The vasorelaxant effect of hydrogen sulfide is enhanced in streptozotocin-induced diabetic rats
- First Online:
- Cite this article as:
- Denizalti, M., Bozkurt, T.E., Akpulat, U. et al. Naunyn-Schmiedeberg's Arch Pharmacol (2011) 383: 509. doi:10.1007/s00210-011-0601-6
- 198 Views
Hydrogen sulfide (H2S) is an endogenous gas which has potent relaxant effect in vascular and nonvascular smooth muscles. In the present study, we have investigated how streptozotocin (STZ)-induced diabetes affected the relaxant effect of H2S in rat isolated thoracic aorta and mesenteric and pulmonary arteries. Diabetes was induced by IV injection of STZ (35 mg/kg). Insulin treatment was applied by using insulin implants. At the end of 4 and 12 weeks, the thoracic aorta and mesenteric and pulmonary arteries were isolated, and the relaxation responses to sodium hydrogen sulfide (NaHS), diazoxide, and acetylcholine were evaluated. The mRNA and protein levels of H2S-synthesizing enzymes were also examined by RT-PCR and Western Blot. The relaxation response to NaHS in the arteries isolated from both 4 and 12 week-diabetic rats was increased when compared with that obtained from the control group. Glibenclamide inhibited the relaxation response to NaHS in the arteries isolated from either diabetic or non-diabetic group of rats. Concurrent treatment of insulin to STZ-injected rats prevented the potentiation of the relaxant effect of NaHS in the arteries. However, acetylcholine and diazoxide-induced relaxation responses were not altered in diabetic group of rats. The mRNA and protein levels of H2S-synthesizing enzymes were also not altered in diabetic rats. STZ-induced experimental diabetes in rats resulted in the potentiation of the relaxation response to H2S in vascular tissues. The potentiated relaxation to H2S in diabetic arteries may play a role in vascular complications frequently seen in severe diabetes.
KeywordsHydrogen sulfideDiabetesRelaxationVascular smooth muscleIn vitro
Hydrogen sulfide (H2S) is known as a toxic gas which is endogenously synthesized from l-cysteine. In mammalian tissues, the biosynthesis of H2S is catalyzed by the pyridoxal-5-phosphate-dependent enzymes, cystathionine-ß-synthetase (CBS), and cystathionine-γ-lyase (CSE; Wang 2002). CBS is highly expressed in the central nervous system, liver, and kidney, whereas CSE is expressed mostly in the vascular system including thoracic aorta, tail artery, mesenteric artery, pulmonary artery, and portal vein and also in the liver, kidney, small intestine, and pancreas (Abe and Kimura 1996; Hosoki et al. 1997; Bao et al. 1998; Zhao et al. 2001; Kimura 2002). Furthermore, H2S is detected in rat, human, and bovine brain tissues (50–160 μM) and also in rat and human blood (10–100 μM; Warenycia et al. 1989; Abe and Kimura 1996; Richardson et al. 2000; Zhao et al. 2001).
H2S has potent relaxant effect in vascular and nonvascular smooth muscles (Hayden et al. 1989; Hosoki et al. 1997; Zhao et al. 2001; Teague et al. 2002). It has been shown that H2S produces relaxation response in isolated arteries and veins of rat in vitro and lowers rat blood pressure in vivo (Zhao et al. 2001; Zhao and Wang 2002). Endothelium is not involved in the vasorelaxant effect of H2S, whereas ATP-sensitive potassium channels are responsible for the H2S-induced relaxation in vascular smooth muscle (Hosoki et al. 1997; Zhao et al. 2001). The mechanism of the relaxation response in nonvascular smooth muscles, i.e., guinea-pig ileum and rat uterine, has not been identified yet (Sidhu et al. 2001; Teague et al. 2002).
