Antinociceptive Properties of Neurosteroids: A Comparison of Alphadolone and Alphaxalone in Potentiation of Opioid Antinociception

Winter, L.; Nadeson, R.; Tucker, A. P.; Goodchild, C. S.

Anesthesia & Analgesia:
doi: 10.1213/01.ANE.0000075835.73967.F3

In this study, we investigated the antinociceptive and sedative effects of the opioids fentanyl, morphine, and oxycodone given alone and in combination with two neurosteroids: alphadolone and alphaxalone. An open-field activity monitor and rotarod apparatus were used to define the sedative effects caused by opioid and neurosteroid compounds given alone intraperitoneally to male Wistar rats. Dose-response curves for antinociception were constructed using only nonsedative doses of these drugs. At nonsedating doses, fentanyl, morphine, and oxycodone all caused dose-dependent tail flick latency (TFL) antinociceptive effects. Because neither neurosteroid altered TFL, electrical current was used as the test to determine doses of neurosteroid that caused antinociceptive effects at nonsedative doses. Alphadolone 10 mg/kg intraperitoneally caused significant antinociceptive effects in the electrical test but alphaxalone did not. All three opioid dose-response curves for TFL antinociception were shifted to the left by coadministration of alphadolone even though alphadolone alone had no effect on TFL. Alphaxalone given alone had no antinociceptive effects at nonsedative doses and it had no effect on opioid antinociception. Neither neurosteroid caused sedative effects when combined with opioids. We conclude that coadministration of alphadolone, but not alphaxalone, with morphine, fentanyl, or oxycodone potentiates antinociception and that this effect is not caused by an increase in sedation.

In Brief

IMPLICATIONS: Sedative effects of drugs that cause pain relief can limit the effective dose and also confuse results from animal experiments. This study separates the sedative and analgesic effects for two neurosteroids and three opioids given alone and in combination. The neurosteroid, alphadolone, increased the pain-relieving effects of opioids without any sedation.

Author Information

Monash University Department of Anaesthesia, Monash Medical Centre, Clayton, Victoria, Australia

This work was supported by a scholarship from the Australian Pain Society. Neurosteroid compounds were supplied by Jurox, Rutherford, NSW.

Use of new analgesics is the subject of US patent 6,048,848 filed by Goodchild and Nadeson.

Accepted for publication April 23, 2003.

Address correspondence and reprint requests to Ms. Lara Winter, Monash University Department of Anaesthesia, Monash Medical Centre, 246 Clayton Rd., Clayton, Victoria, Australia 3168. Address e-mail to

Article Outline

Selye (1) demonstrated that certain pregnanes and androstanes have potent anesthetic and anticonvulsant effects occurring within a few minutes of administration. Since these early findings, it has become apparent that some steroids can interact directly with a surface membrane receptor complex to cause a change in central nervous system excitability (2). The discovery of fast-acting steroid sedatives led to the development of CT1341 (Althesin®) as an IV anesthetic. This contained two active steroids, alphaxalone and alphadolone, with an aqueous vehicle, Cremophor EL (polyethoxylated castor oil surfactant). It was stated in the literature of the time that the anesthetic component of Althesin® was mainly attributed to alphaxalone, whereas alphadolone, with one-third the anesthetic activity of alphaxalone, was present to improve solubility (3). Althesin® was removed from clinical practice because of major anaphylactoid reactions that were attributed to the vehicle. The same preparation of the two neurosteroids continued to be available for veterinary use (Saffan®).

