NEUROVASCULAR MEDICINE - Pursuing Cellular Longevity for Healthy Aging Part 10 pptx

528 NOVEL CELLULAR PATHWAYS

The local RAS is activated under diabetes

(Anderson 1997). We have recently found that

angiotensin II (Ang II) stimulated intracellular ROS

generation in retinal pericytes through an interac￾tion with type 1 receptor. Further, Ang II decreased

DNA synthesis and simultaneously upregulated

VEGF mRNA levels in pericytes, both of which

were blocked by treatment with telmisartan, a com￾mercially available Ang II type 1 receptor blocker,

or an antioxidant NAC (Yamagishi, Amano, Inagaki

et al. 2003; Amano, Yamagishi, Inagaki et al. 2003).

These results suggest that Ang II-type 1 receptor

interaction could induce pericyte loss and dysfunc￾tion through intracellular ROS generation, thus

being involved in diabetic retinopathy. Since Ang II

induces the VEGF receptor, KDR, expression in reti￾nal microvascular ECs, the retinal RAS might aug￾ment the permeability- and angiogenesis-inducing

activity of VEGF, thus implicated in the progression of

diabetic retinopathy as well (Otani, Takagi, Suzuma

et al. 1998).

Blockade of the RAS by inhibitors of angiotensin

converting enzyme or Ang II type 1 receptor antag￾onists can reduce retinal overexpression of VEGF

and hyperpermeability and neovascularization in

experimental diabetes (Babaei-Jadidi, Karachalias,

Ahmed et al. 2003; Anderson 1997; Yamagishi,

Amano, Inagaki et al. 2003). Funatsu et al. (Funatsu,

Yamashita, Nakanishi et al. 2002) recently found that

the vitreous fl uid level of Ang II was signifi cantly

correlated with that of VEGF, and both of them were

signifi cantly higher in patients with active prolifer￾ative diabetic retinopathy than in those with qui￾escent proliferative diabetic retinopathy (Amano,

Yamagishi, Inagaki et al. 2003). These fi ndings fur￾ther support the concept that Ang II contributes to

development and progression of proliferative dia￾betic retinopathy in combination with VEGF. In the

EUCLID Study, the angiotensin-converting enzyme

inhibitor, lisinopril, reduced the risk of progression

of retinopathy by approximately 50% and also sig￾nifi cantly reduced the risk of progression to prolif￾erative retinopathy although retinopathy was not a

primary end point and the study was not suffi ciently

powered for eye-related outcomes (Otani, Takagi,

Suzuma et al. 1998). The interaction of the RAS and

AGE–RAGE system has also been proposed. We have

found that Ang II potentiates the deleterious effects

of AGEs on pericytes by inducing RAGE protein

expression (Yamagishi, Takeuchi, Matsui et al. 2005).

In vivo, AGE injection stimulated RAGE expression

in the eye of spontaneously hypertensive rats, which

was blocked by telmisartan. In vitro, Ang II-type 1

receptor- mediated ROS generation elicited RAGE

gene expression in retinal pericytes through NF-κB

the characteristic changes of the early phase of dia￾betic retinopathy, in streptozotocin-induced diabetic

rats. These observations suggest that SDH-mediated

conversion of sorbitol into fructose and the resultant

ROS generation may play a role in the pathogenesis of

diabetic retinopathy. Since fructose is a stronger gly￾cating agent than glucose, intracellular AGEs forma￾tion via the SDH pathway might be involved in glucose

toxicity to retinal pericytes (Rosen, Nawroth, King

et al. 2001).

There is a growing body of evidence that gener￾ation of ROS is increased in diabetes. High glucose

concentrations, via various mechanisms such as

glucose autoxidation, increased the production

of AGEs, activation of PKC, and stimulation of the

polyol pathway, and it enhanced ROS generation

(Rosen, Nawroth, King et al. 2001; Bonnefont￾Rousselot 2002).Increased ROS generation has been

found to regulate vascular infl ammation, altered

gene expression of growth factors and cytokines,

and platelet and macrophage activation, thus play￾ing a central role in the pathogenesis of diabetic vas￾cular complications (Yamagishi, Edelstein, Du et al.

