Corpora amylacea deposition in the hippocampus of patients with mesial temporal lobe epilepsy: A new role for an old gene?

BACKGROUND: Mesial temporal lobe epilepsy (MTLE) is the most common medically refractory epilepsy syndrome in adults, and hippocampal sclerosis (HS) is the most frequently encountered lesion in patients with MTLE. Premature accumulation of corpora amylacea (CoA), which plays an important role in the sequestration of toxic cellular metabolites, is found in the hippocampus of 50–60% of the patients who undergo surgery for medically refractory MTLE-HS. However, the etiopathogenesis and clinical importance of this phenomenon are still uncertain. The ABCB1 gene product P-glycoprotein (P-gp) plays a prominent role as an antiapoptotic factor in addition to its effl ux transporter function. ABCB1 polymorphism has been found to be associated with downregulation of P-gp expression. We hypothesized that a similar polymorphism will be found in patients with CoA deposition, as the polymorphism predisposes the hippocampal neuronal and glial cells to seizure-induced excitotoxic damage and CoA formation ensues as a buffer response. MATERIALS AND METHODS: We compared five single nucleotide polymorphisms in the ABCB1 gene Ex06+139C/T (rs1202168), Ex 12 C1236T (rs1128503), Ex 17-76T/A (rs1922242), Ex 21 G2677T/A (rs2032582), Ex26 C3435T (rs1045642) among 46 MTLE-HS patients of south Indian ancestry with and without CoA accumulation. RESULTS: We found that subjects carrying the Ex-76T/A polymorphism (TA genotype) had a fi ve-times higher risk of developing CoA accumulation than subjects without this genotype (Odds ratio 5.0, 95% confi dence intervals 1.34–18.55; P = 0.016). CONCLUSION: We speculate that rs1922242 polymorphism results in the downregulation of P-gp function, which predisposes the hippocampal cells to seizure-induced apoptosis, and CoA gets accumulated as a buffer response.


S42
conditions like Alzheimer's disease and Parkinson's disease. [6] The extent of accumulation of CoA has been correlated with seizure duration and interictal psychosis in patients with MTLE-HS. [5] In this regard, the detection of CoA, a pathological marker of neurodegeneration, in the surgical specimens of patients with MTLE-HS is of special interest and deserves a closer scrutiny to elucidate the possible pathomechanisms of HS.
Interestingly, the promotion of CoA formation has been linked to cellular stress, in particular oxidative stress and mitochondrial dysfunction. [6] As recurrent seizures induce signifi cant oxidative stress in the hippocampus, it is likely that CoA formation ensues as a buffer response to seizure-induced excitotoxic damage in the hippocampal neuronal and glial cells. [5] P-glycoprotein (P-gp), an ATPdependent effl ux pump with a broad substrate specifi city, encoded by the ABCB1 gene, is postulated to mediate at least part of the drug resistance. [2,7] The expression of ABCB1 in the epileptic foci of drug-resistant epilepsy was found to be several-fold elevated in both animal and human studies, indicating its signifi cant role in epilepsy.
The C3435T, a major allelic variant of the ABCB1 gene, is proposed to play a crucial role in drug resistance in epilepsy. The C/C genotype carriers reportedly are at a higher risk of pharmacoresistance to AEDs, and this issue has been extensively investigated with confl icting results. [2] The recent meta-analyses have failed to confi rm an association between the ABCB1 C3435T polymorphism and the risk of drug resistance, suggesting a revision in contribution of this polymorphism in the multi-drug transporters hypothesis of drug resistance in epilepsy. [8][9][10] In addition to the effl ux transporter function, P-gp plays an important role in tissue defence through the excretion of toxic compounds and their metabolites, especially in the presence of oxidative stress. [11,12] Hence, the overexpression of ABCB1 gene products might be a tissue response to the oxidative stress of epilepsy.
Uwai et al. reported an intron 16 polymorphism (Ex-76 T/A) that resulted in downregulation of P-gp function in renal carcinoma cells. [13] The Ex-76 T/A has been shown to be in linkage disequilibrium with G2677T and the alleles of Ex-76 T/A formed major haplotypes with G2677T and C3435T. [14] We hypothesized that a similar polymorphism will be found in patients with CoA deposition as the polymorphism predisposes the hippocampal neuronal and glial cells to seizure-induced excitotoxic damage and CoA formation ensues as a buffer response. This prompted us to investigate the role of ABCB1 polymorphisms in the premature accumulation of CoA by comparing the distribution of single nucleotide polymorphisms (SNPs) between groups of MTLE-HS patients with and without CoA in their hippocampus.

Selection of subjects
The epilepsy patients were recruited through the Epilepsy Clinic at R. Madhavan Nayar Center for Comprehensive Epilepsy Care, Sree Chitra Tirunal

Institute of Medical Sciences and Technology at
Trivandrum, India. Phenotypic definition of drug resistance is a critical issue in epilepsy genetics. There is no consistent clinical defi nition of multidrug resistance in epilepsy, and agreement among the different defi nitions is strong but imperfect. [15] This has resulted in diverse criteria used by different researchers, or even a lack of explicit criteria in some cases, rendering it diffi cult to compare fi ndings across studies. To address this issue and to avoid population substructuring in the patient population, we used stringent patient selection criteria to maintain genotype homogeneity. Only patients with self-declared Malayalam-speaking south Indian ancestry with pathologically proven HS were recruited who were seizure-free for more than 1 year after surgery.
We recruited 46 consecutive patients with MTLE who had undergone anterior temporal lobectomy and had pathologically verifi ed HS. The details of our noninvasive presurgical evaluation protocol have been described previously. [16] We excluded subjects with HS associated with temporal lobe neoplasms (secondary HS).

Surgical procedure
For MTLE-HS, we performed an en bloc standard anterior temporal lobectomy with amygdalohippocampectomy. [16] After a standard temporal craniotomy, under general anesthesia, which exposed approximately 8 cm of the temporal lobe and the inferior frontal gyrus, a 5-6-cm block of temporal lateral cortex along with superior temporal gyrus was resected depending on temporal lobe dominance and the presence of prominent cortical   [17] Blood samples (5 ml) were drawn into EDTA vials and genomic DNA was isolated from human leukocytes by standard methods.

Statistical methods
We summarized the quantitative data as mean ±

Patient characteristics
Of  Table 2, and is elaborated.

Clinical features
The age at surgery was found to be signifi cantly  Table 3].
Signifi cant differences were observed at the allele level also (P = 0.04). After adjustment of the odds ratios for age at surgery, sex, age at onset, duration of epilepsy and

Discussion
In this study, using a strict phenotypic defi nition, we Interestingly, in a situation of organ damage or disease, changes in the expression levels of ABC transporters have been observed, probably to compensate for the increased load of harmful products of oxidative stress formed during an insult or to compensate for the loss of effl ux pumps in damaged tissue. [12] In addition to its protective function, P-gp has also been implicated in resistance to apoptosis, most importantly by inhibiting Fas-induced caspase-3 activation and by modulating ceramide metabolism. [12,20] Hence, the overexpression of ABCB1 gene products at the site of tissue damage is a physiological phenomenon for tissue defense. Overexpression of ABCB1 in the epileptic tissue, both in animal and human tissue, was one of the key evidence to propose its role in drug resistance in epilepsy. [21,22] The expression of ABCB1 in the epileptic foci of drug-resistant epilepsy patients was shown to be elevated 10-fold. [23] Elevated levels of ABCB1 and multidrug-associated protein (ABCC1/MRP1) have also been associated with pathologies associated with drug refractoriness, such as in tuberous sclerosis and also in epileptic tissue surrounding dysembryoplastic neuroepithelial tumors, focal cortical dysplasia and hippocampal sclerosis. The adjacent nonepileptic tissue from the same patients did not show elevated expression of the transporter. [24] In 2003, Siddiqui et al. [25] reported the C3435T polymorphism in the ABCB1 gene as being associated with resistance to multiple AEDs, and leading to the suggestion, for the fi rst time, that drug resistance in epilepsy might be genetically determined. Since then, several other genetic association studies have attempted to verify this result; however, taken overall, the role of P-gp in drug resistance in epilepsy still remains uncertain. [2] Three recent meta-analyses addressed the issue. [8][9][10] The fi rst report in 1073 Caucasians patients included three studies with the same defi nition of drug responsiveness and drug resistance, [8] and the second one included 11 studies involving 3371 patients with different ethnicities, but the same defi nition of drug responsiveness and drug resistance. [9] Neither of these studies could confi rm this association and stratifi cation of ethnic subgroups in the second meta-analysis also provided no further evidence.
The fi ndings of the third meta-analysis indicate that neither the C allele nor the T allele carriers of the ABCB1 (ILAE) [15] and ethnicity, to address the issue of different defi nitions of drug responsiveness and drug resistance in the patients with various ethnicities, did not show any association under all genetic models. [10] Hence, in view of these data, we may have to revise the role for ABCB1 gene in epilepsy and more importance must be attributed to its role in tissue defense in epileptic tissue. The role of CoA in sequestration of toxic cellular metabolites is being increasingly recognized and is supported by the presence of ubiquitin and heat shock proteins in CoA and its abundant accumulation in neurodegenerative diseases, where oxidative stress and mitochondrial dysfunction are common denominators compared with age-matched controls. [6] In patients with MTLE-HS, premature accumulation of CoA in the hippocampus might be a refl ection of a tissue reaction to buffer the free radicals and other toxic metabolites generated as a result of repeated seizures. Our observation in this and previous studies [5] of the association between older age and longer duration of epilepsy with CoA accumulation in the hippocampus is in agreement with this hypothesis.
Is there a relationship between ABCB1 polymorphisms and CoA accumulation that we observed in our wellcharacterized MTLE-HS patients? We hypothesize a link through the tissue defense function of ABCB1-encoded P-gp. The Ex-76 T/A has been shown to be in linkage disequilibrium with G2677T and the alleles of Ex-76 T/A formed major haplotypes with G2677T and C3435T. [14] Although data to date are mainly available for coding SNPs in exons 21 and 26, Uwai et al. [13] found that Ex-76 substantiate the need to revise its role in epilepsy genetics. However, the association found in our study is modest, most likely due to the small sample size. As the stringent selection criteria resulted in the small sample size, haplotype analysis was not possible. As the G2677T polymorphism (genotypic and allelic) was also found to be associated with CoA deposition and the Ex-76 T/A has been shown to be in linkage disequilibrium with G2677T, the haplotypic analysis would be interesting. Hence, a further study with a larger sample size and P-glycoprotein functional expression studies will be required to establish the relationship between Ex-76 T/A polymorphism and CoA accumulation.