Death-Associated Protein Kinase Expression in Human Temporal Lobe Epilepsy David C. Henshall, PhD,1 Clara K. Schindler, BS,1 Norman K. So, MB, BChir, MRCP,2 Jing-Quan Lan, MD,1 Robert Meller, DPhil,1 and Roger P. Simon, MD1 Experimental and human data suggest programmed (active) cell death may contribute to the progressive hippocampal atrophy seen in patients with refractory temporal lobe epilepsy. Death-associated protein (DAP) kinase is a novel calcium/calmodulin-activated kinase that functions in apoptosis mediated by death receptors. Because seizure-induced neuronal death involves both death receptor activation and calcium, we examined DAP kinase expression, localization, and interactions in hippocampal resections from patients with intractable temporal lobe epilepsy (n ⴝ 10) and autopsy controls (n ⴝ 6). Expression and phosphorylation of DAP kinase was significantly increased in epilepsy brain compared with control. DAP kinase and DAP kinase–interacting protein 1 (DIP-1) localized to mitochondria in control brain, whereas levels of both were increased in the cytoplasm and microsomal (endoplasmic reticulum) fraction in epilepsy samples. Coimmunoprecipitation analysis showed increased DAP kinase binding to calmodulin, DIP-1, and the Fasassociated protein with death domain (FADD) in epilepsy samples. Finally, immunohistochemistry determined DAP kinase was coexpressed with DIP-1 in neurons. This study provides the first description of DAP kinase and DIP-1 in human brain and suggests DAP kinase is a novel molecular regulator of neuronal death in epilepsy. Ann Neurol 2004;55:485– 494 Experimentally evoked brief, repetitive, and/or spontaneous seizures, in addition to prolonged seizures (status epilepticus), have been shown to cause neuronal loss in brain.1–3 This may be epileptogenic4 and contribute to seizure-induced neurological impairments5 and the severity of subsequent epilepsy.6 Neuroimaging has corroborated these experimental data, whereby patients with suboptimally managed epilepsy experience progressive reductions in hippocampal volume7–10 and cognitive function.11,12 Accordingly, neuroprotective interventions may benefit patients after brain injuries known to precipitate epilepsy and as adjuncts to anticonvulsant therapy after certain seizures. Neuronal loss after seizures may involve programmed (active) cell death pathways.13,14 Several classes of cell death regulatory proteins are activated by seizures, including caspases, death receptors, and Bcl-2 family proteins, and molecular and pharmacological manipulations of these effectors can reduce postseizure neuronal death by as much as half.15–20 Accordingly, defining the relevant molecular pathways and identifying new cell death modulators is a priority for epilepsy management. Death-associated protein (DAP) kinase is a novel 160kDa calcium/calmodulin-dependent, death domain–containing serine/threonine kinase that has been implicated in the control of apoptosis.21 DAP kinase can be activated downstream of death receptors such as tumor necrosis factor receptor 1 (TNFR1) and Fas and has been implicated in premitochondrial events that promote caspase and Bcl-2–regulated apoptosis,22–24 although antiapoptotic functions have been demonstrated.23 In turn, expression of DAP kinase is negatively regulated by ubiquitination via its E3 ligase, DAP kinase interacting protein 1 (DIP-1).25 DAP kinase is present throughout the central nervous system during development, but adult expression is largely restricted to the resistant cornu ammonis 2 region of the hippocampus and dentate granule cells of the hippocampus.26 –28 Recent studies suggest DAP kinase may be a target in neurological diseases,27,29,30 and work by our group in a rodent epilepsy model has implicated DAP kinase in regulating seizure-induced neuronal death, showing DAP kinase overexpression in relation to injury-causing seizures and recruitment to the TNFR1 complex.28 DAP kinase’s molecular profile and participation in pathways activated by seizures makes it an attractive target in the setting of epileptic From the 1Robert S. Dow Neurobiology Laboratories, Legacy Research; and 2Oregon Comprehensive Epilepsy Program, Neurological Sciences Center, Portland, OR. Address correspondence to Dr Henshall, Robert S. Dow Neurobiology Laboratories, Legacy Clinical Research and Technology Center, 1225 NE 2nd Avenue, Portland, OR. E-mail: firstname.lastname@example.org Received Aug 20, 2003, and in revised form Nov 11. Accepted for publication Nov 11, 2003. Published online Feb 18, 2004, in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/ana.20001 © 2004 American Neurological Association Published by Wiley-Liss, Inc., through Wiley Subscription Services 485 brain injury. However, there have been no reports of DAP kinase involvement in human material in the setting of neurological disease. Accordingly, we examined the expression and activation of DAP kinase in the hippocampus obtained from epilepsy patients who had undergone surgical resection for intractable seizures. 