| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Consiglio Nazionale delle Richerche Institute of Clinical Physiology (G.B., A.M.S., G.L., S.D.T., E.B., E.F., A.G.) and Department of Internal Medicine (A.C., A.N., E.F.), University of Pisa School of Medicine, 56124 Pisa, Italy
Address all correspondence and requests for reprints to: Giuseppina Basta, DBiol, Consiglio Nazionale delle Richerche Institute of Clinical Physiology, Via Moruzzi, 1, 56124 Pisa, Italy. E-mail: lapina{at}ifc.cnr.it.
| Abstract |
|---|
|
|
|---|
-(carboxymethyl)lysine-protein adducts (CML) and S100A12 protein, with their cellular receptor (RAGE) is implicated in the pathogenesis of diabetic vascular complications. RAGE has a circulating secretory receptor form, soluble RAGE (sRAGE), which, by neutralizing the action of advanced glycation end products, might exert a protective role against the development of cardiovascular disease. Objective: The objective of the study was to investigate whether plasma sRAGE levels are associated with glycemic control, proinflammatory factors, or circulating ligands of RAGE such as plasma CML and S100A12 protein.
Study Design: We studied 160 subjects, 84 subjects with type 2 diabetes (aged 60 ± 7 yr) and 76 nondiabetic controls (aged 45 ± 10 yr).
Results: Plasma sRAGE was lower in diabetic patients than controls [141 (53345) vs. 735 (5191001) pg/ml, median (interquartile range), P < 0.0001], whereas CML levels were higher in diabetic patients than controls [67.9 (46.084.7) vs. 43.4 (28.065.0) µg/ml, P < 0.0001]. In stepwise regression analysis of the whole data set, hemoglobin A1c, insulin resistance (as homeostasis model assessment), and C-reactive protein were independently associated with plasma sRAGE, whereas age was not. In a subgroup of 26 diabetic and 24 nondiabetic subjects of similar age (54 ± 3 yr), plasma S100A12 levels were higher in diabetic subjects [49 (39126) vs. 28 (2139) ng/ml]. Moreover, low sRAGE and high S100A12 were strongly associated with increased risk for cardiovascular disease (Framingham score). In this subgroup, the plasma S100A12 level was the only determinant of plasma sRAGE concentration.
Conclusion: Plasma level of sRAGE is down-regulated in chronic hyperglycemia; among its ligands, S100A12 protein, but not CML, appears to be associated with this effect.
| Introduction |
|---|
|
|
|---|
-(carboxymethyl)lysine-protein adducts (CML), are formed in a greatly accelerated degree in diabetes (1, 2). Accumulation of AGEs and the enhanced tissue expression of the main receptor for AGEs (RAGE) may contribute to vascular damage (3, 4). Two classes of RAGE ligands, AGEs and S100A12 proteins, appear to drive receptor-mediated cellular activation and, potentially, perturbation of a variety of homeostatic functions of the vasculature (5, 6). S100A12, also called EN-RAGE (extracellular newly identified receptor for AGE binding protein), is a proinflammatory cytokine expressed mainly by granulocytes and is involved in Ca+-dependent signal transduction events (7). Recently it has been reported that human vascular cells express several RAGE variant proteins, including three novel human RAGE transcripts (8), all encoding truncated soluble forms of RAGE (sRAGE), consisting of only the extracellular ligand binding domain and lacking the cytosolic and transmembrane domains. Due to these characteristics, the sRAGE could represent a naturally occurring competitive inhibitor of the signaling pathways induced by the interaction of AGEs with its cellular receptor; by functioning as a decoy, sRAGE may well contribute to the removal/neutralization of circulating RAGE-ligands (8). In animal models of streptozotocin-induced diabetes or genetically manipulated mice, it has been demonstrated that treatment with sRAGE, produced by recombinant technology, prevents cell-bound RAGE signaling (9, 10, 11). Administration of sRAGE to experimental animals predisposed to exaggerated neointimal expansion significantly limited atherosclerosis and restenosis (12). In humans, clinical studies in nondiabetic subjects have shown that higher levels of endogenous sRAGE may be associated with a lower incidence of coronary artery disease (13), arthritis (14), and hypertension (15). Also, it has been suggested that in type 1 diabetic subjects, higher sRAGE levels confer resistance to AGEs (16) and that plasma sRAGE levels were inversely associated with components of the metabolic syndrome (17).
