FormalPara Key Points

Major depressive disorder (MDD) represents a serious health problem; its neurobiological mechanisms are still being elucidated.

Preclinical data suggest antidepressant potential for modulators of KCNQ voltage-gated potassium (K+) channels.

Initial clinical trials of patients with MDD have shown significant efficacy for the KCNQ2/3 K+ channel opener ezogabine.

Larger randomized controlled trials of KCNQ modulators in depressive disorders are warranted.

1 Need for Mechanistically Novel Antidepressant Agents

Major depressive disorder (MDD) is a chronic and debilitating condition characterized by social and occupational impairment. Major depressive disorder is among the most common disorders, affecting more than 264 million worldwide [1]. Approximately 20% of adults in the USA may experience a major depressive episode at some point in their lifetime [2], while in Europe, recent data suggest a 6.4% overall prevalence of current depressive disorder [3]. Major depressive disorder is a heterogenous disorder with symptoms that vary considerably from patient to patient. Impeding novel treatment discovery, the field lacks a complete understanding of the essential neurobiological mechanisms that lead to depression [4, 5]. One of the first theories regarding the pathophysiology of MDD was the monoamine hypothesis of depression developed approximately 50 years ago. The monoamine hypothesis suggests that depression is associated with lower levels of serotonin, norepinephrine, and dopamine. First-generation antidepressants, including tricyclic antidepressants and monoamine oxidase inhibitors, and second-generation drugs (selective serotonin reuptake inhibitors, serotonin-norepinephrine reuptake inhibitors, atypical antidepressants, and serotonin modulators) target largely the monoamine neurotransmitter system and increase the levels of these signaling molecules in the synaptic space. Subsequent work suggests that an imbalance of the monoaminergic transmission is not a condition sufficient to regulate mood, indicating that monoamines may primarily have a modulatory effect on other neurobiological pathways associated with MDD [6, 7]. Moreover, the latency of response to antidepressant medications (2–4 weeks or longer) and the percentage of patients with MDD who do not respond to treatment (approximately one-third [8]) provide additional support to the notion that the monoaminergic theory of depression is incomplete in describing the neurobiological alterations contributing to the disorder. More recently, clinical studies have identified a rapid-onset antidepressant effect of the glutamate N-methyl-d-aspartate receptor antagonist ketamine, which shows improvement in symptoms of depression within hours [9,10,11]. Since ketamine was first tested as a treatment for depression, scientific investigations have examined the role of the glutamate system in the neurobiology of MDD and its potential as a target for novel therapeutics, paving the way for the possibility of more direct pharmacological targets [12, 13]. In 2019, the US Food and Drug Administration (FDA) approved intranasal esketamine spray as an adjunct to standard antidepressants for treatment-resistant depression. Despite this important advance in treatment, the role of the glutamate system in the pathophysiology of depression is not fully characterized.

In addition to incomplete understanding of the neurobiology of MDD, relatively little is known about the biological or psychological mechanisms of resilience, the ability to maintain a state of mental health well-being in the face of stress or adversity [14,15,16]. A growing body of preclinical evidence suggests that stress-resilient animals develop adaptive regulation of more genes and ion channel functions in the brain as compared with stress-susceptible depressive animals, supporting that resilience is an active stress-coping mechanism, rather than a simple lack of pathological alterations in response to prolonged adverse stimuli [15, 17,18,19,20]. Importantly, it has become clear that enhancement of active resilient mechanisms is able to reverse depressive dysfunctions in the brain and achieve antidepressant efficacy in both rodents and human patients [20,21,22], which opens a conceptually novel therapeutic strategy for depression treatment (see more below).

