The aim of the present study was to investigate the effect of hydrogen sulfide (H2S) signaling by nitric oxide (NO) in isolated rat aortas and cultured human umbilical vein endothelial cells (HUVECs). Both administration of H2S and NaHS, as well as endogenous H2S, reduced NO formation, endothelial nitric oxide synthase (eNOS) activity, eNOS transcript abundance, and l-arginine (l-Arg) transport (all P < 0.01). The kinetics analysis of eNOS activity and l-Arg transport showed that H2S reduced Vmax values (all P < 0.01) without modifying Km parameters. Use of selective NOS inhibitors verified that eNOS [vs. inducible NOS (iNOS) and neuronal NOS (nNOS)] was the specific target of H2S regulation. H2S treatment (100 μmol/l) reduced Akt phosphorylation and decreased eNOS phosphorylation at Ser1177. H2S reduced l-Arg uptake by inhibition of a system y+ transporter and decreased the CAT-1 transcript. H2S treatment reduced protein expression of eNOS but not of nNOS and iNOS. Pinacidil (KATP channel opener) exhibited the similar inhibitory effects on the l-Arg/NOS/NO pathway. Glibenclamide (KATP channel inhibitor) partly blocked the inhibitory effect of H2S and pinacidil. An in vivo experiment revealed that H2S downregulated the vascular l-Arg/eNOS/NO pathway after intraperitoneal injection of NaHS (14 μmol/kg) in rats. Taken together, our findings suggest that H2S downregulates the vascular l-Arg/NOS/NO pathway in vitro and in vivo, and the KATP channel could be involved in the regulatory mechanism of H2S.
- nitric oxide
- nitric oxide synthase
- l-arginine transport
hydrogen sulfide (h2s) is a well-known pungent gas, and its toxicology has been extensively studied (40). Recently, H2S has been found to be endogenously generated in various mammalian tissues by 2 pyridoxal-5′-phosphate-dependent enzymes, such as cystathionine β-synthase (CBS), and cystathionine γ-lyase (CSE) (6), with l-cysteine used as a main substrate. The gas may be a functional regulator in the nervous and cardiovascular systems (21). In the cardiovascular system, CSE may be a key enzyme by which H2S is generated (50). H2S causes a dose-dependent relaxation of the isolated rat aorta (16, 50) and mesenteric arterioles (4), induces transient hypotension in anesthetized rats in vivo (50), reduces proliferation of vascular smooth muscle cells (VSMCs) (8, 46), and exerts a negative inotropic action of the heart in vivo and in vitro (13). Significantly, there is evidence linking the endogenous CSE/H2S pathway to the pathogenesis of arterial hypertension (45, 52), pulmonary hypertension induced by hypoxia (5), septic and endotoxin shock (18), hemorrhagic shock (27), ischemic cardiac disease (2, 11), ischemic brain damage (33), and hyperdynamic circulation in cirrhosis (10). Growing evidence suggests that in addition to nitric oxide (NO) and carbon monoxide (CO) (12), endogenous H2S may be a novel cardiovascular gasotransmitter. As an endogenous ligand, H2S directly opens the KATP channel in VSMCs and relaxes vascular smooth muscles in an endothelium-independent manner (50), which is different from the effect of NO and CO.
Interestingly, the administration of sodium nitroprusside (SNP), an NO donor, potentiates the vasorelaxation effects induced by H2S (16) and upregulates H2S production in rat vascular tissues in a concentration-dependent manner (48). In contrast, pretreatment with H2S attenuates the vasorelaxant effect induced by SNP in aortic rings (49). In rats with septic and endotoxin shock, the plasma level of H2S is positively correlated with that of NO (18). In the brain, H2S acts as an endogenous peroxynitrite scavenger (42). All of these results suggest that NO and H2S signaling are interrelated. Of note, the expression of the H2S-generating enzyme CSE had been observed in VSMCs (50), but not in the vascular endothelium, suggesting that H2S is generated in vascular tissues mainly from VSMCs. Whether and how endogenous H2S from VSMCs affect the endothelial NO production pathway is unknown. In the present study, we examined the effects of H2S on NO production, endothelial NO synthase (eNOS) activity, and expression, and l-Arginine (l-Arg, substrate of NOS) transport in isolated rat aortas and cultured human umbilical vein endothelial cells (HUVECs) to explore the regulatory role of H2S in the vascular l-Arg/NOS/NO pathway.
MATERIALS AND METHODS
All animal care and experimental protocols complied with the Animal Management Rules of the Ministry of Health of the People's Republic of China (documentation no. 55, 2001, Ministry of Health of P.R. China) and the Guide for the Care and Use of the Laboratory Animals of the First Hospital, Peking University. Male Sprague-Dawley (SD) rats (250–300 g) were provided by the Animal Department, Health Science Center of Peking University. All animals were maintained on normal rat chow with unrestricted water and on a 12:12-h light-dark cycle.
