Chromogranins as markers of altered hippocampal circuitry in temporal lobe epilepsy.код для вставкиСкачать
Chromogranins as Markers of Altered Hippocampal Circuitry in Temporal Lobe Epilepsy Susanne Pirker, MD,1,2 Thomas Czech, MD,2 Christoph Baumgartner, MD,3 Hans Maier, MD,4 Klaus Novak, MD,2 Sabine Fürtinger, MSc,1 Reiner Fischer-Colbrie, PhD,1 and Günther Sperk, PhD1 Chromogranins are polypeptides which are widely expressed in the central nervous system. They are stored in dense core vesicles of nerve terminals, from where they are released upon stimulation. Using immunocytochemistry, we investigated the distribution of chromogranin A, chromogranin B, secretoneurin, and, for comparison, dynorphin in hippocampal specimens removed at routine surgery from patients with drug-resistant mesial temporal lobe epilepsy and in autopsy tissues from nonneurologically deceased subjects. In post mortem controls (n ⴝ 21), immunoreactivity for all 4 peptides (most prominently for chromogranin B and dynorphin) was observed in the terminal field of mossy fibers. For chromogranins, staining was observed also in sectors CA1 to CA3 and in the subiculum. Chromogranin B immunoreactivity was found in the inner molecular layer of the dentate gyrus, the area of terminating associational-commissural fibers. Secretoneurin and dynorphin immunoreactivity labeled the outer molecular layer and the stratum lacunosum moleculare of sectors CA1 to CA3, where projections from the entorhinal cortex terminate. In specimens with Ammon’s horn sclerosis (n ⴝ 25), staining for all 3 chromogranins and for dynorphin was reduced in the hilus of the dentate gyrus. Instead, intense staining was observed in the inner molecular layer, presumably delineating terminals of sprouted mossy fibers. Specimens obtained from temporal lobe epilepsy patients without Ammon’s horn sclerosis (n ⴝ 4) lacked this pronounced rearrangement of mossy fibers. In the stratum lacunosum moleculare of sector CA1, secretoneurin and dynorphin immunoreactivity was reduced in sclerotic, but not in nonsclerotic, specimens, paralleling the partial loss of fibers arising from the entorhinal cortex. Instead, presumably sprouted secretoneurin-immunoreactive fibers were found in the outer dentate molecular layer in sclerotic specimens. These changes in staining patterns for chromogranins and dynorphin mark profound plastic and functional rearrangement of hippocampal circuitry in temporal lobe epilepsy. Ann Neurol 2001;50:216 –226 More than a century ago, the selective loss of hippocampal neurons, termed Ammon’s horn sclerosis, was described in mesial temporal lobe epilepsy (TLE).1,2 Typically, neurons of the hippocampal subfields CA1 and CA3 and the hilus of the dentate gyrus become severely damaged, whereas other neurons, such as granule cells and CA2 neurons, are relatively spared. At the same time, mossy fibers, the axons of granule cells, undergo pronounced reorganization as revealed by specific labeling with Timm’s stain or dynorphinlike immunoreactivity (IR).3–5 Mossy fibers normally innervate neurons in the hilus of the dentate gyrus and in the stratum lucidum of CA3. Presumably due to a partial loss of granule cells and/or of postsynaptic targets, they appear to be reduced in TLE. Instead, a prominent innervation of the inner molecular layer of the dentate gyrus by collaterals of mossy fibers is observed.3– 6 The relevance of this “mossy fiber reorganization” for the pathophysiology of TLE is still uncertain. It has been suggested that “self-innervation” of granule cells by their axons could contribute fundamentally to the increased excitability of the hippocampal formation.7 Indeed, synaptic contacts between sprouted mossy fiber terminals and dendrites of granule cells have been found by electron microscopy.8 –10 Therefore, neurotransmitters released from sprouted mossy fiber terminals may act on granule cell dendrites and may have an important impact on the altered signal transduction in the epileptic hippocampus. The main transmitter of mossy fibers is glutamate.11 The neuropeptide dynorphin is a cotransmitter in these From the 1Department of Pharmacology, University of Innsbruck, Innsbruck, 2Department of Neurosurgery, and 3Department of Neurology, University of Vienna, Vienna, and 4Department of Pathology, University of Innsbruck, Innsbruck, Austria. Published online May 25, 2001. Received Dec 5, 2000, and in revised form Apr 2, 2001. Accepted for publication Apr 2, 2001. 216 © 2001 Wiley-Liss, Inc. Address correspondence to Dr Sperk, Department of Pharmacology, Peter-Mayr-Strasse 1a, A-6020 Innsbruck, Austria. E-mail: firstname.lastname@example.org neurons.12 It is present in normal mossy fibers and those sprouted to the inner molecular layer of TLE patients.4,9,13 Also, chromogranins have been identified in mossy fibers of rats and humans.14 –17 Chromogranins are a family of acidic polypeptides, which are widely distributed in nervous and endocrine tissues.14 They comprise chromogranin A (CgA), CgB, and secretogranin II and were originally isolated from secretory granules of chromaffin cells in the adrenal medulla. In the nervous system, they are stored in large dense core vesicles and processed by peptidases to numerous smaller peptides,18 which are released upon nerve stimulation.14 Peptides derived from chromogranins may exert neurotransmitter-like or trophic functions.14 Thus, for secretoneurin cleaved from secretogranin II,19 dopamine release–enhancing and chemoattractant actions have been established.14,20 In the kainic acid model of TLE in rats, conspicuous changes in the gene expression of chromogranin mRNAs have been observed.21,22 We investigated by immunocytochemistry the distribution of CgA, CgB, and secretoneurin in hippocampal specimens from patients with drug-resistant TLE with and without hippocampal