(9), and Ca 2ϩ influx into lymphocytes (10). 5-HT 3 receptor activation mediates emetic and inflammatory responses (11) and may contribute to pain reception, anxiety, cognition, cranial motor neuron activity, modulation of affect, and the behavioral consequences of drug abuse (Refs. 12 and 13; but see Ref. 14).The 5-HT 3A receptor subunit shares structural similarities with members of the superfamily of ligand-gated ion channels (15) and is thought to be a pentameric protein (16, 17) with multiple agonist and allosteric ligand binding sites (2, 11). Both native and recombinant 5-HT 3 receptors reveal rapid and cooperative activation by agonists and desensitization to prolonged application of 5-HT (reviewed in Refs. 1, 2, and 6). With a few exceptions, ligand-gated channels require the association of more than one kind of homologous subunit for function, and subunit composition determines the pharmacological (18) and kinetic (e.g. desensitization (19,20)) profile of heteromeric receptors. While the 5-HT 3A subunit expressed in heterologous systems functions efficiently as homomers, different voltagedependence, desensitization, and pharmacological properties between recombinant and native 5-HT 3 receptors suggest that native 5-HT 3 receptors may exist as heteromers (3,(21)(22)(23)(24)(25)(26)(27). Although the 5-HT 3A gene encodes splice variants in mouse (28, 29) and guinea pig (30), most of the characteristics of these variants are similar (29,31,32), and the subtle pharmacological differences (22, 31, 32) cannot completely account for the differences observed between recombinant homomers and native 5-HT 3 receptors. In fact, co-expression of both splice variants in oocytes could not reproduce the responses observed in the cell line from which the splice variants were cloned (22). Biochemical studies on porcine brain have revealed the existence of at least four proteins (52-71 kDa) closely associated with the 5-HT 3A subunit that may represent antigenically distinct channel subunits (33).The cloning of a 5-HT 3B subunit that modified the pharmacological and single channel characteristics of the 5-HT 3A subunit when co-expressed in heterologous expression systems was recently reported (34). We extend this report and provide evidence for the co-existence of both 5-HT 3B and 5-HT 3A receptor subunits in native cells, a requirement for heteromeric association. We show that 5-HT 3B has no effect on the function of ␣22, ␣34, ␣42, and ␣7 nicotinic ACh receptors. Furthermore, the characterization of the pharmacology and function of heteromeric receptors in Xenopus oocyte and mammalian expression systems reported here differs from that described (34), and these differences have important consequences for the function of heteromers in vivo.* The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.The nucleotide sequence (s) 1 The abbreviations used...