2 Molecular Mediators of Resilience to Stress

In addition to human studies, preclinical work allows the identification of the detailed mechanisms underlying MDD that can be then translated into humans, leading to the identification of new targets for drug discovery [23]. Numerous animal models have been established to elucidate the neurobiological mechanism underlying MDD and to test novel antidepressant treatments. A well-validated chronic stress animal model of depression is the chronic social defeat stress (CSDS) model, in which the animal is exposed to repeated and chronic social stress [17, 24]. This model aims to translate animals back into the life events (e.g., loss of social status) that are associated with an increased risk of developing MDD in humans. Chronic social defeat stress induces a depression-like phenotype in a subgroup of mice (susceptible) that is reversed by antidepressant treatment [25,26,27,28]. According to this protocol [17, 24], mice are daily attacked by a larger and aggressive mouse (social defeat). Defeated mice that do not develop a depressed phenotype are considered resilient. The resilient mice show an up-regulation of voltage-gated potassium (K+) channel functions within the ventral tegmental area (VTA) compared with the susceptible and stress-naïve mice (Fig. 1) [17, 20]. The VTA dopaminergic neurons, part of a discrete reward circuit together with the ventral striatum (VS)/nucleus accumbens (NAc) and the ventromedial prefrontal cortex (PFC), play an important role also in mediating stress responses [25, 29]. Ventral tegmental area dopaminergic neurons exhibit two patterns of activity in vivo: low-frequency tonic firing and high-frequency phasic firing [30]. The phasic high-frequency firing of the VTA dopaminergic neurons projecting to the NAc is upregulated by CSDS in the susceptible mice, but not in the resilient mice [31, 32]. Consistently, optogenetic induction of phasic firing of the VTA dopaminergic neurons in this circuit is able to induce a susceptible phenotype in the subthreshold-defeated mice and previously resilient mice. This increase in VTA dopaminergic neuron phasic firing in susceptible mice is induced by an up-regulation of excitatory hyperpolarization-activated cyclic nucleotide-gated (HCN) cation channel-mediated current (Ih) that can be attenuated pharmacologically, which leads to an improvement of the depressive behavior [27]. Long-term treatment with the antidepressant fluoxetine normalizes the hyperactivity (and decreases this excitatory Ih current) in these neurons [27]. Interestingly, the resilient mice show a larger Ih current, in parallel with increased inhibitory K+ channel currents, that stabilizes VTA dopaminergic neurons counterbalancing the hyperactivity of these neurons observed in the susceptible mice (Fig. 1) [20]. Among the upregulated channels in the VTA of resilient mice, a prominent role is displayed by KCNQ3 (Kv7.3), a slow voltage-activated K+ channel [17, 20]. The KCNQ (or KV7) family of K+ channels, which includes KCNQ1-5, are important regulators of neuronal excitability and represent targets for drug discovery for central nervous system (CNS) disorders [33,34,35].

Fig. 1
figure 1

Regulatory function of KCNQ potassium (K+) channels in the ventral tegmental area (VTA) dopaminergic neurons innervating nucleus accumbens (NAc) neurons. A There is baseline expression of hyperpolarization-activated cyclic nucleotide-gated cation (HCN) channels and KCNQ potassium K+ channels in VTA-NAc dopamine neurons in stress-naïve control animals, which contribute to the control levels of Ih and Ik currents and the baseline firing activity of these circuit neurons. The baseline level of these channels is indicated by four HCN and KCNQ channels, respectively on the VTA neuron. B Susceptible subgroup shows increased Ih current without alteration in HCN channel expression, possibly through the phosphorylation by PKC as illustrated in the VTA neuron. The increased Ih current drives the hyper firing activity of VTA-NAc dopamine neurons with no alteration in KCNQ expression, thus no changes in Ik current. C The resilient phenotype displays an even bigger Ih current that concurred with an increased Ik current (illustrated by more KCNQ channels), which maintains the control level of firing activity of VTA-NAc dopamine neurons. The bigger Ih current is possibly induced by stronger Protein kinase C (PKC) phosphorylation (thick arrow). These observations demonstrate a regulatory effect of the KCNQ K+ channel on the firing of these dopamine neurons. D In the susceptible animals, which show various depression-related behaviors, over-expression of KCNQ channel (illustrated by more KCNQ channels) or administration of KCNQ channel openers inhibit the pathophysiological hyperactivity of these neurons and has antidepressant efficacy at a behavioral level

Full size image

Overall, these data provide a rationale for exploring K+ channels, especially the KCNQ subtype, as a target for drug discovery for MDD. Here, we summarize the evidence from animal and human studies on drugs targeting K+ channels as potential novel treatments for MDD. However, as this evidence has been tested primarily in animal models of depression subsequent to CSDS, further studies using other stress-inducing paradigms are warranted to test the antidepressant properties of KCNQ channels.

