Abstract
We previously reported the isolation of a rat cDNA clone encoding a protein with significant sequence homology to the B isoform of human myo-inositol 1,4,5-trisphosphate 3-kinase (IP3 3-kinase B); this protein was thus designated rat IP3 3-kinase B [Thomas, Brake, Luzio, Stanley and Banting (1994) Biochim. Biophys. Acta 1220, 219-222]. However, no IP3 kinase isoform had been shown to generate the physiologically important isoform of inositol tetrakisphosphate, i.e. inositol 1,3,4,5-tetrakisphosphate. We now present direct evidence that the putative rat IP3 3-kinase B is genuinely an IP3 3-kinase. We also show that the enzyme exists both as a peripheral membrane protein tightly associated with the cytosolic face of the extended endoplasmic reticulum network, and as a cytosolic protein. Association of the IP3 3-kinase with membranes is not affected by treatment with brefeldin A, Na2CO3 (pH 11.5), 2 M NaCl, or alteration of [Ca2+]. However, treatment of isolated membranes with 4 M urea leads to dissociation of the kinase from the membrane, implying that membrane association involves specific, conformation-dependent protein-protein interactions. The fact that IP3 3-kinase B is localized exclusively to membranes of Ca2+ stores, is consistent with a model where this kinase plays a role in IP3-dependent Ca2+ release.