To Prevent the Formation of Advanced Glycation Endproducts

ABB Archives of Biochemistry and Biophysics 419 (2003) 31–40 www.elsevier.com/locate/yabbi Minireview Use of aminoguanidine (Pimagedine) to prevent the formation of advanced glycation endproducts

Paul J. Thornalley*

Department of Biological Sciences, University of Essex, Central Campus, Wivenhoe Park, Colchester, Essex CO4 3SQ, UK

Received 3 January 2003, and in revised form 9 June 2003

Abstract

Aminoguanidine (AG) is a prototype therapeutic agent for the prevention of formation of advanced glycation endproducts. It reacts rapidly with a,b-dicarbonyl compounds such as methylglyoxal, glyoxal, and 3-deoxyglucosone to prevent the formation of advanced glycation endproducts (AGEs). The adducts formed are substituted 3-amino-1,2,4-triazine derivatives. Inhibition of disease mechanisms, particularly vascular complications in experimental diabetes, by AG has provided evidence that accumulation of AGEs is a risk factor for disease progression. AG has other pharmacological activities, inhibition of nitric oxide synthase and semicarbazide-sensitive amine oxidase (SSAO), at pharmacological concentrations achieved in vivo for which controls are required in anti-glycation studies. AG is a highly reactive nucleophilic reagent that reacts with many biological molecules (pyridoxal phosphate, pyruvate, glucose, malondialdehyde, and others). Use of high concentrations of AG in vitro brings these reactions and related effects into play. It is unadvisable to use concentrations of AG in excess of 500 lM if selective prevention of AGE formation is desired. The peak plasma concentration of AG in clinical therapy was ca. 50 lM. Clinical trial of AG to prevent progression of diabetic nephropathy was terminated early due to safety concerns and apparent lack of efficacy. Pharmacological scavenging of a-oxoaldehydes or stimulation of host a-oxoaldehyde detoxification remains a worthy therapeutic strategy to prevent diabetic complications and other AGE-related disorders. Ó 2003 Elsevier Inc. All rights reserved.

Aminoguanidine (Pimagedine, AG)1 is a prototype dicarbonyl-directing guanidino group ANHAC(@NH) a,b-dicarbonyl scavenging agent that prevents the for- NH2; the guanidino groups of arginine residues in pro- mation of advanced glycation endproducts (AGEs) teins are key sites of advanced glycation by a,b-dicar- from a,b-dicarbonyl precursors (Fig. 1). The ﬁrst report bonyl compounds [2,3]. These two groups linked together of intervention to prevent AGE formation by AG was provide a reactive bifunctional scavenger of a,b-dicar- the prevention of diabetes-induced arterial wall protein bonyl glycating agents, particularly a-oxoaldehydes such crosslinking [1]. Since then, use of AG to prevent AGE as methylglyoxal, glyoxal, and 3-deoxyglucosone (3-DG) formation in vitro and in vivo has given evidence of the [4]. These a-oxoaldehydes would otherwise form AGEs involvement of advanced glycation in many disease by reaction with arginine and lysine residues [5,6]. The processes and abnormal physiological states. products of the scavenging reaction are substituted AG is nucleophilic agent with two key reaction centers: 3-amino-1,2,4-triazines [4,7,8]. The pharmacological the nucleophilic hydrazine group ANHNH2 and the activity of these products has not been investigated thoroughly but 3-amino-1,2,4-triazine is an inhibitor of * Fax: +44-1206-872-592. inducible nitric oxide synthase [9]. E-mail address: [email protected]. The reaction kinetics of these a-oxoaldehydes with AG 1 Abbreviations used: AGE, advanced glycation endproduct; AG, under physiological conditions, pH 7.4 and 37 °C, in- aminoguanidine; CML, Ne-carboxymethyl-lysine; 3-DG, 3-deoxyg- volved two reaction pathways: the reaction of AG with lucosone; eNOS, iNOS, and nNOS, endothelial, inducible, and highly reactive unhydrated a-oxoaldehyde and reaction neuronal nitric oxide synthase, respectively; PKC, protein kinase C; PLP, pyridoxal-50-phosphate; RAGE, receptor for advanced glycation with the less reactive, a-oxoaldehyde monohydrate or endproducts; SSAO, semicarbazide-sensitive amine oxidase; STZ, cyclic hemiacetal (Fig. 2). The rate of reaction of AG with streptozotocin. glyoxal was ﬁrst order with respect to both reactants; the

Fig. 1. The reaction of AG with glyoxal, methylglyoxal, 3-deoxyglucosone, and a,b-dicarbonyl compounds.

k1 RO O Slow R O and/or OH O H OH HO k-1 H O Hemiacetal Hydrate Free α-oxoaldehyde

NH NH H NNH Slow + H2NNH Fast + 2 NH2 NH2

k3 -H2O -H2O k2

R N N N N

N NH2 R N NH2 6-Substituted 5-Substituted 3-amino-1,2,4-triazine 3-amino-1,2,4-triazine

Fig. 2. Two reaction pathway kinetic models for the scavenging of a-oxoaldehydes by aminoguanidine.

1 1 rate constant kAG;G was 0.892 0.037 M s . The methylglyoxal dehydration MG-H2O MG. The ki- kinetics of the reaction of AG with 3-DG were more netics of these reactions were not inﬂuenced by ionic complex: the rate equation was d½T 0=dt ¼½3 DG strength but the reaction of AG with glyoxal and with ðkAG;3DG½AGþk3DGÞ, where kAG;3DG ¼ð3:23 methylglyoxal under MG-H2O dehydration rate-limited 3 1 1 0:25Þ10 M s and k3DG ¼ð1:73 0:08Þ conditions increased with increasing phosphate buﬀer 105 s1. The kinetics of the reaction of AG with meth- concentration. Kinetic modeling indicated that the rapid ylglyoxal were consistent with reaction of both unhy- reaction of AG with the MG perturbed the MG/MG-H2O drated (MG) and monohydrate (MG-H2O) forms. The equilibrium and the ratio of the isomeric triazine products rate equation was d½T 0=dt ¼fk1kAG;MG=ðk1þ varied with initial reactant concentration [4]. Other 3- to kAG;MG½AGÞ þ kAG;MG-H2Og [MG-H2O][AG], where the 6-carbon dicarbonyl compounds found in physiological rate constant for the reaction of AG with MG, kAG;MG, systems were also scavenged by AG by similar mecha- was 178 15 M1 s1 and for the reaction of AG with nisms (Table 1). These are of interest because they are 1 1 MG-H2O, kAG;MGH2O, was 0.102 0.001 M s ; k1 formed by fragmentation and dehydration of hexoses and and k1 are the forward and reverse rate constants for Amadori products [10,11]. Physiological exposure of this P.J. Thornalley / Archives of Biochemistry and Biophysics 419 (2003) 31–40 33

