We investigated the interaction of the 12 kDa FK506-binding protein (FKBP12) with two ryanodine-receptor isoforms (RyR1 and RyR3) and with two myo-inositol 1,4,5-trisphosphate (IP3) receptor isoforms (IP3R1 and IP3R3). Using glutathione S-transferase (GST)-FKBP12 affinity chromatography, we could efficiently extract RyR1 (42+/-7% of the solubilized RyR1) from terminal cisternae of skeletal muscle as well as RyR3 (32+/-4% of the solubilized RyR3) from RyR3-overexpressing HEK-293 cells. These interactions were completely abolished by FK506 (20 microM) but were largely unaffected by RyR-channel modulators. In contrast, neither IP3R1 nor IP3R3 from various sources, including rabbit cerebellum, A7r5 smooth-muscle cells and IP3R-overexpressing Sf9 insect cells from Spodoptera frugiperda, were retained on the GST-FKBP12 matrix. Moreover, immunoprecipitation experiments indicated a high-affinity interaction of FKBP12 with RyR1 but not with IP3R1. In order to determine the FKBP12-binding site, we fragmented both RyR1 and IP33R1 by limited proteolysis. We obtained a 45 kDa fragment of RyR1 that bound to the GST-FKBP12 matrix, indicating that it retained all requirements for FKBP12 binding. This fragment was identified by its interaction with antibody m34C and must therefore contain its epitope (amino acids 2756-2803). However, no fragment of IP3R1 was retained on the column. These molecular data are in agreement with the lack of correlation between FKBP12 and IP3R1 expression in various cell types. The observation that FKBP12 did not affect IP3-induced Ca2+ release but reduced caffeine-induced Ca2+ release also indicated that mature IP3R1 and IP3R3, in contrast to RyR1 and RyR3, did not display a specific, high-affinity interaction with FKBP12.