|Home | About | Journals | Submit | Contact Us | Français|
We investigated whether a reduced activity in the Rho-A/Rho-kinase pathway could be involved in the impaired vascular reactivity observed in septic shock.
ex vivo animal study.
University research laboratory.
Male Wistar rats.
Rats received an intraperitoneal injection of lypopolisaccharide (LPS, 10 mg/kg) either 6 or 24 h before the onset of our experiments. The effects of Y-27632 (a Rho-kinase inhibitor) were assessed in first order mesenteric rings taken from these animals using wire myograph. The expression of Rho-A, Rho-kinases I and II, and the total and phosphorylated myosin phosphatase targeting subunit 1 (MYPT1) were assessed by western blotting.
The EC50 to Y-27632 was reduced from 2.10 (1.22-3.66) μM (control), to 0.21 (0.09-0.44) μM, and 9.54 (0.82-110.30) nM in LPS-treated groups 6 and 24 h, respectively. The increased potency of Y-27632 was partially reversed by endothelium removal at both 6 and 24 h. Incubation of L-NAME or 1400W (a non-selective and an inducible NOS inhibitor, respectively) normalized the responses to Y-27632 seen 6 h after LPS. However, 1400W had no effect whereas L-NAME caused a partial reduction in the enhanced potency of Y-27632 found 24 h after LPS. The soluble guanylate cyclase inhibitor ODQ was able to bring the Y-27632 response back to normal both 6 and 24 h after LPS. Rho-A, Rho-kinase-I, Rho-kinase-II, and MYPT1 were increased in mesenteric arteries from endotoxemic rats, but the phosphorylated-MYPT1 was significantly reduced. However, incubation with ODQ circumvented the inhibition of MYPT1-phosphorylation found in preparations from LPS-treated animals.
Our findings disclosed an impaired Rho-A/Rho-kinase-mediated phosphorylation of MYPT1 in vessels from endotoxemic animals in a cGMP-dependent manner, suggesting that changes in mechanisms involved in calcium sensitization play a pivotal role in cardiovascular changes observed in septic shock.
Experimental and clinical studies have suggested that the hypotension in septic shock is, at least in part, the consequence of reduced reactivity of microcirculation to endogenous vasoconstrictors (e.g. 1, 2), and these changes in vascular responses may often reduce the efficacy of vasoactive drug support in septic patients. Several events have been associated with the impairment of the vascular reactivity that takes place in experimental models of septic shock, such as inactivation of norepinephrine (3), inactivation of alpha-1 adrenoceptores (4), reduction in intracellular calcium amount and sensitivity (5, 6), increased opening of potassium channels (7-9), and augmented activity of Na+-K+-ATPASE pump (10). Most of these alterations appear to be directly or indirectly (via activation of the soluble guanylate cyclase enzyme) mediated by nitric oxide (e.g. 11, 12), which is produced in large amounts during sepsis, mainly by the inducible isoform of the nitric oxide synthase (iNOS) enzyme.
Among the cell signaling pathways that are crucial to control vascular tone, the Rho-A/Rho-kinase pathway has been scarcely studied in sepsis. Rho-A is a small GTP-binding protein that is involved in the mediation of vascular contraction elicited by several vasoactive agents, including α1-adrenergic agonists (13, 14). Activation of Rho-A leads to stimulation of Rho-kinase which can phosphorylate and subsequently inactivate the myosin light chain (MLC) phosphatase (15, 16), favoring MLC phosphorylation, actin-myosin interaction and cell contraction (for review see 17). Inhibition of Rho-kinase has been shown to reduce leukocyte adhesion in vessels (18) and liver (19) from endotoxemic mouse. Nevertheless, to our knowledge, the functionality of the Rho-A/Rho-kinase in vessels during sepsis has not been evaluated.
