Inhibition of MEK-ERK Signaling Reduces Seizures in Two Mouse Models of Tuberous Sclerosis Complex

Abstract

Tuberous sclerosis complex (TSC) is a monogenic disorder characterized by hyperactivation of the mTOR signaling pathway and developmental brain malformations leading to intractable epilepsy. Although treatment with the recently approved mTOR inhibitor, everolimus, results in clinically relevant seizure suppression in up to 40% of TSC patients, seizures remain uncontrolled in a large number of cases, underscoring the need to identify novel treatment targets. The MEK-ERK signaling pathway has been found to be aberrantly activated in TSC and inhibition of MEK-ERK activity independently of mTOR rescued neuronal dendrite overgrowth in mice modeling TSC neuropathology. Here, we evaluated the efficacy of MEK-ERK inhibition on seizures in two mouse models of TSC. We found that treatment with the MEK inhibitor PD0325901 (mirdametinib) significantly reduced seizure activity in both TSC mouse models. These findings support inhibiting MEK-ERK activity as a potential alternative strategy to treat seizures in TSC.

1. INTRODUCTION

Tuberous Sclerosis Complex (TSC) is a neurodevelopmental disorder affecting 1/6,000 live births (1). TSC is caused by mutations in the TSC1 or TSC2 genes, leading to hyperactivation of the mechanistic target of rapamycin (mTOR) and focal malformations of cortical development known as cortical tubers (2, 3). Approximately 80–90% of individuals with TSC develop epilepsy and more than 60% of these cases are resistant to anti-seizure medications (4). One treatment option for refractory seizures is epilepsy surgery, but even following such an invasive procedure, only 50–60% of the patients will become seizure-free (5–7). Recently, the mTOR inhibitor, everolimus, was approved by the U.S. Food and Drug Administration as adjunctive therapy for TSC-related seizures. Everolimus had a response rate (defined as >50% reduction in seizure frequency) of 40% with a median seizure reduction of 40% in clinical trials, although it led to grade 3/4 adverse events in 40.2% of the patients (8). Improving epilepsy treatment for TSC therefore remains a critical need.

Aberrant upregulation of the extracellular signal-regulated kinase (ERK) signaling pathway has been identified in cortical tubers and slow-growing astrocytomas in patients with TSC (9–13) and conditional Tsc1 knockout mice (14, 15). ERK belongs to the mitogen-activated protein kinase (MAPK) family that is phosphorylated and activated by the MAPK/ERK kinase (MEK) in response to extracellular stimuli (16). The MEK-ERK signaling pathway serves as a critical regulator of cell growth and development and controls diverse functions in the brain, including neurogenesis and neural progenitor cell fate, neuronal maturation, synaptic plasticity, learning and memory formation, and neuroinflammation (17–20). Notably, upregulation of MEK-ERK activity by brain-specific expression of a constitutive active MEK1 results in spontaneous seizure activity in mice (21). Considering these findings altogether, it is conceivable that increased MEK-ERK signaling plays a role in the formation of TSC brain pathology and epilepsy. Consistent with this notion, we previously showed that inhibiting MEK-ERK activity via genetic expression of a dominant-negative MEK1 or treatment with the MEK inhibitor PD0325901 (mirdametinib) corrected abnormal dendritic growth and patterning in Tsc1 null neurons (15). Interestingly, the morphological rescue occurred independently of mTOR activity. Thus, the MEK-ERK pathway may be a plausible treatment target for TSC. Here, we examined whether MEK-ERK inhibition with PD0325901 treatment suppresses seizures using two mouse models of TSC.

2. MATERIALS AND METHODS

2.1. Animals

All procedures were performed in accordance with Yale University and PsychoGenics Institutional Animal Care and Use Committee regulations. Both male and female mice were included. Tsc1^flox/flox; human GFAP-Cre conditional knockout (Tsc1^hGFAP cKO) mice were bred as previously described (22). Mice expressing constitutive active Rheb (Rheb^CA) were generated by in utero electroporation (IUE) in CD-1 mice (Charles River Laboratories) as previously described (23). Briefly, a plasmid solution consisting of pCAG-Rheb^CA [also known as pCAG-Rheb^S16H (24), 2.5–3.5 μg/μl] + pCAG-GFP or pCAG-tdTomato (Addgene #11150 or #83029, respectively; 1.0 μg/μl) was injected into the right lateral ventricle of each embryo at embryonic day (E) 15.5. A tweezer-type electrode was positioned on the embryo head and six 42V, 50 ms pulses were applied using a pulse generator (ECM830, BTX) to electroporate the plasmids into the neural progenitor cells. Expression was targeted to layer 2/3 in the medial prefrontal cortex.