HUVECs (926 cell line) were purchased from the American Type Culture Collection (Manassas, VA). Tris-base, Dowex AG 50 w-x8 (Na+), NADPH, tetrahydrobiopterin (BH4), calmodulin, flavin mononucleotide (FMN), flavin adenine dinucleotide (FAD), l-Arg, l-canavanine, spermidine, pinacidil, glibenclamide, N-ethylmaleimide (NEM), DL-propargylglycine (PAG), l-cysteine, pyridoxal-5′-phosphate, and sodium hydrosulfide (NaHS) were purchased from Sigma (St. Louis, MO). l-[3H]Arg (1.5 TBq/mmol) was from NEN Life Science Products (Boston, MA); and eNOS antibody (SC-654), P-eNOS(Ser1177)-R antibody (SC-12972), iNOS antibody (SC-7271), nNOS antibody (SC-648), phospho-Akt1/2/3 (Ser473) antibody (Sc-7985R), Akt antibody (SC-8312), and β-tubulin antibody (SC-5274) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA).
H2S-saturated solution (0.09 mol/l at room temperature) was made by bubbling with pure H2S gas (offered by Beijing XianHeYu, China) and stored at −70°C.
The specific primers used for sample loading calibration were synthesized by SBS (Beijing, China) and were as follows: eNOS-S, 5′GTTTGTCTGCGGCGATGT-3′; eNOS-A, 5′AAGAAACAGGAAGCGGGTG-3′; cationic amino acid transporter-1 (CAT-1)-S, 5′-TGAAGACAGGGAGTGAGGG-3′; CAT-1-A, 5′-TAGACAAACGGTAGCAGT-3′; GAPDH-S, 5′GGATTTGGTCGTATTGGG3′; and GAPDH-A, 5′GGAAGATGGTGATGGGATT3′. Other chemicals and reagents were of analytical grade.
Organ culture of aortas in vitro.
At the selected time, recipient rats were anesthetized with pentobarbital sodium (30 mg/kg ip) and then decapitated quickly. The entire thoraco-abdominal aorta was quickly removed and placed in cold, sterile PBS (4°C). The organ culture of aortic tissues was carried out as previously described (20), with some modification. Briefly, aortas were cleaned of adherent adipose tissues and collateral vessels under sterile conditions. The aortic tissues were placed in ice-cold PBS, cut into pieces of ∼1 × 1 mm2, and transferred to 24-well plates, containing 1 ml of culture medium [RPMI 1640 with 0.5% (wt/vol) BSA] per well covering the tissue slices completely. Tissues were cultured in a shaking bath (60 beats/min) at 37°C, in an atmosphere of 95% O2-5% CO2. NaHS, the donor of H2S, or H2S gas buffer were added to the wells at a final concentration of 50 μmol/l, then incubated for 2, 4, and 6 h, respectively. For other groups, 1, 10, 100, and 1,000 μmol/l NaHS or H2S gas buffer with or without other drugs (10 mmol/l l-cysteine; 2 mmol/l pyridoxal-5′-phosphate; 2 mmol/l DL-propargylglycine; 10 μmol/l pinacidil, and 10 μmol/l glibenclimide) were added to the wells, which were incubated for 2 h. At the end of incubation, the NO production (content of nitrate and nitrite in medium), tissue NOS activity, and [3H]l-Arg transport were measured. The activity of lactate dehydrogenase (LDH) in the culture medium was measured using a LDH-cytotoxic test kit (44) to determine the tissue viability.
H2S assay by sulfide-sensitive electrode.
Measurement of H2S concentration in culture medium involved use of a sulfide-sensitive electrode (Model 9616; Orion Research, Beverly, MA). In brief, sulfide antioxidant buffer (SAOB) was added into standard sulfide solution or blood samples at a ratio of 1:1 and then stirred thoroughly. Electrodes were rinsed in distilled water, blotted dry, and placed into standards or samples. When a stable reading was displayed, the millivolt value was recorded. The H2S concentration was calculated against the calibration curve of the standard S2− solution (50).
Confluent HUVECs (926 cell line) were put into medium 199 containing 5 mmol/l d-glucose, 10% FCS, 5 mmol/l glutamine, and 0.03 mg/ml gentamycin at 37°C in a 5% CO2 atmosphere (43). Cell viability was determined by trypan blue exclusion assay and by measurement of the activity of LDH in the culture medium (44). Cell viability remained at more than 95% throughout the experiments.
H2S (50 μmol/l) was added after changing to serum-free medium; cells were incubated for 2 h at 37°C in a 95% air-5% CO2 atmosphere, and eNOS enzyme kinetics and l-Arg transport dynamics were measured. For other groups, 100 μmol/l of H2S was added, and cells were incubated for 4 h before gene expression of CAT-1 and eNOS were analyzed by RT-PCR. The protein expression of cellular phospho-Akt and total Akt was detected by Western blot analysis after incubation with H2S (500 μmol/l), pinacidil (10 μmol/l), or glibenclamide (10 μmol/l) for 0.5 h; phospho-eNOS and eNOS protein expression was measured after a 2-h incubation. H2S (100, 500, and 1,000 μmol/l) and/or pinacidil (10 μmol/l) and glibenclamide (10 μmol/l) were added after incubation for 6 h; then eNOS, iNOS and nNOS protein expression was analyzed.
Assay of nitrate plus nitrite.
l-Arg (200 μmol/l) was added to the organ and cell culture medium, which was incubated for different times, according to the experiment protocol; the media were then collected. The assay was performed in a standard flat-bottomed 96-well polystyrene microtiter plate, with PBS used as a control. An amount of 50 μl of nitrate reductase and β-NADPH was added to each well, for final concentrations of 300 U/l and 25 μmol/l, respectively. The plate was incubated at room temperature for 3 h. Excess β-NADPH was consumed by the addition of 50 μl of PBS containing l-glutamic dehydrogenase, α-ketoglutaric acid, and NH4Cl, (for final concentrations of 500 U/l, 4 mmol/l, and 100 mmol/l, respectively) followed by 10-min incubation at 37°C. The nitrite concentration was then measured by the addition of 50 μl Griess reagents, and the absorbance was read at 540 nm by use of a plate reader after 10-min incubation at room temperature (14).