3 General Structure and Function of KCNQ Potassium Channels

Electrical signaling in the neurons occurs thanks to the generation and propagation of action potentials, a consequence of the diffusion of ions across cell membranes through ion channels. Potassium (K+) channels represent a class of ion channels and KCNQ genes encode members of the Kv7 family of K+ channel subunits [36,37,38]. Although voltage-gated K+ channels (KV) are present throughout multiple systems in the body [39], four members of this family, Kv7.2–Kv7.5, are expressed in the CNS in which they contribute to the regulation of neural excitability and modulation of action potential firing frequency [40]. K+ channels are formed by four transmembrane, pore-forming, voltage-sensing α-subunits each consisting of six segments. Four of these segments have a voltage-sensing component and modify the structure of the channel in response to changes in cell membrane potential (voltage gating). These segments surround two additional segments that form an ionic conductance pore. Based on the characteristic of the α-subunits, voltage-gated K+ channels are described as homotertamers (four identical α-subunits), or heteroteramers (two or more distinct α-subunits; as in the case of KV 7.2/KV 7.3—KCNQ2/3). K+ channels are then completed by additional cytoplasmic and transmembrane β-subunits able to modify α-subunit localization and function [41, 42]. KV channels are divided into 12 subfamilies: KV1–KV12. Among these, only a few serve as drug targets for CNS disorders such as epilepsy (KV7) in which KV7 (KCNQ) channels trigger the voltage-dependent K+ current (M-current) that regulates neuronal excitability and action potential firing frequency [43, 44]. Recent work also supports an additional physiological function of KCNQ channels aiming at compensating excessively excited neurons, which is evidenced in hippocampus cells [45] and VTA dopamine neurons [21].

4 KCNQ Potassium Channel as a Target for Treatments for CNS Disorders

The KCNQ family of K+ channels, and specifically KCNQ2 and/or KCNQ3 has been studied for the treatment of epilepsy. Within the members of the KNCQ family, mainly channels from KCNQ2 to KCNQ5 are located in the brain. Specifically, mutations of the KCNQ2 and KCNQ3 genes have been associated with the development of benign familial neonatal seizures, thus suggesting a role in the pathophysiology of epilepsy that could be used as a drug target for the development of new treatments [46]. Ezogabine, also known as retigabine (N-[2-amino-4-(4-fluorobenzylamino)-phenyl] carbamic acid ethyl ester) is a selective KCNQ2/3 channel opener that exerts its effects by increasing the activity of these channels in the CNS and consequentially leading to an overall decrease in neuronal excitability. Ezogabine was developed and then approved by the FDA (in 2011) as an adjunctive treatment of partial-onset seizures in patients aged 18 years or older. Several trials have shown ezogabine has efficacy for the treatment of partial epilepsy [47,48,49].

Ezogabine has also been shown to reduce membrane excitability in amyotrophic lateral sclerosis [50]. Following ezogabine withdrawal from the market in 2017, owing to limited use and declining numbers of patients initiating drug therapy, new drugs targeting the KCNQ channels have been developed. XEN1101 is a novel positive allosteric modulator of the KCNQ2/3 channels currently under development for the treatment of focal epilepsy [51]. A summary of the drugs targeting the KCNQ channels is included in Table 1.

