Single-channel studies on human TREK-1 (hTREK-1) channels to intracellular ischemia related factors

Abstract

TREK-1, a member of the two-pore domain family of potassium channels, majorly contributes to the maintenance of resting membrane potential of a cell and has been reported to respond to ischemic levels of intracellular lactate and acidic pH to provide neuroprotection. There are two N-terminal variants that arise due to Alternative splicing: the shorter variant having a shorter N-terminus than the full-length human TREK-1 (hTREK-1) which is widely expressed in the acute hypoxia sensitive regions of the adult brain like the cerebellum and hippocampus and is upregulated under ischemia. Previous whole-cell patch-clamp experiments on the shorter variant of hTREK1 have shown contradictory results to hypoxia- a condition attributed to ischemia, which has put the neuroprotective role of the hTREK-1 channel into question. Although these experiments were performed on the shorter hTREK-1, there have been no studies on the effect of hypoxia and other ischemia-related factors like lactate, pH, and hemin on the full-length hTREK-1 channel. In the present study, using single-channel inside-out patch-clamp electrophysiology on the full-length hTREK-1 expressed in HEK293 cells, we show that the extended N-terminus of the full-length hTREK-1 channel is required to sense hypoxia. The probability of opening of the full-length hTREK-1 channel reversibly increased on exposure to hypoxia. However, there was a decrease in the open probability of the shorter hTREK-1 channel under similar conditions suggesting that the N-terminus might be responsive to hypoxia or it might interact with the Cterminus of the protein or a sensor situated elsewhere in the channel. Due to the shift in glucose metabolism from aerobic to anaerobic mode upon ischemia, there is an increase in intracellular lactate that has been shown to be a potent modulator of the fulllength hTREK-1 channel. However, the modulatory effect of lactate and hypoxia on the fulllength hTREK-1 has not been reported. We observed a significant increase in the open probability when the channel was first exposed to hypoxia, followed by hypoxia and 20mM lactate together. Interestingly, the extent of increase was reproducible when the channel was exposed to 20mM lactate first, followed by 20mM lactate and hypoxia together. This observation suggests that the sites of action of hypoxia and lactate are independent of each other and additive when presented together. It is known from previous literature that a decrease in intracellular pH due to ischemia leads to activation of the TREK-1 channel. However, the effect of hypoxia and lactate under ischemic pH conditions on the full-length hTREK1 remained elusive. We showed that the channel’s open probability increases with hypoxia under acidic pH 6, which gets further elevated with the addition of 20mM lactate in the medium. Cerebral ischemia is associated with several pathological microenvironmental changes like membrane distortion and changes in intracellular pH. A critical amino acid residue, E306 (E321 in the full-length hTREK-1) in the intracellular C-terminus of the channel has been shown to be involved in mechano-gating of the channel by associating with the inner leaflet of the plasma membrane during intracellular acidosis. Since TREK-1 is mechano-gated and sensitive to intracellular acidosis, we hypothesized that the intactness of the physiological state of the channel by the C-terminus is essential for the hypoxic response in ischemic conditions. Single-channel inside-out recordings from the E321A mutant with a disrupted interaction of the C-terminus with the lipid membrane that traps it in a constitutively open state showed a significant decrease in activity to hypoxia. The importance of the C-terminus that purportedly hangs in the cytosol and the extended N-terminus of the full-length hTREK1 in eliciting the response to hypoxia indicated possible N-C terminus interactions that have been shown in other ion channels. While the non-heme-based oxygen sensing was clear from these isolated insideout single ion channel recordings, it must be pointed out that hypoxia has been previously shown to dimerize HIF (hypoxia-inducible factor) alpha and beta chains through proline residues that are lost under normoxic conditions. Since heme-based oxygen sensing is the common biological mechanism to sense oxygen, I explored the plausible existence of such a mechanism in the full-length hTREK-1 channel. Single-channel recordings in inside-out patches in the presence of bath perfused hemin- a known oxygen sensor and a channel modulator were studied. Bath perfusion of 500nM hemin followed by 500nM hemin and hypoxia together failed to significantly increase the channel activity, although the same channel responded to hypoxia alone by increasing the channel activity. Although beyond the scope of the present thesis work, the non-heme-based hypoxia sensing through N-C terminus interaction in the full-length as a plausible mechanism needs further investigation. A bioinformatics search for a hemin-binding motif that has been indicated in other ion channels was not seen in the full-length hTREK1, indicating that a heme-based oxygen sensing mechanism might be absent in the hTREK1. Finally, the key point that emerges from the study is the polymodal regulation of the hTREK1 where the channel appears to integrate the responses to the ischemia-related factors hypoxia, intracellular lactate, and pH.