Cloning, pharmacological characterisation and tissue distribution of a novel 5-HT4 receptor splice variant, 5-HT4(i)

5-HT4 receptor pre-mRNA is alternatively spliced in human (h) tissue to produce several splice variants, called 5-HT4(a) to 5-HT4(h) and 5-HT4(n). Polymerase chain reaction (PCR) with primers designed to amplify both 5-HT4(a) and 5-HT4(b) amplified three additional bands in different tissues, two representing different mRNA species both encoding 5-HT4(g) and one representing mRNA for a novel splice variant named 5-HT4(i), cloned from testis and pancreas respectively. Primary and nested PCR detected both 5-HT4(g) and 5-HT4(i) in multiple tissues. Whereas 5-HT4(i), was found in all cardiovascular tissues analysed, 5-HT4(g) was mainly present in atria. However, quantitative RT-PCR indicated 5-HT4(g) expression also in cardiac ventricle. The pharmacological profiles and ability to activate adenylyl cyclase (AC) were compared between four recombinant h5-HT4 splice variants (a, b, g and i) expressed transiently and stably in HEK293 cells. Displacement of [H-3]GR113808 with ten ligands revealed identical pharmacological profiles (affinity rank order: GR125487, SB207710, GR113808>SB203186>serotonin, cisapride, tropisetron>renzapride, 5-MeOT>5-CT). In transiently transfected HEK293 cells cisapride was a partial agonist compared to serotonin at 5-HT4(b), 5-HT4(g) and 5-HT4(i) receptors. In membranes from HEK293 cells stably expressing 5-HT4((g)) (3,000 fmol/mg protein) or 5-HT4(i) (500 fmol/mg protein), serotonin and 5-MeOT were full agonists while cisapride was full agonist at 5-HT4(g) and partial agonist at 5-HT4(i), probably due to different receptor expression levels. At both 5-HT4(g) and 5-HT4(i), the behaviour of 5-HT4 receptor antagonists was dependent on receptor level. At high receptor levels, tropisetron and SB207710 and to a variable extent SB203186 and GR113808 displayed some partial agonist activity, whereas GR125487 and SB207266 reduced the AC activity below basal, indicating both receptors to be constitutively active. We conclude that the novel 5-HT4(i) receptor splice variant is phannacologically indistinguishable from other 5-HT4 splice variants and that the 5-HT4(i) C-terminal tail does not influence coupling to AC.