sclerosis and in post mortem tissue from nonneurologically deceased subjects as controls. Our aim was to study chromogranin-IR as an immunocytochemical marker of neuronal elements of the hippocampal circuitry, notably of mossy fibers. Changes in distribution were compared with changes in dynorphin-IR, an established immunocytochemical marker for mossy fibers.4,9 Specimens from four TLE patients without Ammon’s horn sclerosis were included to investigate the relationship between neuronal cell loss and the observed anatomical changes. region. Patients without obvious Ammon’s horn sclerosis were defined as nonsclerotic. They suffered from cortical dysplasia (n ⫽ 1) or cryptogenic TLE (n ⫽ 3). Mean age [⫾ standard error of mean (SEM)] at surgery was 34.5 ⫾ 1.9 years. Twelve females and 17 males were included in the study. The epileptic focus was on the left side in 14 patients and on the right side in 15 patients. Mean duration of epilepsy was 21.4 ⫾ 2.2 and 14.8 ⫾ 5.2 years in patients with and without hippocampal sclerosis, respectively. All patients were taking antiepileptic drugs in mono- or polytherapy. The most frequent antiepileptic drugs were carbamazepine (83%), lamotrigine (28%), valproate (24%), clobazam (21%), and vigabatrin (14%). As control tissues, 21 hippocampi from patients without known history of neurological or psychiatric disease were obtained at routine autopsy (mean age 57.7 ⫾ 3.5 years). Each brain was studied by a neuropathologist, to confirm the absence of a brain lesion. The time from death to fixation of brain specimens ranged from 8 to 36 hours (mean 17.2 ⫾ 3.8 hours). Five cases were females and 16 males. Eleven specimens were taken from the right hemisphere, 5 from the left. In 5 cases, the side investigated was unknown. Causes of death were pneumonia, lung cancer, pharyngeal cancer, laryngeal cancer, breast cancer, liver cancer, liver cirrhosis, melanoma, myocardial infarction, pulmonary embolism, cardiovascular arrest, lymphoma, tracheal cancer, stomach cancer, or renal failure. Surgical specimens of the hippocampal body (middle segment) from TLE patients and autopsy tissue were sectioned perpendicularly to the hippocampal axis into 5-mm-thick blocks and immersed in 4% paraformaldehyde, 50 mM phosphate-buffered saline (PBS, pH 7.4) for 4 to 5 days. After stepwise immersion in sucrose (range 5%–20%) over 2 days, specimens were frozen and stored at –70°C. Microtom sections (40 m) were collected and stored in PBS/0.1% sodium azide at 5°C. Immunohistochemistry Patients and Methods Presurgical evaluation of TLE patients included detailed clinical examinations, prolonged video-electroencephalographic monitoring with scalp and sphenoidal electrodes, and neuropsychological tests, including a Wada test. Neuroimaging studies comprised high-resolution magnetic resonance imaging (MRI) and single-photon emission computed tomography. When the epileptogenic zone was localized to the medial temporal region, patients were referred to selective amygdalohippocampectomy or anteromedial temporal lobe resection. Approval for the study was obtained from the University of Vienna, and informed consent was obtained from patients providing specimens. Surgical specimens of all patients were examined by routine pathology, and in accordance with presurgical examination, they were grouped into those with hippocampal sclerosis (n ⫽ 25) and those without (n ⫽ 4). Specimens with hippocampal sclerosis originated from patients with selectively damaged hippocampus as assessed by MRI and confirmed by the presence of distinct hippocampal sclerosis in the neuropathological examination.23 In these patients, seizures presumably arose from the medial temporal Monoclonal CgA antiserum (clone LK2H10) was purchased from Boehringer-Mannheim (Mannheim, Germany). Antisera for CgB and secretoneurin were raised in rabbits against synthetic peptides corresponding to amino acids 554 –564 of human CgB24 and 154 –186 of rat secretogranin II (differing only in 1 position from the human sequence),25 respectively, and coupled to keyhole-limpet hemocyanin. The antisera recognized the respective chromogranin and processing products but did not cross-react with other chromogranins or a variety of neuropeptides.15,19 The rabbit antisera for dynorphin A(1– 8) and ␣-neoendorphin were donated by Dr. P. Ciofi (INSERM U 156, Neuroendocrinologie Cellulaire, Lille, France). Immunocytochemical staining was abolished by 50 M dynorphin(1– 8) and ␣-neoendorphin, respectively, but not dynorphin A(1–18), met-enkephalin, and leuenkephalin or the other respective opioid peptides (not shown). Antisera did not cross-react with a variety of other neuropeptides.26 Incubations were performed on free-floating sections. For dynorphin and ␣-neoendorphin immunohistochemistry, sections were pretreated with target retrieval solution (pH 6.0; Dako, Vienna, Austria) at 70°C and then at room tempera- Pirker et al: Chromogranins in TLE 217 ture for 20 minutes each. In all instances, the avidin-biotinperoxidase method was used according to the supplier’s protocol (Vectastain Standard ABC kit; Vector, Burlingame, CA). Only for ␣-neoendorphin was a horseradish peroxidase–coupled secondary antibody used (P0448, 1:250; Dako). Depending on the host of the secondary antibody, sections were preincubated for 90 minutes with 10% normal goat serum (Margaritella, Vienna, Austria) when processed for dynorphin, ␣-neoendorphin, CgB, and secretoneurin or with 10% normal horse serum (GIBCO, Paisley, UK) in 50 mM Tris-HCl–buffered saline (pH 7.2, TBS) for CgA immunohistochemistry. This step was followed by incubation with the respective primary antibody (dynorphin 1:20,000, ␣-neoendorphin 1:1,000, CgA 1:2,500, CgB 1:15,000, secretoneurin 1:5,000) for 48 to 72 hours at 4°C in the same buffer. Sections were then incubated in 0.6% H2O2, 