Interactions between GHRH and GABAARs in the brains of patients with epilepsy and in animal models of epilepsy

Growth hormone releasing hormone (GHRH) has recently been shown to increase the level of γ-aminobutyric acid (GABA) and activate GABA receptors (GABARs) in the cerebral cortex. GABA is an inhibitory neurotransmitter that can inhibit seizures. Does GHRH enhance the inhibitory effect of GABA to prevent epilepsy by increasing the GABA level and activating GABARs? In this study, patients with epilepsy and C57/BL6 mice with epilepsy induced by kainic acid (KA) or pentylenetetrazol (PTZ) served as the research subjects. Western blots were used to observe the differences in GHRH expression between the normal group and the epilepsy group, immunofluorescence was performed to explore the localization of GHRH in the brain, and coimmunoprecipitation was used to observe the interaction between GHRH and GABARs. GHRH expression was significantly increased in both patients with temporal lobe epilepsy (TLE) and in two mouse models induced by KA or PTZ compared with that in the normal groups (P < 0.05 or P < 0.01). GHRH was expressed in neurons in both humans and mice. Additionally, GHRH co-localized with presynaptic and postsynaptic sites of inhibitory neurons. Coimmunoprecipitation confirmed that GHRH interacted with GABAAα1 and GABAAβ2 + 3. GHRH may play an important role in inhibiting seizures by activating GABAARs.

Human samples. All clinical samples were randomly selected from our brain tissue bank, which includes 340 tissues from patients with epilepsy and 120 control tissues. In our experiment, samples were collected from 22 patients [13 males and 9 females; mean age 35.23 ± 8.48 years (20-52 years); mean disease course 11.95 ± 6.01 years (5-26 years)] who had undergone resection surgery for intractable epilepsy. All presurgical assessments comprised a detailed history and neurological examination to ensure that these patients were suitable candidates for surgery, as previously described [3]. All patients had taken three or more antiepileptic drugs (AEDs). Twelve histologically normal samples were randomly selected for use as a control group [7 males and 5 females; mean age 32.17 ± 8.31 years (20-52 years)]. The control samples were obtained from patients who had undergone cranial surgery due to increased intracranial pressure caused by head trauma and who had no history of epilepsy, had not been exposed to AEDs and had no apparent signs of CNS disease. Additionally, no significant differences in gender or age were observed between the patients and the control group (P > 0.05). The clinical features of the patients with TLE and the controls are summarized in Tables 1 and 2, respectively.
Mouse models of epilepsy. All adult male C57/BL6 mice weighing 24-28 g were obtained from the Experimental Animal Center of Chongqing Medical University, China, and were randomly divided into a normal control group and an experimental group. All animals were housed in cages with a 12/12 h light/dark cycle and other standard conditions, including a room temperature of 23 ± 1 °C and access to food and water ad libitum. The pentylenetetrazol (PTZ) kindling model was developed using a previously reported protocol 21 1  34  M  8  CBZ, VPA, PHT, TPM  NL  TNl   2  46  M  15  LTG, CBZ, VPA, PHT  NL  TNl   3  38  F  11  OXC, PB, TPM, PHT  NL  TNr   4  32  M  6  CBZ, TPM, PHT, VPA  NL  TNr   5  27  F  13  LTG, CBZ, PB, PHT  Gliosis  TNr   6  40  F  8  CBZ, VPA, TPM  Gliosis  TNl   7 30). After each injection, all animals were immediately observed for 30 min. The evoked seizures were scored using the criteria described by Racine (1972). Mice with at least three consecutive seizures with a score 4 or 5 were considered fully kindled. The kainic acid (KA)-induced chronic epilepsy model was generated as previously described 22,23 . Briefly, stereotaxic injections into the dorsal blade of the CA1 area were made at the following coordinates with respect to the bregma: anteroposterior-1.8 mm, mediolateral-1.5 mm, and 1.5 mm below the dura. Mice were deeply anesthetized and unilaterally injected with 1.0 nmol of KA (Sigma-Aldrich Co., St. Louis, MO, USA), and 50 nl saline were injected into the hippocampus in the contralateral hemisphere. We injected KA for 1 min with a 0.5 μl microsyringe (Hamilton, Reno, NV) 17 . In addition, the microsyringe was left in situ for 5 min after the injection and was then withdrawn slowly to reduce backflow along the injection trace.
