Conditional downregulation of brain- derived neurotrophic factor and tyrosine kinase receptor B blocks epileptogenesis in the human temporal lobe epilepsy hippocampus
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.60392
Source of Support: None, Conflict of Interest: None
Background : Brain-derived neurotrophic factor (BDNF) has been implicated as a potential therapeutic target in temporal lobe epilepsy (TLE). However, whether BDNF exerts an epileptogenic or antiepileptogenic function remains controversial. Materials and Methods : BDNF/tyrosine kinase receptor B (trkB) expression levels were comparatively assessed in the hippocampal tissue of TLE patients with (HS group) and without hippocampal sclerosis (non-HS group) as well as from non-epileptic controls. Results : Immunohistochemistry and immunoblot analysis revealed a marked increase in BDNF/trkB expression in the dentate gyrus and CA3 regions of HS and non-HS groups. The lack of any differences in expression levels was observed between HS and non-HS patients. Meanwhile, treatment with VPA (Valproic acid, anti-epileptic drug) resulted in a significant down-regulation of BDNF/trkB protein expression in sclerotic and non-sclerotic hippocampus (P < 0.001). In contrast, no marked change was noticed in VPA-untreated and OA-treated groups (sodium octanoate). Conclusion : These results suggest that the up-regulation of BDNF/trkB pathway might be at least in part responsible for the epileptogenesis.
Keywords: Brain-derived neurotrophic factor, tyrosine kinase receptor B, temporal lobe epilepsy, valproic acid, hippocampal sclerosis, epileptogenesis
Temporal lobe epilepsy (TLE) is a common, complex disorder, with heterogeneous clinical manifestations and multiple genetic and non-genetic factors.  The hippocampus in many individuals with TLE is characterized by cell loss in specific regions, most prominently in CA1 and CA3, a pattern termed hippocampal (or mesial temporal) sclerosis (HS).  Morphological investigation of resected tissue has revealed that while one group of TLE is associated with HS characterized by segmental or total cell loss and aberrant mossy fiber sprouting, especially in the CA1 and CA3 areas and dentate hilus regions of the hippocampus, , another group is associated with less segmental cell loss (non-HS). Mossy fiber sprouting in the dentate gyrus may be a factor in the pathogenesis of TLE. It has been hypothesized that TLE arises out of the loss of hilar neurons and subsequent deafferentation of the inner molecular layer. ,, Neurotrophic factors may be required to initiate exuberant synapses and therefore may be important for the process of epileptogenesis.
One target of the responsible factors is the brain- derived neurotrophic factor (BDNF), a member of the neurotrophin family, which is highly expressed in many areas of the central nervous system (CNS) and is critically involved in synaptic plasticity, membrane excitability and neurotransmission.  Studies on hippocampal slices have indicated that BDNF can potentiate excitatory transmission, probably by a presynaptic mechanism mediated by tyrosine kinase receptor B (trkB).  Furthermore, while BDNF increases excitation and decreases inhibition in the brain, seizures can induce an up-regulation of both BDNF and trkB receptors. 
Valproic acid (VPA) is a short-chain, branched fatty acid with broad-spectrum anticonvulsant activity. VPA treatment potently protected epileptic animals from hippocampus-dependent cognitive impairment after kainic acid-induced seizures.  To investigate the implicated role of BDNF in TLE, we examined the expression of BDNF and trkB in hippocampi of patients with TLE. We also tested whether VPA would influence BDNF and trkB expression, especially in HS and non-HS, which might be at least in part responsible for preventing the epileptogenesis.
Patient selection and tissue collection
This study was performed between January 2006 and December 2007 on hippocampal tissue obtained from 40 patients suffering from pharmaco-resistant TLE as determined by the First Affiliated Hospital of Harbin Medical University, according to the criteria of the International League Against Epilepsy (ILAE) classification for presurgical evaluation.  This study was approved by the ethics committee of the First Affiliated Hospital of Harbin Medical University (China), and a written informed consent for the study was obtained from each patient. All patients were assessed by comprehensive epilepsy testing, including clinical characterization, Video-electroencephalogram (VEEG), psychiatric and neuropsychological assessments. Hippocampal sclerosis was defined if visual inspection of magnetic resonance imaging (MRI) pictures revealed both hippocampal atrophy and T2 signal abnormalities, and then identified by microscopic analysis by an experienced neuropathologist using neuronal loss in a characteristic pattern (most severe in area CA1 and the dentate hilus) as the criteria.  Patients with bilateral HS or dual pathology (HS accompanied by other epileptogenic lesion) were excluded. After en bloc resection, the hippocampus was microdissected into subregions and total protein was extracted from the CA3 and dentate gyrus subregions. Tissue samples for immunohistochemistry were immediately fixed in 4% formaldehyde and 15% saturated picric acid (pH 7.4) overnight at room temperature and embedded in paraffin. Samples for immunoblotting were freshly frozen on dry ice and stored at -80 °C until further use. Normal hippocampal tissue samples were obtained from the brain storeroom of the First Affiliated Hospital of Harbin Medical University as experimental controls. Ten samples were used for immunohistochemistry study, and ten samples (death due to cerebral trauma, cancer of pancreas, congestive heart disease, acute hepatic failure) for the Western-blot analysis. Considering that trauma increases the expression of BDNF, only one sample was derived from the cerebral trauma.
