Understanding the role of voltage gating of polymodal TRP channels

Abstract

Being able to sense multiple stimuli is a common feature of many transient receptor potential (TRP) channels and is in line with their proposed role in monitoring environment and body conditions. TRPV1 is well known for this polymodal regulation, being activated by noxious heat, acid, capsaicin and a number of cellular mechanisms. Among them, voltage gating has attracted much attention among channel biophysicists and physiologists alike. However, understanding the role of voltage on the overall gating mechanism of TRP channels has stirred some controversy. Voltage-dependent regulation was first described for TRPM4 (Nilius et al. 2003) and later for TRPV1 and TRPM8 with more detailed mechanistic examinations (Voets et al. 2004). Voltage dependence can be fitted using the Boltzmann function: where G is conductance, Gmax is the maximal conductance of the given channel in a given cell, V1/2 is the half-maximal activation potential, z is the valence of the gating charge, F and R are Faraday and gas constants, respectively. At room temperature, both TRPV1 and TRPM8 showed low z values between 0.7 and 0.8e and high V1/2 values of > 90 mV, indicative of shallow voltage dependence and very low open probability at physiological membrane potentials, respectively. Heating and capsaicin induced a parallel shift in the voltage dependence of TRPV1 to more negative values, allowing channel activation at physiological potentials. The opposite is true for cold-activated TRPM8. Heating caused a positive shift while cooling resulted in a negative shift of the voltage dependence. The TRPM8 agonist, menthol, also shifted the voltage dependence to more negative potentials. The opposite effects of temperature on voltage gating between TRPV1 and TRPM8 may seem puzzling. However, this was near perfectly explained by a two-state model, which revealed that TRPV1 had a strong temperature dependence on activation but not deactivation, while TRPM8 had a strong temperature dependence on deactivation but not activation. A key finding derived from this model, which has been under-appreciated by the field, is that because of the small gating charge, a large shift to physiologically relevant voltages is achievable with only a small change in the Gibbs free energy (ΔΔG = zFΔV1/2). Many physiological stimuli could take advantage of this unique property to activate TRPs. Temperature-dependent free energy change is explained by the relationship ΔG = ΔH − TΔS, where ΔG, ΔH, and ΔS are differences in free energy, enthalpy and entropy, respectively, between closed and open states. Ligand-induced free energy change is explained by the change in ΔH between ligand-free and ligand-bound conditions (Voets et al. 2007). An ‘allosteric’ model was proposed as an alternative by Brauchi et al. (2004). The key observation that led to this model is that, for TRPM8, not only was the activation voltage decreased but also Gmax was increased by lowering temperature. The latter effect suggests the existence of a temperature-regulated voltage-independent gating mechanism. Thus, the allosteric model proposes that temperature and voltage are capable of gating TRPM8 independently but they also influence each other in an allosteric fashion. The fundamental question that makes the allosteric model different from the two-state model is whether all effects seen with temperature are due to modulation of voltage gating, i.e. whether voltage is the final gate. In this issue of The Journal of Physiology, Matta & Ahern (2007) present data that challenge the idea of final gating by voltage. The authors studied the voltage-dependent properties of TRPV1 under stimulation by heat, protons, capsaicin and phosphorylation by protein kinase C. All known stimuli of TRPV1 appeared to enhance Gmax at the same time as they negatively shifted V1/2. Interestingly, the TRPV1 antagonist, capsazepine, caused opposite changes. More importantly, at certain high agonist concentrations, the conductance–voltage relationship became flat and did not approach zero at very negative potentials, suggestive of a voltage independent component activated by ligand. This component was estimated by adding another fit parameter, Gmin, in the Boltzmann function. Similar results were obtained for TRPM8. Because of the apparent voltage-independent activation by temperature and ligands, the authors also simulated their data using the allosteric model, with additional parameters added for the ligand-induced transition. Therefore, the polymodal TRP channels seem to be equipped with multiple sensors that work independently but also act in concert to gate the channel. However, it may be too early to claim trophy for the allosteric model. The simplicity and the ability to predict voltage- and time-dependent kinetics still put the two-state model in a superior position as long as all activity is considered voltage dependent. The allosteric model only predicts steady-state activity, and the many fit parameters raise questions as to whether they are all strictly independent. With some approximation, the allosteric model could be converted to the two-state model (Voets et al. 2007). What this all boils down to is whether temperature and/or ligands really alter Gmax and Gmin, a question that challenges the confidence of investigators on the measured conductance values. This can be a serious issue depending on how experiments are performed. Whether steady-state current is reached, whether there is desensitization or sensitization during the series of steps, possible voltage clamp errors, background currents, compensation, and changes in single channel conductance due to temperature or ligand can all affect the accuracy of the results. Given that not all TRP channels are voltage-gated, and some acquire voltage gating through pore blocking by divalent cations while other possess intrinsic voltage sensors within the channel protein, the mechanism and functional significance of voltage gating of TRP channels will continue to spawn further investigation and heated discussions.