MODULATION OF PORE GATING BY ¿SENSOR¿ DOMAINS IN VOLTAGE-GATED K+ CHANNELS.

Many proteins in nature show a modular topology, because it is possible to recognize functional modules responsible of distinguishable functions in the protein. Voltage-gated can be considered modular proteins. The superfamily of voltage-gated channels is composed of channels in which a pore-module is in charge of generating an ion conductance across the cell membrane, and other “sensor”-modules perceive different stimuli and transmit them to the pore, readjusting the conductance in response to changes within the cell. The sensor module in voltage-gated channels is the voltage sensing domain. It is composed of four transmembrane segments and it is able to feel the electrical properties of the membrane, such as changes in potential and, through a mechanical load applied by a short linker, to affect the gating of the pore (i.e. opening or closure of the channel). A more sophisticated regulation is possible thanks to other modules fused to the same channel. Ligand-gated channels usually exhibit a C-terminal domain exposed in the cytoplasm, in contact with all the variable concentration of second messengers, able to modulate the activity of the channel. In this group, cyclic nucleotide-gated channels have a C-terminal domain, composed of a binding domain (CNBD) that respond to difference in concentration of cAMP or cGMP, and a C-linker region, connecting the CNBD to the pore. The CNBD acts as an allosteric domain, and modulate the channel opening upon cAMP binding.The idea that these domains evolved independently before fusing in a single protein, is strengthened by the fact that similar domains are found in a large variety of proteins, that don’t belong to channels family. For example, recently, a new enzyme was discovered, the voltage-sensing phosphatase (VSP) of Ciona intestinalis, whose voltage-sensor domain is fused to a phosphatase. The CNBD of the hyperpolarized cyclic nucleotide-activated channels (HCN), has a conserved structure compared to that of the cAMP-dependent protein kinases. My PhD thesis addresses two different topics, but in both cases I investigated how sensor-modules can give a sophisticated regulation of channel gating. In the first part, I approached the problem about how voltage-dependence originated in voltage-gated channels, in particular I obtained a voltage-gated K + channel fusing two unrelated protein modules: the voltage sensing domain of Ci-VSP and the pore-channel PBCV-1 Kcv. The fusion between a voltage sensor and a potassium channel with a quasi-ohmic behaviour generates a chimaeric protein called KvSynth1, an delayed outward rectifier potassium channel. KvSynth1 retains the pore properties of Kcv (selectivity and filter gating) and the voltage dependence of the Ci-VSP (half activation potential, slope factor, shift of activation curve due to mutations). Moreover, the quality of the rectification is dependent on the length of the linker between the two modules. This highlights a mechanic role of the linker in transmitting the movement of the sensor to the pore, and shows that electromechanical coupling can occur without co-evolution of the two domains. In the second part, the allosteric modulation of the cyclic nucleotide binding domain of HCN channels has been studied on the basis of our findings in the crystal structure of the CNBD of the isoform 4 of HCN (hyperpolarized cyclic nucleotide-activated) channels. HCN are the molecular determinant of the If current, responsible of the autonomic regulation of the heart. In the structure of HCN4 CNBD a putative binding site for cyclic nucleotides in the C-linker region was found. Occupancy of this binding site by the prokaryote second messenger c-di-GMP can completely revert the effect of cAMP in the micromolar range. Docking a large set of molecules in the binding pocket, another compound was identified, (N’-biphenyl-2-yl-N-[1-(3-cyanobenzyl)piperidin-4-yl]-N-(pyridin-3-ylmethyl)urea), able to give the same effect on cAMP modulation. The effect of these molecules is restricted to HCN4; this isoform selectivity underlies that, although the C-terminus of the three isoforms is structured in a similar way, the modulation can be different. Some different features of HCN1, HCN2 and HCN4 were already analysed previously. These results highlight the presence of an second modulatory pathway in HCN channels, indicate a potential drug binding site for heart rate modulation and advance understanding of the mechanism of efficacy of cAMP binding in HCN channels.