2001; Yamagishi, Edelstein, Du et al. 2001; Yamagishi,

Okamoto, Amano et al. 2002; Spitaler, Graier 2002;

Yamagishi, Inagaki, Amano et al. 2002; Yamagishi

S, Amano S, Inagaki et al. 2003). Further, we have

recently found that high glucose–induced mito￾chondrial overproduction of superoxide serves as a

causal link between elevated glucose and hyperglyce￾mic vascular damage in ECs (Nishikawa, Edelstein,

Du et al. 2000; Brownlee 2001). Normalizing levels

of mitochondrial ROS prevent glucose-induced for￾mation of AGEs, activation of PKC, sorbitol accu￾mulation, and NF-κB activation. These observations

suggest that the three main mechanisms implicated

in the pathogenesis of diabetic vascular complica￾tions might refl ect a single hyperglycemia-induced

process, thus providing a novel therapeutic tar￾get for diabetic angiopathies. Recently, Hammes

et al. (Hammes, Du, Edelstein et al. 2003) have dis￾covered that the lipid-soluble thiamine derivative

benfotiamine can inhibit the three major biochem￾ical pathways as well as hyperglycemia-associated

NF-κB activation (Hammes, Du, Edelstein et al.

2003). They showed that benfotiamine prevented

experimental diabetic retinopathy by activating

the pentose phosphate pathway enzyme, transketo￾lase, in the retinas, which converts glyceraldehyde￾3-phosphate and fructose-6-phosphate into pentose￾5-phosphates and other sugars (Hammes, Du,

Edelstein et al. 2003). Thiamine and benfotiamine

therapy is reported to prevent streptozotocin-induced

incipient diabetic nephropathy as well (Babaei-Jadidi,

Karachalias, Ahmed et al. 2003).

Chapter 21: Diabetic Vascular Complications 529

(Cooper, Bonnet, Oldfi eld et al. 2001). Evidence has

implicated the TGF-β system as a major etiologic

agent in the pathogenesis of glomerulosclerosis and

tubulointerstitial fi brosis in diabetic nephropathy

(Sharma, Ziyadeh 1995; Aoyama, Shimokata, Niwa

2000; Wang, LaPage, Hirschberg 2000).

AGEs induce apoptotic cell death and VEGF

expression in human-cultured mesangial cells, as

the case in pericytes (Yamagishi, Inagaki, Okamoto

et al. 2002). Mesangial cells occupy a central ana￾tomical position in the glomerulus, playing crucial

roles in maintaining structure and function of glo￾merular capillary tufts (Dworkin, Ichikawa, Brenner

1983). They actually provide structural support for

capillary loops and modulate glomerular fi ltration

by its smooth muscle activity (Dworkin, Ichikawa,

Brenner 1983; Kreisberg, Venkatachalam, Troyer

1985; Schlondorff 1987). Therefore, it is conceivable

that the AGE-induced mesangial apoptosis and dys￾function may contribute in part to glomerular hyper￾fi ltration, an early renal dysfunction in diabetes.

Several experimental and clinical studies support the

pathological role for VEGF in diabetic nephropathy.

Indeed, antibodies against VEGF have been found to

improve hyperfi ltration and albuminuria in strepto￾zotocin-induced diabetic rats (De Vriese, Tilton, Elger

et al. 2001). Inhibition of VEGF also prevents glomer￾ular hypertrophy in a model of obese type 2 diabetes,

the Zucker diabetic fatty rat (Schrijvers, Flyvbjerg,

Tilton et al. 2006). Further, urinary VEGF levels are

positively correlated with the urinary albumin to cre￾atinine ratio and negatively correlated with creatinine

clearance in type 2 diabetic patients (Kim, Oh, Seo

et al. 2005). These observations suggest that urinary

VEGF might be used as a sensitive marker of diabetic

nephropathy. VEGF overproduction elicited by AGEs

may be involved in diabetic nephropathy.

