General anesthesia traditionally has been characterized by the loss or decrease of certain functions present in the awake state, namely, consciousness, and reflex activities, and the presence of amnesia and analgesia. In subanesthetic concentrations, nitrous-oxide-induced analgesia can be partially reversed by the narcotic antagonist, naloxone, in rats and mice [1-4] and humans [5,6]. The stereospecificity of this effect of naloxone has been demonstrated recently . Rats and mice developed tolerance to the analgesic action of nitrous oxide after prolonged exposure [2,8]. In addition, one-way cross-tolerance was observed, i.e., mice and rats made tolerant to morphine were also tolerant to the analgesic action of nitrous oxide, but animals tolerant to nitrous oxide were not tolerant to morphine . Also, nitrous-oxide-induced analgesia was attenuated in mice whose opioid receptors were blocked by naloxonazine, an "irreversible" opioid receptor antagonist . In rats, after prolonged exposure to nitrous oxide, opioid receptor density in the brainstem decreased as measured with radiolabeled dihydromorphine, indicating a down regulation of opioid receptors . These observations suggest that the analgesic action of nitrous oxide could be explained on the basis of an interaction with the opioid receptor-endogenous opioid peptide system.
Reports concerning the attenuating action of narcotic antagonists on anesthesia with more potent anesthetics, such as halothane, have been controversial. We have reported that intravenous (IV) injection of naloxone resulted in a decrease in the depth of anesthesia, indicated by electroencephalogram arousal and an increased incidence of response to tail-clamping in rats during anesthesia with constant inspired concentrations of cyclopropane, halothane, and enflurane . Arndt and Freye  observed in dogs anesthetized with halothane, that perfusion of the fourth cerebral ventricle with mock cerebrospinal fluid (CSF) containing naloxone caused a lightening of anesthesia, i.e., they responded to painful stimulation more force-fully, rolled and opened their eyes, and, finally, responded to verbal commands. The circulatory depressant action of halothane was also reversed. However, several reports refuted the idea that general anesthetics act on the opioid receptor-endogenous opioid peptide system. In three studies [13-15] naloxone did not antagonize the anesthetic action of halothane in rats, although two of these studies used loss of righting reflex as the indicator of halothane anesthesia [13,14], rather than an analgesic end-point as in other studies. Similar results were reported by Pace and Wong  using halothane in dogs. Harper et al.  and Pace and Wong  did use the tail clamp technique, an analgesic end-point, to determine minimum alveolar anesthetic concentration (MAC) values. Although tail clamping involves pain, motor responses may be attenuated more by the "anesthetic" concentrations of halothane. More recently, Levin et al.  observed in rats that naloxone does not change the tail flick ED50 nor MAC values of halothane, enflurane, and isoflurane. Based on these results, these authors concluded that the opiate receptor-endogenous opioid peptide system is not involved in the analgesic action of general anesthetics. The controversy regarding the possible involvement of the opioid receptor-endogenous opioid peptide system in the mechanism of anesthetic-induced analgesia is based primarily on the ability, or lack thereof, of naloxone or other narcotic antagonists to antagonize the effects of general anesthetics. All of these reports are based on indirect evidence both supporting and refuting this hypothesis. It would appear that only more direct evidence will further resolve the controversy. Therefore, we measured concentrations of six naturally occurring endogenous opioid peptides in CSF obtained from the chronically cannulated third cerebral ventricle of dogs breathing air or nitrous oxide.
This study was approved by our institutional animal care and use committee. Eight conditioned mongrel dogs of either sex weighing 9-12 kg were used. For several days prior to implantation of a catheter into the third cerebral ventricle, they were trained to lie quietly on the laboratory floor and to accept a dog face mask, breathing air. Under pentobarbital anesthesia (36 mg/kg, IV), the trachea was intubated and the head held rigidly with a stereotaxic apparatus (David Kopf Instruments, Tujunga, CA). Under sterile conditions, the scalp was incised and a small burr hole made in the midline 16-18 mm behind the interaural line. A custom made steel cannula (19-gauge, 32 mm in length), mounted on the stereotaxic apparatus, was advanced slowly to a depth at which CSF appeared at the hub (range, 24-29 mm from the surface of the skull). The cannula was then fixed to the skull using dental cement and anchoring screws. A custom made, heparin-bonded , silicone rubber catheter (Dow-Corning, Midland, MI; 0.64 mm outside diameter, 0.3 mm inside diameter) was inserted through the cannula, such that the tip of the catheter extended 3-5 mm beyond the tip of the cannula. The catheter was fixed to the cannula with silicone cement. The external end of the catheter was plugged until use. Animals received 1 L of lactated Ringer's solution containing 1 g of chloramphenicol, infused IV, during and after the procedure. They also received 1.2 million units of Bicillin Registered Trademark (Wyeth-Ayerst, Philadelphia, PA; penicillin G benzathine, penicillin G procaine), intramuscularly.
