Secondary Logo

Journal Logo

Nondepolarizing Neuromuscular Blockers Inhibit the Serotonin-Type 3A Receptor Expressed in Xenopus Oocytes

Min, Kyeong T. MD*; Wu, Christopher L. MD*; Yang, Jay PhD, MD*†

doi: 10.1213/00000539-200002000-00044

Molecular cloning and sequence comparison indicates a high degree of structural homology between muscle nicotinic acetylcholine (nACh) and serotonin-type 3 (5-HT3A) receptors, both members of the direct ligand-gated family of ion channels. Because of the structural similarities and common evolutionary origin of these receptors, neuromuscular blockers (competitive nACh antagonists) may demonstrate pharmacologic cross talk and exhibit attributes of 5-HT3A receptor antagonists. We examined six clinically-used neuromuscular blockers for their ability to antagonize currents flowing through the 5-HT3A receptors in voltage clamped Xenopus oocytes. The neuromuscular blockers reversibly inhibited the 5-HT3A receptor-gated current in the rank order potency of (IC50 mean ± SEM):d-tubocurarine (0.046 ± 0.003 μM), atracurium (0.40 ± 0.03 μM), mivacurium (15.1 ± 2.93 μM), vecuronium (16.3 ± 2.24 μM), and rocuronium (19.5 ± 2.31 μM). Gallamine was essentially inactive as a 5-HT3A receptor antagonist with an extrapolated IC50 of 1170 μM. We demonstrate that drugs classically known as competitive nACh receptor antagonists also block 5-HT3A receptors. It is likely that certain neuromuscular blockers share pharmacological properties with 5-HT3A receptor antagonists, such as a reduction in postoperative nausea and vomiting. With careful drug selection, pharmacological cross talk could potentially be used to minimize polypharmacy and optimize patient management.

Implications Muscle nicotinic acetylcholine and serotonin-type 3A (5-HT3A) receptors are similar. Therefore neuromuscular relaxants may block 5-HT3A receptors. Our pharmacological study demonstrates that neuromuscular relaxants, as with ondansetron, are 5-HT3A receptor antagonists. It is likely that certain neuromuscular relaxants exhibit ondansetron-like clinical properties, such as reduction in postoperative nausea and vomiting.

*Departments of Anesthesiology and †Pharmacology/Physiology, University of Rochester Medical Center, Rochester, New York

October 22, 1999.

Supported by a fellowship from the Yonsei University School of Medicine, Seoul, Korea, National Institutes of Health Grant RO1GM52325, and funded by the Department of Anesthesiology, University of Rochester Medical Center, Rochester, NY.

Address correspondence and reprint requests to Jay Yang, Department of Anesthesiology; Box 604, University of Rochester Medical Center, 601 Elmwood Ave., Rochester, NY 14642. Address e-mail to

The choice of a neuromuscular blocking drug generally has not been considered an important risk factor for postoperative nausea and vomiting (PONV) (1,2). However, a large meta-analysis of factors affecting PONV revealed varying probabilities of PONV based on the neuromuscular blocking drug administered with the least probability of PONV occurring in those who received alcuronium (3). Alcuronium and d-tubocurarine (d-TC) are structurally related (natural alkaloid derivatives) and d-TC binds to and acts as an antagonist at the serotonin-type 3A (5-HT3A) receptor (4,5). Therefore, neuromuscular blocking drugs, much like 5-HT3A receptor antagonists, may have antiemetic properties and affect the incidence of PONV.

The 5-HT3A receptor is a member of the ligand-gated ion-channel (LGIC) superfamily based on sequence homology and shares many structural similarities with nicotinic acetylcholine (nACh), glycine and γ-aminobutyric acid (GABAA) receptors (6). Thus, certain drugs, such as d-TC which is traditionally known as a specific competitive inhibitor of the nACh receptor, may also act as an antagonist at a related receptor of the same superfamily. For example, d-TC reversibly blocks the 5-HT3A receptor (5–7), anion permeable inhibitory glycine, and GABA receptors as well (8–9).

Because of the potential clinical benefit of 5-HT3A receptor antagonism and potential pharmacologic cross talk of neuromuscular blockers at these receptors, we investigated the effects of clinically used neuromuscular blockers on 5-HT3A receptor function.

