Both authors contributed equally to this study.
Inhibitors of advanced glycation end-products prevent loss of enteric neuronal nitric oxide synthase in diabetic rats
Version of Record online: 17 OCT 2007
© 2007 The Authors
Neurogastroenterology & Motility
Volume 20, Issue 3, pages 253–261, March 2008
How to Cite
Jeyabal, P. V. S., Kumar, R., Gangula, P. R. R., Micci, M.-a. and Pasricha, P. J. (2008), Inhibitors of advanced glycation end-products prevent loss of enteric neuronal nitric oxide synthase in diabetic rats. Neurogastroenterology & Motility, 20: 253–261. doi: 10.1111/j.1365-2982.2007.01018.x
- Issue online: 17 OCT 2007
- Version of Record online: 17 OCT 2007
- Received: 23 June 2007 Accepted for publication: 24 July 2007
- advanced glycation end-product;
- neuronal nitric oxide synthase
Abstract Gastrointestinal dysfunction is common in diabetes, and several studies indicate that loss of neuronal nitrergic inhibition may play an important role in its pathogenesis. However, the mechanisms responsible for this effect remain largely unknown. We have previously shown that advanced glycation end-products (AGEs) formed by non-enzymatic glycation dependent processes, can inhibit the expression of intestinal neuronal nitric oxide synthase (nNOS) in vitro acting via their receptor, receptor for AGEs. We now hypothesized that this effect may also be important in experimental diabetes in vivo. We aimed to evaluate the role of AGEs on duodenal nNOS expression and the effects of aminoguanidine (a drug that prevents AGE formation) and ALT-711 (AGE cross-link breaker) in experimental diabetes. Streptozotocin induced diabetic rats were randomized to no treatment, treatment with aminoguanidine (1 g L−1 daily through drinking water) at the induction of diabetes, or treatment with ALT-711 (3 mg kg−1 intraperitoneally), beginning at week 6. A fourth group was used as healthy controls. We performed real time polymerase chain reaction, Western blotting and immunohistochemistry to detect nNOS expression. AGE levels were analysed using sandwich ELISA. Diabetes enhanced accumulation of AGEs in serum, an effect that was prevented by treatment with aminoguanidine and ALT-711. Further, diabetic rats showed a significant reduction in duodenal nNOS expression by mRNA, protein and immunocytochemistry, an effect that was prevented by aminoguanidine. ALT-711 had similar effects on nNOS protein and immunohistochemistry (but not on mRNA levels). The generation of AGEs in diabetes results in loss of intestinal nNOS expression and may be responsible for enteric dysfunction in this condition. This study suggests that treatment directed against AGEs may be useful for the treatment of gastrointestinal complications of diabetes.
advanced glycation end-products
receptor for advanced glycation end-products
neuronal nitric oxide synthase
bovine serum albumin
advanced glycated bovine serum albumin
Patients with diabetes mellitus commonly experience gastric and intestinal dysfunction. The pathogenesis of these complications remains poorly understood, although the degree of hyperglycaemia appears to be an important determinant in their incidence and severity.1 At a molecular level, much attention has recently been devoted to the potential role of altered nitrergic signalling in the enteric nervous system. Nerves throughout the luminal gastrointestinal tract express neuronal nitric oxide synthase (nNOS), which generates nitric oxide (NO), a key neurotransmitter in the regulation of gastrointestinal motility.2 Impairment of NO production appears to be important in the pathogenesis of gastroparesis as evidenced by mice with genetic deletion of nNOS. Further, several investigators including our laboratory has shown that experimental diabetes in rodents can result in reduced nitrergic signalling from mechanisms that may potentially include reduced nNOS expression and loss of the functionally active dimer form of nNOS.3
