| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the *Section of Neurobiology, Physiology and Behavior, Department of Anesthesiology and Pain Medicine, University of California, Davis, California; McLean Hospital, Belmont, Massachusetts; and the Department of Anesthesiology and Perioperative Care, University of California, San Francisco, San Francisco, California.
Address correspondence and reprint requests to Earl E. Carstens, Section of Neurobiology, Physiology & Behavior, University of California, Davis, Davis, California, 95616. E-mail: eecarstens{at}ucdavis.edu.
 |
Abstract |
|---|
 Top Abstract IntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
-amino butyric acid-type A (GABAA) receptor to enhance the action of GABA. Despite its profound anesthetic effects, it is unclear whether propofol has analgesic properties. Goto et al. (1) reported that propofol, unlike pentobarbital, had no effect on second-phase nocifensive behavioral responses elicited by formalin injection in the hindpaws of rats. Gilron et al. (2), however, showed that propofol suppressed hindpaw formalin-evoked expression of fos-like immunoreactivity (FLI) in spinal neurons, suggesting an important analgesic effect. Studies in humans report divergent results for the analgesic properties of propofol, with authors reporting no analgesic effect (3), analgesia (4,6), or hyperalgesia (5). In the present study we attempted to reconcile these discrepant results. We hypothesized that, in normal mice, propofol would suppress behavioral responses elicited by intraplantar injection of formalin and that this would be associated with reduced FLI in the lumbo-sacral spinal cord. We also tested mice with a genetic point mutation in the ß3 subunit of the GABAA receptor that renders the mice resistant to propofol (7). We hypothesized that formalin-evoked behavioral responses would be affected less, or not at all, by propofol in these mutant mice and that this would be associated with unchanged FLI in the spinal cord as compared with wild-type (WT) mice.
|
METHODS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The mice weighed 30.8 ± 5.2 g (mean ± sd) and were 16–42 wk of age at the time of the experiment. All animals were housed in a room with food and water ad libitum and maintained on a 12-h light-dark cycle at 21°C. The animals were maintained 1 to 5 animals per cage. Each animal was restrainer-trained and habituated to the test environment for 1 wk before testing.
For bolus injections, after placement into a restrainer (IITC Life Science, Woodland Hills, CA), IV bolus injections of propofol (30 mg/kg; Abbott Laboratories, North Chicago, IL) or intralipid (10 mg/mL in equivalent volume; Baxter, Deerfield, IL) were given in the tail via a lateral vein using a 30.5-gauge needle. The 30 mg/kg dose of propofol was chosen because it produced a near-maximal loss of righting reflex and depression of a hindlimb withdrawal reflex in WT mice while having a low incidence of lethality in mutant mice of the same strain as used presently (7). Observation of the vessel blanching with saline injection verified successful needle placement. Two mice were excluded for failure to cannulate the tail vein.
Mice to receive infusions were placed into a restrainer, and a 26 g x 19 mm over-the-needle catheter was placed percutaneously into the lateral tail vein. An initial propofol dose of 30 mg/kg was administered IV via the catheter, after which the mouse was removed from the restrainer. Anesthesia was maintained for 60 min by manual infusion of propofol to effect, i.e., absence of spontaneous movement. The average infusion rate was approximately 4 mg ·kg–1 ·min–1.
Within 1 min after the IV injection or infusion treatment, animals received a single 20 µL subcutaneous plantar injection of either formalin (5%) or saline (0.9% sodium chloride) to the hindpaw. These injections were delivered via a microliter injector (Hamilton Company, Reno, NV) connected by a prefilled 10 mm section of polyethylene 20 tubing to a 30.5-gauge needle.
WT mice were assigned to one of the following groups:
Mutant mice were assigned to one of the following groups:
Groups involving continuous infusion of propofol or vehicle in mutant, or continuous infusion of vehicle in WT mice, were not included to avoid performing infusions in awake (or potentially awake) animals.
