MODULATION OF THYROID-HORMONE ACTION BY MUTANT THYROID-HORMONE RECEPTORS, C-ERBA-ALPHA-2 AND PEROXISOME PROLIFERATOR-ACTIVATED RECEPTOR - EVIDENCE FOR DIFFERENT MECHANISMS OF INHIBITION

Thyroid hormone action is not only determined by hormone availability, but also by target organ sensitivity. A dominant negative interaction is known to occur between thyroid hormone receptors (TRs) and the non -ligand binding splicing variant c-erbA alpha 2 as well as mutant TR b eta 1 from kindreds with resistance to thyroid hormone. We compared th e inhibitory effect of naturally occurring mutant hTR beta 1, artifici ally created hTR alpha 1 mutants, c-erbA alpha 2 and the human peroxis ome proliferator-activated receptor (hPPAR) on three prototypic T3-res ponse elements (TREs), TRE-PAL, DR+4 and TRE-LAP. The inhibitory effec t of mutant hTR alpha 1 and beta 1 occurred only on TRE-LAP and to a m inor degree on DR+4 when equimolar ratios of mutant/wildtype receptor were present. In contrast, the c-erbA alpha 2 splicing variant and the hPPAR inhibited TR action on all three TREs. Gel mobility shift exper iments in the presence of T3 showed increased binding of mutant hTR al pha 1 and beta 1 only to TRE-LAP compared to the binding of wildtype h TRs, thereby explaining their TRE-selective dominant negative potency. Contrarily, equal amounts of c-erbA alpha 2 or hPPAR protein did not bind to either of the three response elements even in the presence of RXR. Since the TR:RXR heterodimers were only partially displaced from DNA in the presence of excess amounts of c-erbA alpha 2, it is likely that the TRE-unspecific dominant negative action of c-erbA alpha 2 is due in part to competition for DNA-binding and for TR-auxiliary protei ns. In contrast, equimolar amounts of hPPAR completely inhibited the D NA-binding of hTR beta 1:RXR heterodimers, but not of TR:TR homodimers , suggesting that hPPAR has a higher RXR-binding affinity and is there fore a potent competitor for intranuclear RXR. Since thyroid hormones and peroxisome proliferators regulate in part a similar subset of targ et genes involved in fatty acid metabolism, these results suggest the possibility of cross-talk among the thyroid hormone and peroxisome pro liferator signalling pathways. In summary, the results suggest that th yroid hormone action can be modulated by at least three different mech anisms: (i) increased binding of mutant hTRs to specific TREs; (ii) ef ficient competition for limiting amounts of RXR through the preferenti al formation of hPPAR:RXR, rather than TR:RXR heterodimers; and (iii) competition for binding to DNA and to auxiliary proteins other than RX R in the case of c-erbA alpha 2.