Table 1 Drugs targeting the KCNQ channels, KCNQ channel specificity and doses per medical condition
Full size table

5 KCNQ Potassium Channel in Animal Models of Depression

As stated above, KCNQ channels play an active role in counterbalancing pathological hyperactivity of VTA dopaminergic neurons in mice resilient to CSDS [20], indicating that KCNQ is able to functionally regulate the firing activity of these neurons. Consistent with this, bath application of the KCNQ opener ezogabine to in vitro VTA slice preparation suppressed the hyper firing activity of VTA dopamine neurons in susceptible VTA of susceptible mice [21]. Moreover, local infusion of ezogabine into the VTA and viral overexpression of KCNQ3, a subunit that is upregulated in the resilient mice, both induce antidepressant effects in susceptible mice (Fig. 1). Further in vivo studies show that systemic administration of ezogabine has similar antidepressant efficacy in susceptible mice, indicating a translational potential of this KCNQ opener [21]. Another study using fasudil as KCNQ channel opener shows that fasudil downregulates VTA dopamine neuron firing and induces antidepressant effect in the CSDS model. Interesting, these efficacies are mainly mediated by KCNQ4 in the VTA because the regulatory effects on the dopamine neuron activity and depressive behaviors are absent in KCNQ knockout mice [52]. A more recent study with a novel KCNQ opener, Lu [53] AA41178, demonstrate an antidepressant-like effect in the forced swim test. These studies suggest that KCNQ openers may have different selectivity on KCNQ subunits, but behaviorally induce consistent antidepressant efficacy in the stress-related models of depression.

6 Evidence of Dysregulation within the Reward System in MDD

Anhedonia (loss of interest or pleasure) is a core symptom of MDD [54], and convergent lines of research link anhedonia to quantitative deficits in reward processing [29]. Results from behavioral studies have shown that MDD is associated with a reduced response to reward [55], which correlates with anhedonia severity [56] and illness chronicity [57]. Notably, pre-treatment reward deficits and anhedonia are related to a poor antidepressant response [57, 58] and a long-term antidepressant response [59] and appear to persist after remission [60]. Interestingly, although conventional antidepressants are known to regulate resilience-related mechanisms [61,62,63], there is no study exploring the effect of FDA-approved antidepressants on KCNQ channels. In humans, the use of neuroimaging with functional MRI (fMRI) has shown the potential to identify the neurobiological underpinnings of MDD, thus potentially expediting the development of novel treatments [64]. Brain functional data from resting-state and task-based fMRI studies suggested that MDD is characterized by a dysfunction within the reward circuitry, in which subjects with MDD show a blunted neural response to reward compared with healthy volunteers at the level of fronto-striatal functioning and connectivity [65]. The reduced response to reward stimuli in MDD appears to involve both the reward anticipation and feedback, and is characterized by striatal hypoactivation [66]. Further, alterations in reward processing have been associated with an increased risk for depression [67] with a recent study suggesting that increased connectivity of the VS predicts MDD [68], in accordance with animal studies that highlight that role of the reward circuit in the pathophysiology of MDD [29].

7 Human Studies on KCNQ Potassium Channel for Depression

The first study evaluating the antidepressant effect of the KCNQ2/3 positive modulator ezogabine in subjects with MDD was an open-label study enrolling 18 patients with depression who received ezogabine up to 900 mg/day orally over the course of 10 weeks [22]. The 900-mg daily dosage was selected based on evidence showing that ezogabine reaches adequate brain concentrations and has demonstrated efficacy for partial-onset seizures at dosages between 600 mg/day and 1200 mg/day [69]. This study assessed both depressive and anhedonic clinical symptoms and reward circuitry change following treatment with ezogabine. After treatment, a significant reduction in depressive (Montgomery–Åsberg Depression Rating Scale mean score change: − 13.7 ± 9.7, p < 0.001) and anhedonic symptom severity (Snaith–Hamilton Pleasure Scale) mean score change: − 6.1 ± 5.3, p < 0.001] was reported. Depressive symptoms decreased significantly from pre-treatment to post-treatment and throughout the study, with a significant improvement starting from week 3 onwards (p < 0.001). On the primary outcome, 44% and 28% of the subjects satisfied criteria for response and remission, respectively. Moreover, the improvement in depression and anhedonia was associated with decreased functional connectivity between the ventral caudate (a key component of the VS that receives dopaminergic innervation from the midbrain, an area affected by KCNQ channel modulation), and clusters within the mid-cingulate cortex and posterior cingulate cortex. In addition, data from a subgroup of these participants (n = 9) from who reward learning data from a computer-based task pre-treatment and post-treatment were available showed increased reward learning following treatment. The drug was overall well tolerated and no increase in suicidal ideation or emergence of suicidal behavior was reported during the trial. The most common adverse events reported were dizziness, confusion, and headache and only three participants did not achieve the targeted dose of 900 mg/day, but were retained in the study at the lower dosages of 750 mg/day (n = 1) and 600 mg/day (n = 2). Despite the encouraging results, the small sample size and the open-label nature of the study limit the generalizability of these findings.