Immediately after formalin or saline injection into the hindpaw, each mouse was placed into a clear plastic chamber (8 x 10 x 8 cm) and videotaped. Behavioral scores were compiled from a 60-min period of observation that began immediately after each treatment. Videotapes were coded and reviewed by an investigator who was unaware of the treatment and were scored in 5-min epochs. A nociceptive score, adapted from Ma et al. (6) was assigned to each 5-min interval using equation 1:
|
|
and 4 categories of pain behavior:
Equation 1 provides a weighted overall score reflecting the percentage of time that the animal displayed various nociceptive behaviors during each 5-min interval.
The same mice previously tested behaviorally with formalin were used here, except that two mutant mice in the intralipid + formalin group were excluded for the absence of FLI. Two hours posttreatment each mouse was anesthetized with sodium pentobarbital (120 mg/kg) and perfused transcardially (20 mL/min) with 100 mL of phosphate-buffered saline (0.2 M) and then 50 mL of 4% paraformaldehyde at an analogous rate. The spinal cord was extracted, postfixed for 8 h, and then transferred to a 30% sucrose solution. The lumbo-sacral enlargement (approximately L3–S2) was blocked and a slit made in the ventral white matter contralateral to the side of formalin injection. The spinal cord was cut transversely from approximately L3–S2 segments in 50-µm thick sections and processed using established methods. In brief, the sections were placed into 24-well containers with phosphate-buffered saline. Every third section (i.e., at 150-µm intervals) was washed and blocked in goat serum (3%) and then incubated in primary c-fos antibody (1:50,000 Arnel) for 2 days. After washing, the sections were placed into solution containing secondary biotinylated (goat-anti-rabbit) antibody. After this, the sections were exposed to an avidin-biotin-peroxidase complex reaction that was enhanced with biotinyl tyramide/H2O2. The reaction products were visualized using a nickel-enhanced diaminobenzidine reaction. The sections were collected on glass slides, cover-slipped and examined by light microscopy. One investigator coded the sections. Two other blinded investigators then quantified FLI. One investigator examined all sections microscopically, took photomicrographs of sections exhibiting FLI, and printed them. A second investigator then counted the numbers of nuclei exhibiting FLI in each section from the printed micrographs. The first investigator then recounted the FLI; there was in most cases good correspondence between counts; in the few instances where counts differed, the average was used. Counts of FLI were made according to their location in the superficial dorsal horn (laminae 1–2), intermediate horn (laminae 3–6) or ventral horn (laminae 7–10) on the sides ipsilateral and contralateral to the site of hindpaw injection.
All behavioral experiments and counts of spinal FLI were conducted by investigators blinded as to the genetic status. All mice were genotyped at the conclusion of the study. Just before perfusion while under deep anesthesia, the distal 1 cm of the tail was removed with sharp scissors and placed into a small vial and stored at –80C. The tissue was lysed overnight with proteinase K, and the remaining tissue was eliminated by centrifugation in a microcentrifuge, and the supernatant was kept. Nucleic acids were precipitated with isopropanol, spinned down in a microcentrifuge and washed with 70% Ethanol. DNA was dissolved in TE or water. A polymerase chain reaction with an annealing temperature of 53°C was performed using the primers RJM-8: 5'-GTT CAG CTT CCA TTC TCA CTG-3' and RJM-24: 5'-GCT ATG GCT TTC TGG TGG AG-3, which detects the presence or absence of the loxP site closely linked to the ß3(N265M) point mutation and allows determination of whether mice are homozygous for the point mutation, heterozygous for the point mutation, or WT. Based on the genotyping results, mice were assigned to WT or mutant categories in the corresponding treatment group.