At least 2 days after anesthesia and surgery, each dog, restrained only by a short leash, was randomly given either room air (control) or 66-75 vol% nitrous oxide, balance oxygen (experimental), to breathe through a face mask using a nonrebreathing system. Gas mixtures were delivered by calibrated flowmeters at total flows sufficient to keep the reservoir bag inflated. The external end of the catheter was connected via sterile polyethylene tubing to a 10-mL plastic syringe mounted on a Harvard withdrawal pump. CSF was withdrawn at a rate of 1.5-1.7 mL/h. When a sufficient volume of CSF had been obtained, the inspired mixture was changed and an additional sample of CSF obtained or else the external end of the catheter was replugged for use at a later time. Cannula placement was verified in each animal at autopsy. Catheter deadspace was less than 0.25 mL.
After volume measurement, CSF samples were transferred to polypropylene tubes, placed in a boiling water bath for 10 min, and centrifuged at 1000g for 10 min. Supernatants were frozen at -70 degrees C until assayed for concentrations of endogenous opioid peptides.
Specimens were coded such that the circumstances under which each had been obtained (control or experimental) were not known until assays were completed.
For peptide assays, CSF samples were thawed and passed through Sep-Pak Registered Trademark C18 cartridges (Waters, Milford, MA), which had been pretreated with acetonitrile and water, at a rate of 2 mL/min. The cartridges were washed with 5 mL of 0.05 M phosphate buffer, pH 2.7, and the peptides eluted with 2 mL of 40% acetonitrile in 0.05 M phosphate buffer for 3 min. Sep-Pak Registered Trademark eluates were dried under reduced pressure (Speed-Vac Registered Trademark, Savant Instruments, Farmingdale, NY), reconstituted in 200 micro Liter water, and injected onto a reversed phase high-performance liquid chromatograph (Hewlett-Packard, San Diego, CA; model 1084B, equipped with fraction collector) for separation of individual peptides, by means of a previously developed set of chromatographic conditions for the separation of met5-enkephalin (ME), dynorphin A (DYN-A), dynorphin B (DYN-B), met5-enkephalin-arg (6-phe)7 (MERF), and beta-endorphin (BE).
Using mixtures of synthetic opioid peptides (Peninsula Laboratories, Belmont, CA) at concentrations which could be detected by ultraviolet absorbance at 210 nm, conditions were adjusted to produce a physical separation of the endogenous opioids. The solid phase consisted of a Microsorb Trademark, 3 mu, C18, 10-cm analytical column and guard column (Rainin Instruments, Woburn, MA). The mobile phase was a mixture of acetonitrile and 0.05 M phosphate buffer, pH 2.7, containing 0.1% triethylamine  at a flow rate of 0.5 mL/min in a linear gradient from 15% to 50% acetonitrile over 42 min at a column temperature of 35 degrees C.