Thus, using small first order mesenteric arteries obtained from endotoxemic rats, we confirmed our hypothesis that a reduction in the activity of the Rho-A/Rho-kinase pathway could be involved in the impairment of vascular reactivity existing in sepsis. Considering the higher amounts of NO that are produced in sepsis and the suggested ability of this mediator to physiologically inhibit the Rho-kinase pathway in different tissues, including arteries (20, 21), we also addressed the potential relationship between the NO/guanylate cyclase and the Rho-A/Rho-kinase pathways during endotoxemia. The results obtained support our main hypothesis revealing that although Rho-A/Rho-kinase components are up-regulated in mesenteric arteries from endotoxemic rats, its ability to inhibit the myosin phosphatase activity was markedly blunted, disclosing that impairment in mechanisms involved in calcium sensitization may play a role in cardiovascular changes existing in septic shock.
Male Wistar rats (250-280 g) provided by Harlan Laboratories (Indianapolis, IN, USA) were kept in our animal room under standard laboratory conditions with a constant 12 h light/dark cycle and controlled temperature (22 ± 2 °C), with water and food ad libitum. All procedures were conducted in accordance with internationally accepted principles and were previously approved by the Institutional Animal Care and Use Committee from Medical College of Georgia (GA, USA).
Rats received a single intraperitoneal injection of lypopolisaccharide (LPS, 10 mg/kg) either 6 or 24 h before the onset of our experiments. Control experiments were performed with animals treated with saline (vehicle to LPS; 0.1 ml/100 g).
Animals were killed in a saturated CO2 chamber and immediately the entire intestine with vascular arcades was excised and kept in cold physiological salt solution (PSS, composition in mM: NaCl 130.3, KCl 4.7, CaCl2·2H2O 1.6, KH2PO4 1.18, MgSO4 1.17, D-glucose 5.5, NaHCO3 14.9, and EDTA 0.03). First order branches stemming from the superior mesenteric artery were cut into 1 mm long segments and mounted in a wire myograph (Danish Myo Technology, Denmark) for measurement of isotonic force. Preparations were exposed to a basal tension of 7.5 mN and kept under temperature-controlled (37 °C) and continuously aerated (5% CO2/95% O2) 5 ml of PSS. An interval of 45 min with frequent changes of PSS (each 15 min) without addition of any drug was allowed for stabilization. After this interval all preparations were exposed to KCl (120 mM) and the maximal contraction recorded. The vessels were washed and after a new stabilization period of 45 min the integrity of the endothelial layer was verified by the ability of acetylcholine (1 μM) to relax (at least 80%) vessels contracted by phenylephrine (10 μM). In the experiments shown in figure 3, endothelium was removed by gentle rubbing inside the vessel with a stainless-steel wire. The removal of the endothelium was confirmed by the absence of relaxation induced by acetylcholine in phenylephrine-contracted vessels.
Endothelium-intact or denuded mesenteric rings obtained from LPS-treated (both groups 6 and 24 h) or control rats were contracted with phenylephrine (30 μM) and during the maintained contraction were exposed to cumulative concentrations of Y-27632 (1 nM-100 μM) or acetylcholine (1 nM-1 μM). In some experiments, endothelium-intact mesenteric rings obtained from control animals were incubated with the nitric oxide donor SNAP (70 μM) for 10 min. After this period of incubation, preparations were washed and the vasodilatory responses to Y-27632 or acetylcholine were measured 15 min after the removal of SNAP.
Arterial rings with functional endothelium from LPS-treated (both groups 6 and 24 h) or control animals were incubated in the presence of Y-27632 (3 μM for 10 min) before exposure to cumulative concentrations of phenylephrine (1 nM-300 μM). Time-control experiments (without addition of Y-27632) were conducted for comparison.
In this set of experiments endothelium-intact preparations from LPS-treated (both groups 6 and 24 h) or control animals were incubated with L-NAME (100 μM), indomethacin (10 μM), 1400W (1 μM), zinc protoporphryn IX (500 nM), or 7-nitroindazole (7-NI, 100 μM) for 40 min, or ODQ (10 μM) for 10 min. Each preparation was exposed to only one of these drugs. After the incubation period the vessels were contracted with phenylephrine and the relaxation induced by Y-27632 (1 nM-100 μM) evaluated. Control experiments with incubation of vehicle only (usually 6 μl of Krebs solution or DMSO) were carried out for each experimental group and used for comparison.