Table 1 Kinetic constants for the scavenging of a-oxoaldehydes by aminoguanidine at pH 7.4 and 37 °C

3 1 1 1 3 1 1 6 1 a-Oxoaldehyde k1 (10 s ) k2 (M s ) k3 (10 M s ) kScavenging (10 s )

Glyoxal ——— 44.6 Methylglyoxal 0.798 178 15 102 1 144.6 Hydroxypyruvaldehyde 2.720 117 5 22.7 9.7 58.4 Erythrosone 1.360 53.2 4.3 14.2 5.4 20.7 3-Deoxyerythrosone 0.728 27.1 1.3 4.37 1.22 13.5 Ribosone 0.647 15.8 0.7 7.11 1.88 8.5 3-Deoxyribosone 0.425 9.95 1.19 2.10 0.47 5.0 Glucosone 0.144 4.46 0.50 2.13 0.26 2.3 3-DG ——— 17.5

From [13]. kScavenging is the pseudo-first-order rate constant computed for a-oxoaldehyde scavenging by 50 lM AG. The reaction kinetics of glyoxal and 3-DG fitted a different kinetic model [4]. range of a-oxoaldehydes occurs by slow endogenous concentrations, relative to the conditions in vivo [21]. formation, absorption from ingested food and beverages, Increasing the AG concentration favors scavenging of and from dialysis fluids for subjects with endstage renal the a,b-dicarbonyl compound not only because of mass disease on peritoneal dialysis. Similar pentose dicarbonyl, action effects but also because AG may react with the L-xylosone, and 3-deoxy-xylosone are formed from the highly reactive unhydrated form of a,b-dicarbonyl degradation of L-ascorbic acid [12]. The most efficient compounds before hydration. By this effect, the IC50 for scavenging occurred for methylglyoxal [10,13]. AG is ki- inhibition of protein glycation by methylglyoxal in hu- netically competent to scavenge the a-oxoaldehydes man plasma could be decreased to ca. 3 lM if methyl- studied and decrease related AGE formation in vivo. This glyoxal was formed in situ in the presence of AG. The effect is limited, however, by the rapid renal elimination of rates of reaction of methylglyoxal with water and AG AG; the pharmacokinetic half-life was 1.4 h [14] but are equal at ca. 250 lM AG under physiological condi- increased with development of renal insufficiency [15]. a- tions. Above this concentration, therefore, AG will be a Oxoaldehyde scavenging by AG in vivo has been con- very potent scavenger of methylglyoxal generated in firmed by the detection of the triazine products [16]. situ. Similar effects are likely to be operative for other Scavenging of -oxoaldehydes decreases the formation of a,b-dicarbonyl compounds. AGEs, including associated protein crosslinking. AG AG is also a potent and irreversible inhibitor of prevented glycation-induced crosslinking of rat tail ten- semicarbazide-sensitive amine oxidase (SSAO). The Ki don collagen in vitro [17] and of experimental diabetic rats value of the rat aortal enzyme is ca. 1–2 lM with ben- in vivo [18]. zylamine as substrate [22]. SSAO catalyzes the conver- One of the problems of deploying AG to prevent the sion of methylamine to formaldehyde and aminoacetone formation of AGEs is, however, the high reactivity and to methylglyoxal—although this is thought to be a minor competence of AG in scavenging other carbonyl-con- source of methylglyoxal [23]. It may thereby suppress taining compounds. This becomes significant in vitro the formation of carbonyl compounds from a range of when very high concentrations of AG are used. There amine substrates that would otherwise participate in are also other pharmacological activities of AG associ- glycation reactions. ated with non-covalent, high affinity binding interac- At high concentrations of AG, there is also significant tions. reaction of AG with other carbonyl compounds. In- AG is an inhibitor of nitric oxide synthases (NOS): a vestigating the effects of pyruvate reversing metabolic potent inhibitor of the inducible form (iNOS), pseudohypoxia in hyperglycemia [24], we found that AG IC50 ¼ 31 lM; and a weaker inhibitor of neuronal NOS reacted with pyruvate under physiological conditions to (nNOS), IC50 170 lM, and endothelial NOS (eNOS), form a hydrazone adduct (Fig. 3A). The deduced IC50 IC50 330 lM [19]. Since the IC50 for inhibition of value of AG to compete with cellular metabolism of protein glycation by methylglyoxal in human plasma by pyruvate was ca. 7 mM (R. Ng and P.J. Thornalley, AG was 203 lM [20], it is likely that in all cases where unpublished observation). Derangement of pyruvate AG has been used to prevent glycation reactions, AG metabolism by AG may account for the inhibition of was also competent to inhibit iNOS and probably also glucose-stimulated insulin secretion from pancreatic nNOS. In some studies, low concentrations of AG b-cells by 2–10 mM AG [25,26]. This may also contrib- (1 lM) had anti-glycation effects but these studies em- ute to suppression of oxidative stress linked to mito- ployed low protein concentrations in cell-free systems chondrial dysfunction [27] by decreasing pyruvate where the kinetic competition for a,b-dicarbonyl com- oxidation in the Krebs cycle; for example, 1–5 mM AG pounds between protein and AG had been biased to- decreased hyperglycemia-induced cytotoxicity in retinal wards reaction with AG by use of low protein pericytes in vitro [28]. Similarly, there was a slow 34 P.J. Thornalley / Archives of Biochemistry and Biophysics 419 (2003) 31–40

