Secondary Logo

Journal Logo

The Effects of Propofol in the Area Postrema of Rats

Cechetto, David F. PhD; Diab, Tom; Gibson, Candace J. PhD; Gelb, Adrian W. MBChB, DA, FRCPC*

doi: 10.1097/00000539-200104000-00027
Anesthetic Pharmacology: Research Report

Propofol has an antiemetic effect that may be mediated by γ-aminobutyric acid (GABA) influences on the serotonin system, the mechanism of which is not known. We used three techniques, immunohistochemistry, High Performance Liquid Chromatography, and electrophysiology, to define propofol’s effects on the rat’s brainstem. Paired male Wistar rats received propofol, 20 mg/kg/hr, or Intralipid® for 6 h. The brains were then subjected to immunohistochemical analysis of serotonin. In a separate experiment after a propofol or Intralipid® infusion, cerebrospinal fluid (CSF) was extracted from the fourth ventricle and analyzed for the amount of serotonin and 5-hydroxyindoleacetic acid. Electrophysiological neuronal recordings were made in the area postrema (AP) in response to propofol with and without a GABA or serotonin antagonist. Results showed that immunohistochemical staining for serotonin in the propofol rats was significantly increased (28 ± 12%) in the dorsal raphe and decreased in the AP (17 ± 6%) compared with control. There were no significant changes in the isoflurane-anesthetized animals. Both serotonin and 5-hydroxyindoleacetic acid in the CSF of the fourth ventricle at the level of the AP were significantly reduced by 63% and 36%, respectively. Both propofol and pentobarbital injections reduce AP neuronal activity, but only the propofol response was blocked by bicuculline, a GABA antagonist. We conclude that the reduced levels of serotonin in the AP and the CSF may explain the antiemetic property of propofol. Propofol may also directly act on AP neurons via a GABAA receptor to reduce their activity.

Departments of Anatomy and Cell Biology and Physiology and the *Department of Anaesthesia, University of Western Ontario, London, Ontario

Supported in part by the Heart and Stroke Foundation of Ontario and the Department of Anesthesia, University of Western Ontario.

Presented at the Meeting of the Society for Neuroscience, Miami, Florida, November 14, 1994.

DFC is a Career Investigator of the Heart and Stroke Foundation of Ontario.

December 28, 2000.

Address correspondence and reprint requests to David F. Cechetto, PhD, Department of Anatomy and Cell Biology, University of Western Ontario, London, Ontario N6A 5C1. Address e-mail to

IMPLICATIONS: Propofol may produce its antiemetic effect by depleting the area postrema of serotonin as well as by a direct γ-aminobutyric acid-mediated inhibition.

When propofol is used, postoperative nausea and vomiting is less frequent than when standard barbiturate anesthesia is administered (1). Furthermore, subhypnotic doses of propofol administered after surgery resulted in an antiemetic effect, and the antiemetic properties of propofol after radiation or chemotherapy have been demonstrated (2,3). Propofol binds to a specific site on the γ-aminobutyric acid (GABA)A receptor to potentiate GABA-activated chloride flux (4). In addition, 5-HT release is reduced by GABA mediated inhibition (5,6). The amount of serotonin (5-HT) acting at the area postrema may play a role in the control of emesis (7), yet the influence of propofol on 5-HT concentration has yet to be determined. In addition, it is possible that propofol may be acting in the area postrema via an interaction with 5-HT3 receptors to produce antiemesis. A direct effect of propofol on neuronal activity in the area postrema has also not been demonstrated.

This investigation consists of three studies examining the mechanisms by which propofol might exert effects on regions of the brainstem of the rat that are related to nausea and vomiting. In the first study the effects of propofol on the immunohistochemical labeling of 5-HT in the rat brainstem were examined. Changes that were observed were determined to be specific to propofol by comparison to isoflurane anesthesia. The second study measured the amount of 5-HT and 5-hydroxyindoleacetic acid (a metabolite of serotonin) in the cerebrospinal fluid (CSF) of the fourth ventricle near the area postrema in response to propofol. In the third study, electrophysiological neuronal recordings were made in the area postrema of rats to determine the effects of propofol on neuronal activity and to determine whether GABAA and 5-HT receptors are involved.

Back to Top | Article Outline

Materials and Methods

This study was approved by the University Animal Care Committee. Paired male Wistar rats (n = 5 in each propofol group) weighing 300 to 350 g were anesthetized with pentobarbital (50 mg/kg IP). The femoral artery and vein were cannulated for blood pressure recording and anesthetic infusion, respectively. The cannulas were plugged and chronically implanted by pulling them underneath the skin and exposing them dorsally at the neck. All incisions were closed and the rats were allowed to recover for at least 7 days.

