| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Department of Anesthesia and Intensive Care, Lund University Hospital, Sweden
Address correspondence and reprint requests to Mikael Bodelsson, MD, PhD, Department of Anesthesia and Intensive Care, Lund University Hospital, SE-221 85 Lund, Sweden. Address e-mail to mikael.bodelsson{at}anest.lu.se
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
IMPLICATIONS: Anesthesia with sevoflurane is accompanied by vasodilatation. We studied the effects of sevoflurane on isolated human arteries and veins. In contrast to previous animal studies, our results show that sevoflurane at larger concentrations promotes endothelium-dependent vasodilatation via an enhancement of the vascular smooth muscle response to relaxing second messengers.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Studies have shown that volatile anesthetics inhibit endothelium-dependent relaxation in conductance and resistance arteries in vitro, but the underlying mechanisms seem to differ between the different compounds (7–10). Sevoflurane attenuates the relaxation induced by acetylcholine (ACh) or histamine in rat aorta (8), rabbit and rat mesenteric artery (9–11), and rabbit carotid artery (12). In these preparations, ACh and histamine stimulate the endothelium to release NO and EDHF, and both pathways are inhibited by sevoflurane.
The aim of the present study was to investigate the effect of sevoflurane on endothelium-dependent relaxation in human omental arteries and veins. It has been shown that substance P (SP) relaxes isolated human omental vessels endothelium-dependently via both NO synthesis and hyperpolarization (13). SP was therefore chosen as a stimulator of endothelium-dependent smooth muscle relaxation in this study.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The vessel segments were cut into 2–4-mm-long ring segments and placed into 2-mL organ baths on 2 steel rods through the lumen. One of the rods was attached to a Grass FTO3C force displacement transducer (Grass Medical Instruments, Quincy, MA) for measurement of isometric force. The force was recorded on a Grass polygraph model 7D (Grass Medical Instruments). Six segments in separate organ baths with KRS were run in parallel. The temperature in the baths was thermostatically maintained at 37°C, and the KRS was continuously aerated with a gas mixture containing 92% O2 and 8% CO2 at a rate giving PCO2 5 kPa and a pH value of 7.4 (14). The vessel segments were gradually stretched to a stable resting force of 6 mN during an equilibration period of 60–90 min to obtain the optimal tension level (13). Then KCl (9 x 10–2 M) was added, and the resulting contraction was registered. Subsequently, the KCl was washed out by repetitive changes of the KRS during which the contraction was allowed to return to baseline.
Sevoflurane was vaporized by Sigma Elite vaporizers (Sigma, St Louis, MO) and administered via the O2/CO2 gas mixture. The resulting concentration of sevoflurane in KRS was analyzed with a gas-liquid chromatograph using a head space technique (14) and was adjusted as to achieve a sevoflurane concentration in the KRS corresponding to 1%, 2%, or 4%. The minimum alveolar anesthetic concentration (MAC) for sevoflurane in adults is 1.48%–2.52% (15).
First, we investigated the effect of sevoflurane on the relaxation induced by SP. Sevoflurane 0%, 1%, 2%, or 4% was administered to separate organ baths. Endothelin-1 (ET-1; 10–9 M) was added, and the resulting contraction was registered. If required, the ET-1 concentration was increased stepwise until the contraction was equal to 70%–130% of the contraction induced by 9 x 10–2 M of KCl. Sevoflurane at 1%, 2%, or 4% did not affect the ET-1 concentration required to achieve this contraction level. Within 10 min, the contraction reached a stable plateau (precontraction), and at least 10 min after the introduction of sevoflurane, SP was added cumulatively in log10 units (10–12–10–6 M). The resulting relaxation was registered, and concentration-response curves for SP were constructed. The endothelium was regarded as damaged if SP did not induce relaxation in the control segment, and the patient was excluded from the study. Pilot experiments had shown that a consistent relaxation response to SP was obtained for two subsequent concentration-response experiments separated by washout. If the experiment was repeated a third time on the same segment, the response to SP was decreased because of tachyphylaxis. Thus, we could not study the effect of all three sevoflurane concentrations on each segment. We therefore used the protocol with different sevoflurane concentrations in different organ baths.
