Inhibitors of Advanced Glycation End Product (AGE) Formation and Accumulation
What Are AGEs?
Advanced glycation end products are formed as a result of non-enzymatic biochem- ical reactions called the Maillard reaction which includes the reaction of glucose with amino residues on proteins and lipids (Lutgers et al. 2006; Mulder et al. 2006) (Fig. 1). Early glycation products such as methylglyoxal can be transformed into more stable AGEs, which may be irreversibly cross-linked with proteins or DNA and may subsequently alter the organ function and structure. AGEs are a heterogeneous and complex group of modifications (Maillard 1912).
In type 2 diabetic patients, ischemic heart disease and hypertension correlate with circulating AGE levels, suggesting that they may be potential biomarkers of diabetic cardiovascular risk (Sugiyama et al. 1998). It has been demonstrated that serum AGE levels are not completely associated with glycaemic control, as assessed by HbA1c in the clinical setting (Monnier et al. 1999; Steffes et al. 2003). This may explain the progression of diabetic complications in some patients with relatively good glycaemic control.
The Diabetes Control and Complications Trial (DCCT) study demonstrated that AGE levels were a better predictor of progression to complications than HbA1C. Differences in AGEs accounted for over a third of the variance (Monnier et al. 1999). Thus other factors, such as ROS, may contribute to AGEs production and accumulation in patients with adequate glycaemic control (Baynes and Thorpe 1999; Babaei-Jadidi et al. 2003).
It is yet to be determined which AGE modifications are the most pathogenic in disease. AGE cross-linked moieties, such as pentosidine, have intrinsic fluorescence. Tissue and plasma fluorescence can then be used as indicator AGE modifications with increased fluorescence within the kidney (Soulis-Liparota et al. 1995; Watson et al. 2011), retina (Stitt et al. 2002), skin (Lutgers et al. 2006; Genuth et al. 2005) as well as other organs affected by diabetic microvascular disease (Soulis et al. 1997a). This has been shown to increase with diabetes progression.
Changes in renal and hepatic function are also linked to increases in tissue fluorescence, reflecting the role these organs play in clearing AGEs from the body (Makita et al. 1994). Furthermore, in type 1 and type 2 diabetic patients, circulating levels of fluorescent AGEs correlate with complications (Miura et al. 2003; Kalousova et al. 2006).
Other AGEs, such as N-carboxymethyllysine (CML), are not cross-linked and do not fluoresce however they have are elevated in the serum of type 1 diabetic patients and in diabetic rodent models (Watson et al. 2011; Makita et al. 1991).
Type 2 diabetic patients also demonstrated increased levels in circulating CML (Wautier et al. 2003) and the AGE precursor dicarbonyl methylglyoxal (Kilhovd et al. 2003). Elevations in CML levels have been associated with the presence of microvascular complications, including retinopathy and nephropathy (Beisswenger et al. 1995).
Measurement of AGEs
Several AGEs and their intermediates have been identified and can be measured in serum and urine including CML, pentosidine and the dicarbonyl precursor, methylglyoxal. Circulating levels of AGEs have been measured in diabetic patients as well as in experimental models of diabetes. Our group has demonstrated that the measurement of total AGEs within the serum of diabetic patients does not necessar- ily correlate with progressive diabetic renal dysfunction (Coughlan and Forbes 2011).
By contrast, we have found that circulating high-molecular-weight AGEs in type 2 diabetic patients correlate with the decline in renal dysfunction (Penfold et al. 2010). Furthermore, Coughlan et al. demonstrated that urinary levels of AGEs are strongly associated with progressive renal decline in both type 1 and type 2 diabetics and may be used as a biomarker (Coughlan and Forbes 2011). This has been verified by others who have shown that circulating AGEs are associated with progression of other diabetic complications (Makita et al. 1992).
Accumulation of AGEs in the Skin
The cross-linking of AGEs in the skin is associated with the generation of autofluorescence which may be measured in skin by a non-invasive AGE reader. The accumulation of low-molecular AGEs within the skin was shown to be higher in patients with diabetes correlating with glomerular filtration rate (GFR); those with a higher GFR were found to have a lower fluorescence (Thomas et al. 2005a). Hartog et al. found that diabetic patients on dialysis exhibited higher levels of skin autofluorescence when compared to a control group (Hartog et al. 2005).
More recently it has been demonstrated that skin autofluorescence in a healthy population is directly associated with age, smoking, waist circumference and diet (Kellow et al. 2018). In disease states associated with increased ROS production, there is a significant increase in skin AGE autofluorescence.
AGE Receptors
In addition to the direct deleterious effects of AGE accumulation in tissues, there are receptor-mediated effects. Vascular, renal, neuronal and haematopoietic cells are all known to express receptors for AGEs (Goldin et al. 2006).