Previous studies have shown that Saffan® could inhibit the activity of spinal cord neurones by an action at gamma-amino butyric acid (GABA)A receptors (4–6). Subsequent investigations in awake, intact rats showed that Saffan® could cause antinociceptive effects when assessed by the electrical current threshold (ECT) test but not the tail flick test, and that these effects were attributed to the alphadolone content of the mixture (7,8). Furthermore, these researchers showed that alphadolone caused these antinociceptive effects by an action at spinal cord GABAA receptors, even when the drug was given by a nonspinal parenteral route (intraperitoneally [ip]). A similar antinociceptive profile has been shown for the benzodiazepine midazolam given intrathecally, whereby it causes spinally mediated antinociception by an action at spinal cord GABAA receptors, revealed by noxious electrical current but not tail flick latency (TFL) (9,10). Subsequent studies showed that although midazolam, given alone intrathecally does not affect TFL, it does potentiate the TFL antinociceptive effects of intrathecally administered morphine (11). This paradigm may therefore be applied to the investigation of combination experiments between opioids and neurosteroids. Like midazolam, alphadolone has no effect on TFL at doses reported to be antinociceptive in the electrical nociceptive test. Therefore, alphadolone causing an increase in TFL thresholds for opioid antinociception can properly be called potentiation (12). We set out to investigate the possibility that such potentiation of opioid antinociception could occur with either neurosteroid alphadolone or alphaxalone given ip at doses that did not cause sedation.

Back to Top | Article Outline


This work was done with permission of the Monash University Standing Committee on Ethics in Animal Experimentation (SCEAE Project No. 93017). In all experiments, attention was given to ethical guidelines for the investigation of experimental pain in conscious animals (13). All experiments were performed on male Wistar rats (180–200 g) between the hours of 9:00 am and 5:00 pm. All drugs used were administered via the ip route.

Initial experiments were performed to identify the largest doses of the two neurosteroids and three opioids that did not cause sedation. Individual drug doses that were not sedative were tested for antinociceptive effects. Finally, experiments were performed in which each neurosteroid was administered in combination with each of the opioids. This was to determine whether opioid antinociception could be increased by coadministration of a neurosteroid without increasing sedation.

Back to Top | Article Outline
Tests of Sedation

All drugs were assessed for their effect on motor coordination and sedation using the rotarod treadmill (7650 accelerator rotarod; Ugo Basile, Italy). The training procedure was conducted on the day before the experiment. Training consisted of the animals being placed on the rotating drum at a steady slow speed of four revolutions per minute. Every time an animal fell, it was immediately placed back on the wheel until 15 min of continuous running had elapsed. This training procedure was repeated after the animal had been allowed to rest for at least 30 min. Once trained, these animals were used in one experiment per day for a maximum of four consecutive days.

Before experimentation, each animal was retrained for 15 min of continuous running on the steadily rotating wheel set at a speed of 4 revolutions per minute. For each experiment, an animal was injected with the test drug and 5 min later placed onto the rotating wheel in an isolated section. At this point, a timer started and the wheel accelerated from 4 revolutions per minute with a steady acceleration of 20 revolutions per minute every minute thereafter. The rotarod performance time was the time taken for the animal to fall from the wheel after the wheel began accelerating. Animals were removed from the drum after 2 min to prevent fatigue. Earlier studies have shown that all normal untreated rats run continuously for 2 min. Therefore, 2 min was used as the control cut-off or maximal runtime for the test. During the test period of 30 min, the animals’ runtimes were measured on the wheel 3 times, at 10-min intervals. The shortest runtime of the 3 measured during the 30-min test period was identified for each experiment.

Back to Top | Article Outline
Activity Monitor.

Wistar rats were observed in a commercially available open-field arena in which locomotor and exploratory activity could be monitored in darkness by the breaking of infrared beams arranged in a grid pattern over the entire area (MedAssociates Inc., St. Albans, VT). In these experiments, the animals were injected with a drug and 5 min later they were placed in the activity monitor in which their activity was monitored for 20 min. To avoid habituation to the activity monitor, animals were only used twice in this test for sedation and each experiment was separated by at least 48 h; vehicle control experiments performed earlier identified that two tests performed in such a manner produced the same results as in naïve rats used only once. Rest time was defined as the time spent with no new infrared beam interruptions. Each set of experiments was performed with matching vehicle controls, thus producing comparison data to determine significant sedative effects of the opioid and neurosteroid compounds.

