The mechanism of thyrotropin-releasing hormone (TRH)-induced ether-a-go-go-related gene (erg) K+ current modulation was investigated with the perforated-patch whole-cell technique in clonal somatomammotroph GH3/B6 cells. These cells express a small endogenous erg current known to be reduced by TRH. GH3/B6 cells were injected with cDNA coding for rat erg1, erg2, erg3 and HERG K+ channels. The corresponding erg currents were isolated with the help of the specific erg channel blockers E-4031 and dofetilide and their biophysical properties were determined. TRH (1 M) was able to significantly reduce the different erg currents. The voltage dependence of activation was shifted by 15 mV (erg1), 10 mV (erg2) and 6 mV (erg3) to more positive potentials without strongly affecting erg inactivation. TRH reduced the maximal available erg current amplitude by 12% (erg1), 13% (erg2) and 39% (erg3) and accelerated the time course of erg1 and erg2 channel deactivation, whereas erg3 deactivation kinetics were not significantly altered. The effects of TRH on HERG currents did not differ from those on its rat homologue erg1. In addition, coinjection of rat MiRP1 with HERG cDNA did not influence the TRH-induced modulation of HERG channels. Rat erg1 currents recorded in the cell-attached configuration were reduced by application of TRH to the extra-patch membrane in the majority of the experiments, confirming the involvement of a diffusible second messenger. Application of the phorbol ester phorbol 12-myristate 13-acetate (PMA; 1 M) shifted the voltage dependence of erg1 activation in the depolarizing direction, but it did not reduce the maximal current amplitude. The voltage shift could not be explained by a selective effect on protein kinase C (PKC) since the PKC inhibitor bisindolylmaleimide I did not block the effects of TRH and PMA on erg1. In addition, cholecystokinin, known to activate the phosphoinositol pathway similarly to TRH, did not significantly affect the erg1 current. Various agents interfering with different known TRH-elicited cellular responses were not able to completely mimic or inhibit the TRH effects on erg1. Tested substances included modulators of the cAMP-protein kinase A pathway, arachidonic acid, inhibitors of tyrosine kinase and mitogen-activated protein kinase, sodium nitroprusside and cytochalasin D. The results demonstrate that all three members of the erg channel subfamily are modulated by TRH in GH3/B6 cells. In agreement with previous studies on the TRH-induced modulation of the endogenous erg current in prolactin-secreting anterior pituitary cells, the TRH effects on overexpressed erg1 channels are not mediated by any of the tested signalling pathways.