Introduction

It has been proposed that glycative modification of proteins causes changes at the cellular and tissue levels, which contribute to a variety of age-related pathologies. Glycation was first described by Louis Camille Maillard, in the beginning of the twentieth century, as the browning reaction between reactive monosaccharides and amino acids during food processing. This process was later identified as a pathway leading to the formation of advanced glycation end products (AGEs) (Henle 2005). Since then, many studies have demonstrated that Maillard browning contributes to the color, taste, and aroma of foods; furthermore, the products from this reaction also affect the nutritional and toxicological properties of food (Vlassara and Uribarri 2004). The importance of Maillard products and AGEs was not recognized by the food industry and medical community until a quarter of century later (Finot 2005; Bengmark 2007).

AGEs are a byproduct of the non-enzymatic reaction of reducing sugars with proteins, lipids, and nucleic acids. This reaction alters their function by disrupting molecular conformation, promoting cross-linking, altering enzyme activity, reducing their clearance, and impairing receptor recognition (Stirban et al. 2013). To date, AGEs have been identified as a strong contributor to diabetic complications (Mullarkey et al. 1990; van Boekel 1991; Koschinsky et al. 1997) and neurodegenerative diseases such as Alzheimer’s disease (AD) (Sasaki et al. 1998; Bengmark 2007). Unsurprisingly, hyperglycemia, a condition often associated with heightened AGE levels, is a risk factor for cognitive impairment and dementia (Biessels et al. 2006; Brands et al. 2007; Crane et al. 2013). Diets rich in fruits and vegetables are often recommended to reduce the risk of non-communicable diseases and have a positive effect on blood sugar (Sargeant et al. 2001; Slavin and Lloyd 2012; Pem and Jeewon 2015). Consumption of berry fruits has been shown to decrease inflammation (Zafra-Stone et al. 2007), improve cardiovascular health (Sayegh et al. 2014; Wightman and Heuberger 2015), protect neurons against high-energy and charged particles (Shukitt-Hale et al. 2013; Poulose et al. 2014), and improve age-related declines in cognition (Willis et al. 2009; Krikorian et al. 2010; Malin et al. 2011). The antioxidant and anti-inflammatory properties of the bioactive phytochemicals found in fruits and vegetables are believed to contribute to the observed health benefits. Interestingly, recent studies suggest that many of these phenolic phytochemicals exhibit anti-glycative activity in addition to their antioxidant activity, and these two activities may have synergistic health benefits. This review describes AGE formation and the potential role of dietary bioactives in combating their deleterious effects.

Metabolic Sources of AGEs

The formation of AGEs can occur both in vivo and ex vivo in the presence of sugars and protein. Endogenous AGEs derive from two major pathways, non-enzymatic glycosylation and auto-oxidative glycosylation. Non-enzymatic glycosylation, also known as the Maillard reaction, glycation, and the “classic” or Hodge pathway, is initiated by the addition of sugar’s carbonyl group to the amino group of a protein to form Schiff’s base intermediates, which then undergo further rearrangement into ketoamine Amadori products. This reaction can be divided into early (reversible) and late (irreversible) stages. In the early stage, stable Amadori products are formed from the rearrangement of glycosylamines; reversibility of the products is dependent on the concentration of the compound and incubation time. Amadori products then undergo dehydration, condensation, fragmentation, oxidation, and cyclization; these processes result in the formation of AGEs, a heterogeneous group of side-chained amino acids (Fig. 1) that are characterized by their structures, the ability to fluoresce, and the ability to initiate cross-linking (Lin 2006).

