| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
Department of Anesthesiology, Medical College of Wisconsin, Milwaukee, Wisconsin
Address correspondence and reprint requests to Anthony Hudetz, MD, PhD, Department of Anesthesiology, Medical College of Wisconsin, 8701 Watertown Plank Rd., Milwaukee, WI 53226. Address e-mail to ahudetz{at}mcw.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Implications: Excess nitric oxide and prostaglandins after lipopolysaccharide treatment in the brain augment halothane-induced increases in cerebral blood flow.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Earlier studies indicating a role for NO in volatile anesthetic-induced cerebral vasodilation were targeted to the constitutive isoforms of NO synthase (NOS) (5–7). Results from most of the studies obtained with NOS inhibitors were interpreted according the classical signal transduction hypothesis stating that the anesthetic-induced vasodilation was achieved by an increase in NO production via increased NOS activity.
In parallel with these investigations, a significant body of literature has been published suggesting that in certain conditions, for example hypercapnia, NO may play a different, "permissive" role (8–10). The permissive role means that, in the presence of a NOS inhibitor, administration of a NO donor or cyclic guanosine 3',5'-monophosphate (cGMP) analog in suitable concentrations can restore the normal cerebrovascular reactivity (11). Such results have been seen repeatedly with various vasodilators. These results indicate that vasodilation was not mediated by the classical "enhanced NO generation" mechanism, but through another yet uncertain signal transduction mechanism that required a certain level of NO to be present to produce the normal degree of vasodilation.
One should note that the previously postulated "permissive" role for NO does not necessarily imply that NO acts by an "all-or-none" type gating mechanism for vasodilatation (to an agent other than NO itself), but it may be compatible with a modulatory or "gain-control" mechanism. There has been no systematic study to determine whether NO donors exert a modulatory effect of cerebral vasodilation caused by another agent. In particular, it is unclear whether an increase of NO concentration more than the normal level may influence the cerebral vasodilatory response to other agents such as volatile anesthetics.
A condition in which this effect may be relevant is endotoxemia. Under conditions that follow infection or trauma, inducible NOS (iNOS) is increased (12). After treatment with endotoxin, NO and products of inducible cyclooxygenase (COX-2) elicited vasodilation of the cerebral vascular bed (13–15). However, it is unclear whether endotoxin-induced increases in brain NO levels alter the cerebrovascular response to volatile anesthetics.
One could speculate that if NO played a graded modulatory role in volatile anesthetic-induced vasodilation, then an increase of NO might augment the vasodilatory response to anesthetics. This modulatory effect of NO would not depend on the calcium-calmodulin mediated activation of constitutive NOS and would be exerted regardless of the enzymatic source of NO (notwithstanding that differences in compartmentalization might influence the results). Preliminary data obtained in our laboratory using moderate supernormal doses of a NO donor support the plausibility of this hypothesis. Alternatively, one could hypothesize that the large amounts of NO and COX-2-derived mediators would produce maximal or near-maximal cerebral vasodilation, thus preventing a further dilatory effect of volatile anesthetics. The effect of supernormal NO concentration such as that seen after iNOS expression cannot, therefore, be predicted, and is of interest for experimental study.