Fig. 3. Aminoguanidine adducts and derivatives. (A) AG adduct with pyruvate, (B) AG adduct with glucose, (C) N-(2-acetamidoethyl)aminoguanidine, and (D) AG adduct with PLP.

reaction of AG with glucose to form a b-D-glucopyr- AG has antioxidant activity. At 100 lM, AG de- anosyl AG adduct (Fig. 3B), described by Feather and creased oxidant-induced apoptosis, reactive oxygen co-workers [29]. This reaction does not compete effec- species, and lipid peroxidation in retinal Muller cells [35] tively with glycolytic metabolism for glucose but at high concentrations of AG, it may compete with protein for glucose in glycation reactions. By comparing the rate of reaction of glucose with AG and with serum albumin under physiological conditions, the IC50 value of AG for inhibition of early protein glycation by glucose is ca. 60 mM. The dose–response relationships of these pharmacological activities (Table 2) can be summarized in a titration plot (Fig. 4). From this deduction, it is pre- dicted that use of millimolar concentrations of AG will produce scavenging of pyruvate and prevention of early glycation. Indeed, 10 and 25 mM AG did decrease early glycation of albumin [30]. AG has been used at concentrations >100 mM in some studies, for example [31– 33]. High concentrations of AG may also slowly degrade to form hydrogen peroxide, leading to the suggestion of a pro-oxidant effect [34]. The rate of peroxide formation for 40 mM AG was <0.1% of the rate of peroxide formation by cellular respiration in mitochondria, Fig. 4. Dose–response relationships for pharmacological effects of aminoguanidine. Key: inhibition of a. SSAO, b. iNOS, c. nNOS, d. however, and therefore AG degradation is not a signif- eNOS, e. a-oxoaldehyde-mediated glycation, f. scavenging of pyruvate, icant source of oxidative stress at pharmacological and g. interception of glucose to prevent early glycation. The dotted concentrations in vivo (see Table 2). line is the typical peak plasma concentration of AG in vivo (50 lM).

