Glycative stress as a cause of macular degeneration

Abstract

People are living longer and rates of age-related diseases such as age-related macular degeneration (AMD) are accelerating, placing enormous burdens on patients and health care systems. The quality of carbohydrate foods consumed by an individual impacts health. The glycemic index (GI) is a kinetic measure of the rate at which glucose arrives in the blood stream after consuming various carbohydrates. Consuming diets that favor slowly digested carbohydrates releases sugar into the bloodstream gradually after consuming a meal (low glycemic index). This is associated with reduced risk for major age-related diseases including AMD, cardiovascular disease, and diabetes. In comparison, consuming the same amounts of different carbohydrates in higher GI diets, releases glucose into the blood rapidly, causing glycative stress as well as accumulation of advanced glycation end products (AGEs). Such AGEs are cytotoxic by virtue of their forming abnormal proteins and protein aggregates, as well as inhibiting proteolytic and other protective pathways that might otherwise selectively recognize and remove toxic species. Using in vitro and animal models of glycative stress, we observed that consuming higher GI diets perturbs metabolism and the microbiome, resulting in a shift to more lipid-rich metabolomic profiles. Interactions between aging, diet, eye phenotypes and physiology were observed. A large body of laboratory animal and human clinical epidemiologic data indicates that consuming lower GI diets, or lower glycemia diets, is protective against features of early AMD (AMDf) in mice and AMD prevalence or AMD progression in humans. Drugs may be optimized to diminish the ravages of higher glycemic diets. Human trials are indicated to determine if AMD progression can be retarded using lower GI diets. Here we summarized the current knowledge regarding the pathological role of glycative stress in retinal dysfunction and how dietary strategies might diminish retinal disease.

Keywords: Aging; Autophagy; Glycation; Methylglyoxal; Microbiome; Proteostasis; Retina; Serotonin; Ubiquitin proteolytic system.

Fig. 1.. Consuming higher glycemic index diets increases glycation, compromises proteolytic editing, leads to dysbiosis, and is associated with loss of retinal integrity.

Left side: (A) Our working hypothesis is that consuming higher glycemia diets is associated with more glycative stress and accumulation of AGEs in retinal tissues (red line), decreased proteolytic capacity (blue line), dysbiosis, a more lipid-rich metabolome, and reversing this practice can slow progression of early/intermediate AMD (see text). Right side: Images of retina without (lower), and with (middle) drusen (arrow), and advanced AMD (showing bleeding), indicators of normal, early-intermediate AMD, and advanced AMD. Adapted from (Toomey et al., 2018) (original images from the Project MACULA website (http://projectmacula.org) courtesy of Christine A. Curcio Ph.D.). (B–E) Advanced AMD, (retro illumination, OCT, illustration and histologic), (F–I) Early AMD, images as in B-E. (J–M) Normal retina, images as in B-E. Green in panels D and H indicates basal deposits. Yellow/brown in panels D and H indicates drusen. Blue in panels L, H, D indicates infoldings. Yellow circles within the RPE indicate lipofuscin accumulation. Neovascularization, loss of photoreceptors and RPE degeneration is diagrammed in panel D. The data presented in Section 5 describe retinal features (AMDf) comparable to illustrations D, E, H, J, L, M.

Fig. 2.. AGEs form from glucose or its metabolites upon aging and/or stress and accumulate systemically.

(A) Formation of AGEs from glucose or its metabolites such as dicarbonyls (methylglyoxal, glyoxal, or 3-deoxyglucosone (3-DG)). Aging and oxidative stress promote the accumulation of AGEs throughout the body. AGEs are highlighted in orange. CML: Nε-(carboxymethyl)-lysine; CMA: Nε-(carboxymethyl)-arginine; 3-DG: 3-deoxyglucosone; 3-DGH: GH-1,2,3: Glyoxal-derived hydroimidazolone; MGH-1,2,3: Methylglyoxal-derived hydroimidazolone; CEL: Nε-(carboxyethyl)-lysine; CEA: Nε-(carboxyethyl)-arginine; MOLD: Methylglyoxal-derived lysine dimer. (B) Brunescence increases upon ageing in costal cartilage (reprinted with permission of Dr. Baynes) (Dyer et al., 1991). (C) Levels of Nε-(carboxymethyl)-lysine (CML) increase exponentially in lens crystallins from diabetic (□) and non-diabetic (◆) subjects as a function of age (From Tessier et al. (Tessier et al., 1999). (D) Normal lens from young donor (left) and brunescent cataractous lens from older donor (right). Also see Nandi et al. (Nandi et al., 2020).