In a second series of experiments on the same segments as above, the effect of sevoflurane on the two components of SP-mediated relaxation was investigated. After thorough washout during at least 40 min, relaxation experiments with SP on ET-1 precontracted segments were repeated, as previously described, in the presence of the NO synthase inhibitor, L-NG-nitroarginine methyl ester (L-NAME; 3 x 10–4 M), and/or KCl (3 x 10–2 M), which prevents hyperpolarization of the vascular smooth muscle (16). L-NAME and KCl were added to all the organ baths at least 10 min before the addition of SP.
In a third series of experiments, the endothelium was removed by gentle injection of the O2/CO2 gas mixture through the vessel lumen for 5 min during the equilibration period (17). To confirm that the endothelium had been successfully removed, SP (10–7 M) was added to the vessel segments precontracted with ET-1. If SP induced relaxation, the gas injection was repeated, and the presence of endothelium was subsequently tested until SP did not induce any relaxation. The organ baths were rinsed several times. The effect of sevoflurane (0%, 1%, 2%, or 4%) on SP-induced relaxation in the denuded vessels was then investigated in the presence of both L-NAME and KCl.
In a fourth series of experiments, the effect of sevoflurane (0%, 1%, 2%, or 4%) on the endothelium-independent relaxation induced by glyceryl trinitrate (GTN; 10–10–10–4 M) was studied in the same manner as for SP in the first series. The relaxation experiments were performed in the presence of KCl. In a separate series, similar experiments were also performed in endothelium-denuded segments.
Segments of omental artery, 4–5 mm in length, were equilibrated in 2 mL of aerated KRS with KCl (3 x 10–2 M) for 30 min. Sevoflurane (0%, 1%, 2%, or 4%) was thereafter administered to separate vessel segments for 10 min. SP (10–7 M) was added, and after 90 s, the segments were frozen in liquid nitrogen and kept at –80°C. The segments were ground to a fine powder under liquid nitrogen. The powder was dispersed in 200 µL of ice-cold hydrochloric acid 0.1 M followed by centrifugation at 11,000g for 5 min at room temperature. The content of guanosine 3', 5'-cyclic monophosphate (cGMP) in the supernatant was measured with enzyme-linked immunosorbent assay (Direct Cyclic GMP Enzyme Immunoassay Kit, Assay Designs Inc, Ann Arbor, MI) after acetylation in accordance with the manufacturer’s recommendations. Total protein concentration in the supernatants was determined on the basis of the Biuret reaction combined with colorimetric detection of the cuprous cation using a reagent containing bicinchonic acid (BCA Protein Assay Kit, Pierce, Rockford, IL). The levels of cGMP in the vessel segments are expressed as picomole per milligram protein.
The following compounds were used: SP acetate, L-NAME, ET-1 (Sigma), sevoflurane (Abbott, Sweden), and GTN (Orion Pharma, Sollentuna, Sweden). All substances were dissolved or diluted in distilled water. The KRS contained (mM): Na+ 143, K+ 4.6, Cl– 126, Ca2+ 2.5, HCO3– 25.0, Mg2+ 0.79, SO42- 0.79, H2PO4– 1.2, glucose 5.5, and EDTA 0.024.
The SP-induced relaxation is expressed as percentage of the precontraction induced by ET-1. When the addition of SP did not affect the ET-1-induced precontraction, the relaxation is given as 0%. When the addition of SP relaxed the vessel segments to the tension level before the addition of ET-1, the relaxation is given as 100%. When values from more than one similar experiment were obtained from the same patient, the mean was calculated before further analysis and presentation, and therefore, the number of determinations (n) equals the number of patients. The –log10 concentration required to achieve half-maximum relaxation (pEC50) was calculated using linear interpolation in the concentration-response curve. All values are expressed as mean ± SEM. Two-way repeated-measurement analysis of variance (ANOVA), followed by Dunnett post hoc test, was used to compare the concentration-response curves. One-way repeated-measurement ANOVA was used for evaluation of differences in pEC50 values. Friedman repeated-measures ANOVA on ranks followed by Dunnett post hoc test was used for the evaluation of the levels of cGMP. P < 0.05 was considered statistically significant.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
|
GTN induced a relaxation in the artery segments that was enhanced by sevoflurane at 4% compared with control in intact segments and by sevoflurane at 1% and 4% in endothelium-denuded segments (not shown; n = 6 and 5 for intact and endothelium-denuded segments, respectively).