Table 2 Approximate median effective concentration values for pharmacological effects of aminoguanidine—shown graphically in Fig. 4 Pharmacological effect Median effective concentration (lM) Comment Reaction with glucose 59,000 Competition with protein glycation Reaction with pyruvate 7000 Competition with glycolysis Reaction with a-oxoaldehydes 3–200 Competition with protein glycation (3 lM if unhydrated aldehyde is scavenged) Inhibition of eNOS 330 Competitive enzyme inhibition Inhibition of nNOS 170 Competitive enzyme inhibition Inhibition of iNOS 31 Competitive enzyme inhibition Inhibition of SSAO 2 Competitive enzyme inhibition (benzylamine as substrate) P.J. Thornalley / Archives of Biochemistry and Biophysics 419 (2003) 31–40 35 and decreased lipid peroxidation in STZ diabetic rats Aminoguanidine therapy of experimental diabetic [36]. It is unlikely that these effects are due to scavenging nephropathy of fatty acid oxidation products [37], since the kinetics of scavenging are ca. 1000-fold slower than for scavenging Diabetic nephropathy is a renal insufficiency syn- of a-oxoaldehydes [4]. Higher concentrations (1–10 mM) drome characterized by initial development of hyperfil- inhibited copper-catalyzed ascorbate oxidation [38] and tration. Proximal tubular scarring with impairment of prevented chemiluminescence associated with the leu- protein degradation gives rise to proteinuria and mi- kocyte respiratory burst with evidence of scavenging of croalbuminuria in incipient nephropathy. Glomerular hydrogen peroxide, hypochlorite, and peroxynitrite [39]. basement membrane thickening, mesangial expansion, The prevention of oxidant-induced apoptosis may ra- and glomerulosclerosis leads to a steady decline in glo- ther be due to suppression of apoptosis triggered by merular filtrate rate over 10 years and increased albu- a-oxoaldehyde-mediated DNA damage. Consistent with minuria in overt nephropathy, culminating in endstage this, a-oxoaldehydes accumulate in oxidative stress [40], renal disease with requirement for maintenance of di- glycation of DNA in cultured cells was increased by alysis or kidney transplantation. AGE accumulation glutathione depletion [41] and decreased by AG with protein crosslinking and specific receptor-mediated (100 lM) [42]. cellular responses in the kidney have been linked to the The concentrations of AG achieved in diabetic rats development of diabetic nephropathy [47,48]. AG in- given 1 g/L AG in drinking water were reported [14]. hibited the development of albuminuria and mesangial Average peak plasma concentration was 47 lM. The expansion in STZ diabetic rats on insulin maintenance concentrations achieved in tissues were: aorta 26 nmol/g, therapy. Typically, AG was added in drinking water retina 30 nmol/g, sciatic nerve 38 nmol/g, lens 140 nmol/g, (1 g/L), equivalent to an oral dose of ca. 900 mg/kg/day. kidney (whole) 740 nmol/g, and skin 315 nmol/g. This There was a concomitant decrease in diabetes-associated was equivalent to a dose of 398 mg/kg. Since aminogua- tissue fluorescence and decrease in AGEs (CML and nidine partitions into the tissue water fraction, tissue others) by immunohistochemistry with anti-AGE anti- concentrations were ca. 40–60 lM in the aorta, retina, bodies [49,50]. Similar studies have shown, however, and sciatic nerve and 200, 1060, and 450 lM in the lens, that although AG (0.5 g/L in drinking water or 50 mg/ kidney, and skin. The high concentration of AG in skin kg/day i.p.) inhibited the development of microalbu- and kidney implies that AG may be kinetically competent minuria in STZ diabetic rats, the diabetes associated at these sights to scavenge nascent unhydrated a-oxoal- increase in CML and pentosidine in skin collagen was dehydes upon their formation; the effective IC50 is much not prevented [51]. The scavenging of the CML pre- lower—ca. 3 lM for methylglyoxal, for example [4]. cursor glyoxal by AG may be particularly effective in the kidney, since the concentration of AG in whole kidney was high—approximately 1 mM in whole kidney and Therapeutic intervention with aminoguanidine to prevent concentrated by reabsorption of water to ca. 8 mM in the development of diabetic complications the renal tubule lumen [14]. Glyoxal may also be a more important source of CML in the kidney where there is Successful therapeutic intervention with AG, at least glomerulosclerosis and lipid oxidation than in skin in experimental diabetes, provided the first evidence that where the oxidative degradation of fructosamine deg- a compound preventing AGE formation in diabetes radation may be a more important precursor. could prevent the development of diabetic complica- Hyperfiltration is a hallmark of developing incipient tions. Brownlee [43] has proposed a unifying mechanism nephropathy. The effect of AG on hyperfiltration is not for the development of vascular complications of clear. In some studies, Cooper and co-workers claim diabetes in which the accumulation of methylglyoxal- that AG inhibited hyperfiltration [50] and in others it derived AGEs particularly was implicated as a down- was without effect [52,53]. This may have been due to stream consequence of oxidative stress induced by variability of onset of hyperfiltration and AG may delay mitochondrial dysfunction in hyperglycemia. AG inter- hyperfiltration only. In the same model, AG decreased venes in this mechanism to decrease MG concentration overexpression of pro-sclerotic growth factors, TGF-b1 and MG-mediated formation of AGEs. Superoxide and PDGF-B, and type IV collagen in the kidney [54]. formation in diabetes has also been linked to the acti- AG therapy did not prevent increase in kidney/body vation of vascular NADPH oxidase (which may occur weight ratio or focal glomerular thickening and glo- via RAGE-mediated cell activation [44]) and uncoupling merulosclerosis [55,56]. The complexity of the pharma- of eNOS from nitric oxide formation [45]. AG may also cological activity of AG was elucidated further when decrease superoxide formation by suppressing RAGE Jerums et al. [57] found that AG therapy prevented the activation and by inhibition of uncoupled eNOS increase of protein kinase C (PKC) activity in glomeruli (Ki ¼ 830 lM) [46]—at least in the kidney where AG of diabetic rats. This may be linked to AGE receptor- concentrations are highest. mediated stimulation of oxidative stress and consequent 36 P.J. Thornalley / Archives of Biochemistry and Biophysics 419 (2003) 31–40 membrane translocation of PKC [58]; a similar inter- brane thickening, however, that was resistant to inhibi- pretation may apply to the effects of AG on retinopathy tion by AG [69]. Primary intervention in hypertensive [59]—see below. Indeed, AG therapy of STZ diabetic diabetic rats with AG also decreased acellular capillary rats prevented diabetes-induced increased expression of formation and microthrombus deposition [70]. Second- the receptor for AGEs, RAGE, in renal cortex and ary intervention with AG (0.5 g/L in drinking water) medulla [60]. decreased the progression of vascular occlusion, endo- In the diabetic, spontaneously hypertensive rat, AG thelial proliferation and prevented pericyte loss [71]. AG decreased the development of albuminuria [61]. In the (25 and 100 mg/kg) prevented atrophy of myelinated diabetic transgenic (mREN-2)27 rat, AG (1 g/L in optic nerve [72]. AG (1 g/L in drinking water) prevented drinking water) suppressed glomerulosclerosis but did the decline in NOS-positive retinal neurones of STZ not prevent hyperfiltration or development of albumin- diabetic rats [73]. A five-year study of diabetic dogs uria [62]. AG prevented some proximal tubular degen- showed that AG therapy (20–25 mg/kg) prevented the eration but it was not sufficient to preserve the albumin development of retinopathy. There were decreased reti- degrading capacity of the tubular epithelium and nal microaneurysms, acellular capillaries, and pericyte thereby prevent albuminuria. Early loss of renal prote- loss with respect to diabetic controls [74]. olytic processing during development of incipient nephropathy in STZ diabetic rats was prevented by AG [63]. Hyperfiltration exposes renal tubules to increased Aminoguanidine therapy of experimental diabetic volumes of glomerular filtrate and hence increased neuropathy amounts of glycated albumin and a-oxoaldehydes. If hemodynamic changes in diabetes are marked, it ap- Diabetic neuropathy is morbidity related to diabetes- pears that the renoprotective effect of AG is over- induced abnormalities of both somatic and autonomic whelmed. nerves. There are deficits in endoneurial blood flow, In Otsuka Long-Evans Tokushima Fatty rats—a nerve conduction, and nerve structural changes. The model of type 2 diabetes—AG (1 g/L in drinking water) most common clinical manifestation is symmetric sen- prevented the development of albuminuria, mesangial sorimotor polyneuropathy—a leading cause of foot expansion, and glomerular basement membrane thick- ulceration and amputation [75,76]. AG (25 and 50 mg/ ening [64]. Primary intervention with AG (10 mg/kg s.c.) kg/day i.p.) prevented decreased nerve blood flow and was not effective in the prevention of albuminuria in a improved nerve conduction velocity in STZ diabetic rats primate model of type I diabetes. Glomerular basement [77]. AG did not affect the incidence or ultrastructural membrane thickening was prevented but hyperfiltration appearance of neuroaxonal dystrophy, a feature of au- was increased by AG. AG therapy of controls also led to tonomic neuropathy, in STZ diabetic rats [78]. A study the development of albuminuria [65]. in STZ diabetic baboons showed no evidence of im- provement by AG of declining nerve conduction velocity and autonomic dysfunction (heart rate response to Aminoguanidine therapy of experimental diabetic reti- positional change) [79]. nopathy