We examined the cerebral hyperemic effect of halothane at various times after iNOS/COX-2 induction by intracerebroventricular administration of bacterial lipopolysaccharide (LPS). We also tested the effects of the iNOS inhibitor, aminoguanidine (16), the COX-2 inhibitor, NS398 (17), or the enzyme expression inhibitor, dexamethasone (18), on the regional cerebrocortical blood flow (rCBF) response to halothane after LPS. Because LPS releases vasodilators other than NO, additional experiments were performed to determine whether a direct acting NO donor, diethylamine NONOate (DEA-NONOate), potentiates halothane-induced hyperemia.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
General Surgical Procedures
Experiments were performed in male Sprague-Dawley rats weighing between 265 to 350 g. Anesthesia was induced by intraperitoneal (i.p.) sodium-pentobarbital (60 mg/kg; Sigma Chemical Co., St. Louis, MO). The animals were tracheostomized, paralyzed with pancuronium bromide (1 mg/kg i.p.), and artificially ventilated (SAR-830; CWE, Ardmore, PA) with 30% O2 in N2. Anesthesia was maintained throughout the experiment by inhalation of 0.6% halothane (Anaquest Inc., Madison, WI). Body temperature was maintained at 37° ± 1°C by using a water-circulated heating mat. One of the femoral arteries was cannulated to facilitate the measurement of arterial pressure and arterial blood gases. Arterial PO2, PCO2, and pH were measured with a blood gas/pH analyzer (ABL-300; Radiometer, Brønshøj, Denmark). A femoral vein was cannulated for the infusion of drugs. Arterial blood pressure, end-tidal carbon dioxide tension (ETCO2), inspired and expired oxygen, and halothane concentrations were monitored continuously. As reported previously (8), a 30-gauge stainless steel cannula (HTX-30; Small Parts, Miami Lakes, FL) was placed into the left lateral ventricle for intracerebroventricular (ICV) injection with the bregma chosen as the stereotaxic point (AP = -0.3 mm, LAT = +1.2 mm, DV = -4.5 mm). ICV infusion was performed at the rate of 1 µL/min using a microinfusion pump (model 55–2222; Harvard Apparatus, Holliston, MA) with microsyringe (25 µL, Hamilton Company, Reno, NV) fitted with polyethylene tubing (PE-10). We have demonstrated previously (8) that Evan’s blue dye injected via this route is uniformly distributed over the bilateral cortical surface within a few minutes after the injection. After surgery, a 1-h equilibration period was allowed for stabilization of the preparation.
Assessment of Cerebral Blood Flow
Regional cerebrocortical perfusion was measured by laser-Doppler flowmetry (PF3; Perimed, Stockholm, Sweden). Although the perfusion is not strictly equivalent to rCBF, we used the cerebrocortical perfusion for the assessment of rCBF in this study. The techniques have been described previously (8). Briefly, the head of the rat was placed in a stereotaxic apparatus (model 900; David Kopf, Tujunga, CA). A burr hole of approximately 2.0-mm diameter was created in the right parietal cranium using a low-speed air drill until only a thin translucent plate of cranium remained. The LDF probe (PF316, Perimed, Jarfalla, Sweden), with a tip diameter of 1 mm, was lowered into the well using a micro manipulator without touching the thin bone plate and was positioned in an area devoid of visible pial vessels. Proper placement of the probe was indicated by perfusion values in the range of 110 to 160 perfusion units. A drop of mineral oil was used to provide optical coupling between the LDF probe and the tissue.
Experimental Protocols
Rats were assigned to the following experimental groups. Group 1 (n = 5) received ICV artificial cerebrospinal fluid (ACSF) alone. Rats in Group 2 (n = 8) received ICV LPS (Ecoli, 055:B5; Sigma) at 1 mg/kg. LPS was dissolved in ACSF. The chosen dose of 1 mg/kg LPS was larger than that used in previous reports, 200 ng of LPS intracisternally (19); 100 ng/mL LPS was superfused into cranial windows (15). We chose this dose for ICV injection to achieve sufficient LPS concentration over the cerebral cortex where rCBF was measured based on our former study (13). In that study, we found that the a 1-mg/kg dose yielded a maximal effect without concurrent systemic hypotension. In Group 3 (n = 6), the iNOS selective inhibitor, aminoguanidine (100 mg/kg; RBI, Natick, MA), was injected i.p. 2 h before and 2 h after ICV administration of LPS (1 mg/kg). Group 4 (n = 6) was treated with the selective COX-2 inhibitor, NS-398 (5 mg/kg; BIOMOL, Plymouth Meeting, PA) i.p., 2 h before and 2 h after ICV administration of LPS (1 mg/kg). In Group 5 (n = 5), the rats received dexamethasone (4 mg/kg i.p.; Sigma) 4 h before, and immediately after, ICV administration of LPS (1 mg/kg) to prevent induction of iNOS and COX-2.