As in the artery segments, sevoflurane in the absence of L-NAME and KCl did not affect the SP-induced relaxation in the vein segments compared with control (not shown). In the presence of L-NAME, the SP-induced relaxation was also unaffected by sevoflurane (Fig. 2A).
However, in the presence of KCl, sevoflurane at 4% enhanced the SP-induced relaxation (Fig. 2B). In the presence of both L-NAME and KCl, sevoflurane at 2% and 4% enhanced the relaxation induced by SP (Fig. 2C).
GTN induced a relaxation in the vein segments that was enhanced by sevoflurane at 2% and 4% compared with control in the intact segments and by sevoflurane at 4% in the endothelium-denuded segments (not shown; n = 5 for both intact and endothelium-denuded segments).
SP did not relax endothelium-denuded vessel segments, neither in the absence nor in the presence of sevoflurane (not shown; n = 3 arteries and n = 3 veins). Sevoflurane did not affect the pEC50 values for SP and GTN, neither in the arteries nor in the veins.
SP increased the levels of cGMP in the artery segments to 4.4 ± 1.8 pmol/mg compared to control (0.2 ± 0.04 pmol/mg). Sevoflurane did not affect the SP-induced increase in the cGMP levels (not shown; n = 7).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Sevoflurane did not affect the SP-induced relaxation in arteries and veins not treated with any inhibitor of endothelium-dependent relaxation. In the presence of the NO synthase inhibitor L-NAME, the relaxation induced by SP in human omental arteries and veins is primarily mediated via EDHF (13), which induces hyperpolarization caused by the opening of K+ channels in the smooth muscle cell membrane (20). Sevoflurane enhanced the SP-induced relaxation in the arteries in the presence of L-NAME. This is contrary to the findings in arteries of experimental animals. Akata et al. (9) demonstrated that sevoflurane at 1 MAC inhibits both the NO-mediated and the hyperpolarization-mediated components of the ACh-induced relaxation in rabbit mesenteric arteries. In the rabbit carotid artery, sevoflurane selectively inhibits the ACh-induced release of EDHF (12).
In the presence of KCl (3 x 10–2 M), the relaxation induced by SP in human omental arteries and veins is primarily caused by endothelial release of NO (13). Sevoflurane enhanced the SP-induced relaxation in both the arteries and the veins in the presence of KCl. This is also in contrast to the findings in animal models. Nakamura et al. (8) showed that sevoflurane probably inactivates NO or inhibits its action in the rat aorta. Furthermore, Yamaguchi and Okabe (10) demonstrated that sevoflurane attenuates the endothelium-dependent relaxation produced by ACh in rabbit mesenteric artery. This attenuation was almost abolished in the presence of superoxide dismutase. They concluded that sevoflurane generates oxygen-free radicals that inactivate NO directly.
In the presence of both L-NAME and KCl, the SP-induced relaxation was small. However, the relaxation induced by SP was still enhanced by sevoflurane at 4% in the arteries and by sevoflurane at 2% and 4% in the veins. Alternative explanations for this observation could be: (a) SP induces another endothelial vasoactive factor different from NO and EDHF, such as prostacyclin, in the presence of sevoflurane, (b) in the presence of sevoflurane, SP interacts with tachykinin receptors on the vascular smooth muscle cells, and (c) there is still a small fraction of NO and EDHF released from the endothelium. Wallerstedt and Bodelsson (13) demonstrated that prostanoid synthesis is not involved in the SP-induced relaxation in human omental vessels. Furthermore, a recent study has shown that sevoflurane reduces the prostacyclin production of cultured human umbilical vein endothelial cells (21).
