INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

LIPOPOLYSACCHARIDE (LPS) induces hypotension and suppresses the contractile responses to various vasoconstrictors in vivo or in vitro and causes an increase in cGMP within the vascular wall (4, 10, 16, 49). It has been shown that nitric oxide (NO) activates soluble guanylyl cyclase (GC), thus producing relaxation of vascular and nonvascular smooth muscle through a rise in intracellular cGMP levels (13, 30). Recently, NO has been suggested to play an important role in animals with endotoxemia (15, 27, 44). This is attributed to the induction of a Ca²⁺-independent isoform of NO synthase (NOS) in the vascular smooth muscle cells that produces large amounts of NO, which, in turn, increases cGMP and then causes hypotension and vascular hyporeactivity (25).

The vascular hyporeactivity of endothelium-denuded aortic rings from rats treated with LPS is only partially overcome by N-nitro-L-arginine methyl ester (L-NAME), an inhibitor of NOS activity (19, 40, 45, 49). In addition to NO, carbon monoxide (CO) has similar effects on platelets and vascular smooth muscle (i.e., inhibition of platelet aggregation and vascular smooth muscle relaxation) through the activation of soluble GC (6, 22). Our previous results have already shown that the hyporesponsiveness to norepinephrine (NE) in aortic rings from rats treated with LPS is because of the activation of soluble GC, which is partially mediated by NO, but not CO, indicating that LPS induces the production of other mediator(s) that activates soluble GC in the vascular smooth muscle (49). However, there is no in vivo evidence to support this hypothesis.

The combination of dexamethasone (Dex; 3 mg/kg iv at 30 min before LPS) and aminoguanidine (AG; 15 mg/kg iv at 2 h after LPS), which we chose in this study, was based on our previous studies (39, 48) and the fact that Dex prevents the induction of Ca²⁺-independent NOS (29, 39) and AG inhibits the activity of Ca²⁺-independent NOS, once formed (48) in rats with endotoxemia. It has been shown that Ca²⁺-independent NOS activity was undetectable until rats were treated with LPS for 2 h (42). Therefore, we chose this time point to administer rats with AG to completely inhibit the formation of NO in rats with endotoxemia. Here, we also investigated possible mechanisms of other soluble GC-activating mediator(s) in the vascular hyporeactivity caused by endotoxin ex vivo.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Hemodynamic measurement. Ten-week-old male Wistar-Kyoto rats, whose stock originated from the Charles River Breeding Laboratories in Japan, were purchased from the Department of Laboratory Animal Science of the National Defense Medical Center. Rats were anesthetized by intraperitoneal injection of urethan (1.2 g/kg). The trachea was cannulated to facilitate respiration, and rectal temperature was maintained at 37°C with a homeothermic blanket (Harvard Apparatus, South Natick, MA). The right carotid artery was cannulated and connected to a pressure transducer (P23 ID, Statham, Oxnard, CA) for the measurement of phasic and mean arterial blood pressure (MAP) and heart rate (HR), which were displayed on a Gould model TA5000 polygraph recorder (Gould, Valley View, OH). The left jugular vein was cannulated for the administration of drugs.

Determination of nitrate in the plasma. The blood sample was centrifuged (7,200 g for 3 min) at room temperature to prepare plasma, and the plasma was kept in a 20°C freezer. At a later stage, plasma samples were thawed and deproteinized by incubating them with 95% ethanol (4°C) for 30 min. The samples were subsequently centrifuged for a further 7 min at 13,000 g. It is noted that the nitrate concentration depicted in the study is actually the total nitrite and nitrate concentrations in plasma. The amounts of nitrate in the plasma (2 µl) were measured by adding a reducing agent (0.8% VCl₃ in 1 N HCl) to the purge vessel to convert nitrate to NO, which was stripped from the plasma by using a helium purge gas. The NO was then drawn into the Sievers NO analyzer (Sievers 280 NOA, Sievers, Boulder, CO). Nitrate concentrations were calculated by comparison with standard solutions of sodium nitrate (Sigma Chemical, St. Louis, MO).