1mm thickness prepared from the remaining sample for biochemical analysis. Autopsy control tissue (n ⫽ 6 per group) was obtained from the University of Maryland Brain and Tissue Bank (see Table for details). Hippocampal specimens were from individuals without known neurological disease and were also fresh-frozen, en bloc samples. Materials and Methods Human Brain Tissue Samples Western Blotting This study was approved by the Legacy Health System Institutional Review Board, and informed consent was obtained from all patients. Clinical data are shown in the Table. All patients (n ⫽ 10) were referred for surgical resection of the temporal lobe by an epileptologist after neurological assessment, videoelectroencephalogram recording, and magnetic resonance imaging/neuroimaging. Each patient was determined to have medically intractable epilepsy with a history of recurring seizures. However, no patient had experienced status epilepticus during the year in which their surgery was performed. Seizure frequency for each patient was typically in the range of 1 to 2 per week (range, 2–30 per month). All patients were taking anticonvulsant medication before surgery. All patients underwent left or right temporal lobe resection, and the hippocampus and in some cases adjacent temporal cortex were obtained. Specimens were immediately frozen in liquid nitrogen and stored at ⫺70°C until use. The hippocampus was separated from adjacent temporal cortex and was analyzed in its entirety without further microdissection of subfields. Specimens were first sectioned for immunohistochemistry and then coronal slabs of approximately Western blotting was performed as previously described.31 In brief, samples were homogenized and lysed in buffer containing a protease inhibitor cocktail and cleared by centrifugation. Protein (50g) then was separated using 12 to 15% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE), transferred to polyvinylidene difluoride membranes (BioRad, Hercules, PA) and incubated with monoclonal antibodies against DAP kinase or phospho308DAP kinase (1:1,000; Sigma, St. Louis, MO) and a custom-made rabbit polyclonal antibody to DIP-1 (also known as mind bomb32) that previously has been described.25 As controls, membranes were also immunoblotted with antibodies against ZIP kinase (Biomol, Plymouth, PA), and against phospho-p44/42 mitogen-activated protein kinase (p44/42MAPK) and phosphoglycogen synthase kinase-␤ (pGSK␤) (Cell Signaling Technology, Beverly, MA). Membranes then were incubated with horseradish peroxidase–conjugated secondary antibodies (1:2,000 dilution) followed by chemiluminescence detection (NEN Life Science Products, Boston, MA) and exposure to X-OMAT film (Kodak, Rochester, NY). Images were collected with a Dage 72 camera and gel bands were analyzed Table. Data for Controls and Epilepsy Patients Sample Age (yr) Sex Cause of Death PMI (hr) 50 53 15 13 20 23 M F M M M M Cardiac arrest Pulmonary thromboembolus Traffic accident Suicide (hanging) Gunshot wound (chest) Traffic accident 8 5 12 5 8 8 Control specimens C1 C2 C3 C4 C5 C6 Sample Age (yr) Epilepsy patients E1 57 E2 16 E3 24 E4 29 E5 E6 E7 E8 E9 E10 a 35 40 15 47 55 47 Sex Cause/Risk Factor Seizure Type MRI/Pathology Lobectomya F F F M Unknown Febrile seizures Encephalitis Febrile seizures and status epilepticus Febrile seizures Febrile seizures Febrile seizures Unknown Neocortical lesion Encephalitis CP CP CP CP MTS MTS MTS MTS R R R L CP/GTCS CP/GTCS CP/GTCS CP CP CP MTS MTS MTS MTS MTS MTS L L L R R L F F F F M M Side of hippocampal resection. (right or left). PMI ⫽ postmortem interval; CP ⫽ complex partial; GTCS ⫽ generalized tonic-clonic seizures; MRI ⫽ magnetic resonance imaging; MTS ⫽ mesial temporal sclerosis. 486 Annals of Neurology Vol 55 No 4 April 2004 using gel-scanning integrated optical density software (Bioquant, Nashville, TN). Subcellular Fractionation Subcellular fractionation was performed on a selection of control (C1, C2, C3) and epilepsy specimens (E2, E3, E4) according to previous methods with modifications.33,34 In brief, samples were homogenized in a mannitol/sucrose buffer containing a protease inhibitor cocktail. Mitochondria were separated from the homogenate by centrifugation at 10,000g followed by purification through a Percoll gradient at 41,000g. The crude cytosol fraction was centrifuged at 100,000g to separate microsomes and cytosol. Fractions were then subjected to immunoblot detection of DAP kinase and DIP-1 as described above. Fraction quality was verified with markers for the cytoplasm (␣-tubulin; Santa Cruz Biotechnology, Santa Cruz, CA), mitochondria (cytochrome IV oxidase; Molecular Probes, Eugene, OR; tBid; R & D Systems, Minneapolis, MN) and the endoplasmic reticulum (caspase12; Cell Signaling Technology). Immunoprecipitation Immunoprecipitation was performed as previously described.20,28 In brief, samples from the same specimens used for fractionation were homogenized in lysis buffer containing 1% NP-40 and then protein samples (0.5mg) were incubated overnight at 4°C with 2 to 5g of immunoprecipitation antibodies; anti-actin and anti-Bad (Santa Cruz Biotechnology), anti-calmodulin and anti-FADD (Upstate Biotechnology, Lake Placid, NY), or anti–DIP-1. Samples then were incubated with protein A/G agarose beads (Santa Cruz Biotechnology) for 1 hour at 4°C. The protein–bead complex was collected by centrifugation and boiled, and samples were subjected to Western blotting with anti–DAP kinase. Positive (50g whole-cell lysate) and negative (omitting the immunoprecipitation antibody) controls confirmed specificity. Also, we ensured IgG levels (antibody loading) and levels of the bait (calmodulin, FADD, DIP-1) were expressed and/or immunoprecipitated equivalently between control and epilepsy samples. Immunohistochemistry and DNA Fragmentation Analysis Immunohistochemistry and DNA fragmentation analysis was performed as previously described.28,31 In brief, brain sections (12m) were preblocked in 5% goat serum and then incubated with antibodies against DAP kinase (1:100), DIP-1 (1:500), or neuronal nuclear protein (NeuN; Chemicon, Temecula, CA). Sections then were washed and incubated in a 1:500 dilution of goat anti–mouse or anti–rabbit Cy3 or fluorescein isothiocyante (FITC; Jackson Immunoresearch, Plymouth, PA). Finally, sections were mounted in 4⬘,6 diamidino-2-phenylindole (DAPI; Vector Labs, Burlingame, CA) to assess nuclear morphology, and immunolabeling was studied using a Leica microscope equipped for epifluorescent illumination. Excitation/emission wavelengths were 340/425nm (blue), 500/550nm (green), and 580/ 630nm (red). Images were collected using an Optronics DEI-750 three-chip camera equipped with a BQ 8000 sVGA frame grabber and analyzed using an image analysis system (Bioquant). Analysis of cells exhibiting DNA fragmentation was performed using fluorescein-linked terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL; Roche Molecular Biochemicals, Indianapolis, IN) as previously described.28 TUNEL counts were performed on ten ⫻10 lens fields, and positive cells were further examined under ⫻100 lens magnification to verify and image nuclear features. For phenotype identification, 100 DAP kinase or DIP1–immunopositive cells from sections counterstained with NeuN were examined from three patient or control sections under ⫻40 lens magnification. Data Analysis Data are presented as mean ⫾ standard error of the mean. Data were analyzed using a Mann–Whitney U test (StatView software; SAS Institute, Cary, NC). Significance was accepted at p value less than 0.05. Results Increased DAP Kinase Expression in Hippocampi of Patients with Temporal Lobe Epilepsy Immunoblotting detected constitutive expression of DAP kinase at its predicted weight of 160kDa in whole-cell lysates from all six control brain samples (Fig 1A). Increased levels of DAP kinase were seen in 8 of 10 epilepsy patient samples, and this was statistically significant compared with control (see Fig 1A, B). Because DAP kinase function requires phosphorylation,35 we used a phosphospecific antibody to assess whether DAP kinase was activated. Phosphorylated DAP kinase was detected in 8 of 10 epilepsy specimens examined, and this was significantly different from controls in which phosphorylated DAP kinase was not detected (see Fig 1A, C). Last, we examined expression of DIP-1, the E3 ligase known to regulate expression of DAP kinase.25 We detected low-level constitutive expression of the approximately 110kDa form of DIP-1 in all control brain samples. DIP-1 levels were higher than control in 5 of 10 epilepsy samples, but this difference did not reach statistical significance (see Fig 1A and data not shown). Expression of ZIP Kinase Is Significantly Lower in Epilepsy Brain To verify the specific nature of the changes for DAP kinase, we examined expression of ZIP kinase,36 another member of the DAP kinase–related family. ZIP kinase was expressed in all control samples at approximately 55kDa (Fig 2A). Expression of ZIP kinase was lower in 9 of 10 epilepsy brain samples, and this was statistically significant compared with controls (see Fig 2A,B). A lower weight band was also apparent in epilepsy specimens, suggestive of proteolysis. Although the identity of this fragment is uncertain, ZIP kinase immunoblotting using hippocampal lysates from our ex- Henshall et al: DAP Kinase in TLE 487 tissue), we examined phosphorylation of two other proteins. Reprobing immunoblots from controls showed that phosphorylated p44/42 MAPK and GSK␤ could be detected in these samples (see Fig 2C). Although levels were quite variable, this was not directly related to postmortem interval (see Table). Subcellular Localization of Death-associated Protein Kinase and DIP-1 For subcellular distribution analysis, quality of fractionation was confirmed by the selective presence of ␣-tubulin in the cytoplasm, cleaved Bid (tBid) in mitochondria (epilepsy samples only), and caspase-12, a neuronally expressed endoplasmic reticulum–specific caspase,37 in the microsomal fraction (Fig 3A). In control brain, DAP kinase (see Fig 3B, C) and DIP-1 (see Fig 3D, E), were present almost exclusively within the mitochondrial fraction of the cell. In contrast, levels of both were significantly increased within the cytoplasm and microsomal fractions from epilepsy brain (see Fig 3B–E). Mitochondrial levels of DAP kinase and DIP-1 were not different between control and epilepsy samples (see Fig 3B–E). Fig 1. Death-associated protein (DAP) kinase expression in human brain. (A) Representative immunoblots from control (C) and epilepsy (E) hippocampi showing increased DAP kinase (DAPK) expression (top) and activation/phosphorylation (center) in epilepsy specimens. Molecular weight markers (kDa) are depicted to the left. Note, phospho(p)DAPK also detects a lower weight band (pDAPK*) at approximately 100kDa. DIP-1 levels in epilepsy samples were less noticeably different from control (bottom). Immunoblot quantification graphs below show that (B) DAP kinase