By contrast to gastroparesis, diabetic intestinal dysfunction has received little attention, even though it is relatively common. The term diabetic enteropathy is often used to explain disturbances in bowel function such as chronic diarrhoea which occurs in 15% or more of diabetic patients as reported in large prospective population based studied.4 A role for decreased nNOS expression has been suggested by experimental studies in animals5–7 and by isolated human reports.8 nNOS deficiency is expected to slow gastrointestinal transit. Paradoxically, this could result in diarrhoea because of secondary bacterial overgrowth which is a known problem in diabetics.9–11 It has also been shown that myenteric nNOS expression in the intestine, unlike the stomach, is not dependent on vagal innervation.5 It therefore remains unknown how diabetes induces changes in nNOS expression. Potential mediators of this effect include advanced glycation end-products (AGEs), which are a heterogeneous group of molecules formed by non-enzymatic attachment of reducing sugars to the amino groups of various proteins through a series of complex intermediary reactions including Schiff bases and amadori products. N-carboxymethyl-lysine, pentosidine and methylglyoxal derivatives are classical examples for AGEs. Although some AGE formation occurs even in physiological states, this process is significantly accelerated in uncontrolled diabetes due to the abundant availability of glucose. Although glycation itself can lead to structural and functional changes in the target protein, perhaps a more important consequence may be the ability of the conjugate to activate the receptor for AGEs (RAGE). RAGE is a member of the immunoglobulin superfamily of cell surface molecules, capable of recognizing not only AGE but a variety of other ligands including fibrillar amyloid, amphoterin and S100/calgranulins (including EN-RAGE).12,13 Serum and tissue levels of both AGEs, and other potential ligands for RAGE are elevated in diabetes, both in the serum and within tissues and have been linked to many other complications of diabetes mellitus including those affecting the blood vessels, kidneys, nerves and retina.14
We have recently shown that RAGE is also expressed by myenteric neurons in the intestine and that its activation in vitro can suppress nNOS expression and NO release.15 The aim of the present study was to test the hypothesis that AGE–RAGE signalling was also important in the modulation of intestinal nNOS expression in an in vivo model of diabetes. To this end we examined the expression of duodenal nNOS in streptozotocin (STZ) induced diabetes in control rats and in response to two drugs, aminoguanidine, a well characterized and widely studied inhibitor of AGE formation and a ALT-711, a novel compound that acts as an AGE-cross-link breaker and capable of reversing the glycated state of a molecule.
Materials and methods
STZ, aminoguanidine and bovine serum albumin (BSA) were obtained from Sigma Chemical Co (St Louis, MO, USA). ALT-711 was kindly provided by Alteon Inc. (Montvale, NJ, USA). Reagents for sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and Western blot analysis were purchased from Bio-Rad (Hercules, CA, USA). Western blot stripping buffer was from Pierce (Rockford, IL, USA).
Eight-week-old male Sprague–Dawley rats were housed in pairs in cages with wire mesh bases into controlled environment at a constant room temperature (22 °C), humidity, and 12 h light–darkness cycles. Rats were fed standard chow, and water was given ad libitum. Experiments were approved by the Institutional Animal Care Committee at the University of Texas Medical Branch.
Rats were divided into four groups of six animals each. Group I rats were maintained as healthy controls. Diabetes was induced in Groups II, III and IV as described below. Rats in group II were maintained as disease controls. Diabetic rats in Group III received aminoguanidine (1 g L−1 daily), a dose similar to that shown to be previously effective in diabetic peripheral neuropathy16 in drinking water from day 2 of induction to the end of the experimental period whereas those in Group IV received ALT-711 (3 mg kg−1 daily) by intraperitoneal injection beginning at the sixth week of induction through the entire experimental period.
A single intraperitoneal injection of 55 mg kg−1 STZ dissolved in freshly prepared 50 mmol L−1 citrate buffer (pH 4.0) immediately before administration was given to the rats. Blood concentrations of glucose were measured 48 h later (day 0) in blood obtained from the cut tip of the tail and measured by Accu-chek advantage meter (Roche diagnostics, Indianapolis, IN, USA). Only rats with a blood glucose concentration above 200 mg dL−1 were included in diabetic groups. STZ induced diabetes causes mortality during the experimental period.17 Hence, diabetic rats received minimal doses (4 U) of insulin (Lantus; Aventis Pharmaceuticals Inc., Frankfurt, Germany) intra-muscularly on every second day to maintain body mass and improve survival during the experimental period.18 Animals were sacrificed 12 weeks after induction of diabetes. Blood was centrifuged to obtain serum for measuring AGEs.