Determination of nitrate in vascular smooth muscle. Thoracic aortas from all groups of animals were removed at 360 min after vehicle or LPS injection. They were divided into two parts of experiments, one for the measurement of nitrate in vascular smooth muscle and the other for the contractile responses to NE (as in the following). It is noted that two rat aortas of each group were pooled together (i.e., n = 1) to measure the nitrate in the vascular smooth muscle. After fat and connective tissues were cleaned, aortas were longitudinally trimmed to expose the endothelium. The endothelium was gently removed using a wooden stick, and then the aorta was frozen in liquid nitrogen. Aortas were stored for no more than 2 wk at 80°C before assay. At a later stage, frozen samples were thawed and cut into pieces, and then these samples were homogenized on ice with an polytron PT MR 3000 homogenizer (Kinematic) in a buffer composed of the following (in mM): 50 Tris · HCl, 0.1 EDTA, 0.1 EGTA, 12 2-mercaptoethanol, and 1 phenylmethylsulfonyl fluoride (pH 7.4). Aortic homogenates were then further incubated with diethyldithiocarbamate (1 mM) for 10 min. This reaction was stopped by adding 1% TCA (4°C), and subsequently, these samples were centrifuged for 7 min at 13,000 g. The amounts of nitrate in the supernatant (30 µl, ~60 µg protein) were injected into the NO analyzer as described above.

Organ bath experiments. At the end of in vivo experiments, the rat thoracic aorta was excised and placed in cold physiological salt solution (PSS; 4°C). Control aortas were also obtained from untreated rats. Fat and connective tissues were trimmed from the aorta, and aortas were then cut into rings (3-4 mm in width). Rings of each vessel were rubbed gently to remove the endothelium. Later, the lack of relaxation to ACh (1 µM) after precontraction of rings with NE (100 nM) was considered as evidence that the endothelium had been successfully removed. The rings were mounted in organ baths (20 ml) filled with warmed (37°C) oxygenated (95% O₂-5% CO₂) PSS (pH 7.4) consisting of the following (in mM): 118 NaCl, 4.7 KCl, 1.2 KH₂PO₄, 1.25 MgSO₄, 2.5 CaCl₂, 25 NaHCO₃, and 11 glucose. In addition, 5.6 µM indomethacin was added to the solution to prevent the production of prostanoids in the experiment. Isometric force was measured with Grass FT03-type transducers (Grass Instruments, Quincy, MA) and recorded on a Grass model 7D polygraph recorder (Grass Instruments). The rings were left to equilibrate for 1 h under an optimal resting tension of 2.0 g and washed every 15 min. After we tested whether the endothelium had been successfully removed, drugs were removed from the bath by several washes with PSS, and the tension was allowed to return to baseline. The concentration-response curves to NE (1 nM to 1 µM) were obtained before and after the inhibitor of soluble GC, methylene blue (10 µM for 10 min); the inhibitor of NOS, L-NAME (0.3 mM for 15 min); the inhibitor of NO-sensitive GC, 1H-[1,2,4]oxidazolo[4,3-a]quinoxalin-1-one (ODQ; 1 µM for 15 min); the inhibitor of K⁺ channel, tetraethylammonium (1 mM for 15 min); the specific inhibitor of large-conductance Ca²⁺-activated K⁺ (K_Ca) channel, charybdotoxin (0.1 µM for 15 min); the specific inhibitor of small-conductance K_Ca channel, apamin (1 µM for 15 min); or the combination of L-NAME plus tetraethylammonium or ODQ plus charybdotoxin.

Determination of tissue cGMP content. The rings used to evaluate vascular hyporeactivity to NE were removed from organ baths and further incubated for 30 min in PSS containing IBMX (an inhibitor of phosphodiesterase, 0.1 mM) in a 37°C water bath. To stop the reaction, these rings were placed in 1 ml of 10% TCA at 4°C. After homogenization with a motor-driven glass homogenizer and sonication, the samples were centrifuged (2,500 g for 15 min at 4°C), and the supernatant was transferred to a 25-ml conical glass centrifuge tube to which 5 ml of water-saturated diethyl ether were added. The tube was stoppered and agitated with a vortex mixer at room temperature for 1 min. After the two phases were separated, the upper ether layer was discarded and aspirated off. The ether extraction was repeated three more times. After the last extraction, the tube was placed in an 80°C water bath for 5 min to drive off all traces of ether. Aliquots, 0.5 ml, from the remaining aqueous phase were then lyophilized and stored in a 70°C freezer. For the measurement of cGMP, the lyophilized residues were dissolved in 0.05 M sodium acetate, pH 6.2, and radioimmunoassayed with a cGMP assay kit (New England Nuclear, Boston, MA). Protein was determined by the method of Lowry (20) after the TCA precipitate had been dissolved in 1-2 ml of 0.5 N NaOH.