Diabetic retinopathy is a microvascular complication Other eﬀects of aminoguanidine in experimental diabetes of diabetes characterized by pericyte loss, basement membrane thickening, and development of microaneu- AG (25 mg/kg/day i.p.) delayed but did not prevent rysms and acellular capillaries in the capillary bed of the the development of cataract in moderately hyperglyce- retina. Weakened blood vessels become occluded or mic ([plasma glucose] <15 mM) STZ diabetic rats but ruptured, leading to hypoxia and angiogenesis or hem- not with severe hyperglycemia [80]. AG (0.5 g/L in orrhage. The clinical manifestation is progression drinking water) prevented erectile dysfunction in Wistar through background to proliferative retinopathy with diabetic rats [81]. AG (500 mg/kg/day in food) decreased eventual visual impairment. AGE accumulation and pancreatic glucose toxicity in db/db mice and suppressed speciﬁc receptor-mediated cellular responses in the ret- the development of diabetes [82]. This implicates a-ox- ina have been linked to the development of diabetic oaldehydes in glucose toxicity. Indeed, a-oxoaldehydes nephropathy [66,67]. AG (25 mg/kg i.p. or 0.5 g/L in such as methylglyoxal and 3-DG formed from glycolytic drinking water) therapy of diabetic rats for 26 weeks intermediates were increased in postprandial hypergly- prevented accumulation of AGEs and abnormal endo- cemia [83] and hence these are the glycation-related thelial cell proliferation, and decreased pericyte loss. factors most likely linked to the progression from im- After 75 weeks, AG also decreased the appearance of paired glucose tolerance to type 2 diabetes. AG pre- acellular capillaries and microaneurysms [68]. There was vented leukocyte dysfunction in diabetes. This activity some acellular capillary formation and basement mem- was linked to the decrease in AGEs in plasma protein P.J. Thornalley / Archives of Biochemistry and Biophysics 419 (2003) 31–40 37

[84]. AG (1 g/day) decreased serum cholesterol, trigly- time from randomization to the doubling of baseline cerides, and low density lipoprotein by 10–28% in dia- serum creatinine. Efficacy was evaluated by comparing betic patients [32]. AG (500 lM) did not prevent the the combined low and high AG dose groups with pla- generation of myofibrils in cardiomyocytes in a model of cebo control for the primary outcome. Over 1700 pa- diabetic cardiomyopathy [85]. tient-years of study accumulated in ACTION I. The combined AG dose group showed a tendency towards slower doubling of serum creatinine but this was not Aminoguanidine therapy and the aging process statistically significant. However, the combined AG dose group had lower levels of triglycerides, LDL cho- AG therapy (1 g/L in drinking water) of rats de- lesterol, and urinary protein. In ACTION II, 599 pa- creased age-related disorders: increased blood pressure tients were enrolled but half way through the study, the and decline in glomerular filtration rate, glomeruloscle- External Safety Monitoring Committee recommended rosis, nephron loss, proteinuria cardiac hypertrophy, the early termination of the ACTION II trial due to and aortal stiffness [86,87]. Renal markers of oxidative safety concerns and apparent lack of efficacy. Reported stress, F2-isoprostanes, and lipid peroxidation were not side effects of AG in clinical therapy were gastrointes- decreased [88]. AGE detected immunochemically in tinal disturbance, abnormalities in liver function tests, cardiac, aortal, and renal tissues were claimed to de- flu-like symptoms, and a rare vasculitis [95]. Clinical creased [86,87] or unchanged [88]. AG (1 g/L in drinking therapeutic development was questioned further when water) did not decrease early glycation or glycoxidation kidney tumours were found in AG-treated diabetic rats adducts of skin and tendon collagen in the ageing rat [56]. [89]. The concentration of AGEs in Drosophila melano- gaster was higher in senescent (75 days old) than young (10 days old) flies. Intervention with AG decreased AGE Aminoguanidine derivatives accumulation but was without significant effect on lifespan. A high concentration of AG (40 mM) decreased N-(2-Acetamidoethyl)aminoguanidine (Fig. 3C) has mean lifespan [90]. Dietary AG decreased the age-de- been developed as an improved inhibitor of glycation pendent accumulation of the AGE pentosidine in nat- and weaker inhibitor of NOS (unpublished data). It urally hyperglycemic broiler hens to a similar extent to inhibited the development of microalbuminuria in STZ that achieved by dietary restriction. This had no effect in diabetic rats and in the diabetic transgenic (mREN-2)27 reproductive performance [91,92]. AG supplementation rat similarly to AG [50,62]. AG reacted with pyridoxal- (4 mM) delayed the onset of senescence in dermal 50-phosphate (PLP) to form a hydrazone adduct fibroblast-like cells from senescence-accelerated mice (Fig. 3D) in vitro and in vivo such that oral adminis- [93]. Anti-aging effects of aminoguanidine may be linked tration of AG to mice decreased liver and kidney PLP to decreased protein glycation but there may also be a levels significantly [96]. A pyridoxal adduct of AG has benefit from inhibition of iNOS that is activated during been prepared and it also prevented the development of repeated infections and leads to increased protein ni- nephropathy in STZ diabetic mice with better efficacy tration-implicated in the nitric oxide hypothesis of aging than AG [97]. [94] and age related increased iNOS expression in the kidney associated with hyperfiltration and glomerular sclerosis [88]. Concluding remarks

Prevention of diabetic complications by AG in ex- Clinical trial of aminoguanidine perimental diabetes has provided the ﬁrst evidence that AGE accumulation is a risk factor of diabetic compli- Clinical trials of AG in overt diabetic nephropathy cations. The use of AG as a pharmacological agent to (ACTION) were performed: ACTION I was conducted prevent AGE formation is complicated, however, by the in patients with type 1 diabetes mellitus; ACTION II in concurrent and more potent inhibition of NOS and patients with type 2 diabetes mellitus. These were dou- SSAO—Fig. 5. Controls are necessary to exclude the ble-blinded, multiple-dose, placebo-controlled, ran- participation of these interactions in the pharmacologi- domized clinical trials designed to evaluate the safety cal response observed. The prevention of AGE accu- and eﬃcacy of AG in retarding the rate of progression mulation and RAGE expression in experimental of renal disease in patients with overt diabetic ne- diabetes provides support for the importance of a-ox- phropathy. In ACTION I and II, 33% of patients were oaldehydes scavenged by AG in AGE-linked morbid- randomized to placebo and 67% randomized to either a ity—particularly methylglyoxal, glyoxal, and 3-DG. At low or high dose of AG (50–300 or 100–600 mg/day). concentrations of AG achieved in vivo, these and pos- The primary endpoint in both ACTION I and II was the sibly an antioxidant response (that may be mediated by 38 P.J. Thornalley / Archives of Biochemistry and Biophysics 419 (2003) 31–40