On the experimental day, the arterial cannula was connected to a pressure transducer and the venous cannula to an infusion pump containing the propofol (Zeneca, Mississauga, Canada) or vehicle (Intralipid 10%; Kabi Pharmacia, Baie d’Urfé, PQ, Canada) as control. Propofol or Intralipid was infused through the venous cannula (20 mg · kg−1 · hr−1) for 6 h. The rate of propofol infusion chosen caused no more than a 10 mm Hg reduction in blood pressure and kept the animals lightly anesthetized with little or no response to painful stimulation. Caution was taken to not disturb the control animals so that they slept throughout the study period. In the experimental rats, the rectal temperature was constantly monitored and kept at 37.5°C by a heating pad. Four additional animals were administered isoflurane 1.5% i.e., 1.1 minimum alveolar anesthetic concentration or allowed to breathe room air for 6 h. Isoflurane as a commonly used general anesthetic was used in this comparison to determine that the change in serotonin was not a nonspecific effect of general anesthetics. Nonanesthetized animals breathing air were the controls. The rectal temperature was monitored and maintained at 37.5°C in the isoflurane animals.

After 6 h all the paired experimental and control animals were given a bolus (20–30 mg) dose of propofol and then immediately (within 1 min) perfused transcardially with phosphate-buffered saline (10 mM, pH 7.4), followed by 4% in sodium acetate (0.1 M, pH 6.5) and finally by 4% paraformaldehyde in sodium borate (0.1 M, pH 9.5–11). A propofol bolus was given to both the control and isoflurane animals, because it has quick onset and less than a minute would elapse from the time of administration to when the perfusion would begin. The brains were removed and cytoprotected with 30% sucrose in the last fixative for 12 h. The brains were cut into 40-μm transverse sections on a freezing microtome and the sections were distributed sequentially into series.

Although immunohistochemistry is semiquantitative, and although more quantitative methods exist for serotonin measurement, it was used in these experiments to localize changes in specific anatomical structures in the brainstem (i.e., to enable the assessment of a number of regions simultaneously). This is not possible with other biochemical approaches. Each of the four series of sections was immunoreactively stained using antibodies to serotonin (5-HT) using the corresponding antisera (dilution 1:1000; INCSTAR, Stillwater, MN). Propofol and Intralipid, or isoflurane and air, brain sections were processed simultaneously and with the same solutions and reaction times using the peroxidase antiperoxidase technique. The brains from the isoflurane and air animals were similarly processed. Briefly, sections were mixed with 10% normal sheep serum containing 0.3% Triton-x-100 and 2% bovine serum albumin for 45 min before they were incubated at 4°C for 48 h in the serotonin antisera containing 1% normal sheep serum and 0.3% Triton-x-100. Then, the sections were rinsed in Tris-buffered saline (TBS) (0.05 M, pH 7.4), incubated with sheep antirabbit antiserum (1:50) in 1% normal sheep serum and 0.3% Triton-x-100 for 90 min, rinsed with TBS, incubated with PAP complex (1:150) in 1% normal sheep serum and 0.3% Triton-x-100 for 90 min, rinsed in TBS, reacted with 3,3′-diaminobenzidine tetrahydrochloride (0.5 mg/mL in TBS) and H2O2 (0.3 μL/mL in distilled H2O) for 13 to 30 min and rinsed. All incubations were performed on a slowly moving shaker bath. The sections were mounted on gelatinized glass slides, cover slipped and prepared for analysis. The controls for these antisera included blocking the immunoreactive staining with excess antigen and exclusion of the primary antibody from the processing sequence.

Each section of the brain was initially analyzed using a microscope under bright-field illumination. Areas with immunoreactive staining for serotonin were chosen for further analysis using a computer-assisted imaging analysis system (Mocha; Jandel Scientific, San Rafael, CA) The average intensity of staining over specific regions was quantified and these values were compared between corresponding propofol and control sections. All measures were made by an investigator who was unaware of which brain sections were being analyzed. The background illumination was adjusted to be the same for all sections by subtracting the relative intensity measure from a sample immediately outside the tissue nearest to the brain region being analyzed. The measured average intensity would provide a means of comparing the amount of immunocytochemical staining generated in the peroxidase antiperoxidase reaction.