The aim of the present study was to investigate whether plasma sRAGE levels are associated with glycemic control, proinflammatory factors, or circulating ligands of RAGE such as plasma CML and S100A12 protein.
| Subjects and Methods |
|---|
|
|
|---|
A total of 160 subjects, 84 with type 2 diabetes and 76 nondiabetic subjects, were studied (Table 1
). Nondiabetic subjects were recruited according to the following criteria: 1) absence of diabetes on the oral glucose tolerance test (18), 2) absence of metabolic or other diseases, and 3) no treatment with drugs known to affect glucose tolerance. Diabetic patients were in good general health without evidence of cardiac, hepatic, renal, or chronic inflammatory diseases or other chronic diseases, as determined by history, examination, and routine blood chemistry. Patients had normal serum creatinine (i.e. <1.2 mg/dl) and were not taking any medication known to affect glucose tolerance; 21 diabetic subjects were on renin-angiotensin-system (RAS) blockade. Hypertension was defined as blood pressure greater than 140/90 mm Hg or the use of antihypertensive medications. Twenty-three nondiabetic subjects were newly diagnosed with essential hypertension, and none was on antihypertensive treatment. Forty-four diabetic subjects were also hypertensive, of whom 29 were on antihypertensive treatment. No subject was involved in strenuous physical activity, and body weight was stable (±2 kg) for at least 3 months before the study.
|
Laboratory methods
Blood samples were drawn in tubes without additives, containing heparin or citrate (for routine biochemistry). Fasting plasma glucose, glycosylated hemoglobin (HbA1c), fasting plasma insulin, total cholesterol, triglycerides, white blood cells (WBCs), C-reactive protein (CRP), and creatinine were determined as described previously (18, 19). For CML, sRAGE, and S100A12 determination, blood samples were collected in tubes containing Na2EDTA and aprotinin, centrifuged at 4 C and immediately divided in aliquots. Plasma and serum samples were stored at 80 C until analysis.
Determination of plasma CML levels
Plasma CML levels were measured in triplicate by an in-house competitive ELISA using the mouse F(ab)2 fragment anti-AGE monoclonal antibody (clone 6D12) (ICN Biochemical Division, Aurora, OH), as described previously (20). The lower limit of detection of CML was 0.5 µg/ml.
Determination of plasma sRAGE levels
Plasma sRAGE levels were determined using an ELISA kit (DuoSet ELISA development kit; R&D Systems, Minneapolis, MN) containing the basic components required for the development of double-sandwich ELISAs. Briefly, 96-well microplates (Nunc-Immuno Plates Maxisorp, Nunk, Roskilde, Denmark) were coated with monoclonal antihuman RAGE (1 µg/ml) in coating buffer [10 mmol/liter PBS (pH 7.2)] and incubated overnight at room temperature (RT). Four rinses with washing buffer followed each incubation step. After blocking with reagent buffer (1% BSA in PBS) at RT for 1 h, 100 µl of sample diluted in reagent diluent was added and incubated at RT for 2 h. Recombinant human RAGE was used as standard in a concentration range of 31.24000 pg/ml. Then 100 µl of biotinylated goat antihuman RAGE (200 ng/ml) was added and incubated at RT for 2 h. Next, 100 µl of streptavidin-horseradish peroxidase (1:200) was added and incubated for 20 min at RT. Finally, 100 µl of tetramethybenzidine substrate (Sigma, St. Louis, MO) was added. After 530 min, 50 µl of 2 mol/liter sulfuric acid was added to stop the reaction and OD at 450 nm determined by an ELISA plate reader.