Fig. 5. Schematic summary of the pharmacological activities of aminoguanidine. anti-glycation effects) are the significant pharmacologi- [6] P.J. Thornalley, Clin. Lab. 45 (1999) 263–273. cal activities. The use of millimolar concentrations of [7] J. Hirsch, E. Petrakova, M.S. Feather, Carbohydr. Res. 232 AG to achieve a pharmacological response in vitro (1992) 125–130. [8] H.-J.C. Chen, A. Cerami, J. Carbohydr. Chem. 12 (1993) 731–742. cannot be reconciled with specific inhibition of AGE [9] J. Kolls, J.M. Xie, R. LeBlanc, T. Malinski, S. Nelson, W. Summer, formation. Other effects—metal ion binding, reaction S.S. Greenberg, Proc. Soc. Exp. Biol. Med. 205 (1994) 220–225. with pyruvate and/or other carbonyl metabolites, and [10] S. Agalou, N. Karachalias, B. Tucker, P.J. Thornalley, A. suppression of early glycation may be involved. Dawnay, in: S. Horiuchi, N. Taniguchi, F. Hayase, T. Kurata, The high chemical reactivity of AG is important in T. Osawa (Eds.), The Maillard Reaction in Food Chemistry and Medical Science: Update for the Postgenomic Era, Elsevier, intercepting physiological a-oxoaldehydes before AGE Tokyo, 2002, pp. 181–182. formation. The failure of AG to achieve a successful [11] D.V. Zyzak, J.M. Richardson, S.R. Thorpe, J.W. Baynes, Arch. clinical trial may be attributed to its rapid renal clear- Biochem. Biophys. 316 (1995) 547–554. ance, moderate a,b-dicarbonyl scavenging at pharma- [12] D.B. Shin, M.S. Feather, Carbohydr. Res. 208 (1990) 246–250. cological concentrations in vivo, and toxicity. [13] S. Agalou, N. Karachalias, A.B. Dawnay, P.J. Thornalley, in: S. Horiuchi, N. Taniguchi, F. Hayase, T. Kurata, T. Osawa (Eds.), Hydrophobic derivatives of AG may improve AG de- The Maillard Reaction in Food Chemistry and Medical Science: rivative binding to serum albumin (and other sites) to Update for the Postgenomic Era, Elsevier, Tokyo, 2002, pp. prolong its pharmacokinetic half-life but this may ex- 513–515. acerbate toxicity. Possibly, a prodrug of AG activated in [14] J.R. Nyengaard, K. Chang, S. Berhorst, K.M. Reiser, J.R. vascular sites where a-oxoaldehyde scavenging is Williamson, R.G. Tilton, Diabetes 46 (1997) 94–106. [15] E.F. Foote, Z.M. Look, P.K.W.F.H.C.E. Giles, Am. J. Kidney required may be a valuable future development. Phar- Dis. 25 (1995) 420–425. macological scavenging of a-oxoaldehydes or stimula- [16] A. Araki, M.A. Glomb, M. Takahashi, V.M. Monnier, in: J. tion of host a-oxoaldehyde detoxification remains a OÕBrien, H.E. Nursten, M.J.C. Crabbe, J.M. Ames (Eds.), The worthy therapeutic strategy to prevent diabetic compli- Maillard Reaction in Foods and Medicine, RSC Publ., Cam- cations and other AGE-related disorders. bridge, UK, 1998, p. 400. [17] M.-X. Fu, K.J. Wells-Knecht, J.A. Blackledge, T.J. Lyons, S.R. Thorpe, J.W. Baynes, Diabetes 43 (1994) 676–683. [18] H. Oxlund, T.T. Andreassen, Diabetologia 35 (1992) 19–25. References [19] W.K. Alderton, C. Cooper, R.G. Knowles, Biochem. J. 357 (2001) 593–615. [1] M. Brownlee, H. Vlassara, A. Kooney, P. Ulrich, A. Cerami, [20] T.W.C. Lo, T. Selwood, P.J. Thornalley, Biochem. Pharmacol. 48 Science 232 (1986) 1629–1632. (1994) 1865–1870. [2] N. Ahmed, O.K. Argirov, H.S. Minhas, C.A. Cordeiro, P.J. [21] T.J. Lyons, W. Li, M.C. Wells-Knecht, R. Joki, Diabetes 43 Thornalley, Biochem. J. 364 (2002) 1–14. (1994) 1090–1095. [3] N. Ahmed, P.J. Thornalley, Biochem. J. 364 (2002) 15–24. [22] P.H. Yu, D.M. Zuo, Diabetologia 40 (1997) 1243–1250. [4] P.J. Thornalley, A. Yurek-George, O.K. Argirov, Biochem. [23] P.J. Thornalley, Mol. Aspects Med. 14 (1993) 287–371. Pharmacol. 60 (2000) 55–65. [24] J.R. Williamson, K. Chang, M. Frangos, K.S. Hasan, T. Ido, T. [5] T.P. Degenhardt, S.R. Thorpe, J.W. Baynes, Cell. Mol. Biol. 44 Kawamura, J.R. Nyengaard, M. van den Enden, C. Kilo, R.G. (1998) 1139–1145. Tilton, Diabetes 42 (1993) 801–813. P.J. Thornalley / Archives of Biochemistry and Biophysics 419 (2003) 31–40 39