The receptors for AGE are important modulators of the deleterious effects of these compounds. Receptors for AGEs are generally identified as either inflammatory (the receptor for AGEs (RAGE), AGE-R2) or clearance type receptors (AGE-R1, AGE-R3, CD36, Scr-II, FEEL-1 and FEEL-2) (Sourris and Forbes 2009; Alikhani et al. 2005; Schrijvers et al. 2004; Singh et al. 2001; Vlassara 1997; Vlassara and Bucala 1996).
Receptor for Advanced-Glycation End Products (RAGE)
The receptor for AGEs (RAGE) is a member of the immunoglobulin superfamily, expressed on the surface of monocytes, proximal tubular cells (Morcos et al. 2002), neurons, macrophages, and glomerular epithelial cells (podocytes) (Wendt et al. 2003a), mesangial, endothelial, smooth muscle, and fibroblast cells (Schmidt et al. 1994a; Wautier and Guillausseau 2001; Bierhaus et al. 2005a; Koulis et al. 2015). RAGE is a multiligand receptor which is capable of binding to a number of ligands other than AGEs including amyloid A, s100A8-9, amyloid-β-peptides, calgranulins, and amphoterin (HMGB1) (Yan et al. 1997; Bierhaus et al. 2005a; Yan et al. 2006).
Its major physiological role is thought to be in host–pathogen defence (Bierhaus et al. 2005a). The RAGE gene is located on chromosome 6 adjacent to the HLA locus in both human and mouse (Wautier and Guillausseau 2001; Bierhaus et al. 2005a) and its transcription is known to be both constitutive and inducible.
RAGE is expressed during embryogenesis whilst it is generally down-regulated in adult life in most tissues (Bierhaus et al. 2005a). In chronic diseases such as Alzheimers, ageing, and diabetes, RAGE is known to be elevated (Gao et al. 2008; Son et al. 2017).
The RAGE protein consists of three immunoglobulin-like regions, one v-domain and two c-domains, in addition to transmembrane and cytoplasmic regions (Neeper et al. 1992; Schmidt et al. 1994b). There are a number of isoforms of RAGE, which lack either the cytoplasmic or extracellular domains. These include soluble RAGE (sRAGE), thought to be the result of proteolytic shedding of RAGE from the cell surface (Humpert et al. 2007).
Soluble RAGE binds AGEs with a high affinity and has been considered as a decoy receptor for AGEs (Schlueter et al. 2003). Endothe- lial cells are known to secrete an isoform of RAGE (es-RAGE), which is a c-terminal splice variant of RAGE and lacks a trans-membrane and effector domain. Finally, NT-RAGE lacks an amino terminus; however, its function is still unclear (Yonemura and Tsukita 1999; Bierhaus et al. 2005a, b; Bohlender et al. 2005). Diabetic mice genetically manipulated to over-express RAGE have significant glomerulosclerosis (Yamamoto et al. 2001; Inagi et al. 2006). By contrast, it has been shown that RAGE knockout (KO) mice have less vascular and renal injury with diabetes (Myint et al. 2006; Soro-Paavonen et al. 2008; Coughlan et al. 2009; Sourris et al. 2010; Watson et al. 2012; Koulis et al. 2014). sRAGE treatment significantly attenuated diabetes- associated atherosclerosis development in animal models (Bucciarelli et al. 2002).
AGE-Lowering Therapies
Direct Targeting of AGEs
A number of AGE-lowering compounds have been developed including aminoguanidine, and AGE formation inhibitor had progressed to clinical trials but had to be retracted due to severe side effects. ALA, a putative cross-link inhibitor did not progress into further clinical validation due to closure of the company.
Inhibitors of AGE formation, including aminoguanidine (Soulis-Liparota et al. 1991) and OPB-9195 (Miyata et al. 2000a) have been shown to reduce AGE accumulation by scavenging free reactive carbonyl groups (Brownlee et al. 1986; Miyata et al. 2000b; Khalifah et al. 2005; Booth et al. 1997).
The anti-hyperglycaemic agent metformin can also trap reactive carbonyl groups (Beisswenger et al. 1999). Aspirin also has also been shown to scavenging free carbonyls groups, as well as decreasing AGE levels by targeting preformed intermediates via chelation of copper and other transition metals which can contribute to ROS production (Urios et al. 2007).
LR-90 (methylene bis [4,40-(2chloropheylureido phenoxysiobutyric acid)] is a compound proven to be effective in reducing renal and circulating AGE accumula- tion (Figarola et al. 2003, 2008). It attenuates AGE accumulation via its potent metal chelating abilities and its interaction with reactive carbonyl species (Figarola et al. 2003).