Assay of NOS activity and eNOS enzyme kinetics.
Aortic samples were fast frozen and ground into a fine powder in liquid nitrogen by use of a mortar and pestle before sonicating in 5 volumes of homogenization buffer (50 mmol/l Tris·HCl, pH 7.2, 0.1 mmol/l EDTA, 0.1 mmol/l EGTA, 0.1% l-mercaptoethanol, 10 μg/ml leupeptin, 5 μg/ml pepstatin A, 1 mmol/l PMSF, and 0.5 mmol/l DTT). NOS activity was measured by monitoring the conversion of l-[3H]Arg to l-[3H]citrulline (23). For routine assay of total NOS (tNOS), 0.1 mg of aortic tissue protein was incubated in a reaction buffer containing 1 mmol/l l-[3H]Arg (containing 1 mCi [3H]), 1 mmol/l NADPH, 300 IU/ml calmodulin, 5 μmol/l BH4, 2 mmol/l CaCl2, 10 μmol/l FAD (31), and 10 μmol/l FMN for 30 min at 37°C. For assay of iNOS activity, the reaction buffer was Ca2+-free. The reaction was stopped by adding 50 μl of ice-cold Tris·HCl buffer (pH 5.5) containing 1 mmol/l l-citrulline, 1 mmol/l EGTA, and 1 mmol/l EDTA. The reaction substance was applied to a 1 ml column of Dowex AG50WX-8 (Na+) to absorb l-[3H]Arg. The NOS activity was then quantified by determining the radioactivity on liquid scintillation spectroscopy (Beckman, LS6500) and presented as nmol l-[3H]citrulline·min−1·mg protein−1. eNOS activity was calculated as the difference between total NOS activity and iNOS activity.
The measurement of eNOS enzyme kinetics in HUVECs was as previously described (24), with minor modification. In brief, the cells were washed twice with ice-cold buffer containing (in mmol/l) 100 NaCl, 25 NaH2PO4, and 80 Na2HPO4, at pH of 7.5. The cells were pelleted and resuspended in ice-cold 50 mmol/l Tris·HCl buffer (pH 7.4) containing l mmol/l EDTA, 5 mmol/l mercaptoethanol, 10 μg/ml pepstatin A, 10 μg/ml leupeptin, 90 μg/ml PMSF, and 1 μmol/l BH4. The cells were disrupted by repeated freeze-thawing in liquid nitrogen. The cell lysate (50 μl) was added to 50 μl of buffer as described above, except with different concentrations (10, 20, 40, 80, 160, and 320 μmol/l) of l-[3H]Arg. The other reactions of eNOS activity assay were further processed as described above.
Assay of l-Arg uptake in aortic tissues and l-Arg transport kinetics in HUVECS.
l-Arg uptake and transport kinetics analysis were performed by measured intracellular incorporation l-[3H]Arg. On completion of aortic incubation, the aortic tissues were washed 3 times with ice-cold Krebs-Henseleit (KH) buffer (in mmol/l): 131 NaCl, 5.6 KCl, 25 NaHCO3, 1 NaH2PO4, 5 d-glucose, 20 HEPES, 2.5 CaCl2, 1 MgCl2, at pH 7.4. Aortic l-Arg uptake was measured by incubation with KH buffer containing 100 μmol/l l-Arg and 10 μCi/ml l-[3H]Arg at 37°C for 15 min. Aortic tissues were washed with ice-cold 10 mmol/l unlabeled l-Arg in PBS, then lysed with 200 μl methanoic acid. Total radioactivity of l-[3H]Arg was measured by liquid scintillation counting (37).
Transport kinetics of l-Arg into endothelial cells was measured by the method of Leoncini et al. (25), with minor modifications. Briefly, kinetic experiments were performed in cells incubated for 15 min with KH buffer containing l-Arg (10 to 320 μmol/l) and 2 μCi/ml l-[3H]Arg. Transport was terminated by removing the media and washing the cells three times with ice-cold 10 mmol/l unlabeled l-Arg in PBS. Cells were lysed with 0.5% Triton X-100 in 0.5 mol/l NaOH, and radioactive activity [3H]l-Arg taken by cells was assayed by liquid scintillation counting.
RT-PCR assay of gene expression of eNOS and CAT-1 in HUVECs.
The mRNA levels of eNOS and CAT-1 were measured by RT-PCR, as described previously by our laboratory (19). Total RNA from cells was extracted with Trizol reagent (Applygen Technologies, Beijing, China). RT-PCR was performed in a total volume of 25 μl. After denaturing at 95°C for 5 min, we ran PCR at 94°C for 30 s, 61°C for 30 s, and 72°C for 40 s for 30 cycles. The PCR products were separated on a 1.5% agarose gel and stained with ethidium bromide. The optical density of the bands of eNOS cDNA (553 bp) and CAT-1 cDNA (250 bp) was measured by use of the Gel Documentation System (Bio-Rad, Hercules, CA). The PCR products were amplified again with the human GAPDH primers at 94°C for 30 s, 55°C for 30 s, and 72°C for 30 s for 20 cycles, and the optical density of the GAPDH band (205 bp) was measured. The ratio of eNOS or CAT-1 to GAPDH was considered as the relative amount of eNOS or CAT-1 gene expression.