Back to Top | Article Outline
Tests of Nociception

Nociceptive testing was performed once per day on each rat for a maximum of five consecutive occasions. The animals were placed in a Plexiglass restrainer that was covered to exclude distracting sights and sounds. ECT and TFL tests were used to assess nociceptive thresholds every 5 min. Because ip administration was used, to avoid pharmacokinetic differences between drugs, time response relationships were constructed for all drugs. For example, we plotted antinociceptive responses after ip injection every 5 min until their response had occurred, reached a maximum, and began to wane. From these measurements, we decided on which time points were used for calculations.

Back to Top | Article Outline
ECT Test.

This test was only performed with the neurosteroids to determine doses that caused antinociception. Wire electrodes were applied to the surface of the skin 1 and 5 cm from the base of the tail. These electrodes were connected alternately to a constant current electrical stimulator (50 Hz; 2-ms pulses; 0.5-s train) as previously described (14). By using the up/down method with repeated stimuli every 10–20 s, we defined the minimal electrical current that caused vocalization. This was done every 5 min. Responses were standardized by dividing the thresholds measured after drug administration by those obtained before drug injection. Thus, the drug response was expressed as a ratio of control (predrug) value:MATH

Back to Top | Article Outline
TFL Test.

The tip of the tail was painted black. The heat from an infrared beam on the tail flick unit was focused onto the blackened part of the tail, and the time taken for the animal to remove its tail from the source of the heat was measured. TFL increase was standardized by calculating the percentage of maximum possible effect (%MPE) as follows:MATH

The cut-off time of 10 s was the maximal time allowed for the animal to flick its tail away from the heat source before the lamp was automatically switched off to avoid damage to the tail. Values from individual experiments were combined to produce mean ± sem response for each dose used and these values plotted on dose-response curves (15).

Back to Top | Article Outline
Nociceptive Testing Protocol.

Both ECT and TFL were used in the experiment in which the neurosteroids were administered alone. The thresholds were measured every 5 min and the order of tests was TFL then ECT. Three stable consecutive control readings were obtained (see Fig. 1). Using only nonsedative doses (identified from previous sedation tests), a drug or drug combination was then injected ip and nociceptive thresholds were measured every 5 min for the following 30 min. The 3 postinjection readings used for calculations began 5 min after the final injection. Two ip injections were given in all experiments as shown in Figure 1. For experiments in which either neurosteroid or opioid was given alone, the other injection was vehicle. For combination experiments, the neurosteroid was given followed by an opioid 5 min later as shown in Figure 1.

At least 24 h elapsed between experiments on the same rat. As in previous studies using this model, the order of doses administered to individual rats was randomized. A specific dose that was the first to begiven in the experiment series was compared with the responses obtained for that same dose in experiments performed later in the week on different rats. There was no difference in the magnitude of these responses and thus there was no evidence of tolerance or residual drug effect from the previous day’s experiments. The maximal number of experiments in which nociceptive thresholds were measured in any individual rat was five.

Back to Top | Article Outline
Experimental Paradigm

The activity monitor was used to determine the sedative effects of the following drugs in 100 rats: alphadolone-CD (1–200 mg/kg, n = 5–10 replicate experiments at each dose; Jurox, Rutherford, NSW), alphaxalone (10–50 mg/kg, n = 5–10 replicate experiments at each dose; Jurox), fentanyl citrate (2.5–80 μg/kg, n = 5–8 replicate experiments at each dose; Sublimaze, Janssen-Cilag), morphine sulfate (0.4–12.8 mg/kg, n = 5–8 replicate experiments at each dose; David Bull Laboratories), oxycodone hydrochloride (0.125–2 mg/kg, n = 5–8 replicate experiments at each dose; Endone, Boots, North Ryde, NSW).

The rotarod was also used to determine sedative effects of the same drug dose ranges in 70 rats: alphadolone (1–200 mg/kg, n = 5–8 replicate experiments at each dose), alphaxalone (10–50 mg/kg, n = 5 replicate experiments at each dose), fentanyl (2.5–80 μg/kg, n = 5–10 replicate experiments at each dose), morphine (0.4–12.8 mg/kg, n = 5–8 replicate experiments at each dose), oxycodone (0.125–2 mg/kg, n = 5 replicate experiments at each dose).