Fig. 3.. AGE modification of substrates decreases proteolytic potential and susceptibility to proteolysis.

(A) MG-H1 in RPE after incubation with 1 mM MGO for 2 h. After MGO was removed, the cells were collected at indicated times, lysed and the lysate was probed by immunoblotting using anti-MG-H1 antibody. (B) Protein degradation in RPE after MGO exposure. Proteins in RPE were radiolabeled with ³H-Tyr for 60 h, and then, the cells were exposed to MGO for 2 h. After MGO was removed, the rate of protein degradation was determined based on TCA-soluble radioactivity. Right panel: The rate of degradation was calculated from the slope between 0 and 8 h. (C) Glycation of ¹²⁵I-labeled αA162 decreases its ubiquitin–proteasome system-dependent degradation. (D) Degradation of ¹²⁵I-labeled αA162 using Ubc4 modified by increasing concentrations of MGO. (E) Glycative stress–dependent attenuation of degradation in live cells. Hela cells constitutively expressing the ubiquitin-proteasome system (UPS) reporter substrate Ub-G76V-GFP were first incubated with 10 μM MG132 for 2 h to inhibit the UPS and allow for substrate accumulation. Then, they were treated with 10 μM MG132 and ±2 mM MGO for an additional 2 h to allow accumulation of the substrate. After MG132 and MGO were washed off, the cells were incubated with 100 μgmL⁻¹ of cycloheximide to inhibit new protein synthesis. The cells were observed by fluorescence microscopy. (F) AGE modification of ubiquitin–proteasome system (UPS) components reduces ubiquitination and protein degradation. Ubiquitin conjugates are generated by glycated ubiquitin. Ubiquitin was first glycated by MGO or glucose-6-phosphate (G6P). The MGO or G6P was removed, and the glycated forms of ubiquitin or unmodified ubiquitin were added to rabbit reticulocyte lysate. After 2 h incubation at 37 °C, samples were probed by immunoblotting using anti-ubiquitin antibody. (G) Degradation of ¹²⁵I-labeled α-A crystallin using unmodified ubiquitin, the conjugation competent but proteolytically incompetent K6W mutant ubiquitin, G6P- or MGO-modified ubiquitin, or in a reticulocyte lysate. Adapted from Uchiki et al. (Uchiki et al., 2012).

Fig. 4.. In vitro and in vivo AGE-modified proteins are degraded by UPS and autophagy pathway and AGE stress induces interaction between the UPS and autophagy.

(A) Retinal pigment epithelial cells (RPE) were incubated with 1 mM methylglyoxal (MGO) for 2 h. After MGO was removed, cells were incubated with DMEM containing no inhibitor, 5 μm epoxomicin (proteasome inhibitor), 50 μm chloroquine (that changes the lysosomal pH thus inhibiting autophagic degradation), or 10 mM 3-methyladenine (3-MA; inhibitor of autophagy via its inhibitory effect on class III PI3K) and harvested at 8, 14, and 20 h. Glycation was probed using an anti-MG-H1 antibody. Numbers below the GAPDH standard indicate the relative levels of MG-H1 at the times indicated to MG-H1 levels at 0 h, corrected for protein load. The arrow indicates AGEs that do not enter the running gel and accumulate over time. (B) AGEs and ubiquitin conjugates in whole RPE (left) and enriched lysosomal (right) preparations with and without exposure to glycative stress. RPE were exposed to 4 mM MGO for 2 h with or without subsequent incubation with proteasome (100 nM PS-341) or/and lysosome inhibitors (100 μM leupeptin). The cell lysate was separated by differential centrifugation to obtain a lysosomal fraction. MG-H1, ubiquitin, and lysosomal-associated membrane protein (LAMP) 1 were observed using appropriate antibodies on Western blots. 0 h represents samples from the cells harvested immediately after MGO exposure. The presence of LAMP1 indicates lysosomes. (C) Suppression of UPS degradation leads to accumulation of endogenous AGEs in aged rats, but not in young rats. Representative Western blots for MG-H1 in young (3–4 months old) and aged (24–26 months old) rat hippocampus after proteasome inhibition. Note the increased amount of MG-H1 in the aged group at 6 h and 14 h after lactacystin-injection (right). p62 and phospho-p62 are shown as autophagic markers and GAPDH as a loading control. Adapted from Uchiki et al. (Uchiki et al., 2012) and Aragones et al. (Aragonès et al., 2020a).