Western blot analysis.
Protein isolation and Western blot analysis were carried out as previously described (3). Briefly, a cellular lysate was prepared from the HUVECs and separated by SDS-PAGE (on 7.5% gels). The proteins were transferred to nitrocellulose membranes (Schleicher and Schuell, Dassel, Germany) by electroblotting (Bio-Rad). Proteins were detected with polyclonal anti-eNOS, anti-P-eNOS, anti-iNOS, anti-nNOS, anti-Akt, and anti-P-Akt antibodies and a monoclonal anti-tubulin antibody. After incubation with the appropriate secondary antibodies conjugated to horseradish peroxidase, the membrane was washed, and color was developed by the use of an enhanced chemiluminescence (7) kit (Applygen Technologies, Beijing, China).
The aortic alterations of l-Arg/NOS/NO pathway by injection H2S to in vivo rats.
Rats were intraperitoneally injected with NaHS (3.5 μmol/kg ip each 30 min, for a total of four injections, total H2S 14 μmol/kg, n = 6 rats) or normal saline (n = 6 control rats). Two hours after the first injection, rats were anesthetized by intraperitoneal injection with pentobarbital sodium (30 mg/kg). The blood was collected in heparinized syringes from the abdominal aortas and transferred to tubes. The plasma was separated to assay nitrate plus nitrite content. The aortic tissues were isolated, and its NOS activity and l-Arg uptake were measured according to the above methods.
The results are expressed as means ± SD. Comparisons across more than two groups were analyzed by one-way ANOVA followed by the Student-Newman-Keuls test. A two-tailed P value <0.05 was considered statistically significant.
H2S inhibited NO generation of aortic tissues.
After l-Arg administration to cultured aortic tissues, the level of nitrate plus nitrite, both stable products of NO, continuously accumulated in the medium up to 6 h (Fig. 1A). After administration of H2S (50 μmol/l) and NaHS (50 μmol/l), the NO products in the medium were decreased, by 63% and 47% (P < 0.01) after 2 h, 48% and 30% (P < 0.01) after 4 h, and 32% and 19% (P < 0.05) after 6 h, respectively, compared with that in controls (Fig. 1A). According to the Fig. 1A results, we selected 2 h as a time of best inhibitory effect and observed the effects of various concentrations (from 1 to 1,000 μmol/l) of H2S on NO release at this time. Administration of both H2S and NaHS significantly reduced NO production in a concentration-dependent manner after a 2-h incubation (Fig. 1B). The IC50 value was 19.5 μmol/l for H2S [95% confidence interval (CI), 7.06–53.09 μmol/l] and 15.9 μmol/l for NaHS (95% CI, 7.99–31.59 μmol/l). Although the IC50 of H2S was a little higher than that of NaHS, the inhibitory effect of H2S was stronger than that of NaHS (Fig. 1C). The results showed that exogenous H2S inhibits vascular NO production in vitro.
We felt it important to explore the effect of endogenously generated H2S. Administration of l-cysteine (l-Cys, the substrate of CSE) and pyridoxal-5′-phosphate (PLP, a cofactor of CSE) increased H2S release from incubated aortic tissues with H2S concentration in the culture medium reaching 37.5 ± 2.7 μmol/l (Fig. 1D); baseline H2S concentration in the culture medium was undetectable. Plasma H2S concentration was 45.6 ± 10.59 μmol/l, according to values reported by Zhao et al. (50). At the same time, l-Cys and PLP treatment reduced NO production by 43% compared with that in controls (P < 0.01, Fig. 1E). The effect of l-Cys and PLP treatment on NO production was blocked by preincubation with an inhibitor of CSE, DL-propargylglycine (PAG) (Fig. 1E), which prevented endogenous H2S formation (the H2S concentration in culture medium could not be detected by sensitive sulfur electrode, as shown in Fig. 1D).
The KATP channel is an important molecular target of H2S in cardiovascular tissues (4, 47, 50). To assess the role of KATP channel in H2S-induced decrease of NO production, the KATP channel opener pinacidil and KATP channel inhibitor glibenclamide were used. Pinacidil reduced NO production by 34% and H2S (100 μmol/l) level by 30% (P < 0.01) (Fig. 1F). The activity of both agents was largely blocked by pretreatment with glibenclamide (10 μmol/l). These results suggest that the KATP channel is involved in inhibition of H2S-induced NO generation.
H2S downregulated aortic eNOS activity and changed its kinetics.
eNOS is the principle form of NO synthase in normal arterial endothelium. On incubation, aortic tissue slices with H2S (50 μmol/l) and NaHS (50 μmol/l), reduced eNOS activity by 56% and 47% (P < 0.01) after 2 h, 39% and 27% (P < 0.01) after 4 h, and 12% and 11% (P > 0.05, Fig. 2A) after 6 h, respectively, compared with that in controls. In addition, H2S and NaHS (from 1 to 1,000 μmol/l) markedly decreased aortic eNOS activity in a concentration-dependent manner after 2 h incubation (Fig. 2B). The IC50 values of H2S and NaHS for eNOS activity were 11.49 μmol/l (95% CI, 9.53–19.87 μmol/l) and 31.97 μmol/l (95% CI, 11.93–85.70 μmol/l), respectively. H2S was stronger in inhibiting eNOS activity than NaHS (Fig. 2C). The results are similar to those with NO generation, which suggests that the H2S reduced vascular NO generation, at least in part, by inhibiting eNOS activity.
