 27. Moncada S, Palmer RMJ, Higgs EA: Nitric C, Litwak C, Heinecke JW: A hydroxyl sal of the diabetic endothelial dysfunction oxide: physiology, pathophysiology and radical-like species oxidizes cynomolgus by pharmacological inhibition of poly- pharmacology. Pharmacol Rev 43:109– monkey artery wall proteins in early dia- (ADP-ribose) polymerase activation. Circ 142, 1991 betic vascular disease. J Clin Invest 107: Res 89:684–691, 2001 28. NathanC,XieQW:Nitricoxidesynthases: 853–860, 2001 53. Pacher P, Liaudet L, Garcia Soriano F, roles, tolls, and controls. Cell 78:915– 42. Marfella R, Quagliaro L, Nappo F, Ceri- Mabley JG, Szabo E, Szabo C: The role of 918, 1994 ello A, Giugliano D: Acute hyperglycemia poly(ADP-ribose) polymerase activation 29. Baox KJ, Thiel BA, Stuehr DJ: Macrophage induces an oxidative stress in healthy sub- in the development of myocardial and en- nitric oxide synthase subunits. J Biol Chem jects (Letter). J Clin Invest 108:635–636, dothelial dysfunction in diabetes. Diabetes 268:21120–21129, 1993 2001 51:514–521, 2002 30. Du XL, Edelstein D, Dimmeler S, Ju Q, Sui 43. Ceriello A, Quagliaro L, Catone B, Pascon 54. Cuzzocrea S, Riley DP, Caputi AP, C, Brownlee M: Hyperglycemia inhibits R, Piazzola M, Bais B, Marra G, Tonutti L, Salvemini D: Antioxidant therapy: a new endothelial nitric oxide synthase activity Taboga C, Motz E: The role of hypergly- pharmacological approach in shock, in- by posttranslational modification at the cemia in nitrotyrosine postprandial gen- flammation, and ischemia/reperfusion Akt site. J Clin Invest 108:1341–1348, 2001 eration. Diabetes Care 25:1439–1443, injury. Pharmacol Reviews 53:135–159, 31. Benz D, Cadet P, Mantione K, Zhu W, 2002 2001 Stefano GB: Tonal nitric oxide and health: 44. Ceriello A, Quagliaro L, D’Amico M, Di 55. Nassar T, Kadery B, Lotan C, Da’as N, a free radical and scavenger of free radi- Filippo C, Martella R, Nappo F, Berrino L, Kleinman Y, Haj-Yehia A: Effects of the cals. Med Sci Monit 8:RA1–RA4, 2002 Rossi F, Giugliano D: Acute hyperglyce- superoxide dismutase mimetic com- 32. Hink U, Li H, Mollnau H, Oelze M, mia induces nitrotyrosine formation and pound tempol on endothelial dysfunction Matheis E, Hartmann M, Skatchkov M, apoptosis in perfused heart from rat. Dia- in streptozotocin-induced diabetic rats. Thaiss F, Stahl RAK, Warnholtz A, Mein- betes 51:1076–1082, 2002 Eur J Pharmacol 436:111–118, 2002 ertz T, Griendling K, Harrison DG, For- 45. Mihm MJ, Jing L, Bauer JA: Nitrotyrosine 56. Coppey LJ, Gellett JS, Davidson EP, Dun- stermann U, Munzel T: Mechanisms causes selective vascular endothelial dys- lap JA, Lund DD, Salvemini D, Yorek MA: underlying endothelial dysfunction in di- function and DNA damage. J Cardiovasc Effect of M40403 treatment of diabetic abetes mellitus. Circ Res 88:14–22, 2001 Pharmacol 36:182–187, 2000 rats on endoneurial blood flow, motor 33. Spitaler MM, Graier WF: Vascular targets 46. Frustaci A, Kajstura J, Cimenti C, Jako- nerve conduction velocity and vascular of signalling in diabetes mellitus. niuk I, Leri A, Maseri A, Nadal-Ginard B, function of epineurial arterioles of the sci- Diabetologia 45:476–494, 2002 Anversa P: Myocardial cell death in hu- atic nerve. Br J Pharmacol 134:21–29, 34. Beckman JS, Koppenol WH: Nitric oxide, man diabetes. Circ Res 87:1123–1132, 2001 superoxide, and peroxynitrite: the good, 2000 57. Packer L, Valenza M, Serbinova E, Starke- the bad, and ugly. Am J Physiol 271: 47. Kajstura J, Fiordaliso F, Andreoli AM, Li Reed P, Frost K, Kagan V: Free radical C1424–C1437, 1996 B, Cimenti S, Medow MS, Limana F, scavenging is involved in the protective 