The intraassay and interassay coefficients of variation were less than 5.9% and less than 8.2%, respectively. The lower limit of detection of sRAGE was 21.5 pg/ml. According to the manufacturer, no significant cross-reactivity with EN-RAGE, high-mobility group box 1, S100A10 or S100Baa, and CML was observed.
Determination of plasma S100A12 levels
Concentrations of S100A12 were determined using a double-sandwich ELISA system established in our laboratory. The 96-well microplates (Nunc-Immuno Plates) were coated with 50 ng/ml antihuman S100A12 antibody (R&D Systems) in 50 mmol/liter sodium bicarbonate buffer (pH 9.6) and incubated for 16 h at 4 C. Four rinses with washing buffer followed each incubation step. After blocking with 100 µl blocking buffer [1% BSA in 20 mmol/liter HEPES (pH 7.4)] at RT for 1 h, 100 µl of sample in at least three dilutions using blocking buffer containing 1 mmol/liter CaCL2 and 0.05% Tween 20 was added and incubated at RT for 2 h. Recombinant S100A12 protein (21, 22) was kindly supplied by Dr. Olga Moroz and used as standard in a concentration range of 0.150 ng/ml. Then 100 µl of biotinylated goat antihuman S100A12 (R&D Systems) (100 ng/ml) was added and incubated at RT for 2 h. Next, 100 µl of streptavidin-horseradish peroxidase was added and incubated for 20 min at RT. Finally, 100 µl of tetramethybenzidine substrate (Sigma) was added. After 530 min, 50 µl of 2 mol/liter sulfuric acid was added to stop the reaction and OD at 450 nm determined by an ELISA plate reader. The lower limit of detection of S100A12 was 0.05 ng/ml. The intraassay and interassay coefficients of variation were less than 7.9% and less than 8.0%, respectively.
Statistical analysis
Data are given as the mean ± SD. Variables with a skewed distribution are expressed as median and interquartile range. Group differences were analyzed by Student t test, Mann Whitney U test, and
2 test for normally distributed, nonnormally distributed, and noncontinuous variables, respectively.
Relationships between natural logarithm of plasma sRAGE, Ln(sRAGE), and other factors were first assessed by Pearson correlation analysis; for these, variables with a skewed distribution were log transformed. Normal distribution was checked using the Z test. All variables significantly associated with Ln(sRAGE) were included in the multivariate analysis plus gender and BMI as confounding factors. We then considered, one at a time, the explanatory variables by stepwise regression analysis by adding to the model at each step the variable that explained most variation in Ln(sRAGE) and stopping when the addition of an extra variable made no significant improvement in the amount of variation explained. In Tables 2
and 4
, we reported only significant associations as simple or partial (in stepwise analysis) correlation coefficients and amount of variation explained (R2 adjusted). In Table 4
, we used a hierarchical approach by performing the multivariate analysis with all variables that were significant in simple correlation analysis (model 1) and sequentially excluding the variables that resulted as significant (models 2 and 3) in the multivariate approach. The 10-yr cardiovascular disease (CVD) risk was calculated by the Framingham score (23).
|
|
| Results |
|---|
|
|
|---|
In the whole data set, in simple correlation analysis, sRAGE levels were inversely associated with age, fasting plasma glucose, HbA1c, HOMA, triglycerides, and CRP levels (Table 2
and Fig. 1
). No correlation was found with BMI, total cholesterol, blood pressure, white blood cell count, creatinine, or CML levels (Table 2
). The reciprocal relationship between sRAGE and HbA1c was the strongest, with patients and controls falling on the same regression line (Fig. 1
). When the diabetic patients and controls were analyzed separately, fasting plasma glucose, HbA1c, and HOMA remained significantly associated with sRAGE in diabetic patients, whereas CRP levels remained inversely correlated with sRAGE levels in controls (Table 2
). In stepwise multiple regression analysis, only HbA1c, HOMA, and CRP remained independent determinants of plasma sRAGE concentrations. When analyses were performed separately in the two groups, HbA1c and CRP in nondiabetic subjects, and HbA1c in diabetic patients, were identified as independent predictors of plasma sRAGE (Table 2
).