Statistical analysis. All values are expressed as means ± SE of n observations, where n represents the number of animals or preparations performed. All statistical evaluation was performed using ANOVA followed by a multiple-comparison test (Scheffé's test), except for the measurement of aortic nitrate or cGMP content, which was performed using Student's unpaired t-test. A value of P < 0.05 was considered to be statistically significant.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Complete inhibition of the production of plasma nitrate (an indicator of NO) by a combination of Dex plus AG in rats treated with endotoxin. The basal levels of plasma nitrate ranged from 8.18 ± 0.34 to 8.31 ± 0.40 µM and were not significantly different among any of the animal groups studied. In SOP animals treated with vehicle, no significant change in plasma nitrate was observed during the experimental period (Fig. 1). The administration of LPS (10 mg/kg iv) caused a time-dependent increase in plasma nitrate from 8.3 ± 0.4 to 73.1 ± 4.8 µM (n = 11, P < 0.05) within 6 h.

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Effects of dexamethasone (Dex) plus aminoguanidine (AG) on production of plasma nitrate in rats with endotoxemia. Depicted are changes of plasma nitrate in rats subjected to saline (, n = 6) or Escherichia coli lipopolysaccharide (LPS; 10 mg/kg iv) for 6 h. Groups of LPS-treated animals were pretreated with vehicle (, n = 11) or Dex (3 mg/kg iv at 30 min before LPS) followed by AG (15 mg/kg iv at 2 h after LPS; , n = 9). LPS was administered at time 0. Data are expressed as means ± SE of n observations. * P < 0.05, significant differences when compared with control at same time point.  P < 0.05, significant differences between treated and untreated LPS groups.

Partial inhibition of the production of nitrate in vascular smooth muscle by a combination of Dex plus AG in rats treated with endotoxin. There was almost undetectable nitrate in the endothelium-denuded aortas obtained from the SOP group (n = 5) (Fig. 2). Endotoxemia for 6 h caused a signifcant increase in aortic nitrate (n = 6) that was inhibited by pretreatment of LPS rats with Dex followed by AG (n = 5, P < 0.05). However, a slightly higher nitrate level was observed in the aorta obtained from the Dex plus AG-treated LPS rats than that from the SOP group (n = 5, P < 0.05).

View larger version (39K): [in this window] [in a new window]   Fig. 2.   Effects of Dex plus AG on production of nitrate in aortic smooth muscle obtained from rats with endotoxemia. Depicted are changes of aortic nitrate in rats subjected to saline (open column, n = 5; sham-operated control, SOP) or E. coli LPS (10 mg/kg iv) for 6 h. Groups of LPS-treated animals were pretreated with vehicle (cross-hatched columns, n = 6) or Dex (3 mg/kg iv at 30 min before LPS) followed by AG (15 mg/kg iv at 2 h after LPS) (hatched column, n = 5). Changes of aortic nitrate were measured in aortas without endothelium. Data are expressed as means ± SE of n observations. * P < 0.05, significant differences when compared with SOP group.  P < 0.05, significant differences between treated and untreated LPS groups.

Effects of Dex plus AG on the LPS-induced changes of MAP and HR in the anesthetized rat. At the end of the 20-min stabilization period, mean values for MAP ranged from 116 ± 3 to 119 ± 4 mmHg and for HR from 292 ± 13 to 309 ± 9 beats/min and were not significantly different among any of the animal groups studied. The injection of vehicle (saline) did not cause any significant changes on MAP and HR during the experimental period. In contrast, the injection of rats with LPS (10 mg/kg iv) induced a fall in MAP from 116 ± 3 to 96 ± 4 mmHg (n = 11, P < 0.05) within 15 min. At 1 h after LPS injection, MAP had recovered to 114 ± 5 mmHg. Thereafter, there was a significant further fall in MAP to 79 ± 6 mmHg at 6 h (n = 11, P < 0.05) (Fig. 3A). The injection of LPS was also followed by a substantial significant increase (at 2-4 h) in HR followed by a decrease to the control value (at 6 h) (Fig. 3B).