The involvement of H2S in the pathological states of the cardiovascular system has been suggested, although its physiological role is still not known. The impaired biosynthesis of H2S has been shown in experimental models of spontaneous and hypoxic pulmonary hypertension and hemorrhagic shock in rats (Chunyu et al. 2003; Mok et al. 2004; Yan et al. 2004). Another disease state with the potential role of H2S is diabetes mellitus, since this disease is associated with altered synthesis of endogenous mediators that contribute to the impairment of vascular reactivity (De Vriese et al. 2000). Increased formation of H2S and expression of endogenous H2S-synthesizing enzymes have been demonstrated in liver and pancreas of streptozotocin (STZ)-induced diabetic rats (Yusuf et al. 2005). However, possible accompanying changes in the functional effects of H2S have not been studied in this model of experimental diabetes. In the present study, we have investigated how STZ-induced diabetes in rats affected the relaxant effect of H2S in vascular smooth muscles. For this purpose, the relaxation response elicited by the H2S donor sodium hydrogen sulfide (NaHS) was evaluated in thoracic aorta and mesenteric and pulmonary arteries isolated from non-diabetic, 4, and 12 week-diabetic and insulin-treated diabetic rats.
The study was approved by Hacettepe University Animal Ethics Committee (No: 2008/28-3). Male Sprague–Dawley albino rats (200–250 g) were used in the present study.
Rats were divided into non-diabetic (control), 4- and 12 week diabetic, and insulin-treated diabetic group of rats. Diabetes was induced by tail vein injection of STZ (35 mg/kg) dissolved in 0.1 M citrate buffer (pH 4.5), whereas the control group were injected citrate buffer alone by the same route. Glucose (10%) was added into the drinking water of STZ-treated rats during the first 24 h to reduce hypoglycaemic phase following STZ injection. Insulin-treated diabetic group of rats received insulin implants after STZ injection. Insulin implants were placed under the skin in the abdomen under ketamine/xylazine anesthesia (45/5 mg/kg), and every implant released insulin at a dose of 2 Units in 24 h.
Blood glucose levels of the rats in each group were measured using blood glucose test strips after 72 h of STZ injection, and blood glucose levels over 250 mg/dl were accepted as diabetic. Measurement of blood glucose was repeated after 4 and 12 weeks of treatment in each group.
The rat thoracic aorta and mesenteric and pulmonary arteries were isolated, and rings of approximately 2–3 mm in length were prepared. Subsequently, the rings were mounted under a resting tension of 1.5 g (thoracic aorta) and 1 g (mesenteric and pulmonary arteries) in organ baths filled with Krebs–Henseleit solution. Tissues were equilibrated for 1.5 h and washed by Krebs–Henseleit solution every 15 min before each experimental procedure. Isometric changes in tension were recorded with an isometric force transducer (MAY95-transducer data acquisition system). The composition of the Krebs–Henseleit solution (in millimolars) was NaCl, 113; KCl, 4.7; MgSO4, 1.2; CaCl2, 2.5; KH2PO4, 1.2; NaHCO3, 25.0; and glucose, 11.6 and this was gassed with 95% O2–5% CO2 at 37°C and pH 7.4.
NaHS is used as an H2S donor, and its aqueous solution was introduced directly into the organ bath by an automated pipette. NaHS dissociates to Na+ and HS- in aqueous solution and then HS- associates with H+ to form H2S (Hosoki et al. 1997).
At the beginning of each experiment, KCl (60 mM)-induced contractions were elicited in thoracic aorta and mesenteric and pulmonary arteries isolated from non-diabetic (control), diabetic, and insulin-treated diabetic group of rats. After a 30-min washout period, the arteries were precontracted by 1–3 μM phenylephrine (60–80% of the maximum) and then the relaxation response to NaHS (0.1–3 mM) was evaluated by cumulative addition to the organ bath. In some experiments, NaHS (0.1–1 mM) response was elicited in the presence of ATP-sensitive potassium channel blocker glibenclamide (10 μM). The relaxation response to KATP activator diazoxide (100 nM to 100 μM) is also evaluated in 4 week-diabetic rats. Each concentration–response curve to NaHS was obtained in individual preparations.
The functionality of the endothelium was also determined after obtaining the NaHS-induced relaxation response. For this purpose, after an hour of washout period, the arteries were again precontracted by phenylephrine (1–3 μM), and then the concentration-dependent relaxation response to acetylcholine (10 nM to 0.1 mM) was elicited by cumulative addition to the organ bath.