20% methanol in TBS for 20 minutes to reduce endogenous peroxidase activity and then with secondary biotinylated antibodies (1:200, Vectastain) against mouse (CgA) or rabbit (CgB, secretoneurin, dynorphin) and the avidin-biotinhorseradish peroxidase solution (1:100, Vectastain) for 30 minutes at room temperature. They were then reacted with 0.4 mM 3,3⬘-diaminobenzidine (Sigma, Munich, Germany) and 0.01% H2O2 in TBS for 4 to 6 minutes, mounted on slides, air-dried, dehydrated, and coverslipped. After each incubation step, except the preincubation with normal goat or horse serum, 3 5-minute washes with TBS were included. All buffers and antibody dilutions (except for washing after target retrieval solution, the peroxidase block, and diaminobenzidine) contained 0.4% Triton X-100. In each experiment, sections without primary antibody were included; they did not show immunopositive elements. Sections from control and epilepsy cases were processed simultaneously. Age and length of post mortem time did not affect immunostaining. Data Analysis Staining intensities for CgB- and dynorphin-IR were evaluated using a calibrated image analysis system (MetaMorph3; Visitron Systems, Munich, Germany). After imaging the sections, gray values were evaluated in the inner and outer thirds of the molecular layer and converted to relative optical densities. Data are expressed as ratios of the relative optical densities in these areas (Table). Differences in cell counts and ratios of relative optical densities between autopsies and sclerotic and nonsclerotic specimens were calculated by analysis of variance and further compared between individual groups using a Dunnett test. Cell counts were performed in 20-m-thick Nissl-stained sections using an ocular grid at 200⫻ magnification. Areas of 250 ⫻ 50 and 200 ⫻ 200 m were evaluated in the granule cell and CA2 pyramidal cell layer, respectively. Hilar interneurons were counted in an area of 500 ⫻ 500 m. Data were calculated as neurons/mm3 ⫾ SEM as described.27 Granule cell dispersion was defined as a pattern in which the normally close apposition of granule cells was dissolved. The clear boundary between the granule cell layer and the molecular layer was not observed, and granule cells were scattered into the molecular layer.28 218 Annals of Neurology Vol 50 No 2 August 2001 Table. Ratios of Relative Optical Densities for Chromogranin B (CgB) and Dynorphin in Sclerotic and Nonsclerotic Specimens Ratios of Relative Optical Densities (IML/OML) Tissue CgB Dynorphin n Autopsy Sclerotic Nonsclerotic 1.52 ⫾ 0.05 2.07 ⫾ 0.10a 1.69 ⫾ 0.10 0.67 ⫾ 0.02 2.10 ⫾ 0.09a 0.72 ⫾ 0.12 21 25 4 Ratios of relative optical densities of CgB and dynorphin immunoreactivity in the inner (IML) vs. the outer molecular layer (OML) are shown (mean ⫾ SEM). Note that in autopsies CgB labels more intensely the IML and dynorphin the OML. In sclerotic, but not in nonsclerotic, tissue, labeling of the IML considerably increases for both peptides (compared with OML), reflecting the zone of sprouted mossy fibers. Statistics were performed by analysis of variance with subsequent individual comparisons by a Dunnett test. p ⬍ 0.001 in hippocampal sclerosis vs. controls. a Results Pathology and Dispersion of Granule Cells In accordance with previous observations,27 extensive cell losses were detected in sectors CA1 and CA3/CA4 of sclerotic, but not nonsclerotic, specimens. There was a 77 ⫾ 7.8 % ( p ⬍ 0.01) reduction of neurons in the dentate hilus of sclerotic specimens, whereas the number of neurons in nonsclerotic specimens was reduced by 8 ⫾ 4.7% (not significant) only. Considerable cell losses were also seen in areas relatively resistant to epilepsy-induced cell damage, such as granule cells (55 ⫾ 2.0%) and CA2 pyramidal cells (38 ⫾ 2.2 %, both p ⬍ 0.01) in specimens with hippocampal sclerosis but not significantly in nonsclerotic tissues (10 ⫾ 4.1 and 6 ⫾ 2.6%, respectively). In 9 patients with hippocampal sclerosis, granule cell dispersion was observed.28 The granule cell layer was wider than in other TLE patients and bilaminar in 3 patients. None of the post mortem controls or nonsclerotic hippocampi investigated showed a dispersion of granule cells. Distribution of CgA-IR, CgB-IR, Secretoneurin-IR, and Dynorphin-IR in the Hippocampus of Post Mortem Controls CGA. In controls, CgA-IR was characterized by prominent staining of CA2 pyramidal cells and stratum lucidum of CA3 (Fig 1a). Punctate structures were seen throughout the hilus and stratum pyramidale (Figs 1a, 2a). In the dentate gyrus, some granule cells were CgA-IR (Fig 2a). The molecular layer was nearly devoid of staining (Figs 1a, 2a). Single CgA-IR perikarya and fibers, however, were seen in the molecular layer and in the hilus of the dentate gyrus (not shown). Faint, diffuse staining was present in the stratum lacunosum moleculare; it was slightly more intense in area Fig 1. Chromogranin A (CgA; a, b), chromogranin B (CgB; c, d), secretoneurin (SN; e, f), and dynorphin (Dyn; g, h) immunoreactivity (IR) in the hippocampus of specimens with Ammon’s horn sclerosis and in post mortem controls. In controls (a, c, e, g), IR for all 4 peptides (notably for CgB and dynorphin) is present in the dentate hilus and extends to the strata lucidum and pyramidale of sector CA3, representing the terminal field of mossy fibers. Arrowheads mark borders of the granule cell layer and borders between inner and outer parts of the dentate molecular layer and the granule cell layer, respectively. In autopsies, the inner molecular layer is immunoreactive for CgB (c). Note the prominent dynorphin-IR (g) and secretoneurin-IR (e) bands labeling the outer molecular layer and extending to the stratum lacunosum moleculare of CA3 and CA1. The stratum radiatum shows CgB-IR (c) but not dynorphin-IR (g). In sclerotic specimens (b, d, f, h), these