The ubiquitous flavonoid quercetin is an atypical KCNQ potassium channel activator

Figures

• 

  Fig. 1. Caper extract KCNQ-dependently hyperpolarizes cells. All error bars indicate SEM. n number of oocytes. a Topological representation of a Kv channel showing two of the four subunits that comprise a channel. PH, pore helix. VSD, voltage sensing domain. b Schematic of heteromeric composition of KCNQ2/KCNQ3 (left) and KCNQ1/KCNE1 (right) channels. c Image of caper bush (Capparis spinosa). d Image of pickled capers used in this study. e Mean TEVC current traces for water-injected Xenopus oocytes as indicated, in the absence (Control) or presence of 1% caper extract (n = 4). Dashed line here and throughout indicates the zero-current level. Right inset, the voltage protocol used here and throughout the study (used with either 10 or 20 mV prepulse increments) unless otherwise indicated. f Mean tail current versus prepulse voltage, for traces as in e (n = 4). g Scatter plot of unclamped membrane potential (EM) for cells as in e (n = 4). Statistical analyses by two-way ANOVA. h Mean TEVC current traces for KCNQ1 and KCNQ2/KCNQ3 in the absence (Control) or presence of 1% caper extract (n = 4–6). Arrow, current decay. Circles, regions used for close-up images in k. i Left, mean tail current;…

• 

  Fig. 2. Quercetin but not rutin activates KCNQ2/3 channels. All error bars indicate SEM. n number of oocytes. a Chemical structures (red indicates oxygen) and electrostatic surface plot (red, negative; blue, positive) of quercetin. Arrows/circle, carbonyl group. b Chemical structures (red indicates oxygen) and electrostatic surface plot of rutin. Circled region indicates quercetin moiety. c Mean KCNQ2/KCNQ3 traces in the absence (Control) and presence of 100 µM rutin (n = 5). d Mean tail current and normalized tail currents (G/Gmax) versus prepulse voltage relationships for the traces as in c. e Mean KCNQ2/KCNQ3 traces in the absence (Control) and presence of 100 µM quercetin (n = 4). f Mean tail current and normalized tail currents (G/Gmax) versus prepulse voltage relationships for the traces as in e. g Current fold-change in KCNQ2/3 tail currents in the presence and absence of quercetin or rutin (n = 4–5). h Scatter plot of effect of rutin (100 µM) on unclamped membrane potential (EM) of KCNQ2/3-expressing cells as in c. Statistical analysis by two-way ANOVA. i Scatter plot of effect of quercetin (100 µM) on unclamped membrane potential (EM) of KCNQ2/3-expressing cells as in e.…

• 

  Fig. 3. Quercetin activates KCNQ2/3 by an atypical mechanism. All error bars indicate SEM. n number of oocytes. a Mean TEVC traces of homomeric KCNQ2 and KCNQ3* in the absence (Control) and presence of 100 µM quercetin (n = 4–6). b Mean tail current and normalized tail currents (G/Gmax) versus prepulse voltage relationships for the traces as in a. c Scatter plot of unclamped membrane potential (EM) for cells in A. Statistical analysis by two-way ANOVA. d Structural model with close-up (boxed) of KCNQ3 based on Cryo-EM structure of Xenopus KCNQ1 with KCNQ3-W265 (red) and KCNQ3-R242 (yellow) highlighted. e Topological representation with close-up (boxed) of KCNQ3 showing two of the four subunits that comprise a channel, and approximate locations of KCNQ3-W265 (red) and KCNQ3-R242 (yellow). VSD, voltage sensing domain. f Mean TEVC traces of heteromeric KCNQ2-W236L/KCNQ3-W265L (KCNQ2/Q3-WL/WL) and heteromeric KCNQ2-R213A/KCNQ3-R242A (KCNQ2/Q3-RA/RA) in the absence (Control) and presence of 100 µM quercetin (n = 4–6). g Mean tail current and normalized tail currents (G/Gmax) versus prepulse voltage relationships for the traces in f. h Scatter plot of unclamped membrane potential (EM) for…