Fig. 5.. Glycative stress results in more substrates for- but limited activity of ubiquitin and autophagic proteolytic pathways-, resulting in accelerating cytotoxicity.

(1) When levels of damaged proteins are low, proteins remain in native conformations and are stable. (2) Glycative stress causes protein conformational change. (3) Altered soluble proteins are ubiquitinated and sent to the (4) proteasome for degradation to small peptides. (5) Upon stress, or when there is insufficient proteolytic capacity, aggregates of ubiquitinated proteins accumulate. These may include proteins in unfolded states, aggregated states or native states. These may include AGEs along with unmodified proteins, some including ubiquitin conjugates (box). Some may oligomerize and cross-link forming the higher mass aggregates. Undegraded conjugates may also accumulate if there is insufficient proteasomal, including deubiquitinating, activity. (6) p62 may recognize and present ubiquitinated species to autophagosomes, which fuse with lysosomes. (7) When autophagic capacity is sufficient, cargos are reduced to amino acids. (8) However, if autophagic or lysosomal activity is compromised, undigested materials may accumulate. (9) Aging and stress cause further accumulation of cargos. (10) When proteolytic capacity is insufficient or substrates become resistant to lysosomal proteases, cargos accumulate and lysosomes enlarge. The ubiquitin–proteasome system and the lysosomal proteolytic system can degrade the damaged proteins and toxicity is averted (top). Such is the case in many tissues during youth. Steps 5 and 10 are associated with aging and glycative stress. Under chronic glycative stress, glycated proteins accumulate. Accumulated oligomerized altered proteins may impair the proteolytic machinery, setting up a vicious cycle of stress, limited proteolytic editing, and further damage to the proteome, resulting in some of the disease-related accumulation of AGEs and conjugates observed in vivo. The accumulated aggregated proteins are brunescent.

Fig. 6.. Scheme of the glycation-derived cellular stress responses.

(1) Stressors: AGEs bind to diverse extracellular molecules, inhibiting their normal functions; (2) Detection of glycative stress: receptors in the plasma membrane, such as RAGE, play a crucial role in sensing AGEs, initiating intracellular cascades and signaling pathways; (3) Stress transducers: hyperglycemia impacts on vital intracellular pathways including the polyol pathway, PKC (Protein Kinase C), AGEs pathway, and the hexosamine pathway leading to dysmetabolism and increasing oxidative stress; (4) Stress responses: AGEs derived-pathological consequences include mitochondrial dysfunction and translocation of transcription factors into the nucleus, leading to development of vascular abnormalities, inflammation, apoptosis and necrosis.

Fig. 7.. Glycemic Index is defined as the incremental area under the blood glucose response curve

(AUC) within a 2-h period elicited by a portion of food containing 50 g of available CHO, relative to the AUC elicited by 50 g glucose. Thus, the GI is a kinetic parameter that reflects the potency of foods to raise blood glucose and rates of glucose clearance. The GI of a particular diet is determined by averaging the GI values of the food items, statistically weighted by their carbohydrate contribution. Diets with the same amount of CHO but that have different GI are to be distinguished from diets that have higher or lower amounts of CHO but for which GI may be the same. The use of the GI is predicated on the GI being a property of the food, not a property of the subject in whom it was measured. GI values obtained for the same foods are roughly similar in an ethnically and physiologically wide variety of subjects. Issues regarding the use of GI are reviewed in (Chiu et al., 2011). Adapted from www.GIsymbol.com, with permission from Dr. J. Brand-Miller.

Fig. 8.. Consuming higher GI diets is associated with higher levels of AGEs

(A) in retina, liver, lens and brain (Adapted from (Uchiki et al., 2012)). (B) The HG diet leads to increased glucosepane in plasma. Neural retina (C) and RPE (D) have higher levels of methylglyoxal-derived hydroimidazolone 1/3 (MG-H1/H3) and CEL. Levels of these AGEs were indistinguishable in HGxoLG and LG neural retina and RPE.

