STRUCTURAL STUDIES ON THE REGULATORY DOMAIN OF THREE HCN (HYPERPOLARIZATION-ACTIVATED CYCLIC NUCLEOTIDE-GATED) CHANNEL ISOFORMS

Hyperpolarization-activated cyclic nucleotide gated (HCN) chan- nels underlie the If /Ih cation currents that control pacemaker activity in heart and brain. HCN channels are dually ac- tivated by membrane hyperpolarization and binding of cAMP to their cyclic nucleotide binding domain (CNBD). Binding of cAMP shifts the activation curve of HCN2 and HCN4 by 17 mV, but that of HCN1 by only 2-4 mV. Tetramerization of the CNBD is seemingly part of the cAMP-induced allosteric con- formational changes that increase the open probability of the channel pore. We have obtained the crystal structures of the CNBD of the three isoforms, but the analysis revealed a very conserved structure between HCN1 versus HCN2 and HCN4, except for a loop of β-roll, previously shown to regulate the binding affinity of HCN4. We measured the binding affin- ity of the CNBD for the cAMP and the different propensity of the regulatory domain to tetramerize in absence or presence of the ligand. We confirm that tetramerization is the primary ef- fect of cAMP binding, and the first step in the transmission of this signal, that eventually removes the inhibition imposed by the CNBD on the channel. Accordingly, cAMP- binding releases HCN2 and HCN4 from inhibition, but has little or no effect on HCN1. Our data demonstrate that in HCN1 the CNBD is already tetrameric at basal cAMP concentrations contrary to HCN2 and HCN4. HCN1 shows this peculiar behavior despite its cAMP- binding affinity is in the same range of the affinity found in HCN2 and HCN4. This can be explained by two different 3 affinity states (high and low). HCN1 is, at low cAMP concen- trations, already switched to the low affinity conformation, while the high affinity state is not measurable because the binding site is already occupied. Our results offer a logical explanation for the behavior of HCN1 and an experimental support to the leading hypothesis that ligand-induced tetramerization removes tonic inhibition from the pore. In addition, some more inter- esting information arose from the crystal structure, highlighting an additional electron density close to the tetramerization inter- face of the proteins. We investigated a range of molecules that could bind the proteins in that pocket and potentially alter the functionality of the channel.