LR-90 has shown renoprotective benefits in experimental models of diabetes and attenuated both glomerulosclerosis and albuminuria (Figarola et al. 2003, 2008). In addition, we have demonstrated that LR-90 attenuated diabetes-associated ath- erosclerosis in experimental models of diabetes (Watson et al. 2010).
Vitamin B derivatives such as benfotiamine, thiamine and pyridoxamine have been trialled and have shown initial beneficial AGE-lowering effects in the clinical context (Table 1). Pyridoxamine prevents the formation of AGEs from Amadori intermediates (Khalifah et al. 2005; Booth et al. 1997) and cleaves 3-deoxyglucosone reactive carbonyl intermediates (Chetyrkin et al. 2008).
The inhibitory actions of pyridoxamine on AGE accumulation are associated with improvements in renal function in experimental models (Degenhardt et al. 2002) and a decrease in diabetes-associated atherosclerosis (Watson et al. 2011). In a phase II study in patients with diabetic renal disease, pyridoxamine has also been shown to be renoprotective (Williams et al. 2007).
The liposoluble derivatives of vitamin B1, benfotiamine and thiamine, also exhibit AGE-lowering properties. In contrast to pyridoxamine, benfotiamine and thiamine can decrease the formation of reducing sugars and polyol pathway intermediates (Berrone et al. 2006). Both have also been shown to be beneficial in experimental models of diabetic nephropathy (Babaei-Jadidi et al. 2003; Karachalias et al. 2003).
Type 2 diabetic patients consuming a high-AGE diet treated with benfotiamine showed reduced levels of circulating AGE levels, as well as lower levels of ROS (Stirban et al. 2006). Another study failed to show a positive effect of benfotiamine in type 1 diabetic patients however (Du et al. 2008).
Carnosine, a naturally occurring dipeptide in the brain other tissues, is another antioxidant. Carnosine also reacts with aldehydes, including aldose and ketose sugars, which attenuates AGE formation (Alhamdani et al. 2007; Hipkiss and Chana 1998; Hipkiss et al. 1998). It has been found to have renoprotective effects experi- mental diabetic nephropathy models.
Cleavage of Preformed AGEs
AGEs form nonreversible covalent cross-links within and between tissue proteins and other organic compounds. N-Phenacylthiazolium bromide (N-PTB) (Vasan et al. 1996) and alagebrium (Forbes et al. 2003) novel therapies postulated to cleave or at last reduce cross-linking, allowing glycated proteins to be removed via scaven- ger receptors and renal excretion, although the exact mechanism is yet to be determined.
Specifically, N-PTB cleaved α-dicarbonyl intermediates and reduced AGE formation (Vasan et al. 1996). Unexplainable increases in blood pressure and associated toxicity are seen with PTB; thus this was not translated into the clinic (Cooper et al. 2000). Alagebrium treatment has been shown to attenuate human- isolated systolic hypertension and symptoms of diastolic heart failure (Little et al. 2005). Our own studies have shown that alagebrium had both reno- and atheroprotective actions by reducing circulating and tissue AGE accumulation (Watson et al. 2011, 2012).
Targeting AGE Precursors: Methylglyoxal
In addition to advanced glycation end products (AGEs), more recently it has been shown that the precursors of AGEs, early reactive dicarbonyls, may also play a role in the development of diabetic complications but also in ageing, neurodegenerative diseases and Alzheimer’s disease. There are a number of early or intermediate AGEs, known as α-dicarbonyls including, glyoxal, 3-deoxyglucosone and methylglyoxal. They are formed as by-products of glycolysis, lipid peroxidation and during the degradation of AGEs (Thomas 2011).
These intermediates can be rapidly transformed into irreversible AGEs and AGE cross-links. Indeed, our studies have shown that increasing MGO levels independent of diabetes is associated with increased vascular inflammation and plaque development with plaque area similar to that observed in diabetes (Watson et al. 2012; Tikellis et al. 2014). Type 2 diabetic patients also have elevated levels of reactive dicarbonyls in both their serum and urine (Waris et al. 2015) (Fig. 3).
Conclusion
AGE lowering has been shown to be an effective measure to reduce chronic disease, ageing and complications of diabetes via effects on ROS formation, PKC and NF-κB activation and inflammation. A number of currently used therapeutic agents such as blockers of the RAAS reduce AGEs as a side effect of their main action, in addition to BP reduction and attenuation of ROS dysregulation and inflammation.
Although these agents lower AGEs, the AGE-lowering effect is usually modest. Thus there is an urgent need to develop more effective agents with potent AGE-lowering properties.
Only a handful of direct AGE inhibitors have been developed and are investigated at this stage in RCTs, such as pyridoxamine and vitamin B derivatives. Ongoing RCTs need to confirm their end-organ protection in chronic diseases including diabetes. Epalrestat