Role of KCNQ potassium channels in stress-induced deficit of working memory

Abstract

The prefrontal cortex (PFC) mediates higher cognition but is impaired by stress exposure when high levels of catecholamines activate calcium-cAMP-protein kinase A (PKA) signaling. The current study examined whether stress and increased cAMP-PKA signaling in rat medial PFC (mPFC) reduce pyramidal cell firing and impair working memory by activating KCNQ potassium channels. KCNQ2 channels were found in mPFC layers II/III and V pyramidal cells, and patch-clamp recordings demonstrated KCNQ currents that were increased by forskolin or by chronic stress exposure, and which were associated with reduced neuronal firing. Low dose of KCNQ blockers infused into rat mPFC improved cognitive performance and prevented acute pharmacological stress-induced deficits. Systemic administration of low doses of KCNQ blocker also improved performance in young and aged rats, but higher doses impaired performance and occasionally induced seizures. Taken together, these data demonstrate that KCNQ channels have powerful influences on mPFC neuronal firing and cognitive function, contributing to stress-induced PFC dysfunction.

Keywords: KCNQ; Prefrontal cortex; Pyramidal neurons; Stress; Working memory; cAMP-PKA.

Graphical abstract

Fig. 1

Expression of KCNQ2 channels in rat mPFC. A. Pyramidal neurons in layers II-VI are immunolabeled against KCNQ2. B and C Note the reactive pyramidal somata and principal apical dendrites (arrows) in layer V (B) and in layers II/III (C). D. Omission of the anti-KCNQ2 primary antibody eliminated all labeling; asterisks mark non-reactive pyramidal neurons. Scale bars, 50 μm (a); 10 μm (b); 20 μm (c and d).

Fig. 2

Identification of KCNQ currents in rat mPFC pyramidal cells. A and B. Voltage-clamp recordings of I_M-KCNQ currents in mPFC pyramidal neurons before bath application of the specific KCNQ channel blocker (XE991, 10 μM, Fig. 2A) and of the specific KCNQ channel opener (Retigabine, 10 μM, Fig. 2B). Each panel shows an example of a mPFC pyramidal neuron filled with an inert fluorescent dye (Alexa 488) during whole-cell recordings and a representative trace of I_M-KCNQ currents in response to a depolarizing voltage step from cell's resting membrane potential (~−63 mV) to a holding potential of −40 mV for 1 s followed by a hyperpolarizing step of 20 mV for 1 s. This protocol elicits a steady state or outward current (I_Out) driven by potassium channels opening, including KCNQ channels (indicated by the broken line) followed by the typical relaxation of KCNQ currents at the onset of the hyperpolarization pulse. These recordings were performed in the presence of Na⁺, Ca²⁺ and HCN blockers. Note the decrease (A) and the increase (B) of KCNQ currents (represented by I_Out) after bath application of XE991 and Retigabine, respectively. Graphs showing the individual and mean values of I_Out (**p < 0.01, ***< 0.001, paired Student's t-test).

Fig. 3

Pharmacological inhibition of KCNQ channels increases the firing rate of mPFC pyramidal cells. Current-clamp recordings of action potentials (APs) in mPFC pyramidal neurons before and after specific KCNQ channel blockers, XE911 (A) and linopirdine (B). A. mPFC pyramidal neuron filled with an inert fluorescent dye (Alexa 488) during whole-cell recordings with a representative trace of APs in response to intracellular depolarizing current (150 pA, 300 ms) before and after XE911 (10 μM). B. Another specific KCNQ channels blocker, linopirdine also increased the number of APs evoked by a depolarizing square pulse (150 pA, 300 ms) Note the significant increase in the APs number and the earlier onset of APs under two different KCNQ blockers. *p < 0.05, **p < 0.01, paired Student's t-test.

Fig. 4

Pharmacological opening of KCNQ decreased firing rate of mPFC pyramidal cells. A, Representative traces of regular and bursting firing type of mPFC pyramidal neuron before and after bath application of retigabine (10 μM) in response to a short pulse (300 ms) of intracellular depolarizing current. Both traces top (regular) and bottom (bursting) were evoked by same level of stimulation (150 pA). The cell filled with an inert fluorescent dye (Alexa 488) during whole-cell recording corresponds to the bursting type of neuron. Note the progressive decrease in the number of APs following retigabine application over time in a regular firing type of neuron B, Representative trace of a regular and bursting firing type of mPFC pyramidal neuron in response to a longer pulse of stimulation (5 s, 250 pA) with an example of the regular firing neuron type filled with an inert fluorescent dye (Alexa 488) during whole-cell recordings. Longer pulse of stimulation confirmed that retigabine-decreasef the firing frequency in both type of neurons, as shown in the bar graph where the number of APs has been normalized to control (pre-drug treatment). “*”p < 0.05, **p < 0.001 and *** < 0.0001, paired Studen's t-test.