The surgical procedure was similar to that of the first study. Once anesthetized, the rats were secured in a stereotaxic frame. The head was flexed 75–80° and the occipital bone exposed. Three holes were drilled into the occipital bone, one at the midline near the base and the other two lateral to the midline near the center. Screws were tightened into the two lateral holes. A 3.8-cm, 23-gauge ventricular guide cannula with a plug was lowered approximately 2 mm into the hole at the base. Dental acrylic secured the guide cannula to the screws. The incisions were sutured closed and 20,000 IU of Benzathine penicillin G was given as prophylactic antibiotic. The rats recovered from surgery for at least 7 days.

On the experimental day, the arterial cannula was connected to a pressure transducer and the venous cannula was connected to an infusion pump. Propofol or Intralipid was infused through the venous cannula (20 mg · kg−1 · hr−1) for 6 h. The rectal temperature of the experimental animals was constantly monitored and kept at 37.5°C by a heating pad.

After the infusions, CSF was extracted. The CSF was immediately frozen at −80°C. CSF analyses were optimized for quantification of 5-hydroxyindoleacetic acid (5-HIAA) and 5-HT on a High Performance Liquid Chromatography system with electrochemical detection. The mobile phase contained 0.5 M sodium acetate with 0.1 mM EDTA, pH 4.6 and 7% methanol. The mobile phase was run overnight at 0.2 mL/min to equilibrate over the column and flow was increased to 1.5 mL/min for amine analysis. The column was a 5 μm C-18 reverse phase column with a glassy carbon electrode set at a potential of +0.78 V versus a silver/silver chloride reference electrode. Twenty-five or 50 μL of CSF was acidified with 5 μL of 1 N perchloric acid before injection on the High Performance Liquid Chromatography system. Amounts were quantified by direct comparison against the external standards, 5-HT (100 pg with a retention time of 4 min) and 5-HIAA (1 ng with a retention time of 6 min). The internal standard DHBA ran at 4.5 min. Peak heights of the amines and their metabolites were measured in millimeters and expressed as a ratio with respect to DHBA (e.g., [5-HT peak height (mm) in sample/DHBA peak height (mm) in sample]/[5-HT peak height (mm) in external standard/DHBA peak height (mm) in external standard]) to determine the amount (ng) of amine in the sample. Sensitivity of the 5-HT and the 5-HIAAA assays were 1 pg and 5 pg respectively.

The quantity of 5-HT was measured in picograms per milliliter and 5-HIAA was measured in nanograms per milliliter. The means from the Propofol group were compared with those of the Intralipid group using an unpaired t-test with P < 0.05 considered significant.

Male Wistar rats, 300–350 g, were anesthetized with 3 percent halothane in air and tracheotomized and ventilated with 1.5 percent halothane in air. Polyethylene catheters were placed in the femoral artery and vein for the continuous recording of blood pressure and injection of drugs, respectively. Body temperature was maintained at 37°C by a rectal thermistor coupled to a heating pad. The rat was placed in a Kopf stereotactic frame with the head flexed 75–80°. The skin and neck muscles were reflected laterally. An incision was made in the atlanto-occipital membrane to expose the area postrema, which is a triangular mass of darker tissue between the bilateral gracilis nuclei. After the surgery, the inspired halothane was reduced to 1% in air (<1 minimum alveolar anesthetic concentration) and the rat was allowed 30 min to stabilize.

A glass microelectrode with a 1 mm diameter shaft was pulled to 3–5 MΩ resistance (approximately 1 μm tip), filled with 3.0 M NaCl, and lowered into the area postrema using a Narishige microdriver (Narishige International, East Meadow, NY). Extracellular single and multiple unit recordings of neuronal activity were amplified, displayed on an oscilloscope, and fed into a window discriminator connected to a computer to monitor spontaneous frequency of firing and to generate continuous-time histograms. The analysis software (IPEE; Yim Software; London, ON) records the number of neurons firing within 2-s intervals and reports them as spikes per bin.

During the experiment, the microelectrode was lowered into the area postrema in small increments (1 μm). As soon as spontaneous neuronal activity was encountered the drugs were applied as indicated below. To allow time for recovery from each of the drug applications, only one or two neuronal recordings were made in each animal. In some cases, the second recording in one animal was made by moving the microelectrode deeper (at least 200 μm) into the area postrema until another spontaneously firing neuron was encountered or removing the electrode and inserting it into a new tract in the area postrema located at least 200 μm away from the first tract. Once a site at which spontaneous neuronal activity was located within the area postrema and after a stabilization period (10 min), baseline recordings were made.