RNA and protein expression studies
The time-course expression of CBS and CSE were determined at the RNA and protein levels in the intact thoracic aorta by reverse transcriptase coupled quantitative reverse transcription-polymerase chain reaction (RT-PCR) and Western Blot experiments (CSE only), respectively. RNA isolation from the aortic samples is achieved as described previously (Yuzbasioglu et al. 2010). Briefly, tissue samples were rapidly disrupted using a bead-beater in a 2-ml screw-cap tube containing ceramic beads and 1 ml of the Trizol reagent (Invitrogen) and simultaneous RNA extraction was accomplished upon manufacturer’s recommendations. The RNA integrity and the quality of the extracted total RNA was assessed via denaturing agarose gel electrophoresis and UV absorbance measurements. cDNA synthesis from 1 μg of total RNA was accomplished by using Improm II (Promega) reverse transcriptase following manufacturer’s protocols. The amplification f the target genes [Cbs and Cth (for CSE)] was done by using the SYBR Green technique. Equal amount of cDNA was used for the real-time amplification of the target genes using Jumpstart SYBR Green mix (Sigma-Aldrich) on a Rotorgene 6000 (Corbett Life Science, Australia) fluorometric PCR instrument. The primer pairs and the reaction conditions were designed so as to include at least one intron to avoid misamplification of any contaminated genomic DNA. The sequences of the primer pairs and the reaction conditions are available upon request by e-mail. The amplification products were double-checked for specificity by using meting curve analyzes and agarose gel electrophoresis. The expression of β-actin was used to normalize the expression of the gene of interest. The relative expressions of the diabetic rats (n = 6) are further normalized to the median expression levels of the control samples (n = 5) to achieve relative fold change values. One-way analysis of variance (ANOVA) test is utilized to assess the significance of the expression levels (p < 0.05).
Protein extraction from the aortic tissue samples (for control and diabetic groups, n = 3 each) were performed as described previously (Kocaefe et al. 2010). The total protein concentrations were determined by bicinchoninic acid protein assay (Pierce, Rockford, IL), and 80 μg of total protein was loaded onto the 12% sodium dodecyl (lauryl) sulfate-polyacrylamide gel electrophoresis gel. Following the transfer of proteins onto the nitrocellulose membrane, ponceau-S staining was performed to verify equal loading. The probing of the membrane was achieved using rabbit polyclonal anti Cth (CSE; Abnova) antibody. Signal generation was accomplished by using appropriate horseradish–peroxidase-conjugated secondary antibody using chemiluminescence detection system (SuperSignal West Femto maximum sensitivity substrate—thermo-scientific).
The relaxation response to NaHS, diazoxide, and acetylcholine was expressed as the percentage of phenylephrine-induced precontraction, and phenylephrine contraction was expressed as the percentage of KCl (60 mM)-induced contraction. Data are represented as mean±standard error of mean (SEM). Statistical analysis was done by ANOVA/Newman–Keuls, and Student’s t test by using GraphPad Prism4 software. P < 0.05 was considered as significant.
Drugs and solutions
Sodium hydrogen sulfide, phenylephrine hydrochloride, diazoxide, acetylcholine hydrochloride, papaverine hydrochloride, streptozotocin, and glibenclamide were purchased from Sigma (St. Louis, MO). Insulin implants (Linplant®) were purchased from Linshin Canada, Inc. All drugs were prepared in distilled water except glibenclamide which was dissolved in dimethyl sulfoxide (DMSO), diazoxide in methanol, and streptozotocin which was dissolved in citrate buffer. DMSO and methanol at concentrations that match its dilutions did not alter the control responses.