Protein expression was evaluated in the whole mesenteric arterial bed (except the superior mesenteric artery) taken from LPS- or saline-treated animals. In some experiments, after removal from animals the vascular bed was kept in warmed (37 °C) PSS solution continuously aerated with the carbogenic mixture for 30 min, and then incubated with ODQ (10 μM) or vehicle (DMSO) for 10 min, followed by a single concentration of phenylephrine (30 μM) for 5 min. After this period of incubation, tissues were quickly frozen in liquid nitrogen until be processed in western blot experiments. Samples containing 90 μg of protein were processed for analyses of Rho-A, ROCK-I, ROCK-II, iNOS, soluble guanylate cyclase β1 subunit, MYPT1, and phosphorylated-MYPT1 as previously described (22). Immunoreactivitity was detected by enhanced chemiluminescence autoradiography. The bands were quantified by densitometry using Scion Image software (Scion Co., Frederick, MD).
Phenylephrine hydrochloride, acetylcholine chloride, indomethacin, Nω-Nitro-L-arginine methyl ester hydrochloride (L-NAME), 7-nitroindazole (7-NI), 1H-![1,2,4]oxadiazolo[4,3-alpha]quinoxalin-1-one (ODQ), dimethyl sulfoxide (DMSO), N-([3-(aminomethyl)phenyl]methyl)ethanimidamide dihydrochloride (1400W), zinc protoporphryn IX, lipopolysaccharide (from Escherichia coli - 0128:B12), sodium orthovanadate (Na3VO4), phenylmethylsulphonyl fluoride, protease inhibitor cocktail, and all salts used to prepare the PSS were purchased from Sigma (St. Louis, MO, USA). (R)-(+)-trans-4-(1-Aminoethyl)-N-(4-Pyridyl)cyclohexanecarboxamide dihydrochloride monohydrate (Y-27632) was obtained from Tocris (Ellisville, MO, USA). All reagents used to prepare the SDS-PAGE were acquired from Bio-Rad (Hercules, CA, USA). All primary antibodies were purchased from BD Biosciences (San Jose, CA, USA), except anti-phospho-MYPT1 that was obtained from Upstate (Lake Placid, NY, USA), and anti-β-actin that was from Sigma (St. Louis, MO, USA). Indomethacin was dissolved in 0.5% sodium bicarbonate and ODQ was prepared in DMSO. All other drugs and nutritive solutions were prepared in fresh distilled deionized water immediately before the experiments.
Results are expressed as mean ± standard error of mean of 5-9 experiments. Statistical significance was determined through one-way analysis of variance (ANOVA) followed by Bonferroni’s t test. A p value less than 0.05 was considered statistically significant. Graphs were drawn and statistical analyses were performed using GraphPad Prism version 5.01 for Windows (GraphPad Software, San Diego, CA, USA, www.graphpad.com).
The concentration response curve for relaxation induced by Y-27632 was left-shifted in small mesenteric rings from animals treated with LPS, both 6 and 24 h before, without change in the maximal response (Figure 1A). The EC50 (95% confidence intervals) was reduced from 2.10 (1.22-3.66) μM in control to 0.21 (0.09-0.44) μM, and 9.54 (0.82-110.30) nM in LPS-treated groups 6 and 24 h, respectively (p < 0.05). Nevertheless, the responses to acetylcholine in arteries from LPS-treated rats did not differ from control values (Figure 1B). Similarly, responses to Y-27632 but not to acetylcholine were increased in vessels previously exposed to SNAP, a nitric oxide donor (Figure 1C and D, respectively).