35. Guzik TJ, Mussa S, Gastaldi D, Sadowski Nadal-Ginard B, Leri A, Anversa P: IGF-1 affect of L-propionyl-carnitine against J, Ratnatuga C, Pillai R, Channon KM: overexpression inhibits the development ischemia-reperfusion injury of the heart. Mechanisms of increased vascular super- of diabetic cardiomyopathy and angioten- Arch Biochem Biophys 288:533–537, 1991 oxide production in human diabetes sin II–mediated oxidative stress. Diabetes 58. Broderick TL, Haloftis G, Paulson DJ: mellitus: role of NAD(P)H oxidase and 50:1414–1424, 2001 L-propionyl-carnitine enhancement of endothelial nitric oxide synthase. Circula- 48. Garcia Soriano F, Virag L, Jagtap P, Szabo substrate oxidation and mitochondrial tion 105:1656–1662, 2002 E, Mabley JG, Liaudet L, Marton A, Hoyt respiration in the diabetic rat heart. J Mol 36. Brodsky SV, Gao S, Li H, Goligorsky MS: DG, Murthy KG, Salzman AL, Southan Cell Cardiol 2:331–340, 1996 Hyperglycemic switch from mitochon- GJ, Szabo C: Diabetic endothelial dys- 59. Vanella A, Russo A, Acquaviva R, Campisi drial nitric oxide to superoxide produc- function: the role of poly(ADP-ribose) A, Di Giacomo Sorrenti V, Barcellona ML: tion in endothelial cells. Am J Physiol 283: polymerase activation. Nat Med 7:108– L-propionyl-carnitine as superoxide scav- H2130–H2139, 2002 113, 2001 enger, antioxidant, and DNA cleavage 37. Cosentino F, Hishikawa K, Katusic ZS, 49. Garcia Soriano F, Vira`g L, Szabo C: Dia- protector. Cell Biol Toxicol 16:99–104, 2000 Lu¨ scher TF: High glucose increases nitric betic endothelial dysfunction: role of re- 60. Felix C, Gillis M, Driedzic WR, Paulson oxide synthase expression and superox- active oxygen and nitrogen species DJ, Broderick TL: Effects of propionyl-L- ide anion generation in human aortic production and poly(ADP-ribose) poly- carnitine on isolated mitochondrial func- endothelial cells. Circulation 96:25–28, merase activation. J Mol Med 79:437–448, tion in the reperfused diabetic rat heart. 1997 2001 Diabetes Res Clin Pract 53:17–24, 2001 38. Zou MH, Shi C, Cohen RA: Oxidation of 50. Park KS, Kim JH, Kim MS, Kim JM, Kim 61. Cotter MA, Cameron NE, Keegan A, the zinc-thiolate complex and uncoupling SK, Choi JY, Chung MH, Han B, Kim SY, Dines KC: Effects of acetyl and propionyl- of endothelial nitric oxide synthase by Lee HK: Effects of insulin and antioxidant L-carnitine on peripheral nerve function peroxynitrite. J Clin Invest 109:817–826, on plasma 8-hydroxyguanine and tissue and vascular supply in experimental dia-

1594 DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 Ceriello

betes. Metabolism 44:1209–1214, 1995 dependent repression of the inducible ni- nopathy in diabetic patients with hyper- 62. Packer L, Kraemer K, Rimbach G: Molec- tric oxide synthase gene. Mol Cell Biol 20: cholesterolemia.DiabetesResClinPract56: ular aspects of lipoic acid in the preven- 4699–4707, 2000 1–11, 2002 tion of diabetes complications. 73. Takemoto M, Liao JK: Pleiotropic effects 83. Heart Protection Study Collaborative 17:888–895, 2001 of 3-hydroxy-3-methylglutaryl coenzyme Group: MRC/BHF Heart Protection Study 63. Coppey LJ, Gellett JS, Davidson EP, Dun- A reductase inhibitors. Arterioscler Thromb of cholesterol lowering with simvastatin lap JA, Lund DD, Yorek MA: Effect of Vasc Biol 21:1712–1719, 2001 in 20,536 high-risk individuals: a ran- antioxidant treatment of streptozotocin- 74. Varin R, Mulder P, Tamion F, Richard V, domised placebo-controlled trial. Lancet induced diabetic rats on endoneurial Henry JP, Lallemand F, Lerebours G, 360:7–22, 2002 blood flow, motor nerve conduction ve- Thuillez C: Improvement of endothelial 84. Takemoto M, Node K, Nakagami H, Liao locity, and vascular reactivity of epineur- function by chronic angiotensin-convert- Y, Grimm M, Takemoto Y, Kitakaze M, ial arterioles of the sciatic nerve. Diabetes ing enzyme inhibition in : Liao JK: Statins as antioxidant therapy for 50:1927–1937, 2001 role of nitric oxide, prostanoids, oxidant preventing cardiac myocyte hypertrophy. 