Reconstituted CSF samples and mixtures of synthetic peptide standards dissolved in 200 micro Liter of water, in concentrations approximating those found in the CSF samples, were injected into the chromatograph. Fractions of column effluent bracketing the known retention time of each peptide were collected, dried under reduced pressure, and reconstituted in water or buffer in volumes appropriate for radioimmunoassay (RIA) of each of the five peptides. DYN A and B could not be resolved under these conditions. Therefore, the fraction of column effluent containing these two peptides was again dried under reduced pressure, reconstituted in 200 micro Liter of water, and rechromatographed using a gradient from 18% to 23% acetonitrile in 0.05 M phosphate buffer over 30 min, which completely resolved the two dynorphins. RIA of the separated peptides was performed using commercially available kits, following instructions included therein. For ME, MERF, and BE, RIA kits were purchased from Incstar Corporation (Stillwater, MN) and those for DYN A and B were purchased from Peninisula Laboratories. Authentic peptide standards were purchased from Peninsula Laboratories. Synthetic MERF was used as standard in the ME RIA for determination of MERF in CSF samples. For each peptide, samples derived from the same animal were determined in the same RIA, except when the necessity for further sample dilution prevented this. For ME, the assay was sensitive to 50 fmol/mL; for DYN A, DYN B, MERF, and BE the RIA sensitivities were: 28 fmol/mL, 45 fmol/mL, 0.1 pmol/mL, and 3 fmol/mL, respectively. The intraassay coefficients of variation and recoveries for the entire analysis, including the Sep-Pak extraction and high-performance liquid chromatography separation steps as well as the RIAs were, respectively, 22% and 68% for ME; 8% and 66% for DYN A; 12% and 81% for DYN B; 19% and 83% for MERF; and 9% and 86% for BE. Reported values were corrected for recovery accordingly. Endogenous opioid peptide concentrations in control and experimental CSF samples were compared using paired Student's t-test.
(Table 1) shows the actual volumes of the obtained CSF samples and CSF concentrations of ME. Values ranged from 0.09 or less to 0.65 (mean +/- SEM, 0.30 +/- 0.07) pmol/mL in dogs breathing room air. These increased markedly in samples obtained during nitrous oxide inhalation, ranging from 14.0 to 76.3 (42.4 +/- 8.1) pmol/mL. The difference was highly significant as the changes represent increases in every animal ranging from 28 to more than 400 times the control value.
CSF concentrations of MERF are shown in Table 2. They ranged from 3.8 to 22.6 (14.5 +/- 2.5) pmol/mL in animals breathing room air, whereas during nitrous oxide inhalation, the values ranged from 2.1 to 139 (57.6 +/- 17.8) pmol/mL. This difference was also statistically significant. Increases were found in seven of eight dogs and ranged from 1.5 to 8.3 times above control values. A decrease was measured in one dog.
DYN A concentrations in CSF from animals breathing room air were not different from those in samples obtained with the animals breathing nitrous oxide, with a mean value of 1.36 in the former, and 1.07 pmol/mL in the latter group Table 3. Similarly, DYN B concentrations in CSF showed inconsistent and insignificant changes between the two conditions, averaging 194 in controls and 47.5 fmol/mL during nitrous oxide inhalation Table 4.
CSF concentrations of BE, likewise, were not significantly changed by the inhalation of nitrous oxide, averaging 3.82 in air breathing controls and 15.3 pmol/mL in the same animals breathing nitrous oxide Table 5.
Behaviorally, animals breathing room air through a dog face mask appeared normal, standing or lying quietly during CSF withdrawal. During nitrous oxide breathing the animals sometimes became excited. Three of the eight dogs became tremulous and agitated and needed to be manually restrained in order to continue the inhalation of nitrous oxide; one was noted to be mildly excited; the remaining four dogs were quiet and calm. Hyperventilation, interspersed with brief episodes of panting, occurred in all animals when breathing nitrous oxide.
In each animal the steel cannula was located in the anterior portion of the third cerebral ventricle, slightly below the roof and just forward of the massa intermedia of the thalamus.
A limited number of previous studies have examined CSF concentrations of endogenous opioid peptides, although none have used the dog as a model nor simultaneously measured opioid peptides deriving from all three known precursor molecules: proenkephalin, prodynorphin, and proopiomelanocortin (POMC), which generate all the known naturally occurring endogenous opioid peptides via posttranslational processing. Our methodology is also more specific for each of the measured peptides compared with most previous studies because of the physical separation of each peptide using high-performance liquid chromatography prior to performance of the more sensitive RIA. Using this approach one need be less concerned with the issue of antibody cross-reactivities, a potential problem when RIA is used alone, resulting in the necessity of characterizing such measurements as immunoreactivities. Because we were attempting to measure the effects of an anesthetic, use of an awake conscious control state seemed most desirable. In this regard we attempted to minimize stress to the animals through training and familiarity with surroundings prior to CSF sampling.