and (C) phosphoDAP kinase levels are significantly higher in epilepsy (Epil, n ⫽ 10) specimens versus control (Con, n ⫽ 6). *p ⬍ 0.05, **p ⬍ 0.01 compared with control. perimental rat model28 showed the same weight band appears 24 to 72 hours after seizure cessation (data not shown). To ensure the lack of DAP kinase phosphorylation in controls was not caused by a general absence of phosphorylated proteins in such samples (eg, because of protracted collection times associated with autopsy 488 Annals of Neurology Vol 55 No 4 April 2004 Fig 2. ZIP kinase is downregulated in human epileptic brain. (A) Representative immunoblots from control and epilepsy hippocampi show that expression of ZIP kinase (ZIPK) is lower in epilepsy samples. (B) Graph quantifying ZIP kinase differences. *p ⬍ 0.05 versus control. (C) Detection of phosphorylated proteins in control samples. Control blots were reprobed with antibodies against phosphorylated p44/42 MAPK and GSK␤. Death-associated Protein Kinase Interacts with Calmodulin, FADD, and DIP-1 in Epilepsy Brain We next examined the interaction of DAP kinase with proteins related to its activation and function. In control brain, DAP kinase was not immunoprecipitated by calmodulin (Fig 4A). However, in epilepsy samples, we detected modest but significant binding of DAP kinase to calmodulin (see Fig 4A). DAP kinase binding to actin was approximately fourfold higher in epilepsy brain; however, sample variability precluded this difference reaching statistical significance (data not shown). Next, we tested whether DAP kinase interacts with the death receptor adaptor FADD, a phosphoprotein that links activated death receptors to caspase-8. We detected very little binding of these two proteins in control brain; however, in epilepsy brain, DAP kinase was found to strongly interact with FADD (see Fig 4B). Finally, we found DAP kinase was not present in control brain DIP-1 immunoprecipitates (see Fig 4C). However, in epilepsy brain we detected significant binding of DAP kinase to DIP-1 (see Fig 4C). Because DAP kinase interactions were consistently elevated in epilepsy samples, we performed an additional assay to ensure the specificity of these changes. Binding of 14-3-3 to the cell death promoter Bad (Bcl2–associated death protein) is normally higher in control than seizure hippocampi in the rat.17 Performing this same binding assay using our human material, we confirmed that 14-3-3 binding to Bad immunoprecipitates is higher in control human brain than epilepsy specimens (data not shown). Fig 3. Subcellular localization of DAP kinase and DIP-1. (A) Verification of fraction quality for cytoplasm (cyto), mitochondria (mito), and microsomes (micro). (B) Immunoblots showing DAP kinase localization in each fraction from control and epilepsy brain. (C) Graph quantifying differences in location of DAP kinase between control and epilepsy brain. (D) Immunoblots showing DIP-1 localization in each fraction from control and epilepsy brain. (E) Graph quantifying differences in location of DIP-1 between control and epilepsy brain. Data are from n ⫽ 3 per group. *p ⬍ 0.05 compared with control. Henshall et al: DAP Kinase in TLE 489 Death-associated Protein Kinase and DIP-1 Immunohistochemistry Finally, we used immunohistochemistry to determine which cells express DAP kinase and to support our coimmunoprecipitation findings. DAP kinase–expressing cells were detected in all epilepsy samples examined, particularly within the granule cells of the dentate gyrus (Fig 5C, D, G). Counterstaining using the neuronal phenotype marker NeuN showed 99% of cells expressing DAP kinase were neurons. Examples of DAP kinase– expressing neurons are shown in Figure 5E to H. We next examined DNA fragmentation in these sections. With a single exception, TUNEL-positive cells were detected on sections from each epilepsy patient (range, 0 – 4). However, average TUNEL counts were low (1.7 ⫾ 0.4) and were not significantly different from numbers in controls (1.0 ⫾ 0.3; range, 0 –3). Where present, they appeared as intensely stained, lobed, and/or blebbing, in line with features of a programmed cell death/apoptotic phenotype (see Fig 5B). No TUNEL-positive cells labeled for DAP kinase. DIP-1–expressing cells were more abundant in control and epilepsy brain than were DAP kinase–expressing cells (Fig 6). Counterstaining with NeuN showed that 97% of DIP-1–expressing cells were neurons in control brain, and 93% were neurons in epilepsy specimens (see Fig 6A–D). Although many DIP-1–expressing cells did not express detectable DAP kinase (data not shown), DAP kinase–expressing cells invariably colabeled for DIP-1 in both control (98%) and epilepsy (96%) sections examined (see Fig 6E–L). However, within individual cells in control and epilepsy brain the colocalization of these proteins was quite limited, with DIP-1 largely appearing in a nuclear distribution and the area immediately surrounding it, whereas DAP kinase appeared in the soma region outside the nucleus (see Fig 6E–L). Where colocalization was present this usually appeared as aggregates within the cytoplasm. Fig 4. Death-associated protein (DAP) kinase binding interactions. Representative immunoblots (WB) showing increased DAP kinase binding to (A) calmodulin (CaM), (B) FADD, and (C) DIP-1 immunoprecipitates in epilepsy brain compared with control. Graphs below each panel show quantification of these differences. Data are from n ⫽ 3 per group. *p ⬍ 0.05, **p ⬍ 0.01 compared with control. 