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Effects of Dex plus AG on mean arterial blood pressure (MAP) and heart rate (HR) in rats with endotoxemia. Depicted are changes in MAP (A) and HR (B) in rats subjected to saline (, n = 6) or E. coli LPS (10 mg/kg iv) for 6 h. Groups of LPS-treated animals were pretreated with vehicle (, n = 11) or Dex (3 mg/kg iv at 30 min before LPS) followed by AG (15 mg/kg iv at 2 h after LPS; , n = 9). LPS was administered at time 0. Data are expressed as means ± SE of n observations. * P < 0.05, significant differences when compared with control at same time point.  P < 0.05, significant differences between treated and untreated LPS groups.

Effects of Dex plus AG on the pressor responses to NE in vivo and ex vivo. To normalize values of the pressor response to NE to a similar level between each group, the pressor response to NE before the injection of saline or LPS (i.e., at time 0) was calculated as 100%. In the SOP group, the injection of vehicle had no significant effects on the pressor responses to NE during the experimental period. In contrast, LPS treatment significantly attenuated the pressor responses to NE (1 µg/kg iv) at 2, 4, and 6 h after its administration (n = 11, P < 0.05; Fig. 4). Administration of Dex did not cause significant changes of the pressor responses to NE (before, 38 ± 3 mmHg · min; after, 37 ± 3 mmHg · min). In addition, in Dex-pretreated animals subjected to LPS, Dex partially prevented the development of vascular hyporeactivity to NE at 2 h. When Dex-pretreated LPS-treated rats were treated with AG (at 2 h after LPS injection), the pressor responses to NE at 6 h after LPS injection were restored to 76 ± 5% (n = 9) of the pre-LPS values, whereas they remained reduced (42 ± 11%, n = 11) in LPS-treated animals (Fig. 4).

View larger version (49K): [in this window] [in a new window]   Fig. 4.   Effects of Dex plus AG on pressor responses to norepinephrine (NE) in rats with endotoxemia. Depicted are changes in pressor responses to NE (1 µg/kg iv) in rats subjected to saline (open columns, n = 6) or E. coli LPS (10 mg/kg iv) for 6 h. Groups of LPS-treated animals were pretreated with vehicle (crosshatched columns, n = 11) or Dex (3 mg/kg iv at 30 min before LPS) followed by AG (15 mg/kg iv at 2 h after LPS) (hatched columns, n = 9). Data are expressed as means ± SE of n observations. * P < 0.05, significant differences when compared with control at same time point.  P < 0.05, significant differences between treated and untreated LPS groups.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Effects of Dex plus AG on vascular tone induced by NE in rats with endotoxemia. Depicted are changes in contractile responses to NE (1 nM to 1 µM) in endothelium-denuded aortic rings obtained from saline-injected (, n = 6) or LPS-injected rats. Groups of LPS-treated animals were pretreated with vehicle (, n = 10) or Dex (3 mg/kg iv at 30 min before LPS) followed by AG (15 mg/kg iv at 2 h after LPS) (, n = 8). Data are expressed as means ± SE of n observations. * P < 0.05, significant differences when compared with control at same concentration.  P < 0.05, significant differences between treated and untreated LPS groups at same concentration.

Effects of L-NAME, ODQ, and methylene blue on aortic rings obtained from SOP rats, LPS rats, and LPS rats pretreated with Dex followed by AG. In vitro treatment of rings from LPS rats with L-NAME (0.3 mM) or ODQ (1 µM) partially enhanced the NE-induced contraction, whereas no further enhancement was observed in rings from Dex plus AG-treated LPS rats under such a treatment (Table 1). However, in the presence of methylene blue (10 µM), the contractile responses to NE (1 nM to 1 µM) in rings from both groups were completely restored to the values that are not different from those of the SOP group (Table 1). In contrast, the contractile responses to NE were not significantly affected by L-NAME, ODQ, or methylene blue in rings obtained from the SOP group (Table 1).

Effects of ODQ and methylene blue on vascular reactivity to NE in the presence of NAC in aortas from Dex plus AG-treated LPS rats. Figure 6 demonstrates that the vascular reactivity to NE is not affected by NAC (n = 6) in endothelium-denuded aortic rings obtained from the Dex plus AG-treated group. When the preparations were in the presence of NAC, the vascular reactivity to NE was not significantly enhanced by ODQ (1 or 10 µM; n = 6), but this was restored to normal by methylene blue (10 µM; P < 0.05, n = 6).

