Introduction

TREK-2 (TWIK-related K+ channel-2, K2P10.1) belongs to the two-pore domain potassium (K2P) channel family, which conducts leak or background currents that are essential to diverse cellular functions, including maintenance of the resting potential and regulation of excitability, sensory transduction, ion transport, cell volume metabolism, and apoptosis [13, 20]. The activity of the channel is sensitive to various physiological and pathological factors, such as neurotransmitters, hormones, temperature, unsaturated free fatty acids, mechanical stretch, intra- or extracellular pH, and protein kinases [1, 13, 20]. In addition, TREK-2 is also regulated by volatile general anesthetics, such as diltiazem and mibefradil [20, 24]. Accordingly, by tuning cellular excitability, the function of the channel is associated with diverse physiological, pathological, or pharmacological processes. For example, recent studies implicate that TREK-2 in dorsal root ganglion neurons hyperpolarizes C-nociceptors and limits pathological spontaneous pain [1].

Structurally, K2P channels are unique in that each subunit possesses four transmembrane segments and two pore-forming domains. As a result, mature K2P channels tend to form dimers to constitute the fourfold pore apparatus. Recent characterization of the crystal structures of two members of K2P channels reveals that the elements of pore structure are akin to other types of potassium channels. Along the ion conduction pathway in K2P channels, there also arranges the inner gate (activation gate) and selectivity filter (SF) gate (also called C-type gate, outer gate or inactivation gate) from intracellular side to extracellular side. Importantly, the structure of SF, which is formed by signature sequence TV(I)GY(F)G, is absolutely conserved among all K+ channels [3, 23]. From extracellular side to intracellular side, the key carbonyl oxygens together with the side-chain hydroxyl oxygen of the threonines define four equally spaced K+ binding sites, which are commonly termed as S1∼S4 [2]. In the process of K+ efflux, the SF experiences reversely conformational transition to produce conductive and non-conductive status alternately and highly discriminates K+ over Na+ eventually. Obviously, the SF plays a key role in gating the pore of K+ channels, and thus, it is also referred as the SF gate [9, 10, 12, 13, 34, 35].

Alternative translation initiation (ATI) is an important cellular mechanism contributing to the diversity of protein functions. Accumulating evidence shows that the isoforms generated from ATI mechanism (ATI isoform) exhibits various functional properties [4, 21, 28, 30, 33]. TREK-1 is another member of TREK subclass, the nearest relative of TREK-2. It has been found that the TREK-1 ATI isoforms, which are different in the length of N-terminus (Nt), display a big diversity in both unitary conductance and K+ selectivity [33]. By contrast, such ATI isoforms of TREK-2 only exhibit different conductance, whereas similar selectivity to K+ [30]. According to previous study, the alteration of ion selectivity is associated with the SF conformational transition in TREK-1 [22], but whether such correlation also exists in TREK-2 is still unknown. Detailed sequence mapping shows that the intracellular Nt is involved in the regulation of unitary conductance level of TREK-2 channel [30]. However, whether the Nt affects the SF status of TREK-2 and the possible relationship between unitary conductance and SF conformation still remain to be explored. In this study, we report that Ba2+ inhibits TREK-2 currents by binding to the assumed K+ binding site 4 (S4). Then, using Ba2+ and extracellular pH (pHo) as SF gate probing tools, we reveal that all the ATI isoforms and wild-type (WT) TREK-2 display similar responses to the alterations of extracellular Ba2+ and pHo, suggesting that they adopt roughly similar SF conformation in S4 site and the position affected by pHo response. Accordingly, the resembled ion selectivity in ATI isoforms seems originated from their similar SF configuration, which seems support the idea that neither the SF conformation nor the role of Nt is involved in the regulation of the unitary conductance in TREK-2.

Materials and methods

Molecular biology

The cDNA encoding human TREK-2 (NM_138318.1) for this work was amplified from a TREK-2 plasmid (OriGene) by using PCR, and was cloned into pGH19 vector (a high-expression level vector in Xenopus laevis oocyte) using BglII(TaKaRa) and Xba I (TaKaRa), as described previously [22]. Mutants were produced using the MutanBEST kit (TaKaRa) according to the manufacturer’s manual. All mutations were confirmed by DNA sequencing. cRNA was transcribed in vitro using the RiboMAXTM large-scale RNA production systems kit (Promega).