After removal of the endothelium, SP did not relax the vessels, neither in the absence nor in the presence of sevoflurane. This excludes the possibility that SP interacts with tachykinin receptors on the vascular smooth muscle cells in the presence of sevoflurane. Furthermore, it supports the hypothesis that small amounts of NO and EDHF are released from the endothelium even in the presence of both L-NAME and KCl and that sevoflurane enhances the release or action of them.
It is not clear why sevoflurane affects the SP-induced relaxation only in the presence of inhibitors of the relaxation. A plausible reason could be that each of the NO and EDHF systems mediates a relaxation that is more than half of the total relaxing capacity in this model. In other words, the two systems seem to back up for each other. As demonstrated, both systems have to be inhibited for a major effect to be registered (13). This could indicate the presence of a relaxation reserve, i.e., less than a maximum release of endothelium-derived relaxing substances is required to mediate a maximum relaxation. In the absence of inhibitors, any further enhancement of the relaxation, such as in the presence of sevoflurane, might be obscured by the massive relaxation produced by the relaxation reserve. Thus, the vasodilating effects of sevoflurane might be more pronounced in vascular beds with an impaired endothelial function, such as with atherosclerosis (22).
At this stage, we could not tell whether sevoflurane acts on endothelial release of NO and EDHF or if it modulates the smooth muscle response to these compounds. NO relaxes vascular smooth muscle cells via activation of intracellular guanylyl cyclase, which catalyzes the production of the relaxing second messenger cGMP. GTN relaxes smooth muscle after transformation to NO via a direct action on guanylyl cyclase in the muscle cells without any involvement of the endothelium (23). We performed the experiments with GTN in the presence of KCl (3 x 10–2 M) to isolate the direct action on guanylyl cyclase because NO (exogenous and endogenous) can open Ca2+-dependent K+ channels directly without the involvement of cGMP and thereby also causing hyperpolarization (24). Sevoflurane at 4% enhanced the relaxation induced by GTN compared with control in both the arteries and the vein segments regardless of the presence of the endothelium. This is also in contrast to the results obtained from animal models in which the endothelium-independent relaxation induced by NO-donors such as GTN and sodium nitroprusside is not affected by sevoflurane (8–10,12). It suggests that, in human arteries and veins, the site of action of sevoflurane is not located in the endothelium, but rather in the smooth muscle cells where it enhances the action of relaxing mediators such as cGMP. This was supported by our observation that sevoflurane did not affect the SP-induced production of cGMP.
The net effect of an anesthetic on arterial blood pressure depends on the sum of the simultaneous actions of the anesthetic on, for example, myocardial contractility, central cardiovascular regulatory mechanisms, sympathetic neuromuscular transmission, vascular smooth muscle, and endothelium. The vasodilatation induced by sevoflurane in humans seems to be due not only to an effect on the smooth muscle cell contractility (14), but also to effects on the endothelium-mediated relaxation, as demonstrated in the present study. It must be noted that the effects of sevoflurane on arteries demonstrated in the present study were found to be significant only at the largest concentration used (4%). This corresponds to approximately 2 MAC in humans, which is in the upper range of the clinically used concentrations. Furthermore, the arteries used are too large to contribute significantly to the regulation of the systemic vascular resistance, implying that the results must be interpreted with caution. However, in some of the vein experiments, 2% sevoflurane was enough for a significant effect to be registered. This could indicate that the effects of sevoflurane on veins are clinically more relevant than the effects on arteries. The veins used in the present study indeed represent functional capacitance vessels.
In conclusion, we have demonstrated that sevoflurane promotes endothelium-dependent relaxation in human omental arteries and veins, probably via an enhancement of the response of the smooth muscle cells to relaxing mediators.
|
Acknowledgments |
|---|
The authors are grateful to Cristina D. Ciornei, MSc, and Ingrid Berkestedt, MD, for excellent technical assistance.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  I. Berkestedt, A. Nelson, and M. Bodelsson Endogenous antimicrobial peptide LL-37 induces human vasodilatation Br. J. Anaesth., June 1, 2008; 100(6): 803 - 809. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