View larger version (19K): [in this window] [in a new window]   Fig. 6.   Effects of 1H-[1,2,4]oxidazolo[4,3-a]quinoxalin-1-one (ODQ) and methylene blue on vascular tone induced by NE in presence of N-acetyl-L-cysteine (NAC) in endotoxemic rats pretreated with Dex plus AG. Depicted are changes in contractile responses to NE (1 nM to 1 µM) in absence (, n = 6) or presence of NAC (10 mM) in endothelium-denuded aortic rings obtained from LPS-treated animals pretreated with Dex (3 mg/kg iv at 30 min before LPS) followed by AG (15 mg/kg iv at 2 h after LPS). Groups of aortic rings in presence of NAC were pretreated without (, n = 6) or with ODQ (, 1 µM, n = 6; , 10 µM, n = 6) or methylene blue (, 10 µM, n = 6). Data are expressed as means ± SE of n observations. * P < 0.05, significant differences when compared with group of aortic rings in absence of NAC.  P < 0.05, significant differences when compared with group of aortic rings in presence of NAC alone.

cGMP levels of aortic rings obtained from SOP rats, LPS rats, and Dex plus AG-treated LPS rats. The basal level of aortic cGMP was 28 ± 4 pmol/mg protein in the SOP group (n = 6) and was not affected by treatment of rings with L-NAME (n = 5), ODQ (n = 5), or methylene blue (n = 5) in vitro (Fig. 7). Six hours of endotoxemia was associated with a substantial increase in the aortic cGMP content (n = 10, P < 0.05), which was partially inhibited by L-NAME (n = 8) or ODQ (n = 6) and completely inhibited by methylene blue (n = 8), but not affected by tetraethylammonium (n = 4, data not shown) in vitro. This higher cGMP content caused by endotoxemia was significantly reduced in rings obtained from endotoxemic animals pretreated with Dex followed by AG (n = 8, P < 0.05). This reduced content was not attenuated by L-NAME (n = 7) or ODQ (n = 5) but was further inhibited by methylene blue (n = 7) to a level similar to normal (Fig. 7).

View larger version (28K): [in this window] [in a new window]   Fig. 7.   Effects of N-nitro-L-arginine methyl ester (L-NAME), ODQ, and methylene blue on aortic cGMP level in normal rats, endotoxemic rats, and endotoxemic rats pretreated with Dex plus AG. Depicted are changes in cGMP levels before (open columns) and after L-NAME (0.3 mM, crosshatched columns), ODQ (1 µM, hatched columns), or methylene blue (10 µM, solid columns) in endothelium-denuded aortas obtained from rats subjected to saline (SOP, n = 5 or 6) or E. coli LPS (10 mg/kg iv) for 6 h. Groups of LPS-treated animals were pretreated with vehicle (LPS, n = 6-10) or Dex (3 mg/kg iv at 30 min before LPS) followed by AG (15 mg/kg iv at 2 h after LPS) (Dex/AG/LPS, n = 5-8). Data are expressed as means ± SE of n observations. * P < 0.05, significant differences when compared with SOP group.  P < 0.05, significant differences between treated and untreated LPS groups under same condition. # P < 0.05, significant differences between before and after treatment of inhibitors in same group.

Effects of tetraethylammonium or charybdotoxin on aortic rings obtained from LPS rats or LPS rats pretreated with Dex then plus AG. In vitro treatment of rings from LPS rats with tetraethylammonium (1 mM) or charybdotoxin (0.1 µM), but not with apamin (1 µM; unpublished observations), partially enhanced the NE-induced contraction, whereas this contraction was further completely restored by the treatment of rings with a combination of L-NAME (0.3 mM) plus tetraethylammonium (n = 8, P < 0.05; Fig. 8A) or a combination of ODQ (1 µM) plus charybdotoxin (n = 6, P < 0.05; Fig. 8B). However, in rings obtained from Dex plus AG-treated LPS rats, the contractile responses to NE were also completely restored by tetraethylammonium (n = 7, P < 0.05; Fig. 9A) or charybdotoxin (n = 6, P < 0.05; Fig. 9B), but not significantly enhanced by ODQ (n = 5; Table 1).