Electrophysiology

X. laevis were purchased from Nasco (USA). Oocytes were surgically removed from adult females and treated with 0.5 mg/ml collagenase type II (Gibco, USA), defolliculated and incubated at 19 °C for 1∼3 days before use. The oocytes were injected with ∼5 ng cRNA (46 nl in volume) per cell. Whole-cell currents were measured 1∼3 days after injection by the two-electrode voltage clamp technique, using an Axoclamp2B amplifier (Axon Instruments, Union City, CA). For two-electrode voltage clamp experiments, single oocyte was placed in a 0.3-ml perfusion chamber and impaled with two standard microelectrodes (1–2.5 MΩ resistance), which were filled with 3 M KCl. Recordings were performed in standard physiological extracellular solution, which contained (in mM): KCl, 5; NaCl, 93; MgCl2, 1; CaCl2,1.8; HEPES, 5; and adjusted to pH 7.4, with NaOH. HEPES buffer was replaced by Tris in pH 8.5–9.3 solutions and MES in pH 6.5 solutions. When required, potassium ions in bath solutions were isotonically replaced by sodium ions. BaCl2 was diluted from 1 M stock and added to the various solutions as indicated. Recordings were performed under constant perfusion at room temperature. Oocyte membrane potential was held at −80 mV and currents were elicited by a ramp from −120 to +60 mV for 150 ms with a 2-s interpulse interval. Data were sampled at 2 kHz and filtered at 0.5 kHz with Clampex 10.2 software (Axon Instruments). The experiments were approved by the Animal Care and Ethics Committee of Southern Medical University and Beijing Institute of Pharmacology and Toxicology.

Data analysis

Concentration-response curves were fitted to the Hill equation according to the following parameters: I = 1/(1 + ([X]o/IC50)h), where I represents the recorded current, IC50 is the concentration of X (represents Ba2+ or H+ in this research) required for achieving half-maximal effects, and h is the Hill coefficient. Time constants were calculated by fitting the data to the following equation: I = I o + A × e − T/τ, where I is the current value, I o is the inhibited current at equilibrium, T is the elapsed time after Ba2+ application, A is a constant and τ is the apparent forward time constant in seconds. The fitting was separately performed for each experiment using Origin 8.0 software (OriginLab).

Statistical significance was assessed with Student’s t test using GraphPad Prism 5.0 (GraphPad Software, Inc.) software. A significant difference was considered when p < 0.05.

Results

Ba2+ inhibits TREK-2 currents in a concentration- and time-dependent manner

Our previous studies indicate that Ba2+ inhibits the activity of TREK-1 channels [22]. TREK-1 and TREK-2 belong to the same subclass of K2P family and share 78 % homology at protein level [18]. We thus suspect that TREK-2 channels might display a similar response with TREK-1 upon the application of Ba2+. To verify the hypothesis, we expressed TREK-2 channels in X. laevis oocytes, and the currents were elicited by voltage ramps from −120 to 60 mV. Indeed, as shown in Fig. 1a, on the application of increasing concentration of external Ba2+ ([Ba2+]o), the currents through TREK-2 channels were dramatically inhibited. At a test potential of 0 mV, the inhibition by the application of varied [Ba2+]o was fitted by Hill equation. The half-maximal inhibitory concentration (IC50) and Hill coefficient were estimated to be 0.84 ± 0.06 mM and 1.61 ± 0.12, respectively (Fig. 1b). The Hill coefficient was bigger than 1, implicating that the interaction between TREK-2 and Ba2+ is positively cooperative.

Fig. 1
figure 1

TREK-2 currents are inhibited by externally applied Ba2+ in a concentration- and time-dependent manner. a Representative current–voltage relationship for TREK-2 expressed in Xenopus oocytes in the absence and presence of different [Ba2+]o. The concentration of extracellular K+ was 5 mM. Currents were elicited by a voltage ramp from −120 to 60 mV, from a holding potential of −80 mV. b Dose response for inhibition of wild-type TREK-2 (WT) channels at 0 mV with the protocol in A. Data points were mean ± SE (n ≥ 6). The solid line represents the best fit of the data to the Hill equation. c The time course of binding and unbinding of 1 mM Ba2+ to TREK-2 channel from a representative TREK-2-expressing oocyte. d The time course of inhibition to TREK-2 channel by different [Ba2+]o from a representative TREK-2-expressing oocyte. e The time constant (τblock, s) of inhibition to TREK-2 current by different [Ba2+]o (mean ± SE, n ≥ 6). *p < 0.05 or **p < 0.01 versus 0.5 mM [Ba2+]o