2.2. Overview of Experimental paradigm

Tsc1^hGFAP cKO mice were randomly split into 3 groups that received either vehicle, 1.5 mg/kg PD0325901, or 6 mg/kg PD0325901 once daily from postnatal day (P) 21 to P54–55. Mice were implanted with EEG electrodes at P20–23 and monitored with continuous (24/7) video-EEG recording from P35 to P54–55 (totaling 34–35 days of treatment, 19–20 days of recording). Brain tissue was collected at P54–55 for western blot analysis. Rheb^CA mice were implanted with EEG electrodes between 2–4 months of age and monitored with video-EEG recording for 5 days to establish pre-treatment baseline activity. Mice were then randomly split into 2 groups that received either vehicle or 5 mg/kg PD0325901 once daily for 10 days while undergoing video-EEG recording (totaling 10 days of treatment, 15 days of recording). Only mice that survived through the end of the experiment with complete EEG recording were included in the data analysis.

2.3. Drug treatment

For Tsc1^hGFAP cKO mice, PD0325901 (MedChem Express) was dissolved in 0.5% methylcellulose and 0.2% Tween 80 and administered daily via oral gavage (p.o.). For Rheb^CA mice, PD0325901 (SelleckChem) was dissolved in DMSO, 0.5% hydroxypropyl methylcellulose, and 0.2% Tween 80 and administered daily via intraperitoneal (i.p.) injections. Both administration routes resulted in decreased phospho-ERK1/2 (p-ERK1/2) levels in the brain.

2.4. Video-EEG recording and analysis

For EEG electrode implantation, mice were anesthetized with 2.5–5% isoflurane and positioned on a stereotaxic frame using ear bars. A rostro-caudal midline incision was made in the skin to expose the skull surface. Four pilot holes (two bilateral holes 1 mm anterior to bregma and two bilateral holes 5 mm posterior to bregma, each 1.5 mm lateral to the sagittal suture) were tapped through the skull to the dura mater using a 23-gauge needle. A prefabricated EEG headmount (Pinnacle Technology, Inc. #8201-EEG, 2-channel EEG with common reference, bipolar montage) was attached on top of the skull with superglue and four stainless steel screws (Pinnacle Technology, cat. no. 8209) were threaded into the pilot holes. Silver conductive paint was applied around the screw threads to ensure a solid connection with the headmount. The entire implant was insulated using dental acrylic. For pre- and post-operative care, mice were administered 5 mg/kg rimadyl and 0.1 mg/kg buprenorphine (i.p. injection) 15–30 minutes pre-surgery and treated with 5 mg/kg rimdayl once every 24 hours for 48 hours post-surgery.

Mice were allowed to recover in their home cages for at least 6 days before continuous video-EEG recording. During recording, mice were housed in individual cages with ad libitum access to food and water in light- and temperature-controlled rooms. Synchronous vEEG recording was acquired using a tethered three-channel EEG system (Pinnacle Technology, Inc. #8200-K1-iSE3) and Sirenia Acquisition software (Pinnacle Technology, Inc.). Seizures were manually scored using Seizure software (Pinnacle Technology, Inc.) and EDF browser. All data were analyzed blinded to experimental groups. Electrographic seizures were identified by their characteristic pattern of discrete periods of rhythmic spike discharges that evolved in frequency and amplitude lasting ≥10 seconds (22–24). Video data were inspected for myoclonic jerks, tonic-clonic activities, and convulsions. For each Tsc1^hGFAP cKO mouse, the average daily seizure frequency was calculated by dividing the total number of seizures during the recording period by the total number of recording days (i.e., 19–20 consecutive days). For each Rheb^CA mouse, the average daily seizure frequency during baseline and the last 5 days of treatment was calculated by dividing the total number of seizures during the respective recording periods by the total number of recording days (5 consecutive days for baseline, 5 consecutive days for treatment). Data are shown as the number of seizures per 24 hours.