View larger version (16K): [in this window] [in a new window]   Fig. 8.   Effect of tetraethylammonium or charybdotoxin on NE-induced contraction in rats treated with LPS alone. Depicted are changes of contractile responses to NE in absence or presence of tetraethylammonium (A; , 1 mM) or charybdotoxin (B; , 0.1 µM) in endothelium-denuded aortic rings obtained from rats subjected to LPS (10 mg/kg iv; , n = 8 in A and n = 6 in B). Contractile responses to NE were also performed on endothelium-denuded aortic rings obtained from rats subjected to saline only (i.e., normal controls, , n = 6). In A,  represents aortic rings in presence of L-NAME and tetraethylammonium (n = 8), whereas  in B represents aortic rings in presence of ODQ and charybdotoxin (n = 6). Data are expressed as means ± SE of n observations. * P < 0.05, significant differences when compared with control at same concentration. # P < 0.05, significant differences between absence and presence of tetraethylammonium or charybdotoxin at same concentration. § P < 0.05, significant differences between L-NAME plus tetraethylammonium and tetraethylammonium alone or between ODQ plus charybdotoxin and charybdotoxin alone at same concentration.

View larger version (14K): [in this window] [in a new window]   Fig. 9.   Effect of tetraethylammonium or charybdotoxin on NE-induced contraction in endotoxemic rats pretreated with Dex plus AG. Depicted are changes of contractile responses to NE in absence () or presence () of tetraethylammonium (1 mM; n = 7) (A) or in absence () or presence () of charybdotoxin (0.1 µM; n = 6) (B) in endothelium-denuded aortic rings obtained from LPS rats subjected to Dex (3 mg/kg iv at 30 min before LPS) followed by AG (15 mg/kg iv at 2 h after LPS). Contractile responses to NE were also performed on endothelium-denuded aortic rings obtained from rats subjected to saline only (i.e., normal controls;  in A, n = 6). Data are expressed as means ± SE of n observations. * P < 0.05, significant differences when compared with control at same concentration. # P < 0.05, significant differences between absence and presence of tetraethylammonium or charybdotoxin at same concentration.

High-K⁺-induced contractions on aortic rings obtained from LPS rats or LPS rats pretreated with Dex and then AG. Figure 10 shows that high K⁺ (90 mM) causes contractions in endothelium-denuded aortic rings obtained from SOP, LPS-treated, and Dex plus AG-treated LPS groups. There was no significant vascular hyporeactivity to high K⁺ in rings obtained from Dex plus AG-treated LPS rats when compared with SOP rats, whereas a reduction of high-K⁺-induced contraction was observed in preparations obtained from LPS-treated rats. In addition, in the presence of L-NAME (0.3 mM) or ODQ (1 µM), this hyporesponsiveness to high K⁺ was restored to normal. However, the treatment of rings with L-NAME or ODQ had no significant effects on the contraction induced by high K⁺ in SOP rats and Dex plus AG-treated LPS rats.

View larger version (48K): [in this window] [in a new window]   Fig. 10.   Effects of L-NAME or ODQ on vascular tone induced by high K+ in rats with endotoxemia. Depicted are changes in contractile responses to high K+ (90 mM) in endothelium-denuded aortic rings obtained from saline-injected (open columns, n = 5) or LPS-injected rats. Groups of LPS-treated animals were pretreated with vehicle (crosshatched columns, n = 9) or Dex (3 mg/kg iv at 30 min before LPS) followed by AG (15 mg/kg iv at 2 h after LPS) (hatched columns, n = 5-7). After control experiment was performed, i.e., in absence of L-NAME or ODQ, L-NAME (0.3 mM) or ODQ (1 µM) was added into bath for 15 min, and then high-K+-induced contraction was repeated. Data are expressed as means ± SE of n observations. * P < 0.05, significant differences when compared with sham control group.  P < 0.05, significant differences between treated and untreated LPS groups. # P < 0.05, significant differences between absence and presence of L-NAME or ODQ in LPS-treated alone rats.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

In this study, we successfully demonstrated that pretreatment of LPS rats with Dex followed by AG completely inhibits the production of nitrate (an indicator of NO) in the plama (Fig. 1), which is associated with a protection of delayed hypotension, but further enhances the tachycardia induced by LPS in the anesthetized rat. The enhancement of HR by Dex plus AG in LPS rats may be because of an inhibition of the production of NO, which plays an inhibitory role in sinus discharge rate and AV nodal conduction (8), and then decreases the HR. It has been shown that injection of LPS in rats resulted in expression of mRNA of Ca²⁺-independent NOS by myocytes cultured from the ventricular myocardium (2) and increased activity of Ca²⁺-independent NOS in homogenized ventricular myocardium (36). These results further support the speculation that inhibition of NO formation in cardiac tissues causes an increase of HR. However, this increased HR in LPS rats pretreated with Dex followed by AG did not further increase the MAP above the SOP group. This may be because of a partial, but not complete, restoration of pressor responses to NE by Dex plus AG. In addition, it is possible that the blood pressure might be restored in part by increased cardiac output resulting from reflexively elevated HR. Thus the HR will remain elevated to support blood pressure as long as some decrease in vascular resistance persists after Dex plus AG treatment. This explanation is also in accord with the hyporeactivity observed in vitro.