Fig. 5

cAMP-PKA signaling increases KCNQ currents and reduces the firing of mPFC pyramidal cells. Voltage-clamp recordings of I_M-KCNQ currents (A) and current-clamp recordings (B) of mPFC pyramidal neurons, filled with an inert fluorescent dye (Alexa 488) during whole-cell recordings, before and after bath application of the cAMP-PKA signaling activator, forskolin (10 μM). Raising cAMP with forskolin increased KCNQ currents and decreased firing in mPFC pyramidal cells. A. Representative trace of KCNQ currents in response to a depolarizing voltage step from cell's resting membrane potential (~−63 mV) to a holding potential of ~−40 mV for 1 s followed by a hyperpolarizing step of 20 mV for 1 s. Note the increase in I_Out amplitude under forskolin quantified in the summary graph. The relaxation amplitude of KCNQ current was also quantified in this set of experiment in response to 3 different hyperpolarizing voltage steps (20, 40 and 50 mV). Note the significant increase in the KCNQ relaxation magnitude at all hyperpolarizing voltage steps tested (**p < 0.01, paired Student's t-test, n = 3) B. Current-clamp recordings of APs in response to intracellular depolarizing current (250 pA, 300 ms) in a regular firing type of mPFC pyramidal neurons showing a significant decrease of the number of APs in presence of forskolin (*p < 0.05, **p < 0.01, paired Student's t-test, n = 7).

Fig. 6

Blocking KCNQ channels with Linopirdine reversed the Forskolin-induced decrease in mPFC pyramidal neuronal firing. Representative current-clamp recordings of regular type (A) and bursting (B) type of firing recorded in gap free mode before and after bath application of forskolin alone and then followed by co-application of forskolin with linopirdine (10 μM each). Note that Forskolin decreases firing in both (regular and bursting) type of mPFC pyramidal neurons. This decrease was reversed by addition of KCNQ channels blocker linopirdine. Histograms showing the number of actions potentials (APs) and the number of inter-spike interval (within 500 ms bin period) before and after bath application of forskolin alone and co-application of forskolin with linopirdine. (***p < 0.001 and **p < 0.01, paired Student t-test).

Fig. 7

Chronic stress exposure increases KCNQ currents in mPFC pyramidal cells A. Representative trace of KCNQ currents in response to a depolarizing voltage step from cell's resting membrane potential (~−63 mV) to a holding potential of ~−40 mV for 1 s followed by a hyperpolarizing step of 20 mV for 1 s in control and in stressed group of animals. Note that chronic stress exposure significantly increased I_Out amplitude (**p < 0.01, unpaired Student's t-test) and the relaxation amplitude of KCNQ currents induced by 20 mV hyperpolarizing step (**p < 0.01, unpaired Student's t-test), and by 30 and 40 mV hyperpolarizing steps (*p < 0.05, unpaired Student's t-test). B and C Current-clamp recordings of APs in response to intracellular depolarizing current (450 pA, 300 ms) in regular-firing (B) and in bursting firing (C) type of mPFC pyramidal neurons in control and stressed group of animals. Graphs showing the effect of stress on the number of action potentials with increasing depolarizing currents. Note that chronic stress exposure reduced firing in the regular-type (**p < 0.01 and *p < 0.05, unpaired Student's t-test) but not in the bursting-type of mPFC pyramidal cells. All recorded cells presented in this figure were filled with an inert fluorescent dye (Alexa 488). Interestingly, the regular firing-type of cells that were affected by chronic stress exposure were located in both (II/III) and (V) cell layers.

Fig. 8

Blockade of KCNQ channels within mPFC improves delayed alternation performance and prevents stress-induced working memory deficits A. Infusion of the KCNQ blocker, linopirdine (0.001 μg/0.5 μL) into the rat mPFC improves working memory performance in young adult rats (**p < 0.02; n = 6). B. Infusion of the PKA inhibitor, Rp-cAMPS, into the rat mPFC blocked the delayed alternation deficits induced by the pharmacological stressor, FG7142, in young adult rats. A dose of Rp-cAMPS was chosen for each rat that had no effect on its own to prevent additive effects (**p < 0.001 compared to vehicle; †p < 0.05 compared to FG7142; n = 5). C. Infusion of the KCNQ blocker, linopirdine, into the rat mPFC blocked the delayed alternation deficits induced by the pharmacological stressor, FG7142, in young adult rats. A dose of linopirdine was chosen for each rat that had no effect on its own to prevent additive effects (**p < 0.001 compared to vehicle; †p < 0.02 compared to FG7142; n = 6).

Fig. 9

Systemic administration of the KCNQ blocker, XE991, has an inverted U dose-response on working memory performance in young adult rats A. Systemic injection of XE991 (0.001–2.5 mg/kg, i.p.) produced an inverted U dose-response curve on performance of the delayed alternation task in young adult rats (n = 12). A low dose was identified for each animal that improved performance. In contrast, the 2.5 mg/kg dose impaired performance. This high dose was chosen due to its enhancing effects on long term memory consolidation in mice (*p < 0.05; **p < 0.001 compared to saline control). B. Examples of XE991 dose-response curves in two young adult rats. Note that the drug appears to be more potent in rat 4 than in rat 5. Both are improved by lower doses and markedly impaired by the 2.5 mg/kg dose. The dashed lines represent the vehicle baselines for each individual rat.

Fig. 10

Low dose KCNQ blockade improves working memory performance in aged rats A. Systemic injection of an optimal low dose of XE991 (0.005–0.5 mg/kg, i.p.; n = 8) significantly improved performance of the delayed alternation task in aged rats (**p < 0.01). B. Example of an XE991 dose-response curve in aged rat 9. Note that the first time 0.1 mg/kg dosage was administered it markedly improved performance (noted by the “1”), but the second time this dose was given, it impaired performance (noted by the “2”).