2.5. Western blotting

Whole brain tissue from Tsc1^hGFAP cKO mice and microdissected cortices containing the electroporated region from Rheb^CA mice were rapidly dissected in ice-cold phosphate buffered saline (PBS; 137 mM NaCl, 2.7 mM KCl, 4.3 mM Na₂HPO₄, 1.47 mM KH₂PO₄, pH 7.4) and rapidly frozen in dry ice. All samples were stored at −80°C until used. The collected tissue was homogenized in 100 mM Tris-HCl (pH 7.4), 0.32 M sucrose, 1 mM EDTA, 5 mM HEPES, and a cocktail of protease and phosphatase inhibitors. Protein samples were resolved by SDS-PAGE and transferred onto PVDF membranes. Membranes were blocked in 5% nonfat milk, 1 mM Na₃VO₄, and PBS+0.1% Tween 20 for 1 hour. Membranes were incubated in primary antibodies [p-ERK1/2 (1:2000, Cell Signaling Technology, CST #4370), ERK1/2 (1:2000, CST #4695), GAPDH (1:5000, CST #5174), β-tubulin (1:50–0, CST #2146)] overnight at 4°C and in secondary antibody [HRP-linked anti-Rabbit IgG (1:5000, CST #7074)] for 1 hour at room temperature. Immunoreactive bands were developed with enhanced chemiluminescence reagents and captured on autoradiography films that were digitized using an image scanner (Fig. 1) or using a digital imaging system (Fig. 2). Optical densities of immunoreactive bands were measured using Image J software (National Institutes of Health). Each band was normalized to the GAPDH or β-tubulin levels within the same lane for loading control. p-ERK1/2 levels were subsequently normalized to total ERK1/2 levels. Data are expressed as percent of the vehicle-treated group.

Figure 1: PD0325901 treatment decreases seizures in Tsc1^hGFAP cKO mice.

(A). Schematic of the experimental paradigm. Tsc1^hGFAP cKO mice were treated with vehicle, 1.5 mg/kg PD0325901, or 6 mg/kg PD0325901 once daily from P21 to P54–55 (treatment days D1 to D34–35). Mice were monitored with continuous video-EEG recording starting at P35 (D15). Brain tissue was collected at the end of the experiment on P54–55 (D34–35) for western blot analysis. (B) Schematic of the EEG montage. (C) Representative EEG trace showing electrographic seizure activity in a Tsc1^hGFAP cKO mouse. (D) Quantification of seizure frequency (average daily seizures) in vehicle- and PD0325901-treated Tsc1^hGFAP cKO mice. *p<0.05 by Kruskal-Wallis test with Dunn’s post-hoc test, n=12 (vehicle), 12 (1.5 mg/kg PD0325901), and 15 (6 mg/kg PD0325901) mice. (E) Animal bodyweight over the course of treatment. Top graph shows the weights of female mice and bottom graph shows the weights of male mice. *p<0.05 by mixed-effects analysis with Tukey’s post-hoc test, n=8 female/4 male (vehicle), 6 female/6 male (1.5 mg/kg PD0325901), and 8 female/7 male (6 mg/kg PD0325901) mice. (F) Representative western blots showing p-ERK1/2, ERK1/2, and GAPDH levels in whole brain lysates from vehicle- and PD0325901-treated Tsc1^hGFAPCKO mice. (G) Western blot quantification of p-ERK1/2/ERK1/2 levels (normalized to GAPDH). *p<0.05, ***p<0.001 by one-way ANOVA with Tukey’s post-hoc test, n=7 (vehicle), 12 (1.5 mg/kg PD0325901), and 14 (6 mg/kg PD0325901) mice. (H) Scatterplot of seizures/24 hours vs. p-ERK1/2/ERK1/2 levels. r=0.3820, p=0.0283 by Spearman correlation, n=7 (vehicle), 12 (1.5 mg/kg PD0325901), and 14 (6 mg/kg PD0325901) mice. Error bars are ±SEM.

Figure 2: PD0325901 treatment decreases seizures in Rheb^CA mice.