Nonsedative doses revealed antinociception. The antinociceptive effects for the neurosteroids were assessed in 35 rats: alphadolone (0.06–10 mg/kg, n = 5–8 replicate experiments at each dose), alphaxalone (10 mg/kg, n = 5 replicate experiments at each dose).

All nonsedative doses of opioids were tested for antinociception using TFL in 15 rats: fentanyl (2.5–40 μg/kg, n = 5 replicate experiments at each dose), morphine (0.4–6.4 mg/kg, n = 5–6 replicate experiments at each dose), and oxycodone (0.125–1 mg/kg, n = 5 replicate experiments at each dose). These curves were repeated in the presence of 10 mg/kg of alphadolone for each of fentanyl (2.5–40 μg/kg, n = 5–10 replicate experiments at each dose), morphine (0.4–6.4 mg/kg, n = 5–7 replicate experiments at each dose), and oxycodone (0.125–1 mg/kg, n = 5–10 replicate experiments at each dose) in 20 rats. Opioids were also tested in combination with a nonsedative dose of alphaxalone (10 mg/kg) in 10 rats: fentanyl (5–20 μg/kg, n = 5 replicate experiments at each dose), morphine (0.8–3.2 mg/kg, n = 5 replicate experiments at each dose), and oxycodone (0.25–0.5 mg/kg, n = 5 replicate experiments at each dose).

The neurosteroid-opioid combinations were also tested for sedation in the activity monitor and rotarod in 20 rats: fentanyl 20 mg/kg (n = 5 replicate experiments), morphine 3.2 mg/kg (n = 5 replicate experiments), and oxycodone 0.25 mg/kg (n = 5 replicate experiments).

Back to Top | Article Outline
Vehicle Controls

Vehicle control experiments were performed for all individual drugs in all behavioral tests. The neurosteroids were dissolved in phosphate buffer with cyclodextrin. Saline was used to dissolve the opioids.

Back to Top | Article Outline
Data Analysis

For dose-response curves, the ECT responses in the tail and TFL values (%MPE) were combined for each dose of agonist and plotted as means ± sem. The TFL responses (%MPE) in the experiments investigating combinations were also combined and expressed as means ± sem for each drug treatment; these were plotted as dose-response curves. These dose-response curves for the opioid-neurosteroid combinations were compared with the dose-response curves for the respective opioid used alone using two-way analysis of variance (ANOVA) followed by the Dunnett’s t-test for post hoc comparisons. Runtime results for the rotarod were combined for each dose of drug or drug combination and expressed as means ± sem. These were compared with the control maximal runtime of 2 min using one-way ANOVA followed by the Dunnett’s t- test for post hoc comparisons. Resting time results from the activity monitor were combined for each drug dose or combination and expressed as means ± sem. Statistical evaluation was performed by means of a one-way ANOVA, followed by the Dunnett’s t-test for post hoc comparisons. For all statistical comparisons, a value of P < 0.05 was considered statistically significant.

Back to Top | Article Outline


Vehicle Controls

Neither saline nor cyclodextrin/phosphate buffer caused significant sedative (activity monitor 982 ± 17, n = 16 compared with results shown below; rotarod results shown below) or antinociceptive effects (ECT 1.02 ± 0.27, n = 5; TFL 5.02% ± 4.75%, n = 5). The responses for antinociception obtained from the first experiment performed in a series was compared with the same experiment performed at the end of the 5-day test period and revealed no significant differences. Thus, the experimental paradigm did not cause habituation or tolerance and no residual drug effect was present from one day’s testing to the next.

Back to Top | Article Outline
Dose-Response Relationships for Neurosteroids and Opioids Given Alone
Tests of Sedation.

All rats injected with vehicle control solutions tested on the rotarod had a mean runtime of 120 s (sem ± 0, n = 10); i.e., all vehicle controls showed the normal runtime of a nonsedated rat. The vehicle-treated control rats tested in the activity monitor had a mean rest time of 937 s (sem ± 22 s; n = 53).