Back to Top | Article Outline


A cDNA clone encoding the full-length murine 5-HT3A receptor (kindly provided by L. Tecott, University of California San Francisco) was subcloned into a custom oocyte expression vector. This expression vector is a conventional pBS II/KS+ plasmid modified by us such that the multiple cloning site is flanked by the Xenopus globin 5′ and 3′ untranslated sequence. The construct was linearized by a Sac I digestion, proteinase K and SDS treated to prepare a RNAse free template, and cRNA synthesized (Message Machine; Ambion, TX) in vitro from the T3 polymerase site following the manufacturer’s recommended protocol. Approximately 50 ng of cRNA was injected into Stage IV oocytes by using a microinjector (Drummond Scientific, Broomall, PA). Thus the receptors formed are 5-HT3A homoligomeric receptors. The protocol for oocyte harvest was approved by the University of Rochester Committee on Animal Care.

After a 24–72 h incubation period, oocytes were placed into a recording chamber approximately 500 μL in volume and continuously perfused with frog Ringer’s solution (in mM: 115 NaCl, 2 KCl, 1.3 Na2HPO4, 1.8 CaCl2; pH 7.4) at 3 mL/min. The oocytes were penetrated with two glass microelectrodes with resistance of approximately 1 M Ohm when filled with 2 M KCl solution. Electrophysiological recordings were obtained by the two-electrode voltage clamp technique. Feedback gain on the voltage clamp amplifier (Oocyte Clamp; Warner Instruments, Hamden, CT) was adjusted for optimal stability by using 10-mV voltage steps until a slightly underdamped response current was attained. Serotonin, with or without various drugs, was bath applied by a computer-controlled solenoid valve. The bath solution exchange time constant was approximately 0.5 s and orders of magnitude slower than the biochemical interactions between ligand and receptors. Preliminary experiments indicated that preapplication or simultaneous application of the competitive antagonists with serotonin made no difference on the resultant current magnitude. Therefore, the currents measured during the 20–30 s drug co-application can be considered as the equilibrium current resulting from steady state action of the competitive antagonists. The timing of drug application and current digitization were controlled by Clampex v 5.2 (Axon Instruments; Burlingame, CA). With 2.5 μM serotonin as the agonist, currents evoked by repeated applications were stable, with no evidence of cumulative desensitization with a drug application interval of 90 s. At higher serotonin concentrations, the application interval was increased to 300 s.

Clinical formulation of neuromuscular blockers (mivacurium, atracurium, rocuronium, and vecuronium), d-TC, and gallamine (RBI, St. Louis, MO) were diluted in frog Ringer’s solution. The final concentration of benzyl alcohol present in the highest concentration of mivacurium and atracurium examined was <0.009%.

Peak currents induced by the drug applications were measured and dose responses fit (Sigmaplot; Jandel Corp., CA) to the equation I = Imax/(1 + {ED50/[C]}n) or I = Icont/(1 + {IC50/[A]}n) where n = Hill coefficient and [C] and [A] are concentrations of agonist (serotonin) and antagonist (neuromuscular blockers), respectively. Statistical significance of the estimated IC50 values for the different neuromuscular blockers was determined by a Student’s t-test with Bonferoni correction for multiple comparisons. Symbols on the dose-response figures represent mean ± SEM.

Back to Top | Article Outline


Because the apparent inhibitory potency of a competitive antagonist depends on the relative amount of agonist used, we first characterized the serotonin concentration response of the expressed 5-HT3A receptor. Serotonin applications induced concentration-dependent currents (EC50 = 2.52 ± 0.17 μM, Hill coefficient = 2.09 ± 0.24, based on 17 cells) in oocytes injected with the 5-HT3A cRNA (Figure 1A). A specific 5-HT3A agonist, 2Me5-HT, also induced currents in injected oocytes, not in control oocytes (data not shown). Serotonin-induced currents were reversibly blocked by the nonspecific antagonist metoclopramide (1 μM), and 5-HT3A receptor specific antagonists, ondansetron (30 nM) and MDL72222 (1 μM). In contrast, 5-HT1 and 5-HT2 receptor specific antagonists, methiothepin or ketanserin (10 μM for both), only had minimal effects on the currents (Figure 1B–F). The IC50 for both antagonists for their respective receptors is in the nM range (10,11). These pharmacological properties confirm that the observed current is mediated by the 5-HT3A receptor and is consistent with a previous report (6).