Pentobarbital was chosen for comparison to propofol in these experiments because it can be administered IV and is a GABAA agonist similar to propofol, although it binds to a distinct site. The doses of propofol and pentobarbital represent one-fifth that required to achieve full anesthesia. These doses do not alter the arterial blood pressure, which could activate baroreceptors and indirectly change neuronal activity in the area postrema. Furthermore, previous reports have indicated that propofol has an antiemetic effect at subanesthetic doses (3). Pentobarbital injections were also included to determine if the responses to propofol could be attributed to a general anesthetic effect only.

In 14 rats, 24 neurons were recorded and observed for changes in neuronal activity after IV injection of propofol (0.5 mg in 0.05 mL Intralipid®). In six animals, sodium pentobarbital (3.3 mg in 0.05 mL phosphate-buffered saline) was injected IV and the change in neuronal activity recorded for 11 neurons. Neuronal activity was observed for up to 30 min after the injection of the anesthetics.

In seven rats, eight neurons were recorded in which propofol was injected IV before and after the application of bicuculline (1 μM in 50 μL of phosphate buffered saline) onto the surface of the area postrema. In all cases, the local application of bicuculline onto the area postrema was accomplished by lowering the tip of the needle of a 100-μL Hamilton to a point immediately above the area postrema and slowly infusing 50 μL of the drug onto the surface. Control experiments indicated that the vehicle for the drug, 50 μL phosphate-buffered saline, had no effect on the activity of any of the neuronal recordings. There is a natural depression on the floor of the fourth ventricle of the rat in which the 50 μL injections pooled over the area postrema restricting flow and diffusion to other regions of the brainstem. In three rats, three neurons were recorded in which pentobarbital was injected IV before and after local injection of bicuculline onto the area postrema.

In four rats, tryptophan (400 mg/kg IP) was administered 45 min before the neuronal recordings to increase endogenous 5-HT levels. In these animals, nine neurons were initially tested to determine if they retained the inhibitory response to propofol. The response of these neurons was then recorded after an injection of ondansetron (0.01 to 0.1 mg/kg IV) and an additional IV test dose of propofol.

After the recordings, the sites of neuronal recordings in the area postrema were lesioned with an electrolytic current through the recording electrode (25 microamp, 1 s on/1 s off for 15 min). The rats were killed by transcardial perfusion with phosphate-buffered saline followed by 10% formaldehyde. The brains were removed and fixed in 10% sucrose and 10% formaldehyde for at least 12 h. The medulla was sectioned on a freezing microtome at a thickness of 50 μm. The sections were stained with thionin for histological examination and identification of the recording sites. The lesions in the area postrema were located using light microscopy and plotted on a series of representative coronal sections of the rat medulla.

The neuronal activity was sent to a window discriminator connected to a computer with software (IPEE; Yim Software) which permitted online analysis of the changes in firing frequency. Continuous histograms of the firing frequency were generated with this software. Each bin (histogram bar) represents the number of spikes recorded in a 2-s period. A change in firing frequency after the application of a drug is determined to have occurred when five consecutive bins are more than 1 sd from the baseline firing frequency. Baseline firing frequency is determined from the 100 s immediately preceding the application of the drug. The magnitude of the absolute response was determined by calculating the total number of spikes minus baseline and then multiplying this value by the response duration.

For the immunohistochemical results, a mean was derived from the optical density measurements for serotonin at specific regions of the brain in each animal. The means from a discrete region in the propofol animal were expressed as a percentage of the mean intensity of the respective region in the control animal. A two-tailed Student’s t-test was applied to the data. The null hypothesis was that the percentage data did not differ from 100% and a probability of 0.05 or below was considered significant.

For the electrophysiological results, a paired t-test was used to determine the significance of changes in responses to propofol or pentobarbital before and after pretreatment with bicuculline or ondansetron.

Back to Top | Article Outline


None of the drug administrations altered the arterial blood pressure or heart rate. Five-HT-like immunoreactivity was most intense in the dorsal raphe, median raphe nucleus, raphe obscuras nucleus, area postrema, nucleus of the solitary tract, anterior and posterior paraventricular thalamic nuclei, and locus coerelus. However, the amount of serotonin in the propofol rats was significantly altered in only two brain regions, the area postrema and the dorsal raphe (Fig. 1). The area postrema had a 17 ± 6% (P < 0.05) decrease in average staining intensity in the propofol animals compared with the Intralipid controls. There was no similar significant change in the serotonin staining in the isoflurane anesthetized compared to the unanesthetized animals. The pattern of 5-HT staining in the area postrema included an intense terminal field on the dorsal and ventral edges (Fig. 1). In these peripheral regions, fiber and terminal bouton staining was predominant. There appeared to be fewer 5-HT positive fibers and terminal boutons in the dorsal region compared with the ventral. The body of the area postrema consisted of a homogeneous 5-HT staining and contained only occasional terminal boutons (Fig. 1). Individual 5-HT-like cell bodies were not seen within the area postrema. These patterns were consistent in both propofol and control animals; however, the staining intensities differed.

