Heteromeric neuronal nicotinic acetylcholine receptors (nAChRs) expressed from an and a β subunit have long been thought to have a stoichiometry of 2 and 3β subunits, as suggested by reporter mutation studies on 4 β2 and 3 β4 nAChRs expressed in Xenopus oocytes (Cooper et al., 1991; Boorman et al., 2000). Recently, 4β2 nAChRs have been shown to be capable of forming functional receptors with stoichiometries of both 2 :3β and 3 :2β , when stably transfected into HEK293 cells (Nelson et al., 2003).

We tested whether peripheral-type 3 β4 neuronal nicotinic receptors expressed in a mammalian cell line have the same stoichiometry as in oocytes. Wild-type or mutant human 3 and β4 subunits (in a 1:1 ratio) were transfected into HEK293 cells. Dose-response curves to ACh were obtained for each receptor type by whole-cell patch-clamp recording. Inserting a leucine to threonine mutation in the middle of the second transmembrane domain (L9'T) increased the receptor agonist sensitivity for all receptor types. The effect was more marked when the mutation was present in the 3 subunit than when it was in the β4 subunit: the EC₅₀ value decreased from its wild-type value of 86 ± 7.8 µM (n=6) to 18 ± 1.8 µM for 3 β4 ^L9'T (n=4) and 4 ± 0.4 µM for 3 ^L9'T β4 (n=4).

Thus, unlike in oocytes, 3β4 nAChRs expressed in a mammalian system contain on average more a than b subunits in the pentamer when expressed in the 1:1 ratio. This raises the possibility that more than one type of 3 β4 receptor is formed within heterologous expression systems. We decided to check if we could force the assembly of one stoichiometry by changing the :β ratio to either 9:1 or 1:9, in order to overexpress either the 3 or the β4 subunit in oocytes. Receptors expressed from the 1:9 ratio (overexpression of β4) had an ACh EC₅₀ that was very similar to that of receptors expressed from a 1:1 ratio (EC₅₀ = 133 ± 8.2; nH = 1.77 ± 0.17, n=4 and EC₅₀ =134 ± 6 µM nH = 1.73 ± 0.10, n=10, respectively). Overexpression of the 3 subunit, however produced a receptor population with a higher EC₅₀ (EC₅₀ = 300 ± 17; nH = 2.32 ± 0.21, n=4). These results were confirmed by the markedly different EC₅₀ values for 3 + β4 L9’T given by the two ratios (EC₅₀: 0.75 ± 0.05 µM vs 38.8 ± 1.4 µM).

In conclusion, different types of 3β4 receptors can be formed in heterologous expression systems. Which type of receptor is predominant depends on the expression system used. Because heterologous expression systems are used in the characterisation of pharmacological agents for drug screening, it will be important to determine which expression system most faithfully reproduces the properties of native nAChRs.

Boorman et al. (2000) J. Physiol. 529 (3), p565-77.
Cooper et al. (1991) Nature350 (6315), p235-38.
Nelson et al., (2003) Mol. Pharmacol.63 (2), 332-41.

This work has been sponsored by a Wellcome Trust grant (064652) and the MRC (to SB).