Mechanism of the modulation of BK potassium channel complexes with different auxiliary subunit compositions by the omega-3 fatty acid DHA

Abstract

Large-conductance Ca(2+)- and voltage-activated K(+) (BK) channels are well known for their functional versatility, which is bestowed in part by their rich modulatory repertoire. We recently showed that long-chain omega-3 polyunsaturated fatty acids such as docosahexaenoic acid (DHA) found in oily fish lower blood pressure by activating vascular BK channels made of Slo1+β1 subunits. Here we examined the action of DHA on BK channels with different auxiliary subunit compositions. Neuronal Slo1+β4 channels were just as well activated by DHA as vascular Slo1+β1 channels. In contrast, the stimulatory effect of DHA was much smaller in Slo1+β2, Slo1+LRRC26 (γ1), and Slo1 channels without auxiliary subunits. Mutagenesis of β1, β2, and β4 showed that the large effect of DHA in Slo1+β1 and Slo1+β4 is conferred by the presence of two residues, one in the N terminus and the other in the first transmembrane segment of the β1 and β4 subunits. Transfer of this amino acid pair from β1 or β4 to β2 introduces a large response to DHA in Slo1+β2. The presence of a pair of oppositely charged residues at the aforementioned positions in β subunits is associated with a large response to DHA. The Slo1 auxiliary subunits are expressed in a highly tissue-dependent fashion. Thus, the subunit composition-dependent stimulation by DHA demonstrates that BK channels are effectors of omega-3 fatty acids with marked tissue specificity.

Fig. 1

Differential effects of DHA on Slo1 BK channels with various auxiliary subunits. (A) Structure of DHA (docosahexaenoic acid). (B) Representative currents elicited by pulses to different voltages in inside-out patches without Ca²⁺ through channels composed of Slo1 alone, Slo1+β1, Slo1+β2, Slo1+β2 ∆2–19, Slo1+β4, and Slo1+LRRC26 (γ1). In each panel, currents before (blue) and after (red) application of DHA (3 µM) are shown. Pulses were applied every 2 s except for Slo1+β2, which was stimulated every 10 s. For Slo1+β2, 1-s prepulses to −100 mV preceded depolarization pulses. For Slo1+LRRC26 (γ1), the holding voltage was −80 mV. (C) Representative peak IV curves from Slo1+β1, Slo1+β2 ∆2–19, Slo1+β4, and Slo1+LRRC26 (γ1). Results before (blue) and after (red) application of DHA (3 µM) are shown. All results were obtained in the virtual absence of Ca²⁺.

Fig. 2.

Changes in gating properties of Slo1+β1 (A), Slo1+β2 ∆2–19 (B), and Slo1+β4 (C) caused by DHA (3 µM) without Ca²⁺. For each channel type (from Top to Bottom), the fractional increase in peak outward current by DHA, normalized GV curves before (blue) and after (red) application of DHA, time constants of current relaxation before (blue) and after (red) application of DHA, and the fractional increase in time constant by DHA are shown. n = 4–11. In the GV panels, the smooth curves are Boltzmann fits to the results with V_0.5 = 160.0 ± 1.0 mV and Q_app = 0.90 ± 0.04 (control) and 102.7 ± 1.7 mV and 0.93 ± 0.05 (DHA) in A, V_0.5 = 156.9 ± 1.0 mV and Q_app = 1.05 ± 0.04 (control), and 131.6 ± 1.2 mV and 0.96 ± 0.04 (DHA) in B, and V_0.5 = 216.3 ± 2.2 mV and Q_app = 0.88 ± 0.07 (control) and 150.2 ± 2.3 mV and 0.82 ± 0.06 (DHA) in C. All results were obtained in the virtual absence of Ca²⁺.

Fig. 3.

Changes in steady-steady activation properties by DHA in different Slo1 channel complexes. (A) ∆V_0.5 and fractional changes in Q_app in the channels indicated. The ∆V_0.5 values for Slo1+β1 and Slo1+β4 differ from those for Slo1+β2 ∆2–19 and Slo1+LRRC26 (γ1), which are in turn different from those for Slo1 alone (P < 0.05). The Q_app ratio values were statistically indistinguishable among the channel types. (B) Structural organization of β subunits. (C) ∆V_0.5 and fractional changes in Q_app in Slo1+β1–β2 chimeric constructs. The ∆V_0.5 results for the Middle two chimeras are different from the Top and the Bottom chimera, which are in turn different from each other (P < 0.05). (D) ∆V_0.5 and fractional changes in Q_app in the chimeric constructs targeting the N terminus and TM1. Based on the ∆V_0.5 values, the Upper two channels, the third and fourth channels, and the Lower channel constitute three statistically distinct groups (P < 0.05). In C and D, pink regions in the structural cartoons represent β1 segments, and light blue areas represent β2 segments. All results are from inside-out experiments with 3 µM DHA in the virtual absence of Ca²⁺.

Fig. 4.

Positions 11 and 18 in β1 are important for DHA sensitivity. (A) N-terminal and TM1 sequences of β1, β2 ∆2–32, and β4. In B–G, ∆V_0.5 and ratios of Q_app in β1-to-β2 mutants (B), β1 R11 mutants (C), β1 C18 mutants (D), β2-to-β1/β4 mutants (E), β4-to-β2 mutants (F), and charge-pair double mutants (G) are shown. The gray shaded areas represent the mean ± SEM in Slo1+β1 and Slo1+β2 ∆2–19 (B–E, and G) and Slo1+β4 and Slo1+β2 ∆2–19 (F) for comparison. NF, nonfunctional—currents from cells transfected with these constructs were indistinguishable from those transfected with Slo1 alone. In B, only the ∆V_0.5 values for Slo1+β1 R11A are different from all others and Slo1+β1 (P < 0.05) but indistinguishable from those for Slo1+β2 ∆2–32 (P < 0.05). In C, only the ∆V_0.5 values for R11K are different from those for other mutants, which in turn are indistinguishable from those for Slo1+β2 ∆2–32 (P < 0.05). In D, the ∆V_0.5 values for C18E are greater than those for Slo1+β2 ∆2–32 but smaller than those for Slo1+β1 (P < 0.05). In E, β2∆n = β2 ∆2–32. (H) Changes in gating of Slo1+β2 ∆2–32 A42R:L49C possessing β2-to-β1 mutations in the β2 background by DHA (3 µM). From Upper to Lower, representative currents before (blue) and after (red) application of DHA, fractional increases in peak current size, GV curves, time constants of current relaxation, and fractional increases in time constant of current relaxation are shown. The gray areas, when present, represent the results (mean ± SEM) obtained from Slo1+β1. n = 4–8. In the GV panel, the smooth curves are Boltzmann fits to the results with V_0.5 = 192.1 ± 0.8 mV and Q_app = 0.98 ± 0.03 (control) and 129.9 ± 2.4 mV and 0.87 ± 0.07 (DHA). (I) Changes in gating of Slo1+β2 ∆2–32 A42R:L49C possessing β2-to-β4 mutations in the β2 background by DHA (3 µM). Graphs shown are as in H. Gray areas, when present, represent the results (mean ± SEM) obtained from Slo1+β4. n = 5–7. In the GV panel, the smooth curves are Boltzmann fits to the results with V_0.5 = 202.2 ± 1.0 mV and Q_app = 0.99 ± 0.03 (control) and 148.9 ± 1.7 mV and 0.94 ± 0.06 (DHA). All results are from inside-out experiments with 3 µM DHA without Ca²⁺.