The vascular hyporeactivity to various endogenous vasoconstrictor agents is one of the most important factors that causes hypotension or death in septic animals or patients (14, 47). In the past decade, NO has been proposed to regulate the vascular tone importantly (24). Recently, growing evidence has shown that an increased NO formation by smooth muscle cells plays a substantial role in the decrease of total peripheral resistance after endotoxin application or hemorrhagic shock (7, 43). This increase is primarily because of the induction of a Ca²⁺-independent NOS, which has been demonstrated in most blood vessels (31, 37) or cultured vascular cells (7, 29) and is associated with hypotension and vascular hyporesponsiveness to various vasoconstrictor agents. However, more recent studies demonstrate that inhibition of NOS only partially restores the vascular hyporeactivity to NE caused by endotoxin ex vivo (19, 40, 45, 49). It is noteworthy that there is an NO-independent activation of soluble GC by interleukin-1 (IL-1) in vascular smooth muscle cells in vitro (3). In addition, treatment of rats with the endogenous IL-1 receptor antagonist (IL-1ra) in vivo before endotoxin treatment reduced the vascular hyporeactivity of the thoracic aorta ex vivo, and this effect is only in part because of prevention of the induction of NOS by IL-1ra (40). Thus the production of IL-1 by LPS in vivo is associated with the pathway of NO-independent activation of soluble GC. There is evidence that an enhanced formation of oxygen free radicals in endotoxin shock is also observed (12), and these free radicals may also activate soluble GC (32). Indeed, our previous results have demonstrated that activation of soluble GC by factor(s) other than NO or CO contribute to the vascular hyporeactivity to NE in the aortas of rats treated with endotoxin (49).

Here, we further confirm this hypothesis by in vivo studies showing that Dex plus AG completely inhibited the overproduction of NO in rats treated with LPS, but this combination only partially restores the hyporeactivity to NE in vivo and ex vivo. This vascular hyporeactivity, however, was not improved by in vitro treatment of aortic rings with L-NAME, an inhibitor of NOS. In contrast, in aortic rings obtained from rats treated with LPS only, the vascular hyporeactivity to NE was enhanced by inhibitors of NOS in vitro (14, 49). There are two possibilities that should be considered concerning L-NAME-insensitive NO formation contributing to vascular hyporeactivity. One possibility is that the dose of Dex plus AG and L-NAME applied in this study may not be sufficient to completely inhibit vascular NO formation. This is unlikely, since the L-NAME-insensitive NO formation was inhibited by methylene blue, although methylene blue had been regarded as an inhibitor of NOS (21). The latter argument can be clarified by using ODQ, an inhibitor of NO-sensitive GC (11). Our results demonstrated that ODQ did not further significantly improve the vascular hyporeactivity to NE in rings from the Dex plus AG-treated LPS rat, suggesting that this is not because of insufficient dose of Dex plus AG on the rat. The other possibility is LPS-induced formation of slowly decomposing NO storage form in the aorta (26), which may account for L-NAME-insensitive, cGMP-dependent vasorelaxation. Our results demonstrated that the nitrate level in the aorta (without the endothelium) homogenate determined by the NO analyzer was slightly, but significantly, higher in Dex plus AG-treated LPS rats when compared with that in the SOP control group (Fig. 2). It is quite possible that NO generated independently of NOS contributes to vasorelaxation. To examine the functional effect of NO that reacts with non-heme iron to form protein-bound dinitrosyl nonheme iron complexes (DNIC) on vascular reactivity, NAC, an agent that decomposes the DNIC (26), was incubated with rat aortas to evaluate the NE-induced contraction in vitro. However, the vascular hyporeactivity to NE in aortas from Dex plus AG-treated LPS rats was not significantly enhanced in the presence of NAC. In addition, this response was not attenuated by ODQ (even at 10 µM) but was further restored to the normal level by methylene blue. Thus the NO stored as protein-bound DNIC may only contribute to a minor component in vascular hyporeactivity induced by endotoxemia. This suggestion is supported by the fact that ODQ did not completely inhibit the aortic cGMP level elevated by the residual NO from pathways independent of NOS.