490 Annals of Neurology Vol 55 No 4 April 2004 Discussion With evidence accumulating that temporal lobe epilepsy is a progressive disease for brain injury, the impetus to understand the molecular determinants of seizure-induced neuronal loss intensifies. This study provides the first description to our knowledge of DAP kinase, a novel regulator of cell death, and its E3 ligase DIP-1 in human brain. Our data show that DAP kinase is overexpressed and likely activated in hippocampi obtained from patients with intractable epilepsy. Because DAP kinase is able to regulate apoptosis in a variety of cell types including neurons, these data may have significant implications for the treatment of seizure-induced neuronal death and the design of neuroprotective adjuncts to treat epilepsy or epilepsyprecipitating brain injuries. Seizures, particularly when prolonged, cause selective Fig 5. Death-associated protein (DAP) kinase immunohistochemistry. (A) DAPI-stained, low-power (⫻2.5 lens) field photomicrograph from an epilepsy en bloc specimen showing visible blade of the dentate gyrus. (B) Examples of TUNEL-positive nuclei from epilepsy patient sections. Note lobed/blebbing features suggesting programmed cell death. (C) Single DAP kinase–expressing cell (labeled with Cy3) from the hilar region of the dentate gyrus of an epilepsy patient. (D) Panel showing expression of DAP kinase in a group of cells within the dentate granule cell layer. (E–H) Phenotype of DAP kinase–expressing cells in epilepsy brain. Counterstaining of sections with NeuN (labeled with FITC) shows DAP kinase is expressed by neurons. Arrows indicate immunostained cells. Scale bar ⫽ 300m (A), 6m (B), 10m (C, D), 12m (E–H). Fig 6. DIP-1 immunohistochemistry and coexpression in DAPK-positive cells. (A–D) Immunohistochemistry showing phenotype of DIP-1–expressing cells in epilepsy brain. Counterstaining of sections with NeuN (labeled with Cy3) shows DIP-1 (labeled with FITC) is expressed by neurons. Arrows indicate neurons staining for DIP-1, whereas an arrowhead identifies a DIP-1–expressing cell that is not NeuN labeled. (E–L) Colocalization of DIP-1 with DAP kinase in control (middle panels) and epilepsy (bottom panels) brain. Note DAP kinase largely stains the cytoplasm, whereas DIP-1 appears restricted to a nuclear area. Scale bar ⫽ 12m (A–D), 8m (E–L). degeneration of neurons within the hippocampus that is mediated in part by induction of programmed (active) cell death pathways.13,14 Although studies have challenged whether cell loss is a requirement for the development of spontaneous seizures (epilepsy),6,38 – 40 new data demonstrate that seizure-induced neuronal loss can directly cause epilepsy in certain cases.4 Neuronal loss also may underlie seizure-induced cognitive dysfunction5 and may exacerbate seizure severity in experimental epilepsy.6 Accordingly, targeting apoptosis- Henshall et al: DAP Kinase in TLE 491 regulating proteins to prevent seizure-induced neuronal death as an adjunct to anticonvulsive treatment may provide a means to reduce the impact of seizures on brain. Although still controversial,39,41,42 several regulators of programmed cell death, including caspases and Bcl-2 family proteins, appear to be involved in seizure-induced neuronal death.16,17,19,33 The initiation site for these pathways may be surface-expressed death receptors such as TNFR1, which rapidly assemble after seizures, and recruit adaptor proteins, such as FADD, and initiator caspases.20,28 An ideal death effector target therefore might be one responsive to both the inevitable calcium overload in the wake of seizure activity and apoptotic pathways mediated by death receptors. DAP kinase fits this profile because its activity is stimulated by calcium/calmodulin and it regulates apoptosis in response to TNF/Fas.22,35 Our analysis of hippocampi from patients with intractable temporal lobe epilepsy provides the first report of DAP kinase upregulation in a human neurological disease. Two additional experiments, showing increased DAP kinase phosphorylation and binding to calmodulin, support DAP kinase being functionally active in these specimens. We base this interpretation on evidence that stimulation of the full-length (wild type) DAP kinase by calcium/calmodulin increases DAP kinase phosphorylation, phosphorylation of DAP kinase substrates, and apoptosis.35,43 In most,22,24,29,35 but not all,23,25 systems, DAP kinase is proapoptotic, and experimental models show its expression is increased by cell death–inducing but not benign seizures.28 Accordingly, DAP kinase may be a molecular determinant of the progressive hippocampal damage seen in temporal lobe epilepsy patients. In turn, DAP kinase might be a therapeutic target to treat neuronal injury caused by brief or prolonged seizures. Promisingly, DAP kinase inhibition can abrogate neuronal death in response to stimuli known to be generated by seizures.29,44 Because DAP kinase may contribute to injury outcome after stroke,27,30 it also might be a target relevant to other neurological diseases. DAP kinase has been reported previously to localize to the microfilaments of the cytoskeleton and mediate structural rearrangements of the cell during apoptosis