Figure 1

Figure 1

The dorsal raphe had a 28 ± 12% (P < 0.01) increase in average staining intensity in the propofol animals compared with the control (Fig. 1). There was no significant effect of the isoflurane on the serotonin staining in the dorsal raphe. Intense staining was found within cell bodies and terminals throughout the dorsal raphe in both propofol and control animals. Staining intensity was greatest in the ventral portion and was a result of the aggregation of 5-HT-like immunoreactive soma, fibers, and terminals. The staining pattern within the dorsal raphe consisted of many large cell bodies and dense fibers and terminals (Fig. 1).

The amount of 5-HT in the CSF of the propofol animals (397 ± 55.6 pg/mL;n = 3) was reduced (P < 0.01) compared with the control group (1024.7 ± 80.9 pg/mL;n = 3;Fig. 2). The amount of 5-HIAA (a product of 5-HT catabolism) in the CSF of propofol animals (72.2 ± 2.4 ng/mL;n = 4) was less (P < 0.01) than the control group (111.6 ± 9.9 ng/mL;n = 4;Fig. 2).

Figure 2

Figure 2

The majority of neuronal sites (16 of 24 sites) recorded from within the area postrema demonstrated a reduction in neuronal activity in response to propofol. A typical example of this response is shown in Figure 3. Propofol resulted in an average decrease of 60% ± 6 from the initial firing frequency. Of the remaining neurons in the area postrema tested for a response to propofol, six showed no change whereas two exhibited increased activity. The duration of these responses was 1 to 5 min. The location of these neurons in the area postrema is shown in Figure 4.

Figure 3

Figure 3

Figure 4

Figure 4

Six of the 11 neurons in the area postrema tested for a response to pentobarbital exhibited a decrease in activity. An example of this is shown in Figure 3. The activity in these neurons was reduced an average of 66 ± 6% from the initial firing frequency. Three sites showed an increase in activity (27 ± 6%) and two sites responded with no change. The duration of these responses was 0.5 to 5 min. Figure 4 indicates the sites in the area postrema at which these neurons were recorded. The inhibitory response of pentobarbital was the same magnitude as that of propofol and it was not necessary to do a dose response to determine the optimum amounts of the two drugs for bicuculline and ondansetron experiments.

Local infusion of bicuculline onto the surface of the area postrema did not result in any significant change in the basal firing rate of 11 neurons recorded in the area postrema. Eight of these neurons that responded with a significant inhibition to propofol before bicuculline were also tested for a response to propofol 1 min after the infusion of bicuculline. In six of these, the bicuculline completely abolished the inhibitory response to propofol. An example of this block of the propofol response by bicuculline is shown in Figure 5A. In the remaining two neuronal recordings, the inhibitory response to propofol was greatly reduced after bicuculline application. The inhibitory response to propofol injection recovered approximately 50 min after the administration of the bicuculline. The location of these neurons in the area postrema is indicated in Figure 4.

Figure 5

Figure 5

The response to pentobarbital injection was inhibited (66 ± 6%) before bicuculline application. Of the neurons with robust responses to pentobarbital, there was no evidence of any effect of bicuculline on the response. The inhibitory response (61 ± 8%) to pentobarbital in these neurons was not significantly changed by bicuculline infusion onto the area postrema (Fig. 5B). This is in contrast to propofol, in which all of the responses were greatly affected by being either completely or mostly blocked. The lack of effect of bicuculline on the pentobarbital response was significant (P < 0.0001) compared with the inhibition of bicuculline on the propofol response.

After a minimum of 45 min after the IP injection of tryptophan (to increase endogenous 5-HT levels), injection of ondansetron (0.05 mg/kg IV) significantly inhibited the activity of neurons in the area postrema. The neuronal activity was inhibited an average of 42 ± 8%. The onset of this inhibition was slow (3 to 5 min;Fig. 6) and lasted for more than 1 h.