The nociceptive behavioral score was summed for the Phase 1 period (0–10 min) and the Phase 2 period (20–60 min) and compared among the different groups using analysis of variance followed by Student-Newman-Keuls for multiple comparison. Counts of spinal FLI were taken from the 5 sections per animal exhibiting maximal FLI, in L4–5 and S1 segments. The FLI counts among the groups were compared according to total FLI for all sections and laminar locations (i.e., superficial dorsal horn, intermediate horn and ventral horn) using analysis of variance followed by Student-Newman-Keuls multiple comparisons. A P < 0.05 was considered significant.
|
RESULTS |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
vs 
; Fig. 1C). Phase 2 behavior was not affected by bolus propofol injection in WT mice and was very similar to that observed in WT mice receiving intralipid vehicle (Fig. 1A, C). The WT mice receiving the propofol infusion were anesthetized for the entire 60-min infusion and displayed very low levels of nociceptive behavior (Fig. 1A, 
; Fig. 1D). Control WT mice receiving intralipid plus intraplantar saline exhibited little or no nociceptive behavior (Fig. 1A, 
).
|
Propofol did not anesthetize the mutant mice, which demonstrated nociceptive behavior during Phase 1 and Phase 2 that did not significantly differ from behavior in mutant mice receiving intralipid control (Fig. 1B, vs ; and Fig. 1E). Mutant mice receiving propofol plus intraplantar saline exhibited little or no nociceptive behavior (Fig. 1B, ).
All treatment groups exhibited robust FLI in the ipsilateral lumbo-sacral spinal cord (Fig. 2A–E) except for controls receiving intraplantar saline (Fig. 2F) or no treatment. Most FLI was observed in the ipsilateral superficial dorsal horn, with some in deeper dorsal horn and intermediate zone and with a small amount observed in the contralateral cord (Fig. 2).
|
Figures 2A–C show representative sections through the lumbo-sacral enlargement from WT mice receiving a bolus injection of propofol plus intraplantar formalin. Note the robust FLI, primarily in the superficial dorsal horn. In WT mice, a propofol bolus did not significantly affect FLI (Fig. 3A; open bars) as compared with vehicle (intralipid) plus intraplantar formalin (Fig. 3A; filled bars), or continuous infusion of propofol plus intraplantar formalin in animals that did not exhibit any nociceptive behavior (Fig. 3A; hatched bars). FLI was significantly greater in superficial dorsal horn as compared with the intermediate horn and ventral horn (P < 0.05). Overall, total counts of FLI in the propofol bolus + intraplantar formalin (64 ± 19 cells/section), intralipid + intraplantar formalin groups (57 ± 19 cells/section) and the propofol infusion group (total 58 ± 11 cells/section) did not differ significantly from each other but were significantly (P < 0.05) greater compared with the group receiving intralipid + intraplantar saline (Fig. 3A, gray bars). The low FLI counts in the latter group suggest that the intraplantar injection itself did not contribute significantly to FLI counts in the other treatment groups.
|
Both the mutant group that received a bolus injection of propofol and intraplantar formalin and the group that received a bolus injection of intralipid and intraplantar formalin exhibited FLI in the superficial dorsal horn (Fig. 2D, E, respectively) that did not significantly differ between groups (Fig. 3B, open and closed bars). The mutant control group that received propofol and intraplantar saline exhibited significantly less FLI (Fig. 3B, gray bars). Two additional untreated control mutant mice exhibited very low levels of FLI (total 1 cell/section in ipsilateral cord).
