Regulation of Bestrophin Cl Channels by Calcium: Role of the C Terminus

Human bestrophin-1 (hBest1), which is genetically linked to several kinds of retinopathy and macular degeneration in both humans and dogs, is the founding member of a family of Cl⁻ ion channels that are activated by intracellular Ca²⁺. At present, the structures and mechanisms responsible for Ca²⁺ sensing remain unknown. Here, we have used a combination of molecular modeling, density functional–binding energy calculations, mutagenesis, and patch clamp to identify the regions of hBest1 involved in Ca²⁺ sensing. We identified a cluster of a five contiguous acidic amino acids in the C terminus immediately after the last transmembrane domain, followed by an EF hand and another regulatory domain that are essential for Ca²⁺ sensing by hBest1. The cluster of five amino acids (293–308) is crucial for normal channel gating by Ca²⁺ because all but two of the 35 mutations we made in this region rendered the channel incapable of being activated by Ca²⁺. Using homology models built on the crystal structure of calmodulin (CaM), an EF hand (EF1) was identified in hBest1. EF1 was predicted to bind Ca²⁺ with a slightly higher affinity than the third EF hand of CaM and lower affinity than the second EF hand of troponin C. As predicted by the model, the D312G mutation in the putative Ca²⁺-binding loop (312–323) reduced the apparent Ca²⁺ affinity by 20-fold. In addition, the D312G and D323N mutations abolished Ca²⁺-dependent rundown of the current. Furthermore, analysis of truncation mutants of hBest1 identified a domain adjacent to EF1 that is rich in acidic amino acids (350–390) that is required for Ca²⁺ activation and plays a role in current rundown. These experiments identify a region of hBest1 (312–323) that is involved in the gating of hBest1 by Ca²⁺ and suggest a model in which Ca²⁺ binding to EF1 activates the channel in a process that requires the acidic domain (293–308) and another regulatory domain (350–390). Many of the ~100 disease-causing mutations in hBest1 are located in this region that we have implicated in Ca²⁺ sensing, suggesting that these mutations disrupt hBest1 channel gating by Ca²⁺.