For pharmacological studies, hippocampal specimens from surgery were immediately transferred into cold sucrose solution containing the following (in mM) 234 sucrose, 11 glucose, 24 NaHCO 3 , 2.5 KCl, 1.25 NaH 2 PO 4 , 2 MgSO 4 , and 0.5 CaCl 2 , equilibrated with a 95%-5% mixture of O 2 and CO 2 . Horizontal slices (400 µm) were cut on a VT 1000S Vibratome (Leica, Bensheim, Germany) at 4°C in sucrose solution. Hippocampal specimens were immediately immersed in sucrose solution containing sodium valproate (VPA; Sigma, 0.3 mM, similar to the therapeutic dose) or its pharmacologically inactive analog, sodium octanoate (OA; Sigma) for 3 h (hours) or 9 h. Slices were thereafter stored at -80°C for Western - blot analysis.
Histology and histochemistry
Fixed slices were incubated in phosphate-buffered-saline (PBS, 0.01M) containing 10% methanol and 3% H 2 O 2 for 1 h then incubated in avidin-biotin-horseradish peroxidase complex (ABC Elite kit; Vector laboratories Inc, Burlingame, CA) in PBS (1:100; pH 7.3) containing 2% Triton X for 48 to 72 h at 4°C. The slices were then reacted with 3, 3'-diaminobenzidine tetrahydrochloride (0.06%) and H 2 O 2 (0.003%) in PBS (pH 7.4). Sections were fixed and stained with hematoxylin and eosin (HE).
To test the expression of BDNF and trkB in patients with TLE, immunohistochemical staining was first performed on different groups. The different subareas in the hippocampus were defined according to the nomenclature of Amaral and Insausti.  Selected hippocampal subareas for analysis were: CA1, CA2, CA3, hilus, and dentate gyrus. Three-micron-thick slices were washed three times, for 10 min each, in 0.1 M and endogenous peroxidase activity blocked with 3% H 2 O 2 . Tissue sections from all patients were stained simultaneously to minimize variation in the immunohistochemical reactions. Afterwards, sections were incubated in one of the primary antibody solutions (BDNF, 1:200 and trkB 1:150, respectively; SANTA CRUZ, CA, USA), overnight at 4°C.  After three washes in 0.1 M PBS for 5 min each, sections were incubated with secondary antibodies (goat anti-rabbit IgG, 1:200 dilution; Vector Laboratories, Burlingame, CA). The resultant immuno peroxidase complexes were developed by incubation in 0.5% 3, 3'-diaminobenzidine hydrochloride (DAB; Sigma, Saint Louis, MO, USA) in PBS. For each antibody, control experiments were performed in which the primary antibody was omitted.
The results of the immunohistochemical staining were read separately by two pathologists who were blinded to the clinical parameters of the individual cases. BDNF and trkB immunostaining showed brown coloration of plasma and nuclei indicating positive staining. Unstained plasma and nuclei took blue color. The score of BDNF and trkB = Number of cells showing positive staining/Total number of cells x 100%. The number of BDNF and trkB expressing positive cells was estimated as a percentage of the total number of cells per section and scored in four categories: 0, < 5%; 1, 5-25%; 2, 26-50%; 3, >50%. The intensity of staining was graded such that negative control, (score 0), weak staining, (score 1), moderate staining, (score 2), strong staining, (score 3). Then score of percentage of cells stained and the intensity of cell staining were combined such that: A score of 0 was (-), a score of 1 was (±), a score of 2-3 was (+), a score of 4-6 was (++), and a score of 7-9 was (+++).
Western blot analysis
Fresh hippocampal specimens were lysed on ice for 10 min in a lysis buffer (400 mM KCl, 50 mM HEPES, 1.5 mM EDTA, 20% glycerol, 0.5% NP 4 O, 20 mM NaF, 10 mM Na 2 molybdate, 10 µM Na 3 ortho, and 1 mM dithiothreitol) containing a mixture of protease inhibitors (1 mM PMSF, 0.2 mg/ml bacitracin, 0.2 mg/ml aprotin, 5 µg/ml leupeptin, and 5 µg/ml pepstatin A). Specimens were then homogenized and centrifuged at 1000 x g for 30 min. Protein concentrations were determined by the Bradford method. The supernatant was boiled for 8 min in 2 x SDS sample buffer (62.5 M Tris-HCl, 10% glycerol, 2% SDS, 0.05% bromphenol blue, 1% ß-mercaptoethanol) and centrifuged at 1000 x g for 1 min at 4°C. Samples were electrophoresed on 15% SDS polyacrylamide gels, transferred onto nitrocellulose membrane using a semi-dry system at 24 V for 435 min. Membranes were stained by Ponceau S, nonspecific binding blocked by 0.02 mM PBS containing 5% skim milk and 0.05% Tween-20 for 1 h, incubated overnight with affinity-purified rabbit polyclonal antibody for BDNF (1:200 dilution) and trkB (1:150 dilution) in the same buffer solution. After three 10-min washes in PBS, the membrane was incubated with a horseradish peroxidase-conjugated anti-rabbit IgG (1:6000 dilution; Promega Corporation, Madison,Wisconsin, USA) for 1 h. Following a 30 min rinse with buffer, immunoreactive bands were dried and visualized by enhanced chemiluminescence in the dark for 30 min, using a stabilized substrate for alkaline phosphatase (Promega Corporation, Madison, Wisconsin, USA).
Data are expressed as mean ± standard deviation of values. Statistical analyses were performed using the One-way ANOVA, Chi-squared test and Wilcoxon rank sum test. Densitometric measurements were analyzed for significant group differences using one-way ANOVA. A value of P < 0.05 was considered statistically significant. Data were analyzed by SAS 8.2 software package (SAS Institute Inc., Cary, North Carolina, USA) and the program for one-way Anova test in SAS is proc anova; the program for Chi-squared test is proc freq; and the program for Wilcoxon rank sum test is proc npar1way Wilcoxon.
The clinical characteristics of the control and TLE patients (HS and non-HS) are summarized in [Table 1]. Forty patients fulfilled the inclusion criteria. HS was assessed in 20 patients (28.5 ± 8.7 years of age); 20 patients were assessed to be non-HS (22.2 ± 9.3 years of age). The mean duration of epilepsy was 14.7 ± 9.8 years in HS patients and 11.0 ± 7.1 years in non-HS patients. There were no differences between HS and non-HS patients with regard to the side (right or left) of resection, age and gender (P > 0.05).
Increased expression of brain-derived neurotrophic factor and tyrosine kinase receptor B in patients with temporal lobe epilepsy
To assess the expression levels of BDNF and trkB in the patients with TLE in comparison with controls, an immunohistochemical study was performed. In the hippocampal tissue, an increase of BDNF immunoreactivity was evident in the cytoplasm and proximal dendrites of CA3 pyramidal neurons [Figure 1]. In addition, increases of trkB immunoreactivity have been present in cytoplasm and nucleolus. Meanwhile, the glial cells as identified by glial fibrillary acidic protein (GFAP) staining in adjacent sections were negative for BDNF and trkB.
The control group displayed low level of BDNF immunoreactivity. In the TLE group (HS and non-HS), however, a significant increase in BDNF was observed in dentate gyrus compared with the control hippocampus (P < 0.01, [Table 2]). Similar results were observed in the CA3 region (data not shown). trkB displayed a similar immunoreactivity distribution. Quantification of BDNF and trkB revealed a robust increase in the HS and non-HS group compared with the control group. However, there was no statistically significant difference between the HS and non-HS groups (P >0.05, [Table 2]).
Elevated brain-derived neurotrophic factor and tyrosine kinase receptor B protein levels in temporal lobe epilepsy patients
To assess whether comparable changes in protein product levels of BDNF and trkB occurred in the hippocampus, a semiquantitative analysis was carried out by Western-blotting. A drastic increase in BDNF and trkB proteins was observed in the dentate gyrus [Figure 2]a and CA3 regions (data not shown) of the TLE group. Results for the hippocampal tissues were consistent with immunohistochemical analysis [Figure 2], showing a significant increase of BDNF protein (14 KDa) in the HS and non-HS groups. Meanwhile, trkB levels were significantly up-regulated (145 KDa) compared with the control group ( P < 0.05). There were no obvious differences between the HS and non-HS groups ( P >0.05, [Figure 2]b).
Valproic acid treatment results in down-regulated brain-derived neurotrophic factor and tyrosine kinase receptor B protein levels
We tested the hypothesis that VPA treatment leads to the decrease of BDNF and trkB expression and explored the mechanism relating to the increase of BDNF and trkB in epileptogenic patients. Because BDNF synthesis would require time to change, 1 h VPA incubation did not produce any significant changes in BDNF expression, VPA (0.3 mM) decreased BDNF expression by 39% and 53% in HS and non-HS patients at 3 h, respectively, and 85% and 77% at 9 h compared with the VPA-untreated group (control). The levels of trkB attenuated 3.7 and 2.0 fold by 3 h and 6.4 and 5.0-fold by 9 h of VPA treatment compared with OA-treated and control groups (P < 0.001, [Figure 3]b). In contrast, changes of BDNF and trkB protein were not evident between the HS and non-HS group (P = 0.15, [Figure 3]b).
The majority of patients with TLE have HS as a pathological abnormality underlying TLE. Surgical resection of the epileptogenic tissue, in particular the hippocampus, most often results in seizure control, implicating the hippocampus in the generation and propagation of seizures in these patients.  In hippocampal specimens of patients with HS, the mossy fibers send collaterals into the inner molecular layer of the dentate gyrus and form multiple asymmetric synapses with the dendritic structures of granule cells. 
Seizure-induced up-regulation of BDNF , in the hippocampus is a common phenomenon across various experimental models of epilepsy. Studies from areas CA1 and CA3 in hippocampal slices have indicated that BDNF can potentiate excitatory transmission, probably by a presynaptic mechanism mediated by trkB receptors.  A similar mechanism is also implicated in the dentate gyrus.  However, the evidence for specific contributions of BDNF and trkB to TLE (HS and non-HS) remains controversial. Some studies have revealed that in comparison with non-HS cases, patients with HS show increased BDNF, nervre growth factor(NGF) and NT-3 mRNA.  Moreover, granule cell BDNF mRNA levels correlated inversely with neuronal density in the Ammon's horn and correlated positively with supragranular mossy fiber sprouting. These results indicate that in the chronically damaged human hippocampus, granule cells express neurotrophin mRNAs, and that the mRNA levels correlate with either neuronal loss in the hippocampus or aberrant supragranular mossy fiber sprouting. However, the present study shows obviously up-regulated BDNF and trkB immunoreactivity in the hippocampal neurons (dentate gyrus and CA3), with no pronounced differences in intensity between the HS and non-HS groups. Consistent with our immunohistochemical data, BDNF and trkB were significantly up-regulated in hippocampal tissue from the HS and non-HS groups compared with the control group. VPA is used to treat epilepsy. It alters gene transcription, through the inhibition of histone deacetylase and the activation of transcription factors and signaling cascades. , The results presented here demonstrate significant down-regulation of BDNF and trkB protein expression in the hippocampal regions of patients with HS and non-HS TLE. In contrast to the VPA-treated group, no marked change was noticed in VPA-untreated or OA-treated groups, indicating that BDNF-mediated activation of trkB enhances the excitatory actions on hippocampi of TLE patients.
A large body of evidence suggests that BDNF is involved in TLE, although it is not entirely clear whether its function is pro-epileptogenic or anti-epileptogenic. In human subjects with epilepsy, BDNF and other neurotrophins have been reported to play a role in morphological changes leading to hyperexcitability of the hippocampus. In particular, BDNF and nerve growth factor mRNA levels are increased in kindling and other seizure models, whereas NT-3 mRNA level is decreased. The increased neuronal activity leads to a secondary increase in BDNF/trkB levels and initiates further potentiation.  Recent studies have shown that interfering with BDNF signal transduction inhibits the development of the epileptic state in vivo and infusion of BDNF into hippocampus is followed by seizures.  These events ultimately would escalate by positive feedback and reach seizure threshold.  Our main contribution to this topic is the finding that drastic attenuation of BDNF/trkB levels occurs in the hippocampus when treated by VPA, implying the presence of BDNF-evoked excitatory synaptic currents in human TLE hippocampus.
The significant elevation of BDNF/trkB expression levels, in sclerotic and non-sclerotic hippocampal specimens suggests that epileptogenecity is promoted by BDNF activation of trkB. Further studies on the mechanism by which VPA regulates the function of BDNF/trkB in TLE patients are essential for a better understanding of human epilepsy.
We wish to thank Rui Re, Ben-chang Li (Harbin Medical University) for technical assistance. This work was supported by Grants-in-aid for Science Foundation of Heilongjiang province of China (GB06C40307).
[Figure 1], [Figure 2], [Figure 3]
[Table 1], [Table 2]