[25] Y. Tasaka, F. Nakaya, H. Matsumoto, Y. Omori, Endocr. J. 41 [54] D.J. Kelly, R.E. Gilbert, A.J. Cox, T. Soulis, G. Jerums, M.E. (1994) 309–313. Cooper, J. Am. Soc. Nephrol. 12 (2001) 2098–2107. [26] D. Zhaner, W.J. Malaisse, Diabetes Nutr. Metab. 5 (1992) [55] C.D. Agardh, U. Stenram, O. Torffvit, E. Agardh, J. Diabetes 43–46. Complic. 16 (2002) 395–400. [27] T. Nishikawa, D. Edelstein, X. Liang Du, S. Yamagishi, T. [56] P.S. Oturai, R. Rasch, E. Haselager, P.B. Johansen, H. Yokoy- Matsumura, Y. Kaneda, M.A. Yorek, D. Beede, P.J. Oates, H.-P. ama, M.K. Thomsen, B. Myrup, A. KofoedEnevoldsen, T. Hammes, I. Giardino, M. Brownlee, Nature 404 (2000) Deckert, APMIS 104 (1996) 259–264. 787–790. [57] T.M. Osicka, Y. Yu, S. Panagiotopoulos, S.P. Clavant, Z. [28] R. Chibber, G.M. Molinatti, J.S.K. Wong, D. Mirrlees, E.M. Kiriazis, R.N. Pike, L. Pratt, L.M. Russo, B.E. Kemp, W.D. Kohner, Diabetes 43 (1994) 758–763. Comper, G. Jerums, Diabetes 49 (2000) 87–93. [29] J. Hirsch, E. Petrakova, M.S. Feather, C.L. Barnes, Carbohydr. [58] V. Scivittaro, M.B. Ganz, M.F. Weiss, Am. J. Physiol. Renal Res. 267 (1995) 17–25. Physiol. 278 (2000) F676–F683. [30] J.R. Requena, P. Vidal, J. Cabezascerrato, Diabetes Res. Clin. [59] T.M. Osicka, Y. Yu, V. Lee, S. Panagiotopoulos, B.E. Kemp, G. Pract. 19 (1993) 23–30. Jerums, Clin. Sci. 100 (2001) 149–257. [31] M. Kuzuya, T. Asai, S. Kanda, K. Maeda, X.W. Cheng, A. [60] Y.F. Huang, S.Y. Lin, J.H. Zhou, Chin. Med. J. 111 (1998) Iguchi, Diabetologia 44 (2001) 436. 698–704. [32] R. Bucala, Z. Makita, G. Vega, S. Grundy, T. Koschinsky, A. [61] D. Edelstein, M. Brownlee, Diabetologia 35 (1992) 96–97. Cerami, H. Vlassara, Proc. Natl. Acad. Sci. USA 91 (1994) [62] J.L. Wilkinson-Berka, D.J. Kelly, S.M. Koerner, K. Jaworski, B. 9441–9445. Davis, V. Thallas, M.E. Cooper, Diabetes 51 (2002) [33] P. Mene, C. Pascale, A. Teti, S. Bernardini, G.A. Cinotti, F. 3283–3289. Pugliese, J. Am. Soc. Nephrol. 10 (1999) 1478–1486. [63] T.M. Osicka, Z. Kiriazis, L.M. Pratt, G. Jerums, W.D. Comper, [34] P.M. Ou, S.P. Wolff, Biochem. Pharmacol. 46 (1993) Diabetologia 44 (2001) 230–236. 1139–1144. [64] A. Yamauchi, A. Takei, Z. Makita, S. Nakamoto, N. Ohashi, H. [35] I. Giardino, A.K. Fard, D.L. Hatchell, M. Brownlee, Diabetes 47 Kiguchi, T. Ishii, T. Koike, T. Sartuta, Diabetes Res. Clin. Pract. (1998) 1114–1120. 34 (1997) 127–133. [36] S.H. Ihm, H.J. Yoo, S.W. Park, J. Ihm, Metab. Clin. Exp. 48 [65] A.M. Birrell, S. Hefferman, P. Kirwan, S. McLennan, A.G. Gillin, (1999) 1141–1145. D.K. Yue, J. Diabetes Complic. 16 (2002) 301–309. [37] Y. Al-Abed, R. Bucala, Chem. Res. Toxicol. 10 (1997) [66] M. Porta, F. Bandello, Diabetologia 45 (2002) 1617–1634. 875–879. [67] M. Lorenzi, C. Gerhardinger, Diabetologia 44 (2001) [38] D.L. Price, P.M. Rhett, S.R. Thorpe, J.W. Baynes, J. Biol. Chem. 791–804. 276 (2001) 48967–48972. [68] H.-P. Hammes, S. Martin, K. Federlin, K. Geisen, M. Brownlee, [39] G. Yildiz, A.T. Demiryurek, I. Sahen-Erdemli, I. Kanzik, Br. J. Proc. Natl. Acad. Sci. USA 88 (1991) 11555–11558. Pharmacol. 124 (1998) 905–910. [69] H.-P. Hammes, S. Syed, M. Uhlmann, A. Weiss, K. Federlin, K. [40] E.A. Abordo, H.S. Minhas, P.J. Thornalley, Biochem. Pharmacol. Geisen, M. Brownlee, Diabetologia 38 (1995) 269–273. 58 (1999) 641–648. [70] H.-P. Hammes, M. Brownlee, D. Edelstein, M. Saleck, S. Martin, [41] T.K. Shires, J. Tresnak, M. Kaminsky, S.L. Herzog, B. Truc- K. Federlin, Diabetologia 37 (1994) 32–35. Pham, FASEB J. 4 (1990) 3340–3345. [71] H.-P. Hammes, D. Strodter, A. Weiss, R.G. Bretzel, K. Federlin, [42] K. Shimoi, A. Okitau, M.H.L. Green, J.K. Lowe, T. Ohta, K. M. Brownlee, Diabetologia 38 (1995) 656–660. Kaji, H. Terato, H. Ide, N. Kinae, Mutat. Res. Fundam. Mol. [72] M. Inoue, N. Ohgiya, M. Yamamoto, Brain Res. 800 (1998) 319– Mech. Mutagen. 480 (2001) 371–378. 322. [43] M. Brownlee, Nature 414 (2001) 813–820. [73] E. Roufail, T. Soulis, E. Boel, M.E. Cooper, S. Rees, Diabetologia [44] M.P. Wautier, O. Chappey, S. Corda, D.M. Stern, A.-M. 41 (1998) 1419–1425. Schmidt, J.L. Wautier, Am. J. Physiol. 280 (2001) E685–E694. [74] T.S. Kern, R.L. Engerman, Diabetes 50 (2001) 1636–1642. [45] U. Hink, H. Li, H. Mollnau, M. Oelze, E. Matheis, M. Hartmann, [75] D. Ziegler, D. Luft, Int. Rev. Neurobiol. 50 (2002) 451–463. M. Skatchkov, F. Thaiss, R.A.K. Stahl, A. Warnholtz, T. [76] P.J. Thornalley, Int. Rev. Neurobiol. 50 (2002) 37–57. Meinertz, K. Griendling, D.G. Harison, U. Fostermann, T. [77] M. Kihara, J.D. Schmelzer, J.F. Poduslo, G.L. Curran, K.K. Munzel, Circ. Res. 88 (2001) 14–22. Nickander, P.A. Low, Proc. Natl. Acad. Sci. USA 88 (1991) [46] D.J. Wolff, A. Lubeskie, Arch. Biochem. Biophys. 316 (1995) 6107–6111. 290–301. [78] R.E. Schmidt, D.A. Dorsey, L.N. Beaudet, K.M. Resier, J.R. [47] H.-H. Parving, J.T. Harrington, J. Lewis, G. Boner, E.J. Lewis, G. Williamson, R.G. Tilton, Diabetes 45 (1996) 284–290. Viberti, R. Gilbert, I. Raz, M. Cooper, K. Rossing, Kidney Int. 60 [79] A.M. Birrell, S.J. Hefferman, A.D. Ansselin, S. McLennan, D.K. (2001) 2041–2055. Church, A.G. Gillin, D.K. Yue, Diabetologia 43 (2000) [48] A. Heidland, K. Sebekova, R. Schinzel, Am. J. Kidney Dis. 34 110–116. (2001) S100–S106. [80] S. Swamy-Mruthinti, K. Green, E.C. Abraham, Exp. Eye Res. 62 [49] T. Soulis-Liparota, M. Cooper, D.X. Papazoglou, B. Clarke, G. (1996) 505–510. Jerums, Diabetes 40 (1991) 1328–1334. [81] J.J. Cartledge, I. Eardley, J.F.B. Morrison, BJU Int. 87 (2001) [50] J.M. Forbes, T. Soulis, V. Thallas, S. Panagiotopoulos, D.M. 402–407. Long, S. Vasan, D. Wagle, G. Jerums, M.E. Cooper, Diabetologia [82] V. Piercy, N. Toseland, N.C. Turner, Metabolism 47 (1998) 44 (2001) 108–114. 1477–1480. [51] T.P. Degenhardt, M.-X. Fu, E. Voss, K. Reiff, R. Neidlein, K. [83] P.J. Beisswenger, M.E. Wood, S.K. Howell, A.D. Touchette, Strein, S.R. Thorpe, J.W. Baynes, R. Reiter, Diabetes Res. Clin. R.M. OÕDell, B.S. Szwergold, Diabetes Care 24 (2001) Pract. 43 (1999) 81–89. 726–732. [52] T. Soulis, M.E. Cooper, S. Sastra, V. Thalias, S. Panagiotopoulos, [84] P. Sannomiya, M.A. Olivereira, Z.B. Fortes, Br. J. Pharmacol. O.J. Bjerrum, G. Jerums, Diabetologia 40 (1997) 1141–1151. 122 (1997) 894–898. [53] J.M. Forbes, M.E. Cooper, V. Thalias, W.C. Burns, M.C. [85] D. Dyntar, M. Eppenberger-Eberhardt, K. Maedler, M. Fruschy, Thomas, G.C. Brammar, F. Lee, S.L. Grant, L.A. Burrell, G. H.M. Eppenberger, G.A. Spinas, M.Y. Donath, Diabetes 50 Jerums, T.M. Osicka, Diabetes 51 (2002) 3274–3282. (2001) 2105–2113. 40 P.J. Thornalley / Archives of Biochemistry and Biophysics 419 (2003) 31–40