Moreover, we have recently found that AGE–RAGE

interaction stimulates MCP-1 expression in mesangial

cells through ROS generation (Yamagishi, Inagaki,

Okamoto et al. 2002). Increased MCP-1 expression

associated with monocyte infi ltration in mesangium

has been observed in the early phase of diabetic

nephropathy as well (Banba, Nakamura, Matsumura

et al. 2000). Plasma MCP-1 was positively correlated

with urinary albumin excretion rate in type 1 diabetic

patients (Chiarelli, Cipollone, Mohn et al. 2002). AGE

accumulation in glomerulus could also be implicated

in the initiation of diabetic nephropathy by promot￾ing the secretion of MCP-1.

AGE formation on extracellular matrix proteins

alters both matrix–matrix and cell–matrix interac￾tions, involved in the pathogenesis of diabetic glom￾erulosclerosis. For example, nonenzymatic glycations

of type IV collagen and laminin reduce their ability

activation. Further, Ang II augmented AGE-induced

pericyte apoptosis, the earliest hallmark of diabetic

retinopathy. Further, we have recently found that

telmisartan blocks the Ang II-induced RAGE expres￾sion in ECs as well (Nakamura, Yamagishi, Nakamura

et al. 2005). Telmisartan could decrease endothelial

RAGE levels in patients with essential hypertension.

Taken together, these observations provide the func￾tional interaction between the AGE–RAGE system and

the RAS in the pathogenesis of diabetic retinopathy,

thus suggesting a novel benefi cial aspect of telmisar￾tan on the devastating disorder. We posit a table that

presents the etiologies of diabetic retinopathy and its

possible therapeutic agents (Table 21.2).

ROLE OF AGES IN DIABETIC

NEPHROPATHY

Diabetic nephropathy is a leading cause of ESRD and

accounts for disabilities and the high mortality rate in

patients with diabetes (Krolewski, Warram, Valsania

et al. 1991). Development of diabetic nephropathy

is characterized by glomerular hyperfi ltration and

thickening of glomerular basement membranes, fol￾lowed by an expansion of extracellular matrix in

mesangial areas and increased urinary albumin excre￾tion rate (UAER). Diabetic nephropathy ultimately

progresses to glomerular sclerosis associated with

renal dysfunction (Sharma, Ziyadeh 1995). Further, it

has recently been recognized that changes within tub￾ulointerstitium, including proximal tubular cell atro￾phy and tubulointerstitial fi brosis, are also important

in terms of renal prognosis in diabetic nephropathy

(Ziyadeh, Goldfarb 1991; Lane, Steffes, Fioretto et al.

1993; Taft, Nolan, Yeung et al. 1994; Jones, Saunders,

Qi et al. 1999; Gilbert, Cooper 1999). Such tubular

changes have been reported to be the dominant lesion

in about one-third of patients with type 2 diabetes

(Fiorreto, Mauer, Brocco et al. 1996). It appears that

both metabolic and hemodynamic factors interact to

stimulate the expression of cytokines and growth fac￾tors in glomeruli and tubules from the diabetic kidney

Table 21.2 Diabetic Retinopathy

Etiology Cellular Pathway Treatment Regimen

AGE–RAGE VEGF Pimagedine

ROS ICAM-1 Amadorins

Polyol pathway MCP-1 OPB-9195

PKC PAI-1 sRAGE

RAS Angiopoietins PEDF

Benfotiamine

Telmisartan

530 NOVEL CELLULAR PATHWAYS

malondialdehyde-lysine accumulate in the expanded

mesangial matrix and thickened glomerular base￾ment membranes of early diabetic nephropathy, and

in nodular lesions of advanced disease, further sug￾gesting the active role of AGEs for diabetic nephropa￾thy (Suzuki, Miyata, Saotome et al. 1999).