The kinetics of eNOS regulation by H2S were studied in HUVEC cultures. Incubation of HUVEC with NaHS (50 μmol/l, 2 h) inhibited the maximal activity of eNOS (Vmax, 75.1 ± 2.9 vs. control 113.0 ± 4.0 pmol·min−1·mg protein−1, P < 0.01) without modifying the Km value for l-Arg (33.6 ± 4.4 vs. control 39.3 ± 3.0 μmol/l, P > 0.05, Fig. 2D), thus significantly decreasing the enzyme activity efficiency (ratio of Vmax to Km: 2.24 ± 0.66 vs. 2.88 ± 1.33, P < 0.01). We investigated the effect on aortic eNOS activity of endogenously generated H2S by adding l-Cys and PLP to HUVECs. eNOS activity was reduced by 48% (P < 0.01; Fig. 2E), an effect that was attenuated by pretreatment with PAG (a CSE inhibitor) (48). l-canavanine, a selective iNOS inhibitor (26), and spermidine, an nNOS inhibitor, alone or together, did not affect the inhibition effects of H2S (100 μmol/l) on NOS activity (Fig. 2F), which indicates that H2S specifically regulates eNOS activity. Like H2S, pinacidil inhibited eNOS activity, which was not blocked by the iNOS or nNOS inhibitors used. However, glibenclamide blocked inhibition of eNOS activity by H2S and pinacidil. Collectively, these results suggest that a KATP channel is involved in the impact of H2S on eNOS activity.
Both Ca2+-dependent and -independent pathways are involved in the regulation of eNOS activity. There is evidence that H2S modulates Ca2+ influx in cardiovascular cells (1), suggesting that H2S may inhibit eNOS activity through a Ca2+-dependent pathway. Whether a Ca2+-independent pathway is involved in the inhibition of eNOS activity by H2S is unknown. We detected phosphorylation of Ser1177, which activates eNOS by a Ca2+-independent pathway (17), in eNOS protein. Incubation with H2S (500 μmol/l) for 2 h inhibited the phosphorylation of eNOS Ser1177 without changing eNOS expression (Fig. 2G; P < 0.01). Pinacidil also slightly inhibited eNOS phosphorylation. Pretreatment with glibenclamide (10 μmol/l) partially blocked the effects of H2S and pinacidil on eNOS phosphorylation (Fig. 2G). The results confirmed the role of a KATP channel in the regulation of H2S in eNOS activity.
Akt (also called protein kinase B) mediates eNOS phosphorylation of Ser1177 in vitro and in vivo and plays a role in both calcium-dependent and -independent eNOS activation (15). H2S treatment for 30 min significantly reduced Akt phosphorylation without changing total Akt protein expression (Fig. 2H). Pinacidil also reduced Akt phosphorylation, and glibenclamide blocked the effects of H2S and pinacidil. This suggests that Akt signaling is involved in the regulation of H2S to eNOS phosphorylation, which partially modulates by opening KATP channels.
H2S inhibited l-[3H]Arg uptake by aortic tissues and changed l-[3H]Arg transport kinetics in HUVECs.
To determine the impact of H2S on transcellular transport of l-Arg, we measured l-[3H]Arg uptake in aortic tissues. H2S (50 μmol/l) and NaHS (50 μmol/l) decreased l-[3H]Arg uptake by 58% and 50% (P < 0.01) after 2 h, 39% and 30% (P < 0.01) after 4 h, and 12% and 8% (P > 0.05) after 6 h, respectively (Fig. 3A). In addition, incubation with H2S and NaHS for 2 h inhibited the l-[3H]Arg uptake in a concentration-dependent manner (1 to 1,000 μmol/l; Fig. 3B). The IC50 values were 84.08 μmol/l (95% CI: 27.28–259.20 μmol/l) and 39.15 μmol/l (95% CI: 10.46–146.50 μmol/l) for H2S and NaHS, respectively, the inhibitory effect of H2S being stronger than that of NaHS (Fig. 3C).
To explore the regulation of l-Arg transport activity by H2S, we investigated the kinetics of l-Arg transport in HUVECs. NaHS (50 μmol/l, 2 h) decreased the Vmax of l-[3H]Arg transport (17.5 ± 1.2 vs. control 20.6 ± 1.1 pmol/105 cells/min, P < 0.01) without modifying the Km value (84.4 ± 14.8 vs. 73.0 ± 10.4 μmol/l, P > 0.05) and thus reduced significantly the transport efficiency (Vmax/Km: 0.21 ± 0.08 vs. 0.28 ± 0.11, P < 0.01) (Fig. 3D).
In addition, administration of l-Cys and PLP inhibited l-Arg uptake, which was blocked by PAG pretreatment, as shown in Fig. 3E. The results suggest that H2S, whether exogenously administered or endogenously generated, inhibits l-Arg transport.