and autophagy.35,45 Although DAP kinase binds actin in the hippocampus (our data and Henshall and colleagues28), we found it only within the mitochondrial compartment in control brain. These data might reflect mitochondria as a site of (presumably protective) DAP kinase sequestration. Indeed, other proapoptotic proteins are sequestered to mitochondria,46,47 being released in response to apoptotic stimuli. Alternatively, DAP kinase may target mitochondrial proteins, identification of which could lend insight into its in vivo substrates, which remain largely unknown. Surprisingly, mitochondria were not the site of DAP kinase 492 Annals of Neurology Vol 55 No 4 April 2004 accumulation in epilepsy brain, and whereas the observed cytoplasmic increases might have been predicted from previous work on its localization35 its elevation in the microsomal fraction was unexpected. The endoplasmic reticulum, a major constituent of the microsomal fraction, is a potent initiator of apoptosis when stressed,48 and data have emerged implicating this organelle in seizure-induced neuronal death.49,50 Indeed, we have shown TNFR activating factor 2 and apoptosis signal–regulating kinase 1, which both mediate endoplasmic reticulum stress pathways,51,52 are activated and/or recruited to the TNFR1 pathway after experimental seizures.20 The function of DAP kinase within the endoplasmic reticulum is unclear, but myosin light chain kinase, which bears substrate and functional similarity has been shown to coordinate TNFR1 movement within cells,53 and TNFR1 has been shown to traffic to the microsomes during apoptosis.54 Because we observed increased binding of DAP kinase to FADD in the same samples, it is possible that a DAP kinase–containing death signaling complex (DISC) is forming in the cytoplasm or microsomal fraction. Indeed, this occurs after experimental seizures,28 and cleaved caspase-8 can be detected in our epilepsy patient specimens (D.C. Henshall, C.K. Schindler, unpublished observation). We might further speculate that FADD may be a substrate for DAP kinase. FADD is a phosphoprotein, and its phosphorylated form interacts with activated Fas receptors,55 suggesting DAP kinase may act on FADD to influence Fas signaling in brain. Further studies are now required to elucidate the significance of these localization changes in terms of their relation to DISC formation, substrates/targets and specificity to the environment of the epileptic brain. DIP-1 was identified recently as the in vivo regulator of DAP kinase levels through its ability as an E3 ligase to ubiquitinate and thereby downregulate DAP kinase expression.25 Here, we provide the first description of DIP-1 in human brain, showing a close relationship with DAP kinase in which they were coexpressed in neurons, similarly distributed in subcellular compartments, and could be coimmunoprecipitated. However, because DAP kinase levels were raised in epileptic brain despite their binding, the function of DIP-1 may be compromised in these patients. Nevertheless, promoting DIP-1 function could yet provide a means to target the DAP kinase pathway. Certain limitations should be considered in the interpretation of our findings. Although our autopsy hippocampi were well matched for sex and age, they would not reflect anticonvulsant medication or surgical/anesthesia effects. However, control animals in our experimental seizure model undergo a surgical manipulation for placement of cannulas/electrodes and receive anticonvulsant, and this does not activate cell death pathways.16,17 Additional validation for our find- ings comes from internal controls (eg, lower ZIP kinase expression in epilepsy brain compared with control) and previous work showing expected levels of constitutive and nonconstitutive genes in autopsy brain.31 Second, increased phosphorylation of DAP kinase within its calmodulin regulatory domain can be inhibitory,43 implying increased DAP kinase phosphorylation may not be as reliable a marker of activity as calmodulin binding. Finally, antiapoptotic functions of DAP kinase have been reported.23,25 Our data do not preclude such a role, which could underlie the infrequency of TUNEL in epileptic brain reported here and previously.31 If this is not the case, single time point sampling may have underestimated TUNEL, or the chronically epileptic brain may counter DAP kinase by adjusting its repertoire of antiapoptotic proteins. Indeed, Bcl-2 levels are increased in epilepsy specimens,31 and Bcl-2 can inhibit DAP kinase–induced neuronal death.22 In conclusion, we provide the first evidence for the overexpression and activation of DAP kinase in a human neurological disease, epilepsy. DAP kinase’s place within proapoptotic pathways activated by seizures mean such changes could underlie the molecular drive to progressive neuronal loss in patients with refractory epilepsy. Accordingly, these data lend insight into both the impact of epilepsy on brain and strategies for the treatment of injurious seizures. This work was supported by grants from the NIH (National Institute of Neurological Disorders and Stroke, NS39016, D.C.H., R.P.S.; NS41935, D.C.H.). We thank Drs Rosenban and Abtin for surgical collection of specimens and the University of Maryland Brain and Tissue bank for autopsy specimens. We also thank Dr P. J. Gallagher for the DIP-1 antibody. References 1. Cavazos JE, Das I, Sutula TP. Neuronal loss induced in limbic pathways by kindling: evidence for induction of hippocampal sclerosis by repeated brief seizures. J Neurosci 1994;14: 3106 –3121. 