Fig. 1
figure 1

Pathways for the formation of advanced glycation end products (AGEs). In the early stages of reaction, monosaccharides, such as glucose, fructose, and galactose, interact non-enzymatically with peptides, proteins, and nucleic acid via Schiff’s base, to form a relatively stable Amadori product. The reactions leading up to Amadori formation are reversible depending on the concentration and incubation time. The late stage reaction, an irreversible reaction, involves dehydration, hydrolysis, condensation, fragmentation, and oxidation of Amadori product to form AGEs. Glycolytic intermediates and lipid peroxidation can contribute to the formation of AGEs in vivo

Full size image

Alternatively, the oxidation of monosaccharides catalyzed by transition metals, such as copper and iron, generates dicarbonyl products that can react with proteins and lead to the formation of AGEs (Wells-Knecht et al. 1995). Therefore, this pathway is termed auto-oxidative glycosylation. The rate of reaction for both of the pathways is affected by various factors, such as temperature, pH, carbonyl/amine ratio, and time of exposure (Poulsen et al. 2013). AGE formation can occur under anaerobic conditions; however, the presence of reactive oxygen species (ROS) significantly increases the rate of formation (Baynes and Thorpe 2000). Some of the endogenous AGEs that have been well researched include N epsilon-(carboxymethyl) lysine (CML), pentosidine, pyrraline, imidazolone, and crossline (Brownlee 1995; Kikuchi et al. 2003).

Dietary Sources of AGEs

In addition to endogenous sources, AGEs are also present in our daily diet. Foods comprised of animal products contain AGEs; however, cooking accelerates the formation of AGEs in food, especially food rich in fats and proteins (Goldberg et al. 2004). Furthermore, modern cooking practices such as grilling, broiling, roasting, searing, and frying substantially accelerate the levels of AGEs (Goldberg et al. 2004). Relative dietary AGE (dAGE) content per serving in some commonly consumed foods is summarized in Table 1. It is evident that the foods which contain high amounts of fats or oils are prone to produce excessive AGEs upon processing or cooking. Fruits, whole grains, and milk are generally low in dAGEs. The harmful effects of dAGEs were largely ignored because AGEs are resistant to enzymatic digestion, and the oral bioavailability of AGEs absorbed from the gastrointestinal tract is low—approximately 10 % of total AGEs present in food (Koschinsky et al. 1997; He et al. 1999). However, recent studies have established that supplementation of AGE-rich diets in humans and animals has significantly contributed to heightened levels of endogenous AGEs (Koschinsky et al. 1997; He et al. 1999).

Table 1 Dietary AGE content in commonly consumed foods, based on carboxymethyl lysine content and assessed using ELISA technique (Uribarri et al. 2010)
Full size table

A comprehensive list of dAGEs and guiding food choices to reduce dAGE intake are provided in the new database for dAGEs (Uribarri et al. 2010).

Dicarbonyl intermediates of the Maillard reaction such as glyoxal and methylglyoxal (MG), as well as Amadori compound CML, play a role in aging and disease pathologies (Frye et al. 1998; Bair et al. 2010). However, exogenous MG and CML contents have been shown to vary widely depending upon the type of cooking practice followed (Uribarri et al. 2010). For example, milk contains nearly negligible amounts of MG and CML, while butter and cheese have more than 1000-fold higher amounts of these dicarbonyl byproducts. Similarly, there is a 100-fold increase in levels of MG in pan-fried steak, compared to that of raw meat. Acidic (low pH) food constituents, such as vitamin C or vinegar marinade, has been shown to reduce the formation of AGEs by arresting the glycation process (Uribarri et al. 2010). Besides food, it has been hypothesized that AGEs are formed during the curing and aging of tobacco (Nicholl and Bucala 1998). A study by Cerami et al. (1997) demonstrated that both the aqueous extracts of tobacco and cigarette smoke contain highly reactive AGEs that readily cross cell membranes and, in the form of cigarette smoke, are absorbed through the lungs, leading to rapid induction of AGE-modified protein in vivo.