staining patterns become rearranged: staining in the polymorph cell layer adjacent to the granule cell layer is reduced; instead, pronounced labeling of the inner molecular layer of the dentate gyrus is seen (arrows). Note augmented CgB-IR (d) and secretoneurin-IR (f) in hilar interneurons of epileptic specimens. H, hilus; o, outer two-thirds of molecular layer; I, inner third of molecular layer; lm, stratum lacunosum moleculare; sr, stratum radiatum. Scale bar ⫽ 500 m Pirker et al: Chromogranins in TLE 219 Fig 2. Chromogranin A (CgA), chromogranin B (CgB), secretoneurin (SN), and dynorphin (Dyn) immunoreactivity (IR) in dentate gyrus of post mortem controls (a–d) and in nonsclerotic (e–h) and sclerotic specimens (i–l) of temporal lobe epilepsy patients. CgA-IR labels some granule cells and punctate structures in the dentate hilus of controls (a). CgB-IR (b) and dynorphin-IR (d) label prominently granule cells, fiber bundles, individual fibers, and cells in the hilus. Labeling of granule cells is enhanced in epileptic specimens for CgA (e, i), CgB (f, j), and dynorphin (h, l). Whereas the inner molecular layer is lightly labeled only by CgB-IR in autopsies (b), dense CgB-IR (j) and dynorphin-IR (l) are seen there in sclerotic specimens. Generally, labeling is reduced in the dentate hilus, where well-stained cells become visible (j, l). Secretoneurin-IR (c) and dynorphin-IR (d) label the outer molecular layer with a clear-cut border to the inner molecular layer in autopsies, a pattern which is lost in sclerotic (k, l), but not in nonsclerotic (g, h), specimens. Note a plexus of secretoneurin-IR fibers in the supragranular molecular layer of autopsies (arrow in c), secretoneurin-IR cells in the nonsclerotic (g) and sclerotic hilus (k), and fibers extending to the outer molecular layer in sclerotic specimens (k). H, hilus; o, outer two-thirds of molecular layer; i, inner third of molecular layer. Scale bar ⫽ 100 m. 220 Annals of Neurology Vol 50 No 2 August 2001 CA1 and gradually declined in sectors CA2 and CA3 (Fig 1a). Pronounced CgB-IR was observed in the terminal field of mossy fibers and throughout the hippocampus proper (especially in the strata pyramidale and radiatum) and the subicular complex of all autopsies (Figs 1c). Bundles of CgB-IR fibers, presumably mossy fibers, arising from the granule cell layer, formed a dense network in the adjacent hilus (Figs 1c, 2b, 3a,d) and dissolved to more punctate structures deeper in sector CA4 (Fig 3d) and the strata lucidum and pyramidale CA3 (Fig 1c). Weaker, diffuse staining was observed in the inner third of the dentate molecular layer and in the stratum radiatum of area CA3 (Figs 1c, 2b; Table). At high magnification, single fibers were found throughout the molecular layer (not shown). Numerous granule cells were CgB-IR; also, CA2 pyramidal cells were CgB-IR, though less so than for CgA (Figs 1c, 2b, 3a). CGB. SECRETONEURIN. In post mortem controls, secretoneurinIR was present in the outer two-thirds of the molecular layer, the supragranular layer, and the polymorph cell region and extended to the entire CA3 field terminating at the border to CA2 (Fig 1e). IR structures were especially dense in the supragranular layer (Fig 2c, arrow). In the molecular layer, secretoneurin-IR exhibited a laminar distribution (Figs 1e, 2c). The outer two-thirds of the molecular layer were considerably more intensely stained than the inner third (Fig 1e). Single immunopositive fibers were seen throughout the molecular layer, and single moderately stained neurons were found in the molecular and granule cell layers at high magnification (Figs 2c, 3c). In the polymorph cell layer, secretoneurin-IR was found in numerous fibers (presumably mossy fibers) and in single perikarya (Fig 3c). In the hippocampus proper, diffuse secretoneurinpositive staining was observed throughout the stratum lacunosum moleculare and equally in the strata pyramidale and lucidum of area CA3 (Fig 1e). Secretoneurin-IR fibers were present in CA1 and even more numerous in regions CA2 and CA3. Single moderately stained neurons were seen in sectors CA1 to CA3, especially in the strata oriens, pyramidale, and radiatum (not shown). DYNORPHIN. The distribution patterns of dynorphinand ␣-neoendorphin-IR were essentially identical. We therefore refrain from describing ␣-neoendorphin-IR. In autopsies, dark dynorphin-IR was observed in the terminal field of mossy fibers and in granule cells (Fig 1g). Mossy fibers were marked by dark dynorphin-IR structures of punctate appearance, which were particularly dense in the dentate hilus and strata lucidum and pyramidale of CA3 (Figs 1g, 2d). In addition, single somata and many thin fibers were observed in the hilus (Fig 2d). Whereas the inner third of the dentate molecular layer was devoid of dynorphin-IR, its outer two-thirds were labeled by diffuse staining with a sharp IR band delineating the border to the inner layer (Figs 1g, 2d; Table). Single dark fibers and perikarya were found in all parts of the molecular layer. Light diffuse staining was observed also in the subgranular layer and the stratum lacunosum moleculare of sectors CA1 to CA3 (Fig 1g) CgA-IR, CgB-IR, Secretoneurin-IR, and DynorphinIR in Specimens with Hippocampal Sclerosis Marked changes in CgA-IR were observed in sclerotic hippocampi. In most cases (23/25), the number of granule cells labeled appeared to be increased compared with autopsies (Fig 2i). In contrast to post mortem controls, faint CgA-IR was present in the molecular layer of most cases with hippocampal sclerosis (22/25), presumably contained in terminals of sprouted mossy fibers (arrow in Fig 1b). In addition, scattered fibers and somata were observed in all parts of the molecular layer and hilus (Fig 2i). The intensity and width of CgA-IR in the inner molecular layer varied between individual