HUVECs transport arginine through two Na+-independent systems. System y+ is sensitive to N-ethylmaleimide (NEM) and relates to the expression of CAT1 and CAT2B. System y+L relates to the expression of y+LAT2, y+LAT1, and 4F2hc (35). To identify which type of transporter system acts as a main target of H2S, we investigated l-Arg uptake with or without NEM (0.5 mmol/l, pretreatment 20 min). NEM reduced l-Arg uptake by 32% compared with control (Fig. 3F, P < 0.01). NEM pretreatment plus H2S (100 μmol/l) did not change the l-Arg influx compared with the effect of NEM alone (Fig. 3F; P > 0.05). The KATP channel opener pinacidil slightly reduced l-Arg uptake (12% decrease compared with control) as did H2S, but NEM intensified the inhibitory effects of pinacidil on l-Arg uptake compared with NEM treatment alone (Fig. 3F, P < 0.01). Simultaneously, glibenclamide blocked, in part, the inhibitory effects of both H2S and pinacidil.
H2S down-regulated CAT-1 and eNOS gene transcript level in HUVECs.
To determine whether H2S modulates the transcript level of CAT-1, mRNA expression of the l-Arg transporter and of eNOS was determined by RT-PCR in HUVEC after incubation for 4 h with H2S (100 μmol/l). Transcript levels of CAT-1 and eNOS were 40% and 43%, respectively (all P < 0.01, Fig. 4) less than that in controls.
H2S inhibited eNOS but did not change iNOS and nNOS protein expression in HUVECs.
To confirm whether H2S modulates eNOS, eNOS protein expression was assayed by Western blot analysis in HUVECs after 6-h H2S incubation. H2S (100, 500, and 1,000 μmol/l) significantly inhibited eNOS protein expression but not in a concentration-dependent manner, and the inhibitory effect was partially blocked by glibenclamide preincubation (10 μmol/l, Fig. 5A). Fig. 5B confirmed the above phenomena and showed pinacidil-induced inhibition of eNOS protein expression and block of this by glibenclamide.
Induction of iNOS protein expression by H2S was not observed (Fig. 5C). nNOS is another constitutive NOS that is expressed in endothelial cells. Neither H2S nor pinacidil altered nNOS expression (Fig. 5D).
NaHS inhibited l-Arg/NOS/NO pathway in rats in vivo.
To demonstrate the in vivo effects of H2S on the vascular l-Arg/NOS/NO pathway, we analyzed plasma levels of nitrite and nitrate after intraperitoneal administration of 14 μmol/kg NaHS in rats. At 2 h after NaHS injection, the plasma content of nitrate plus nitrite was 21% (24.44 ± 0.80 vs. 31.00 ± 2.66 μmol/l less than in controls, P < 0.01, Fig. 6A). Aortic eNOS activity was reduced by 42% (23.06 ± 5. 08 vs. 39.50 ± 8.67 pmol·min−1·mg protein−1, P < 0.01, Fig. 6B), and l-Arg uptake decreased by 30% (88.35 ± 14.31 vs. 125.40 ± 18.96 pmol·min−1·mg protein−1, P < 0.01, Fig. 6C). iNOS activity was not induced by H2S injection in the in vivo experiments (Fig. 6B).
Recently, the novel endogenous gas H2S has been recognized as another cardiovascular gasotransmitter that exerts important cardiovascular effects similar to those of NO and CO. However, H2S, as opposed to NO and CO, is specifically released from VSMCs but not endothelium, opens the KATP channel, induces cellular membrane hyperpolarization, and causes vessel dilation in an autocrine/paracrine manner (40, 41, 50), which differs from the effect of NO and CO. Some studies have reported that NO enhanced endogenous H2S production (48) and its vascular dilation effect (16), whereas H2S inhibited vascular relaxation induced by NO (49). In the present study, we found that endogenous H2S downregulated the vascular l-Arg/NOS/NO pathway. This was mediated in part through the KATP channel. Taken together, these results suggest a delicate crosstalk between endothelial and VSMCs by a complex interaction of the gasotransmitters NO and H2S (Fig. 7).
The physiological serum concentration of H2S in Sprague-Dawley (SD) rats has been reported to be 45.6+/−14.2 μmol/l (50). In the present study, an exogenously administered physiological concentration of H2S or its donor, NaHS, inhibited NO release from cultured aortic tissues. The inhibitory effect of H2S weakened with incubation time but remained detectable for 6 h or more, consistent with anticipated changes of H2S concentration in the medium (as a gasotransmitter, H2S could be metabolized in the cell quickly) (44). According to the concentration-dependent inhibitory effects of H2S, the IC50 value of H2S was calculated as approximately one-half of the physiological serum concentration. Endogenous H2S generation from aortic tissues is 3.6 nmol·min−1·g wet tissue−1 (48). SNP (from 10 to 1,000 nmol/l) increased aortic H2S production in a concentration-dependent manner; 1 μmol/l SNP can increase H2S by 2 nmol·min−1·g wet tissue−1 (48). One possibility is a negative feedback regulation in which accumulated NO enhances H2S release from VSMCs during the short term and, over accumulative H2S, inhibits NO formation, achieving dynamic balance in the ultimate.