Cellular Effects of AGEs

Protein glycation promotes cellular dysfunction through multiple mechanisms. Glycation causes structural changes to protein, which can inhibit specific functions, such as those of the cytoskeleton and antioxidant enzymes (McLean et al. 1992; Ledesma et al. 1994; Yan and Harding 1997). Furthermore, changes in protein structure can also result in cross-linkage, aggregation, and precipitation of proteins (Eble et al. 1983; McPherson et al. 1988; Vlassara and Uribarri 2004; Bengmark 2006). For example, arteriosclerosis is a disease where the cytotoxicity of glycation involved in its pathogenesis is well understood. In arteriosclerosis, cross-linkage of proteins increases tissue remodeling, thickening and stiffening of the basement membranes, as well as serving as a reactive site where circulating serum proteins, such as lipoproteins, are covalently trapped, leading to decreased vessel dilation capability (Cerami et al. 1997; Singh et al. 2001). Moreover, AGE modification of ApoB prevents normal uptake, thus reducing cholesterol efflux (reviewed by Singh et al. 2001). Recent evidence has also suggested that protein glycation increases oxidative damage directly (Lee et al. 1998; Yim et al. 2001) and indirectly accelerates oxidative damage (Qian et al. 1998; Sobal et al. 2000).

Accumulation of AGEs, such as CML, has been reported in the mitochondrial matrix of aged animals (Bakala et al. 2003). It has been reported that glycation of renal mitochondrial proteins induces a significant increase in superoxide production (Rosca et al. 2005). Methylglyoxal (an AGE intermediate) has been shown to suppress the tricarboxylic acid cycle (TCA cycle) and electron respiratory chain (Ray et al. 1994; Rosca et al. 2002, 2005). Furthermore, a recent study by Speer et al. (2003) has suggested that MG inhibits mitochondrial permeability transition by covalently modifying the mitochondrial permeability transition pore. Thus, it is apparent that the glycation process is a potent initiator of mitochondrial dysfunction.

Furthermore, ex vivo-derived AGEs can induce cellular alterations once absorbed into cells. For example, tobacco-derived AGEs have been shown to accumulate in the lens and coronary artery vascular walls (Nicholl and Bucala 1998). Food-derived AGEs have been reported to induce ROS, suppress antioxidant reserves (e.g., GSH/GSSG ratio), and induce cytokine production in cultured endothelial cells (Cai et al. 2002). Similarly, diabetic animals fed an AGE-rich diet developed diabetic vascular or renal pathology, whereas their age-matched controls on an AGE-poor diet remained nearly free of such pathology (Zheng et al. 2002; Lin et al. 2003). These findings suggest that food-derived AGE precursors can induce significant modification of macromolecules, causing them to be toxic to target cells.

The interaction between AGEs and their receptors has been reported to activate various signaling pathways that can: (A) significantly influence cellular functions (Heidland et al. 2001; Ott et al. 2014), (B) enhance oxidative and nitrosative stress (Cai et al. 2006, 2007; Ott et al. 2014), (C) induce cellular proliferation (Huang et al. 1999; Guh et al. 2001), (D) upregulate expression of genes involved in tissue remodeling, regeneration, and inflammatory response (reviewed by Singh et al. 2001), (E) increase production of reactive oxygen species (ROS) in macrophages, microglia, and endothelial cells (Schmidt et al. 1994; Wautier et al. 1994; Yan et al. 1994; Lander et al. 1997; Basta et al. 2005), and (F) induce apoptosis (Nitti et al. 2005). Furthermore, chronic overproduction of AGEs leads to chronic inflammatory responses and abnormal growth of certain cells (reviewed by Singh et al. 2001).

AGEs and Cognitive Function

Glucose is the brain’s primary source of energy. Over time, glucose undergoes a cascade of non-enzymatic reactions with proteins, lipids, and nucleic acids and ultimately forms various types of AGEs, which accumulate in brain regions associated with learning and memory (i.e., cerebral cortex and hippocampus) (Thangthaeng et al. 2008). A hyperglycemic environment, as seen in diabetes mellitus (DM), facilitates the formation of AGEs, thus increasing the risk of cognitive impairment and dementia (Biessels et al. 2006). A cross-sectional study by Brands and colleagues comparing cognitive performance between subjects with type 2 diabetes mellitus (T2DM) to age-matched controls showed that those with T2DM performed significantly, but modestly, worse than controls on cognitive tasks, particularly on tasks that required high levels of mental efficiency (Brands et al. 2007). Moreover, longitudinal studies of subjects with T2DM showed an accelerated cognitive decline as compared to age-matched subjects without T2DM (Knopman et al. 2001; Yaffe et al. 2012). Interestingly, high peripheral AGE levels were associated with greater cognitive decline regardless of DM (Yaffe et al. 2011).