As expected, we observed reduced contractile responses to both phenylephrine (Figure 2A) and KCl (Figure 2B) in preparations obtained from endotoxemic animals. Although the incubation of Y-27632 (3 μM) for 10 min did not cause a significant effect in phenyleprhine-induced contractions in mesenteric rings from control (not treated with LPS) animals (Figure 2C), this Rho-kinase inhibitor reduced the maximal response to phenylephrine by 25 and 45% in rings obtained from LPS-treated animals 6 and 24 h before, respectively (Figures 2D and E).
Removal of endothelium caused a slight right-shift in the relaxation induced by Y-27632 in small mesenteric rings obtained from control animals (Figure 3A), and partially prevented the increased responses seen in preparations obtained from both 6 and 24 h LPS-treated groups (Figures 3B and C, respectively). The non-selective nitric oxide synthase (NOS) inhibitor L-NAME (100 μM) also caused minor effects on Y-27632-induced relaxation in control preparations (Figure 4A). The incubation of L-NAME was able to normalize the relaxation induced by Y-27632 in rings obtained from animals treated with LPS 6 h before (Figure 4B) and caused a partial reversal in the 24 h group (Figure 4C). In addition, the selective inhibitor of the inducible NOS isoform, 1400W (1 μM), did not alter the responses to Y-27632 in control preparations (Figure 5A) and circumvented the enhanced effects of Y-27632 at 6 h (Figure 5B), but not at 24 h (Figure 5C) after LPS administration. Moreover, although ODQ induced only a slight right-shift in the effect of Y-27632 in the control group (Figure 6A), this soluble guanylate cyclase inhibitor completely overcame the changes in Y-27632-induced relaxation both 6 and 24 h after LPS administration (Figure 6B and C, respectively). On the other hand, exposure to 7-NI (a selective neuronal NOS inhibitor), indomethacin (a cyclooxygenase inhibitor), or zinc protoporphryn IX (a heme oxygenase 1 inhibitor), did not alter Y-27632-induced relaxation in control or LPS-treated groups (data not shown).
Measurement of iNOS levels in the mesenteric arterial bed revealed that this enzyme is highly expressed 6 hours after LPS-injection, but only small amounts still detectable 24 h after (Figure 7A). Also, at least at the times investigated, we did not find differences in the amounts of sGC protein between the LPS-treated and control groups (Figure 7B). Protein levels of Rho-A, ROCK-I and ROCK-II were significantly increased in the homogenates from mesenteric arterial bed obtained from animals injected with LPS, both 6 and 24 h before (Figure 8). Our western-blot analysis also revealed that the mesenteric arterial bed from LPS-treated animals presented increased levels of total MYPT1 protein (Figure 9A). However, the amounts of phosphorylated-MYPT1 were significantly reduced both 6 and 24 h after induction of endotoxemia (Figure 9B).
Western blot assays revealed that incubation of the arterial mesenteric bed for 10 min with ODQ did not cause any significant change in the phosphorylation of MYPT1 in vessels from control animals. However, arteries from LPS-treated rats, both 6 and 24 h before the experiment, presented increased amounts of phosphorylated-MYPT1 after exposure to ODQ, when compared with preparations incubated with vehicle only (Figure 10).
The first finding of this study was that the relaxation induced by Y-27632, a selective and equipotent inhibitor of both isoforms of Rho-kinase, ROCK-I and ROCK-II (23), is enhanced in small mesenteric arteries from LPS-treated animals, whereas acetylcholine-induced relaxation was unaltered (Figure 1A and B). Similarly, previous exposure of mesenteric vessels from non-endotoxemic rats to the nitric oxide donor SNAP, also increased the effects of Y-27632 but not of acetylcholine (Figure 1 C and D). In addition, the Rho-kinase antagonist Y-27632 was a much more potent inhibitor of contractions induced by phenylephrine in mesenteric rings from endotoxemic rats than in control (from animals not treated with LPS) preparations (Figure 2). Taken together, these results reinforced our hypothesis that the functionality of the Rho-A/Rho-kinase in septic shock, a condition associated with high amounts of nitric oxide, could be impaired by mechanisms involving the activity of the NO/guanylate cyclase pathway.