A number of studies have demonstrated that amin￾oguanidine decreased AGE accumulation and plasma

protein trapping in the glomerular basement mem￾brane (Matsumura, Yamagishi, Brownlee 2000). In

streptozocin-induced diabetic rats, aminoguanidine

treatment for 32 weeks dramatically reduced the level

of albumin excretion and prevented the development

of mesangial expansion (Soulis-Liparota Cooper,

Papazoglou et al. 1991). Furthermore, aminoguani￾dine treatment was found to prevent albuminuria

in diabetic hypertensive rats without affecting blood

pressure (Edelstein, Brownlee 1992). Whether inhi￾bition by aminoguanidine of inducible nitric oxide

synthase (iNOS) could contribute to these renopro￾tective effects remains to be elucidated. However,

methylguanidine, which inhibits iNOS but not AGE

formation, was reported not to retard the develop￾ment of albuminuria in diabetic rats (Soulis, Cooper,

Sastra et al. 1997). These observations suggest that the

benefi cial effects of aminoguanidine could be medi￾ated predominantly by decreased AGE formation

rather than by iNOS inhibition. A recent randomized,

double-masked, placebo-controlled study (ACTION I

trial) revealed that pimagedineR (aminoguanidine)

reduced the decrease in glomerular fi ltration rate and

24-hour total proteinuria in type 1 diabetic patients

(Bolton, Cattran, Williams et al. 2004). Although the

time for doubling of serum creatinine, a primary end

point of this study, was not signifi cantly improved by

pimagedineR treatment (P = 0.099), the trial provided

the fi rst clinical proof of the concept that blockade of

AGE formation could result in a signifi cant attenua￾tion of diabetic nephropathy.

We have found that OPB-9195, a synthetic thiazoli￾dine derivative and novel inhibitor of AGEs, prevented

the progression of diabetic nephropathy by lowing

serum concentrations of AGEs and their deposition

of glomeruli in Otsuka–Long–Evans–Tokushima–

Fatty rats, a type 2 diabetes mellitus model animal

(Tsuchida, Makita, Yamagishi et al. 1999). OPB-9195

was also found to retard the progression of diabetic

nephropathy by blocking type IV collagen produc￾tion and suppressing overproduction of two growth

factors, TGF-β and VEGF.

Recently, Degenhardt and Baynes et al. (Degen￾hardt, Alderson, Arrington et al. 2002) reported that

pyridoxamine inhibited the progression of renal dis￾ease and decreases hyperlipidemia and apparent

redox imbalances in diabetic rats. Pyridoxamine and

aminoguanidine had similar effects on parameters

to interact with negatively charged proteoglycans,

increasing vascular permeability to albumin (Silbiger,

Crowley, Shan et al. 1993). Furthermore, AGE forma￾tion on various types of matrix proteins impairs their

degradation by matrix metalloproteinases, contribut￾ing to basement membrane thickening and mesan￾gial expansion, hallmarks of diabetic nephropathy

(Brownlee 1993; Mott, Khalifah, Nagase et al. 1997).

AGEs formed on the matrix components can trap and

covalently cross-link with the extravasated plasma

proteins such as lipoproteins, thereby exacerbating

diabetic glomerulosclerosis (Brownlee 1993).

AGEs stimulate insulin-like growth factor-I, -II,

PDGF and TGF-β in mesangial cells, which in turn

mediate production of type IV collagen, laminin, and

fi bronectin (Matsumura, Yamagishi, Brownlee 2000;

Yamagishi, Takeuchi, Makita 2001). AGEs induce

TGF-β overexpression in both podocytes and proxi￾mal tubular cells as well (Wendt TM, Tanji N, Guo J,

et al. 2003; Yamagishi, Inagaki, Okamoto et al. 2003).

Recently, Ziyadeh et al. (2000) reported that long￾term treatment of type 2 diabetic model mice with

blocking antibodies against TGF-β suppressed excess

matrix gene expression, glomerulosclerosis, and pre￾vented the development of renal insuffi ciency. These

observations suggest that AGE-induced TGF-β expres￾sion plays an important role in the pathogenesis of

glomerulosclerosis and tubulointerstitial fi brosis in

diabetic nephropathy (Raj, Choudhury, Welbourne

et al. 2000; Yamagishi, Koga, Inagaki et al. 2002).

