Direct neurotransmitter activation of voltage-gated potassium channels

Abstract

Voltage-gated potassium channels KCNQ2-5 generate the M-current, which controls neuronal excitability. KCNQ2-5 subunits each harbor a high-affinity anticonvulsant drug-binding pocket containing an essential tryptophan (W265 in human KCNQ3) conserved for >500 million years, yet lacking a known physiological function. Here, phylogenetic analysis, electrostatic potential mapping, in silico docking, electrophysiology, and radioligand binding assays reveal that the anticonvulsant binding pocket evolved to accommodate endogenous neurotransmitters including γ-aminobutyric acid (GABA), which directly activates KCNQ5 and KCNQ3 via W265. GABA, and endogenous metabolites β-hydroxybutyric acid (BHB) and γ-amino-β-hydroxybutyric acid (GABOB), competitively and differentially shift the voltage dependence of KCNQ3 activation. Our results uncover a novel paradigm: direct neurotransmitter activation of voltage-gated ion channels, enabling chemosensing of the neurotransmitter/metabolite landscape to regulate channel activity and cellular excitability.

Fig. 1

Neuronal KCNQs contain an evolutionarily conserved GABA-binding pocket. a Sequence alignment of S5 residues 265–275 (human numbering) in KCNQ5 or the closest equivalent in representative organisms from the clades shown in b. Yellow, tryptophan essential for drug binding (KCNQ5-W270; KCNQ3-W265). Organism descriptors: black, single KCNQ gene in genome, no W270 equivalent; dark blue, multiple KCNQ genes, no W270 equivalent; red, multiple KCNQ genes, at least one containing W270 equivalent. b Phylogeny (distances not intended to equate to time) indicating evolutionary relationships from ancestral Protista to Chordates and other major clades. Text colors as in a; gray, no KCNQ genes. c Upper, chimeric KCNQ1/KCNQ3 structural model (red, KCNQ3-W265; cyan, KCNQ3-L314); lower, topological representation (pentagon, KCNQ3-W265). d Electrostatic surface potentials (red, electron-dense; blue, electron-poor; green, neutral) and structures calculated and plotted using Jmol. e Close-up side views of KCNQ structure as in c, showing results of SwissDock unguided in silico docking of GABA, retigabine, and ML-213. f View from extracellular side of KCNQ structure showing W265 accessibility. g View from extracellular side of Kv1.2-Kv2.1 “paddle-chimera” structure showing S5 accessibility (*) despite lipid presence (lipid spacefills shown). h Mean ³H-GABA binding, normalized to channel surface expression, for KCNQ channels expressed in Xenopus oocytes; n = 47 (Q1), 14 (Q2), 48 (Q3), 30 (Q4), 44 (Q5) in two to four batches of oocytes; ***P = 0.0003; ****P < 0.0001; versus KCNQ1. Inset, sequence alignment of S5 265–275 (KCNQ5 numbering) for human KCNQ1–5. Yellow, KCNQ5-W270. Error bars indicate SEM. i Mean ³H-GABA binding to KCNQ3 expressed in Xenopus oocytes, with (+) versus without (−) competition from unlabeled (cold) GABA (1 mM); n = 20 per group; **P = 0.007. Error bars indicate SEM. j Saturation binding studies using ³H-GABA and oocyte-expressed KCNQ3. Ten different ³H-GABA concentrations were used (0.1–1000 nM; n = 20 oocytes per group). Mean non-specific binding was quantified in parallel experiments similar except for the addition of 1 mM cold GABA. Mean specific binding was calculated by subtracting mean non-specific from total binding. Error bars indicate SEM