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The striking difference between the lack of effect of propofol on FLI and the behavioral effects of propofol confirms results from other studies that have examined analgesic properties of anesthetics. We previously showed that halothane and isoflurane, up to concentrations (1.5%) that ablate movement in response to a supramaximal stimulus, did not significantly alter FLI (8). When isoflurane was administered at 1.8%, however, FLI decreased. Sun et al. (9) showed that both halothane and nitrous oxide failed to suppress FLI in rats, although fentanyl decreased FLI by 40%–50%. In seeming contrast, Gilron et al. (2) showed that propofol given before formalin injection suppressed FLI, but those investigators used a dilute formalin solution (1.5%); possibly, propofol can provide analgesia for this presumably weaker nociceptive stimulus. Sevoflurane also reduced formalin-evoked nociceptive behavior and spinal FLI (10). On the other hand, pentobarbital, which also acts at the GABAA receptor, failed to affect FLI (2). Xenon, which has predominant actions at the N-methyl-d-aspartate (NMDA) receptor (11), suppressed Phase 1 and Phase 2 behavior as well as FLI, the latter by 50% (6). Xenon presumably produced this effect because of its action at the NMDA receptor. However, nitrous oxide (12) and ketamine (13) also act at the NMDA receptor, and yet these anesthetics did not alter FLI in response to formalin injection (9,14). Thus, anesthetics that have similar receptor effects can dissimilarly affect FLI expression after formalin injection. Finally, there is evidence that propofol might produce analgesia at a spinal site and hyperalgesia at a supraspinal site, actions that presumably counteract each other and might result in little or no analgesic properties (15).
The inability of propofol to reduce formalin-evoked FLI in spinal neurons is at odds with neurophysiological data indicating that propofol (1 mg/kg IV) produced significant, short-lasting depression of nociceptive dorsal horn neurons in goats (16). We recently verified that systemically administered propofol dose-dependently inhibited noxious heat-evoked responses of rat dorsal horn neurons (17). The discrepancy between the ability of propofol to inhibit nociceptive responses and its inability to affect formalin-evoked FLI in spinal dorsal horn neurons remains unexplained.
It is particularly noteworthy that continuous infusion of propofol effectively eliminated any formalin-evoked nociceptive behavioral responses in WT mice, yet did not prevent FLI in dorsal horn neurons. Given the widely held assumption that FLI reflects robust neuronal excitation, this result suggests that activity in nociceptive dorsal horn neurons is not necessarily associated with nocifensive behavioral responses. Similarly, we (8,18) reported that isoflurane given in concentrations at or more than 1 MAC, the minimum alveolar concentration necessary to prevent movement in response to supramaximal noxious stimulation, did not depress nociceptive responses of dorsal horn neurons or FLI. It is speculated that propofol and isoflurane may exert their immobilizing effects at a stage of motor processing beyond the dorsal horn, such as ventral horn premotor neurons involved in generating patterned limb movements.
There was a trend toward lower behavior scores in mutant mice receiving propofol (Fig. 1E), but this failed to reach statistical significance with the fairly large numbers of animals tested (8 and 12 per group), consistent with an earlier report that propofol and etomidate were ineffective in producing anesthesia in mice having the same mutation (7). It is possible that had we studied more animals, this would have become significant, although it is unclear what mechanism would be responsible. In addition to its major effect on the GABAA receptor, propofol also affected the glycine receptor (19). Perhaps the presence of a mutant GABAA receptor subtype led to altered regulation of other receptors (such as the glycine receptor), and this might result in altered sensitivity to propofol in the mutant as compared with WT mice.
The effects of propofol on human pain are unclear. Wilder-Smith et al. (3) determined that propofol infusions did not affect thermal pain thresholds. Using an argon laser stimulus, Anker-Moller et al. (4) demonstrated that propofol injection (0.25 mg/kg) depressed acute pain, while others have reported that propofol increased responses to noxious thermal stimulation (5). A major confounding factor when determining whether a hypnotic drug affects the pain threshold is the altered consciousness that is the drug’s main pharmacological effect. Hence, if analgesia (or hyperalgesia) occurs at or near the same dose as sedation and unconsciousness, then the patient would be unable to accurately report the drug effect on pain.
In summary, we have shown that propofol did not affect FLI in lumbo-sacral spinal cord after intraplantar formalin injection, even when it was administered as an infusion that prevented nociceptive behavior. These data suggest that propofol does not have significant analgesic properties.
|
ACKNOWLEDGMENTS |
|---|
|
Footnotes |
|---|
Supported, in part, by NIGMS 47818 and DE 13685, NIH GM 57970, GM61283, P01-GM47818, and DE 13685.
|
REFERENCES |
|---|
|
TopAbstractIntroductionMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