In vitro methylene blue, an inhibitor of soluble GC, completely restored this hyporeactivity in both groups, suggesting that hyporesponsiveness to NE caused by LPS is likely due to activation of soluble GC, which is partially mediated by NO. This hypothesis was confirmed by our results of aortic cGMP content, which manifested that in aortas from rats treated with LPS, the cGMP level was significantly increased and partially attenuated by L-NAME or ODQ in vitro, whereas this was completely inhibited by methylene blue in vitro (Fig. 6). Treatment of LPS rats with Dex plus AG also significantly prevented the formation of cGMP in aortas. The residual cGMP level was not further decreased by in vitro treatment of these rings with L-NAME or ODQ, but this was completely inhibited by methylene blue. Indeed, it has been shown that in vivo administration of methylene blue was able to restore both vascular cGMP and pressor responses to NE to control levels in LPS-treated rats and in patients with septic shock (28, 33). Although some studies showed that methylene blue prevented the activation of soluble GC by generating superoxide anion (46) or hydroxyl radical (17), however, the latter is decomposed from H₂O₂ in the presence of free metal ions (e.g., iron) in the cell (34). A recent important finding demonstrated that H₂O₂ activated glibenclamide-sensitive K⁺ channels, suggesting that activation of the K⁺ channel contributes to oxidant injury (also seen in sepsis) (9).

In addition, several studies have demonstrated that cGMP activates the K_Ca channel (1, 23, 41) because cGMP-dependent, but not cGMP-independent, vasodilatation is inhibited by tetraethylammonium, an inhibitor of classical K⁺ channel, and charybdotoxin (but not apamin), an inhibitor of large-conductance K_Ca channels. Indeed, Yao et al. (51) have recently reported that a gene encoding a K⁺ channel (Kcn1) isolated from rabbits is specifically regulated by cGMP and plays an important role in mediating the effects of substances, such as NO, which increase intracellular cGMP. In contrast, one study showed that NO directly activated K_Ca channels in cell-free membrane patches without requiring cGMP (5), suggesting that there are at least two types of K_Ca channels present in the vascular smooth muscle cell (i.e., one is regulated by cGMP, and the other is regulated by NO). In the present study, we demonstrate that the pressor response to NE in rings obtained from LPS rats is partially enhanced by tetraethylammonium or charybdotoxin in vitro, whereas that response in rings from Dex plus AG-treated LPS rats is further restored to the level as observed in the SOP group. Similarly, the vascular hyporeactivity to high K⁺ was observed in aortic rings obtained from LPS-treated rats but not from Dex plus AG-treated LPS rats. This hyporesponsiveness to high K⁺ in LPS rats was further restored to normal by L-NAME or ODQ in vitro. In addition, this restoration is similar to the effect of methylene blue on the vascular hyporeactivity to NE in rings obtained from LPS rats. These results strongly indicate that LPS directly or indirectly causes vasorelaxation by activation of large-conductance K_Ca channels, which can be activated by cGMP and inhibited by tetraethylammonium, charybdotoxin, or high K⁺.

One could argue that other K⁺ channels, e.g., ATP-sensitive K⁺ channels (K_ATP channels) activated by the hydroxyl radical, were also regulated by cGMP (18). This, however, is not the case, because in vitro glibenclamide, a specific inhibitor of K_ATP channels, did not ameliorate the vascular hyporeactivity to NE in rings obtained from rats treated with endotoxin (50). Therefore, persistent activation of K_Ca channels (most likely) either by cGMP or NO, resulting in vasodilatation, contributes to the mechanism of endotoxin-induced septic shock, which is associated with hypotension, peripherial vasodilatation, and a reduced response to vasoconstrictor agents (38). In addition to the production of NO from macrophages, vascular smooth muscle cells, and glomerular mesangial cells by LPS or cytokines (35), other soluble GC-activating factor(s) seems also to be released by the stimulation of above cells or animals with endotoxin.

In conclusion, regardless of the precise mediator(s), in addition to NO, of endotoxin-induced septic shock, the activation of large-conductance K_Ca channels seems to be one of the mechanisms accounting for the hypotension, peripherial vasodilatation, and vascular hyporeactivity to vasocontrictor agents in animals treated with endotoxin. Therefore, it should be noted that inhibition of the activation of soluble GC elicited by NO or any soluble GC-activating factor produced in the blood vessel of animals or patients with endotoxemia would be a novel class of agents for the therapy of septic shock.