Four hours after ICV injection of LPS, the rCBF responses to 1.0 minimum alveolar anesthetic concentration (MAC) and 1.5 MAC halothane were tested (1.1 and 1.6 vol%) (20). Each anesthetic concentration was maintained for approximately 10 min. Arterial blood pressure was supported during incremental halothane administration by an infusion of methoxamine (2–20 nmol/min) as described previously (2,7).
Another series of experiments was performed in 12 rats, wherein the rCBF responses to 1.0 MAC and 1.5 MAC halothane were tested in the presence of ICV ACSF (1 µL/min, n = 7), the NO donor, DEA-NONOate (1 mM in ACSF at a rate of 1 µL/min, n = 7), and NO-independent vasodilator, adenosine (5 mM in ACSF at a rate of 1 µL/min, n = 5) (21). The effect of DEA-NONOate lasted for approximately 30 min, and for adenosine, it lasted for 15 min. Each treatment was applied in each rat in alternating order at 1 h apart. Aminoguanidine, NS-398, and dexamethasone were dissolved in 1 mL of peanut oil using a sonicator. We previously reported that the injection of peanut oil alone had no effect on rCBF (8), and also confirmed in preliminary experiments that aminoguanidine, NS-398, and dexamethasone had no effect on rCBF over time.
Expression of iNOS and COX-2
Total RNA was isolated from the cerebral cortex by using TRIzol reagent (GIBCO BRL, Gaithersburg, MD). The concentration of total RNA in each sample (n = 4) was measured by a spectrophotometer at a wavelength of 260 nm (GeneQuant; Pharmacia Biotech, Piscataway, NJ). The RNA was reverse transcribed (RT) by incubating 1 µg of RNA for 40 min at 42°C with 2.5 U/µL MuLV reverse transcriptase with 2.5 µmol/L random hexamers, 1 mmol/L dNTP, 5 mmol/L MgCl2, and 1 U/µL RNase inhibitor (GeneAmp, Perkin-Elmer, Foster City, CA) in a volume of 10 µL. The entire RT-reaction was amplified by polymerase chain reaction (PCR) in a 50 µL PCR reaction containing 2 mmol/L MgCl2, 1.25 U Taq DNA polymerase, and 0.2 µmol/L of specific primers for iNOS, COX-2, or glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as a control for RT-reaction. The sequences of these primers (Operon, Alameda, CA) have been reported previously (22,23) and were as follows: iNOS forward, 5'-ACAACGTGGAGAAAACCCCAGGTG-3'; iNOS reverse, 5'-ACAGCTCCGGGCATCGAAGACC-3'; COX-2 forward, 5'-GAAGTGGGGGTTTAGGATCATC-3'; COX-2 reverse, 5'-CCTTTCACTTTC- GGATAACCA-3'; GAPDH forward, 5'-CACGGCAAG- TTCAATGGCACA-3'; GAPDH reverse, 5'-GAATT- GTGAGGGAGAGTGCTC-3'. The iNOS reactions were cycled 35 times at 96°C for 30 s, 65°C for 60 s, and 72°C for 90 s and yielded a single band corresponding to a 565 base pair cDNA fragment. The COX-2 reactions were cycled 35 times at 96°C for 30 s, 60°C for 60 s, and 72°C for 90 s and yielded a single band corresponding to a 381 base pair cDNA fragment. The GAPDH reactions were cycled by using the same conditions as iNOS or COX-2, and produced a single band corresponding to a 970 base pair cDNA fragment. RNA extracted from the spleen of rats treated with LPS (10 mg/kg i.p.) was used as positive control of either iNOS or COX-2 gene expression. Twenty microliters of the RT-PCR products was electrophoresed on a 1% agarose gel and visualized by ethidium bromide staining under ultraviolet light. The ratio of the intensities of iNOS or COX-2 bands was assessed by a fluoroimager (Vistra, Sunnyvale, CA) and normalized with the intensity of GAPDH band as reported previously (24).