Figure 6

Figure 6

Nine neurons were tested with propofol before and after the administration of ondansetron. The propofol inhibitory response in six of these neurons was totally abolished and greatly reduced in three others after the ondansetron (Fig. 6). Overall, the inhibitory response to the propofol injection was attenuated by 91 ± 2%. The inhibitory response to propofol recovered approximately 3 h after the ondansetron administration (Fig. 6).

Back to Top | Article Outline


The results of our study demonstrated that the intensity of 5-HT-like immunoreactivity in the area postrema of animals receiving propofol was reduced when compared with control. The area postrema is the chemoreceptive trigger zone of the vomiting center located in the medulla and is responsible for mediating emesis caused by a variety of different stimuli (8). Lesions of the area postrema result in attenuation of radiation-induced and chemotherapy-induced taste aversion or emesis in animals (9,10). Other investigations have confirmed the presence of 5-HT-containing cells in the rat area postrema (11–13) and these 5-HT-containing neurons in the area postrema have been implicated in the control of emesis (7,14). In addition, 5-HT antagonists acting at a specific receptor (5-HT3) are effective in the prevention and reduction of radiation- or drug-induced vomiting (14,15). Antiemetic drugs, such as ondansetron, act as antagonists at 5-HT3 receptors (16) and the area postrema contains the largest concentration of these receptors in the brain (17). Furthermore, Barnes et al. (7) showed that inhibition of cisplatin-induced emesis occurred after 5-HT depletion from the area postrema of the ferret. Therefore, a reduction in the amount of 5-HT in the area postrema by propofol may be (partly) responsible for propofol’s antiemetic properties. The reduced levels of 5-HT in the area postrema might occur in two different ways.

First, it is possible that the area postrema receives serotonergic innervation that is depleted of 5-HT or from which its release is inhibited. However, Matsuura et al. (18) showed that the area postrema is not innervated by serotonergic fibers. Our immunoreactivity is supportive of this lack of 5-HT innervation of the area postrema because the 5-HT-like fibers and terminals were primarily located at the border of the area postrema without penetration into the region containing the cell bodies of this structure.

A second possibility, summarized in Figure 7, is that reduced levels of 5-HT in the CSF make less 5-HT available to be sequestered by the area postrema. Five-HT is hydrophilic and does not readily cross the blood-brain-barrier, and the area postrema does not appear to directly synthesize 5-HT (8). Five-HT is however released into the CSF by supraependymal fibers that surround the walls of the ventricles. These fibers originate from the raphe nuclei, which are the major source of 5-HT in the brain (19), but these fibers do not reach the area postrema (18,20). Evidence indicates that the area postrema cells can sequester 5-HT from the CSF (21–23) and it has been suggested that the 5-HT in the CSF released from supraependymal fibers are absorbed by “5-HT” cells in the area postrema (8).

Figure 7

Figure 7

Our results demonstrate that after a six-hour propofol infusion, the amount of 5-HT and its metabolite, 5-hydroxyindoelacetic acid, in the CSF of the fourth ventricle is reduced compared with the control animal. Our study also showed that the dorsal raphe had a significant increase in 5-HT-like immunoreactivity. These results suggest that the serotonergic supraependymal fibers may either be releasing less 5-HT to the CSF or the synthesis of 5-HT may be reduced by propofol. It has been demonstrated that 5-HT release from the dorsal raphe in rats can be reduced by the inhibitory amino acid GABA (5,6). Furthermore, these effects were totally antagonized by the GABAA receptor antagonist bicuculline (6). Because it is known that propofol acts through enhancement of a GABAA receptor mechanism (24–26), it is likely that propofol contributes to the 5-HT accretion in the dorsal raphe. The reduction in 5-HT release into the CSF from the serotonergic supraependymal fibers could then also account for the concomitant reduction in 5-HT sequestration by the area postrema.

Our electrophysiological results indicate that both propofol and pentobarbital inhibit neuronal activity in the area postrema of rats. The dose of these anesthetics was approximately one-fifth that required to adequately anesthetize the animal. Consequently, there were no changes in arterial blood pressure or heart rate in response to their administration. Thus, it is not likely that the changes in neuronal activity were indirectly caused by activation of cardiopulmonary afferents such as arterial baroreceptors or chemoreceptors. Both propofol and pentobarbital readily cross the blood-brain-barrier and the most likely explanation for the change in neuronal firing is a direct action on neurons in the area postrema or other central sites with inputs to the area postrema. However, previous investigations have demonstrated that propofol, but not pentobarbital, has antiemetic properties (2,27). Therefore, our electrophysiological results would suggest that a simple reduction in neuronal activity in the area postrema is not sufficient to produce an antiemetic action.