Figure 1

Figure 1

Coapplication of 2.5 μM serotonin with clinically used neuromuscular nACh blockers decreased the current magnitude (Figure 2, A–F). The neuromuscular blocker inhibition of serotonin-gated current was reversible with washout of the antagonist. The IC50 for inhibition of 5-HT3A receptor mediated current by d-TC, atracurium, mivacurium, vecuronium and rocuronium was (in μM): 0.046 ± 0.003, 0.40 ± 0.03, 15.1 ± 2.93, 16.3 ± 2.24 and 19.5 ± 2.31, respectively (based on 4 to 8 oocytes). Gallamine, a ganglionic nACh blocker, was a poor 5-HT3A antagonist with an extrapolated IC50 of 1170 μM (Figure 2G, Table 1). The estimated IC50 values for the respective neuromuscular blockers were significantly different (P < 0.05) from each other, except between the three steroidal neuromuscular blockers (mivacurium, vecuronium, and rocuronium). Atracurium antagonism of serotonin-induced current was reversed even in the presence of atracurium by increasing the serotonin concentration, indicating that the antagonism was competitive (Figure 3A). The reversal potential for the serotonin-induced current determined by a ramp-voltage command (Figure 3B) remained unaltered by atracurium (−8.5 ± 0.3 mV for control, n = 4 oocytes versus −8.9 ± 0.4 mV, n = 4 oocytes). The degree of blockade measured as the ratio of current remaining after blockade/control current was also voltage-independent: 0.15 ± 0.2 (−70 mV); 0.14 ± 0.1 (+ 40 mV) ondansetron; 0.14 ± 0.2 (−70 mV); 0.13 ± 0.04 (+ 40 mV) atracurium. These observations indicate that these antagonists work from the extracellular side and that their action has no effect on channel-ion selectivity.

Figure 2

Figure 2

Table 1

Table 1

Figure 3

Figure 3

Figure 3C shows the correlation between potency of neuromuscular blockers as a neuromuscular blocker (vertical axis) and their potency as a 5-HT3A blocker (horizontal axis). The one-half maximal blocking concentrations for the neuromuscular blockers were calculated from pharmacokinetic variables assuming: 1) total distribution of a given dose into the volume of distribution; 2) no drug to drug interactions; and 3) only free drug is pharmacologically active (see Table 1 legend). The dotted line indicates equipotency of nACh and 5-HT3A antagonism. Neuromuscular blockers (mivacurium, vecuronium, rocuronium, and gallamine) falling to the right of the line display relatively more antagonism of the nACh than 5-HT3A receptor. Atracurium falls right on the line (equipotent nACh and 5-HT3A antagonist) and d-TC falls to the left of the line, thus indicating relatively more antagonism of 5-HT3A than nACh receptor.

Back to Top | Article Outline


Members of the LGIC superfamily include nACh, 5-HT3A, GABAA, and glycine receptors (12). The LGICs share a common evolutionary ancestor, and the receptors are made of similar subunits; therefore, there is a structural basis for potential pharmacological cross talk among different LGIC receptors and ion channels. The 5-HT3A and nACh receptors are selective for cations (excitatory) and are structurally closer to each other than to the other LGICs, GABAA, and glycine, which are selective for anions and inhibitory (12). However, d-TC antagonism of receptor members of this superfamily extends to both cation and anion selective channels.

Although the precise structural determinants for d-TC antagonism of various members of LGICs remains unknown, studies of nACh receptors implicate a dodecapeptide in the vicinity of the highly conserved pair of adjacent cystines immediately extracellular to the first transmembrane portion of the receptor subunit polypeptide as the critical region (13). In fact, the d-TC binding pocket seems to be formed at the interface of two different subunits in a heteroligomeric nACh receptor, resulting in two distinct sites with binding affinities differing by 100-fold (14). Further analysis of d-TC interaction with homoligomeric receptors, such as 5-HT3A or glycine α where heterogenous subunit interface should not exist, may reveal a single homogenous population of binding sites.

In our study, we demonstrate an example of pharmacological cross talk where drugs classically known for their antinicotinic cholinergic action also block 5-HT3A receptor mediated current. Previous studies (4–9) have also shown that d-TC acts as an antagonist at the 5-HT3A, GABA, and glycine receptors. We found that neuromuscular blockers exhibited varying potencies as 5-HT3A receptor antagonists with d-TC and atracurium displaying relatively more antagonism than vecuronium, rocuronium, and gallamine. We also noted that d-TC showed relatively more antagonism of the 5-HT3A than nACh receptor. The fact that d-TC is an equal or superior antagonist at the 5-HT3A than nACh receptor has also been documented in other studies (5,7). The d-TC antagonism of 5-HT3A receptor was competitive, was voltage independent, and had no effect on the reversal potential. These properties are shared with ondansetron, a prototypic 5-HT3A receptor antagonist, and are consistent with the presumed site of action of these drugs outside the transmembrane voltage field (i.e., acting from the extracellular side). Recent studies (15,16) indicate that d-TC potency as a serotonin receptor antagonist differs depending on species and whether the receptors are expressed as a homoligomer or as a heteroligomer in combination with the 5-HT3B receptor subunit. Curiously, metoclopramide and ondansetron potency was not dependent on the presence of a 5-HT3B subunit. Whether these factors effect other neuromuscular blocker antagonism of 5-HT3 receptors is not known.