In vivo, the administration of AGE-albumin to

normal healthy mice for 4 weeks has been found to

induce glomerular hypertrophy with overexpression

of type IV collagen, laminin B1, and TGF-β genes

(Yang, Vlassara, Peten et al. 1994). Furthermore,

chronic infusion of AGE-albumin to otherwise

healthy rats leads to focal glomerulosclerosis, mesan￾gial expansion, and albuminuria (Vlassara H,

Striker LJ, Teichberg et al. 1994). Recently, RAGE￾overexpressing diabetic mice have been found to show

progressive glomerulosclerosis with renal dysfunc￾tion, compared with diabetic littermates lacking the

RAGE transgene (Yamamoto, Kato, Doi et al. 2001).

Further, diabetic homozygous RAGE null mice failed

to develop signifi cantly increased mesangial matrix

expansion or thickening of the glomerular basement

membrane (Wendt, Tanji, Guo et al. 2003). Taken

together, these fi ndings suggest that the activation of

AGE–RAGE axis contributes to expression of VEGF

and enhanced attraction/activation of infl ammatory

cells in the diabetic glomerulus, thereby setting the

stage for mesangial activation and TGF-β production;

processes that converge to cause albuminuria and

glomerulosclerosis.

AGEs including glycoxidation or lipoxidation pro￾ducts such as Nε-(carboxymethyl)lysine, pentosidine,

Chapter 21: Diabetic Vascular Complications 531

ROLE OF AGES IN CVD

Atherosclerotic arterial disease may be manifested

clinically as CVD. Deaths from CVD predominate

in patients with diabetes of over 30 years’ duration

and in those diagnosed after 40 years of age. CVD

is responsible for about 70% of all causes of death

in patients with type 2 diabetes (Laakso 1999). In

Framingham study, the incidence of CVD was 2 to

4 times greater in diabetic patients than in general

polulation (Haffner, Lehto, Ronnemaa et al. 1998).

Conventional risk factors, including hyperlipidemia,

hypertension, smoking, obesity, lack of exercise,

and a positive family history, contribute similarly

to macrovascular complications in type 2 diabetic

patients and nondiabetic subjects (Laakso 1999).

The levels of these factors in diabetic patients were

certainly increased, but not enough to explain the

exaggerated risk for macrovascular complications in

diabetic population (Standl, Balletshofer, Dahl et al.

1996). Therefore, specifi c diabetes-related risk fac￾tors should be involved in the excess risk in diabetic

patients.

A variety of molecular mechanisms underlying

the actions of AGEs and their contribution to diabetic

macrovascular complications have been proposed

(Stitt, Bucala, Vlassara 1997; Bierhaus, Hofmann,

Ziegler et al. 1998; Schmidt, Stern 2000; Vlassara,

Palace 2002; Wendt, Bucciarelli, Qu et al. 2002).

AGEs formed on the extracellular matrix results in

decreased elasticity of vasculatures, and quench nitric

oxide, which could mediate defective endothelium￾dependent vasodilatation in diabetes (Bucala, Tracey,

Cerami 1991). AGE modifi cation of low-density lipo￾protein (LDL) exhibits impaired plasma clearance

and contributes signifi cantly to increased LDL in

vivo, thus being involved in atherosclerosis (Bucala,

Mitchell, Arnold et al. 1995). Binding of AGEs to

RAGE results in generation of intracellular ROS

generation and subsequent activation of the redox￾sensitive transcription factor NF-κB in vascular wall

cells, which promotes the expression of a variety of

atherosclerosis-related genes, including ICAM-1, vas￾cular cell adhesion molecule-1, MCP-1, PAI-1, tissue

factor, VEGF, and RAGE (Stitt, Bucala, Vlassara 1997;

Bierhaus, Hofmann, Ziegler et al. 1998; Schmidt,

Stern 2000; Tanaka, Yonekura, Yamagishi et al. 2000;

Vlassara, Palace 2002; Wendt, Bucciarelli, Qu et al.

2002). AGEs have the ability to induce osteoblas￾tic differentiation of microvascular pericytes, which

would contribute to the development of vascular cal￾cifi cation in accelerated atherosclerosis in diabetes as

well (Yamagishi, Fujimori, Yonekura et al. 1999). The

interaction of the RAS and AGEs in the development

of diabetic macrovascular complications has also

been proposed. AGE–RAGE interaction augments

measured, supporting a mechanism of action involv￾ing AGE inhibition (Degenhardt, Alderson, Arrington

et al. 2002). Although the results of AGE inhibitors

in animal models of diabetic nephropathy are prom￾ising, effectiveness of these AGE inhibitors must be

confi rmed by multicenter, randomized, double-blind

clinical studies.