Effects of streptozotocin and insulin treatment on plasma glucose levels and body weights of the rats
Blood glucose (mg/dl)
Body weight (g)
Control (4 weeks)
102.33 ± 5.28
253.33 ± 12.5
STZ (4 weeks)
419.82 ± 18.19*
232.73 ± 10.54
STZ + insulin (4 weeks)
79.20 ± 12.57**
255.00 ± 19.79
Control (12 weeks)
106.71 ± 7.59
311.66 ± 14.70
STZ (12 weeks)
321.55 ± 5.98*
261.66 ± 10.46*
STZ + insulin (12 weeks)
81.16 ± 22.28**
308.57 ± 26.58
Effects of streptozotocin and insulin treatment on phenylephrine-induced contraction response
Control (4 weeks)
84.43 ± 3.97
100.47 ± 7.31
82.94 ± 5.00
STZ (4 weeks)
77.59 ± 5.61
87.17 ± 1.53
82.45 ± 7.76
STZ + insulin (4 weeks)
91.09 ± 2.28
105.40 ± 9.81
79.52 ± 4.51
Control (12 weeks)
84.47 ± 4.86
96.52 ± 7.24
83.52 ± 6.08
STZ (12 weeks)
86.55 ± 9.84
113.26 ± 10.42
80.84 ± 8.15
STZ + insulin (12 weeks)
81.93 ± 6.85
102.12 ± 10.67
80.32 ± 5.55
The relaxant effect of H2S has been shown in various tissues including the vascular and nonvascular smooth muscles (Hosoki et al. 1997; Zhao et al. 2001; Teague et al. 2002; Zhao and Wang 2002). In the present study, we have also shown that H2S donor NaHS elicited relaxation response in isolated rat thoracic aorta and mesenteric and pulmonary arteries. Furthermore, glibenclamide inhibited this relaxation response confirming the previous studies that ATP-sensitive potassium channels are involved in H2S-induced vasorelaxation (Hosoki et al. 1997; Zhao et al. 2001; Teague et al. 2002; Zhao and Wang 2002). We have used NaHS as H2S donor in the present study. It has been reported that ∼30% of NaHS in solution is present as free H2S gas (Zhao and Wang 2002). In addition, this ratio is also dependent on the temperature of the medium. The 1 mM NaHS yields approximately 333 μM of H2S at 20°C and 185 μM at 37°C (Dombkowski et al. 2004; Lee et al. 2007). According to the temperature (37°C) used in our experiments, the maximum relaxation response observed at 1 mM NaHS corresponds to ∼185 μM H2S, and this concentration is comparable to that of its physiological levels (Warenycia et al. 1989; Abe and Kimura 1996; Richardson et al. 2000; Zhao et al. 2001).
In the present study, we have investigated the effect of diabetes in H2S-induced relaxation response in isolated arteries since the biosynthesis of H2S is altered in experimental models of this disease (Yusuf et al. 2005). Diabetes was induced by STZ injection to rats and evidenced by the increase in blood glucose level. We have demonstrated that the relaxation response to H2S donor NaHS was potentiated in the thoracic aorta and mesenteric and pulmonary arteries isolated from either 4- or 12 week-diabetic rats. Furthermore, the potentiated relaxation to NaHS was also inhibited by KATP blocker glibenclamide, and the potentiation was not observed in the arteries isolated from insulin-treated diabetic rats. Yusuf et al. (2005) have reported that STZ-induced diabetes in rats was associated with a marked increase in pancreas and liver H2S synthesis, and these changes were also reversed by concurrent insulin treatment (Yusuf et al. 2005). Plasma H2S concentration was found similar in both STZ-induced diabetic and non-diabetic rats (Yusuf et al. 2005). On the other hand, it has been shown that exposure to H2S reduces insulin secretion in β cells by activating KATP channels (Ali et al. 2007). This study proposed that the elevated level of H2S-synthesizing enzymes in pancreas may play an important role in the pathophysiology of diabetes by modulating the insulin release. However, the effect of H2S on the function of vascular smooth muscle has not been identified, and our study is the first to demonstrate the enhanced relaxant effect of H2S in STZ-induced diabetic arteries.