Fig. 2

KCNQ3 and KCNQ5 are activated by GABA in Xenopus oocytes. All error bars in figure indicate SEM. a Mean tail current versus prepulse voltage relationships recorded by TEVC in Xenopus laevis oocytes expressing homomeric KCNQ1–5 channels in the absence (black) and presence (purple) of GABA (n = 4–8). Voltage protocol as in c. b Mean GABA dose response at −60 mV for KCNQ1–5, quantified from data as in a (n = 5–8). c TEVC recordings using a current-voltage family protocol (upper left inset) in oocytes expressing KCNQ2/3 channels in the absence (black) and presence (purple) of GABA (10 µM). Dashed line indicates zero current level in this and all following current traces. d Averaged current traces at −60 mV for KCNQ2/3 channels in the absence (black) and presence (purple) of GABA (10 µM) (n = 10). Voltage protocol upper inset. e Left, mean tail current; right, mean normalized tail current (G/Gmax), both measured at arrow in voltage protocol (center) from traces as in c, with versus without GABA (10 µM) (n = 10). f Mean voltage dependence of KCNQ2/3 current fold-increase by GABA (10 µM), plotted from traces as in c (n = 10). g Mean dose response of KCNQ2/3 channels at −60 mV for GABA (calculated EC₅₀ = 0.85 µM; n = 10) and glutamate (no effect; n = 5). h GABA hyperpolarizes resting membrane potential (E_M) of unclamped oocytes expressing KCNQ2/3. Left, effects of 10 µM GABA; right, dose response; n = 10, ****P < 0.0001. i Exemplar −60 mV KCNQ2/3 current before (left, black), during wash-in of GABA (purple) and after wash-out (right, black). j Exemplar −60 mV KCNQ2/3 current before (left, black), during wash-in of GABA (purple), wash-in of XE991 (red), and after wash-out (right, black). Membrane potential was clamped at −60 mV except for a 2-min pulse to +60 mV during the early phase of GABA wash-in. k Mean activation (left) and deactivation (right) rates for KCNQ2/3 before (control) and after wash-in of GABA (n = 10); **P < 0.01; ***P = 0.0007. Activation rate was quantified using voltage protocol as in c. Deactivation rate was quantified using voltage protocol shown above

Fig. 3

GABA directly opens KCNQ2/3 channels. All error bars in figure indicate SEM. a TEVC of water-injected Xenopus laevis oocytes showing no effect of GABA (100 µM) on endogenous currents or (lower right) resting membrane potential (E_M) (n = 4–5). b TEVC of Xenopus laevis oocytes expressing KCNA1 (Kv1.1) showing no effect of GABA (1 mM) on peak current (left) or normalized tail current (right) (n = 4). c–f GABA effects on KCNQ2/3 in oocytes do not require GABA_B receptor activity. Mean KCNQ2/3 tail currents in oocytes showing GABA activates KCNQ2/3 in the presence of pertussis toxin (2 µg/ml), saclofen (100 µM), or CGP55845 (100 µM) and that neither baclofen (100 µM), saclofen, nor CGP55845 alter KCNQ2/3 current independently (n = 5–7). g Upper, tail current; lower, normalized conductance; showing mean GABA response of oocyte-expressed KCNQ2/KCNQ3-W265L (n = 5). h Upper, tail current; lower, normalized conductance; showing mean GABA response of oocyte-expressed KCNQ2-W236L/KCNQ3-W265 (Q2/Q3-WL/WL) (n = 5). i Mean current fold-changes (upper) and dose responses (lower) for channels as indicated; KCNQ2/KCNQ3 results (purple line) and KCNQ3* results (red line) from Fig. 2 shown for comparison; n = 5–10. j Mean ³H-GABA binding for KCNQ2/3 versus KCNQ2-W236L/KCNQ3-W265L channels expressed in Xenopus oocytes; n = 73 (Q2/Q3), 29 (KCNQ2-W236L/KCNQ3-W265L) in two to four batches of oocytes; ****P < 0.0001

