In the past ten years, the understanding of endothelial dysfunction in CKD has greatly expanded. Important findings include the discovery of circulating EPC, reflecting endothelial repair capacity, and the demonstration of the specific toxicity of uremic compounds . The relationship between circulating EPC levels and uremic toxins in CKD patients is not well understood .
AGEs have been identified in the blood and tissues of patients with ESRD . Oxidative and carbonyl stress have been identified as factors in uremia . Endogenous AGEs undergo cellular proteolysis and are released into plasma as AGE free adducts for urinary excretion . These glycation free adducts include free fructoselysine (FL), free hydroimidazolones, CML, CEL, pentosidine, and other AGEs . The normal high renal clearance of AGE free adducts is impaired markedly in CKD and thus considered as potential uremic toxins .
AGEs have been demonstrated to impair endothelial functions in several ways. They up-regulate adhesion molecule expression  and increase endothelial layer permeability . In vitro, AGEs prompt intracellular generation of hydrogen peroxide and activation of nicotinamide adenine dinucleotide phosphate (NADPH) oxidase . AGEs enhance apoptosis, depress EPC migration and tube formation in a concentration-dependent manner, and increase RAGE expression in these cells [18–20]. The negative impact of AGEs on bone marrow mesenchymal stem cells (MSCs) has been reported by us previously .
Mixtures of AGEs have been prepared for study after a long, complex, in vitro preparation process, resulting in “AGE-modified” proteins of unknown composition . Few studies have used the complete range of genuine uremic AGE compounds. [Glorieux et al. 29] reported that genuine AGE compounds, including CML and CEL, could activate leukocyte response and hence play a role in uremia related atherogenesis. [Wang et al. 27] found that CML could accelerate progression of atherosclerotic calcification in diabetes. In the present study, the biologic effects of CML and CEL were studied, with specific emphasis on EPC function.
We believe the concentrations of CML and CEL used in this study requires special comment. EUTox reported the maximum concentration of CML is 6.9 mg/L in CKD patients in 2003 . In their follow-up publication in 2007 , EUTox reported the maximum concentration (Cmax) of CML to be 111–221 nM (≈ 40 μg/L) and CEL to be 336–817 nM (≈ 200 μg/L) in CKD patients . Recently, EUTox reported the Cmax of CML as 18.5 mg/L in CKD patients . These different concentrations of CML and CEL might be explained by their use of different analytical methods, ELISA [17, 24] and liquid chromatography-mass spectrometry (LC-MS) . The quantification of AGEs using ELISA is problematic. Because of uncertain epitope specificity of antibodies employed, and the use of highly modified standard antigens dissimilar to the minimally modified antigens in physiological samples, AGE detection using ELISA does not usually provide AGE levels in absolute concentration, but rather in arbitrary units with or without normalisation to a reference AGE protein standard . LC-MS analysis is now becoming the standard method for quantitation of glycation adducts . We believe the findings published in 2007 were correct, according to our previous analysis (mean serum concentration of CML in ESRD patients without dialysis therapy ≈ 200 μg/L using LC-MS analysis, n = 5, data not published). For this reason we used CML and CEL in the 250–1000 μg/L range in this study. We believe previously published concentrations (2003, 2012) were higher than our study because of their use of ELISA to detect AGE.
MAPKs are a family of serine/threonine kinases comprised of extracellular signal-regulated kinase (ERK), c-Jun N-terminal kinase (JNK), and p38 MAPK (p38) . Growth factor-induced proliferation of EPC has been demonstrated to be regulated by activation of the MAPK signaling pathway [44, 45]. The MAPK signaling pathway activation has been demonstrated to be biphasic [46–50], with an early rapid increase in MAPK activation from 5 to 60 minutes, followed by a late second wave of MAPK activation of lower amplitude beginning at 7–10 hours. This second wave of MAPK activity was sustained for hours after growth factor stimulation and was critical for cellular progression from G1 into S phase [46, 48, 49, 51]. In this study, the activity of MAPKs in EPCs treated with CML or CEL for 24 hours was tested. The reduced growth response to CML or CEL in EPCs was accompanied by significantly less phosphorylation of MAPKs, suggesting that CML and CEL may reduce EPCs proliferation via inhibition of the MAPK signaling pathway.
Our study found that CML and CEL had no effects on AGE receptor (RAGE) expressions in EPCs. This is consistent with previous studies, which found that CML-modified proteins were unable to bind to RAGE . This may be because proteins modified by AGEs to a high extent (30–40 AGE residues per molecule) are competent AGE receptor ligands, whereas proteins modified by AGEs to a minimal extent (1–2 AGE residues per molecule) are not [53, 54].