|
|
|
| Discussion |
|---|
|
|
|---|
We found HbA1c to be a significant independent factor inversely associated with plasma sRAGE levels in both diabetic and nondiabetic subjects. CRP levels, on the other hand, were inversely associated only in the nondiabetic group, possibly because in diabetic subjects CRP had a narrow distribution (Fig 1
). Recently another study has described HbA1c and CRP as major factors determining plasma sRAGE but only in nondiabetic subjects (17). However, both studies confirm that sRAGE levels are associated with glycemic control and inflammation; the discrepancy may be due to the fact that we measured total sRAGE levels, whereas Koyama et al. (17) measured a specific variant of sRAGE. Importantly, we have shown the inverse relationship between plasma sRAGE and S100A12, a relevant proinflammatory molecule strongly linked with RAGE signaling. This finding suggests that S100A12 could be involved in a negative-feedback mechanism in the RAGE signaling pathway. Further studies will be needed to elucidate the cellular sources of sRAGE and confirm that S100A12 and other cytokines are involved in its regulation/production.
A role of AGEs and oxidative stress in the development of insulin resistance has been hypothesized (27, 28). Several epidemiological studies have shown a relationship between hyperglycemia and CVD in the general population (27); additionally, postprandial hyperglycemia has been indicated as an independent predictor of increased intima-media thickness in both nondiabetic (29) and diabetic subjects (30). The lower levels of sRAGE detected in diabetic patients might indicate an enhanced propensity to diabetic vascular complications, given that RAGE ligands can interact freely with cell-bound RAGE signaling (4). In support of this possibility, we found that subjects between 47 and 61 yr of age with low sRAGE levels carried a higher CVD risk (by Framingham score). These data together with the inverse association between sRAGE and atherosclerosis previously published (13, 17) suggest a role for sRAGE as a novel marker in the development of CVD.
Recent studies have shown an inverse relationship between sRAGE and blood pressure (15, 17). In the present series, no relationship was found between sRAGE and any measure of blood pressure (Fig. 1
), in either the whole group or subgroups regardless of antihypertensive treatments. In the study by Geroldi et al. (15), the relationship was present only with pulse pressure, and in the study by Koyama et al. (17), only diastolic pressure correlated in the diabetic group, suggesting that subject selection may account for the differences. We also investigated the possible effect of RAS blockade on sRAGE levels because it has been postulated that angiotensin-converting enzyme inhibition reduces the accumulation of AGE products in diabetes partly by increasing production and secretion of plasma sRAGE (31). In our diabetic patients, the sRAGE values of those treated with angiotensin-converting enzyme inhibitors tended to be higher than those of untreated patients, but only a prospective trial can definitively prove this predicament.
The mechanism by which plasma sRAGE levels are decreased in diabetic patients is unknown. One possibility is that hyperglycemia inhibits sRAGE production directly or via increased cytokines and/or hyperglycemia-induced AGEs. Alternatively, lower sRAGE levels could be due to an increased clearance of AGE ligand/sRAGE complexes (32). However, plasma CML levels were found to not correlate with sRAGE levels, despite being higher in diabetic patients than controls, suggesting that CML may not be involved in sRAGE modulation. On the other hand, non-CML AGEs, representing the extent of glycation, could be inversely associated to sRAGE; in fact, it has been reported that non-CML AGE levels are associated with both fasting glucose and HbA1c levels in diabetic patients (33), and we have found a strong inverse association between sRAGE and HbA1c.
In the subgroup of subjects between 47 and 61 yr of age, we also determined the plasma levels of the RAGE-ligand, S100A12, a molecule closely linked with RAGE signaling, which has been reported to be elevated in many chronic inflammatory diseases (34, 35, 36). As previously described (37), plasma S100A12 levels were related to subclinical inflammation (CRP) and glycemic control. Thus, it can be hypothesized that hyperglycemia itself might stimulate S100A12 protein production, either directly in granulocytes, monocytes, or other cells or indirectly via hyperglycemia-produced cytokines, e.g. IL-6, because recently it has been reported in human macrophages (38). The inverse correlation between S100A12 and sRAGE suggests that insulin resistance, by promoting hyperglycemia and inflammation, may up-regulate S100A12 release, which in turn decreases plasma sRAGE levels.