The electrophysiological results could be potentially complicated by the fact that the animals were maintained on halothane anesthesia. However, the halothane was maintained at the minimal level required for anesthesia in unstimulated animals (< 1 minimum alveolar anesthetic concentration). Our results may therefore reflect an interaction of anesthetics. However, there is no other way to do these experiments because neuronal recording in the area postrema in awake animals is virtually impossible.

The selectivity of action of propofol in the area postrema is indicated by the results in which only propofol but not isoflurane affected the 5-HT levels in the dorsal raphe and area postrema. In addition, only the inhibitory neuronal response by propofol, but not pentobarbital, was blocked by local application of the GABAA antagonist, bicuculline. This indicates that the effects of propofol on the area postrema neurons are not caused by a nonspecific anesthetic effect. Nevertheless, the effects of pentobarbital can be blocked by bicuculline in other regions of the brain (28), indicating that some pentobarbital effects are GABA-mediated. The lack of effect of bicuculline on pentobarbital in the area postrema contradicts a unifying theory of anesthesia through ubiquitous GABAA receptor action. Because there is regional variation in the expression of multiple subunit genes for the GABAA receptor, and thus not all GABAA receptors in the brain are the same, it is likely that anesthetics acting at GABAA receptors will also have regional effects (29).

The area postrema contains the largest concentration of 5-HT3 receptors in the brain (17). Propofol may interact with 5-HT3 receptors in the area postrema to exert an antiemetic effect. In our experiments, ondansetron, a 5-HT3 antagonist, resulted in a decrease in neuronal activity in the area postrema. This could represent an attenuation of the endogenous 5-HT mediated activity in the area postrema that is a result of activation of 5-HT3 receptors. More importantly, after the ondansetron administration, the inhibitory effect of propofol on area postrema neuronal activity was completely abolished in most recordings. It is possible that propofol, via GABAA receptors, has a direct inhibitory action within the area postrema on 5-HT3 receptors. However, there is no evidence to support this possibility.

In summary, our proposed mechanisms for the interaction of propofol and 5-HT in the area postrema is as follows (Fig. 7): Propofol inhibits the release of 5-HT from the dorsal raphe nucleus via an enhancement of the GABAA synaptic activity. This results in a reduction in the 5-HT released into the CSF from the serotonergic supraependymal fibers that, in turn, causes a reduction in the 5-HT sequestered by the area postrema. In addition, the results with ondansetron, which would have antagonized the action of 5-HT3 receptors in the area postrema, indicates the possibility of a direct GABAA mediated action of propofol on 5-HT3 receptors in the area postrema.