In vivo multichannel electroencephalogram (EEG) recordings. EEG recordings were captured from the different groups (control group and epilepsy group) in the chronic phase of the KA-induced epilepsy model. We implanted a microwire array into the hippocampus of each mouse and performed multichannel EEG recordings in vivo. We used an OmniPlex ® Dneural Data Acquisition System (Plexon, Dallas, TX, USA) to record EEGs, and defined an electrophysiological seizure as a seizure with a high frequency (>5 Hz) and high amplitude (>2 times the baseline) that lasted for more than 5 s.  Tissue preparation. Humans. All human samples were collected from patients in the operating room, immediately frozen in liquid nitrogen and stored at −80 °C until they were used for Western blotting. The remaining samples were sectioned at a 10 μm thickness and stored at −20 °C for immunofluorescence analysis.
Mice. For the immunofluorescence analysis, the brain tissues were fixed with 4% paraformaldehyde and then successively incubated in 20% and 30% sucrose in PBS for 24 h. The tissues were then sectioned at 10 µm at −20 °C for the laser confocal microscopy analysis. For Western blotting, the neocortex and hippocampus were flash-frozen in liquid nitrogen and then stored at −80 °C until further use.   Western blot and coimmunoprecipitation. Western blot. The tissue samples were collected from humans and mice for Western blot analysis using a previously described protocol 4,25 . Briefly, total proteins were extracted according to the manufacturer's protocol (Keygen Biotech, Nanjing, China). SDS-PAGE gels (5% stacking gel; 10% separating gel) were used to separate total protein lysates (40 μg per lane) and electrophoretically transferred to polyvinylidene fluoride membranes (PVDF) (Millipore Corporation, USA). PVDF membranes were incubated with 5% nonfat milk for 1 h at room temperature (RT) to prevent non-specific binding. Later, the membranes were incubated with rabbit anti-GHRH (1:500, catalog No: ab187512, Abcam, Cambridge, MA, USA) and anti-GAPDH (1:3,000, Proteintech, Wuhan, China) antibodies overnight at 4 °C. The secondary antibody (peroxidase-conjugated goat anti-rabbit IgG (1:3,000, Proteintech, Wuhan, China) was incubated with the PVDF membranes for 1 h at RT on the next day. Enhanced chemiluminescence (ECL) reagent (Thermo, Marina, CA, USA) and a Fusion FX5 image analysis system (Vilber Lourmat, Marne-la-Vallée, France) were used to visualize the bands. Finally, Quantity One software (Bio-Rad, CA, USA) was used to measure the resulting optical density (OD) values, which were normalized to GAPDH expression.

Patients Age (y) Sex (M/F) Resection tissue Pathology
Coimmunoprecipitation. Coimmunoprecipitation was performed according to the method used in our previous publication 4 . Protein extracts from mouse hippocampal tissues were homogenized and then mixed with immunoprecipitation (IP) lysis buffer. Equal amounts of proteins were incubated with 2 μl of rabbit IgG (Abcam, Statistical analysis. We analyzed the data using Student's t-test to evaluate the differences between two groups (two-independent samples, normal distribution and equal variances) with SPSS18.0 software (SPSS, Inc., Chicago, IL, USA). Each experiment was repeated three times. The χ 2 test was used to compare the gender-specific differences between patients with TLE and controls. P < 0.05 or P < 0.01 was considered a statistically significant difference. The data are presented as the Mean ± SEM. Data were also analyzed using GraphPad Prism software (GraphPad Prism Software, San Diego, CA, USA).
Data availability. The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Results
Clinical characteristics. In the study, all patients in the TLE group had a seizure recurrence history of at least 5 years and had taken three or more antiepileptic drugs (AEDs). In addition, the TLE group had a mean age of 35.23 ± 8.48 years (20-52 years) and consisted of 22 patients (13 males and 9 females); the average duration of epilepsy was 11.95 ± 6.01 years (5-26 years) ( Table 1). The control group had a mean age of 32.1 7 ± 8.31 years (20-52 years) with 7 men and 5 women (Table 2). No significant differences in age and gender were observed between the TLE and control groups (P > 0.05).

Increased GHRH expression in patients with TLE. First, we performed Western blotting to investigate
GHRH expression levels in the cortices of patients with epilepsy (n = 22) and controls (n = 12); GAPDH expression was assessed as an internal loading control and was present as a 36 kDa band (Fig. 1a). In our experiment, a 12 kDa GHRH band was detected (Fig. 1a), and the mean OD ratio for GHRH was significantly increased in the TLE group (1.43 ± 0.10) compared to that in the control group (0.81 ± 0.06) (Fig. 1a). GHRH expression levels were normalized by calculating the OD ratio of GHRH to GAPDH (GHRH/GAPDH). These values were significantly different (P < 0.01) (Fig. 1b). All data are presented as the Mean ± SEM.
Increased GHRH expression in mouse epilepsy models. Next, we examined GHRH expression in mouse models of epilepsy induced by KA (Fig. 2a) and PTZ (Fig. 2b) using Western blotting. The evoked seizures were scored in accordance with the criteria reported by Racine (1972). Mice with at least three consecutive seizures with scores of 4 or 5 were considered fully kindled. Moreover, EEG recordings were captured from both the control and epilepsy group in the chronic phase of the KA-induced model (Fig. 2a). Similar to the human samples, the GHRH expression levels were significantly increased in the hippocampus and cortex of the epileptic mice compared to the levels in control mice. In the KA-induced epilepsy mouse model, the GHRH expression levels in the hippocampus and cortex of epileptic mice were 1.17 ± 0.11 and 1.44 ± 0.08 compared to 0.85 ± 0.09 and 0.75 ± 0.11 in the control mice, respectively (Fig. 3a,c). All data are shown as the mean ± SEM. Similarly, in a mouse model of PTZ-kindled epilepsy, the GHRH expression levels in the hippocampus and cortex of epileptic mice were 1.21 ± 0.05 and 1.16 ± 0.05 compared with 0.85 ± 0.04 and 0.88 ± 0.04 in the control mice, respectively (Fig. 4a,c). GAPDH was also used as an internal control, and the mean OD value of GHRH was normalized to GAPDH (GHRH/GAPDH). These values were significantly different (P < 0.05 or P < 0.01) (Figs 3b,d and 4b,d). Data are expressed as the mean ± SEM.

Localization of GHRH in the human neocortex and in the mouse hippocampus and cortex.
Immunofluorescence was performed on cortical tissues obtained from adult human brains to examine the localization of GHRH. GHRH (green) was co-expressed with MAP2 (red, a marker of dendrites) in the cytoplasm of neurons (yellow) (Fig. 5a), but was not co-expressed with GFAP (red, a marker of astrocytes) in the human neocortex (Fig. 5b). Furthermore, we investigated the localization of GHRH in the mouse hippocampus and cortex. GHRH (green) was localized in both the cortex and the whole hippocampus (Fig. 6a,d), and the expression overlapped with MAP2-positive (red) neurons. However, GHRH (green) was not co-localized with GFAP (red) in astrocytes (Fig. 6b,e). In addition, we further studied the localization of GHRH in specific neurons and subcellular structures in the human neocortex. GHRH was co-expressed with GAD67 (red, a marker of inhibitory neurons) (Fig. 7a) and VGAT (purple, a marker of presynaptic specializations from inhibitory neurons) and Gephyrin (red, a marker of inhibitory specializations from postsynaptic neurons) (Fig. 7d). Finally, we obtained consistent results in the mouse hippocampus and cortex. GHRH staining overlapped with GAD67 (red) (Fig. 7b,c), VGAT (purple) and Gephyrin (red) staining (Fig. 7e,f).  Interaction between GHRH and GABAARs. Immunoprecipitations were conducted using mouse hippocampal tissues to further examine the interactions in which GHRH participates in inhibitory synapses in the hippocampus. GABAAα1 (Fig. 8a) and GABAAβ2 + 3 (Fig. 8b) were coimmunoprecipitated with anti-GHRH antibodies, indicating that GHRH interacted with GABAARs. This interaction was verified by the reciprocal coimmunoprecipitation of GHRH with anti-GABAAα1 (Fig. 8a) and anti-GABAAβ2 + 3 antibodies (Fig. 8b).

Discussion
The primary finding of this study is that GHRH expression was significantly increased in the brains of both patients with TLE and in two mouse models of epilepsy induced by either KA or PTZ. Second, most of the GHRH-positive cells co-expressed neuronal markers, but not glial cell markers, in both human and mouse tissues. In addition, GHRH co-localized with presynaptic and postsynaptic inhibitory neurons. Interestingly, we confirmed that GHRH and GABAARs interacted. Therefore, we propose that GHRH may exert an antiepileptic effect.
GHRH has been shown to increase GABA levels and activate GABARs to promote NREM sleep, which contributes to the regulation of sleep by GH. Peterfi Z. et al. discovered both NREM sleep amount and double-labeled Fos + glutamic acid decarboxylase (GAD) cell counts in the MnPN and VLPO were significantly increased in rats that received intracerebroventricular injection of GHRH compared with saline-treated rats. While the opposite results were obtained in the octreotide (somatostatin analog OCT) and the high dose of GHRH antagonist-treated groups. All the results identified GABAergic neurons as potential targets of the sleep-regulatory actions of GHRH in the MnPN and VLPO 13 . Furthermore, a study using an amyloid precursor protein (APP) transgenic mouse model showed that mice express GHRH-dependent factors that protect against amyloid β Aβ) deposition. GHRH also increases the level of IGF-1, thereby collectively improving Alzheimer's disease (AD)-and obstructive sleep apnea (OSA)-related cognitive dysfunction 14,26 . In addition, a clinical research unit at the University of Washington School of Medicine conducted a randomized, double-blind, placebo-controlled study that administered a 1 mg subcutaneous injection of a human GHRH analogue daily. After 20 weeks, GABA levels were increased in the brain (P < 0.04), and IGF-1 levels were positively correlated with GABA levels (r = 0.47; P = 0.001). However, myo-inositol (MI, an AD-related marker) levels decreased (P = 0.002), which may be one of the mechanisms by which GHRH and IGF-1 improve cognitive function during normal aging 11 .
The GABAA receptor (GABAAR) is a part of the ligand-gated ion channel complex that mediates the opening of the chloride channel. GABAARs are the most important part of inhibitory neurotransmission in the brain and are composed of the following subunits: α1-6, β1-3, γ1-3, δ, ε, π, θ, and ρ1-3, among which, the α1-3, β2-3 and γ2 subunit compositions of five dimers play a major role 6,27 . Changes in the expression of these subunits and alterations in the function of the receptor induced by mutation, tumors or hypoxia may interfere with inhibitory transmission by GABAARs, resulting in epilepsy 15,28 . For example, the alpha 1 subunit D219N mutation has been identified in patients with idiopathic generalized epilepsy 29 , and an A322D mutation was detected in patients with juvenile myoclonic epilepsy 30 . Fluoxetine has been shown to antagonize brain injury by increasing GABAAR levels in the dentate gyrus and CA1-CA2 areas of rats with pilocarpine-induced epilepsy 31 . As shown in our previous studies, ubiquitin-like protein 1 (plic-1), which is enriched in inhibitory synapses and is associated with GABAARs and the ubiquitin proteasome system (UPS), activates GABAARs, thereby inhibiting seizures 32 .
In this study, GHRH expression was first assessed by Western blotting and was significantly increased both in patients with TLE and in two mouse models of epilepsy induced by either KA or PTZ. Moreover, immunofluorescence staining revealed the co-localization of most of the GHRH-positive cells with neurons, but not glial cells, in both humans and mice. We incubated sections with antibodies against inhibitory neurons, inhibitory synaptic markers (GAD67, VGAT and Gephyrin) and GHRH and examined their immunofluorescence localization to further determine whether GHRH was related to inhibition. GHRH co-localized with presynaptic and postsynaptic inhibitory neurons. Then, coimmunoprecipitations were performed to examine whether GHRH interacted with GABAR at inhibitory synapses and participated in the inhibitory function. Surprisingly, coimmunoprecipitation experiments confirmed interactions between GHRH and GABAAα1 and GABAAβ2 + 3. Therefore, we propose that GHRH participates in anti-epileptic processes by increasing GABA levels and activating GABAARs. However, the results in this study are descriptive and do not provide evidence showing that GHRH contributes to an antiepileptic effect. Further studies are needed to test whether GHRH directly interacts with GABAARs to further activate them, thereby enhancing the inhibitory effect to prevent epilepsy.