In addition to endogenous AGEs, dietary AGEs (dAGEs) have been identified as a potentially modifiable cause of cognitive decline. Cai et al. (2014) demonstrated that serum methylglyoxal (sMG) was positively correlated with dAGEs, and higher sMG levels predicted accelerated cognitive decline in older adults (Cai et al. 2014; Perrone and Grant 2015). Similarly, long-term supplementation with MG impaired psychomotor function and object recognition performance in C57/BL6 mice when compared with age-matched controls (Cai et al. 2014). Moreover, recent studies by Perrone and Grant examining the relationship between estimated dAGEs and the incidence of Alzheimer’s disease (AD) reported a significant positive correlation between dAGEs and AD incidence, suggesting dietary AGEs as a risk factor for AD (Perrone and Grant 2015).

Taken together, elevated levels of endogenous AGEs or dAGEs may trigger cognitive dysfunction. For example, the glycated form of amyloid-β induced a higher degree of cognitive impairment, dendritic spine deterioration, suppression of long-term potentiation (LTP), and down-regulation of synaptic proteins when compared with its non-glycated form (Chen et al. 2014). This may be because the glycated form is a better ligand for the receptor for AGEs (RAGEs) than the non-glycated form, leading to greater neuronal death (Li et al. 2013). Not surprisingly, when an accelerated aging mouse model (SAMP 8) was treated with anti-glycation compound J147 for 7 months, these mice were found to have preserved cognitive function and restored molecular markers of synaptic function, vascular pathology, and inflammation at 10 months of age when compared with age-matched controls (Currais et al. 2015).

Anti-glycative Effects of Berries

Dietary berry fruits have been associated with healthful benefits, which are partially attributed to the antioxidant properties of the phytochemicals they contain (for further review see Miller and Shukitt-Hale 2012; Shukitt-Hale et al. 2015; Skrovankova et al. 2015). An extract of Vaccinium vitis-idaea L. (V. vitis-idaea L., commonly known as mountain cranberry or lingonberry), an ingredient of Cree traditional herbal medicine used to treat diabetic symptoms, was shown to inhibit the formation of fluorescent AGE species and CML adducts of bovine serum albumin (CML-BSA) in vitro (Beaulieu et al. 2010). Later, other berry fruit extracts were shown to exhibit anti-glycative activity with black bearberry, crowberry, common juniper, and bog cranberry being among the most potent (Harris et al. 2014). Using in vitro systems, Liu et al. (2011) also demonstrated that cranberry phytochemicals display anti-glycative activity, especially fractions enriched with procyanidins.

Commonly consumed berries such as cranberries, strawberries, and blueberries also display anti-glycative activity. In an in vitro study conducted by Parengkuan et al. (2013), glucose incubated with either human serum albumin (HSA) or bovine skin collagen type I (BSC I) was used to test the anti-glycative activity of various fruit extracts. Both strawberry (IC50 HSA = 0.288 mg/ml, IC50 BSCI = 0.607 mg/ml) and blueberry (IC50 HSA = 0.293 mg/ml, IC50 BSCI = 0.207 mg/ml) extracts significantly inhibited glycation of albumin and collagen. The potency of strawberry and blueberry extracts was slightly lower but comparable to the positive control, aminoguanidine (IC50 HSA = 0.0603 mg/ml, IC50 BSCI = 0.232 mg/ml). The authors proposed that the bioactive phytochemicals in the berries were responsible for the anti-glycative activity observed.