samples. In one-third of patients, staining was light and restricted to the inner third of the dentate molecular layer (Figs 1b, 2i). In another third of sclerotic specimens (9/25), slight CgA-IR extended to the middle and outer thirds of the molecular layer (not shown). In contrast to increased CgA-IR in granule cells and the molecular layer, the extent of IR staining was often reduced in the dentate hilus and stratum lacunosum moleculare of CA1 (Fig 1b). As in controls, pyramidal cells of CA2 were well labeled (Figs 1b, 3i). Their number was, however, significantly reduced, by about 38%, compared to controls and nonsclerotic specimens. Single somata and fibers were seen throughout the hippocampus proper. In the dentate hilus, the stratum lucidum, and less frequently other parts of the hippocampus proper, numerous plaque-like structures were labeled for all 3 chromogranins and dynorphin (Fig 3e, j). They may result from degenerating (mossy) fibers. Similar IR clusters were detected in sclerotic specimens by immunocytochemistry for glutamate receptors 2 and 3.29 Interestingly, CgA-IR and secretoneurin-IR have been observed in senile plaques of Alzheimer’s patients.30 CGA. In patients with hippocampal sclerosis, increased CgB-IR was observed in granule cells and well-stained neurons were detected in parts of the hilus devoid of CgB-IR mossy fibers (Figs 1d, 2j, 3b,g). In the hilus and the hippocampus proper, the same neuronal struc- CGB. Pirker et al: Chromogranins in TLE 221 Fig 3. High-power photomicrographs of chromogranin B (CgB; a, b, d, e, g, h), secretoneurin (c, f), and chromogranin A (CgA; i, j) immunoreactive (IR) structures in the hippocampus of autopsies and specimens with hippocampal sclerosis. Modest CgB-IR is seen in granule cells of autopsies (a) with conspicuous fiber bundles arising from the layer (a) forming a dense network adjacent to the granule cell layer (upper part of d) and dissolving into more punctate structures and IR clusters in CA4 (lower part of d). In sclerotic specimens, the number of fiber bundles is drastically reduced and numerous CgB-IR plaque-like structures (e) and neurons (g) are seen. Plaque-like structures are also detected for CgA (j) and may reflect degenerating (mossy) fibers. Secretoneurin-IR neurons are seen in the hilus of autopsies (c). Their labeling is augmented in sclerotic specimens, and they may give rise to sprouted fibers (f). CgA-IR labels well CA2 pyramidal cells in controls (not shown) and sclerotic specimens (i). g, granule cell layer. Scale bars ⫽ 150 m (a, b, i, j) and 50 m (c–h). tures were present as in controls. However, their abundance (but not their staining intensity) was considerably reduced (Fig 1d). Bundles of impressively 222 Annals of Neurology Vol 50 No 2 August 2001 stained fibers, presumably mossy fibers, were less frequently observed than in controls (Figs 2j, 3b,e). In contrast, single fibers arising from the granule cell layer were clearly detected in the subgranular layer (Fig 3b). The most conspicuous change was the strong increase in CgB-IR in the inner third of the dentate molecular layer, likely labeling sprouted mossy fiber terminals [Figs 1d (arrow), 2j], also reflected by a significant increase in the ratio of relative optical densities in the inner vs. the outer molecular layers (Table). The CgB-IR band covered 16% to 56% (average 34 ⫾ 4%) of the molecular layer. Thus, in 18 of the 25 cases investigated, the moderate to dark IR band occupied the inner third of the molecular layer (Fig 2j). In 6 cases, only a narrow dark IR band was observed in the supragranular layer. On average, specimens with granule cell dispersion28 showed significantly ( p ⬍ 0.05) broader labeling of the molecular layer (42 ⫾ 4%) than those without dispersion. In 1 specimen with granule cell dispersion, the IR band extended to the middle and in another specimen to the entire molecular layer. There was an inverse correlation between the extent of the molecular layer labeled by sprouted CgBpositive fibers and the number of neuropeptide Y mRNA-positive neurons in the dentate hilus (not shown), also indicating a correlation of the extent of neurodegeneration and mossy fiber sprouting. In most specimens, densely packed fibers were seen in the outer molecular layer, which may arise from neurons in the dentate hilus (Fig 3h). No clinical parameters, such as febrile seizures, duration of epilepsy, or other possible risk factors, were found to correlate with either the extent of neurodegeneration or sprouting (not shown). SECRETONEURIN. Secretoneurin-IR was considerably enhanced in the molecular layer of the dentate gyrus, especially in its inner third, the area of sprouted mossy fiber collaterals (Figs 1f, 2k). In contrast, in the stratum lacunosum moleculare of CA1, secretoneurin-IR was less prominent than in controls (Fig 1f). Superimposed over diffuse immunostaining, numerous dark fibers were seen in the molecular layer (Fig 2k), presumably originating from intensely stained neurons in the dentate hilus (Fig 3f). The number of fibers was increased compared to controls. In the polymorph cell region, secretoneurin-IR was reduced in structures presumably formed by mossy fibers (Fig 2h). It was, however, frequently present in distinct fibers and somata throughout the hilus and hippocampus proper of surgical specimens (Figs 1f, 2k 3f). In the stratum lacunosum moleculare, other than in controls, staining was seen only in areas CA2 and CA3, not in CA1 (Fig 1f). DYNORPHIN. In all sclerotic specimens, dynorphin-IR was detected in the molecular layer of the dentate gyrus with a distribution similar to that of CgB-IR. In most cases (14/25), a dark dynorphin-IR band was present in the inner third of the molecular layer, whereas the outer portions were only lightly stained (Figs 1h, 2l), as also indicated by a markedly