Measurement of the iNOS Activity
Calcium independent (iNOS) activity was measured by the conversion of (3H)-L-arginine to (3H)-L-citrulline by using a radioactive HPLC method originally described by Carlberg (25). Cerebral cortical tissue of each sample (n = 4) was homogenized in 20 mM N-2 hydroxyethelpiperazine-N-2-ethanesulfonic acid (HEPES) buffer (pH 7.4). After centrifuging the homogenates at 9000g for 10 min at 4°C, aliquots of homogenate (150 µg protein) were incubated with (3H)-L-arginine (0.2 µCi, 20 µM; Amersham, Arlington Heights, IL) in 100 µL of 20 mM HEPES calcium-free buffer containing 0.5 mM ethyleneglycotetraacetic acid, 1 mM NADPH, 2.5 µM FAD, 1 µM FMN, and tetrahydrobiopterin for 5 min at 37°C, and reaction was stopped by adding 50 µL of 20 mM ethylenediaminetetraacetic acid solution (pH 5.5) and frozen in liquid nitrogen. Products were separated by reverse-phase HPLC on a LC-18 DB column (Supelco, Bellefonte, PA). Products were monitored by using an on-line radioactive flow detector (A-100; Radiomatic Instruments, Median, CT). Results were expressed as picomole citrulline produced/milligram protein minute. All chemicals using iNOS assay except (3H)-L-arginine were purchased from Sigma.
Measurement of COX-2 Activity
COX-2 activity was determined by measuring the concentration of PGE2 by using enzyme immunoassay kit (Cayman Chemical, Ann Arbor, MI) (26). Cerebral cortical tissue of each sample (n = 4) was homogenized in HEPES buffer. Aliquots of homogenates (150 µg protein) were incubated with tracer (PGE2: acetylcholinesterase conjugate) and PGE2 monoclonal antibody in a 96-well microtiter plate precoated goat antimouse antibody. The plate was developed with Ellman’s reagent containing the substrate to acetylcholinesterase and the amount of the product of this reaction was detected by plate reader (Micro Reader; Bio-Tek Inst, Winooski, VT) at 410 nm. The concentration of PGE2 in each sample was then calculated according to a standard curve obtained from various known concentrations of PGE2 (1 pg/mL–10 ng/mL).
Drug Specificity and Semiquantitative Analysis of PCR Data
The selectivity of NS-398 and aminoguanidine has been tested previously by other investigators. Brian et al. (15) reported that NS-398 did not affect the cerebrovasodilatory response to either bradykinin or adenosine 5'-diphosphate. Niwa et al. (27) reported that the cerebral vasodilatory response to acetylcholine was not affected by NS-398. Iadecola et al. (28) reported that aminoguanidine did not change the cerebrovascular reactivity to hypercapnia. Further, both NS-398 and aminoguanidine inhibit their respective inducible enzymes with minimal inhibition of the constitutive isoforms (16,17). We previously addressed the efficacy of these inhibitors, showing the following: 1) aminoguanidine inhibited iNOS activity; 2) NS-398 reduced COX-2-induced PGE2 production; 3) dexamethasone inhibited activity of both enzymes; and 4) dexamethasone attenuated expression of iNOS and COX-2 mRNA in the brain of rats (13). We have also shown (13) that there was a linear relationship between the fluorescent intensity of PCR products for iNOS, COX-2, and GAPDH and the amount of RNA (0.25–2.0 µg) extracted from the cerebral cortex of ICV LPS-treated animals. This method verifies the use of RT-PCR method for semiquantitative comparisons of each mRNA levels.