The potential antidepressant effect of the KCNQ2/3 modulator ezogabine has been tested subsequently in a randomized placebo-controlled clinical trial exploring its effect on clinical symptoms and brain response during reward anticipation [70], a translational measure of target engagement that appears to normalize with antidepressant treatment [71, 72]. In this study, 45 patients with depression were randomly assigned to receive ezogabine (up to 900 mg daily) or placebo in a 1:1 approach and underwent fMRI during a reward task prior and following the treatment period; clinical measures of depression, anhedonia, and global symptom severity were also collected at weekly visits. Ezogabine compared with placebo was associated with a large improvement in depression as measured by the Montgomery–Åsberg Depression Rating Scale (p < 0.001) and the Quick Inventory of Depressive Symptomatology-Self-Report (p = 0.002) and equivalent to a medium-large effect size (differences in the change in means for the Montgomery–Åsberg Depression Rating Scale and Quick Inventory of Depressive Symptomatology-Self-Report were d = 0.76 and d = 0.56, respectively). The ezogabine group showed also a significant improvement in hedonic capacity as measured by the Snaith–Hamilton Pleasure Scale (p < 0.001), the anticipatory and consummatory subscales of the Temporal Experience of Pleasure Scale (p < 0.001 and p = 0.05, respectively), and other anhedonia measures. The ezogabine group showed also a trend towards an increase in VS response to reward anticipation compared with placebo post-treatment compared to baseline (p = 0.07), albeit not reaching statistical significance.

Interestingly, this study had a similar design to a recent randomized controlled trial aiming to test the anti-anhedonic effect of the selective kappa opioid receptor antagonist JNJ-67953964 across a range of mood and anxiety disorders [73], including the use of fMRI during a reward task to assess the effect of treatment on the VS. Specifically, both studies used fMRI activation during a reward task to assess the effect of treatment on the VS as a primary endpoint, while measures of clinical improvement were listed as secondary outcomes. This design is particularly suited for early phase trials of novel interventions for mental disorders that aim to translate basic science into clinical hypotheses with the main purpose of gathering knowledge on the relationship between the underlying processes of mental disorders and the mechanisms of action of the proposed intervention. However, despite the similar methodology, while ezogabine was associated with beneficial effects on several secondary clinical outcomes, it did not show a statistically significant effect on the VS, possibly reflecting a false-negative finding due to the limited sample size. Overall, these initial findings encourage larger randomized controlled trials investigating longer term efficacy or tolerability to establish the potential of the KCNQ2/3 K+ channel as a target for the development of novel medications to treat depression and anhedonia. Further, future studies are also needed to compare the effect of ezogabine on depressive and anhedonic symptoms to current FDA-approved antidepressants.

8 Role of KCNQ channel openers in other psychiatric disorders and other potassium channels under study

KCNQ channel openers have been explored as pharmacological targets for CNS disorders, such as epilepsy. Ezogabine showed efficacy in several animal models of epilepsy and has been marketed as an adjunctive treatment of partial-onset seizures in patients aged 18 years and older [9,10,11]. Currently, other KCNQ channel openers are under development. A novel positive allosteric modulator (XEN1101) of the KCNQ2/3 channels is currently under development for the treatment of focal epilepsy [51], with a mechanism of action similar to ezogabine. A recently concluded phase I clinical trial on XEN1101 by Premoli et al. [74] found an effect of XEN1101 on corticospinal excitability using transcranial magnetic stimulation. The authors enrolled 20 healthy volunteers using a cross-over design and completed transcranial magnetic stimulation‐evoked electroencephalogram potentials prior to and at regular intervals following the administration of the drug. Similar to other antiepileptic drugs, XEN1101 was shown to suppress cortical and corticospinal excitability, thus demonstrating its activity in the CNS. A randomized, double-blind, placebo-controlled, multicenter, phase II clinical trial of XEN1101 is currently underway to evaluate the clinical efficacy, safety, and tolerability of the drug as adjunctive treatment for focal epilepsy in adults (ClinicalTrials.gov number: NCT03796962).