increased ratio of relative optical densities in the inner vs. the outer molecular layers (Table). No distinct border was detected between the inner and the outer molecular layers (Fig 1h). Similar to controls, many granule cells were intensely stained (Fig 2l). In the polymorphic cell region, dynorphin-IR somata were considerably less numerous than in autopsies (Fig 2l), being detected in 5 specimens only. The extent of punctate IR was reduced in the hilus and the CA3 sector (Figs 1h, 2l). Dynorphin-IR was also reduced in the stratum lacunosum moleculare of CA1 in most cases (Fig 1h). Essentially the same distribution pattern was observed for ␣-neoendorphin-IR (not shown). Chromogranin-IR and Dynorphin-IR in Specimens without Hippocampal Sclerosis In TLE patients without hippocampal sclerosis, IR patterns in most instances were more like those in autopsies than in sclerotic specimens. In particular, no decrease in presumed mossy fiber staining was seen in the dentate hilus (Fig 2e-h), only marginal additional staining was detected in the inner molecular layer for dynorphin-IR, and the outer molecular layer was discretely labeled for secretoneurin and dynorphin (Fig 2g,h). However, labeling of hilar interneurons was enhanced as in sclerotic specimens, though without significant cell loss. Numbers of CgA-IR granule cells and hilar neurons were increased compared to controls (Fig 2e). CGA. In 3 patients, very faint and diffuse CgB-IR was found throughout the molecular layer (Fig 2f). In contrast to controls, dark punctate staining was observed in the supragranular layer at the crest and tips of the dentate gyrus. Staining in the hilus was slightly lighter compared with autopsies. Pyramidal cells of area CA2 were stained in 3 of 4 cases (not shown). CGB. Secretoneurin-IR was similarly distributed, although generally less intensive, in nonsclerotic specimens compared with autopsies (Fig 2g). Labeling was, however, enhanced in hilar neurons (Figs 2g, 3f). SECRETONEURIN. DYNORPHIN. Staining patterns were similar to controls, though often more intense (Fig 2h). Specifically, no prominent dynorphin-IR band was seen in the inner molecular layer. In contrast to autopsies, punctate staining was seen at the crest of the dentate gyrus in the supragranular layer of 3 cases. In the fourth case, light diffuse staining was observed throughout the molecular layer. IR in the hilus was more intense than in controls. Pirker et al: Chromogranins in TLE 223 Discussion Our data provide evidence for a wide and heterogeneous distribution of chromogranins within the human hippocampal formation and for profound changes in TLE. Comparisons between post mortem tissues and surgical specimens have to be drawn with caution for two reasons: mean age of patients at surgery was significantly lower than that of deceased subjects and the time required from isolation of the hippocampal tissue during surgery until its fixation was considerably shorter (three to five hours) than the mean post mortem interval prior to autopsy (17.2 hours). However, no correlation between post mortem times or age at death with the intensity of immunolabeling could be detected. In addition, tissue samples from patients without hippocampal sclerosis to some extent served as controls since they were obtained using the same protocol. Our surgical samples obtained from the hippocampal body (middle segment) were not large enough to investigate possible differences in the rostral–caudal axis. Autopsy specimens were investigated at the same part of the hippocampus. Chromogranin-IR and Dynorphin-IR as Neuroanatomical Markers in the Post Mortem Hippocampus Although significant overlap in the IR patterns was seen for all chromogranins and dynorphin, characteristic differences were also noted. There is an interesting segregation of IR in the molecular layer: whereas CgB-IR faintly labels the entire inner third of the molecular layer following the pattern of associationalcommissural fiber innervation,31 secretoneurin-IR forms a narrow band of intensely labeled individual fibers in the innermost supragranular layer, possibly arising from hilar neurons. The outer two-thirds of the molecular layer (with its distinct border to the inner third) and the stratum lacunosum moleculare are labeled for secretoneurin and dynorphin, outlining terminal areas of entorhinal cortical afferents. High concentrations of dynorphin mRNA in the entorhinal cortex32 indicate that the peptide may be contained in the respective nerve terminals. Typical staining patterns for chromogranins were also seen in the hippocampus proper. Pyramidal cells, especially those in CA2, were preferentially labeled for CgA. The pattern of CgB-IR in the hippocampus proper and the subiculum follows the areas innervated by Schaffer collaterals and axon terminals of CA1 pyramidal cells. This is consistent with the expression of CgB mRNA throughout the pyramidal cell layer of the rat.22 224 Annals of Neurology Vol 50 No 2 August 2001 Reorganization of the Hippocampal Formation in TLE The most conspicuous characteristic of specimens with hippocampal sclerosis is the dramatic reorganization of mossy fibers, as indicated by marked reductions in IR of all three chromogranins (most prominently visualized for CgB) and dynorphin in the dentate hilus and strong increases in IR of chromogranins and dynorphin in the inner molecular layer. In the 4 nonsclerotic TLE specimens investigated, only discrete labeling of the inner molecular layer was detected, excluding extensive sprouting into this area. This indicates that sprouting is associated with neurodegeneration, supported by a high correlation between the area of the inner molecular layer occupied by sprouted mossy fibers and the extent of neurodegeneration in the sclerotic specimens.4,33 Although mossy fiber sprouting may not be