Akil et al.  measured concentrations of enkephalin-like activity, using radioreceptor and bioassays, in third ventricular CSF from eight human patients with intractable pain and found a mean of 0.69 pmol/mL, similar in magnitude to our control canine value of 0.30 pmol/mL, despite differing methodologies and species. In that study, focal electrical stimulation of periventricular brain sites resulted in both significant analgesia and increases in enkephalin-like material averaging 46%-83% above control values; this was among the first demonstrations of in vivo release of an endogenous opioid peptide. Other studies demonstrated release of endogenous opioid peptide-like compounds into CSF in response to manual or electroacupuncture [21-23]. In contrast to these studies showing an increase in CSF concentrations of endogenous opioid peptides associated with a state of analgesia, other studies showed that noxious stimulation also caused an increase in CSF concentrations of endogenous opioid peptides. Concentrations of ME-like material were found to increase in aqueductal and spinal perfusates of rats after intraarterial injection of bradykinin and high-intensity sciatic nerve stimulation . Likewise, tooth pulp stimulation was found to increase concentrations of ME-like material in cisternal CSF of cats anesthetized with halothane . Thus, it would appear that endogenous opioid peptide concentrations in CSF can be increased both by stimulation which produces analgesia, as well as by noxious stimulation, presumably an attenuating response to the latter situation.
Several earlier studies seemed to indicate that analgesic electrical stimulation of periventricular brain sites in humans caused an increase in BE-like immunoreactivity in ventricular CSF [26,27]. However, subsequent studies found that such changes were artifactual, resulting from interference by iodinated contrast media, which causes a release of BE, rather than a release caused by electrical stimulation [28-30].
Possible effects of nitrous oxide on the endogenous opioid peptide system have been examined in several previous studies. Way et al.  measured concentrations of BE-like and leu5-enkephalin (LE) immunore-activities in ventricular CSF of patients before and during anesthesia which consisted of thiopental, 70% nitrous oxide in oxygen, and halothane. They found that BE-like immunoreactivity was unchanged and that LE was undetectable. The cannulae from which CSF was obtained in these patients were placed in a lateral cerebral ventricle, the structures surrounding which are not known to contain endogenous opioid peptides, in marked contradistinction to the third cerebral ventricle, tissues surrounding which are among the richest in the neuraxis in their content of endogenous opioid peptides. In all likelihood, one would need to postulate a reversal of normal CSF flow through the foramen of Monro from the third ventricle into the lateral ventricle in order to see any changes in endogenous opioid peptides when sampling from the lateral ventricle.
Morris and Livingston  exposed rats to 80% nitrous oxide-20% oxygen for either 1 or 18 h and then measured ME immunoreactivities in five discrete brain areas. They found no significant differences in any of the areas examined with either 1- or 18-h exposures to nitrous oxide compared to parallel control animals. Although the areas examined (dorsal raphe, medial thalamus, periaqueductal grey, raphe magnus, and locus coeruleus) are all considered important in the mediation of nociception, it is not clear that nitrous oxide would be expected to have an effect on any or all of these brain areas. Similarly, it is also not clear what effect nitrous oxide would be expected to have on tissue opioid peptide concentrations. An increase, a decrease, or no change might be explained depending on the turnover rates of both the peptide end-products as well as precursor molecules. Currently, methods for studying turnover of endogenous opioid peptides are lacking.
Quock et al.  measured immunoreactivities of both ME and LE in ventriculocisternal perfusates of rats anesthetized with urethane. They found an approximately twofold increase in ME immunoreactivity in fractions of perfusate obtained when animals were exposed to 75% nitrous oxide-25% oxygen, compared to animals breathing room air. BE-like immunoreactivity was not changed by exposure to nitrous oxide in this study. As noted by these authors, the effects of urethane anesthesia upon central nervous system release of ME, if any, are unknown. Likewise, there are marked physiologic differences between a ventriculocisternal perfusion model in anesthetized rats and chronically cannulated and acclimated conscious dogs. These differences might account for the much larger increases in ME seen in the present study. Qualitatively, our results are in agreement with theirs with respect to the peptides measured in common. BE concentrations were unchanged in both studies.