64. Hetizer T, Finckh B, Albers S, Krohn K, stress, and bradykinin. Circulation J Clin Invest 108:1429–1437, 2001 Kohlschutter A, Meinertz T: Beneficial ef- 102:351–356, 2000 85. Mu¨ nzel T, Keaney JF: Are ACE inhibitors fects of alpha-lipoic acid and ascorbic acid 75. Zhang H, Schmeisser A, Garlichs CD, a “magic bullet” against oxidative stress? on endothelium-dependent, nitric oxide- Plotze K, Damme U, Mugge A, Daniel WG: Circulation 104:1571–1574, 2001 mediated vasodilation in diabetic patients: Angiotensin II-induced superoxide anion 86. Nickenig G, Harrison DG: The AT1-type relation to parameters of oxidative stress. generation in human vascular endothelial angiotensin receptor in oxidative stress Free Radic Biol Med 31:53–61, 2001 cells: role of membrane-bound NADH-/ and atherogenesis. Part I: oxidative stress 65. Beckman JA, Goldfine AB, Gordon MB, NADPH-oxidases. Cardiovasc Res 44:215– and atherogenesis. Circulation 105:393– Garrett LA, Creager MA: Inhibition of 222, 1999 396, 2002 protein kinase Cbeta prevents impaired 76. Wassmann S, Baumer AT, Strehlow K, 87. Gabriely I, Yang XM, Cases JA, Ma XH, endothelium-dependent vasodilation van Eickels M, Grohe C, Ahlbory K, Ro- Rossetti L, Barzilai N: Hyperglycemia caused by hyperglycemia in humans. Circ sen R, Bohm M, Nickenig G: Endothelial modulates angiotensinogen gene ex- Res 90:107–111, 2002 dysfunction and oxidative stress during es- pression. Am J Physiol 281:R795–R802, 66. Szabo C, Mabley JG, Moeller SM, Shi- trogen deficiency in spontaneously hyper- 2001 manovich R, Pacher Virag L, Soriano FG, tensive rats. Circulation 103:435–441, 2001 88. Chaturvedi N, Sjolie AK, Stephenson JM, Van Duzer JH, Williams W, Salzman, 77. Shiojiri T, Wada K, Nakajima A, Katay- Abrahamian H, Keipes M, Castellarin A, Groves JT: Part I: Pathogenetic role of per- ama K, Shibuya A, Kudo C, Kadowaki T, Rogulja-Pepeonik Z, Fuller JH: Effect of oxynitrite in the development of diabetes Mayumi T, Yura Y, Kamisaki Y: PPAR- lisinopril on progression of retinopathy in and diabetic vascular complications: gamma ligands inhibit nitrotyrosine normotensive people with type 1 diabe- studies with FP15, a novel potent per- formation and inflammatory mediator tes. Lancet 351:28–31, 1998 oxynitrite decomposition catalyst. Mol expressions in adjuvant-induced rheuma- 89. Heart Outcomes Prevention Evaluation Med 8:571–580, 2002 toid arthritis mice. Eur J Pharmacol 448: Study Investigators: Effects of ramipril 67. Kowluru RA: Diabetes-induced elevations 231–238, 2002 on cardiovascular and microvascular out- in retinal oxidative stress, protein kinase 78. Dobrucki LW, Kalinowski L, Dobrucki comes in people with diabetes mellitus: C and nitric oxide are interrelated. Acta IT, Malinski T: Statin-stimulated nitric results of the HOPE study and MICRO- Diabetol 38:179–185, 2001 oxide release from endothelium. Med Sci HOPE substudy. Lancet 355:253–259, 68. Zingarelli B, Salzman AL, Szabo C: Ge- Monit 7:622–627, 2001 2000 netic disruption of poly (ADP-ribose) 79. Christ M, Bauersachs J, Liebetrau C, Heck 90. Lewis EJ, Hunsicker LG, Clarke WR, Berl synthetase inhibits the expression of M, Gunther A, Wehling M: Glucose in- T, Pohl MA, Lewis JB, Ritz E, Atkins RC, P-selectin and intercellular