Body weight and blood glucose
Body weight and blood glucose concentrations were subsequently measured every 2 weeks in overnight fasted animals through out the period.
Enzyme-linked immunosorbent assay for AGEs
Wells [96-well Enzyme-linked immunosorbent assay (ELISA) plate; FALCON, Franklin Lakes, NJ, USA] were coated with polyclonal anti-AGE antibody (AGE102; 10 μg mL−1; Biologo, Kronshagen, Germany) in 50 mmol L−1 carbonate buffer (pH 9.6) overnight at room temperature as previously described.19 The wells were then washed with phosphate-buffered saline (PBS) containing 0.05% Tween 20 and blocked at room temperature with PBS containing 0.25% BSA. After washing, the wells were incubated with the standards (AGE-BSA as described below; diluted 1 : 10–1 : 100 000) or samples (rat serum diluted in PBS 1 : 10–1 : 10 000) at room temperature for 3 h. After washing, the wells were incubated with monoclonal anti-AGE antibody (clone 6D12; 0.5 μg mL−1; Biologo) for 2 h at room temperature followed by antimouse IgG-HRP (1 : 7500; Bio-Rad) for 1 h at room temperature. The wells were washed again and developed with peroxidase substrate (Alpha diagnostic, San Antonio, TX, USA) for 20 min. After adding the stopping solution (Alpha diagnostic) the yellow colour was read at 450 nm. The absorbance obtained using AGE-BSA was used as standard. Advanced glycated bovine serum albumin (AGE-BSA) were produced by incubating BSA (50 mg mL−1) with 1 mol L−1 glucose in PBS in sterile conditions at 37 °C for 12 weeks. Excess unbound glucose was then removed using dialysis against a high volume of PBS then stored at −80 °C until needed.
Quantification of nNOS mRNA by real time reverse transcription PCR
Total RNA from the duodenum was extracted by using Trizol reagent (Invitrogen, Carlsbad, CA, USA) phenol–chloroform extraction method. First-strand cDNA was then generated using TaqMan Reverse Transcription Regents (Applied Biosystems, Foster City, CA, USA). Quantitative real-time polymerase chain reaction (PCR) was carried out using the ABI Prism 5700 Sequence Detector with TaqMan Universal PCR Master Mix Kit (Applied Biosystems). β-III tubulin was measured as a reference gene.
The following sequence-specific primers and probes were designed using primer express software 2.0 (Applied Biosystems) – for rat nNOS: forward – 5′-ACGGACCCGACCTCAGAGA-3′; reverse – 5′-CGAGGCCGAACACTGAGAAC-3′ and probe: 5′-6FAM-AAGTACTGGACCCCTGGCCAATGTGA-TAMRA-3′; for β-III tubulin: forward – 5′-GGGCCTTTGGACACCTATTCA-3′; reverse – 5′-GCCCTTTGGCCCAGTTGT-3′ and probe: 5′-6FAM-CCTGACAACTTTATCTTCGGTCAGAGTGGTG-TAMRA-3′. PCR conditions were 50 °C for 2 min, 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s, 55 °C for 15 s and 60 for 1 min. Data were normalized to β-III tubulin and relative quantification of gene expression was performed using the 2 [ΔΔ-C (T)] or 2-(DDCt) relative quantification method (to the difference on normalized number of cycles to threshold).
