Résumé Ion channels generate the electrical signals with which the nervous system uses to sense the world, process information, create memories and control behavior. K2P channels serve as a hub for the generation and regulation of the negative resting membrane potential and neuronal excitability. They display low basal activity but can be further activated by various stimuli and are notably involved in mood regulation, pain perception, PUFA-dependent neuroprotection against ischemia and migraine. These properties suggest that K2Ps are attractive pharmacological targets and provide motivation for developing a deeper understanding of their function and organization within the plasma membrane. The absence of specific pharmacology and biochemical properties of K2P have made it difficult to determine their functions and subunit composition. To overcome these problems, we first designed a light-gated dominant-negative form of K2P2.1 (TREK1, TREK1-PCS) which allows natively expressed channels to be targeted and reversibly regulated by light. Using this tool, we showed that TREK1 typically considered to be a leak channel, contributes significantly to the hippocampal GABAB current which is involved in sIPSP and how phospholipids act specifically on TREK channels. Second, to address K2P channel composition, we developed and used a Single Molecule fluorescence Pulldown (SiMPull) assay, enabling the direct visualization of individual protein complexes to identify associated subunits and determine their stoichiometry. We found that heteromerization is not rare between K2P channel members and a link between heteromerization, migraine induction and a new genetically mechanism. We demonstrate that migraine-associated frameshift mutations of TRESK (K2P18.1) lead to the production, in addition to the expected nonfunctional Cterminal truncated TRESK, of a second protein fragment TRESK-MT2 due to a mechanism called frameshift mutation-induced Alternative Translation Initiation (fsATI). We show that by co-assembling with and inhibiting TREK1 and TREK2, another subfamily of K2P channels, TRESK-MT2 increases trigeminal sensory neuron excitability, leading to a migraine-like phenotype. These finding identify TREK1 and TREK2 as potential targets in migraine and suggest that fsATI should be considered as a distinct class of mutations (Royal et al 2018, Neuron 101-232:245 ; http://www.cnrs.fr/fr/decouverte-de-nouveaux-mecanismes-lorigine-de-la-migraine).