Full size image

Moreover, Ba2+ blockade on TREK-2 was time dependent. At the presence of 1 mM Ba2+, the currents were decreased gradually as the time elapsed until the steady-state blockade was achieved. When washout was performed, the activity of TREK-2 was also gradually recovered (Fig. 1c). As [Ba2+]o was enhanced to 2 and 4 mM, the binding rate of the ions was accelerated (Fig. 1d). The time course of blocking was fitted with a mono exponential function, and the apparent forward time constant for Ba2+ inhibition (τblock) was estimated. When the [Ba2+]o was increased from 0.5 to 6 mM, the τblock decreased from 40.86 ± 3.82 to 13.77 ± 0.75 s (Fig. 1E, p < 0.01). However, we did not observe the correlation between exit rate of Ba2+ and its concentration (data not shown). Taken together, above data clearly demonstrate that Ba2+ inhibits the currents of TREK-2 channels in a concentration- and time-dependent manner.

External K+ blunts the blocking effect of Ba2+ on TREK-2

It has been established that Ba2+ has a similar radius as K+ and is able to block the activity of several types of potassium channels through fitting into the selectivity filter (SF) [7, 14, 25]. If that is also the case in TREK-2, the effects of Ba2+ could be blunted by elevated concentration of extracellular K+ ([K+]o) [8, 22, 26, 27]. To address this hypothesis, the access degree of Ba2+ to the channel in different [K+]o was assessed. As the time-dependent curves illustrated in Fig. 2a, enhancement of [K+]o from 0 to 40 mM dramatically interfered Ba2+ binding to TREK-2. As a result, the ability of 1 mM [Ba2+]o to block TREK-2 was blunted from 96.26 ± 0.27 % at K+-free condition to 52.48 ± 3.61, 21.57 ± 1.99, and 10.66 ± 1.69 % at 5, 20, and 40 mM [K+]o, respectively. These results reveal that the binding of Ba2+ to TREK-2 channels is strongly affected by [K+]o, which implicates that the binding site of Ba2+ may reside within the SF of TREK-2.

Fig. 2
figure 2

Effects of different [K+]o on Ba2+ inhibition of TREK-2 currents. a Effects of different [K+]o on the time course of blockade by 1 mM Ba2+ in a representative oocyte expressing TREK-2 channels. b Average TREK-2 current decreased by Ba2+ (in %) in different [K+]o. Data represent mean ± SE (n ≥ 6). ***p < 0.01 versus 0 mM [K+]o

Full size image

Ba2+ blocks TREK-2 by binding at the S4 site within the selectivity filter

Variation of [K+]o is tightly coupled to conformational change at SF, i.e., K+-free extracellular solution evokes a non-conductive state of SF [35]. Since Ba2+ binding to TREK-2 is associated with such K+-dependent SF conformational transition and, especially, Ba2+ blocks TREK-1 by overlapped binding at the S4 site within SF [22], we next investigated whether the ions bind to the S4 site of TREK-2. The putative threonines (positions 172 and 281) comprising the S4 site of TREK-2 are strictly conserved compared with several types of K+ channels (Fig. 3a). To test our hypothesis, mutants containing the threonine-to-serine transition at positions 172 (T172S) and 281 (T281S) were constructed. As the time-dependent curves shown in Fig. 3b, the response of the mutants to 2 mM [Ba2+]o were drastically blunted. Accordingly, the IC50 for Ba2+ block to T172S and T281S was 5.32 ± 1.27 and 4.9 ± 0.65 mM, respectively (Fig. 3c). Obviously, the threonine mutations increased the IC50 for Ba2+ by approximately sixfolds, suggesting that Ba2+ blocks TREK-2 currents by occupying theS4 site within SF.