(A) Schematic of the experimental paradigm. Rheb^CA mice were generated by in utero electroporation (IUE) at embryonic day (E) 15.5. Mice at 2–4 months of age were monitored with continuous video-EEG recording for 5 days to establish baseline activity. Mice were then treated with either vehicle or 5 mg/kg PD0325901 once daily for 10 days while undergoing video-EEG recording. B) Schematic of the EEG montage. (C) Representative EEG trace showing electrographic seizure activity in a Rheb^CA mouse. (D) Quantification of seizure frequency (average daily seizures) in vehicle- and PD0325901-treated Rheb^CA mice. The graphs show pairwise comparisons between baseline (days D1–5) and treatment (D11–15) within individual animals. Group means are indicated by the bars. *p=0.0059 by Wilcoxon matched-pairs signed-rank test, n=11 (vehicle) and 11 (PD0325901) mice. E) Representative western blots showing p-ERK1/2, ERK1/2, and β-tubulin levels within microdissected cortices containing the electroporated region from vehicle- and PD0325901-treated Rheb^CA mice. (F) Western blot quantification of p-ERK1/2/ERK1/2 levels (normalized to β-tubulin). *p=0.0440 by unpaired t-test, n=5 (vehicle) and 3 (PD0325901) mice. Error bars are ±SEM.

2.6. Statistical analysis

All statistical analyses were performed using GraphPad Prism 9 software. Group differences were assessed by parametric and nonparametric tests as appropriate. The specific tests are indicated in the results and figure legends. Sample size (n) refers to the number of animals. The significance level was set at p< 0.05. All data are presented as mean ± standard error of the mean (SEM).

3. RESULTS

3.1. PD0325901 treatment decreases seizures in Tsc1^hGFAP cKO mice

Tsc1^hGFAP cKO mice with global inactivation of the Tsc1 gene in neurons and glial cells develop progressive epilepsy by 1 month of age and die prematurely between 2–4 months of age (25, 26). Most of the observed seizures are characterized by forelimb clonus and tonic stiffening of the trunk or extremities, and occasionally, loss of upright posture (25, 26). To examine the effects of PD0325901 treatment on seizure frequency, we treated Tsc1^hGFAP cKO mice with vehicle or PD0325901 starting at P21 for 34–35 days. Mice were monitored for seizure activity by continuous video-EEG recording during the last 19–20 days of treatment (Fig. 1A–C). Vehicle-treated Tsc1^hGFAP cKO mice had an average seizure frequency of 0.89 seizures/24 hours (Fig. 1D). PD0325901 treatment at a dose of 1.5 mg/kg bodyweight/day did not significantly change the seizure frequency (mean of 0.56 seizures/24 hours), whereas PD0325901 treatment at a dose of 6 mg/kg significantly reduced the seizure frequency (mean of 0.14 seizures/24 hours) compared to vehicle-treated mice [p=0.0300 by Kruskal-Wallis test; Dunn’s post-hoc test: p<0.05 (vehicle vs. 6 mg/kg PD0325901), n=12 (vehicle), 12 (1.5 mg/kg PD0325901), 15 (6 mg/kg PD0325901) mice, Fig. 1D]. No sex effects on the treatment response were observed across the groups [sex × treatment effect: p=0.2638 by two-way ANOVA, n=8 female/4 males (vehicle), 6 female/6 male (1.5 mg/kg PD0325901), 8 female/7 male (6 mg/kg PD0325901) mice].

Tsc1^hGFAP cKO mice experience bodyweight loss by 7 weeks of age (26). PD0325901 treatment at 1.5 mg/kg and 6 mg/kg did not affect the weight of female mice during the duration of the treatment, suggesting that neither treatment doses improved or exacerbated their weight loss [time × treatment effect: p=0.0594, time effect: p<0.0001, treatment effect: p=0.6242 by mixed-effects analysis, n=8 (vehicle), 6 (1.5 mg/kg PD0325901), 8 (6 mg/kg PD0325901) mice, Fig. 1E–top]. In comparison, the mean weight of male mice was significantly improved at 24 days of treatment (by 6 weeks of age) in both PD0325901-treated groups compared to the vehicle-treated group [time × treatment effect: p=0.0006, time effect: p<0.0001, treatment effect: p=0.1787 by mixed-effects analysis; Tukey’s post-hoc test: p<0.05 (vehicle vs. 1.5 mg/kg PD0325901), p<0.05 (vehicle vs. 6 mg/kg PD0325901) at day 24, n=4 (vehicle), 6 (1.5 mg/kg PD0325901), 7 (6 mg/kg PD0325901) mice, Fig. 1E–bottom].