In a further study, Quock et al.  measured concentrations of ME immunoreactivity in brain regions of rats exposed to 75% nitrous oxide-25% oxygen and found significant 12%-18% increases in the brainstem, spinal cord, hypothalamus, and corpus striatum, but not in the cerebral cortex or diencephalon, compared to air exposed control animals. Interpretation of these results concerning turnover and release of ME is difficult, the more so because the RIA method used would not distinguish between end-product pentapeptides and longer chain intermediate or precursor molecules containing the ME sequence. The increase in tissue concentrations found in this study is not reconciled easily with their previous study  showing a release of ME immunoreactivity into ventriculocisternal perfusate. One might expect release into CSF to be reflected by lower tissue concentrations. As yet unknown changes in the posttranslational processing of precursor molecules could account for this apparent disparity. Turnover studies, if methods were available, might aid greatly in clarifying the situation.
Zuniga et al.  measured tissue concentrations of BE immunoreactivity in rats exposed to nitrous oxide and found increases in the medial basal hypothalamic region and periaqueductal gray, but not in the remainder of the diencephalon. Lack of concomitant changes in alpha-melanocyte stimulating hormone (alpha-MSH) suggested differences in posttranslational processing of the POMC precursor molecule, from which both BE and alpha-MSH are derived. Essentially the same results were obtained when dispersed cells from the rat basal hypothalamus, attached to cytodex beads, were found to secrete BE, but not alpha-MSH, into the superfusion medium when the medium was changed from one equilibrated with room air to one equilibrated with 80% nitrous oxide-20% oxygen . Whether secretion of BE by these cells would be expected to increase concentrations of BE in third ventricular CSF is not known. Our finding that BE concentrations were unchanged in third ventricular CSF of dogs, along with the previously mentioned lack of change of BE in ventriculocisternal perfusate of rats , suggests that the BE-containing cells in the arcuate region of the hypothalamus may have processes of passage to terminal fields more caudal in the neuraxis, rather than more localized BE-secretory activity. Immunocytochemical studies demonstrating BE-reactive fibers within the immediate subependymal zone but never penetrating the ependymal cell layer  would support such an interpretation.
Our results suggest the preferential involvement of the proenkephalin-derived series of endogenous opioid peptides in the mediation of nitrous oxide-induced analgesia. Of the five peptides measured, only the two derived from proenkephalin, ME and MERF, were found to increase. Posttranslational processing of one proenkephalin molecule may yield between five and seven molecules of ME but only one each of MERF and LE . Although the details and stoichiometry of proenkephalin processing are not known, the increases in ME far in excess of the other two end-products (MERF and LE) are consistent with the known primary structure of their common precursor. DYN A and DYN B, both derived from prodynorphin, and BE, derived from POMC, were unchanged.
Inherent in this discussion is the supposition on our part that nitrous oxide produces analgesia in the dog, as we did not attempt to measure analgesia in this study. However, such an assumption seems valid because nitrous oxide reduces the MAC of halothane in dogs, to largely the same extent as in other animals , and, as noted earlier [1-4,7] produces analgesia in rodents. The small size of rodents precluded using them for this study despite a better documented analgesic effect of nitrous oxide.
Increases in CSF concentrations of proenkephalin-derived peptides infer a release from brain tissue. Whether this represents an "overflow" of locally released peptide acting locally, or whether of physiologic importance in terms of peptide transport to more distant sites of action, is unknown. ME and LE were originally shown to produce naloxone reversible analgesia lasting 10-12 min after injection of 100-200 micro gram amounts into the lateral ventricles of chronically cannulated rats after a 1- to 6-min latency period . A later study in mice confirmed the naloxone reversible analgesic effect of ME after intracerebroventricular injection and, in addition, found that MERF also had these effects but was eight times more potent than ME . If there are large differences in analgesic potency between ME and MERF, then assessing the relative contribution of these peptides in mediating nitrous oxide-induced analgesia becomes very difficult, as large quantitative increases in CSF concentrations, i.e., as seen with ME, may be offset to unknown degrees by lesser increases, i.e., as seen with MERF, in the more potent peptide.