Back to Top | Article Outline


1. Stark RD, Binks SM, Dutka VN, et al. A review of the safety and tolerance of propofol (’Diprivan’). Postgrad Med J 1985; 61: 152–6.
2. Borgeat A, Wilder-Smith OH, Wilder-Smith CH, et al. Propofol improves patient comfort during cisplatin chemotherapy: a pilot study. Oncology 1993; 50: 456–9.
3. Borgeat A, Wilder-Smith OH, Saiah M, Rifat K. Subhypnotic doses of propofol possess direct antiemetic properties. Anesth Analg 1992; 74: 539–41.
4. Trapani G, Latrofa A, Franco M, et al. Propofol analogues: synthesis, relationships between structure affinity at GABAA receptor in rat brain, and differential electrophysiological profile at recombinant human GABAA receptors. J Med Chem 1998; 41: 1846–54.
5. Kalen P, Strecker RE, Rosengren E, Bjorklund A. Regulation of striatal serotonin release by the lateral habenula-dorsal raphe pathway in the rat as demonstrated by in vivo microdialysis: role of excitatory amino acids and GABA. Brain Res 1989; 492: 187–202.
6. Becquet D, Hery M, Francois-Bellan AM, et al. Glutamate, GABA, glycine and taurine modulate serotonin synthesis and release in rostral and caudal rhombencephalic raphe cells in primary cultures. Neurochem Int 1993; 23: 269–83.
7. Barnes JM, Barnes NM, Costall B, et al. Reserpine, para-chlorophenylalanine and fenfluramine antagonise cisplatin-induced emesis in the ferret. Neuropharmacology 1988; 27: 783–90.
8. Borison HL. Area postrema: chemoreceptor circum-ventricular organ of the medulla oblongata. Prog Neurobiol 1989; 32: 351–90.
9. Ossenkopp KP. Taste aversions conditioned with gamma radiation: attenuation by area postrema lesions in rats. Behav Brain Res 1983; 7: 297–305.
10. Borison HL, Borison R, McCarthy LE. Role of the area postrema in vomiting and related functions. Fed Proc 1984; 43: 2955–8.
11. Newton BNV, Maley B, Trauig HH. The distribution of met-enkephalin (ME), serotonin (5-HT) and substance P (SP) immunoreactivities in the area postrema (AP) of the rat and cat. Proc Soc Neurosci 1983; 9: 293.
12. Pickel VM, Armstrong DM. Ultrastructural localization of monoamines and peptides in rat area postrema. Fed Proc 1984; 43: 2949–51.
13. Lanca AJ, van der Kooy D. A serotonin-containing pathway from the area postrema to the parabrachial nucleus in the rat. Neuroscience 1985; 14: 1117–26.
14. Miner WD, Sanger GJ, Turner DH. Evidence that 5-hydroxytryptamine3 receptors mediate cytotoxic drug and radiation-evoked emesis. Br J Cancer 1987; 56: 159–62.
15. Higgins GA, Kilpatrick GJ, Bunce KT, et al. 5-HT3 receptor antagonists injected into the area postrema inhibit cisplatin-induced emesis in the ferret. Br J Pharmacol 1989; 97: 247–55.
16. Butler A, Hill JM, Ireland SJ, et al. Pharmacological properties of GR38032F, a novel antagonist at 5-HT3 receptors. Br J Pharmacol 1988; 94: 397–412.
17. Kilpatrick GJ, Jones BJ, Tyers MB. The distribution of specific binding of the 5-HT3 receptor ligand [3H]GR65630 in rat brain using quantitative autoradiography. Neurosci Lett 1988; 94: 156–60.
18. Matsuura T, Takeuchi Y, Kojima M, et al. Immunohistochemical studies of the serotonergic supraependymal plexus in the mammalian ventricular system, with special reference to the characteristic reticular ramification. Acta Anat Basel 1985; 123: 201–19.
19. Steinbusch HW. Distribution of serotonin-immunoreactivity in the central nervous system of the rat-cell bodies and terminals. Neuroscience 1981; 6: 557–618.
20. Takeuchi Y, Sano Y. Serotonin distribution in the circumventricular organs of the rat: an immunohistochemical study. Anat Embryol (Berl) 1983; 167: 311–9.
21. Torack RM, Finke EH. Evidence for a sequestration of function within the area postrema based on scanning electron microscopy and the penetration of horseradish peroxidase. Z Zellforsh Mikrosk Anat 1971; 118: 85–96.
22. Dow RC, Laszlo I, Ritchie IM. Cellular localization of the uptake of 5- hydroxytryptamine in the area postrema of the rabbit after injection into lateral ventricle. Br J Pharmacol 1973; 49: 580–7.
23. Yamamoto M, Chan-Palay V, Palay SL. Autoradiographic experiments to examine uptake, anterograde and retrograde transport of tritiated serotonin in the mammalian brain. Anat Embryol Berl 1980; 159: 137–49.
24. Concas A, Santoro G, Mascia MP, et al. The general anesthetic propofol enhances the function of gamma-aminobutyric acid-coupled chloride channel in the rat cerebral cortex. J Neurochem 1990; 55: 2135–8.
25. Peduto VA, Concas A, Santoro G, et al. Biochemical and electrophysiologic evidence that propofol enhances GABAergic transmission in the rat brain. Anesthesiology 1991; 75: 1000–9.
26. Hara M, Kai Y, Ikemoto Y. Propofol activates GABAA receptor-chloride ionophore complex in dissociated hippocampal pyramidal neurons of the rat. Anesthesiology 1993; 79: 781–8.
27. Schaer H, Prochacka K. Recovery, amnesia and affective state following propofol in comparison with thiopental. Anaesthetist 1990; 39: 306–12.
28. Yamada KA, Moerschbaecher JM, Hamosh P, Gillis RA. Pentobarbital causes cardiorespiratory depression by interacting with a GABAergic system at the ventral surface of the medulla. J Pharmacol Exp Ther 1983; 226: 349–55.
29. Olsen RW, Sapp DM, Bureau MH, et al. Allosteric actions of central nervous system depressants including anesthetic on subtypes of the inhibitory -aminobutyric acidA receptor-chloride channel complex. Ann NY Acad Sci 1991; 625: 145–154.
© 2001 International Anesthesia Research Society