To examine the extent of MEK-ERK pathway inhibition following PD0325901 treatment in Tsc1^hGFAP cKO mice, we collected whole brain tissue at the end of the study and performed western blotting to determine the protein levels of p-ERK1/2 and ERK1/2. Since active MEK phosphorylates ERK1/2, the ratio of p-ERK1/2 to total ERK1/2 levels was used as a readout of MEK-ERK activity. Consistent with the seizure data, PD0325901 treatment at 6 mg/kg but not 1.5 mg/kg significantly decreased the p-ERK1/2/ERK1/2 ratio compared to vehicle-treated Tsc1^hGFAP cKO mice [p=0.0002 by one-way ANOVA; Tukey’s post-hoc test: p<0.001 (vehicle vs. 6 mg/kg PD0325901), p<0.05 (1.5 vs. 6 mg/kg PD0325901), n=7 (vehicle), 12 (1.5 mg/kg PD0325901), 14 (6 mg/kg PD0325901) mice, Fig. 1F, G]. Further, plotting the number of seizures as a function of the p-ERK1/2/ERK1/2 ratio revealed a significant positive correlation between seizure frequency and MEK-ERK activity levels [r=0.3820, p=0.0283 by Spearman correlation, n=7 (vehicle), 12 (1.5 mg/kg PD0325901), 14 (6 mg/kg PD0325901) mice, Fig. 1H]. Collectively, these data support that PD0325901 treatment decreases seizure activity in a dose-dependent manner in Tsc1^hGFAP cKO mice.

3.2. PD0325901 treatment decreases seizures in Rheb^CA mice

The positive data in Tsc1^hGFAP cKO mice led us to test the efficacy of PD0325901 in another mouse model of TSC that recapitulates the focal nature of cortical tubers. In this model, focal expression of Rheb^CA in developing neurons by IUE results in mTOR hyperactivation and typical neurological features of TSC, including neuronal hypertrophy, ectopic neuron placement, and seizures (23, 24). We first monitored 2–4 months old Rheb^CA mice with continuous video-EEG recording for 5 days to establish baseline seizure activity. Mice were then treated with vehicle or 5 mg/kg PD0325901 once daily for the following 10 days, and the treatment effects on seizures were assessed during the last 5 days of treatment (Fig. 2A–C). All Rheb^CA mice in this study displayed electrographic seizures that varied from 0.2 to 15.2 seizures/24 hours, except for one mouse in the drug-treated group which had 50.8 seizures/24 hours, Fig. 2D). We previously reported that such variability in the seizure frequency does not necessarily result from differences in the number of electroporated cells (23). Rather, this may occur from a combination of other parameters, including (but not limited to) the amount of electroporated plasmids, plasmid expression, and size of the targeted region. There were no significant changes in the seizure frequencies between baseline and treatment in the vehicle-treated group (5.4 vs. 3.9 seizures/24 hours, p=0.3633 by Wilcoxon matched-pairs signed-rank test, n=11 mice, Fig. 2D). In contrast, mice that received PD0325901 treatment had a significant reduction in seizure frequency compared to baseline (11.5 vs. 3.2 seizures/24 hours, p=0.0059 by Wilcoxon matched-pairs signed-rank test, n=11 mice, Fig. 2D). Even after excluding the mouse with high seizure frequency (50.8 seizures/24 hours) from the drug-treated group analysis, the effect of PD0325901 remained significant (p=0.0117 by Wilcoxon matched-pairs signed-rank test). No sex effects on the treatment response were observed in the vehicle-treated (sex × treatment effect: p=0.6037 by two-way repeated measures ANOVA, n=5 females, 6 males) or PD0325901-treated (sex × treatment effect: p=0.2207 by two-way repeated measures ANOVA, n=7 females, 4 males) groups. PD0325901 treatment also significantly reduced the p-ERK1/2/ERK1/2 ratio in Rheb^CA mice [p=0.0440 by unpaired t-test, n=5 (vehicle), 3 (PD0325901) mice, Fig. 2E, F], similar to in the Tsc1^hGFAP cKO mice. Together, these findings support that PD0325901 treatment decreases seizures and MEK-ERK activity in Rheb^CA mice.