• 

  Fig. 4. Quercetin but not rutin promotes KCNQ1 activation and inactivation. All error bars indicate SEM. n number of oocytes. a Mean TEVC traces of homomeric KCNQ1 in the absence (Control) and presence of 100 µM quercetin (n = 4) or 100 µM rutin (n = 5). b Mean tail current and normalized tail currents (G/Gmax) versus prepulse voltage relationships for the traces in a. c Scatter plot of unclamped membrane potential (EM) for cells in A. Statistical analysis by two-way ANOVA. d Human KCNQ1 closed-state structural model showing lack of predicted quercetin binding proximal to R243. e Human KCNQ1 open-state structural model showing lack of predicted quercetin binding proximal to R243. f Mean TEVC traces of homomeric KCNQ1-R243A in the absence (Control) and presence of 100 µM quercetin (n = 5). g Mean tail current and normalized tail currents (G/Gmax) versus prepulse voltage relationships for the traces in f. h Scatter plot of unclamped membrane potential (EM) for cells in f. Statistical analyses by two-way ANOVA.

• 

  Fig. 5. Quercetin speeds KCNQ1 activation and inactivation. All error bars indicate SEM. n number of oocytes. a Voltage protocol and mean TEVC current traces of KCNQ1 comparing the relationship between shorter prepulse durations and normalized tail current at −80 mV (n = 4) in the absence and presence of 100 µM quercetin. b Voltage protocol and mean TEVC current traces of KCNQ1 comparing the relationship between longer prepulse durations and normalized tail current at −80 mV (n = 4) in the absence and presence of 100 µM quercetin. c Mean normalized peak tail current versus prepulse duration in the absence (Ctrl) or presence of 100 µM quercetin quantified from traces in a, b and fit with a single exponential function (n = 4). d Voltage protocol and mean TEVC current traces of KCNQ1 comparing the tail current recovery “hook” currents at +40 mV (solid arrow) and −80 mV (open arrow) in the absence and presence of 100 µM quercetin (n = 5). e Close-up view of inactivation recovery “hooks” taken from initial tail current portion of mean KCNQ1 current traces in the absence (Control) and presence of 100 µM quercetin. Red line indicates −30 mV tail pulse after +40 mV prepulse, used for…

• 

  Fig. 6. Quercetin requires extracellular S4 charged residues R228 and R231 for KCNQ1 current augmentation. All error bars indicate SEM. n number of oocytes. a Human KCNQ1 closed-state structural model36 showing predicted quercetin binding to the top of the VSD (red circles). b Human KCNQ1 open-state structural model36 showing lack of predicted quercetin binding to the top of the VSD (black circles). c Human KCNQ1 closed-state structural model close-ups36 showing predicted quercetin binding to the top of 3 of the VSDs. d Close-up of top of one VSD from the cryo-EM-derived human KCNQ1-calmodulin structure showing predicted docking of quercetin close to the R231 sidechain. e Close-up of top of another VSD from the cryo-EM-derived human KCNQ1-calmodulin structure showing predicted H-binding (green line) of quercetin to the R231 sidechain. f Mean TEVC current traces showing effects of quercetin (100 µM) on KCNQ1-R228A (n = 5). g Mean peak tail current versus prepulse voltage relationship for the traces as in f (n = 5). h Scatter plot of mean normalized tail hook for the traces as in f (n = 4–5). i Scatter plot of unclamped membrane potential (EM) for cells as in f (n = 5). Statistical…

• 

  Fig. 7. Quercetin requires S6 residue F340 for modulation of KCNQ1 activation and inactivation. All error bars indicate SEM. n number of oocytes. a Human KCNQ1 open-state structural model close-up36 showing predicted quercetin binding to F340. b Mean TEVC current traces showing effects of quercetin (100 µM) on KCNQ1-F340V (n = 5). c Close-up of inactivation recovery “hook” currents from traces as in b. d Mean peak tail current (left) and normalized peak tail current (G/Gmax) (right) versus prepulse voltage relationships for the traces as in b (n = 5). e Scatter plot of unclamped membrane potential (EM) for cells as in b (n = 5). Statistical analysis by two-way ANOVA.