2. Bengzon J, Kokaia Z, Elmer E, et al. Apoptosis and proliferation of dentate gyrus neurons after single and intermittent limbic seizures. Proc Natl Acad Sci USA 1997;94:10432–10437. 3. Bouilleret V, Loup F, Kiener T, et al. Early loss of interneurons and delayed subunit-specific changes in GABA(A)-receptor expression in a mouse model of mesial temporal lobe epilepsy. Hippocampus 2000;10:305–324. 4. Sayin U, Osting S, Hagen J, et al. Spontaneous seizures and loss of axo-axonic and axo-somatic inhibition induced by repeated brief seizures in kindled rats. J Neurosci 2003;23: 2759 –2768. 5. Kotloski R, Lynch M, Lauersdorf S, et al. Repeated brief seizures induce progressive hippocampal neuron loss and memory deficits. Prog Brain Res 2002;135:95–110. 6. Zhang X, Cui SS, Wallace AE, et al. Relations between brain pathology and temporal lobe epilepsy. J Neurosci 2002;22: 6052– 6061. 7. Kalviainen R, Salmenpera T, Partanen K, et al. Recurrent seizures may cause hippocampal damage in temporal lobe epilepsy. Neurology 1998;50:1377–1382. 8. Tasch E, Cendes F, Li LM, et al. Neuroimaging evidence of progressive neuronal loss and dysfunction in temporal lobe epilepsy. Ann Neurol 1999;45:568 –576. 9. Bernasconi A, Tasch E, Cendes F, et al. Proton magnetic resonance spectroscopic imaging suggests progressive neuronal damage in human temporal lobe epilepsy. Prog Brain Res 2002; 135:297–304. 10. Briellmann RS, Berkovic SF, Syngeniotis A, et al. Seizureassociated hippocampal volume loss: a longitudinal magnetic resonance study of temporal lobe epilepsy. Ann Neurol 2002; 51:641– 644. 11. Jokeit H, Ebner A. Long term effects of refractory temporal lobe epilepsy on cognitive abilities: a cross sectional study. J Neurol Neurosurg Psychiatry 1999;67:44 –50. 12. Hermann BP, Seidenberg M, Bell B. The neurodevelopmental impact of childhood onset temporal lobe epilepsy on brain structure and function and the risk of progressive cognitive effects. Prog Brain Res 2002;135:429 – 438. 13. Bengzon J, Mohapel P, Ekdahl CT, et al. Neuronal apoptosis after brief and prolonged seizures. Prog Brain Res 2002;135: 111–119. 14. Liou AKF, Clark RS, Henshall DC, et al. To die or not to die for neurons in ischemia, traumatic brain injury and epilepsy: a review on the stress-activated signaling pathways and apoptotic pathways. Prog Neurobiol 2003;69:103–142. 15. Morrison RS, Wenzel HJ, Kinoshita Y, et al. Loss of the p53 tumor suppressor gene protects neurons from kainate-induced cell death. J Neurosci 1996;16:1337–1345. 16. Henshall DC, Bonislawski DP, Skradski SL, et al. Cleavage of Bid may amplify caspase-8-induced neuronal death following focally evoked limbic seizures. Neurobiol Dis 2001;8:568 –580. 17. Henshall DC, Araki T, Schindler CK, et al. Activation of Bcl2-associated death protein and counter-response of Akt within cell populations during seizure-induced neuronal death. J Neurosci 2002;22:8458 – 8465. 18. Roy M, Hom JJ, Sapolsky RM. HSV-mediated delivery of virally derived anti-apoptotic genes protects the rat hippocampus from damage following excitotoxicity, but not metabolic disruption. Gene Ther 2002;9:214 –219. 19. Narkilahti S, Nissinen J, Pitkanen A. Administration of caspase 3 inhibitor during and after status epilepticus in rat: effect on neuronal damage and epileptogenesis. Neuropharmacology 2003;44:1068 –1088. 20. Shinoda S, Skradski SL, Araki T, et al. Formation of a tumour necrosis factor receptor 1 molecular scaffolding complex and activation of apoptosis signal-regulating kinase 1 during seizureinduced neuronal death. Eur J Neurosci 2003;17:2065–2076. 21. Cohen O, Kimchi A. DAP-kinase: from functional gene cloning to establishment of its role in apoptosis and cancer. Cell Death Differ 2001;8:6 –15. 22. Cohen O, Inbal B, Kissil JL, et al. DAP-kinase participates in TNF-alpha- and Fas-induced apoptosis and its function requires the death domain. J Cell Biol 1999;146:141–148. 23. Jin Y, Blue EK, Dixon S, et al. Identification of a new form of death-associated protein kinase that promotes cell survival. J Biol Chem 2001;276:39667–39678. 24. Raveh T, Droguett G, Horwitz MS, et al. DAP kinase activates a p19ARF/p53-mediated apoptotic checkpoint to suppress oncogenic transformation. Nat Cell Biol 2001;3:1–7. 25. Jin Y, Blue EK, Dixon S, et al. A DAPK interacting protein, DIP-1, is an E3 ubiquitin ligase that promotes TNF-induced apoptosis and regulates the cellular levels of DAPK. J Biol Chem 2002;277:46980 – 46986. Henshall et al: DAP Kinase in TLE 493 26. Sakagami H, Kondo H. Molecular cloning and developmental expression of a rat homologue of death-associated protein kinase in the nervous system. Brain Res Mol Brain Res 1997;52: 249 –256. 27. Yamamoto M, Takahashi H, Nakamura T, et al. Developmental changes in distribution of death-associated protein kinase mRNAs. J Neurosci Res 1999;58:674 – 683. 28. Henshall DC, Araki T, Schindler CK, et al. Expression of death-associated protein kinase and recruitment to the tumor necrosis factor signaling pathway following brief seizures. J. Neurochem. 2003;86:1260 –1270. 