Fig. 3
figure 3

Ba2+ inhibits TREK-2 current through binding at S4 site of the selectivity filter. a Multiple sequences alignment of the SF regions of TREK-2 (P1 and P2), TREK-1(P1 and P2), KcsA, and Kv1.5. The positions of the conserved threonine residues among them and the SF GY/FG signature sequences were boxed. b Comparison of the relative 2 mM Ba2+ inhibitory effects on WT, the T172S mutant, and the T281S mutant. The currents recorded from representative oocytes expressing the indicated channels. c Dose–response curves of the WT, T172S, and T281S mutants. Data points were mean ± SE (n ≥ 6). The data were fitted with Hill equation

Full size image

The isoforms generated by alternative translation initiation (ATI isoform) mechanism adopts similar conformation in S4 site

Because of “leaky scanning” of ribosomes to initiation site [16], mRNA of TREK-2 channels generates three isoforms (Fig. 4a). These isoforms display largely divergent conductance, but exhibit similar selectivity for K+ [30]. Given that the SF configuration of K+ channels is directly associated with ion selectivity, we thus suspect that the ATI isoforms might have similar conformation in SF. As established above, Ba2+ inhibits TREK-2 activity through docking at S4 site, and thus, these alkaline metal ions could be used as specific tools to define the conformational alteration in S4 site. We then measured the Ba2+ response of ATI isoforms.

Fig. 4
figure 4

Response of TREK-2 isoforms generated by ATI mechanism to externally applied Ba2+. a The Nt sequence of the ATI isoforms (WT and M3) and the mutants that harboring the M-to-L transition of the second and/or third initiation sites (M2L3 and M1L2L3). The potential initiation sites located at positions 1, 60, and 72 were boxed. b Comparison of the inhibition percentage for the indicated channels in the presence of varying [Ba2+]o. Data points were mean ± S.E (n ≥ 6). c IC50 values of [Ba2+]o inhibition for TREK-2 isoforms and mutated channels. d The access rate (τblock, s) of Ba2+ to TREK-2 isoforms and mutants at different [Ba2+]o (mean ± SE, n ≥ 5). e The exit rate (τunblock, s) of Ba2+ to TREK-2 isoforms and mutants at different [Ba2+]o (mean ± SE, n ≥ 5)

Full size image

Mutants were constructed such that only one specific initiation site could be used, as described previously [30]. The wild-type human TREK-2 was referred as WT (also termed as M1M2M3). The M2L3 were constructed by deleting the segment before the second initiation site and mutating the third initiation site to leucine simultaneously. The Nt up to the third initiation site was deleted to produce M3. Both of the second and third methionines (M2 and M3) were mutated to leucines simultaneously to generate M1L2L3 (Fig. 4a). As shown in Fig. 4b, the concentration response curves for all the mutants displayed similar Ba2+ blocking profile with those of WT channels. Accordingly, their IC50s did not show significant change compared with WT channels, with a value of 0.95 ± 0.11, 0.91 ± 0.02, and 0.96 ± 0.05 mM for M2L3,M3 and M1L2L3, respectively (Fig. 4c, p > 0.05). Moreover, in the presence of different [Ba2+]o, no significant difference was found in the access and exit rate of Ba2+ (τunblock, calculated by fitting the data to mono exponential equation) between WT channels and ATI isoforms (Fig. 4d, e, p > 0.05). These resembled behaviors of ATI isoforms upon the application of Ba2+ demonstrate that they roughly adopt similar conformation in the S4 site. In addition, these results suggest that the Nt domain is not involved in the regulation of the conformational transition in S4 site.

The ATI isoforms experience similar conformational transitions induced by extracellular H+ within SF

TREK-1, the nearest relative of TREK-2 in K2P family, is inhibited by extracellular acidification via a mechanism resembling C-type inactivation [8]. Both crystal structural analysis of KcsA (potassium channels from S. lividans) [11] and functional characterization of other K+ channels [12, 22, 31, 37, 38] indicate that structural collapse (or barrier formation) occurs at the outer of SF during C-type inactivation. TREK-2 is also inhibited by alkaline pHo, and these effects are blunted by high [K+]o [29], which implicates that pHo modulates TREK-2 via manipulating SF. Despite that the His151 has been identified as a pHo sensor [29], the relative location of SF involved in the regulation (pHo position) is still unknown. Thus, to define the possible relationship between the relative position regulated by pHo and S4 site, we further evaluated the response of S4 site mutants to alteration of pHo. S4 is the innermost site of SF; hence, if pHo acts on the S4 site or inner part of SF, the response of S4 mutants to pHo will be altered. However, under physiological potassium concentration (5 mM), the degree of pHo effects was resembled between S4 mutants and WT channels (Fig. 5a). The fitting to Hill equation revealed a half block (pH50) of 7.67 ± 0.35, 8.12 ± 0.09, and 7.95 ± 0.12 for WT, T172S, and T281S, respectively (Fig. 5b, p > 0.05). Since mutations of S4 site do not affect pHo response of TREK-2 channels, the conformational transition evoked by pHo fluctuation is other than the S4 site, that is, the outer position of SF.

