Modulation by the BK accessory β4 subunit of phosphorylation-dependent changes in excitability of dentate gyrus granule neurons

Abstract

Large-conductance voltage- and calcium-activated potassium (BK) channels are large-conductance calcium- and voltage-activated potassium channels critical for neuronal excitability. Some neurons express so called fast-gated, type I BK channels. Other neurons express BK channels assembled with the accessory β4 subunit conferring slow gating of type II BK channels. However, it is not clear how protein phosphorylation modulates these two distinct BK channel types. Using β4-knockout mice, we compared fast- or slow-gated BK channels in response to changes in phosphorylation status of hippocampus dentate gyrus granule neurons. We utilized the selective PP2A/PP4 phosphatase inhibitor Fostriecin to study changes in action potential shape and firing properties of the neurons. In β4-knockout neurons, Fostriecin increases BK current, speeds up BK channel activation and reduces action potential amplitudes. Fostriecin increases spiking during early components of an action potential train. In contrast, inhibition of BK channels through β4 in wild-type neurons or by the BK channel inhibitor Paxilline opposes Fostriecin effects. Voltage clamp recordings of neurons reveal that Fostriecin increases both calcium and BK currents. However, Fostriecin does not activate BK α channels in transfected HEK293 cells lacking calcium channels. In summary, these results suggest that fast-gating, type I BK channels lacking β4 can increase neuronal excitability in response to reduced phosphatase activity and activation of calcium channels. By opposing BK channel activation, the β4 subunit plays an important role in moderating firing frequency regardless of changes in phosphorylation status.

Figure 1. Protein phosphatase inhibition reduces action potential amplitudes in β4 knockout neurons

(A) Example action potential waveforms recorded from WT and β4 KO neurons during 210 pA current injection. Representative 10^th action potentials from an action potential train of control neurons (solid gray line), neurons treated with 250 nM Fostriecin (solid dark line), and neurons treated with Fostriecin and 5 μM Paxilline (dashed line). (B–D) Plots of averaged action potential properties including action potential amplitude (B), half-width (C) and fAHP amplitude (D). First columns represent control; second columns are for Fostriecin application (intracellular), third columns are for Fostriecin (intracellular) with Paxilline in the bath.

Figure 2. Fostriecin block of phosphatase activity causes BK channel-dependent increases in instantaneous firing frequency

(A, B) Examples of action potential trains elicited by 210 pA current injection for WT and KO neurons, respectively. Traces from control (left) are compared with traces recorded after application of Fostriecin alone (middle) or Fostriecin with Paxilline (right). (C) Summarized firing frequency from a one second current (210 pA) injection in WT and β4 KO neurons, respectively. (D, E) Instantaneous firing frequency is for the first ten action potentials during a 210 pA current injection.

Figure 3. Fostriecin block of phosphatase activity increases BK currents in β4 KO neurons

(A) Example traces before (upper panel) and 10 minutes after application of BK channel blocker Paxilline (middle panel). Lower panel shows subtracted net BK current. (B) Example isolation of BK current from a Fostriecin treated cell. Vertical scale represents 1 nA current (vertical) and 10 ms time (horizontal). (C) Current density is summarized for BK current as a function of voltage. Filled symbols are for control, empty symbols are for Fostriecin-treated cells. Numbers in parentheses represent number of measured cells. (D) Activation time of isolated BK current is shorter with Fostriecin treatment. Inset shows example isolated BK current traces during first 5 millisecond of a +60 mV voltage step.

Figure 4. Fostriecin block of phosphatase activity increases voltage-dependent calcium currents in the granule neurons of the β4 KO mice

Representative whole-cell currents from control (A) and from a cell treated with intracellular 250 nM Fostriecin (B) are summarized as current density as a function of command voltage (C). Traces shown are in 20 mV intervals from −60 mV to +60 mV. The numbers in parentheses represent number of measured cells. All measurements were done in the presence of blockers of voltage-dependent potassium and sodium currents (see Methods). An interpolation was used to connect the data points.

Figure 5. Cadmium block of calcium influx occludes effects of protein phosphatase inhibition

(A) Example β4 KO action potential waveforms recorded in cadmium alone (200 μM cadmium, grey line) or cadmium with Fostriecin (dashed line). Representative 10^th action potentials from an action potential train during 210 pA current injection. (B-D) Plots of averaged action potential properties including action potential amplitude (B), half-width (C), fAHP amplitude (D), frequency (E) and instantaneous frequency (F).

Figure 6. cAMP reproduces effects of Fostriecin on action potential amplitude and half-width in KO neurons

(A) Example action potential waveforms recorded from WT and β4 KO neurons during 210 pA current injection. The plots represent 10^th action potentials from action potential trains of control (un-treated) neurons (solid gray line) and from neurons treated with 1 mM intracellular cAMP (solid dark line). Co-application of BK channel specific blocker 5 μM Paxilline with cAMP (dashed line) is also shown. Scale bars are 20 mV (for voltage axis) and 2 ms (for time axis). (B–D) Plots show averaged action potential properties including action potential amplitude (B), half-width (C) and fAHP amplitude (D). Data are shown for the 10^th action potential during a 210 pA current injection. Taken from left, first columns represent control, second (middle) columns are for cAMP in the pipette (intracellular), third columns are for cAMP (pipette) and Paxilline in the bath solution.