In summary, most previous studies concerning the possible interactions of nitrous oxide with the opioid receptor-endogenous opioid peptide system in mediating analgesia have been both indirect and controversial. In an effort to shed more direct light on this controversy we measured concentrations of five naturally occurring endogenous opioid peptides, derived from all three known precursor molecules, in third ventricular CSF of chronically cannulated and trained dogs, obtaining paired samples of CSF taken with the animals breathing either room air or nitrous oxide. Inhalation of nitrous oxide was found to cause a striking increase in ME, and a significant increase in MERF. Both of these peptides are derived from proenkephalin, suggesting a selective involvement of this family of neuropeptides in the mediation of nitrous-oxide-induced analgesia. Concentrations of the other three peptides measured failed to change, suggesting a lack of involvement of either the prodynorphin or proopiomelanocortin families of peptides.
Whether these results are unique to nitrous oxide, whether they apply only to those anesthetics known to have good analgesic properties when used in subanesthetic concentrations, or whether they apply to all inhalational anesthetics are unknown. Results of such studies may improve our understanding of mechanisms of analgesia and anesthesia, as well as delineate selective effects of different anesthetics on different neurotransmitter-receptor systems within the central nervous system.
1. Berkowitz BA, Ngai SH, Finck AD. Nitrous oxide analgesia: resemblance to opiate action. Science 1976;194:967-8.
2. Berkowitz BA, Finck AD, Ngai SH. Nitrous oxide analgesia: reversal by naloxone and development of tolerance. J Pharmacol Exp Ther 1977;203:539-47.
3. Lawrence D, Livingston A. Opiate-like analgesic activity in general anesthetics. Br J Pharmacol 1981;73:435-42.
4. Smith EH, Rees JMH. The effects of naloxone on the analgesic activities of general anesthetics. Experientia 1981;37:289-90.
5. Yang JC, Clark WC, Ngai SH. Antagonism of nitrous oxide by naloxone in man. Anesthesiology 1980;52:414-7.
6. Chapman CR, Benedetti C. Nitrous oxide effects on cerebral evoked potential to pain: partial reversal with a narcotic antagonist. Anesthesiology 1979;51:135-8.
7. Quock RM, Graczak LM. Influence of narcotic antagonist drugs upon nitrous oxide analgesia in mice. Brain Res 1988;440:35-41.
8. Berkowitz BA, Finck AD, Hynes MD, Ngai SH. Tolerance to nitrous oxide analgesia in rats and mice. Anesthesiology 1979;51:309-12.
9. Finck AD, Samaniego E, Ngai SH. Irreversible opioid receptor blockade decreases the analgesic effects of ketamine and nitrous oxide in mice. Anesthesiology 1988;69:A604.
10. Ngai SH, Finck AD. Prolonged exposure to nitrous oxide decreases opiate receptor density in rat brainstem. Anesthesiology 1982;57:26-30.
11. Finck AD, Ngai SH, Berkowitz BA. Antagonism of general anesthesia by naloxone in the rat. Anesthesiology 1977;46:241-5.
12. Arndt JO, Freye E. Perfusion of naloxone through the fourth cerebral ventricle reverses the circulatory and hypnotic effects of halothane in dogs. Anesthesiology 1979;51:58-63.
13. Bennett RB. Naloxone fails to antagonize the righting response in rats anesthetized with halothane. Anesthesiology 1978;49:9-11.
14. Smith RA, Wilson M, Miller KW. Naloxone has no effect on nitrous oxide anesthesia. Anesthesiology 1978;49:6-8.
15. Harper MH, Winter PM, Johnson BH, Eger EI II. Naloxone does not antagonize general anesthesia in the rat. Anesthesiology 1978;49:3-5.
16. Pace NL, Wong KC. Failure of naloxone and naltrexone to antagonize halothane anesthesia in the dog. Anesth Analg 1979;58:36-9.
17. Levin LL, Winter PM, Nemoto EM, et al. Naloxone does not antagonize the analgesic effects of inhalation anesthetics. Anesth Analg 1986;65:330-2.
18. Eberle JW, Manton JR, Meals CR, et al. Cross-linked heparin binding of a membrane oxygenator system. J Biomed Mater Res 1973;7:145-53.
19. Rivier JE. Use of trialkyl ammonium phosphate (TAAP) buffers in reverse phase HPLC for high resolution and high recovery of peptides and proteins. J Liquid Chromatogr 1978;1:343-66.
20. Akil H, Richardson DE, Hughes J, Barchas JD. Enkephalin-like material elevated in ventricular cerebrospinal fluid of pain patients after analgetic focal stimulation. Science 1978;201:463-5.