To investigate the role of endothelium-derived substances in the increased potency of Y-27632, we performed experiments using endothelium-denuded mesenteric rings. Our results showed that removal of endothelium was not able to normalize the relaxation induced by Y-27632, but caused a significant reduction in its potency both 6 and 24 h after LPS (Figure 3B and C), suggesting that substances derived from endothelium might be accountable for at least part of this phenomenon. Because the classical cyclooxygenase inhibitor indomethacin and the heme oxygenase 1 inhibitor zinc protoporphryn IX did not change the effects of Y-27632 in any of our experimental groups (data not shown), the involvement of prostanoids or carbon monoxide in our findings were discarded.
The neuronal nitric oxide synthase (nNOS) inhibitor 7-NI did not influence the effects of Y-27632 (data not shown). However, incubation with L-NAME, a non-selective nitric oxide synthase inhibitor, as well with 1400W, an iNOS inhibitor, was able to completely restore the responses to Y-27632 at 6 h after LPS (Figure (Figure4B4B and and5B,5B, respectively). These results showed that, at least in this early stage of endotoxemia, the enhancement in the relaxation induced by Y-27632 is dependent on NO produced by the iNOS isoform present in both smooth muscle and endothelial cells. Importantly, enhanced production of NO mainly by iNOS has been assigned as a key event in the circulatory failure existing in septic shock (24), but the cellular mechanisms involved are still not completely understood. Interestingly, NO released by iNOS appears to be crucial in the enhanced effect of Y-27632 found 6 h but not 24 h after LPS, since in this later stage of endotoxemia 1400W had no effect (Figure 5C). This lack of effect by 1400W happened at a time when the levels of iNOS found in the mesenteric arterial bed were reduced to amounts very close to that seen in control animals (Figure 7A). Moreover, the ability o L-NAME (but not 1400W or 7-NI), as well as endothelium-removal to partially reverse the enhanced potency of Y-27632 seen 24 h after LPS (Figures (Figures3C3C and and4C),4C), suggests that endogenous NO coming from eNOS or sources other than iNOS (or nNOS) is involved in this phenomenon at later stages of endotoxemia. Since eNOS can generate superoxide anions (25), which were associated with several damages in sepsis (26), we are currently investigating how are the expression and activity of eNOS, as well as the importance of superoxide generation in our results. In addition, large amounts of NO, either exogenous or endogenous (produced by iNOS during endotoxemia), can cause long-lasting impairment in the vascular responsiveness (12, 27). These long-lasting effects of NO may be consequence of its ability to complex with endogenous thiols, being subsequently released for longs periods (28, 29).
In spite of the putative importance of the endothelium-derived hyperpolarizing EDHF in microcirculatory systems, such as the mesenteric bed (for review see 30), the ability of the nitric oxide donor SNAP to reproduce the state of hypersensitivity to Y-27632, as well as the full (at 6 h) or parcial (at 24 h) reversion caused by NO synthase inhibitors suggest that the enhancement in the effects of Y-27632 seen in endotoxemic small mesenteric arteries is not dependent of EDHF.
Using the selective soluble guanylate cyclase (sGC) inhibitor ODQ, we found that this enzyme plays a crucial role in the enhanced effect of Y-27632 existing in vessels from endotoxemic rats. ODQ caused a slight right-shift in the effects of Y-27632 in control preparations, but was able to bring the Y-27632 response back to normal levels in those from LPS-treated rats (both groups 6 and 24 h), as shown in Figure 6. Several studies have investigated the role of sGC in sepsis and septic shock. Most of them demonstrated that inhibition of the sGC can ameliorate the hypotension and the hyporesponsiveness to vasoactive agents in experimental models of sepsis (e.g. 12, 31-33). It has been shown that the protein levels of sGC in homogenates from lungs are reduced 8 h after the administration of LPS (34). However, this same study (34) demonstrated that the mRNA for sGC is increased 24 h after LPS. We investigated if increased protein levels of sGC in mesenteric vessels could explain the increased potency of Y-27632 described in our study. However, sGC protein levels were not different in the mesenteric bed from LPS-treated animals when compared to control tissue (Figure 7B), showing that the importance of the cGMP in the inhibition of the Rho-A/Rho-kinase pathway is not related to an increased expression of the sGC enzyme.