Data Acquisition and Statistical Analysis
Baseline rCBF was defined as the average of laser-Doppler flow recorded over a selected 1-min period under steady-state condition at 4 h after treatment with LPS. Percent changes of rCBF from this baseline were calculated for each increment of halothane administration. All data were expressed as mean ± SD. Analysis of variance followed by the Student-Neuman-Keuls test was used for between-group comparisons. P < 0.05 was considered to be significant.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The effect of endotoxin on volatile anesthetic-induced cerebral vasodilatation has not been studied before. Nevertheless, the results obtained herein are consistent with findings in other vascular beds. Grissom et al. (29) reported that exposure of rat aortic rings to LPS inhibited phenylephrine-induced contraction and subsequent application of halothane produced more than additive decrease in tension. In the same study, a NOS inhibitor (L-NNA) partially, but not completely, attenuated the LPS enhancement of halothane’s effect. These results suggested that an induction of iNOS, and possibly other vasodilatory systems, i.e., the PG system, played a role in sensitizing vascular smooth relaxation to halothane. In this context, our results demonstrate for the first time that NO and PGs produced by their respective inducible enzyme isoforms can augment halothane-induced cerebral hyperemia.
The observed changes in iNOS and COX-2 mRNA levels and enzyme activities in the cerebral cortex were consistent with the behavior of rCBF and suggest that LPS produced significant increases in NO and PG production that contributed to an increase in baseline rCBF. Moreover, the effect of LPS on halothane-induced hyperemia was prevented by the iNOS inhibitor, aminoguanidine, and partially reversed by the COX-2 inhibitor, NS-398. Because NS-398 conferred a smaller effect than aminoguanidine, these results support the view that the primary role in halothane-induced hyperemia was played by NO after endotoxin treatment. This assertion is supported by the present results that the NO donor, DEA-NONOate, enhanced the halothane-induced hyperemia while the NO-independent vasodilator, adenosine, did not.
Conversely, aminoguanidine and NS-398 were equally effective in reversing the effect of endotoxin on baseline rCBF, suggesting a similar contribution of NO and PGs to baseline rCBF. This suggests that after LPS administration, the mechanisms of modulation of baseline rCBF and of halothane-augmented rCBF were different. The reason that aminoguanidine, NS-398, and dexamethasone did not fully reverse the endotoxin-induced increase in baseline rCBF is unclear. Because we have previously demonstrated that these inhibitors completely inhibit iNOS and COX-2 at the dose used in the present study (13), we speculate that NO derived from constitutive NOS, in particular neuronal NOS (nNOS) (30) and/or other vasodilators, were involved in affecting baseline rCBF.
A small continuous increase in rCBF was observed over four hours in control animals. We assume that, despite systematic sterilization, some contamination of ACSF by endotoxin was possible, or that an up-regulation of iNOS by the continuous administration of halothane occurred. For example, Zuo and Johns (31) showed up-regulation of endothelial NOS (eNOS) and iNOS after a four-hour exposure to halothane or isoflurane.
It is noteworthy that dexamethasone suppressed not only LPS-induced increase in COX-2 expression but also constitutive COX-2 expression. A possible interpretation of these data is that some of the COX-2 expression in control animals is modulated by inflammatory mechanisms. Another possible interpretation is that COX-2 participates in synaptic signaling in functional hyperemia, which was recently reported (27). Dexamethasone may have suppressed this property of constitutive COX-2 expression.
The finding that NS-398 only attenuated the effect of LPS on halothane-induced hyperemia whereas aminoguanidine completely reversed the response is puzzling. If PGs were in fact contributing to the halothane response, only partial inhibition after aminoguanidine would be expected. A possible explanation for this discrepancy could be that aminoguanidine had some inhibitory effect on COX-2. However, there is no indication from the extensive literature for such an effect of aminoguanidine. A more likely explanation is the possible interaction between the NO and PG systems. Pelligrino and Wang (11) described at least four levels of possible cross-talk between the NO-cGMP and prostaglandin-cyclic guanosine 3',5' monophosphate (cAMP) systems. If NO and PGs partially shared a common vasodilatory signaling pathway, such that NO was the principal mediator and PGs supported or augmented the NO effect, this type of convergence in the pathways could explain the observed effects of the inhibitors.