In another study, Brancaleone et al (2008) have shown that endogenous H2S production was impaired in aortic tissue homogenates along with the progression of the disease in non-obese diabetic NOD mice. However, Cbs and Cth genes and protein expressions and thus, the level of H2S-synthesizing enzymes were not altered in STZ-induced diabetic rat aorta in the present experiments. The relaxation to exogenous H2S was also evaluated in the aorta isolated from NOD mice, and the potentiation of the relaxation response was reported along with the disease severity. Thus, the enhanced vascular relaxant effect of H2S demonstrated in STZ-induced diabetic rat arteries in the present experiments seems to be comparable to that elicited in the isolated aorta from NOD mice with severe disease despite of the differences in diabetic model and species. Furthermore, the potentiated relaxant effect of H2S in diabetic vascular smooth muscles may be regardless of the level of H2S or synthesizing enzymes since plasma H2S level was found unchanged in STZ-induced diabetic rats while found reduced in NOD mice (Yusuf et al. 2005; Brancaleone et al. 2008). This finding may be considered as a protective mechanism against vascular complications of diabetes since increased relaxation to H2S may compensate the impaired response to endogenous mediators and may reduce the vascular tone in this metabolic disease.
Endothelium-dependent vasodilatation is generally used as a parameter to assess the endothelial function of arteries in different pathophysiologies. Conflicting results indicating normal, impaired, or enhanced endothelium-dependent responses in different models of diabetes have been reported (De Vriese et al. 2000). In the present study, we did not demonstrate any change in the acetylcholine-induced relaxation in STZ-induced diabetic arteries. Thus, enhanced relaxation to H2S in the present study does not seem to be mediated through altered endothelial function.
KATP channels which are reported to be responsible for the vasorelaxant effect of H2S are distributed in a variety of tissues including cardiomyocytes, smooth muscle cells, and pancreatic β-cells (Fujita and Kurachi 2000). The impairment of vascular KATP-channel function was observed in coronary microvessels obtained from diabetic patients (Weintraub 2003). The present finding that NaHS-induced relaxation response was enhanced in diabetic arteries may be due to the increased sensitivity of vascular KATP channels to H2S or to increased expression of these channels in diabetic rats. Thus, we have evaluated the effect of KATP-channel opener diazoxide in diabetic arteries in the present study. Diazoxide-induced relaxation response was not altered in aorta and mesenteric and pulmonary arteries isolated from 4-weeks of STZ-induced diabetic rats. This finding indicates that either the increased sensitivity or increased expression of vascular KATP channels is not involved in the enhanced relaxation response elicited by NaHS in isolated diabetic arteries. There are possibly additional mechanisms since KATP channels are not the only signaling mechanism in the vasorelaxant effect of H2S. Lee et al. (2007) have shown that H2S decrease the intracellular pH of vascular smooth muscle cells which is associated with its vasorelaxant effect. Thus, altered tissue acidification in diabetes may contribute to the enhanced vasorelaxant effect of H2S. It has been demonstrated that elevated extracellular glucose concentrations regulates the Na+/H+ exchanger activity in vascular smooth muscle cells (Williams and Howard 1994). Furthermore, increased activity of Na+/H+ exchanger have been proposed for the altered regulation of intracellular pH in mesenteric vasculature in diabetes (Jandeleit-Dahm et al. 2000). Therefore, altered intracellular pH of vascular smooth muscle cells in diabetes may cause the enhanced vasorelaxant effect of H2S in diabetic arteries. It has also been reported that H2S inhibits cytochrome c oxidase which causes energy deficit and intracellular acidosis to induce relaxation, and thus, diabetes as a metabolic disease may also interfere with the activity of cytochrome c oxidase and leads to the potentiation of the vascular relaxation to H2S (Kiss et al. 2008).
In conclusion, the present study demonstrated that experimental diabetes induced by STZ injection to rats resulted in enhanced relaxation response to H2S in thoracic aorta and mesenteric and pulmonary arteries. However, STZ-induced diabetes model does not affect the expression of the Cbs and Cth genes in the rat aorta and the enhanced H2S-induced relaxation response is not due to any alterations in H2S synthesizing enzymes. Since H2S is an endogenously synthesized gaseous mediator in the vascular tissues, its enhanced relaxant effect in diabetic arteries may indicate its role in the vascular complications of this disease.
This study is supported by Hacettepe University Research Foundation.
Some of this work was presented as a poster in the “Federation of the European Pharmacological Societies (EPHAR) 2008 Congress”, 13–17 July 2008, Manchester, UK.
Conflict of interest