Neuromuscular blockers are one class of drugs routinely used in general anesthesia; however, they often are not considered to be an influence on the incidence of PONV. Although some studies (17,18) comparing different neuromuscular blockers have found no difference in incidence of PONV, a larger meta-analysis of factors affecting incidence of PONV in patients undergoing gynecological surgery found that use of a neuromuscular blocker to be one of the most important factors in developing PONV with the greatest probability occurring in those who received pancuronium and alcuronium (3). A logistic regression analysis of two variables in 13 categories was performed on data derived from 1442 patients in a multicenter study. Well recognized factors influencing PONV (use of ondansetron prophylaxis, type of surgery, and age) were confirmed as contributory factors in this study.

Of relevance to our study is the observation by Haigh et al. (3) that the adjusted probability of development of PONV varied with the specific neuromuscular blocker used. The adjusted probability of development of PONV was found to be as follows in greatest-to-least order: pancuronium, vecuronium, atracurium, and alcuronium, a curare-like alkaloid of long duration of action not clinically available in the United States. This correlates with our findings of relative potency of neuromuscular blockers as 5-HT3A receptor antagonists. Neuromuscular blockers exhibiting the least-to-greatest amount of 5-HT3A receptor antagonism were: gallamine, rocuronium, mivacurium, vecuronium, atracurium, and d-TC. Compared with steroidal compounds, the benzylisoquinolinium neuromuscular blockers, d-TC, and atracurium, were the more potent 5-HT3A receptor antagonists. Whether alcuronium, which is structurally similar to d-TC, blocks the 5-HT3A receptor-like d-TC is not known. The decrease in adjusted probabilities of nausea and vomiting caused by the use of alcuronium was equivalent to 8 mg of ondansetron. Although neuromuscular blockers per se are not thought to play an important role in development of PONV (1–3), there are no other studies with comparable statistical power that have examined the potential antiemetic action of neuromuscular blockers.

It is possible that certain neuromuscular blockers share antiemetic properties with 5-HT3A receptor antagonists, such as ondansetron, which are widely recognized for their ability to prevent PONV (19). By judicious drug selection, such pharmacological cross talk could be of benefit clinically. It is unknown whether reversal of neuromuscular blockade with anticholinesterases will reverse any antiemetic effects of the neuromuscular blockers (20,21). The estimated free serum concentrations of d-TC and atracurium attained at the time of bolus administration of these drugs to facilitate intubation are most likely high enough to exhibit 5-HT3A blocking effects. It is also not known whether any residual neuromuscular blocker present at the time of emergence from general anesthesia will have sufficient pharmacological effect on the 5-HT3A receptors. However, if our hypothesis is correct, the results of previous antiemetic studies that allowed use of multiple neuromuscular blockers may be affected if one group received more of a certain type of neuromuscular blocker.

The 5-HT3A receptors are diffusely distributed in both the central and peripheral nervous system and those located in the area postrema of the brain and in the visceral afferent nerves are thought to mediate PONV (22). In addition to its well documented role in mediating nausea and vomiting, 5-HT3A receptors are thought to be involved in many other physiological processes, including peripheral nociception and central antinociception, conditioned aversion response to drugs, anxiety, and cognition (22). Although neuromuscular blockers do not cross intact blood-brain barriers, the area postrema is deficient in a blood-brain barrier, and macromolecules such as albumin (and thus, presumably neuromuscular blockers) enter the central nervous system through these areas (23). The potential antiemetic effect of neuromuscular blockers may be mediated by both central and peripheral 5-HT3A receptors.

In conclusion, we demonstrate that clinically used muscle relaxants, classically known and used as nACh receptor antagonists, are also potent 5-HT3A receptor antagonists. The benzylisoquinolinum class of neuromuscular blockers, in particular d-TC and atracurium, are potent 5-HT3A receptor antagonists exhibiting potency equal to or greater than nACh receptor blockers. PONV remains a significant complication of general anesthesia with an overall incidence of approximately 20%–30% (19). It is possible that certain muscle relaxants share pharmacological properties with 5-HT3A antagonists such as reduction in PONV.