Cross Talk between the AGE–RAGE Axis

and the RAS in Diabetic Nephropathy

Recent experiments have focused on the interaction

of the AGE–RAGE axis and the RAS thought to be

critical to the development of diabetic nephropathy.

Indeed, angiotensin converting enzyme inhibition

reduces the accumulation of renal and serum AGEs,

probably via effects on oxidative pathways (Forbes,

Cooper, Thallas et al. 2002). Long-term treatment

with Ang II receptor 1 antagonist may exert salu￾tary effects on AGEs levels in the rat remnant kid￾ney model, probably due to improved renal function

(Sebekova, Schinzel, Munch et al. 1999). Ramipril

administration has been recently shown to result in a

mild decline of fl uorescent non-carboxymethyllysine￾AGEs and malondialdehyde concentrations in nondi￾abetic nephropathy patients (Sebekova, Gazdikova,

Syrova et al. 2003). Further, we have recently found

that the AGE–RAGE-mediated ROS generation

activates TGF-β-Smad signaling and subsequently

induces mesangial cell hypertrophy and fi bronectin

synthesis by autocrine production of Ang II (Fukami,

Ueda, Yamagishi et al. 2004). In addition, AGEs

induce mitogenesis and collagen production in renal

interstitial fi broblasts as well via Ang II-connective

tissue growth factor pathway (Lee, Guh, Chen et al.

2005). Moreover, olmesartan medoxomil, an Ang II

type 1 receptor blocker, protects against glomeru￾losclerosis and renal tubular injury in AGE-injected

rats, thus further supporting the concept that AGEs

could induce renal damage in diabetes partly via the

activation of RAS (Yamagishi, Takeuchi, Inoue et al.

2005). We posit a table that presents the etiologies

of diabetic nephropathy and its possible therapeutic

agents (Table 21.3).

Table 21.3 Diabetic Nephropathy

Etiology Cellular Pathway Treatment Regimen

AGE–RAGE VEGF Pimagedine

ROS MCP-1 Pyridoxamine

PKC TGF-β OPB-9195

RAS Smad Olmesartan

Hyperfi ltration

532 NOVEL CELLULAR PATHWAYS

fi vefold lower AGE content signifi cantly decreased

serum levels of AGEs, soluble form of VCAM-1 and

C-reactive protein (CRP), compared to equivalent

regular diets (Vlassara, Cai, Crandall et al. 2002).

AGE-poor diets also reduced peripheral mononu￾clear cell tumor necrosis factor-α (TNF-α) expres￾sion at both mRNA and protein levels (Vlassara, Cai,

Crandall et al. 2002). Further, LDL pooled from dia￾betic patients on a standard diet for 6 weeks (high

AGE-LDL) was more glycated and oxidized than

that from diabetic patients on an AGE-poor diet (low

AGE-LDL) (Cai, He, Zhu et al. 2004). High AGE-LDL

signifi cantly induced soluble form of VCAM-1 expres￾sion in human umbilical vein ECs via redox-sensitive

MAPK activation, compared to native LDL or low

AGE-LDL (Cai, He, Zhu et al. 2004). In addition, AGE

pronyl-glycine, a food-derived AGE, was reported to

elicit infl ammatory response to cellular proliferation

in an intestinal cell line, Caco-2, through the RAGE￾mediated MAPK activation (Zill, Bek, Hofmann et al.

2003). These observations suggest the causal link

between dietary intake of AGEs and proinfl amma￾tion and vascular injury, thus providing the clinical

relevance of dietary AGE restriction in the prevention

of accelerated atherosclerosis in diabetes. We have

very recently found that PAI-1 and fi brinogen levels

are positively associated with serum AGE levels in

nondiabetic general population. Food-derived AGEs

may also be associated with thrombogenic tendency

in nondiabetic subjects (Enomoto, Adachi, Yamagishi

et al. 2006).

