Adenine-nucleotide binding sites on the inositol 1,4,5-trisphosphate receptor bind caffeine, but not adenophostin A or cyclic ADP-ribose

Binding of ATP to the inositol 1,4,5-trisphosphate receptor (IP3R) results in a more pronounced Ca2+ release in the presence of inositol 1,4,5-trispho sphate (IP3). We have expressed the cDNAs encoding two putative adenine-nuc leotide binding sites of the neuronal form of IP3R-1 as glutathione S-trans ferase (GST)-fusion proteins in bacteria. Specific [alpha-P-32]ATP binding was observed for the two GST-fusion proteins, representing aa 1710-1850 and aa 1944-2040 of IP3R-1. The ATP-binding sites in both fusion proteins had the same nucleotide specificity as found for the intact IP3R (ATP > ADP > A MP > GTP). Smaller GST-fusion proteins (aa 1745-1792 and aa 2005-2023) disp layed a much weaker ATP-binding activity. CoA, which also potentiated IP3-i nduced Ca2+ release in A7r5 cells, interacted with the ATP-binding sites on the fusion proteins. Such interaction was not observed for 1,N-6-etheno Co A and 9'-dephospho-CoA, which are much less effective in potentiating IP3-i nduced Ca2+ release. Since the adenine-containing compounds adenophostin A, caffeine and cyclic ADP-ribose modulate IP3-induced Ca2+ release, a possib le effect of these compounds on the ATP-binding sites was examined. ATP sti mulated adenophostin A- and IP3-induced Ca2+ release in A7r5 cells with an EC50 of respectively 21 and 20 mu M. Also the threshold concentration of AT P for stimulating the release was similar for the two agonists. Adenophosti n A (100 mu M) and cyclic ADP-ribose (100 mu M) were ineffective in displac ing [alpha-P-32]ATP from the binding sites of both GST-fusion proteins. Caf feine (50 mM), however, inhibited [alpha-P-32]ATP binding to both fusion pr oteins by more than 50%. These data provide evidence for a direct interacti on of caffeine but not of adenophostin A or cyclic ADP-ribose with the aden ine-nucleotide binding sites of the IP3R.