Fig. 5
figure 5

Response of ATI isoforms and mutants to extracellular fluctuation of pH. a Normalized responses of the indicated WT and S4 site mutated channels to pHo changes in 5 mM [K+]o solution. Data (mean ± SE, n ≥ 6) was taken at 0 mV, normalized to activity at pH 6.5. b Summary of the pH50 values for pH inhibition of the indicated TREK-2 channels. Results were shown to as mean ± SE (n ≥ 5). c Exemplar recordings of the response of WT channels and M1L2L3 mutant to the external pHo changes from 6.5 to 9.3. The currents recorded in pH 9.3 were normalized by those recorded in pH 6.5. d Comparison of the pHo response between ATI isoforms and related mutants. Data (mean ± SE, n ≥ 6) was taken at 0 mV, normalized to activity at pH 6.5

Full size image

Next, we investigated whether the ATI isoforms possess similar conformation in outer SF by examining the pHo response of the Nt mutants. Firstly, the gradual current loss induced by pHo elevation from 6.5 to 9.3 in the representative oocytes expressing WT channel and M1L2L3, one of ATI isoform, was very similar (Fig. 5c). Further measurement of the response all the Nt mutants to varied pHo (Fig. 5d) revealed that the pH50 for M2L3, M3, and M1L2L3 was 7.70 ± 0.54, 7.74 ± 0.24, and 7.72 ± 0.18, respectively, a similar range with WT channels (Fig. 5b, p > 0.05). The resembled pHo response between WT channels and Nt mutants demonstrate that the outer SF conformation of ATI isoforms (the position was termed as pHo position in this study) is similar with that of WT channels.

Taken together, our provided data indicate that the SF conformations of ATI isoforms and WT TREK-2 are similar in the S4 site and pHo position, suggesting the role of Nt domain is not associated with the conformation of SF. Since the domain controls the single-channel conductance [30], it seems that the conformational alteration of SF does not contribute to unitary conductance of TREK-2.

Discussion

Ba2+ has been proven to be a valuable tool to learn about the SF configuration of K+ channels, due to its ability to prevent the K+ flux effectively in several types of K+ channels through fitting into the SF [5, 6, 14, 22, 25]. Using Ba2+ as a probe, we have previously confirmed that extracellular K+-free and acidification induce collapsed SF conformation in TREK-1 [22]. Here, by detail characterizing the interaction between Ba2+ and TREK-2 channels, we demonstrate that the ions are also capable of blocking TREK-2 expressed in X. laevis oocytes, in a concentration- and time-dependent manner (Fig. 1). The inhibition is blunted by elevated [K+]o, strongly suggesting that Ba2+ exerts its inhibitory function to TREK-2 via acting on the SF gate (Fig. 2). The decreased sensitivity of S4 site mutants (T172S and T281S) to the application of Ba2+ provides further evidence that the ion binds at S4 site of TREK-2 (Fig. 3). The regulation mode of Ba2+ to TREK-2 supports the notion that the SF is highly conserved between TREK-2 and other types of K+ channels, as illustrated in the crystal structures of K2P channels [3, 23]. Moreover, Ba2+ is also a qualified tool to explore the properties of the SF in TREK-2 channels.

Except for Ba2+, we also investigated the effects of pHo, which is considered as another input modality that acts on the SF of TREK-2. TREK-1 is inhibited by acidic pHo via a mechanism resembling C-type inactivation, in which the structure of outer SF is collapsed [8, 22]. In contrast to the situation of TREK-1, the activity of TREK-2 is inhibited by alkaline pHo [29]. Elevated [K+]o impedes the inhibition [29, 36], which suggests that pHo also acts on the SF of TREK-2 channels. We reveal that destruction of S4 site does not alter the sensitivity of TREK-2 to pHo (Fig. 5a, b), which implicates that the conformational transition induced by pHo occurs at the location (pHo position) where is different from S4 site. As S4 is the innermost K+ binding site, the pHo position should be located at the outer SF, which is consistent with the concept that C-type inactivation involves a localized constriction in the outer vestibule of the selectivity filter. Structural analysis of KcsA reveals that S2 and S3 are lost in the SF configuration of C-type inactivation [11], whereas functional characterization of TASK2 (TWIK-related acid-sensitive potassium channel 2, another member of K2P family) indicates that the energetic barriers at S0/S1 account for channel blockade at acid pHo [38]. The outer SF regulation properties enable pHo to be another facile tool to explore the characteristics of other stimuli that act on the SF of TREK-2 channels.

