In the present study, we used adult human brain tissue (n = 14) from two sources: autopsies (kindly supplied by Dr. R. Alcaraz, Forensic Pathology Service, Basque Institute of Legal Medicine, Bilbao, Spain) and postoperative tissue from patients (Neurosurgery Service, Hospital de la Princesa, Madrid, Spain). The autopsy tissue was obtained at 2–3 h postmortem, from three normal males who died in traffic accidents (aged 23, 49, and 69 years). The human tissue obtained by biopsy through surgical intervention was from the temporal neocortex and hippocampal formation of 11 patients (H48, H65, H109, H123, H225, H239, H240, H241, H242, H247, and H248) diagnosed with intractable mesial temporal lobe epilepsy (sex: 3 males, 8 females; mean age and range: 33.8, 21–65 years; mean age and range of onset: 12.9, 6–18 years; mean and range of duration: 23.9, 4–50 years). According to the Helsinki Declaration, the patient's consent was obtained in all cases (British Medical Journal, 302: 1194, 1991) and all the protocols were approved by the Institutional Ethical Committee (Protocols 4/2002 and 14/2002; Hospital de la Princesa, Madrid, Spain). Video-EEG recording was performed through electrodes located in the scalp and bilaterally in the foramen ovale to locate the epileptic foci. Epileptogenic regions were further identified at the time of surgery through subdural electrocorticographic (ECoG) recordings with a grid of 4 × 5 electrodes and a strip of four electrodes embedded in Sylastic, with a 1.2 mm in diameter and 1-cm center-to-center interelectrode distance (Add-Tech, Medical Instrument Cooperation, Racine, WI, U.S.A.). These electrodes were placed directly over the exposed lateral temporal neocortex or uncus and parahippocampal gyrus, respectively. Recordings were performed with a 32-channel Easy EEG II (Cadwell, Kennewick, WA, U.S.A.) and sampled at 400 Hz with a bandwidth of 1–70 Hz over a minimum period of 20 min. The electrodes that recorded spikes (<80 ms) or sharp waves (80–200 ms) with a mean frequency greater than 1 spike/minute identified the spiking areas. Nonspiking areas were defined as those in which no spikes, sharp waves, or slow activity were detected by the electrodes. Photographs of the electrode locations were taken before removal of the grid and the spiking and nonspiking areas were identified prior to tissue excision. Tailored temporal lobectomy plus amigdalohippocampectomy were performed under electrocorticography guidance in all cases. After surgery, the spiking and nonspiking areas of the lateral neocortex and mesial structures were subjected to standard neuropathological assessment. Hippocampal sclerosis was observed in eight of the 11 patients (H48, H109, H123, H225, H239, H240, H241, and H247). It was characterized by neuronal loss, granule cell dispersion, and mossy fiber proliferation in the dentate gyrus, and by neuronal loss and gliosis in varying degree in the stratum pyramidale of the CA fields (cf., Arellano et al., 2004). In the remaining three patients (H65, H242, and H248) no pathological findings were observed in the resected tissue and the hippocampal formation exhibited apparently normal cytoarchitecture. The lateral neocortex was histologically normal in all cases.