CONCLUSION

In the DCCT-EDIC, the reduction in the risk of

progressive diabetic micro- and macroangipathies

resulting from intensive therapy in patients with

type 1 diabetes persisted for at least several years,

despite increasing hyperglycemia (DCCT-EDIC

Research Group 2000; Writing Team for DCCT-EDIC

Research Group 2003; Nathan, Lachin, Cleary et al.

2003; Nathan, Cleary, Backlund et al. 2005). These

clinical studies strongly suggest that so-called hyper￾glycemic memory is involved in the pathogenesis of dia￾betic vascular complications, AGE hypothesis seems

to be most compatible with this theory. Moreover,

large clinical investigations will be needed to clar￾ify whether the inhibition of AGE formation or the

blockade of their downstream signaling could pre￾vent the development and progression of vascular

complications in diabetes. Until the specifi c remedy

that targets diabetic vascular complications are devel￾oped, multifactorial intensifi ed intervention will be a

promising therapeutic strategy for the prevention of

these devastating disorders.

Ang II-induced smooth muscle cell proliferation and

activation, thus being involved in accelerated ath￾erosclerosis in diabetes (Shaw, Schmidt, Banes et al.

2003). AGEs have been actually detected within ath￾erosclerotic lesions in both extra- and intracellu￾lar locations (Nakamura, Horii, Nishino et al. 1993;

Niwa, Katsuzaki, Miyazaki et al. 1997; Sima, Popov,

Starodub et al. 1997).

In animal models, Park et al. (1998) has demon￾strated that diabetic apolipoprotein E (apoE) null

animals receiving soluble RAGE (sRAGE) display a

dose-dependent suppression of accelerated athero￾sclerosis in these mice. Lesions that formed in ani￾mals receiving sRAGE appeared largely arrested at

the fatty streak stage; the number of complex ath￾erosclerotic lesions was strikingly reduced in diabetic

apoE null mice. The tissue and plasma AGE burden

was suppressed in diabetic apoE null mice receiving

sRAGE, suggesting that the AGE–RAGE-induced oxi￾dative stress generation might participate in AGEs

formation themselves. Treatment with sRAGE did

not affect the levels of established risk factors in these

mice. These observations suggest the active involve￾ment of AGE–RAGE interaction in the pathogenesis

in accelerated atherosclerosis in diabetes. The same

group has recently reported that the AGE–RAGE sys￾tem contributes to the atherosclerotic lesion progres￾sion as well, and RAGE blockade stabilizes the lesions

in these mice (Bucciarelli, Wendt, Qu et al. 2002).

Another study shows a correlation between AGE lev￾els and the degree of atheroma in cholesterol-fed

rabbits, and aminoguanidine has an antiatherogenic

effect in these rabbits by inhibiting AGEs formation

(Panagiotopoulos, O’Brien, Bucala et al. 1998). In

humans, RAGE overexpression is associated with

enhanced infl ammatory reaction and cyclooxyge￾nase-2 and prostaglandin E synthase-1 expression in

diabetic plaque macrophages, and this effect may con￾tribute to plague destabilization by inducing culprit

metalloproteinase expression (Cipollone, Iezzi, Fazia

et al. 2003).

Recently, food-derived AGEs are reported to

induce oxidative stress and promote infl ammatory

signals (Cai, Gao, Zhu et al. 2002). Dietary glycotoxins

promote diabetic atherosclerosis in apoE-defi cient

mice (Lin, Reis, Dore et al. 2002; Lin, Choudhury, Cai

et al. 2003). Further, an AGE-poor diet that contained

four- to fi vefold lower AGE contents for 2 months

also decreased serum levels of AGEs and markedly

reduced tissue AGE and RAGE expression, numbers

of infl ammatory cells, tissue factor, VCAM-1, and

MCP-1 levels in diabetic apolipoprotein E-defi cient

mice (Lin, Choudhury, Cai et al. 2003).

Diet is a major environmental source of pro￾infl ammatory AGEs in humans as well (Vlassara, Cai,

Crandall et al. 2002). In diabetic patients, diets with