The cellular mechanism of iNOS/COX-2-dependent augmentation of the cerebral hyperemic response to halothane awaits clarification. We (7) and others (5,6) have proposed a role for NO from eNOS and nNOS (and to a lesser degree for prostaglandins) in halothane- or isoflurane-induced cerebral vasodilation in vivo. The traditional view of the NOS signaling pathway has been that an increase in volatile anesthetic concentration would stimulate NO production, which in turn would produce vascular relaxation. In fact, Loeb et al. (32) reported significant increases in cerebellar NO concentration measured by a microelectrode during isoflurane administration, which may support the traditional view. However, the signal transduction mechanisms leading to increased NO production are clearly different for iNOS and for constitutive NOS, and the Ca2+-calmodulin stimulation of eNOS and nNOS cannot be invoked to explain the iNOS-related enhancement of the hyperemic response. Rather, we believe that an enhancement of halothane-induced hyperemia after LPS is mainly attributed to an effect of NO on the vasodilatory signal transduction system distal to the site of NO generation.
The possibility that NO may "permit" or "modulate" cerebral vasodilation to agents such as CO2 and 
2-agonist has been proposed (8–10,33). According to the proposed scheme, it is not the source and mechanism of NO generation that would be important, but only the steady-state concentration of NO in the tissue that would influence the vasodilatory response to an agent, in our case, halothane. The validity of interpretation of experiments designed to test a "permissive" role of NO has recently been challenged by Pelligrino et al. (34). These authors suggest that exogenous NO given after constitutive NOS inhibition does not restore the normal state of signaling apparatus; instead, it activates other vasodilatory pathways, such as the miconazole-inhibitable cytochrome P450 pathway. Likewise, it is possible that, in our experiments, NO produced after iNOS induction augmented the halothane-induced vasodilation by actions distinct from the classical actions of NO from constitutive enzymatic sources. Another possible explanation is that the NO from iNOS with a consequent increase in cGMP would reduce the sensitivity of the contractile mechanism to calcium ions (35). Moreover, cGMP may inhibit phosphodiesterase III activity followed by the activation of cAMP, which may decrease the sensitivity of the contractile protein to calcium by altering myosin light chain kinase activity (36). Because halothane has been reported to relax vascular smooth muscle by reducing the intracellular calcium followed by inhibition of the contractile protein (37), the altered calcium sensitivity by either cGMP or cAMP would modulate the cerebral vasodilatory response to halothane. Still another possibility is that the increased NO and PG levels would augment the cerebral hyperemia by dilating a vascular segment that is distinct from that targeted by halothane. For example, dilation of upstream vascular segments would magnify the reduction of total cerebrovascular resistance by halothane even if the effect of halothane on small vessels was unchanged. Clearly, all these hypotheses will have to be tested in further studies to elucidate the underlying mechanism of iNOS/COX-2-dependent augmentation of the cerebral vasodilatory response to anesthetics.
In summary, our study of cerebrocortical blood flow response to halothane in rats suggests that endotoxin-induced increases in NO and PG levels associated with increased expression of iNOS and COX-2 augment the cerebrocortical hyperemic response to halothane.
|
Acknowledgments |
|---|
The authors thank Anita Tredeau for assistance in preparation of this manuscript.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. R. McCrory and S. G. E. Lindahl Cyclooxygenase Inhibition for Postoperative Analgesia Anesth. Analg., July 1, 2002; 95(1): 169 - 176. [Full Text] [PDF]
|
|
|
|
|
|
K. Nagase and K. Ando-Nagase Endotoxin Augments Cerebral Hyperemic Response to Halothane by Inducing Nitric Oxide Synthase Anesth. Analg., April 1, 2001; 92(4): 1075 - 1075. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