This study has some limitations. First, the system so far developed for sRAGE determination cannot distinguish among diverse sRAGE splice variants. Because the proportion of diverse isoforms of sRAGE could differ among individuals, the development of new specific assays will improve the evaluation of sRAGE levels. Second, because S100A12 protein has been measured in a small subgroup of subjects, our study also had insufficient power to evaluate, in multivariate analysis, the relation with sRAGE and the other parameters; this important caveat limits the generalizability of the conclusions, and further studies need to be done to confirm in larger clinical studies these important relationships. Third, whether lower sRAGE levels in diabetic patients are the consequence of diabetes or a causal factor of the disease needs to be elucidated. Prospective and intervention studies are necessary to clarify the cause-effect relationships among sRAGE, hyperglycemia, and S100A12 protein.
| Acknowledgments |
|---|
| Footnotes |
|---|
First Published Online August 22, 2006
Abbreviations: AGE, Advanced glycation end products; CML, N
-(carboxymethyl)lysine-protein adducts; CRP, C-reactive protein; CVD, cardiovascular disease; HbA1c, glycosylated hemoglobin; HOMA, homeostasis model assessment; Ln, natural logarithm; RAGE, AGE receptor; RAS, renin-angiotensin-system; RT, room temperature; sRAGE, soluble RAGE; WBC, white blood cell.
Received November 28, 2005.
Accepted August 14, 2006.
| References |
|---|
|
|
|---|
B. Diabetes 50:27922808
-carboxymethyllisine (CML) are increased in systemic sclerosis. Rheumatology, in pressThis article has been cited by other articles:
![]() |
H. Maillard-Lefebvre, E. Boulanger, M. Daroux, C. Gaxatte, B. I. Hudson, and M. Lambert Soluble receptor for advanced glycation end products: a new biomarker in diagnosis and prognosis of chronic inflammatory diseases Rheumatology, July 9, 2009; (2009) kep199v1. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. Z. Kalea, N. Reiniger, H. Yang, M. Arriero, A. M. Schmidt, and B. I. Hudson Alternative splicing of the murine receptor for advanced glycation end-products (RAGE) gene FASEB J, June 1, 2009; 23(6): 1766 - 1774. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. Katakami, M. Matsuhisa, H. Kaneto, T.-A. Matsuoka, K. Sakamoto, T. Yasuda, and Y. Yamasaki Endogenous secretory RAGE but not soluble RAGE is associated with carotid atherosclerosis in type 1 diabetes patients Diabetes and Vascular Disease Research, September 1, 2008; 5(3): 190 - 197. [Abstract] [PDF] |
||||
![]() |
Z Sternberg, B Weinstock-Guttman, D Hojnacki, P Zamboni, R Zivadinov, K Chadha, A Lieberman, L Kazim, A Drake, P Rocco, et al. Soluble receptor for advanced glycation end products in multiple sclerosis: A potential marker of disease severity Multiple Sclerosis, July 1, 2008; 14(6): 759 - 763. [Abstract] [PDF] |
||||
![]() |
B. I. Hudson, A. M. Carter, E. Harja, A. Z. Kalea, M. Arriero, H. Yang, P. J. Grant, and A. M. Schmidt Identification, classification, and expression of RAGE gene splice variants FASEB J, May 1, 2008; 22(5): 1572 - 1580. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. Kalousova, M. Jachymova, O. Mestek, M. Hodkova, M. Kazderova, V. Tesar, and T. Zima Receptor for advanced glycation end products--soluble form and gene polymorphisms in chronic haemodialysis patients Nephrol. Dial. Transplant., July 1, 2007; 22(7): 2020 - 2026. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |