Increased Expression of Fragile X Mental Retardation Protein in Malformative Lesions of Patients with Focal Cortical Dysplasia

Abstract

Objective:

Focal cortical dysplasia (FCD) accounts for nearly half of all cases of medically refractory epilepsy in the pediatric and adult patient populations. This neurological disorder stems from localized malformations in cortical brain tissue due to impaired neuronal proliferation, differentiation, and migration patterns. Recent studies in animal models have highlighted the potential role of the Fragile X mental retardation protein (FMRP) levels in FCD. The purpose of this study was to investigate the status of FMRP activation in cortical brain tissues surgically resected from patients with FCD. In parallel, this study also investigated protein levels within the PI3K/AKT/mTOR and canonical Wnt signaling pathways.

Methods:

Pathologic tissue from malformative lesions of FCD patients with medically refractory epilepsy were compared to relatively normal control non-epileptic tissue from patients with intracranial neoplasms. A series of western blotting assays were performed to assess key proteins in the PI3K/AKT/mTOR, canonical Wnt signaling pathways, and FMRP.

Results:

There was suppression of S235/236-phophorylated S6, GSK3α, and GSK3β protein levels in samples derived from FCD patients, compared to non-epileptic controls. FCD samples also had significantly greater levels of total and S499-phosphorylated FMRP.

Conclusions:

These findings support our hypothesis that malformative lesions associated with FCD are characterized by high levels of FMRP activation along with dysregulation of both PI3K/AKT/mTOR and canonical Wnt signaling. These novel clinical findings extend previous work in animal models, further suggesting a potential unforeseen role of GSK3α, and GSK3β in the pathophysiology of FCD and refractory epilepsy.

INTRODUCTION

Epilepsy is a neurological disorder characterized by recurrent seizures, which affects nearly 45.9 million people globally with an approximate incidence rate of 2.8 million persons per year [3, 10]. Recurrent seizures are associated with comorbid declines in physical, psychiatric, and cognitive health; therefore, prevention of seizure activity is essential [13, 35]. In the clinical setting, anticonvulsant monotherapy effectively manages the majority of epileptic conditions by reducing the likelihood of future seizures [6]. However, roughly one-third of those with epilepsy remain refractory to conventional treatment [37].

Patients with focal cortical dysplasia (FCD) accounts for nearly 50% of all cases of refractory epilepsy in both pediatric and adult populations [17, 38]. FCD causes localized malformations in cortical brain tissue due to impaired neuronal proliferation, differentiation, and migration patterns [1]. These irregularities give rise to the architectural and cytological hallmarks of FCD, including columnar cell disorganization, laminar disruptions, dysmorphic neurons, and balloon cell formation [17]. Although there is considerable heterogeneity in the pathomorphology of FCD, previous research has shown this disorder to be associated with several key perturbations in molecular signaling. The canonical Wnt signaling pathway is hypothesized to play a role in FCD pathology due to its role in maintaining neural stem cell microdomains, cell differentiation and cell migration patterns [21, 29]. Previous work has also characterized the involvement of the phosphoinositide 3-kinase/AKT/mammalian target of rapamycin (PI3K/AKT/mTOR) intracellular signaling pathway in FCD and epilepsy [22]. One molecular signaling protein that has received attention in the epilepsy field, is the Fragile X mental retardation protein (FMRP). FMRP is expressed ubiquitously throughout the brain where it regulates protein synthesis through its association with polyribosomes in a phosphorylation-dependent manner [7, 20]. Previous studies demonstrated that seizures reduce synaptic localization of the FMRP protein [4] as well as the degree of FMRP phosphorylation [33].

Although the exact mechanism is unclear, it has been proposed that because FMRP activity is regulated downstream of mTOR it may provide negative feedback to upstream elements of the PI3K/AKT/mTOR pathway [15, 36]. FMRP may additionally influence Wnt signaling actions via negative feedback inhibition of Glycogen synthase kinase-3 (GSK3) [26]. Given the role of these signaling pathways in FCD pathophysiology, it is possible that FMRP may also be dysregulated. Previous studies using neuron subset-specific Phosphatase and tensin homolog (NS-Pten) KO mice, an animal model of FCD, we observed increased expression of total and S499-phosphorylated FMRP and alterations in Wnt signaling [24, 25, 34]. Despite this pre-clinical characterization, there has been no previous investigation into whether this effect is recapitulated in human patients with FCD. The main objective of this study was to determine the status of FMRP activation in tissues surgically resected from patients with FCD with drug-resistant epilepsy. In addition, because signaling through the PI3K/AKT/mTOR and canonical Wnt pathways regulate FMRP activation, we also investigated key regulatory proteins in these cascades.

METHODS

Brain Tissue Samples

The brain tissue samples used in this experiment were derived from patients undergoing resective brain surgery. Tissue from malformative lesions of patients with medically refractory epilepsy (FCD) (n = 6) were compared to control tissue from patients with intracranial neoplasms, including glioblastoma (n = 3). Patients providing control tissue had no previous medical history of seizures or epileptic syndromes. Tissues were collected following informed consent from the patients under the Institutional Review Board (IRB) protocol # 1011004282 (Development of a Biorepository for Methodist Research Institute). All tissue samples were de-identified prior to western blotting analyses, in accordance with guidelines from the Purdue University Human Research Protection Institutional Review Board (Protocol # 1507016240), The Health Insurance Portability and Accountability Act of 1996 (HIPAA), and The Code of Ethics of the World Medical Association (Declaration of Helsinki) for experiments involving humans.

Western Blotting Analysis

Tissues were processed for protein extraction and dilutions exactly as previously described [39]. Briefly, small tissue blocks (0.5 cm) containing an undetermined amount of white and grey matter from each sample were homogenized in ice-cold homogenization buffer containing 0.32 M sucrose, 1mM EDTA, 5mM HEPES, and a protease inhibitor cocktail (P3840; Sigma, USA). Equivalent protein concentrations were confirmed using the Bradford Protein Assay (Bio Rad, Hercules, CA, U.S.A.), diluted in Laemmli loading buffer (4x: 0.25M Tris, pH 6.8, 6% SDS, 40% glycerol, 0.04% bromophenol Blue, 200 mM dithiothreitol). Sample processing through SDS-PAGE and western blotting were performed as previously described [23]. For immunoblotting, proteins were transferred to Hybond-P polyvinyl difluoride membranes (GE Healthcare, Piscataway, NJ, U.S.A.), blocked for 1 hour at room temperature in blocking solution (5% nonfat milk diluted in Tris buffered saline [50mM Tris-HCl, pH = 7.4, 150mM NaCl] with 0.1% Tween [TTBS] and 1mM Na₃VO₄), and incubated at 4°C overnight with the following primary antibodies in 5% milk in TTBS: total Akt, S473-phosphorylated Akt, total S6, S235/236-phosphorylated S6, S240/244-phosphorylated S6, total FMRP, S499-phosphorylated FMRP, Axin1, GSK3α, GSK3β, LRP6, DVL2, Wnt3A, Wnt5A/B, β-catenin, HCN1, HCN2, and Kv4.2 (See Table S1 for further antibody details). Membranes were then washed in TTBS (3 × 5 min) and incubated with horseradish peroxidase-labeled secondary antibody, anti-rabbit immunoglobulin G (IgG). Immunoreactive bands were visualized using chemiluminescence with GE ECL Prime (GE Healthcare) and scanned using a ProteinSimple Western Blot Imaging System (ProteinSimple, Santa Clara, CA, U.S.A.)

The optical density of each immunoreactive band was measured using the ProteinSimple AlphaView software. Measurements obtained from all bands of interest were standardized to endogenous actin levels. All experimental data points represent a single tissue sample (n = 1). All groups were normalized to the control group average per western blot.

Statistical Analysis

All data were analyzed using SPSS 23.0 (SPSS, Chicago, IL, U.S.A.) or GraphPad Prism 7 (San Diego, CA, U.S.A.). For all comparisons, the level of significance was set at p < 0.05. We used a Student’s t-test for comparison of the FCD and non-epileptic brain tissue groups. All bars represent the mean and error bars represent the standard error of the mean. To keep the results section manageable, only statistically significant effects are included below but all results can be found on Table 1.

Table 1.

Western Blotting Results Summary

Protein	Focal Cortical Dysplasia	Control	Change	Sig.
Total AKT	98.66 ± 2.261, n=6	100 ± 5.397, n=3	—	—
Total Phospho-AKT (S473)	75.98 ± 25.63, n=6	100 ± 26.04, n=3	—	—
% Total Phospho-AKT (S473)	77.58 ± 25.89, n=6	102.1 ± 28.3, n=3	—	—
Total S6	76.66 ± 7.905, n=6	100 ± 19.9, n=3	—	—
Total Phospho-S6 (S240/244)	58.97 ± 10.13, n=6	100 ± 21.14, n=3	—	—
% Total Phospho-S6 (S240/244)	84.31 ± 21.87, n=6	105 ± 19.8, n=3	—	—
Total S6	85.06 ± 2.393, n=6	100 ± 16.69, n=3	—	—
Total Phospho-S6 (S235/236)	69.8 ± 5.452, n=6	100 ± 11.96, n=3	↓	*
% Total Phospho-S6 (S235/236)	82.19 ± 6.187, n=6	107.8 ± 24.05, n=3	—	—
Total FMRP	243.1 ± 33.08, n=6	100 ± 45.59, n=3	↑	*
Total Phospho-FMRP (S499)	337.8 ± 61.08, n=6	100 ± 45.56, n=3	↑	*
% Total Phospho-FMRP (S499)	140.3 ± 21.8, n=6	141.3 ± 64.09, n=3	—	—
Axin1	100.5 ± 18.32, n=6	100 ± 37.06, n=3	—	—
GSK3a	68.37 ± 4.014, n=6	100 ± 4.166, n=3	↓	**
GSK3b	70.99 ± 3.649, n=6	100 ± 3.659, n=3	↓	**
LRP6	75.72 ± 21.36, n=6	100 ± 39.96, n=3	—	—
DVL2	70.72 ± 17.44, n=6	100 ± 15.26, n=3	—	—
Wnt3A	114.1 ± 17.42, n=6	100 ± 4.659, n=3	—	—
Wnt5A/B	76.14 ± 28.57, n=6	100 ± 63.67, n=3	—	—
b-Catenin	46.86 ± 10.86, n=6	100 ± 35.07, n=3	—	—
HCN2	114.6 ± 10.27, n=6	100 ± 8.232, n=3	—	—
Kv4.2	85.94 ± 20.76, n=6	100 ± 23.8, n=3	—	—
HCN1	103 ± 24.01, n=6	100 ± 26.32, n=3	—	—

Note:

^{* =}p < 0.05,

^{** =}p < 0.01

RESULTS

In evaluating FMRP we found several significant differences between the FCD and non-epileptic groups. There were significant increases in the protein levels of both total FMRP (t_1,7 = 2.52, p < 0.05, r = 0.69) and S499-phosphorylated FMRP (t_1,7 = 2.52, p < 0.05, r = 0.69) (Figure 1 & Table 1). Note that these increases were not present in the ratio of S499-phosphorylated/total FMRP expression levels. We also evaluated two ion channels known to be targeted and regulated by FMRP signaling, HCN1 and Kv4.2. The protein levels of these potassium channels were similar in FDC and non-epileptic tissues (p>0.05).

Figure 1. Western Blotting Changes in FCD Brain Tissue.

In our evaluation of FCD brain tissue samples, we found significant increases in levels of total FMRP (p < 0.05, r = 0.69) and S499-phosphorylated FMRP (p < 0.05, r = 0.69), significant reductions in total levels of the S235/236-phosphorylated S6 protein (p < 0.05, r = 0.71), no change in total levels of ribosomal S6 protein, but a decrease in GSK3a (p < 0.01, r = 0.88), and GSK3b (p < 0.01, r = 0.88). * = p < 0.05; ** = p < 0.001. In addition, a representative loading control (Actin) is shown. The bars represent the mean and the error bars represent the standard error of the mean.

In our evaluation of the PI3K/AKT/mTOR signaling pathway, we found a significant reduction in total levels of the S235/236-phosphorylated S6 protein in FCD samples compared to non-epileptic controls (t_1,7 = 2.70, p < 0.05, r = 0.71) (Figure 1 & Table 1). This reduction was not present in total S6, nor the ratio of S235/236-phosphorylated/total S6 expression levels. There were no significant differences in total Akt, total S473-phosphorylated Akt, or the ratio of S473-phosphorylated/total Akt, S240/244-phosphorylated S6, or the ratio of S240/244-phosphorylated/total S6 between the groups (Table 1). In contrast, in our evaluation of the canonical Wnt signaling pathway, we found significant reductions in the protein levels of both GSK3α (t_1,7 = 4.88, p < 0.01, r = 0.88) and GSK3β (t_1,7 = 4.96, p < 0.01, r = 0.88) in the FCD tissues compared to non-epileptic controls (Figure 1 & Table 1). However, we found no differences in LRP6, Wnt3A, Wnt5A/B, Axin1, DVL2, or β-catenin (Table 1).