Membrane protein identification and deglycosylation
The normal (nonspiking) areas of the lateral neocortex of patient H225 (female, 49-years old) were used for protein extraction. In addition, we used neocortical and hippocampal tissue from three adult Wistar rats sacrificed with an overdose of pentobarbital. Membranes were prepared from freshly isolated rat neocortex and hippocampus, and from normal human neocortex using differential centrifugation. Rat and human tissue were each homogenized, with the aid of a glass homogenizer, in a buffer solution (1 ml per 3 g of tissue) containing (in mM): 200 sucrose, 10 Tris, 10 HEPES, and 1 EDTA (pH 7.2 at 24°). The homogenate was centrifuged at 5,800 ×g for 10 min at 4°C. The supernatant was centrifuged at 48,000 ×g for 30 min at 4°C. The final pellet was resuspended in 0.5% SDS, 100-mM Tris (pH 7.6), 1% mercaptoethanol, 50-mM EDTA with protease inhibitors and stored at −80°C. Protein concentration was determined with a protein assay kit (Bio-Rad, Hercules, CA, U.S.A.) using bovine serum albumin as the standard. For deglycoslyation experiments, 20 μg of membrane protein were denatured by boiling 5 min in a solution containing 0.5% SDS, 100-mM Tris (pH 7.6), 1% mercaptoethanol, 50-mM EDTA and protease inhibitors. Membranes were incubated overnight at 37°C in 100 μl of 0.16% SDS, 0.7% Nonidet P 40, 100-mM Tris (pH 7.6), 1% mercaptoethanol, 50-mM EDTA, protease inhibitors, and 1 unit of N-glycosidase F (Roche, Mannheim, Germany). Enzymatic treatment was terminated by addition of electrophoresis sample buffer (see below). Control samples were processed similarly but incubation was carried out in the absence of N-glycosidase F. Prestained molecular weight markers (New England BioLabs, Beverly, MA, U.S.A.) and membrane protein samples were boiled in sample buffer (2% SDS, 13-mM Tris (pH 6.8), 10% glycerol, 0.1-M DTT (dithiothreitol) and 0.002% bromophenol blue] and then separated by SDS-polyacrylamide Gel Electrophoresis and transferred to polyvinylidene difluoride (PVDF) membranes (Immobilon P, Millipore, Billerica, MA, U.S.A.), using a Mini-Protean System (Bio-Rad) in transfer buffer [192-mM glycine, 25-mM Tris (pH 8.3) and 15% methanol]. The PVDF membrane was blocked in Tris-buffered saline (TBS)-milk [7% nonfat dry milk and 0.05% Tween-20 in TBS (pH 7.4)] for 1 h and then incubated overnight at 4°C in the same solution with the addition of T4 mouse monoclonal anti-NKCC (dilution 1:200) antibodies or rabbit anti-KCC2 (dilution 1:200) affinity purified antibodies. After three 10-min washes in TBS-Tween the membrane was incubated with secondary antibody (horseradish peroxidase-conjugated goat anti-mouse or anti-rabbit IgG, respectively; Jackson ImmunoResearch, West Grove, PA, U.S.A.) for 2 h at 24°C in TBS-Tween. After three washes in TBS, bound antibody was detected using an enhanced chemiluminescence assay (ECL, Amersham Biosciences, Buckinghamshire, United Kingdom).
Specificity of the NKCC and KCC2 antibodies
The NKCC antibody was generated against a fusion protein fragment encompassing the last 310 residues of the carboxy-terminus (S760–S1212) of the human colonic NKCC, and recognizes both NKCC1 and NKCC2 isoforms (Lytle et al., 1995). The hybridoma culture supernatant containing the monoclonal antibodies T4 was obtained from the Development Studies Hybridoma Bank maintained by the University of Iowa (Department of Biological Sciences, Iowa City, IA, U.S.A.) under contract N01-HD-7-3263 from the National Institute of Child Health and Human Development (NICHD). The monoclonal antibody (mAb) was purified by affinity chromatography as described previously (Alvarez-Leefmans et al., 2001). The antibody selectivity for NKCC detection has been characterized extensively, and has been shown to recognize NKCC proteins in a wide variety of cell types (Lytle et al., 1995; Maglova et al., 1998; McDaniel and Lytle, 1999) including rat hippocampal (Marty et al., 2002) and cultured cortical neurons (Sun and Murali, 1999), dorsal root ganglion cells, sensory axons and Schwann cells (Alvarez-Leefmans et al., 2001), astrocytes (Yan et al., 2001), and cultured oligodendrocytes (Wang et al., 2003). The T4 mAb is not NKCC-isoform specific. This mAb recognizes both NKCC1 and NKCC2 and any splice variants of these cotransporters preserving epitopes present in the last 310 residues of the carboxy (C) terminus. However, NKCC2 transcript and proteins are not present in the brain (Gamba et al., 1994; Delpire and Mount, 2002). Hence, T4 immunostaining in the brain is likely to reveal mainly, if not exclusively, NKCC1 and its splice variants with conserved C-terminus (Randall et al., 1997; Plotkin et al., 1997b; Vibat et al., 2001). The rabbit anti-KCC2 polyclonal antibody was generated against a purified fusion protein (B22) containing a 112-amino acid segment of the carboxyl terminus of the rat KCC2 (932–1043). The immune antiserum was purified by affinity chromatography. This antibody specifically recognizes a band of approximately 140-kDa glycoprotein detectable only within the central nervous system and in KCC2 transfected HEK-293 cells (Williams et al., 1999).