The authors thank Dr. Paul Bigeleisen for comments on the manuscript and Ms. Christine Seccombe for help with preparation of the manuscript.

Back to Top | Article Outline


1. Kenny GNC. Risk factors for postoperative nausea and vomiting. Anaesthesia 1994; 49(Suppl):6–10.
2. Rabey PG, Smith G. Anaesthetic factors contributing to postoperative nausea and vomiting. Br J Anaesth 1992; 69(Suppl):40–5.
3. Haigh CG, Kaplan LA, Durham JM, et al. Nausea and vomiting after gynaecological surgery: a meta-analysis of factors affecting their incidence. Br J Anaesth 1993; 71:517–22.
4. Aprison MH, Galvez-Ruano E, Lipkowitz KB. Comparison of binding mechanisms at cholinergic, serotonergic, glycinergic and GABAergic receptors. J Neurosci Res 1996; 43:127–36.
5. Peters JA, Malone HM, Lambert JJ. Antagonism of 5-HT3 receptor mediated currents in murine N1E-115 neuroblastoma cells by + -tubocurarine. Neurosci Lett 1990; 110:107–12.
6. Maricq AV, Peterson AS, Brake AJ, et al. Primary structure and functional expression of the 5-HT3 receptor, a serotonin-gated ion channel. Science 1991; 254:432–7.
7. Vanner S, Surprenant A. Effects of 5-HT3 receptor antagonists on 5-HT and nicotinic depolarizations in guinea-pig submucosal neurones. Br J Pharmacol 1990; 99:840–4.
8. Curtis DR, Game CJ, McCulloch RM. Antagonism of inhibitory amino acid action by tubocurarine. Br J Pharmacol 1974; 52:101–3.
9. Lebeda FJ, Hablitz JJ, Johnston D. Antagonism of GABA-mediated responses by d-tubocurarine in hippocampal neurons. J Neurophysiol 1982; 48:622–32.
10. Awouters F. The pharmacology of ketanserin: the first selective serotonin S2-antagonist. Drug Dev Res 1985; 6:263–300.
11. Martin LL, Sanders-Bush E. Comparison of the pharmacological characteristics of 5-HT1 and 5-HT2 binding sites with those of serotonin autoreceptors which modulate serotonin release. Naunyn Schmiedebergs Arch Pharmacol 1982; 321:165–70.
12. Ortells MO, Lunt GG. Evolutionary history of the ligand-gated ion-channel superfamily of receptors. Trends Neurosci 1995; 18:121–7.
13. Akabas MH, Karlin A. Identification of acetylcholine receptor channel-lining residues in the M1 segment of the alpha-subunit. Biochem 1995; 34:12496–500.
14. Pedersen SE, Cohen JB. D-tubocurarine binding sites are located at alpha-gamma and alpha-delta subunit interfaces of the nicotinic acetylcholine receptor. Proc Natl Acad Sci (USA) 1990; 87:2785–9.
15. Davies PA, Pistis M, Hanna MC, et al. The 5-HT3B subunit is a major determinant of serotonin-receptor function. Nature 1999; 397:359–63.
16. Hope AG, Belelli D, Mair ID, et al. Molecular determinants of (+)-tubocurarine binding at recombinant 5-hydroxytryptamine 3A receptor subunits. Mol Pharmacol 1999; 55:1037–43.
17. Collins KM, Plantevin OM, Docherty PW. Comparison of atracurium and alcuronium in day-case gynaecological surgery. Anaesthesia 1984; 39:1130–4.
18. Zuurmond WWA, van Leeuwen L. Atracurium versus vecuronium: a comparison of recovery in outpatient arthroscopy. Can J Anaesth 1988; 35:139–42.
19. Watcha MF, White PF. Postoperative nausea and vomiting. Anesthesiology 1992; 77:162–84.
20. King MJ, Milazkiewicz R, Carli F, Deacock AR. Influence of neostigmine on postoperative vomiting. Br J Anaesth 1988; 61:403–6.
21. Hovorka J, Korttila K, Nelskyla K, et al. Reversal of neuromuscular blockade with neostigmine has no effect on the incidence or severity of postoperative nausea and vomiting. Anesth Analg 1997; 85:1359–61.
22. Hindle AT. Recent developments in the physiology and pharmacology of 5-hydroxytryptamine. Br J Anaesth 1994; 73:395–407.
23. Broadwell RD, Sofroniew MV. Serum protein bypass the blood-brain fluid barriers for extracellular entry to the central nervous system. Exper Neurol 1993; 120:245–63.
© 2000 International Anesthesia Research Society