H2S endogenously generated by l-Cys plus PLP reduced NO production by 43%. Pretreatment with PAG, a CSE inhibitor (51) and pinacidil, a nonselective KATP channel opener, decreased aortic NO production, similar to the effect of H2S. Glibenclamide, a nonselective inhibitor, blocked the inhibitory effects of H2S and pinacidil. Collectively, these data suggest that endogenously generated H2S downregulates vascular NO production, and opening of the KATP channel could be one of the possible mechanisms.
NO is generated from the conversion of l-Arg to l-citrulline by the enzymatic action of NADPH-dependent NO synthases (NOS). NOS exists in 3 distinct isoforms, a constitutive neuronal NOS (NOS I or nNOS), an inducible NOS (NOS II or iNOS), and a constitutive endothelial NOS (NOS III or eNOS) (1). In the vaculature, NO is produced mainly in the endothelium by constitutively expressed eNOS. The present results suggest that exogenous H2S or its donor NaHS reduced eNOS activity within 4 h in a concentration (from 1 to 1,000 μmol/l)dependent manner. The IC50 values for H2S (11.49 μmol/l) were ∼26% of the physiological serum H2S concentration, and lower than that of H2S-inhibited NO production (19.5 μmol/l). The kinetic study showed that H2S did not modify the enzyme affinity to l-Arg (represented as Km value) but decreased the maximum value of catalytic velocity (Vmax), indicating that H2S reduces eNOS catalytic efficiency. H2S endogenously generated by administration of l-Cys plus PLP also inhibited eNOS activity.
To verify the specificity of H2S-mediated inhibition of eNOS activity, we investigated the effect of single or combined administration of the selective iNOS inhibitor l-canavanine and nNOS inhibitor spermidine. These treatments did not change the inhibitory effect of H2S on constitutive NOS activity. Although iNOS activity could be induced by stimuli such as cytokines, reactive oxygen species, or some drugs (1), we detected very low iNOS activity in normal cultured aortic tissues, and neither H2S nor NaHS incubation affected iNOS activity during the entire incubation period. Additionally, H2S inhibits eNOS transcript level and protein expression. These findings suggest that eNOS is an important regulatory target of H2S-affected NO generation.
l-Arg transport is another key step to limiting NO generation besides NOS catalysis (8). In the present study, exogenous administration of both H2S and NaHS significantly reduced l-[3H]Arg transporter activity within 4 h and returned to control levels after 6 h. Inhibition was concentration dependent in the 1 to 1,000 μmol/l range, and the inhibitory effect of H2S (from 12.4% to 55.3%) was higher than that of NaHS (from 3.6% to 38.8%). The IC50 values were 84 μmol/l for H2S, and approximated 182% of physiological serum concentrations. Intracellular levels of l-Arg are not generally considered rate-limiting for eNOS catalysis (51). Consumption of intracellular l-Arg initiates and stimulates its uptake by cells (30), and at this time, NO generation is dependent on l-Arg uptake. These may explain why the IC50 of H2S for l-Arg uptake was higher than that for NO generation. Kinetic analysis of the l-Arg transport showed that H2S inhibited the Vmax of l-[3H]Arg uptake but did not influence its Km value. Furthermore, administration of l-Cys plus PLP increased endogenous H2S production and reduced l-Arg transport.
l-Arg enters mammalian cells through several membrane-bound cationic amino acid transporters: systems y+, b0, +, B0, +, and y+L (8). Previous studies have demonstrated that Arg transport in the vascular endothelium is mainly attributed to system y+, and to some extent to system y+L (35). NEM, a system y+ selective inhibitor, reduced l-Arg influx per se. After pretreatment with NEM, H2S did not change l-Arg influx, suggesting that the NEM-sensitive system y+ transporter activity contributes to the regulation of l-Arg uptake by H2S. In addition, the present study found that H2S also decreased the CAT1 (encoding the system y+ transporter) transcript level by 34%. All of the above results suggest that endogenous H2S inhibits l-Arg transport under physiological conditions and that the system y+ transporter is the main pathway regulated by H2S.
The effect of injection of NaHS on the l-Arg/NOS/NO pathway was also examined in vivo in rats. According to our previous work, injection of NaHS (14 μmol/kg) after 2 h increased plasma H2S concentration by 67% over baseline (52). In the present study, the plasma level of nitrate plus nitrite was decreased by 21%, and aortic eNOS activity by 42%, and l-Arg by 30% 2 h after injection of NaHS. These in vivo results were similar to those of the in vitro experiments and support the notion that H2S downregulates the vascular l-Arg/NOS/NO pathway (i.e., inhibited l-Arg uptake, eNOS activity, and NO generation). Venous injection of H2S can directly induce vasodilation and cause transient hypotension (16, 50), and we have previously reported that intraperitoneal bolus injection of NaHS reduces arterial blood pressure. Increasing plasma H2S level from 39 to 79 μmol/l reduced arterial blood pressure by 22 mmHg (from 134 to 112 mmHg); but plasma H2S concentration increased from 79 to 123 μmol/l, while arterial blood pressure fell by 6 mmHg (from 112 to 106 mmHg) (52). The present in vivo studies that show inhibition of the l-Arg/NOS/NO pathway by H2S may explain, in part, this phenomenon.