4. DISCUSSION AND CONCLUSION

In the present study, we show that inhibiting MEK-ERK activity with PD0325901 treatment effectively reduces seizures in two mouse models of TSC characterized by diffuse and focal cortical malformations, respectively. Although PD0325901 appears to be more effective in the diffuse model, it is difficult to directly compare the drug efficacy between the two models because of differences in the experimental protocols (treatment age and duration, oral gavage vs. i.p. injection) and baseline seizure frequencies (mean of <1 seizure/day for the diffuse model and >5 seizures/day per day in the focal model). However, despite the differences in the experimental approach and the severity of the seizure phenotype, seizure reduction was observed in both models, underscoring the validity of our findings with respect to the efficacy of PD0325901 treatment for seizures in TSC. Here, we did not examine whether blocking MEK-ERK activity would reduce seizures in other epilepsy models that may be associated with increased MEK-ERK activity, and this could be a subject of future studies.

Mechanistically, it was previously reported that upregulation of the actin-binding protein filamin A leads to dendritic overgrowth and abnormal patterning in Tsc1 null and Rheb^CA neurons (15, 27). Interestingly, these studies reported that increased filamin A expression occurred in an mTOR-independent and MEK-ERK-dependent manner, and suppression of MEK-ERK signaling with PD0325901 treatment rescued the associated dendritic defects. Furthermore, inhibition of filamin A was shown to reduce seizure frequency in Rheb^CA mice (27). Thus, one possible mechanism through which inhibition of MEK-ERK activity reduces seizures is through modulation of filamin A expression. Inhibiting MEK-ERK activity may also rescue other morphological or molecular abnormalities associated with TSC, and future studies aimed to identify these mechanisms could provide additional insights.

In terms of therapeutics, increasing the dose of PD0325901 to further reduce MEK-ERK activity in the brain may improve seizure suppression. However, higher doses of PD0325901 (≥15 mg, twice daily) can lead to musculoskeletal and neurological adverse events, including gait disturbance, memory impairment, confusion, mental status changes, visual disturbances, and muscular weakness, in humans (28) (29). Other common toxicities include diarrhea, rash, fatigue, and nausea (28) (29). In contrast, lower doses of PD0325901 (2 mg, twice daily) seem to be well tolerated with milder side effects (30). Thus, future pharmacokinetic and pharmacodynamic studies to titrate efficacy against toxicity to determine optimal therapeutic doses would be important. One alternative strategy to increase drug efficacy while maintaining at lower doses is to improve PD0325901 brain penetration by modifying the compound to enhance the brain-to-plasma ratio (31). Another option is to examine whether combination therapy with everolimus has synergistic effects that could be beneficial for epilepsy treatment (32).

In conclusion, our findings show that PD0325901 treatment reduces seizures in two mouse models of TSC and suggest targeting MEK-ERK activity as a feasible therapeutic strategy to treat seizures in TSC.

Highlights.

• Inhibition of MEK-ERK activity with PD0325901 treatment reduces seizures in Tsc1 conditional knockout mice

• PD0325901 treatment decreases phospho-ERK1/2 levels in a dose-dependent manner in Tsc1 conditional knockout mice

• PD0325901 treatment reduces seizures and phospho-ERK1/2 levels in an in utero electroporation-based mouse model of TSC characterized by focal expression of constitutive active Rheb

Abbreviations:

TSC

tuberous sclerosis complex

mTOR

mechanistic target of rapamycin

ERK

extracellular signal-regulated kinase

MAPK

mitogen-activated protein kinase

MEK

MAPK/ERK kinase

Tsc1^hGFAP cKO

Tsc1^flox/flox; human GFAP-Cre conditional knockout

Rheb^CA

constitutive active Rheb

IUE

in utero electroporation

p-ERK1/2

phospho-ERK1/2