Tables 1 and 2 show the results in the rotarod and activity monitor tests for alphadolone, alphaxalone, fentanyl, morphine, and oxycodone given alone. Comparison of the rotarod and the activity monitor experiments showed that all the compounds tested caused dose-dependent sedation. It can be concluded from combining the results of both of these tests that no sedative effects occurred with alphadolone or alphaxalone at doses of ≤10 mg/kg, fentanyl at doses of ≤20 μg/kg, morphine at doses ≤3.2 mg/kg, or oxycodone at doses of ≤1.0 mg/kg. These doses represent the largest doses that were used for antinociceptive testing.

Back to Top | Article Outline
Tests of Nociception.

Figure 2 shows the dose-response relationships for the antinociceptive effects of alphadolone (A) and alphaxalone (B) assessed with the ECT and TFL tests. Nonsedative doses of alphadolone up to 10 mg/kg caused a dose-dependent increase in ECT values but had no effect on TFL. By contrast, the nonsedative dose of alphaxalone 10 mg/kg caused no change in ECT or TFL.

Back to Top | Article Outline
Combinations of Opioids and Neurosteroids

Dose-response curves for all 3 opioids assessed with TFL are shown in Figure 3 for the opioids given alone and in combination with alphadolone and alphaxalone (both at the dose of 10 mg/kg). All three opioids caused a dose-dependent antinociceptive effect assessed by TFL. Although alphadolone did not affect TFL when given alone, it significantly potentiated the opioid antinociceptive effects. By contrast, alphaxalone caused no such increase of nociceptive thresholds when combined with any of the opioids.

The rotarod results showed no decrease in runtime for combinations of alphadolone and opioids at doses that showed the largest potentiation of opioid antinociception by the neurosteroid (Table 3). In addition, the results from the activity monitor showed that combinations of alphadolone and opioids caused no change in rest times (Table 4). Thus, the increase in opioid antinociception caused by coadministration of alphadolone is described by the leftward shift of the dose-response curve (Fig. 2). It was not attributed to an increase in sedation caused by coadministration of the opioid with the neurosteroid. Both the rotarod and activity monitor also showed no decrease in runtime or increased rest times for combinations of alphaxalone with opioids (Tables 2 and 3).

Back to Top | Article Outline


The main conclusion from this study is that alphadolone and alphaxalone possess different properties with respect to antinociception. It is clear from published literature that they are very similar in structure and both compounds are IV anesthetics, differing only in potency; alphadolone causes sedation and anesthesia when given IV at one-third the potency of alphaxalone (3). The behavior of these two neurosteroids is quite different when they are administered ip or intragastrically compared with IV injection (7,8). In those experiments, it was shown that alphadolone caused antinociceptive effects with no signs of sedation using observations of righting reflex and crude observations of behavior. By contrast, alphaxalone caused sedation and anesthesia with no signs of antinociception even when given ip. The experiments reported here confirm and extend those observations with a detailed dose-response analysis of sedative and antinociceptive effects.

The search for sedative effects after ip administration was performed in detail using two well-established tests, rotarod and open-field exploration. A nonsedative dose of 10 mg/kg for both alphadolone and alphaxalone was identified. Alphadolone caused dose-dependent antinociceptive effects in the ECT test at doses smaller than those that caused sedation. By contrast, the same dose of alphaxalone caused no such antinociception. TFL was not affected by alphadolone, confirming previous reports (7,8). With respect to the relative potency of the two neurosteroids as anesthetics rather than sedatives, one can see that the doses that lead to inability to negotiate the rotarod and no movement in the activity box are 100–200 mg/kg for alphadolone and 25–50 mg/kg for alphaxalone. This relative potency is quite close to previous reports in which alphaxalone showed three times the anesthetic effects of alphadolone (3).

This study also demonstrates that alphadolone potentiates opioid antinociception. The term “potentiation” has been defined as an increased effect of a drug combination compared with an effect of a single drug, when the second drug in the combination is ineffective when given alone at the same dose (12). The effect of alphadolone on the antinociceptive effects of each of the opioids used in this study is properly called “potentiation,” because alphadolone given alone had no effect on TFL and yet the TFL responses to the opioids were significantly increased by the coadministration of alphadolone. This potentiation was achieved without causing sedation and is such an effect on antinociception rather than being secondary to sedation.