21. Clement-Jones V, Lowry PJ, McLoughlin L, et al. Acupuncture in heroin addicts: changes in met-enkephalin and beta-endorphin in blood and cerebrospinal fluid. Lancet ii:1979;380-3.
22. Sjolund B, Terenius L, Eriksson M. Increased cerebrospinal fluid levels of endorphins after electroacupuncture. Acta Physiol Scand 1977;100:382-4.
23. Zhang A, Pan X, Xu S, et al. Endorphins and acupuncture analgesia. Chin Med J (Engl) 1980;93:673-80.
24. Yaksh TL, Elde RP. Release of methionine-enkephalin immunoreactivity from the rat spinal cord in vivo. Eur J Pharmacol 1980;63:359-62.
25. Cesselin F, Oliveras JL, Bourgoin S, et al. Increased levels of met-enkephalin-like material in the CSF of anaesthetized cats after tooth pulp stimulation. Brain Res 1982;237:325-38.
26. Akil H, Richardson DE, Barchas JD, Li CH. Appearance of beta-endorphin-like immunoreactivity in human ventricular cerebrospinal fluid upon analgesic electrical stimulation. Proc Natl Acad Sci USA 1978;75:5170-2.
27. Hosobuchi Y, Rossier J, Bloom FE, Guillemin R. Stimulation of human periaqueductal gray for pain relief increases immunoreactive beta-endorphin in ventricular fluid. Science 1979;203:279-81.
28. Fessler RG, Brown FD, Rachlin JR, et al. Elevated beta-endorphin in cerebrospinal fluid after electrical brain stimulation: artifact of contrast infusion? Science 1984;224:1017-9.
29. Dionne RA, Mueller GP, Young RF, et al. Contrast medium causes the apparent increase in beta-endorphin levels in human cerebrospinal fluid following brain stimulation. Pain 1984;20:313-21.
30. Harnish PP, Zuniga J, Northington FK, et al. Effect of contrast media on beta-endorphin secretion: an in vitro study. Acta Radiol 1988;29:741-3.
31. Way WL, Hosobuchi Y, Johnson BH, et al. Anesthesia does not increase opioid peptides in cerebrospinal fluid of humans. Anesthesiology 1984;60:43-5.
32. Morris B, Livingston A. Effects of nitrous oxide exposure on met-enkephalin levels in discrete areas of rat brain. Neurosci Lett 1984;45:11-4.
33. Quock RM, Kouchich FJ, Tseng LF. Does nitrous oxide induce release of brain opioid peptides? Pharmacology 1985;30:95-9.
34. Quock RM, Kouchich FJ, Tseng LF. Influence of nitrous oxide upon regional brain levels of methionine-enkephalin-like immunoreactivity in rats. Brain Res Bull 1986;16:321-3.
35. Zuniga JR, Joseph SA, Knigge KM. The effects of nitrous oxide on the central endogenous pro-opiomelanocortin system in the rat. Brain Res 1987;420:57-65.
36. Zuniga JR, Joseph SA, Knigge KM. The effects of nitrous oxide on the secretory activity of pro-opiomelanocortin peptides from basal hypothalamic cells attached to cytodex beads in a superfusion in vitro system. Brain Res 1987;420:66-72.
37. Bloom F, Battenberg E, Rossier J, et al. Neurons containing beta-endorphin in rat brain exist separately from those containing enkephalin: Immunocytochemical studies. Proc Natl Acad Sci USA 1978;75:1591-5.
38. Lewis RV, Stern AS. Biosynthesis of the enkephalins and enkephalin-containing polypeptides. Annu Rev Pharmacol Toxicol 1983;23:353-72.
39. Steffey EP, Eger EI II. Nitrous oxide in veterinary practice and animal research. In: Eger EI II, ed. Nitrous oxide/N2
O. New York: Elsevier, 1985:305-12.
40. Belluzzi JD, Grant N, Garsky V, et al. Analgesia induced in vivo by central administration of enkephalin in rat. Nature 1976;260:625-6.
© 1995 International Anesthesia Research Society
41. Inturrisi CE, Umans JG, Wolff D, et al. Analgesic activity of the naturally occurring heptapeptide [Met]enkephalin-Arg6-Phe7
. Proc Natl Acad Sci USA 1980;77:5512-4.