• 

  Fig. 8. Quercetin requires S4 charged residues R228 and R231 for augmentation of KCNQ1/KCNE1 current. All error bars indicate SEM. n number of oocytes. a Mean current traces showing effects of quercetin (100 µM) on KCNQ1/KCNE1 (n = 5). b Mean peak tail current and current fold-change versus prepulse voltage for traces as in a (n = 5). c Scatter plot of unclamped membrane potential (EM) for cells as in A (n = 5). Statistical analyses by two-way ANOVA. d Human KCNQ1/KCNE1 closed-state structural model36 showing predicted quercetin binding to the top of the VSD. e Human KCNQ1 closed-state structural model close-up36 showing predicted quercetin binding to the top of the VSD. f Human KCNQ1 open-state structural model close-up36 showing predicted quercetin binding to the VSD. g Mean current traces showing effects of quercetin (100 µM) on KCNQ1-R228A/KCNE1 (n = 6). h Mean peak and normalized peak tail current (G/Gmax) versus prepulse voltage for traces as in g (n = 6). i Scatter plot of unclamped EM for cells as in g (n = 6). Statistical analyses by two-way ANOVA. j Mean current traces showing effects of quercetin (100 µM) on KCNQ1-R231A/KCNE1 (n = 5). k Mean peak and normalized peak tail…

• 

  Fig. 9. Quercetin requires S4 charged residue R231 for augmentation of KCNQ1/KCNE3 current. All error bars indicate SEM. n number of oocytes. a Mean TEVC current traces showing effects of quercetin (100 µM) on KCNQ1/KCNE3 (n = 6). b Mean peak tail current (left) and normalized peak tail current (G/Gmax) (right) versus prepulse voltage relationships for the traces as in a (n = 6). c Scatter plot of unclamped membrane potential (EM) for cells as in a (n = 6). Statistical analyses by two-way ANOVA. d Human KCNQ1/KCNE3 structural model49 showing predicted quercetin binding to the top of the VSD. e Human KCNQ1/KCNE3 structural model49 close-up showing predicted quercetin binding to the top of the VSD. f Three predicted docking clusters at the top of the VSD of the cryo-EM-derived human KCNQ1-PIP2-calmodulin/KCNE3 structure37 showing docking of quercetin close to and/or hydrogen bonding (green line) with the R231 sidechain. g Mean TEVC current traces showing effects of quercetin (100 µM) on KCNQ1-R228A/KCNE3 (n = 5). h Mean peak tail current (left) and normalized peak tail current (G/Gmax) (right) versus prepulse voltage relationships for the traces as in g (n = 5). i Scatter plot of unclamped…

• 

  Fig. 10. Quercetin and cilantro aldehyde E-2-dodecenal combine to synergistically inactivate KCNQ1. All error bars indicate SEM. n number of oocytes. a Quercetin dose response for channels indicated, quantified by shift in midpoint voltage dependence of activation (ΔV0.5activation). b Mean current fold-changes for channels indicated, quantified by the quotient of the tail currents in the presence and absence of 100 µM quercetin. c Mean traces, tail and normalized tail current (G/Gmax) and EM as indicated for KCNQ1/KCNE1 and KCNQ3* in the absence (Ctrl) and presence of 100 µM rutin (n = 5). Statistical analysis by two-way ANOVA. d Mean TEVC traces, tail and normalized tail current (G/Gmax) and EM as indicated for KCNQ1/KCNE1 and KCNQ3* in the absence (Ctrl) and presence (R + Q) of 10 µM rutin plus 10 µM quercetin (n = 5). Statistical analysis by two-way ANOVA. e Mean KCNQ1/KCNE1 current increase (%) induced by rutin, quercetin or the combination of both (values in micromolar). f Mean KCNQ3* current increase (%) induced by rutin, quercetin or the combination of both (values in micromolar). g Human KCNQ1 structural model showing predicted quercetin and E-2-dodecenal binding locations. h Mean…