Other studies at the preclinical level have explored the potential of KCNQ channels as a target for the treatment of other psychiatric disorders. Preliminary evidence from animal studies suggests a potential effect of ezogabine on mania, thus with possible use in cases of bipolar disorder. Ezogabine reduced hyperactivity induced by amphetamine and chlordiazepoxide without altering the basal levels of locomotor activity [75]. Other authors replicated this finding and showed that the administration of ezogabine was able to reduce neuronal firing in the VTA, and dopamine release in the NAc [76]. KCNQ channels have been studied also in the context of substance use disorder, including alcohol. Recent evidence suggests that ethanol inhibits KCNQ2/3 channels [77], while ezogabine administration into the NAc significantly reduced voluntary alcohol intake especially in high-drinking rats [78] and systemic treatment with ML213 (KCNQ2 and KCNQ4 opener) also reduced ethanol drinking in rodents [79]. Similarly, in rats exposed to cocaine, methylphenidate, or phencyclidine, the administration of ezogabine inhibited drug-induced locomotor activity and reduced c-Fos expression in the NAc and primary motor cortex, and reduced basal extracellular levels of dopamine metabolites, leading to a reduction in striatal and cortical excitability and thus the CNS effects of these drugs on the animal reward circuit [80]. KCNQ channels have also been investigated in an animal model of cocaine relapse [81]. Researchers showed that cues associated with cocaine were able to trigger drug-seeking behavior in mice, which is mediated by PFC neurons. Among the changes observed in these neurons, a reduction of inhibition mediated by KCNQ channels was reported, indicating the hyperactivity of the PFC neurons. In these animals, administration of ezogabine led to normalized PFC neuronal excitability and reduced drug-seeking behavior. Overall, these data suggest that KCNQ could be an important target for drug development for the treatment for drug and alcohol addiction.

Finally, KCNQ channels appear to regulate PFC connectivity, an area involved in higher cognitive function. Specifically, KCNQ2 channels have been found in medial PFC layers II/III and V pyramidal cells, in which increased KCNQ currents induced by chronic stress exposure have been associated with reduced neuronal firing. Similarly to chronic stress exposure, ezogabine has been shown to increase the open state of KCNQ channels and increase the slow relaxation amplitude of KCNQ channels, leading to a reduction in the firing of medial PFC neurons in animals [82]. These data possibly explain the side effects most commonly associated with ezogabine (confusion and memory impairment), and should also be explored in humans to fully understand the potential of KCNQ channel openers for the treatment of mental disorders.

Of note, other members of the KCNQ family of channels may represent interesting targets for drug development. For example, KCNQ1 channels also appear to be expressed in the CNS [83] and recent evidence suggests a role for KCNQ1 polymorphism in the cognitive deficits and decreased white matter integrity associated with schizophrenia [84].

9 Conclusions

K+ channels appear a promising target for the development of novel therapeutics for the treatment of psychiatric disorders and specifically for depression. The targeting of this channel represents a completely different mechanism of action compared with any currently available antidepressant. Both preclinical and clinical studies provide initial evidence on the antidepressant effect of the K+ channel modulators, especially for KCNQ channel openers in which modulation of these channels can enhance stress resilience in animals and improve depressive and anhedonic symptoms in patients with MDD. Based on these findings, larger randomized controlled trials of KCNQ modulators in depressive disorders are warranted to confirm their potential as a viable treatment for depressive disorders and other stress-related disorders, ultimately leading to the development of a different class of antidepressant medications.