[86] B. Corman, M. Duriez, P. Poitevein, D. Heudes, P. Bruneval, A. [92] M. Iqbal, L.L. Probert, H. Klandorf, Poultry Sci. 76 (1997) 1574– Tedgui, B.I. Levy, Proc. Natl. Acad. Sci. USA 95 (1998) 1579. 1301–1306. [93] H. Fujisawa, T. Nishikawa, B.H. Zhu, Y. Nishimura, M. Shimizu, [87] Y.M. Li, M. Steﬀes, T. Donnelly, C. Liu, H. Fuh, J. Basgen, R. M. Kimoto, K. Higuchi, M. Hosokawa, J. Gerontol. Ser. A Biol. Bucala, H. Vlassara, Proc. Natl. Acad. Sci. USA 93 (1996) Sci. Med. Sci. 54 (1999) B276–B282. 3902–3907. [94] S.M. McCann, J. Lincinio, M.L. Wong, W.H. Yu, S. Karanth, V. [88] J.F. Reckelhoﬀ, B.S. Hennington, V. Kanji, L.C. Racusen, A.M. Rettorri, Exp. Gerontol. 33 (1998) 813–826. Schmidt, S.D. Yan, J. Morrow, L.J. Roberts, A.K. Salahudeen, [95] B.I. Freedman, J.-P. Wuerth, K. Cartwright, R.P. Bain, S. Dippe, Am. J. Hypertens. 12 (1999) 492–498. K. Hershon, A.D. Mooradian, B.S. Spinowitz, Control. Clin. [89] D.R. Sell, J.F. Nelson, V.M. Monnier, J. Gerontol. Ser. A Biol. Trials 20 (1999) 493–510. Sci. Med. Sci. 56 (2001) B405–B411. [96] T. Taguchi, M. Sugiura, Y. Hamada, I. Miwa, Biochem. [90] A.J. Oudes, C.M. Herr, Y. Olsen, J.E. Fleming, Mech. Ageing Pharmacol 55 (1998) 1667–1671. Dev. 100 (1998) 221–229. [97] H. Miyoshi, T. Taguchi, M. Sugiura, M. Takeuchi, K. Yanagis- [91] M. Iqbal, L.L. Probert, N.H. Alhumadi, H. Klandorf, J. awa, Y. Watanabe, I. Miwa, Z. Makita, T. Koike, Horm. Metab. Gerontol. Ser. A Biol. Sci Med. Sci. 54 (1999) B171–B176. Res. 34 (2002) 371–377. 本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP 图书馆。图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页文献云下载图书馆入口外文数据库大全疑难文献辅助工具