UC Davis

Ca2+-dependent regulation of retinal guanylyl cyclase (RetGC), controlled by the sensor proteins GCAP5 and RD3, is important for promoting the visual recovery phase of phototransduction in retinal photoreceptor cells. Particular point mutations in either RetGC, GCAPs or RD3 each disrupt the Ca2+-dependent cyclase activity and are genetically linked to various retinal degenerative diseases, including autosomal cone dystrophy and Leber’s congenital amaurosis. In this thesis, I will present atomic-level structural analyses of both GCAP5 and RD3. Before solving the structure of RD3, I first developed an elaborate procedure to prepare enough RD3 protein required for NMR that involved refolding functional protein from inclusion bodies. The NMR structure of the RD3 protein (described in Chapter 2) reveals a novel fold with a non-canonical 4-helix bundle, which identified key amino acid residues on the RD3 surface that make intermolecular contacts with RetGC. Mutation of these hot spot residues weaken RD3 binding to RetGC. This work on the RD3 protein was published in J. Biol. Chem (2019) 294:2318. In a separate project described in Chapter 3, I used both NMR and EPR-DEER to solve the dimeric structure of GCAP5, which provides insights into the Ca2+/Fe2+-dependent conformational changes in GCAP5 that control the activation of RetGC. GCAP5 contains two non-conserved cysteine residues (Cys15 and Cys17) that are essential for the binding of Fe2+ to GCAP5 that are not observed in GCAP1. In vivo functional studies show that Fe2+ binding to GCAP5 inhibits RetGC activity and GCAP5 is suggested to be a redox sensor in visual phototransduction. Binding and mutagenesis studies reveal that GCAP5 forms a dimer, which binds to a single Fe2+. In essence, a total of four sulfhydryl groups (from Cys15 and Cys17) from each molecule of the GCAP5 dimer chelate the bound Fe2+ with a tetrahedral geometry, and the bound Fe2+ bridges two molecules of GCAP5 in a 1:2 complex. I solved the NMR structure of the Fe2+-free activator state of GCAP5, which is like the structure of GCAP1 except for differences in the N-terminal region that binds to Fe2+ in GCAP5. The N-terminal helix in GCAP5 is one turn longer to stabilize the exposure of Cys15 and Cys17, which are both essential for Fe2+ binding in GCAP5. I used EPR-DEER to measure intermolecular distances in the Fe2+-free GCAP5 dimer that were used to elucidate the dimeric structure. The GCAP5 dimer interface is comprised of mostly hydrophobic residues (H18, Y21, M25, F72, V76 and W93) that are conserved in GCAP1. In addition, the GCAP5 dimer contains an intermolecular salt bridge (between R22 and D71) that is not conserved in GCAP1. The double mutation H18E/Y21E and single mutations (R22D and M25E) both disrupt GCAP5 dimerization and the corresponding mutations in GCAP1 each abolish cyclase activation. Also, a mutation in GCAP1 (H19E) was found in a human patient that has autosomal dominant cone dystrophy. I conclude that Fe2+ binding and dimerization of GCAP5 is critical for the regulation of RetGC and the residues identified at the dimer interface (H18, Y21, M25, F72, V76 and W93) are important targets for future drug design.