Our findings support the concept that eNOS is an important target molecule of H2S, affecting its activity, as well as gene and/or protein expression. Several mechanisms are known to regulate eNOS activity, including the interaction of eNOS with caveolin-1, heat shock protein 90 (Hsp90), or membrane phospholipids, and enzyme translocation and phosphorylation (1). Ser1177 appears to be the most important site of eNOS phosphorylation and is affected by most, if not all, of the diverse stimuli that promote eNOS activation. Phosphorylation of eNOS-Ser1177 increases eNOS sensitivity to Ca2+/calmodulin binding and leads to eNOS activation (28). Our study found that H2S downregulated eNOS phosphorylation at Ser1177. This result implies that H2S inhibits eNOS activity partly through inhibition eNOS phosphorylation at Ser1177. Akt-mediated-phosphorylation of eNOS-Ser1177 is a crucial step of eNOS activation induced by estrogen, insulin, VEGF, statins, and shear stress (15). H2S reduced Akt phosphorylation without influencing total Akt protein, suggesting that the Akt pathway may contribute to regulation of H2S by eNOS-Ser1177 phosphorylation. H2S is an endogenous opener of the KATP channel in many cell types, e.g., (39, 47, 50). The KATP channel inhibitor glibenclamide blocked the effects of H2S on eNOS activity (Fig. 2, G and H), eNOS phosporylation, and Akt phosphorylation. Glibenclamide also blocked inhibition of eNOS phosphorylation by the nonselective KATP channel opener pinacidil. These data imply that H2S may open KATP channels, followed by reduced Akt phosphorylation and inhibition of eNOS phosphorylation, leading to downregulation of eNOS activity.
Our study found that H2S also decreased eNOS transcript abundance. Numerous physiological and pathophysiological stimuli have been identified to modulate eNOS expression. Reactive oxygen species, as a signal molecule, play an important role in eNOS transcription and posttranslational regulation. Effects of H2O2 on eNOS transcription and posttranslational modification were reported in bovine and human endothelial cells. Chronic oxidative stress caused by excessive H2O2 production in vivo evokes a compensatory response involving increased eNOS transcript (36). Our previous work showed that H2S reduced oxidative radical release and scavenged H2O2 directly (11). H2S may reduce the H2O2 signal in eNOS transcription and posttranslation and thereby reduce eNOS transcription. Fig. 5D revealed that KATP channel opening by H2S and pinacidil inhibited eNOS protein expression, effects that were blocked by glibenclamide. These data suggest that KATP channels are involved in the inhibition of eNOS protein expression by H2S. However, the precise mechanisms of inhibition of eNOS protein expression by H2S need to be further investigated.
l-Arg transporter could be an additional target molecule of H2S regulation. l-Arg transport from plasma into cells is mediated by several different classes of CATs. System y+ activity, selective for cationic amino acids only, is encoded by 4 genes (CAT1, CAT2, CAT3, and CAT4) representing different isoforms (35). Systems y+ (CAT-1) and y+L (heavy chain subunit-4F2hc) have been detected in the endothelium (9). Transport of cationic amino acids via CAT-1 is voltage dependent with alterations of membrane potential affecting both Vmax and Km values for influx (22). l-Arg transport is sensitive to the membrane potential; it is stimulated by drugs that cause membrane hyperpolarization and inhibited by those that cause membrane depolarization (51). H2S opens the KATP channel and induces membrane hyperpolarization (50), which may contribute to l-Arg transport activity, and furthermore, it results in an accumulation of the intracellular pool of l-Arg available for eNOS. However, opening KATP channels alone is not enough to explain the mechanisms of H2S-mediated l-Arg influx. Administration of pinacidil reduced l-Arg influx, as did H2S, but NEM could not block the inhibitory effects of pinacidil. Glibenclamide blocked the effect of pinacidil to a lesser extent than H2S, suggesting that the mechanisms involved in the H2S regulation of l-Arg influx may be more complex.
Administration of pinacidil-reduced l-Arg influx as did H2S, but NEM could not block the effects of pinacidil. Glibenclamide blocked almost completely the inhibition effects of pinacidil but only slightly blocked the effects of H2S, which suggests that the mechanisms involved in the H2S regulation of l-Arg influx are complex. Opening KATP channels alone is not enough to explain the mechanisms of H2S-mediated l-Arg influx.
Oh et al. (29) found that H2S could inhibit NO production in LPS-stimulated macrophages through a mechanism that involves heme oxygenase-1(HO-1)/CO. H2S increased pulmonary arterial HO-1 gene transcription and protein expression (32). CO, derived from VSMCs and endothelium, has vasodilatory and antiproliferative effects and acts as a competitive antagonist for NO-mediated sGC activation or displaces internal stores of NO. As a heme ligand, CO potentially inhibits NOS activity and reduces NO production (34, 38).
Taken together, these results suggest that the gasotransmitter family, including NO, CO, and H2S, derived from vascular tissues in a paracrine/autocrine manner, interacts in the regulation of vascular homeostasis. The interaction of these gasotransmitters in distinct locations of the blood vessels could maintain a dynamic balance and form a regulatory "network" (41). Imbalances of the network regulation may contribute to pathogenesis of cardiovascular diseases. Research into the network regulation of gasotransmitters should reveal a novel prospect for understanding the mechanism of cardiovascular diseases and a novel target for prevention and treatment.
This work was supported by The Major State Basic Research Development Program of the People's Republic of China (Grant 2006CB503807), National Natural Science Foundation of People's Republic China (Grant 30400151), National Natural Science Foundation for Distinguished Young Scholars (Grant 30425010), and Grant 20030001035 from the Ministry of Education, China.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society