Fig. 4

GABA activates KCNQ2/3 in CHO cells and native M-current. All error bars in figure indicate SEM. a Exemplar current traces from CHO cells transfected to express KCNQ2/3 channels, recorded using whole-cell patch clamp, showing effects of 100 µM GABA (n = 6). b Mean normalized tail current from recordings as in a (n = 6). c Representative micrographs (left) and whole-cell currents (right) from undifferentiated (upper) versus nerve growth factor (NGF)-differentiated (lower) PC12 cells. Scale bars, 10 µm. d Upper, representative tail currents (using voltage protocol as in c); lower, mean normalized tail currents, recorded from NGF-differentiated PC12 cells bathed in extracellular solution alone (control) or with picrotoxin (100 µM) and CGP55845 (10 µM) to block GABA_A and GABA_B receptors, respectively, alone (black square) or in combination with GABA (100 µM; purple circle) or XE991 (10 µM; open red square); n = 7). e Representative tail currents (using voltage protocol on left) recorded from mouse DRG neurons bathed in extracellular solution containing picrotoxin (100 µM) and CGP55845 (10 µM), alone (black) or in combination with GABA (100 µM; purple) or GABA + XE991 (10 µM; red). f Mean current at −60 mV divided by current at −20 mV in the same DRG neuron, using the protocol as in e, in the absence (control) or presence of GABA (n = 7); **P = 0.005. g Mean GABA-dependent increase in current at −20 mV versus at −60 mV in DRG neurons, using the protocol as in e (n = 7); P = 0.07 between groups. h Mean effect of GABA versus GABA + XE991 on the magnitude of the deactivating current at −60 mV, using the protocol as in e (n = 5); P = 0.1 between groups

Fig. 5

GABA activation is independent of and overrides muscarinic inhibition of KCNQ2/3 in Xenopus oocytes. All error bars in figure indicate SEM. a Mean dose response for acetylcholine (ACh) on KCNQ2/3 activity in oocytes (n = 5). b Mean effects versus voltage for ACh or atropine alone or in combination on KCNQ2/3 activity in oocytes (n = 5). c Lack of effects of dopamine (100 µM) on KCNQ2/3 mean tail currents in oocytes (n = 6). d Atropine blocks KCNQ2/3 inhibition by ACh (measured using tail currents in oocytes; n = 5). e Lack of effects of atropine (100 µM) on KCNQ2/3 mean tail currents in oocytes (n = 5). f Atropine does not prevent GABA activation of KCNQ2/3 measured via tail currents in oocytes (n = 5). Left, mean tail currents, right, current fold-change versus voltage for GABA + atropine versus no drugs. g GABA prevents inhibition of KCNQ2/3 by ACh, measured via tail currents in oocytes (n = 5). Left, mean tail currents, right, current fold-change versus voltage for GABA + ACh versus no drugs. h GABA (10 µM) effects on KCNQ2/3 in oocytes are not prevented by partial depletion/inhibition of synthesis of PIP₂ using wortmannin (30 µM for 3 h) (n = 5). Left, mean tail current; right, GABA dose response at −60 mV; (n = 5). i GABA (100 µM) has no effect on KCNQ1/KCNE1. Left, averaged traces; right, mean tail currents (n = 6). j GABA (100 µM) does not alter resting membrane potential (E_M) of unclamped oocytes expressing KCNQ1-KCNE1 (n = 6)

Fig. 6

GABA and metabolites compete for binding to KCNQ3-W265. All error bars in figure indicate SEM. a Mean effects of BHB and GABOB (10 µM) versus valeric acid (no effects at 5 mM) on KCNQ2/3 in oocytes (n = 5–6). b Mean dose response of KCNQ2/3 activation at −60 mV by GABA (from Fig. 2) and related compounds (n = 5–10). c In silico docking predicts that BHB and GABOB bind to KCNQ3-W265, the latter also H-bonding with L237 in the S4–5 linker. d, e GABOB (d), but not glutamate (e), antagonizes the effects of GABA on KCNQ2/3 activation in oocytes (n = 4–6). Upper, mean tail currents; lower, current fold-change elicited by neurotransmitter combination versus voltage. f GABOB antagonizes the effects of retigabine on KCNQ2/3 activation in oocytes (n = 10). Upper, mean tail currents; lower, current fold-change elicited versus voltage. Data are for KCNQ2/3 in the absence of retigabine (Ctrl), with retigabine (RTG), or with retigabine + GABOB (RTG + GABOB). **P < 0.01; ****P < 0.0001