Fig. 9.. Accumulation of AGEs in the plasma and RPE of Nrf2-null HG mice.

(A) Plasma levels of MG-H1 are increased in the plasma of 18-months-old Nrf2-null mice relative to 24-months-old wild-type mice. (B–C) Plasma pentosidine and fructosyl-lysine are increased in the plasma of Nrf2 null mice fed HG. Adapted from (Rowan et al., 2020).

Fig. 10.. Consuming higher glycemic index diets is associated with a higher prevalence of AMD related phenotypes and multiple biomarkers of AMDf.

(A) Twenty-four-month-old wild-type mice fed HG exhibited significant photoreceptor outer nuclear layer (ONL) thinning relative to mice fed LG diet, especially near the optic nerve head and on the superior hemisphere of the retina. (B) The overall retinal damage score was created by summing the area under the ONL thickness curve and scaled so that the greater the amount of photoreceptor cell loss [lower area under the curve (AUC)], the greater the retinal damage score. (C) Retina damage score change over 24 months in mice fed HG, HGxoLG, and LG diets. The ONL thickness in mice fed HGxoLG animals did not differ from that of LG-fed animals, and there was no further photoreceptor cell loss in the HGxoLG group after their change to the LG diet. However, once retinal damage started, at around age 18 months, the rate of accretion was similar to that observed in the mice fed HG diet. (D–I) Toluidine blue stained sections through the retinas of LG and HGxoLG have normal architecture, whereas HG mouse retinas show the indicated lesions. Square brackets indicate regions with a missing RPE monolayer or hypopigmentation as indicated (F, H, I). Asterisk indicates regions of RPE multilayering (F, H). The arrowhead in (G) points to a subretinal deposit. (J) HG diet induces AMD-related ultrastructural changes and lipofuscin accumulation in the RPE in twenty-four-month-old mice. Electron micrographs show the loss of basal infoldings, vacuoles, formation of large basal laminar deposits along with membranous debris, as well as accumulation of lipofuscin, phagosomes, and lipid deposits in HG RPE. (K) Phagosomes accumulate in RPE from mice fed HG, but not in RPE of mice fed HGxoLG or LG. (L) The number of lipofuscin granules is linearly related to the extent of retinal degeneration. (M, N, O) HG diet leads to increased AGEs and lipids but decreased serotonin in the plasma. Retina damage score is linearly related to plasma lipofuscin, glucosepane, C22:6 lysophosphatidylethanolamine (LPE) and serotonin, positively for the first three and oppositely for serotonin. (P) AMDf are observed earlier in life in Nrf2-null mice (18 months rather than 24 months in wild-type mice), and they are more frequent in HG mice. Abbreviations: bi-basal infoldings, Bm-Bruch’s membrane, m-mitochondria, lfn-lipofuscin granules, v-vacuoles, bld-basal laminar deposits, md-membranous debris, Ph-phagosome, Pl-phagolysosome, ld-lipid droplets.

Fig. 11.. Correlation network of retina features, metabolome, microbiome.

Red plasma, Blue urine, Green microbiota OTU. Circles indicate no diet associations. Triangles denote interaction with diet, Square means an additive effect. Solid edges indicate positive correlations. Dashes indicate negative associations. Adapted from (Rowan et al., 2017).

Fig. 12.. Drugs that regulate glucose limit age-related macular degeneration features.

(A) ONL thickness is highest in LG, lowest in HG. Acarbose diminishes damage due to HG, but Empagliflozin and Fenofibrate do not. (B) RPE pigment is retained in LG and Acarbose- HG mice. (C) Acarbose and empagliflozin diminish vacuolation in RPE in HG mice. Fenofibrate does not.

Fig. 13.. Risk of developing AMD, or AMD progression, due to consumption of high GI diet.

Odds or risk ratios are shown as red dots and lines represent the upper and lower confidence intervals on a logarithmic scale. Blue circle indicates a study that monitored progression of AMD.

Fig. 14.. Odds ratios (95% confidence intervals) of AMD according to quintile groups of two major dietary pattern scores.

Abbreviation: OR, odds ratio; CI, confidence interval. Western diets, as consumed in the USA tend to be HG diets. Oriental diets tend to be LG. Adapted from Chiu et al. (2014).