Since a down-regulation in the Rho-A/Rho-kinase pathway components could explain the increased potency of Y-27632, we evaluated the protein levels of Rho-A and its downstream targets ROCK-I and ROCK-II. Surprisingly, western blot analysis revealed that Rho-A protein levels, as well as ROCK-I and ROCK-II, were significantly increased (2 or more folds) in vessels from LPS-treated animals (Figure 8). To investigate the degree of activity of the Rho-A/Rho-kinase pathway, we also assessed the protein levels of both MYPT1 (the regulatory subunit of the smooth muscle myosin phosphatase), and phosphorylated-MYPT1. The results obtained disclosed that total MYPT1 protein levels (and consequently myosin phosphatase levels) were increased in the mesenteric arterial bed of endotoxemic rats (both groups 6 and 24 h; Figure 9A). The MYPT1 is physiologically phosphorylated by ROCK mainly at the Thr696 residue (35), resulting in reduced activity of the myosin phosphatase, that favors calcium-sensitization and maintenance of smooth muscle contraction (see the schematic diagram in Figure 11). Thus, higher cellular amounts of myosin phosphatase could be enough to render the vascular system less responsiveness to vasoactive stimulus in septic states. However, our results also disclosed that although total MYPT1 was increased, protein levels of phosphorylated-MYPT1 were significantly reduced (Figure 9B), indicating that even with the compensatory up-regulation of the Rho-A/Rho-kinase components (Rho-A, ROCK-I and ROCK-II), this pathway was not able to properly phosphorylate and inhibit the myosin phosphatase.
Physiological activation of sGC has been previously associated with a slight but significant inhibition of the Rho-A/Rho-kinase pathway in non-endotoxemic rat aorta (21). In our experiments, the inhibition of the soluble guanylate cyclase not only normalized the responses to Y-27632 (Figure 6), but also increased the levels of the phosphorylated-MYPT1 in mesenteric arteries from LPS-treated animals (Figure 10). It has been described that 8-bromo-cGMP, an analogue of cGMP, activates the myosin phosphatase leading to calcium-desensitization and vasodilation in vitro (36), and also leads to phosphorylation of Ser852 of MYPT1, that prevents ROCK-mediated phosphorylation and inhibition of myosin phosphatase (37).
Changes in the vascular responsiveness caused by activation of the nitric oxide/guanylate cyclase pathway during sepsis have been frequently associated with K+ channels opening, among other cellular impairments (8, 9,10, 11, 12). Our data, however, clearly discloses that the activity of the Rho-A/Rho-kinase components is suppressed by cGMP-dependent mechanisms in resistance arteries. A better understanding about the relationship between cGMP and the activity of the Rho-A/Rho-kinase in septic shock may contribute to the development of new therapeutic strategies for this condition.
Severe sepsis and septic shock are still being reported world-wide as a condition associated with high mortality rates. Systemic administration of LPS in animals is an experimental model commonly used to study most of the cardiovascular changes existing during septic shock in humans. At least to our knowledge this is the first study describing an impaired Rho-A/Rho-kinase-mediated phosphorylation of MYPT1 in the vascular smooth muscle from endotoxin-treated animals. Our results disclosed that changes in cGMP-dependent mechanisms involved in calcium sensitization may play a pivotal role in the cardiovascular response to septic shock, such as the hypotension and the hyporeactivity to vasoactive drugs.
This work was supported by grants from National Institute of Health (HL-71138 and HL-74167), Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq - Brazil), and Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES - Brazil).