Cloned rat and human TREK-2 expressed in mammalian cells and in Xenopus oocytes exhibits multiple unitary conductance levels that range from ∼52 to ∼220 pS [15, 30]. Nt participates in the regulation of conductance, as illustrated by the behavior of the three ATI isoforms. Full length isoform has small unitary conductance near 52 pS, while short isoforms translated from the second and third initiation sites have large conductance between 185 and 220 pS. Using extracellular Ba2+ and pHo as tools, we investigated whether the Nt commands the conductance via controlling the SF conformation. In the presence of different [Ba2+]o, all the ATI isoforms exhibit similar response, including the access rate, exit rate, and inhibition ratio (Fig. 4). Obviously, the Nt does not affect Ba2+ binding to TREK-2, which suggests the all the ATI isoforms have similar S4 configuration, and the Nt does not influence the S4 conformation. Moreover, we also evaluated the response of all the ATI isoforms to varied pHo. Very similar to the situation of Ba2+, the isoforms also display analogous responses to pHo (Fig. 5c, d). Since pHo acts on the outer vestibule of the SF, we conclude that all the ATI isoforms possess similar outer SF conformation. Together, these data strongly implicate that all the ATI isoforms and WT channels show similar conformation in their SF region, and the Nt of TREK-2 channels is not associated with the SF conformation. We then speculate that in TREK-2 channels, the conformation of the SF has no impact on the single-channel conductance. This concept is in accordance with the phenomenon that all the ATI isoforms possess different conductance but exhibit similar selectivity to K+ [30]. In voltage-gated K+ (Kv) channels, residue substitutes in the vicinity of inner gate or replacement the inner helices (constitute the inner gate) affect the unitary conductance, whereas exchanging the SF between shaker and Kv3.1 does not influence the conductance [19, 32]. These results implicate that the unitary conductance is highly associated with the status of the inner gate [17]. Hence, the SF configuration is mainly conserved to control the selectivity to K+ and to allow K+ flow, rather than regulating the conductance. Our data regarding the function of SF on unitary conductance in TREK-2 channels appear resembled with those of Kv channels. However, this property is not necessarily identical among all the K+ channels, even in the TREK subclass of K2P channels. For example, ATI isoforms of TREK-1 channels also show big divergence in unitary conductance, but the Nt deletant [33], one of the ATI isoforms, displays nonconductive SF conformation compared with its wild-type counterpart [22]. And accordingly, the selectivity for K+ ions of the truncated channels is decreased [33].

In addition to identical selectivity to K+ among TREK-2 isoforms, these isoforms also have similar sensitivity to activators and inhibitors such as arachidonic acid, membrane stretch, proton, and cAMP. It has been established in TREK-1 channels that the active sites of these stimuli is gathered within the proximal region of intracellular C-terminus [13, 20], a relatively highly conserved region between TREK-1 and TREK-2. Therefore, it is easy to understand that the active sites of these input modalities could also be located within the proximal side of the C-terminus. Consequently, all the ATI isoforms have completely identical C-terminus, which might lead to their similar response to above stimuli.

In summary, we found that externally, application of Ba2+ inhibits TREK-2 currents in a time- and concentration-dependent manner. The blocking effects are blunted by elevated [K+]o and mutations of S4 site, suggesting the inhibitory mechanism of the ion is originated from its docking properties within the SF. Furthermore, we identified that pHo acts on the position outer of S4 site. Then, using extracellular Ba2+ and pHo as probing tools, we demonstrate that TREK-2 isoforms generated by ATI mechanism adopts similar conformation in S4 site and the pHo position, implicating that the Nt has a minor effect, if any, on regulating the SF conformation. We thus speculate the SF conformation plays a little role on the single-channel conductance in TREK-2 channels. Our findings characterize the relation between the Nt and the SF status, which may help to understand the detail gating mechanism of TREK-2 channels and K2P channels.