Duodenum of all experimental animals were homogenized in ice-cold lysis buffer (Promega, Madison, WI, USA) containing 125 mmol L−1 Tris (pH 7.8) with H3PO4with 10 mmol L−1 CDTA, 10 mmol L−1 DTT, 50% Triton X-100, 100 μmol proteinase cocktail inhibitor, 1 mmol L−1 phenylmethylsulphonylfluoride (PMSF). After centrifugation (for 2 min, at 4 °C, 12 000 g) the supernatants were collected and protein content was determined (BCA Protein Assay Kit; Pierce). For nNOS protein analysis, samples (300 μg protein) were diluted in 4x SDS loading buffer [0.25 mol Tris–HCl (pH 6.8), 8% SDS, 40% glycerol, 2.5% DTT, 0.05% bromophenol blue], boiled for 5 min, and subjected to 8% SDS-polyacrylamide gel electrophoresis (PAGE) with PROTEAN II xi system (Bio-Rad, Richmond, CA, USA). After electrophoresis, proteins were transferred to PVDF membranes and were incubated in blocking buffer (5% non-fat dry milk in TBS containing 0.1% Tween 20; TBST) for 1 h at room temperature and probed with mouse monoclonal anti-nNOS antibody (BD Transduction Laboratories, San Jose, CA, USA) at a dilution of 1 : 1000 in blocking buffer overnight at 4 °C. After washing in TBST, the blots were incubated with horseradish peroxidase (HRP)-conjugated goat antimouse immunoglobulin (Ig) G antibody (Bio-Rad) at a dilution of 1 : 2000 in TBST containing 2.5% non-fat dry milk for 1 h at room temperature.
For RAGE analysis, Samples (60 μg) were subjected to 10% SDS-PAGE (Bio-Rad) with Mini-PROTEAN II xi system (Bio-Rad) and transferred to PVDF membranes, and incubated in blocking buffer (5% non-fat dry milk in TBST) for 1 h at room temperature and probed with mouse monoclonal anti-RAGE antibody (Chemicon international, Temecula, CA, USA) at a dilution of 1 : 200 in blocking buffer overnight at 4 °C. After washing in TBST, the blots were incubated with HRP-conjugated goat antimouse immunoglobulin (Ig) G antibody (Bio-Rad) at a dilution of 1 : 1000 in TBST containing 2.5% non-fat dry milk for 1 h at room temperature. The immunoreactive bands were visualized using enhanced chemiluminescence (ECL kit; Amersham, Buckinghamshire, UK). The membranes were exposed to X-ray films and subsequently stripped and re-probed with mouse monoclonal γ-tubulin antibody (1 : 2000; Sigma). The intensities of bands were quantified using Alpha digidoc software (San Leandro, CA, USA).
Whole duodenum with myenteric plexus were fixed in Zamboni’s fixative for 10 min at room temperature. The Zamboni-fixed tissue samples were later dehydrated and embedded in paraffin wax at 55 °C according to a previously described method. Approximately 6-μm-thick sections were cut on a microtome and placed in a water bath at 48 °C. Thereafter, sections were transferred onto prewashed microscopic slides, which were dried in an oven at 55 °C for 30 min to enhance attachment of sections. The sections were then de-paraffinized in xylene and processed for immunohistochemistry. After 30-min incubation in the blocking reagent, the appropriate dilution of primary mouse monoclonal anti-nNOS antibody (BD Transduction Laboratories) and negative control reagents were applied. The sections were incubated in primary antibodies for 60 min at room temperature. The slides were then washed and incubated for 30 min with prediluted biotinylated antimouse IgG. After washing in TBST containing Tween 20, the sections were incubated with streptavidin–HRP conjugate for 20 min followed by washing with TBST. The peroxidase activity was visualized by incubating the specimens for 3 min in TBS solution containing 3,3-diaminobenzidine tetrahydrochloride and 0.03% hydrogen peroxide. The slides were later washed, counter-stained with haematoxylin for 30 s, and dehydrated before mounting. The antiserum to nNOS was used at 1 : 7500 dilution. The specificity of the antibody was confirmed by processing tissue samples in the absence of anti-nNOS serum. The number of nNOS positive cells in each myenteric ganglion was counted from five different microscopic fields for each group of rats.
Values are expressed as mean (SE). Where necessary, multiple comparisons between groups were performed using anova and if indicated, a post-test analysis was carried out using Dunnett’s method for comparing all three diabetic groups to healthy controls (group I).