DISCUSSION

Canonical Wnt signaling is crucial for migration and differentiation of cortical neurons in early brain development [29]. In the inactive state, Wnt signaling maintains low levels of cytoplasmic β-catenin through a cytoplasmic destruction complex. Once this complex sequesters β-catenin, GSK3 and CK1α sequentially phosphorylate the protein until a binding site for a ubiquitin ligase is generated [12, 27]. Extracellular co-binding of Wnt glycoprotein ligands to Frizzled and LRP6 activates the pathway, dissociating the cytoplasmic destruction complex and increasing β-catenin levels. Stabilized β-catenin then translocates to the nucleus to join one or more transcription complexes, where it will actively influence gene expression [27]. In the present study, expression of both GSK3 isoforms were significantly reduced in resected brain tissue of FCD patients. Without a concurrent reduction in Wnt3a, these findings suggest hyperactivation of Wnt signaling and an intrinsic loss of regulatory control over β-catenin. This interpretation is consistent with animal studies demonstrating that Wnt3a-mediated hyperactivation of the canonical pathway leads to premature intermediate progenitor cell (IPC) differentiation at the subventricular-intermediate zone border and cortical malformations [29]. Our lab also recently found that Wnt signaling is hyperactivated in the neuron subset-specific Phosphatase and tensin homolog (NS-Pten) knockout mouse, which is a mouse model of FCD [14]. It is also consistent with a previous clinical investigations showing abnormal large neurons and balloon cells in FCD lesions differentially express GSK3 [8]. Therefore, there are several converging lines of evidence for the role of aberrant Wnt signaling in FCD.

The present study found S235/236-phophorylated S6 to be significantly reduced in resected brain tissue of FCD patients, suggesting reduction in PI3K/AKT/mTOR pathway activity. Increased PI3K/AKT/mTOR signaling plays a substantial role in development of FCD, making these findings somewhat surprising. Recent studies have identified brain somatic mutations activating mTOR signaling as a major genetic cause of focal malformations of cortical development (FMCDs), including focal cortical dysplasia (FCD) and hemimegaloencephaly (HME) [9, 18, 30, 32]. In clinical literature, changes in PI3K/AKT/mTOR pathway protein expression seem to differ between FCD subtypes and select pathologic cell types. For example, patients with FCD Type IIb display overall enhanced mTOR signaling and phosphorylation of the S6 ribosomal protein, but this is limited exclusively to balloon and giant cells [2]. While we detected no differences in Akt between FCD patients and controls, FCD Type IIb patients have previously been found to exhibit elevated Akt and PDK1, both upstream molecules of the mTORC1 signaling cascade [28]. Interestingly, there has been no evidence for increased mTORC1 activity in FCD Type I, highlighting the variability between FCD subtypes [28, 31]. Another possibility is that the reduced S235/236-phophorylated S6 could be due to relative enhancement of mTOR activity in controls. Future studies should seek to extend our findings by evaluating differences between FCD subtypes.

The present study found increased levels of both total and S499-phosphorylated FMRP in resected FCD tissue. It has been shown that 10–20% of individuals with Fragile X Syndrome (FXS), characterized by a congenital loss of FMRP, have a co-occuring epileptic phenotype [5, 11, 16]. Studies using animal models have shown that acute seizures disrupt the levels and localization of FMRP [4, 33], but its role in cortical development has only recently been investigated. Appropriate positioning of the cortical plate requires neurons to undergo a morphological multipolar-to-bipolar transition. La Fata et al. (2014) observed that loss of FMRP in Fmr1 KO mice lead to dysregulation of neural cell adhesion molecules (N-cadherin), causing impaired multipolar-to-bipolar transitioning and abnormal cortical plate positioning [19]. Our results potentially extend these findings, suggesting that FMRP could influence cell morphology and function throughout the lifespan in FCD patients. However, the broader cellular and physiological implications of elevated FMRP expression on epileptic phenotypes in FCD patients remains to be investigated.

Altered FMRP expression in resected tissue from FCD patients is largely in line with previous studies from our lab. Using the NS-Pten KO mouse model, we have repeatedly demonstrated increased levels of total and S499-phosphorylated [24, 25, 34]. In the PI3K/AKT/mTOR pathway, FMRP is believed to function as a downstream, counterregulatory molecule that limits mTOR-mediated protein synthesis. Although a definitive mechanism involved in this feedback process remains elusive, it is hypothesized that FMRP may act on the upstream PI3K enhancer protein (PIKE) [36]. FMRP also plays a pivotal role in canonical Wnt signaling via transcriptional repression of GSK3β [26]. Therefore, increases in total and S499-phosphorylated FMRP by 2.5- and 3.4-fold, respectively, may explain the reductions in S235/236-phophorylated S6 and GSK3 expression.

FMRP is a versatile translational regulator of many synaptic proteins and signaling pathways. There is a growing body of evidence implicating dysregulation of this protein in a multitude of neurological disorders. Taken together, our findings expand upon previous studies by demonstrating a potentially unforeseen role of FMRP in the pathophysiology of patients with FCD and refractory epilepsy. While many studies have focused on the association of FMRP with behavior and its neuromolecular correlates, further research is needed to discern the role of this protein in human disease outside of Fragile X syndrome.