The 5-HT₃ receptor is a member of the ligand-gated ion channel superfamily, and is thought to exist as a pentameric protein. The functional receptor can be assembled from 5-HT_3A subunits alone, or from this subunit in combination with other 5-HT receptor subunits (e.g. 5-HT_3B, 5-HT_3B), to yield receptors with differing properties. In addition, sequence analysis has revealed the existence of a differing number of potential phosphorylation sites within the large cytoplasmic loop, which may have biological significance for the expression of the different subunit isoforms.

This study aimed to identify kinases able to phosphorylate the 5-HT_3A and 5-HT_3B subunits, and then elucidate the specific amino acid residues targeted using phosphoamino acid analysis (PAA; Kamps et al., 1989) and site-directed mutagenesis.

cDNA sequences encoding the major cytoplasmic loop domains of the 5-HT_3A (3A) and 5-HT_3B (3B) subunits were expressed as GST-fusion proteins, purified from lysed cultures using glutathione agarose beads, and subsequently labelled with [³²P]-ATP, using a standard in vitro kinase assay (where n 3; MacDonald et al., 1997). PKA, PKC and CaMKII were found to specifically phosphorylate 3A, whereas 3B was only phosphorylated by PKC and CaMKII. No phosphorylation of GST alone was noted by any of the kinases. PAA was used to identify the amino acid residue phosphorylated in each transferred on to PVDF and the [³²P]-labelled bands excised and acid hydrolysed (5.7M HCl, 1hr, 110^oC) before resolution on 2-D cellulose thin layer chromatography (TLC) plates. Autoradiography of TLC plates revealed the identity of residue types phosphorylated in each reaction by comparison of the migration patterns of our [³²P]-labelled samples with those of ninhydrin stained, 'cold' phospho-amino acids.

PAA suggested that PKC phosphorylated a serine residue in both the 3A and 3B constructs. To date, no PAA data for CaMKII or PKA phosphorylation has been resolved.

QuikChange site-directed mutagenesis (Stratagene) was employed to mutate potential kinase target residues to alanine, and the in vitro kinase assay repeated. PKA phosphorylation of 3A was absent in the S389A mutant, whereas phosphorylation of 3B by PKC could be removed by creating a double serine knockout construct, S345A-S368A. The identity of the amino acids phosphorylated by PKC and CaMKII in 3A, and CaMKII in 3B remain to be elucidated. However, this apparent differential phosphorylation may represent a mechanism to differentially modulate 5-HT₃ receptor isoforms.

Future work includes the identification of the remaining kinase substrates in these sequences, followed by in vivo phosphorylation of full-length cloned h5-HT₃ in transiently transfected cell lines.

Kamps M.P. & Sefton B.M. (1989) Anal Biochem., 176:22-27.
McDonald B.J. & Moss, S.J. (1997) Neuropharm., 36:1377-1385.

We thank the Wellcome Trust for financial support.