The mechanism of this interaction has been suggested by earlier work but requires further investigation. Alphadolone produces its antinociceptive effects by spinal cord GABAA receptors, even when the drug is administered ip or intragastrically (7,8). Alphadolone has no effect on TFL when administered alone. This is a property it shares with midazolam, a benzodiazepine that also only affects ECT when it is given intrathecally and again by an action at spinal cord GABAA receptors (9,10). Coadministration of midazolam with morphine intrathecally leads to marked potentiation of morphine TFL antinociception (11). This is a comparable result with that reported here, except in the experiments reported in this article, the drugs were given by a nonspinal route. The most likely mechanism for alphadolone potentiation of opioid antinociception is therefore positive modulation of GABAA receptors in the spinal cord.

We conclude that ip alphadolone, but not alphaxalone, causes antinociceptive effects and that it potentiates opioid-mediated antinociception without sedation. Furthermore, this may represent a drug interaction at the level of the spinal cord thus pointing to the possibility of an improvement in pain relief afforded by opioids without an increase in opioid-related side effects.

Back to Top | Article Outline


1. Selye H. Anesthetic effect of steroid hormones. Proc Soc Exp Biol Med 1941; 46: 116–21. Ref. type: generic.
2. Puia G, Santi MR, Vicini S, et al. Neurosteroids act on recombinant human GABAA receptors. Neuron 1990; 4: 759–65.
3. Stock JPP. A brief history of steroid anaesthesia before Althesin (CT1341). BMJ 1973; 4: 495.
4. Harrison NL, Majewska MD, Harrington JW, Barker JL. Structure-activity relationships for steroid interaction with the gamma-aminobutyric acidA receptor complex. J Pharmacol Exp Ther 1987; 241: 346–53.
5. Harrison NL, Vicini S, Barker JL. A steroid anesthetic prolongs inhibitory postsynaptic currents in cultured rat hippocampal neurons. J Neurosci 1987; 7: 604–9.
6. Mistry DK, Cottrell GA. Actions of steroids and bemegride on the GABAA receptor of mouse spinal neurones in culture. Exp Physiol 1990; 75: 199–209.
7. Nadeson R, Goodchild CS. Antinociceptive properties of neurosteroids. II. Experiments with Saffan and its components alphaxalone and alphadolone to reveal separation of anaesthetic and antinociceptive effects and the involvement of spinal cord GABA(A) receptors. Pain 2000; 88: 31–9.
8. Nadeson R, Goodchild CS. Antinociceptive properties of neurosteroids. III. Experiments with alphadolone given intravenously, intraperitoneally, and intragastrically. Br J Anaesth 2001; 86: 704–8.
9. Edwards M, Serrao JM, Gent JP, Goodchild CS. On the mechanism by which midazolam causes spinally mediated analgesia. Anesthesiology 1990; 73: 273–7.
10. Goodchild CS, Serrao JM. Intrathecal midazolam in the rat: evidence for spinally mediated analgesia. Br J Anaesth 1987; 59: 1563–70.
11. Yanez A, Sabbe MB, Stevens CW, Yaksh TL. Interaction of midazolam and morphine in the spinal cord of the rat. Neuropharmacology 1990; 29: 359–64.
12. Dickenson AH, Sullivan AF. Combination therapy in analgesia: seeking synergy. Curr Opin Anaesthesiol 1993; 6: 861–5.
13. Zimmermann M. Ethical guidelines for investigations of experimental pain in conscious animals. Pain 1983; 16: 109–10.
14. Serrao JM, Stubbs SC, Goodchild CS, Gent JP. Intrathecal midazolam and fentanyl in the rat: evidence for different spinal antinociceptive effects. Anesthesiology 1989; 70: 780–6.
15. Nadeson R, Guo Z, Porter V, et al. gamma-Aminobutyric acidA receptors and spinally mediated antinociception in rats. J Pharmacol Exp Ther 1996; 278: 620–6.
© 2003 International Anesthesia Research Society