a prerequisite for epilepsy (as indicated by the lack of sprouting in nonsclerotic TLE specimens), it may augment recurrent seizures.34 However, we cannot exclude that patients without hippocampal sclerosis at surgery may not have developed hippocampal sclerosis later. It is, however, interesting to note that the average time spent in epilepsy was similar for our three adult patients without Ammon’s horn sclerosis as for those with sclerosis (18.7 vs. 22.4 years). Mossy fiber sprouting has been reported in aged controls.35 We did not observe, however, significant dynorphin-IR in autopsies, including specimens of two older persons (92 and 80 years) otherwise not included in this study. Another interesting observation is the apparent loss of the sharp secretoneurin-IR and dynorphin-IR border between the inner and outer two-thirds of the molecular layer in hippocampal sclerosis. This may be caused by invading mossy fibers and possible reduction of perforant path terminals. The reduced staining in the stratum lacunosum moleculare CA1 to CA3 may be related to loss of entorhinal cortex neurons36 projecting to this area and supports the assumption that the peptides may be contained in these fibers. Dense secretoneurin-IR and CgB-IR fibers, likely arising from hilar interneurons, invade the outer molecular layer of sclerotic TLE specimens, as also shown for neuropeptide Y–containing fibers.13,37 Increased chromogranin-IR in hilar interneurons, in granule cells (CgA and CgB), and in pyramidal cells (CgA) of sclerotic and nonsclerotic specimens may reflect seizureinduced increases in the synthesis of the respective polypeptides. This is consistent with increased chromogranin mRNA concentrations observed in hippocampal neurons of epileptic rats.21,22 Interneurons seem to overexpress chromogranins also in the nonsclerotic hippocampus, reflecting increased activity of presumable GABAergic neurons and possibly contributing to endogenous anticonvulsive mechanisms. Progressive loss of these vulnerable neurons could later propagate neurodegeneration. During epileptic seizures, sprouted mossy fibers and axons of hilar interneurons may release chromogranins or peptides derived from them. Although little is known about the physiological functions of chromogranins, actions of other neuropeptides are well established in animal models of TLE. Thus, neuropeptide Y becomes overexpressed after epileptic seizures and may serve as an endogenous anticonvulsant by suppressing glutamate release from mossy fibers and Schaffer collaterals.38,39 Similarly, dynorphin exerts an inhibitory action on glutamate release through -receptors located presynaptically on mossy fibers and perforant path terminals.40 Inhibitory action has been shown also for dynorphin released from sprouted mossy fibers upon -receptors on perforant path terminals.41 In conclusion, changes in IR of chromogranins mark an impressive rearrangement of neuronal circuitries in patients with TLE-induced Ammon’s horn sclerosis. These include degenerative loss of chromogranin-IR mossy fibers in the dentate hilus and sprouting of mossy fibers. Distortion of the lamination of the molecular layer, as outlined by secretoneurin-IR and dynorphin-IR and losses of IR in the stratum lacunosum moleculare, indicate impaired innervations from the entorhinal cortex. Surviving hilar interneurons exhibit enhanced expression of secretoneurin and presumably give rise to fibers sprouting to the outer parts of the dentate molecular layer. Chromogranins and peptides derived from them may be released during epileptic seizures and may exert a functional role in the epileptic hippocampus. This project was funded by the Austrian Federal Ministry for Science and Transport (GZ 70.039/2-Pr/4/98) and the Austrian Science Foundation. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. We thank Anna Wieselthaler for excellent technical assistance and C. Trawöger and Dr C. Schwarzer for valuable discussion. 18. References 19. 1. Sommer W. Erkrankung des Ammonshornes als ätiologisches Moment der Epilepsie. Arch Psychiatr Nervenkrankh 1880;10: 631– 675. 2. Pfleger L. Beobachtungen über Schrumpfung und Sklerose des Ammonshorns bei Epilepsie. Allg Z Psychiatr 1880;36:359 – 365. 3. Babb TL, Kupfer WR, Pretorius JK, et al. Synaptic reorganization by mossy fibers in human epileptic fascia dentata. Neuroscience 1991;42:351–363. 4. Houser CR, Miyashiro JE, Swartz BE, et al. Altered patterns of dynorphin immunoreactivity suggest mossy fiber reorganization in human hippocampal epilepsy. J Neurosci 1990;10:267–282. 5. Sutula T, Cascino G, Cavazos J, et al. Mossy fiber synaptic reorganization in the epileptic human temporal lobe. Ann Neurol 1989;26:321–330. 6. Mikkonen M, Soininen H, Kalvianen R, et al. Remodeling of neuronal circuitries in human temporal lobe epilepsy: increased 20. 21. 22. 23. expression of highly polysialylated neural cell adhesion molecule in the hippocampus and the entorhinal cortex. Ann Neurol 1998;44:923–934. Wuarin JP, Dudek FE. Electrographic seizures and new recurrent excitatory circuits in the dentate gyrus of hippocampal slices from kainate-treated epileptic rats. J Neurosci 1996;16: 4438 – 4448. Okazaki MM, Evenson DA, Nadler JV. Hippocampal mossy fiber sprouting and synapse formation after status epilepticus in rats: visualization after retrograde transport of biocytin. J Comp Neurol 1995;352:515–534. Zhang N, Houser CR. Ultrastructural localization of dynorphin in the dentate gyrus in human temporal lobe epilepsy: a study of reorganized mossy fiber synapses. J Comp Neurol 1999;405: 472– 490. Frotscher M, Zimmer J. Commissural fibers terminate on nonpyramidal neurons in the guinea pig hippocampus—a combined Golgi/EM degeneration study. Brain Res 1983;265:289 – 293. Ottersen OP, Storm-Mathisen J. Excitatory and inhibitory amino acids in the