29. Pelled D, Raveh T, Riebeling C, et al. Death-associated protein (DAP) kinase plays a central role in ceramide-induced apoptosis in cultured hippocampal neurons. J Biol Chem 2002;277: 1957–1961. 30. Schumacher AM, Velentza AV, Watterson DM, et al. DAPK catalytic activity in the hippocampus increases during the recovery phase in an animal model of brain hypoxic-ischemic injury. Biochim Biophys Acta 2002;1600:128 –137. 31. Henshall DC, Clark RS, Adelson PD, et al. Alterations in bcl-2 and caspase gene family protein expression in human temporal lobe epilepsy. Neurology 2000;55:250 –257. 32. Itoh M, Kim C-H, Palardy G, et al. Mind bomb is a ubiquitin ligase that is essential for efficient activation of notch signaling by delta. Dev Cell 2003;4:67– 82. 33. Henshall DC, Bonislawski DP, Skradski SL, et al. Formation of the Apaf-1/cytochrome c complex precedes activation of caspase-9 during seizure-induced neuronal death. Cell Death Differ 2001;8:1169 –1181. 34. Chen D, Minami M, Henshall DC, et al. Upregulation of mitochondrial base-excision repair capability within rat brain after brief ischemia. J Cereb Blood Flow Metab 2003;23:88 –98. 35. Cohen O, Feinstein E, Kimchi A. DAP-kinase is a Ca2⫹/ calmodulin-dependent, cytoskeletal-associated protein kinase, with cell death-inducing functions that depend on its catalytic activity. EMBO J 1997;16:998 –1008. 36. Kawai T, Matsumoto M, Takeda K, et al. ZIP kinase, a novel serine/threonine kinase which mediates apoptosis. Mol Cell Biol 1998;18:1642–1651. 37. Nakagawa T, Zhu H, Morishima N, et al. Caspase-12 mediates endoplasmic-reticulum-specific apoptosis and cytotoxicity by amyloid-beta. Nature 2000;403:98 –103. 38. Andre V, Ferrandon A, Marescaux C, et al. Vigabatrin protects against hippocampal damage but is not antiepileptogenic in the lithium-pilocarpine model of temporal lobe epilepsy. Epilepsy Res 2001;47:99 –117. 39. Ebert U, Brandt C, Loscher W. Delayed sclerosis, neuroprotection, and limbic epileptogenesis after status epilepticus in the rat. Epilepsia 2002;43:86 –95. 40. Brandt C, Potschka H, Loscher W, et al. N-methyl-D-aspartate receptor blockade after status epilepticus protects against limbic brain damage but not against epilepsy in kainate model of temporal lobe epilepsy. Neuroscience 2003;118:727–740. 41. Fujikawa DG, Ke X, Trinidad RB, et al. Caspase-3 is not activated in seizure-induced neuronal necrosis with internucleosomal DNA cleavage. J Neurochem 2002;83:229 –240. 494 Annals of Neurology Vol 55 No 4 April 2004 42. Puig B, Ferrer I. Caspase-3-associated apoptotic cell death in excitotoxic necrosis of the entorhinal cortex following intraperitoneal injection of kainic acid in the rat. Neurosci Lett 2002; 321:182–186. 43. Shohat G, Spivak-Kroizman T, Cohen O, et al. The proapoptotic function of death-associated protein kinase is controlled by a unique inhibitory autophosphorylation-based mechanism. J Biol Chem 2001;276:47460 – 47467. 44. Stoica BA, Movsesyan VA, Lea PMt, et al. Ceramide-induced neuronal apoptosis is associated with dephosphorylation of Akt, BAD, FKHR, GSK-3beta, and induction of the mitochondrialdependent intrinsic caspase pathway. Mol Cell Neurosci 2003; 22:365–382. 45. Inbal B, Bialik S, Sabanay I, et al. DAP kinase and DRP-1 mediate membrane blebbing and the formation of autophagic vesicles during programmed cell death. J Cell Biol 2002;157: 455– 468. 46. Noshita N, Sugawara T, Fujimura M, et al. Manganese superoxide dismutase affects cytochrome c release and caspase-9 activation after transient focal cerebral ischemia in mice. J Cereb Blood Flow Metab 2002;21:557–567. 47. Cao G, Luo Y, Nagayama T, et al. Cloning and characterization of rat caspase-9: implications for a role in mediating caspase-3 activation and hippocampal cell death after transient cerebral ischemia. J Cereb Blood Flow Metab 2002;22: 534 –546. 48. Ferri KF, Kroemer G. Organelle-specific initiation of cell death pathways. Nat Cell Biol 2001;3:E255–E263. 49. Pelletier MR, Wadia JS, Mills LR, et al. Seizure-induced cell death produced by repeated tetanic stimulation in vitro: possible role of endoplasmic reticulum calcium stores. J Neurophysiol 1999;81:3054 –3064. 50. Kitao Y, Ozawa K, Miyazaki M, et al. Expression of the endoplasmic reticulum molecular chaperone (ORP150) rescues hippocampal neurons from glutamate toxicity. J Clin Invest 2001; 108:1439 –1450. 51. Yoneda T, Imaizumi K, Oono K, et al. Activation of caspase12, an endoplastic reticulum (ER) resident caspase, through tumor necrosis factor receptor-associated factor 2-dependent mechanism in response to the ER stress. J Biol Chem 2001; 276:13935–13940. 52. Nishitoh H, Matsuzawa A, Tobiume K, et al. ASK1 is essential for endoplasmic reticulum stress-induced neuronal cell death triggered by expanded polyglutamine repeats. Genes Dev 2002; 16:1345–1355. 53. Jin Y, Atkinson SJ, Marrs JA, et al. Myosin ii light chain phosphorylation regulates membrane localization and apoptotic signaling of tumor necrosis factor receptor-1. J Biol Chem 2001; 276:30342–30349. 54. Storey H, Stewart A, Vandenabeele P, et al. The p55 tumour necrosis factor receptor TNFR1 contains a trans-Golgi network localization signal in the C-terminal region of its cytoplasmic tail. Biochem J 2002;366:15–22. 55. Kischkel FC, Hellbardt S, Behrmann I, et al. Cytotoxicitydependent APO-1 (Fas/CD95)-associated proteins form a death-inducing signaling complex (DISC) with the receptor. EMBO J 1995;14:5579 –5588.