adhesion mol- creases endothelial-dependent superox- RohdeR,RazI,CollaborativeStudyGroup: ecule-1 in myocardial ischemia/reper- ide formation in coronary arteries by Renoprotective effect of the angiotensin fusion injury. Circ Res 13:85–94, 1998 NAD(P)H oxidase activation: attenuation receptor antagonist irbesartan in patients 69. Szabo C, Virag L, Cuzzocrea S, Scott GS, by the 3-hydroxy-3-methylglutaryl coen- with nephropathy due to type 2 diabetes. Hake P, O’Connor MP, Zingarelli B, Salz- zyme A reductase inhibitor atorvastatin. N Engl J Med 345:851–860, 2001 man A, Kun E: Protection against per- Diabetes 51:2648–2652, 2002 91. Brenner BM, Cooper ME, de Zeeuw D, oxynitrite-induced fibroblast injury and 80. Tsunekawa T, Hayashi T, Kano H, Sumi Keane WF, Mitch WE, Parving HH, Re- arthritis development by inhibition of D, Matsui-Hirai H, Thakur NK, Egashira muzzi G, Snapinn SM, Zhang Z, Shahinfar poly(ADP-ribose) synthase. Proc Natl K, Iguchi A: Cerivastatin, a hydroxymeth- S, RENAAL Study Investigators: Effects of Acad SciUSA95:3867–3872, 1998 ylglutaryl reductase inhibi- losartan on renal and cardiovascular out- 70. Oliver FJ, Menissier-de Murcia J, Nacci C, tor, improves endothelial function in comes in patients with type 2 diabetes and Decker P, Andriantsitohaina R, Muller S, elderly diabetic patients within 3 days. nephropathy. N Engl J Med 345:861–869, de la Rubia G, Stoclet JC, de Murcia G: Circulation 104:376–379, 2001 2001 Resistance to endotoxic shock as a conse- 81. Ceriello A, Taboga C, Tonutti L, Quagli- 92. Parving HH, Lehnert H, Brochner-Mor- quence of defective NF-kappaB activation aro L, Piconi L, Bais B, Da Ros R, Motz E: tensen J, Gomis R, Andersen S, Arner P: in poly (ADP-ribose) polymerase-1 defi- Evidence for an independent and cum- Irbesartan in Patients With Type 2 Diabe- cient mice. EMBO J 18:4446–4454, 1999 ulative effect of postprandial hypertri- tes and Microalbuminuria Study Group. 71. Virag L, Salzman AL, Szabo C: Poly(ADP- glyceridemia and hyperglycemia on N Engl J Med 345:870–878, 2001 ribose) synthetase activation mediates endothelial dysfunction and oxidative 93. Lindholm LH, Ibsen H, Dahlof B, De- mitochondrial injury during oxidant- stress generation: effects of short- and vereux RB, Beevers G, de Faire U, Fyhr- induced cell death. J Immunol 161:3753– long-term simvastatin treatment. Circula- quist F, Julius S, Kjeldsen SE, Kristiansson 3759, 1998 tion 106:1211–1218, 2002 K, Lederballe-Pedersen O, Nieminen MS, 72. Li M, Pascual G, Glass CK: Peroxisome 82. Sen K, Misra A, Kumar A, Pandey RM: Omvik P, Oparil S, Wedel H, Aurup P, proliferator-activated receptor gamma- Simvastatin retards progression of reti- Edelman J, Snapinn S, the LIFE Study

DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003 1595 Causal antioxidant therapy for diabetic complications

Group: Cardiovascular morbidity and blockers, type 2 diabetes, and renoprotec- 186–194, 2002 mortality in patients with diabetes in the tion (Letter). N Engl J Med 346:705–707, 96. Ceriello A, Morocutti A, Mercuri L, Qua- Losartan Intervention For Endpoint re- 2002 gliaro L, Moro M, Damante G, Viberti duction in hypertension study (LIFE): a 95. Onozato ML, Tojo A, Goto A, Fujita T, GC: Defective intracellular antioxidant randomised trial against atenolol. Lancet Wilcox CS: Oxidative stress and nitric ox- enzyme production in type 1 diabetic pa- 359:1004–1010, 2002 ide synthase in rat diabetic nephropathy: tients with nephropathy. Diabetes 49: 94. Ceriello A, Motz E: Angiotensin-receptor effects of ACEI and ARB. Kidney Int 61: 2170–2177, 2000

1596 DIABETES CARE, VOLUME 26, NUMBER 5, MAY 2003