hippocampus. In: Chan-Palay V, Köhler C, eds. The hippocampus: new vistas. New York: Alan R. Liss; 1989:97–117. McGinty JF, Henriksen SJ, Goldstein A, et al. Dynorphin is contained within hippocampal mossy fibers: immunochemical alterations after kainic acid administration and colchicineinduced neurotoxicity. Proc Natl Acad Sci USA 1983;80:589 – 593. de Lanerolle NC, Kim JH, Robbins RJ, et al. Hippocampal interneuron loss and plasticity in human temporal lobe epilepsy. Brain Res 1989;495:387–395. Winkler H, Fischer-Colbrie R. The chromogranins A and B: the first 25 years and future perspectives. Neuroscience 1992; 49:497–528. Marksteiner J, Bauer R, Kaufmann WA, et al. PE-11, a peptide derived from chromogranin B, in the human brain. Neuroscience 1999;91:1155–1170. Munoz DG. The distribution of chromogranin A-like immunoreactivity in the human hippocampus coincides with the pattern of resistance to epilepsy-induced neuronal damage. Ann Neurol 1990;27:266 –275. Somogyi P, Hodgson AJ, DePotter RW, et al. Chromogranin immunoreactivity in the central nervous system. Immunochemical characterisation, distribution and relationship to catecholamine and enkephalin pathways. Brain Res Rev 1984;320: 193–230. Laslop A, Weiss C, Savaria D, et al. Proteolytic processing of chromogranin B and secretogranin II by prohormone convertases. J Neurochem 1998;70:374 –383. Kirchmair R, Hogue-Angeletti R, Gutierrez J, et al. Secretoneurin—a neuropeptide generated in brain, adrenal medulla and other endocrine tissues by proteolytic processing of secretogranin II (chromogranin C). Neuroscience 1993;53:359 –365. Reinisch N, Kirchmair R, Kahler CM, et al. Attraction of human monocytes by the neuropeptide secretoneurin. FEBS Lett 1993;334:41– 44. Mahata SK, Gruber B, Mahata M, et al. Kainic acid seizures in the rat: differential expression of chromogranin A, carboxypeptidase H and peptidylglycine ␣-amidating monooxygenase in subfields of the hippocampal formation. Acta Neuropathol 1993;86:590 –595. Mahata SK, Marksteiner J, Sperk G, et al. Temporal lobe epilepsy of the rat: differential expression of mRNAs of chromogranin B, secretogranin II, synaptin/synaptophysin and p65 in subfield of the hippocampus. Brain Res Mol Brain Res 1992; 16:1–12. Wieser HG, Engel J Jr, Williamson PD, et al. Surgically reme- Pirker et al: Chromogranins in TLE 225 24. 25. 26. 27. 28. 29. 30. 31. 226 diable temporal lobe syndromes. In: Engel J Jr, ed. Surgical treatment of epilepsies. 2nd ed. New York: Raven; 1993:49 – 63. Benedum UM, Lamouroux A, Konecki DS, et al. The primary structure of human secretogranin I (chromogranin B): comparison with chromogranin A reveals homologous terminal domains and a large intervening variable region. EMBO J 1987; 6:1203–1211. Gerdes HH, Phillips E, Huttner WB. The primary structure of rat secretogranin II deduced from a cDNA sequence. Nucleic Acids Res 1988;16:11811. Griffond B, Deray A, Fellmann D, et al. Colocalization of prolactin- and dynorphin-like substances in a neuronal population of the rat lateral hypothalamus. Neurosci Lett 1993;156: 91–95. Mathern GW, Pretorius JK, Mendoza D, et al. Increased hippocampal AMPA and NMDA receptor subunit immunoreactivity in temporal lobe epilepsy patients. J Neuropathol Exp Neurol 1998;57:615– 634. Houser CR. Granule cell dispersion in the dentate gyrus of humans with temporal lobe epilepsy. Brain Res 1990;535:195– 204. de Lanerolle NC, Eid T, von Campe G, et al. Glutamate receptor subunits GluR1 and GluR2/3 distribution shows reorganization in the human epileptogenic hippocampus. Eur J Neurosci 1998;10:1687–1703. Weiler R, Lassmann H, Fischer P, et al. A high ratio of chromogranin A to synaptin/synaptophysin is a common feature of brains in Alzheimer and Pick disease. FEBS Lett 1990;263: 337–339. Deller T, Nitsch R, Frotscher M. Associational and commissural afferents of parvalbumin-immunoreactive neurons in the rat hippocampus: a combined immunocytochemical and PHA-L study. J Comp Neurol 1994;350:612– 622. Annals of Neurology Vol 50 No 2 August 2001 32. Hurd YL, Herman MM, Hyde TM, et al. Prodynorphin mRNA expression is increased in the patch vs matrix compartment of the caudate nucleus in suicide subjects. Mol Psychiatry 1997;2:495–500. 33. Mathern GW, Pretorius JK, Babb TL. Quantified patterns of mossy fiber sprouting and neuron densities in hippocampal and lesional seizures. J Neurosurg 1995;82:211–219. 34. Cronin J, Obenaus A, Houser CR, et al. Electrophysiology of dentate granule cells after kainate-induced synaptic reorganization of the mossy fibers. Brain Res 1992;573:305–310. 35. Cassell MD, Brown MW. The distribution of Timm’s stain in the nonsulphide-perfused human hippocampal formation. J Comp Neurol 1984;222:461– 471. 36. Du F, Whetsell WO Jr, Abou-Khalil B, et al. Preferential neuronal loss in layer III of the entorhinal cortex in patients with temporal lobe epilepsy. Epilepsy Res 1993;16:223–233. 37. Mathern GW, Babb TL, Pretorius JK, et al. Reactive synaptogenesis and neuron densities for neuropeptide Y, somatostatin, and glutamate decarboxylase immunoreactivity in the epileptogenic human fascia dentata. J Neurosci 1995;15:3990 – 4004. 38. Gruber B, Greber S, Rupp E, et al. Differential NPY mRNA expression in granule cells and interneurons of the rat dentate gyrus after kainic acid injection. Hippocampus 1994;4:474 – 482. 39. Vezzani A, Sperk G, Colmers WF. Neuropeptide Y: emerging evidence for a functional role in seizure modulation. Trends Neurosci 1999;22:25–30. 40. Weisskopf MG, Zalutsky RA, Nicoll RA. The opioid peptide dynorphin mediates heterosynaptic depression of hippocampal mossy fibre synapses and modulates long-term potentiation. Nature 1993;362:423– 427. 41. Terman GW, Drake CT, Simmons ML, et al. Opioid modulation of recurrent excitation in the hippocampal dentate gyrus. J Neurosci 2000;20:4379 – 4388.