Calmodulin (CaM) regulation of CaV channels has long fascinated biophysicists, with structure-function analysis mostly focused on an IQ domain in the carboxy-terminus of channels. It is clear that Ca2+-free CaM (apoCaM) preassociates with the IQ domain before Ca2+ entry through channels (see Liu X et al this meeting), and that Ca2+/CaM has the potential to bind the IQ. Hence, Ca2+-dependent inactivation (CDI) of channels has been thought to result by transducing Ca2+-dependent conformational changes of CaM, all while bound to the IQ. By contrast, functional analysis of our structure of Ca2+/CaM complexed with the IQ domain of CaV2.1 (Structure 16:607) hints that Ca2+/CaM may depart from the IQ domain during channel regulation. To generalize this hypothesis, we here alanine scanned the entire IQ domain of CaV1.3, an exemplary prototype of CaV1 channels with high homology to long-studied CaV1.2. Importantly, alanine substitution likely preserves backbone fold, whereas prior studies often used more disruptive mutations targeting specific loci. As before, alanine substitution of the signature isoleucine strongly suppressed CDI. Suprisingly, however, CDI was strongly diminished only at one other residue, upstream of the central isoleucine. Altering many other sites, presumed important for Ca2+/CaM-IQ binding in crystal structures, left CDI unscathed. Moreover, we homology modeled Ca2+/CaM bound to the CaV1.3 IQ domain, based on x-ray structures of CaV1.2. We then computed (Robetta) the energetic cost of alanine mutations (ΔΔG for binding). If Ca2+/CaM-IQ binding begets CDI, plots of CDI versus ΔΔG should define a Boltzmann function. Instead, a highly scattered relationship was produced. Thus, CDI may involve (partial) departure of Ca2+/CaM from the IQ domain of CaV1.3, to interact with alternative sites (Yang, Ben Johny, & Yue, this meeting). This departure would fundamentally transform our understanding of CaM/channel regulation.
Read full abstract