Recent evidence suggests that the antioxidant activities of berry phytochemicals may mediate anti-glycative activities. For example, phenolic acids, such as ferulic acid, resveratrol, chlorogenic acid, caffeic acid, cinnamic acid and its derivatives, as well as ellagic acid have been reported to inhibit AGE formation in vitro through inhibition of the advanced stage of the glycation reaction, which involves an oxidation/reduction step (Sadowska-Bartosz et al. 2014; Sadowska-Bartosz and Bartosz 2015). However, a structural analysis study by Matsuda et al. (2003) indicated that not all antioxidants were able to prevent AGE formation. Flavonoids with strong superoxide anion radical scavenging activity more potently inhibit AGE formation as compared to those with weak superoxide anion radical scavenging activity. Furthermore, a study by Harris et al. (2011) demonstrated that the total phenolic content and free radical scavenging activity of ethanolic extracts from 17 medicinal plants were positively associated with anti-glycative activity. Wei and colleagues have also shown that phenolic compounds removed from their sugar moieties, found in blueberries, blackberries, strawberries, raspberries, cranberries, and Noble muscadine grapes, reduce the production of AGEs by scavenging bicarbonyl products, such as MG formation in vitro (Wang et al. 2011). In berries, procyanidins and catechins are constituents known to quench glyoxal and MG intermediaries (Wang et al. 2011).

Studies examining the effects of polyphenol treatment in diabetic rodent models have reported some positive effects. A study by Ciddi and colleagues treated streptozotocin (STZ)-induced diabetic rats with resveratrol (10 mg/kg) orally for 3 weeks and reported that resveratrol treatment significantly decreased AGE formation in the kidneys relative to non-treated controls (Ciddi and Dodda 2014). Like resveratrol, fisetin, a flavonol commonly found in strawberry, (~40 mg/kg/day) reduced MG-modified proteins in the cerebral cortex, blood, and kidneys of Akita mice, a type I DM mouse model, when fed for 18 weeks. Interestingly, normal C57/BL6 mice, fed with fisetin (~25 mg/kg/day) from 6 weeks of age to 24 weeks of age, also had significantly less MG-modified proteins in their cerebral cortices than age-matched controls (Maher et al. 2011). These results suggest that the anti-glycative activity of fisetin may have beneficial effects in the aging brain.

Berries are also rich in vitamin C and numerous phenolic acids that are known to inhibit the glycation of carbohydrates and proteins, thereby reducing AGE formation (Uribarri et al. 2010). Berry phytochemicals are also known to interact with numerous signaling molecules in the brain, which reduce oxidative burden and improve neuronal signaling (Poulose et al. 2012). Recent evidence also indicates that AGEs interact with amyloid-β plaques and ubiquitinated inclusion bodies, thereby increasing neurotoxicity and increasing toxic protein aggregation in the brains of AD patients (Ko et al. 2015). However, dietary berry fruits have been shown to reduce the amyloid-β burden and enhance autophagic protein homeostasis, leading to improved cognitive function, in rodent models of aging (Poulose et al. 2014, 2015; Shukitt-Hale et al. 2015).

Conclusions

Mounting evidence indicates that AGEs may be a key contributor to pathophysiology and age-related impairment; thus, there is a need to develop potent and effective AGE inhibitors. Clinical trials studying synthetic AGE inhibitors have yielded inconsistent and unconvincing evidence of efficacy and have raised concern regarding the possible toxicity of synthetic AGE inhibitor molecules. Berry fruits phytochemicals are a potential potent natural AGE inhibitor. In vitro studies have shown that many berry fruits and their bioactive phytochemicals display a robust ability to inhibit AGE formation through their antioxidant properties. Additionally, berry fruits phytochemicals may protect AGE-induced neurotoxicity by reducing inflammation signaling, as well as enhancing AGEs degradation through autophagy. As illustrated in Fig. 2, combination of these effects is hypothesized to produce an improvement in cognitive function. Interestingly, to date, no clinical studies have examined the effect of dietary berry fruit on AGE-related mechanisms of neurodegeneration; however, preclinical data strongly suggest an association of berry fruit supplementation with cognitive improvements and prevention of age-associated neurodegeneration. Therefore, natural products with potent anti-glycative activity present a promising approach to reduce the endogenous and exogenous burden of AGE at the cellular and neuronal level, thereby contributing to the possible cognitive improvements.

Fig. 2
figure 2

Schematic representation of the possible mechanisms of berries in reducing AGEs and improving cognitive function

Full size image