**4. Metabolic pathways implicated in hyperglycemia-induced lesion of vasculature**

The Diabetes Control and Complications Trial (DCCT) clinical trial confirmed that chronic hyperglycemia is detrimental in the development and progression of DR, though the exact mechanisms of microvascular lesions due to hyperglycemia are not yet fully understood [45]. A lot of interconnecting biochemical pathways are involved in hyperglycemia-induced vascular pathologies. Four major mechanisms explaining how hyperglycemia causes diabetic complications include: (1) increased glucose flux through the polyol pathway, (2) increased formation of advanced glycation end products (AGE), (3) activation of the protein kinase C (PKC) pathway, and (4) a fourth mechanism has been suggested recently: increased glucose metabolism through the hexosamine pathway [46].

#### **4.1 Increased polyol pathway flux**

Excessive glucose in diabetes is metabolized through the polyol pathway. In the polyol pathway, aldose reductase (AR) reduces glucose to sorbitol (polyol), using nicotinamide adenine dinucleotide phosphate (NADPH) as a cofactor. Subsequently, sorbitol is slowly metabolized into fructose by sorbitol dehydrogenase (SDH) with NAD+ reduced to NADH. Sorbitol is a sugar alcohol and strongly hydrophilic; therefore sorbitol cannot diffuse easily across the cell membrane. It was demonstrated that excessive intracellular storage of sorbitol results in hyperosmotic cellular damage [47]. Increased polyol pathway flux is considered to have several negative effects on retinal cells. Concomitant decrease of NADPH results in less NADPH availability for use by glutathione reductase, which uses NADPH as a cofactor to regenerate intracellular glutathione. Glutathione protects cells by neutralizing reactive oxygen species (ROS). Thereby, reduced NADPH in a hyperglycemic environment could course or exacerbate intracellular oxidative stress. Fructose produced by the polyol pathway is metabolized consequently to fructose-3-phosphate or 3-deoxyglucosone, which are potent glycating substances and can lead to generation of AGEs [48]. The presence of AR was shown in the ganglion retinal cells, Müller cell processes, retinal pigment epithelium, and the pericytes and endothelial cells of retinal capillaries in diabetic models in animals. These studies also pointed out increased apoptosis of pericytes due to AR activity [49, 50]. It was demonstrated in other work, however, the presence of AR in the cytoplasm of pericytes but not in endothelial cells in experimental diabetes [51]. Sato and colleagues observed accumulation of sugar alcohols in pericytes in contrast with similar cultured endothelial cells [52]. It was shown that AR was localized in human retinal pericytes but not in the endothelial cells [53, 54]. This data suggested that a selective degeneration and loss of retinal pericytes may be due to AR activity. Some studies showed that aldose reductase inhibitors (ARIs) were able to reduce the incidence and severity of diabetic retinal lesions occurring in the galactose-fed animals. The administration of ARIs to animal model of diabetes indicated that

ARIs can prevent pericyte loss, formation of microaneurysms, hemorrhages, and abnormal growth of endothelial cells in areas of pericytes loss [55, 56]. Thickening of basement membrane in the retinal capillaries was significantly inhibited by administration of ARI in animal diabetic model [57]. It was shown in human genetic studies that certain polymorphisms of the AR gene are associated with elevated tissue levels of AR and higher risk of diabetic complications [58]. However, sorbinil retinopathy clinical trial, where ARI sorbinil was administred for 2–3 years to adults with insulin-dependent diabetes, had no clinically important effect on DR [59]. Probably, it is important to develop more effective ARIs.

#### **4.2 Increased formation of advanced glycation end products (AGE)**

AGEs are built up at a permanent but slow rate starting at the embryonic development, accumulate through entire life, and linked with aging. However, AGEs formation is markedly accelerated in diabetes because of hyperglycemic environment [60]. AGEs are formed from the non-enzymatic reaction of sugars, such as glucose and fructose with free amino groups of proteins, lipids, and nucleic acids. The initial products of this reaction, such as Schiff bases, which spontaneously reform themselves into Amadori products, are reversible. Further reactions and molecular rearrangements result in the formation of irreversible crossed-linked derivatives termed AGEs, which are composed of a heterogeneous class of molecules that are yellow brown pigments, fluoresce. AGEs capable of forming cross-links with other structures and interact with cells via specific cell-surface AGE-binding receptors (RAGE), triggering inflammatory events, production of growth factors, generation of reactive oxygen intermediates induce oxidative stress [61, 62]. AGEs are toxic because they can modify intracellular proteins, including those involved in the regulation of gene transcription, or transfigure the extracellular matrix proteins, leading to reduction of the cell-to-cell interaction and vascular dysfunction, and also can modify circulating blood proteins. It demonstrated free radical generation by glycation products in vitro [63, 64]. The interaction of AGEs with RAGE has been involved in the development of DR. It was demonstrated that retinal endothelial cells, pericytes, and ganglion cells are expressed RAGE under normal or diabetic conditions in vitro and in vivo [61, 62, 65, 66]. Yatamagishi and colleagues demonstrated pericytes apoptosis mediated via AGEs-RAGE interactions. It was proposed that AGEs-RAGE interactions induced generation of intracellular ROSs, which course overexpression of proapoptotic Bax protein in pericytes [67]. Schmidt studies demonstrated that AGEs after interaction with their cellular receptors are responsible for induction of oxidative stress, activation of nuclear factor kappa-light-chain-enhancer (NF- *k*B) both in vitro and in vivo [61, 62]. NF-*k*B is associated with transcriptional activation of genes associated with inflammatory responses [68]. Interestingly, persistent hyperglycemia leads to a gradual accumulation of AGEs in the BM and in pericytes in diabetic animal models, however, the retinal endothelial cells did not store AGEs. It was suggested that endothelium is capable to uptake AGEs directly from the blood stream through RAGE located on their luminal surface and further transfer the AGEs to subendothelial matrix and to pericytes. The preferential appearance of intracellular AGE deposits within pericytes and BM may affect their functions and lead to progression of DR [65]. Another study by Yamagishi demonstrated that in vitro exposure of retinal pericytes to AGEs retarded pericytes growth and induced apoptosis; moreover, these effects were cell-specific [66]. It was found that administration of an inhibitor of AGEs to diabetic animals prevented accumulation of AGEs in the retinal capillaries and significantly diminished pericyte loss, subsequent formation of microaneurysms, acellular capillaries, and capillary closure [69]. Furthermore, it was shown in vitro that AGEs induced VEGF overproduction by retinal pericytes, that is, additionally disturbed retinal

microvascular homeostasis in concert with pericyte apoptosis [70]. Thereby, accumulation of AGEs significantly contributes to the development of diabetic retinopathy.