Effects on bodyweight and blood glucose level
The body weight and blood glucose levels of all the animals were measured every two weeks throughout the treatment period, and data are shown in Fig. 1. During the experimental period, mean body weight was significantly decreased in all the diabetic rats whereas control rats gained in body weight. Mean blood glucose concentrations were higher in all the diabetic induced animals than in the control animals. Administration of low dose insulin (4 U) did not affect the blood glucose concentration of diabetic induced animals during the experimental period. Further, treatment with aminoguanidine or ALT-711 did not show any significant effect on body weight or blood glucose levels.
Changes in circulating AGE levels
We evaluated serum AGE levels in control and experimental animals and found significant differences among the four groups (P < 0.001 by anova), with elevations in the diabetic group that were not seen in either treatment group as compared with controls (Fig. 2).
Duodenal RAGE expression
Western immunoblotting analysis using mouse monoclonal anti-RAGE antibody revealed immunoreactive bands of protein extracts at approximately 48 kDa (Fig. 3). There was no significant change observed between immunoreactive bands between the groups.
Effect of aminoguanidine and ALT-711 on nNOS mRNA expression
Real time PCR was performed to examine nNOS expression at the mRNA level. Data were normalized against β-III tubulin mRNA and expressed as a percentage of the mean of the control group. Significant differences in mRNA were seen (P = 0.003 by anova), with nNOS mRNA decreasing to 56% of control levels in diabetic rats (Fig. 4), an effect that was prevented by aminoguanidine but not by ALT-711.
Expression of nNOS at protein level
Western immunoblotting of the duodenal tissue homogenate using anti-nNOS antibody revealed a band at 155 kDa (Fig. 5) in control and all experimental groups. nNOS protein expressions differed significantly (P = 0.001 by anova) with a markedly lower value in the diabetic group, an effect not seen in either treatment group (which did not differ from controls).
Immunolocalization of nNOS
Fig. 6A shows myenteric ganglia in the duodenum of rats in the various experimental groups. Although nNOS expression was observed in the myenteric plexus of all experimental groups, the number of nNOS positive cells per ganglia was significantly different amongst the groups (P < 0.001), being reduced by nearly half in untreated diabetic rats compared with healthy controls, an effect that was reversed with treatment by either drug.
Perturbation in NO mediated relaxation of gastrointestinal tract is suggested to play a major role in gastrointestinal dysfunction. Our study shows that diabetes results in the accumulation of serum AGEs. Further, in diabetic rats, there is a significant reduction in myenteric nNOS expression, an effect that can be reversed by either of two strategies aimed at counteracting AGE formation – aminoguanidine given prophylactically or ALT-711 given therapeutically. Although we did not measure intestinal function in our experimental groups, these results provide the first reported evidence for the potential importance of AGE signalling in the pathogenesis of enteric neuropathy.
Activation of RAGE, a multi-ligand member of immunoglobulin superfamily triggers a broad spectrum of intracellular signalling pathways including NF-kB, JAk/stat, p21 ras, MAP kinases that lead to a variety of physiological effects in different biological systems.20–23 Engagement of RAGE with AGEs is also known to affect the expression of inducible NOS (iNOS)24 and endothelial NOS (eNOS).25 A recent study also suggested the involvement of AGEs in the reduction of nNOS expression in the penis of diabetic rats.26 Until recently, there was little or no information on the expression of this system in the gut, although one autopsy study of patients with amyloid polyneuropathy did demonstrate the presence of both RAGE and AGE within myenteric ganglia.27 Furthermore, AGEs do accumulate in the gastric pylorus19 and suppression of nNOS expression28 was also observed in experimental diabetes. In a previous study, our group provided the first definite demonstration of RAGE expression by myenteric neurons and showed that in vitro, activation of this receptor was responsible for downregulation of nNOS in the rat duodenum.15
In the present study, we have validated our initial in vitro findings in an in vivo model of diabetes, using two pharmacological agents (aminoguanidine and ALT-711) with different modes of action. Consistent with the literature, diabetes resulted in the accumulation of AGEs in both serum and the duodenum, at levels that are comparable with that previously reported, using similar assays.19 Both experimental drugs effectively lowered these levels.
Aminoguanidine is an older agent, that has been shown to prevent AGE formation in hyperglycaemic states and effectively treat experimental diabetic neuropathy, retinopathy and nephropathy.29 Aminoguanidine prevents the AGEs formation possibly through the trapping of reactive carbonyl groups in glycating agents such as methylglyoxal, glyoxal and 3-deoxyglucosone.30 However, this drug also has a variety of other effects including acting as an antioxidant31 and a cytoprotective agent by increasing total sulphydryl (SH) content.32 Further, it inhibits aldose reductase33 and can chelate metal ions.34 Perhaps most importantly, at concentrations required for AGE prevention, aminoguanidine inhibits all isoforms of NOS, with iNOS being much more sensitive than nNOS and eNOS.35,36 These effects may account for the discrepant effects of aminoguanidine and ALT-711 on nNOS mRNA expression (see below).
The results of clinical trials with aminoguanidine in diabetic states have been equivocal. Because of its relative lack of specificity and some concern about adverse effects, attention has shifted to newer agents such as ALT-711. ALT-711 is a class of novel cross-link breaker that has been shown to cleave preformed AGEs37 and in the diabetic milieu to reduce AGE accumulation.38 In agreement with previous reports,19,26 in this study, we observed that the late administration of ALT-711 reduced the serum AGEs to control levels. Interestingly, and in contrast to the results with aminoguanidine, restoration of nNOS expression was observed with ALT-711 only at the protein level (immunohistochemistry and Western blot analysis) and not at the level of mRNA.
Other investigators, working in other organs, have generally reported results that are consistent with our findings. Aminoguanidine, but not N-nitro-l-arginine methyl ester (L-NAME), prevents the depletion in the retinal nNOS-expressing neurons, implying that the AGE specific effect may be more important for this effect.39 On the other hand, treatment with ALT-711, but not with aminoguanidine, reversed NOS depletion in the penis of diabetic rats; in addition, serum and penile tissue AGE levels were apparently only reduced with ALT-711, questioning the effectiveness of the aminoguanidine regimen in achieving therapeutic levels.26 In contrast, a recent study reported that after 8 weeks of STZ-induced diabetes, a significant increase in nNOS activity was observed.40 However, the authors did not measure AGE levels and so mechanistic interpretation of these results is difficult.
Given the non-specific nature of aminoguanidine, the results of the experiments with ALT-711 lend confidence to our conclusion that AGE signalling is involved in the suppression of enteric nNOS in diabetes. Further, if ALT-711 is a truly specific anti-AGE agent, its differential effects on nNOS mRNA and protein expression assume significance for two reasons. Firstly, suppression of nNOS gene transcription in diabetes may be due to factors other than activation of the AGE-RAGE pathways, whereas the latter may be more important in post-transcriptional or post-translational modifications. However, the precise reason for these differences is not understood, and further studies will be required to examine these possibilities.
Finally, our experiments do not suggest an increase in RAGE expression in diabetes, in contrast to what has been reported in the literature.41 This may be due to differences in the organ being studied or in the duration of diabetes. Nevertheless, even though enteric RAGE is not upregulated, it is probably expressed at a level adequate to mediate the consequences of elevated AGE levels in diabetes.
In conclusion, our results provide evidence for the accumulation of AGEs and associated nNOS suppression in the duodenum of diabetic induced rats and provide strong support for a cause and effect between these phenomena. These results suggest that countering the AGE–RAGE signalling pathway may be a major target for diabetes related gastrointestinal complications, particularly those that arise from a reduction in nNOS expression.
- 1GI symptoms in diabetes mellitus are associated with both poor glycemic control and diabetic complications. Am J Gastroenterol 2002; 97: 604–11., , , , , .Direct Link: