* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Distinct Neuropathologic Phenotypes After Disrupting the
Holonomic brain theory wikipedia , lookup
Biochemistry of Alzheimer's disease wikipedia , lookup
Development of the nervous system wikipedia , lookup
Biological neuron model wikipedia , lookup
Clinical neurochemistry wikipedia , lookup
Neuroplasticity wikipedia , lookup
Neuroregeneration wikipedia , lookup
Neuroanatomy wikipedia , lookup
Premovement neuronal activity wikipedia , lookup
Neuroeconomics wikipedia , lookup
Neuropsychopharmacology wikipedia , lookup
Eyeblink conditioning wikipedia , lookup
Metastability in the brain wikipedia , lookup
Nervous system network models wikipedia , lookup
Feature detection (nervous system) wikipedia , lookup
Aging brain wikipedia , lookup
Neurogenomics wikipedia , lookup
Spike-and-wave wikipedia , lookup
Environmental enrichment wikipedia , lookup
Epigenetics in learning and memory wikipedia , lookup
Synaptic gating wikipedia , lookup
J Neuropathol Exp Neurol Copyright Ó 2010 by the American Association of Neuropathologists, Inc. Vol. 69, No. 12 December 2010 pp. 1228Y1246 ORIGINAL ARTICLE Distinct Neuropathologic Phenotypes After Disrupting the Chloride Transport Proteins ClC-6 or ClC-7/Ostm1 Sarah N.R. Pressey, MSc, Kieran J. O’Donnell, MSc, Tobias Stauber, PhD, Jens C. Fuhrmann, PhD, Jaana Tyynelä, PhD, Thomas J. Jentsch, PhD, and Jonathan D. Cooper, PhD INTRODUCTION Abstract The proteins ClC-6 and ClC-7 are expressed in the endosomallysosomal system. Because Clcn6-deficient mice display some features of neuronal ceroid lipofuscinosis (NCL), CLCN6 may be a candidate gene for novel forms of NCL. Using landmarks of disease progression from NCL mouse models as a guide, we examined neuropathologic alterations in the central nervous system of Clcn6j/j, Clcn7j/j, and gl mice. gl mice bear a mutation in Ostm1, the A-subunit critical for Clcn7 function. Severely affected Clcn7j/j and gl mice have remarkably similar neuropathologic phenotypes, with pronounced reactive changes and neuron loss in the thalamocortical system, similar to findings in early-onset forms of NCL. In contrast, Clcn6j/j mice display slowly progressive, milder neuropathologic features with very little thalamic involvement or microglial activation. These findings detail for the first time the markedly different neuropathologic consequences of mutations in these two CLC genes. Clcn7j/j and gl mice bear a close resemblance to the progressive neuropathologic phenotypes of early onset forms of NCL, whereas the distinct phenotype of Clcn6-deficient mice suggests that this gene could be a candidate for a later-onset form of mild neurologic dysfunction with some NCL-like features. Key Words: Chloride channel, Chloride/proton exchanger, Grey lethal, Lysosomal storage disorder, Neuronal ceroid lipofuscinosis/ Batten disease. From the Pediatric Storage Disorders Laboratory, Department of Neuroscience (SNRP, KJO, JDC), Centre for the Cellular Basis of Behavior, Institute of Psychiatry, King’s College London, UK; Leibniz Institute for Molecular Pharmacology [FMP] and Max Delbrück Centre [MDC] for Molecular Medicine (TS, JCF, TJJ), Berlin, Germany; and Institute of Biomedicine/ Biochemistry (JT), University of Helsinki, Finland. Send correspondence and reprint requests to: Jonathan D. Cooper, PhD, Pediatric Storage Disorders Laboratory, Department of Neuroscience and Centre for the Cellular Basis of Behavior, MRC Centre for Neurodegeneration Research, James Black Centre, Institute of Psychiatry, King’s College London, 125 Coldharbour Ln, London, SE5 9NU, UK; E-mail: [email protected] Jens C. Fuhrmann is now at Metanomics GmbH, Berlin, Germany. These studies were supported by National Institutes of Health grant NS41930 (J.D.C.), European Commission 6th Framework Research Grant LSHMCT-2003-503051 (J.D.C.), and a UK Medical Research Council Studentship (S.N.R.P. and J.D.C.), the Deutsche Forschungsgemeinschaft to TJJ (Je164/7 and SFB 740), and Academy of Finland 1122404 (J.T.). The following nonprofit agencies also contributed financially to this work: The Batten Disease Support and Research Association (J.D.C.), The Batten Disease Family Association (J.D.C.), and The Natalie Fund (J.D.C.). 1228 Members of the CLC family of chloride channels and transporters have diverse functions in regulating electrical excitability, performing transepithelial transport, maintaining electroneutrality during acidification of organelles, and control of ionic homeostasis (1). Much of this information has come from the disruption of CLC genes in mouse models (1, 2). Of the 9 mammalian CLCs, 4 are associated with inherited disorders, with mutations in the genes encoding ClC-Kb, ClC-1, ClC-5, and ClC-7 that cause Bartter syndrome, myotonia congenita, Dent disease, and osteopetrosis, respectively (3Y6). ClC-6 and ClC-7 form a distinct branch of the CLC gene family, sharing 45% sequence homology (7). ClC-7 requires a A-subunit, Ostm1 (8); mutations in Ostm1 are responsible for the phenotype observed in grey-lethal (gl ) spontaneous mutant mice (9). The gl mice phenotype was ascribed to a severe loss of ClC-7, which is unstable without its A-subunit (8). The phenotypes of Clcn6j/j, Clcn7j/j, and gl mice have unexpectedly suggested links to a group of rare fatal pediatric storage disorders, the neuronal ceroid lipofuscinoses (NCLs or Batten disease) (10, 11). Individuals with NCL present with seizures and visual, intellectual and motor deterioration (12). Because there are no effective therapies for these autosomal recessive disorders, their relentless disease course invariably ends in premature death (12, 13). Disease-causing mutations have been identified in at least 8 genes (13, 14), but there are still variant and adultonset NCL cases with no characterized molecular basis, suggesting that new disease-causing genes remain to be identified. The NCLs are characterized by autofluorescent intralysosomal storage material accumulation consisting of a mixture of lipids and proteins, the most abundant being subunit c of mitochondrial ATPase and saposins (13, 15). This storage material accumulates widely in central and peripheral tissues, but there is no evidence that it has a direct role in pathogenesis. The ultrastructural appearance of the storage material is specific to different forms of NCL, consisting of granular osmiophilic deposits, fingerprint, and curvilinear profiles (13, 15). The storage material that accumulates in the lysosomes in the central nervous system (CNS) of Clcn6j/j, Clcn7j/j, and gl mice resembles that seen in the NCLs; it is electron-dense, granular osmiophilic deposit-like, autofluorescent, and stains positively for subunit c of mitochondrial ATPase (8, 10, 11). Beyond these common features, the neuronal phenotypes of Clcn6j/j and Clcn7j/j mice differ in many respects. J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 Clcn6j/j mice have a normal life span, with mild behavioral abnormalities and decreased pain sensitivity (11). Upregulation of inflammatory genes in the hippocampi of Clcn6deficient mice suggests that there is an immune response in these mice (11), similar to that in the NCLs, but this has not been characterized in detail. gl and Clcn7-deficient mice display disease phenotypes that are very similar to each other and death occurs by 7 weeks (4, 8). Both of these mice share neuropathologic phenotypes with the NCLs including severe retinal and CNS degeneration (4, 8). Clcn7j/j mice show optic nerve degeneration and associated localized glial activation (4, 10, 16), but this has not yet been investigated in gl mice. Clcn7j/j and gl mice also display severe osteopetrosis and growth retardation (4, 8). Both CLCN7 and OSTM1 deficiency lead to osteopetrosis in a subset of patients with juvenile osteopetrosis (4, 9, 17, 18). These patients also have neurologic signs and brain atrophy as a result of deficiency in CLCN7 or OSTM1 (19Y21). Detailed characterization of NCL mouse models has revealed a series of common themes with respect to the brain regions and cell types that are affected, although the precise timing and nature of these events differs among NCL forms (22Y25). Regional cortical atrophy, localized early gliosis, and loss of both interneurons and thalamic relay neurons are strong indicators of disease progression. In light of the similarities in CNS pathology between the NCLs and Clcn6-, Clcn7-, and Ostm1-deficient mice, we examined the progression of neuropathologic phenotypes in these mice to determine their status as potential models for new forms of NCL. These models provide insight into the mechanisms of neuronal storage, the functions of CLC genes in the CNS and the distinct neuropathologic consequences of deficiencies of these 2 CLC genes. MATERIALS AND METHODS Clcn6- and Clcn7-Deficient Mice Clcn6- and Clcn7-deficient mice (Clcn6j/j and Clcn7 ) were both generated by targeted disruption strategies, as previously described (4, 11), Clcn6j/j mice were backcrossed with C57BL/6 control mice and Clcn7j/j mice were on a mixed C57BL/6-129SV background. Grey-lethal mice (gl), which have a spontaneous mutation in Ostm1 (8, 9), were obtained from the Jackson Laboratory (Bar Harbor, ME) and crossed onto the C57BL/6 background. Clcn6j/j, Clcn7j/j, and gl mutant mice and age-matched littermate controls were maintained at the Max Delbrück Centre, Berlin. All animal procedures performed in accordance with national and institutional animal welfare directives. j/j Histologic Analysis The brains of deficient mice and age-matched controls were examined at 3 time points during their life span, when they were presymptomatic, moderately and more severely affected. Clcn6j/j mice have a normal life span and were examined at 55, 92, and 107 weeks. Because Clcn7j/j mice have an earlier onset disease phenotype and die between 6 and 7 weeks, their brains were examined at 1.5, 3, and 4 weeks. gl mice were also examined at 4 weeks. Ó 2010 American Association of Neuropathologists, Inc. Neuropathology of ClC-6 & ClC-7/Ostm1-Deficient Mice Brains were removed and immersion fixed in 4% paraformaldehyde (Sigma-Aldrich, Poole, UK) in 0.1 mol/L phosphate buffer, pH 7.4 for 3 days and then cryoprotected in 30% sucrose, 0.5% sodium azide in 50 mmol/L Tris-buffered saline (TBS), pH 7.6. Frozen coronal sections 40 Km thick through the rostrocaudal extent of the cortical mantle were collected at 1 per well in 96-well plates containing a cryoprotective solution (30% ethylene glycol [Sigma-Aldrich], 15% sucrose, 0.05% sodium azide in TBS; Sigma-Aldrich) (26). All subsequent histologic analyses were performed blind to genotype. Histochemical Staining Every sixth section through each brain was slide-mounted and incubated in 0.05% cresyl fast violet (Merck, Darmstadt, Germany), 0.05% acetic acid in water for 30 minutes at 60-C, rinsed in deionized water, and then differentiated through an ascending series of alcohols before clearing in xylene and coverslipped with DPX (VWR, Poole, UK) (23, 26). Periodic acid Schiff (PAS) staining was used to determine the presence of accumulated carbohydrate storage material (27). Sections were slide-mounted and left to dry overnight and then incubated in periodic acid (Sigma-Aldrich) for 2 minutes and then rinsed with several changes of dH2O. Sections were next incubated in the Schiff reagent for 2 minutes, followed by washing for 2 minutes in running tap water and dehydrated through an alcohol series, cleared in xylene, and coverslipped with DPX. Immunohistochemistry A standard immunoperoxidase protocol was used to detect glial fibrillary acid protein (GFAP), microglia (F4/80), lysosomes (Cathepsin D [CtsD]), and GABAergic interneurons (somatostatin [SOM], parvalbumin [PV], calretinin [CR], and calbindin [CB]) (10, 11, 26, 28). Endogenous peroxidase activity was quenched in 1% H2O2 (VWR) in TBS for 15 minutes. Sections were then rinsed in TBS and blocked in 15% normal serum (Vector, Northampton, UK) in TBS with 0.3% Triton-X (TBS-T; SigmaAldrich) before incubating in the appropriate primary antibody; polyclonal rabbit anti-GFAP (1:4000; DakoCytomation, Ely, UK), rat anti-mouse F4/80 (1:100; Serotec, Oxford, UK), rabbit anti-CtsD (1:1000; Millipore-Upstate, Watford, UK), rabbit antiSOM (1:1000; Bachem, Bubendorf, Switzerland), rabbit antiPV (1:5000; Swant, Bellinzona, Switzerland), rabbit anti-CR (1:10,000; Swant), or rabbit anti-CB (1:20,000; Swant) in 10% normal serum in TBS-T overnight at 4-C. Sections were rinsed in TBS and incubated with the appropriate biotinylated secondary antibodies; swine anti-rabbit (1:1000; DakoCytomation), rabbit anti-rat (1:200; Vector), and goat anti-rabbit (1:1000; Vector) for 2 hours at room temperature. Subsequently, sections were rinsed in TBS, followed by incubation in avidin-biotinperoxidase complex (1:1000, Vectastain Elite ABC kit; Vector) in TBS for 2 hours and then rinsed in TBS. To visualize immunoreactivity (IR), sections were incubated in 0.05% 3,3¶diaminobenzidine tetrahydrochloride (Sigma-Aldrich) containing 0.001% H2O2 in TBS for up to 25 minutes, depending on antigen, and then rinsed in ice-cold TBS. Finally, sections were mounted on gelatin-chrome-coated Superfrost microscope slides (VWR), air-dried overnight, cleared in xylene, and coverslipped with DPX. 1229 Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. Pressey et al J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 A standard immunofluorescence protocol was used to reveal the abundance and distribution of lysosomes (CtsD) within neurons (NeuN) or microglia (CD68). Sections were blocked in 15% normal serum in TBS-T, then probed with the following primary antibodies: rabbit anti-CtsD (1:1000; Millipore-Upstate), mouse anti-NeuN (1:1000; Millipore), and rat anti-CD68 (1:2000; Serotec) in 10% normal serum in TBS-T overnight at 4-C. After rinsing in TBS and incubating with the appropriate fluorescent secondary Alexa Fluor antibodies (Molecular Probes, Invitrogen, Carlsbad, CA) for 2 hours at room temperature, sections were rinsed in TBS and mounted on gelatine-chrome-coated microscope slides and coverslipped with fluoromount-G (Southern Biotech, Birmingham, AL). Regional Volume, Cortical Thickness, and Cell Number Estimation Stereological estimates were obtained using StereoInvestigator software (Microbrightfield, Inc, Williston, VT), with regions of interest (ROIs) defined by referring to the neuroanatomic landmarks described by Paxinos and Franklin (29). Cavalieri estimates of regional volume (23, 28) were obtained using the Nissl-stained sections described. A sampling grid was superimposed over the ROI and the number of points covering the relevant area was assessed. The volume in cubic micrometers was estimated from a series of sections through each ROI. Sampling grids and series used were as follows: for brain subdivisions for Clcn7j/j and gl mice, a 150-Km2 grid was used for all regions; for cortex, 1:12 series; and all other regions, 1:6 series. For Clcn6j/j mice: cortex, 1:12, 250 Km2; striatum, 1:12, 150 Km2; thalamus, 1:6, 150 Km2; and hippocampus, 1:6, 200 Km2. Volumes of representative cortical subfields that serve different functions were also obtained from the same sections. These included the primary somatosensory (S1BF), primary motor (M1), lateral entorhinal (LEnt), and primary visual (V1) cortices, using a 150-Km2 grid on a 1:6 series for all mice. For the estimation of the volume of white matter tracts, a 100-Km2 grid was used for all mice on a 1:12 series for the corpus callosum and a 1:6 series for the internal capsule and cerebral peduncle. Analyses were carried out on an Olympus BX50 microscope (Olympus, Southend-on-Sea, UK) linked to a DAGE-MTI CCD-100 camera (Dage-MTI, Michigan City, IN). The optical fractionator probe was used to estimate cell numbers (30) in Nissl-stained sections of thalamic relay nuclei and lamina IV-VI of the S1BF and in sections through S1BF stained for markers of interneuron populations known to be vulnerable in this brain region (23, 25, 26). Nissl-stained cells were only counted if they had a neuronal morphology and a clearly identifiable nucleus; astrocytes and microglia with small soma and cells with indistinct morphology were not counted. A line was traced around the boundary of the ROI, a grid was superimposed, and cells were counted within a series of dissector frames placed according to the sampling grid size. Nissl+ neurons and SOM+, PV+, CR+, or CB+ interneurons were counted using the 100 and 40 objectives, respectively. Different grid and dissector sizes were determined according to each brain region using a coefficient of error value of less than 0.1 to indicate sampling efficiency. Because of the different brain sizes and ages investigated, these sampling parameters 1230 were different for each mouse model, with the following sampling schemes used: For Nissl-stained sections in Clcn7j/j and gl mice: lateral geniculate nucleus (LGN) grid 125 125 Km, frame 74 42 Km; ventral posterior thalamic nucleus (VPM/VPL), grid 175 175 Km, frame 74 43 Km; medial geniculate nucleus (MGN) grid 200 200 Km, frame 50 35 Km; S1BF lamina IV, grid 150 150 Km, frame 41 26 Km; S1BF lamina V, grid 200 200 Km, frame 60 40 Km; S1BF lamina VI, grid 200 200 Km, frame 60 40 Km. For Nissl-stained sections in Clcn6j/j mice: VPM/VPL grid 225 225 Km, frame 74 42 Km; S1BF lamina IV, grid 150 150 Km, frame 41 26 Km; S1BF lamina V, grid 175 175 Km, frame 60 40 Km; and S1BF lamina VI, grid 200 200 Km, frame 60 40 Km. A 1:6 series was used for these analyses. For all sections stained for PV: grid 250 200 Km, frame 150 90 Km, 1:12 series. For all sections stained for SOM, CR and CB grid 250 180 Km, frame 180 100 Km, 1:12 series. Because of the comparatively low abundance of interneurons in the hippocampus versus the neocortex, stereological methods prove inefficient for estimating hippocampal interneuron numbers without sampling the entire tissue. Therefore, to determine the number of interneurons present in different subfields of the hippocampus, counts of interneurons were made manually in each defined hippocampal subfield, as previously described (26, 31). Data were plotted as the mean number of interneurons per section positive for each marker in each of these hippocampal subfields. Quantitative Analysis of Glial Phenotypes The optical density of GFAP-IR was assessed using semiautomated thresholding image analysis, as previously described (28, 32). Using a 40 objective, 40 nonoverlapping images representing the entire ROI were taken from 4 consecutive sections starting from a defined anatomic landmark. Image Pro Plus image analysis software (Media Cybernetics, Chicago, IL) was used to determine the area of IR for each antigen in each region by applying a threshold that discriminated staining from background in each image. Data were plotted as the mean percentage area of IR per field T SEM for each region. Statistical Analysis Microsoft Excel (Redmond, WA) was used for data collection and SPSS (SPSS, Inc, Chicago, IL) for statistical analysis. To test for significance between genotypes, Student t test or analysis of variance with post hoc Bonferroni analysis were used as appropriate. The mean coefficient of error for all individual optical fractionator and Cavalieri estimates was calculated according to the method of Gundersen and Jensen (33) and was less than 0.1 in all these analyses. RESULTS Clcn7j/j, gl, and Clcn6j/j mice have a lysosomal storage disease that resembles the NCLs (8, 10, 11). Detailed landmarks of disease progression such as selective atrophy, localized glial activation, the loss of interneurons, and thalamocortical neurons are available for many of the forms of murine NCL (22). To assess whether similar events occur in Ó 2010 American Association of Neuropathologists, Inc. Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 Neuropathology of ClC-6 & ClC-7/Ostm1-Deficient Mice FIGURE 1. Distinct cortical and white matter atrophy in Clcn7-and Clcn6<deficient mice. (A, B) Unbiased Cavalieri estimates of regional volume revealed cortex-specific atrophy in severely affected 4-week-old Clcn7j/j mice versus age-matched littermate controls (+/+) that was not evident in 3-week-old mice (A). Hipp = hippocampus. There were no differences in regional volume of severely affected 92-week-old Clcn6j/j mice (B). (C, D) Cavalieri estimates of the primary motor (M1), lateral entorhinal (LEnt), primary visual (V1), and somatosensory barrel field (S1BF) cortex revealed selective atrophy of S1BF in 4-week-old Clcn7j/j mice (C) and of the LEnt in 92-week-old Clcn6j/j mice (D). Both these events occurred late as there was no change in S1BF volume in 3-weekold Clcn7j/j mice (C) or LEnt volume in 55-week-old Clcn6j/j mice (D). (EYG) Cavalieri estimates of white matter tract volumes revealed atrophy of the corpus callosum in Clcn7j/j mice that became evident at 4 weeks (E), and atrophy of the internal capsule that became significant at 4 weeks (F). Clcn6j/j mice showed atrophy of the corpus callosum from 55 weeks (G). *, p G 0.05; **, p G 0.01; ***, p G 0.001; using Student t test. n = 3Y5; error bars = SEM. Ó 2010 American Association of Neuropathologists, Inc. 1231 Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. Pressey et al J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 FIGURE 2. Localized astrocytosis in thalamic relay nuclei and basal ganglia in Clcn7-, but not Clcn6-deficient mice. (A) Immunohistochemical staining for glial fibrillary acidic protein (GFAP) revealed pronounced astrocytosis in the basal ganglia and thalamic nuclei of Clcn7j/j mice at 4 weeks. Thalamic nuclei boundaries are marked with white dashed lines. LGN indicates lateral geniculate nucleus; MGN, medial geniculate nucleus; Rt, reticular thalamic nucleus; VPM/VPL, ventral posterior thalamic nucleus. Basal ganglia: GP indicates globus pallidus; SN, substantia nigra. (B) GFAP immunoreactivity was first evident in Clcn7j/j mice in the VPM/VPL at 1.5 weeks and became progressively more widespread with increased age, extending to the MGN in 3-week-old mice. (C) GFAP+ astrocytes are scattered throughout the thalamus of Clcn6j/j mice with no particular focus. Scale bar = 500 Km. 1232 Ó 2010 American Association of Neuropathologists, Inc. Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 Neuropathology of ClC-6 & ClC-7/Ostm1-Deficient Mice FIGURE 3. Distinct patterns of astrocytosis in the cortex of Clcn7j/j and Clcn6j/j mice. Immunohistochemical staining for glial fibrillary acidic protein (GFAP) revealed progressive activation of astrocytes in the cortex of Clcn7j/j and Clcn6j/j mice. Comparison to adjacent Nissl-stained sections enabled the definition of cortical laminar boundaries. (A) At 1.5 weeks, more prominent GFAP+ astrocytes were present in Clcn7j/j compared with age-matched control mice (+/+); they are progressively more widespread and intensely stained at 3 and 4 weeks. Dense bands of astrocytosis were evident in lamina IV and the boundary between lam V and VI in the somatosensory barrel field cortex (S1BF) from 3 weeks. (BYE) In Clcn6j/j mice, there was a more generalized pattern of astrocytosis with GFAP+ astrocytes scattered throughout the cortex in sensory (B, S1BF) and ventral regions (D, piriform cortex) occurring most frequently in deep laminae. Quantitative thresholding image analysis in the S1BF (C) and piriform cortex (E) of Clcn6j/j mice demonstrated consistent upregulation of GFAP immunoreactivity versus age-matched controls (+/+). *, p G 0.05; ***, p G 0.001; analysis of variance, n = 3Y5; error bars show SEM). Scale bar = 250 Km. Clcn7j/j and Clcn6j/j mice, the CNS of mutant and control mice were examined at presymptomatic, moderately and severely affected stages and were compared with severely affected gl mice. Regional Atrophy in Clcn7j/j and Clcn6j/j Mice The Cavalieri method was used to obtain estimates of regional volumes (23). This analysis revealed a late-onset atrophy of the cortex in Clcn7j/j mice (Fig. 1A), but in no other brain region in severely affected mice at 4 weeks. This cortical atrophy displayed subfield specificity and was only significant in the S1BF at this age; there was no significant difference at earlier ages (Fig. 1C). A similar late-onset atrophy of the LEnt was evident in Clcn6j/j mice (Fig. 1D), which displayed no other regional atrophy, even at 92 weeks (Fig. 1B). These data indicate regionally specific effects on Ó 2010 American Association of Neuropathologists, Inc. the sensory cortex of Clcn7j/j mice and on the archicortex of Clcn6j/j mice toward the end of their disease progression. There was a significant late-onset atrophy of the corpus callosum, the internal capsule, and the cerebral peduncles in severely affected Clcn7j/j mice at 4 weeks (Figs. 1E, F). In Clcn6j/j mice, only the corpus callosum displayed significant atrophy that was evident from 55 weeks (Fig. 1G). Thus, there was widespread atrophy of white matter in Clcn7j/j mice versus more selective effects on the corpus callosum in Clcn6j/j mice. Clcn7j/j Mice But Not Clcn6j/j Mice Display Pronounced Localized Astrocytosis Astrocytosis can be a sensitive marker of ongoing neuronal dysfunction or injury (22, 34, 35). In control mice, GFAP-IR was confined to fibrous astrocytes within the white 1233 Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. Pressey et al 1234 J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 Ó 2010 American Association of Neuropathologists, Inc. Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 matter. Only occasional faintly stained protoplasmic astrocytes were scattered in the thalamus of control mice at all ages (Figs. 2A, C). The GFAP-IR in thalamic nuclei of Clcn7j/j mice appeared at different stages of disease progression according to the sensory modality these nuclei relay (Figs. 2A, B). It started in the VPM/VPL (somatosensory/pain information) and subsequently extended to the MGN (auditory information) (Fig. 2B), LGN (visual information), and reticular thalamic nucleus (Rt) (provides inhibitory input to the whole thalamus) with increased age. Another distinctive feature of Clcn7j/j mice was the presence of intense GFAP-IR within basal ganglia and midbrain nuclei. This was first evident at 3 weeks in the globus pallidus and substantia nigra (not shown) and progressed in density and intensity by 4 weeks (Fig. 2A). This early-onset and progressive astrocytosis in the thalamus of Clcn7j/j mice was accompanied by similar laminar-specific events in the corresponding cortical regions of these mice, which became more widespread with increased age. At 1.5 weeks, numerous faintly stained astrocytes were already evident in the S1BF of Clcn7j/j mice (Fig. 3A). There was an intensely stained band of hypertrophic astrocytes in lamina IV and at the boundary of laminae V and VI by 3 weeks (Fig. 3A), which further increased in intensity by 4 weeks (Fig. 3A). This phenotype was mirrored in the primary visual and auditory cortex of Clcn7j/j mice without the intensely stained band of GFAP-IR in lamina IV and at the border between laminae V and VI (not shown). In marked contrast, Clcn6j/j mice displayed more generalized astrocytosis. In the thalamus of these mice, GFAP+ astrocytes were particularly evident at the level of the VPM/VPL and LGN (Fig. 2C) but were scattered widely with no specific focus. Similarly, numerous GFAP+ astrocytes were also evident throughout the cortical mantle of Clcn6j/j mice and were most densely packed in deeper laminae of sensory cortical regions (Fig. 3B) and in ventral areas of the cortex such as the piriform cortex (Fig. 3D). This phenotype was evident at 55 weeks and subsequently became slightly more pronounced with increased age. Thresholding image analysis revealed that this increase in GFAP-IR in the cortex of Clcn6 -deficient mice was significant at 55 and 92 weeks in the S1BF (Fig. 3C) and piriform cortex (Fig. 3E). A brain from an older mouse killed at 107 weeks displayed no further IR increase (Figs. 3B, D). Collectively, these data reveal markedly different degrees of astrocytosis in Clcn6- and Clcn7-deficient mice, with scattered and diffuse astrocytosis in the thalamus of Clcn6j/j mice and a more prominent cortical astrocytosis evident part way through disease progression, but that does not become more Neuropathology of ClC-6 & ClC-7/Ostm1-Deficient Mice widespread. In contrast, Clcn7j/j mice display more pronounced and rapidly progressing astrocytosis that is initially localized to individual thalamic nuclei and cortical laminae, spreading to become widespread with disease progression. Localized Microglial Activation in Clcn7j/j Mice Microglial activation is a consistent feature of many neurodegenerative disorders (34), but its timing and relationship to neuron loss and astrocytosis varies markedly. F4/80 is commonly used to reveal the progressive morphologic transformation of resting microglia that have numerous thin ramified processes to a brain macrophage-like morphology with an enlarged cell soma (23, 26, 34). In Clcn6j/j mice, a faintly stained network of F4/80+ processes was evident in all brain regions (Figs. 4A, C), even in the oldest mice, with very little difference from the appearance of microglia in control mice (Figs. 4A, C). In marked contrast, intense F4/80-IR was evident in the CNS of 4-week-old Clcn7j/j mice in the same brain regions that exhibited GFAP-IR (Fig. 4). Darkly stained F4/80+ microglia were present in VPM/VPL, LGN, MGN, and Rt nuclei of the thalamus and in the globus pallidus and substantia nigra (Fig. 4A). In the cortical mantle there was an intense band of F4/80+ microglia in lamina IV of S1BF, the same location as the most intense GFAP-IR (Figs. 4C, D). Similar to astrocytosis, microglial activation was also progressive and became more pronounced with increased age (Fig. 4B) but remained more localized than astrocytosis within each region. The microglia mostly had many long ramified processes and a cell soma that was scarcely visible indicating that they had not become fully activated (Fig. 4A, inserts). These data provide further new evidence for marked differences in nature and timing of the neuroimmune responses in Clcn6- and Clcn7-deficient mice. Compared with the pronounced and localized reactive changes in Clcn7j/j mice, these responses are far less prominent in Clcn6j/j mice. Storage Material in Clcn7j/j and Clcn6j/j Mice The ultrastructural appearance of storage material that accumulates in lysosomes of neurons in Clcn7j/j and Clcn6j/j mice resembles that seen in the NCLs (10, 11). In NCLs, this material accumulates throughout the brain, whereas neuron loss and gliosis occur in specific brain regions and cell types (22). The ultrastructural appearance of the storage material has been well characterized in Clcn7j/j and Clcn6j/j mice, and we examined the distribution of this material. Periodic acid Schiff staining revealed progressive accumulation of carbohydrates throughout the cortex of Clcn7j/j FIGURE 4. Microglial activation in Clcn7- but not in Clcn6-deficient mice. (A) Immunohistochemical staining for the microglial marker F4/80 revealed localized upregulation of microglia in basal ganglia and thalamic nuclei of 4-week-old Clcn7j/j mice but not in severely affected 92-week-old Clcn6j/j mice. Microglia in Clcn7j/j mice were highly ramified but were not fully transformed into amoeboid microglia. Thalamic nuclei boundaries are marked with white dashed lines. GP indicates globus pallidus; LGN, lateral geniculate nucleus; MGN, medial geniculate nucleus; Rt, reticular thalamic nucleus; SN, substantia nigra; VPM/VPL, ventral posterior thalamic nucleus. (B) Microglial activation was progressive in Clcn7j/j mice with a low level of F4/80 immunoreactivity evident at 1.5 weeks, followed by intense F4/80 immunoreactivity in thalamic nuclei at 3 weeks. (C) There were activated microglia in lamina IV of the somatosensory barrel field cortex (S1BF) of 4-week-old Clcn7j/j mice, but no F4/80+ microglia were apparent in the cortex of 92-week-old Clcn6j/j mice. (D) Corresponding fields of F4/80 and GFAP immunoreactivity in the S1BF of Clcn7j/j mice (white lines mark laminar boundaries). Scale bar = (AYC) 500 Km; higher-power views in A = 20 Km; (D) 250 Km. Ó 2010 American Association of Neuropathologists, Inc. 1235 Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. Pressey et al 1236 J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 Ó 2010 American Association of Neuropathologists, Inc. Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 mice, in accordance with Kasper et al (10) (Fig. 5A). Small cytoplasmic aggregates were initially present throughout all cortical laminae and subregions but were largest in pyramidal neurons of lamina V (Fig. 5A, inserts); they became more pronounced with increased age. The accumulation of carbohydrates in lamina V did not correlate with the most intense band of astrocytosis and microglial activation in lamina IV, which contains granule neurons that receive thalamic input (Fig. 5A, arrowheads). Compared with the cortex, accumulation of PAS+ material was less prominent and relatively delayed in the thalamus of Clcn7j/j mice (Fig. 5B). Cellular aggregates were spread diffusely not affecting every neuron. This generalized distribution of storage in the thalamus was in contrast to the localized activation of glia within specific thalamic nuclei of these mice (Figs. 2 and 4). We also used the lysosomal marker CtsD as a second marker of the intracellular distribution of storage material. In the cortex of Clcn7j/j mice, large round clumps of CtsD-IR were evident in lamina IV, where glial activation was most intense (Fig. 5A, arrowheads). Costaining with the microglial marker CD68 demonstrated that these large accumulations of storage material were present in microglia (Fig. 6, yellow arrows), a feature seen in many forms of NCL (22, 36). In contrast in lamina V, CtsD staining revealed long, thin accumulations of lysosomes in the proximal axon segment (Fig. 5A, insert). Here, costaining with the neuronal marker NeuN confirmed that the accumulated storage material was present in neurons (Fig. 6, red arrows). Prominent accumulations of CtsD-IR were also apparent within the thalamus of Clcn7j/j mice with large round clumps present in the VPM (Fig. 5B), which was revealed to be storage material in microglia in this nucleus (Fig. 4A). In contrast, in the adjacent posterior nucleus, where very few activated microglia were present, the storage material was visible within neurons as long, thin clumps from 3 weeks (Fig. 5B). In Clcn6j/j mice, CtsD-IR lysosomal aggregates were apparent within neurons throughout the cortex and the thalamus (Fig. 5C). CtsD-IR was evident in the initial axon segment of neurons and was clearly seen in lamina II/III and V of the cortex, in accordance with Poet et al (11). This CtsDIR was confirmed to be in neurons by staining for NeuN (Fig. 6, red arrows). Although microglial activation is not pronounced in these Clcn6j/j mice, where microglia were present, they also showed CtsD-IR (Fig. 6). This widespread distribution of neuronal storage in Clcn6j/j mice resembles Neuropathology of ClC-6 & ClC-7/Ostm1-Deficient Mice the generalized and widespread upregulation of GFAP in these mice. Neuron Loss in the Thalamocortical System of Clcn7j/j and Clcn6j/j Mice Reports of neuron loss have been confined to the hippocampus and cerebellum of Clcn7j/j mice (10). Neurodegeneration has not yet been reported in Clcn6j/j mice (11). To investigate relationships between astrocytosis, microglial activation, and neuron loss, we focused on the thalamic nuclei (VPM/VPL, MGN, and LGN) and cortical laminae (IV, V, and VI) in S1BF that displayed prominent astrocyte and microglial activation. We also investigated cortical interneuron populations that are known to degenerate in NCL models (22, 26). In 4-week-old Clcn7j/j mice, optical fractionator estimates of neuron number in the S1BF revealed a significant loss of neurons in laminae IV, V, and VI (Fig. 7A). This occurred first in laminae that receive projections from the thalamus (lamina IV) or that project back to the thalamus (lamina VI) and only subsequently in lamina V (Figs. 7, AYC). There was also a loss of SOM+, PV+, and CB+ interneurons in the S1BF of Clcn7j/j mice, but CR+ interneurons in this region were spared (Fig. 7D). In severely affected Clcn6j/j mice, significant S1BF neuron loss was confined to lamina VI thalamic feedback neurons, with no loss either of lamina IV and V neurons (Fig. 7F) or of SOM+ and PV+ interneurons (Fig. 7G). In the thalamus of Clcn7j/j mice, there was a delayed and significant loss of somatosensory relay neurons in VPM/VPL (Fig. 7E) but no loss of auditory or visual relay neurons in MGN or LGN, respectively (Fig. 7E) or of any subpopulation of thalamic relay neurons that were severely affected in Clcn6j/j mice (Fig. 7H). Thus, the extent and pattern of thalamocortical neuron loss differ markedly in Clcn7j/j and Clcn6j/j mice. Distinct Hippocampal Pathology in Clcn7j/j and Clcn6j/j Mice Hippocampal pathology has been briefly described in Clcn7j/j mice (10) and is a key feature of murine NCL, with loss of selected inhibitory interneuron populations (22). Therefore, we next compared pathologic changes in hippocampi of Clcn7j/j and Clcn6j/j mice. There was a progressive increase in staining for both GFAP and F4/80 in the hippocampus of Clcn7j/j mice (Fig. 8A) with increased age. Astrocytosis and microglial activation were most intense in FIGURE 5. Storage material is evident throughout Clcn7j/j and Clcn6j/j mice. Histochemical staining and immunohistochemistry (IHC) reveal storage material in the cortex and thalamus of Clcn7j/j and Clcn6j/j mice. (A) Periodic acid Schiff stain reveals storage material in the cortex of Clcn7j/j mice, from 1.5 weeks; staining becomes more intense at 3 and 4 weeks. Storage material is most prominent in lamina V (arrows, inserts). Likewise, immunohistochemical staining with Cathepsin D (CtsD) revealed storage material in the proximal axon segments of neurons in lamina V. This did not correlate with the most intense glial activation which occurred in lamina IV, where CtsD reactivity revealed storage material accumulation in microglia (arrow heads). (B) Cathepsin D (CtsD)+ storage material in microglia (arrowheads) that was evident in the thalamus of Clcn7j/j mice from 3 weeks. CtsD+ storage material is shown in microglia in the ventral posterior medial thalamic nucleus (VPM) and the adjacent posterior thalamic nucleus (Po) where microglia were not evident. Neuronal storage material is visible in long, thin clumps. (C) Storage material in the cortex of Clcn6j/j mice in lamina II/III and V; inserts show lamina V and throughout the thalamus; VPM is shown. Scale bar = 100 Km in cortex strips, 20 Km in cortex inserts, and 10 Km in thalamus images. Ó 2010 American Association of Neuropathologists, Inc. 1237 Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. Pressey et al J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 FIGURE 6. Neuronal lysosomal aggregates in Clcn7j/j and Clcn6j/j mice. (A) Costaining with the markers NeuN for neurons, CD68 for microglia, and cathepsin D (CtsD) for lysosomes shows a punctate distribution of lysosomes (green) in neurons (adjacent to blue nuclear NeuN) in control mice (+/+). (BYD) In Clcn7j/j and Clcn6j/j mice, CtsD reactivity is clustered together in clumps (green). In lamina IV in Clcn7j/j mice, CD68+ microglia are full of CtsD (yellow arrows) (B), whereas clumps of CtsD reactivity adjacent to neuronal nuclei were evident in lamina V (C, red arrows). In Clcn6j/j mice, CtsD reactivity was also clumped in neurons (D, red arrows) and microglia (yellow arrows). Scale bar = 10 Km. CA3 of the pyramidal layer (Fig. 8A, asterisk), and there was an additional region of intense F4/80-IR in the dentate gyrus (Fig. 8A, 2 asterisks). These reactive changes were accompanied by a progressive increase in storage material, with the distribution of the most prominent clumps of PAS reactivity correlating with the intense band of glial upregulation in CA3 (Fig. 8A). Smaller PAS+ aggregates were also present throughout CA1/2 and the dentate gyrus of Clcn7j/j mice. CtsD staining showed both storage material in microglia in CA3 (Fig. 8A, insert) and smaller clumps of neuronal storage throughout this pyramidal cell layer. As with the thalamocortical system, the hippocampal pathology in Clcn6j/j mice was much less pronounced than that of Clcn7j/j mice. GFAP-IR and F4/80-IR were com- 1238 parable to those in control mice (not shown). However, there were clumps of CtsD-IR storage material adjacent to neurons of the pyramidal cell layer in all hippocampal subfields (Fig. 8B), which were already evident in 55-week-old mice. There was a very restricted pattern of hippocampal neuron loss in both Clcn7j/j and Clcn6j/j mice that was only evident at the end stages of disease (Fig. 8C). Despite the loss of CA3 pyramidal neurons evident in Clcn7j/j mice (10), there was no significant loss of PV+, CR+, or CB+ interneurons in any region of the hippocampus of these mice (Figs. 8, EYG). Loss of SOM+ interneurons was confined to the stratum radiatum of aged Clcn7j/j mice (Fig. 8D). Similar results were obtained in aged Clcn6j/j mice, with no significant loss of SOM+ or PV+ interneurons in any subfield Ó 2010 American Association of Neuropathologists, Inc. Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 Neuropathology of ClC-6 & ClC-7/Ostm1-Deficient Mice FIGURE 7. Loss of distinct populations of thalamocortical neurons in Clcn7j/j and Clcn6j/j mice. (AYC) Optical fractionator estimates of Nissl+ neurons in the S1BF of 4-week-old Clcn7j/j mice revealed neuron loss in all lamina examined (IV, V, and VI) (A). Onset of neuron loss was first evident in lamina IV and VI at 3 weeks (B) but was not seen at 1.5 weeks (C). Neuron loss was relatively delayed in lamina V and was only evident at 4 weeks (A, B). (D) Assessment of survival of interneurons in the S1BF of Clcn7j/j mice revealed loss of somatostatin (SOM)+, parvalbumin (PV)+, and calbindin (CB)+ neurons but not of calretinin (CR)+ neurons. (E) Estimates of numbers of Nissl+ thalamic relay neurons revealed the selective loss of neurons from the ventral posterior thalamic nucleus (VPM/VPL) of 4-week-old Clcn7j/j mice. There was no difference in cell survival between Clcn7j/j mice and agematched controls at 3 weeks. (FYH) In Clcn6j/j mice, there was a selective loss of S1BF Nissl+ lamina VI neurons (F), but no effect on cell survival of SOM+ or PV+ interneurons in the S1BF (G) or of Nissl+ neurons in the VPM/VPL (H). *, p G 0.05; ***, p G 0.001, Student t test, n = 4Y5; error bars = SEM. Ó 2010 American Association of Neuropathologists, Inc. 1239 Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. Pressey et al J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 FIGURE 8. Distinct hippocampal pathology in Clcn7j/j and Clcn6j/j mice. (A) Immunohistochemical staining for astrocytes (glial fibrillary acidic protein [GFAP]), microglia (F4/80), and lysosomal storage material (CtsD) and histochemical staining for storage material (PAS; carbohydrates) identify CA3 pyramidal layer (*, inserts) and the dentate gyrus (**) as sites of pathology in Clcn7j/j mice. (B) In Clcn6j/j mice, CtsD immunoreactivity reveals accumulation of storage material in the pyramidal layer in long thin clumps (insert from CA3). (C) Schematic diagram illustrating the hippocampal subfields analyzed for hippocampal survival. (DYG) Counts of interneurons in discrete hippocampal subfields in Clcn7j/j mice reveal no loss of parvalbumin (PV)+, calretinin (CR)+, or calbindin (CB)+ interneurons and very restricted loss of somatostatin (SOM)+ interneurons in the radiatum (D) but not in any other subfield. (H, I) In Clcn6j/j mice, there was no evidence for loss of either SOM+ or PV+ interneurons. Scale bar = 200 Km in lowpower images, 20 Km in inserts (*, p G 0.05, Student t test, n = 3Y4, error bars = SEM). (Figs. 8H, I). Thus, although glial activation and storage material accumulation are prominent in the hippocampus of Clcn7j/j mice, effects on hippocampal interneurons are much less pronounced than in the cortex where there is significant loss of different types of interneurons. Similar Pathologic Events in Clcn7j/j and gl Mice Immunohistochemical staining with markers of glial activation in gl mice showed localized patches of intense 1240 GFAP-IR and F4/80-IR subcortically in the VPM/VPL, LGN, MGN, and Rt thalamic nuclei and in the globus pallidus and substantia nigra (Fig. 9A), similar to that seen in Clcn7j/j mice. In the cortex of gl mice, there was widespread astrocytosis with a dark band of intensely stained and densely packed astrocytes in lamina IV and the boundary of lamina V/VI of the S1BF (Fig. 9B). The distribution of F4/80-IR in the cortex of gl mice mirrored that of GFAP-IR, with an additional intense band of microglial activation in lamina IV of the S1BF (Fig. 9B). The distribution of storage material Ó 2010 American Association of Neuropathologists, Inc. Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 Neuropathology of ClC-6 & ClC-7/Ostm1-Deficient Mice FIGURE 9. Glial activation and storage material in grey-lethal mice resemble that of Clcn7j/j mice. (A, B) Immunohistochemical staining for glial fibrillary acidic protein (GFAP), F4/80, cathepsin D (CtsD), and PAS stain revealed a similar pattern of glial activation and storage material in the thalamus (A) and the cortex (B) of gl and Clcn7j/j mice. Higher magnification images show that microglia were rarely fully transformed in either mouse model (A, inserts). Thalamic nuclei boundaries are marked with white dashed lines. GP indicates globus pallidus; LGN, lateral geniculate nucleus; MGN, medial geniculate nucleus; Rt, reticular thalamic nucleus; SN, substantia nigra; VPM/VPL, ventral posterior thalamic nucleus. Scale bar = (A) 500 Km; insert 20 Km; (B) 250 Km; inserts 20 Km. also closely resembled that of Clcn7j/j mice, with microglia full of CtsD+ lysosomes in lamina IV (Fig. 9B, arrows) and prominent storage material in neurons of lamina V (Fig. 9B, arrowheads). gl mice showed overall atrophy of the cortex (Fig. 10A) that involved atrophy of S1BF, LEnt and V1 but not M1 (Fig. 10B). Consistent with previous qualitative reports (21), white matter structures were also atrophied in gl mice, with significantly reduced volume of the corpus callosum, internal capsule (Fig. 10C) and cerebral peduncle (not shown). Similar to Clcn7j/j mice, optical fractionator estimates of neuron number in gl mice showed a significant loss of cortical neurons in laminae IV, V and VI of the S1BF and of SOM+ and PV+ interneurons in S1BF (Figs. 10D, E) and thalamic relay neurons in VPM/VPL (Fig. 10F). As in Clcn7j/j mice, hippocampal interneurons were spared (Figs. 10G, H). Taken together, these data reveal that the neuropathologic phenotypes of gl and Clcn7j/j mice are remarkably similar. The patterns of glial activation and vulnerability of thalamocortical neuron populations and sparing of hippocampal interneurons suggest that the same regions are pathologic targets of the disease process in these 2 models. There Ó 2010 American Association of Neuropathologists, Inc. was a mild but significant atrophy of the thalamus, hippocampus, LEnt, and V1 in gl mice but not in Clcn7j/j mice. DISCUSSION We provide the first detailed comparative neuropathologic characterization of Clcn6j/j, Clcn7j/j, and gl mice and compare these novel findings to established landmarks of disease progression in the NCLs (Table). Our study has revealed significant new data of the detailed neurodegenerative and reactive phenotypes of CLC-deficient mice. The analysis revealed that the underlying neuropathology in Clcn7j/j and gl mice is remarkably similar, with comparable patterns of selective cell loss and distribution of reactive astrocytes, microglia, and storage material. These selective effects on sensory regions of cortex and the thalamus seen in Clcn7j/j and gl mice closely mirror those seen in certain forms of NCL (Table). We also found a much milder and more slowly progressing pathologic phenotype in Clcn6j/j mice with astrocytosis confined to the deeper layers of the cortex that also displayed neuron loss. Thus, the neuropathologic features of Clcn6j/j mice display less resemblance to those of the NCLs (Table). These data reinforce that 1241 Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. Pressey et al J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 FIGURE 10. Similar patterns of atrophy and neuron loss in grey-lethal and Clcn7j/j mice. (A) Cavalieri estimates of regional volume reveal significant atrophy of cortex, thalamus, and hippocampus of gl mice versus age-matched controls. (B) Cavalieri estimates of cortical subfields reveal selective atrophy of the somatosensory barrel field cortex (S1BF), primary visual (V1), lateral entorhinal (LEnt), but not the primary motor cortex (M1) of gl mice. (C) Assessment of white matter tract volumes reveals atrophy of both the corpus callosum and internal capsule. (DYF) Optical fractionator estimates of neuronal number reveal significant loss of Nissl+ neurons in lamina IVYVI of the S1BF, of somatostatin (SOM)+ and parvalbumin (PV)+ neurons in the S1BF, and of Nissl+ neurons in the VPM/VPL in gl mice. (G, H) Manual counts of SOM+ and PV+ interneurons in hippocampal subfields did not reveal any evidence for neuron loss. *, p G 0.05; **, p G 0.01, Student t test, n = 4; error bars = SEM. 1242 Ó 2010 American Association of Neuropathologists, Inc. Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 Neuropathology of ClC-6 & ClC-7/Ostm1-Deficient Mice TABLE. Comparison of Neuropathologic Phenotypes of Clcn7j/j, gl, and Clcn6j/j Mice to a Typical Neuronal Ceroid Lipofuscinosis Mouse Model Pathologic Feature Late-stage cortical atrophy Selective regional cortical atrophy Atrophy of white matter tracts Intense localized astrocytosis in thalamus and cortex Intense localized microglial activation Neuronal storage material accumulation Storage material spread indiscriminately throughout the brain No direct correlation between storage material distribution and neuron loss Thalamic relay neuron loss Cortical neuron loss Cortical interneuron loss Hippocampal interneuron loss NCL Clcn7j/j / gl Clcn6j/j ( ( ( ( ( ( ( ( ( ( ( ( ( ( S1BF/V1 and LEnt* ( widespread ( ( ( , ( ( VPM/VPL only ( ( , , ( LEnt only ( corpus callosum only , , ( , ( , ( Lamina VI only , , Features shared with typical neuronal ceroid lipofuscinosis (NCL) mouse models are indicated with ( ; features that are not typical are indicated as ,. Clcn7j/j and the naturally occurring gl mutant mice share greater similarities with typical NCL models than do Clcn6j/j mice. *Atrophy of S1BF, V1 and LEnt where evident in gl mice; selective atrophy of S1BF was evident in Clcn7j/j mice. LEnt, lateral entorhinal cortex; S1BF, somatosensory barrel field cortex; V1, primary visual cortex; VPM/VPL, ventral posterior thalamic nucleus. mutations in these two CLC genes have distinctly different pathologic consequences in the CNS. This study not only furthers our understanding of the role of these genes within the brain, but it is also important in their consideration as candidate genes for variant forms of NCL. Mice Deficient for Late Endosomal/Lysosomal CLC Proteins Display Features of NCL Initial characterization of Clcn7j/j, Clcn6j/j, and mice revealed some phenotypic similarities to Clcn3 mouse models of NCL (10, 11, 15, 37, 38). Despite their severe neurodegeneration, Clcn3j/j mice are not thought to be an accurate model of NCL because they lack NCL-like storage material (10). Therefore, we focused on Clcn7j/j and Clcn6j/j mice. In addition, gl mice bear many similarities to Clcn7j/j mice and the Ostm1 gene that is mutated in these mice encodes a A-subunit for ClC-7 (8). Our data provide the first evidence that the underlying neuropathology of gl mice also closely resembles that of Clcn7j/j mice. Furthermore, we found that many of the typical NCL phenotypes are also apparent in Clcn7j/j mice (Table). The early onset and rapid progression of the neurodegenerative phenotype of Clcn7j/j mice most closely resembles that of CtsD -deficient mice (24), which have been identified as modeling congenital NCL (39). Both mice die before 2 months and display particularly pronounced pathologic effects within the somatosensory thalamocortical system, including selective loss of somatosensory relay neurons in VPM/VPL and their cortical targets in lamina IV of S1BF (Figs. 2Y4 and 7) (24). Clcn7j/j mice also display the localized astrocytosis and microglial activation within individual thalamic nuclei and loss of cortical interneuron populations that are also evident in multiple forms of NCL (Table) (23Y26). In contrast to multiple forms of NCL, cortical pathology starts earlier and is more severe than in the thalamus in Clcn7j/j mice. Neuron loss occurs first and most profoundly within the cortex; subsequently, more restricted and late-stage neuron loss was evident in the thalamus of j/j Ó 2010 American Association of Neuropathologists, Inc. Clcn7j/j mice (Fig. 7). Such predominance of cortical pathology was also recently reported in the Cln5-deficient mouse model of Finnish variant late infantile NCL (25) and highlights the distinct timing and staging of pathologic events between models of NCL. Furthermore, despite a loss of pyramidal neurons in CA3 of Clcn7j/j mice (10), hippocampal inhibitory interneurons were spared in these mice (Fig. 8; Table). This is surprising considering that loss of hippocampal interneurons occurs consistently in murine models of infantile, juvenile and variant NCL (26, 28, 31); however, this has not been characterized in CtsD-deficient mice. This study also revealed a close correlation between regions of glial activation and neuron loss in Clcn7j/j mice. In the thalamus, astrocytosis was first seen in the VPM/VPL (Fig. 2), the only thalamic nuclei that later displays neuron loss (Fig. 7). In the cortex, astrocytosis and microglial activation were most pronounced in lamina IV and the boundary of lamina V and VI (Figs. 3 and 4), and neuron loss was evident first in lamina IV and VI (Fig. 7). The premature death of Clcn7j/j mice precludes any investigation of whether neuron loss and glial activation would have become more profound in older mice. Inactivation of Clcn7 only in forebrain neurons, however, produces mice that have a normal life span (16), providing a useful tool to investigate the downstream effects of Clcn7 deficiency in neurons. Clcn7lox/lox; EMX1cre mice display glial activation solely in the brain regions in which Clcn7 was deleted and go on to display a near complete loss of neurons (16). Our data support the hypothesis that both astrocytes and microglia are activated by localized factors released by dysfunctional Clcn7-deficient neurons before they die. Our analysis also revealed that Clcn6j/j mice display few similarities to the well-defined neuropathologic phenotypes of NCL mouse models with no evidence of localized glial responses within the thalamus, loss of thalamic relay neurons or of GABAergic interneuron populations in the cortex or hippocampus (Table) (22). However, these mildly affected mice do show cortical and white matter atrophy, 1243 Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. Pressey et al J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 generalized astrocytosis, and neurodegeneration in lamina VI of the S1BF where astrocytosis was most pronounced. Thus, like Clcn7j/j mice, and distinct from most forms of NCL, the predominant pathologic alterations occur in the cortex of Clcn6j/j mice. Finally, our analysis of gl mice revealed that they share many phenotypes with Clcn7j/j mice, with reactive astrocytes and microglia in the same thalamic nuclei, cortical laminae, and midbrain nuclei (Figs. 9 and 10). Moreover, neuron loss in the S1BF and VPM/VPL occurred to a similar extent in gl mice and Clcn7j/j mice. The only pathologic discrepancy between these mice was in the small, but significant, atrophy of the hippocampus, thalamus, LEnt, and V1 in gl but not in Clcn7j/j mice. Thus, loss of either Ostm1 or ClC-7 has remarkably similar consequences on both neuron populations and glia. Complex Relationship of Storage Material With Pathology The NCLs are characterized by uniform intralysosomal accumulation of autofluorescent storage material with no particular preference for any type of neuron or brain region (13, 15). Indeed, there is no evidence that this storage material has a direct role in NCL pathogenesis. The present results suggest that the relationships between accumulation of storage material, distribution of neuron loss, and glial activation in Clcn7j/j and Clcn6j/j mice are likely to be complex and may differ between and within brain regions. In the cortex and thalamus of Clcn7j/j mice, there was no correlation between storage and neuron loss or glial activation. For example, the most intense band of astrocytosis and microglial activation occurred in lamina IV and neuron loss was an early event at 3 weeks; however, there was very little storage material in this lamina. The opposite was true for lamina V, in which the most marked storage material was observed. Very little microglial activation and neuron loss occurred late, at 4 weeks. In the thalamus of Clcn7j/j mice, despite the pronounced localized glial activation within individual thalamic nuclei, neuronal storage material was present in a generalized pattern throughout the thalamus. The clearest correlation between these pathologic events was, however, apparent in the hippocampus where astrocytosis, microglial activation, storage material, and neuron loss were all most prominent in CA3. The generalized and less pronounced pathology of Clcn6j/j mice makes for a more simple comparison; in the thalamus and hippocampus, the pattern of astrocytosis correlated with the widespread indiscriminate distribution of storage material accumulation with no accompanying neuron loss. However, in the cortex of these mice, there was a similar disconnect between events as in Clcn7j/j mice, with neuron loss restricted to lamina VI and the most prominent storage material in lamina V. Candidate Genes for the NCLs The NCLs are a heterogeneous group of diseases with mutations in a different gene responsible for each disease subtype (12). However, there are variant and adult-onset cases 1244 of human NCL with no identified disease-causing mutation and concerted efforts are being made to identify the underlying genes (http://www.ucl.ac.uk/ncl/RNGC.shtml). Much valuable information on NCL-like disease phenotypes can be obtained from engineered or spontaneous mutant mice (22). For example, Cln8mnd and Cln6nclf were initially identified as candidate NCL mouse models and were subsequently confirmed as bearing mutations in genes linked to 2 variant forms of NCL (40Y43). Similarly, the CtsDj/j mouse was reported to show several NCL-like phenotypes (44, 45), long before mutations in CTSD were associated with congenital and lateronset forms of NCL (39, 46). Reports of mice deficient in chloride transporter genes have raised the possibility of finding another class of genes that might be responsible for novel forms of NCLs (10, 11). Our current data support previous observations that Clcn7j/j mice bear many similarities to NCL models, showing lysosomal storage, retinal and optic nerve degeneration, and selective loss of hippocampal and cerebellar neurons (10). Importantly, mice with ClC-7Yrescued osteoclasts live only a few weeks longer and die of their CNS phenotype (10). Clinically, patients with mutations in CLCN7 or OSTM1 have CNS pathology and associated neurologic manifestations, in addition to osteopetrosis (19, 21). Here, we show for the first time that mice deficient in Clcn7 or Ostm1 bear a close resemblance to the progressive neuropathologic phenotype of early onset forms of NCL (24), with similar populations of neurons showing vulnerability and a comparable distribution of glial activation (Table). Although Clcn7 deficiency does not specifically result in a form of NCL, these data underline that loss of murine Clcn7 results in CNS outcomes that resemble NCL with distinct, yet possibly connected, underlying molecular mechanisms. Indeed, the data provide evidence for syndromic forms of NCL in which other symptoms or disease features are evident in addition to the conventional presentation of NCL. The use of Bsyndromic NCL[ may be novel for these disorders but is analogous to the nomenclature used to describe other inherited diseases that have forms in which additional symptoms are evident, for example, the osteopetrosis evident in CLCN7 deficiency (4, 17, 18) or the blindness in Usher syndrome, a form of syndromic deafness (47Y49). Clcn7j/j mice and Clcn6j/j mice have distinct phenotypes (Table). Indeed, beyond accumulation of autofluorescent storage material, Clcn6j/j mice display very few of the common neuropathologic features of NCL. The cortex is the main site of pathology in Clcn6j/j mice, with no typical early localized thalamic glial activation or cell loss. Clcn6j/j mice display a distinctive accumulation of storage material in the proximal segment of the axon hillock (11), a feature that is not typical of murine NCLs but has been reported in adult-onset human NCL (50, 51). This feature was also seen in Clcn7j/j mice in cortical neurons (lamina V; Fig. 5A, inserts) and the thalamus (Fig. 5B, arrows). As such, this particular phenotype may be specific to these CLC-deficient mice rather than simply reflecting a delayed disease onset. Screening a large panel of adult-onset NCL cases did not reveal any homozygous or compound heterozygous diseasecausing mutations in CLCN6 (11). However, there is currently Ó 2010 American Association of Neuropathologists, Inc. Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 no model for variant late-onset forms of NCL and few histologic data are available from human cases. Indeed, abnormal pain sensitivity, storage material accumulation (11), and our data suggest that Clcn6 may prove to be a candidate gene for late-onset forms of NCL. Our data also reveal the detailed effects of mutations in these genes on the CNS. Like many of the proteins involved in NCL pathogenesis (52), ClC-6, ClC-7, and Ostm1 are expressed within the endosomal/lysosomal system (1) and colocalize with late endosomal/lysosomal markers (4, 8, 10, 11). Lysosomal Clj conductance was thought to play an important role in vesicle acidification by neutralizing the H+ conductance created by the vacuolar H+ ATPase (2). The lack of this Clj current and impaired development of the acidsecreting ruffled border prevents the acidification of the resorption lacuna in Clcn7j/j osteoclasts, resulting in defective bone resorption and osteopetrosis (4). Nevertheless, the pH of lysosomes in cultured neurons is indistinguishable between control and mutant mice (8, 10). The production of neuropathologic phenotypes shown here and the accumulation of storage material in Clcn6j/j, Clcn7j/j, and gl mice (8, 10, 11) suggest that the function of these proteins is significant for late endosomal/lysosomal function in neurons. Indeed, recent evidence from Clcn7j/j kidney cells suggests that this gene is important for lysosomal protein degradation (16). Although the mechanisms that underlie chloride transporter deficiency remain poorly understood, it is apparent that the consequences of mutations in different members of a gene family can differ markedly, even within the same series of interconnected pathways. Our current data from Clcn6j/j and Clcn7j/j mice provide new evidence in support of this hypothesis with widely divergent and dissimilar neuropathologic phenotypes in these mutant mice In conclusion, our data clearly highlight that loss of either Clcn7 or Ostm1 result in similar neuropathologic phenotypes. The broad range of neurologic and neuropathologic similarities between Clcn7j/j, gl mice, and early-onset forms of NCL hints at the involvement of convergent functional pathways in the pathogenesis of these disorders and suggests that mutations in CLCN7 or OSTM1 may underlie syndromic NCL. Our data also highlight that loss of Clcn6 has distinct effects on the CNS and suggest that mutations in CLCN6 could be responsible for NCL-like adult-onset neurologic dysfunction. Evidence for the physiological function of Clcn6 and Clcn7 as Clj/H+ exchangers is now emerging (2, 53, 54), and understanding these processes in neurons will provide crucial insights into the possible mechanisms of cell death and why materials accumulate within lysosomes. REFERENCES 1. Jentsch TJ. CLC chloride channels and transporters: from genes to protein structure, pathology and physiology. Crit Rev Biochem Mol Biol 2008;43:3Y36 2. Jentsch TJ. Chloride and the endosomal-lysosomal pathway: emerging roles of CLC chloride transporters. J Physiol 2007;578:633Y40 3. Koch MC, Steinmeyer K, Lorenz C, et al. The skeletal muscle chloride channel in dominant and recessive human myotonia. Science 1992;257: 797Y800 4. Kornak U, Kasper D, Bosl MR, et al. Loss of the ClC-7 chloride channel leads to osteopetrosis in mice and man. Cell 2001;104:205Y15 Ó 2010 American Association of Neuropathologists, Inc. Neuropathology of ClC-6 & ClC-7/Ostm1-Deficient Mice 5. Lloyd SE, Pearce SH, Fisher SE, et al. A common molecular basis for three inherited kidney stone diseases. Nature 1996;379:445Y49 6. Simon DB, Bindra RS, Mansfield TA, et al. Mutations in the chloride channel gene, CLCNKB, cause Bartter’s syndrome type III. Nat Genet 1997;17:171Y78 7. Brandt S, Jentsch TJ. ClC-6 and ClC-7 are two novel broadly expressed members of the CLC chloride channel family. FEBS Lett 1995;377: 15Y20 8. Lange PF, Wartosch L, Jentsch TJ, et al. ClC-7 requires Ostm1 as a beta-subunit to support bone resorption and lysosomal function. Nature 2006;440:220Y23 9. Chalhoub N, Benachenhou N, Rajapurohitam V, et al. Grey-lethal mutation induces severe malignant autosomal recessive osteopetrosis in mouse and human. Nat Med 2003;9:399Y406 10. Kasper D, Planells-Cases R, Fuhrmann JC, et al. Loss of the chloride channel ClC-7 leads to lysosomal storage disease and neurodegeneration. EMBO J 2005;24:1079Y91 11. Poet M, Kornak U, Schweizer M, et al. Lysosomal storage disease upon disruption of the neuronal chloride transport protein ClC-6. Proc Natl Acad Sci U S A 2006;103:13854Y59 12. Jalanko A, Braulke T. Neuronal ceroid lipofuscinoses. Biochim Biophys Acta 2009;1793:697Y709 13. Haltia M. The neuronal ceroid-lipofuscinoses. J Neuropathol Exp Neurol 2003;62:1Y13 14. Siintola E, Lehesjoki AE, Mole SE. Molecular genetics of the NCLsV Status and perspectives. Biochim Biophys Acta 2006;1762:857Y64 15. Goebel HH, Wisniewski KE. Current state of clinical and morphological features in human NCL. Brain Pathol 2004;14:61Y69 16. Wartosch L, Fuhrmann JC, Schweizer M, et al. Lysosomal degradation of endocytosed proteins depends on the chloride transport protein ClC-7. FASEB J 2009;23:4056Y68 17. Cleiren E, Benichou O, Van HE, et al. Albers-Schonberg disease (autosomal dominant osteopetrosis, type II) results from mutations in the ClCN7 chloride channel gene. Hum Mol Genet 2001;10:2861Y67 18. Quarello P, Forni M, Barberis L, et al. Severe malignant osteopetrosis caused by a GL gene mutation. J Bone Miner Res 2004;19:1194Y99 19. Frattini A, Pangrazio A, Susani L, et al. Chloride channel ClCN7 mutations are responsible for severe recessive, dominant, and intermediate osteopetrosis. J Bone Miner Res 2003;18:1740Y47 20. Maranda B, Chabot G, Decarie JC, et al. Clinical and cellular manifestations of OSTM1-related infantile osteopetrosis. J Bone Miner Res 2008; 23:296Y300 21. Pangrazio A, Poliani PL, Megarbane A, et al. Mutations in OSTM1 (grey lethal) define a particularly severe form of autosomal recessive osteopetrosis with neural involvement. J Bone Miner Res 2006;21:1098Y105 22. Cooper JD, Russell C, Mitchison HM. Progress towards understanding disease mechanisms in small vertebrate models of neuronal ceroid lipofuscinosis. Biochim Biophys Acta 2006;1762:873Y89 23. Kielar C, Maddox L, Bible E, et al. Successive neuron loss in the thalamus and cortex in a mouse model of infantile neuronal ceroid lipofuscinosis. Neurobiol Dis 2007;25:150Y62 24. Partanen S, Haapanen A, Kielar C, et al. Synaptic changes in the thalamocortical system of cathepsin DYdeficient mice: A model of human congenital neuronal ceroid-lipofuscinosis. J Neuropathol Exp Neurol 2008;67:16Y29 25. von Schantz C, Kielar C, Hansen SN, et al. Progressive thalamocortical neuron loss in Cln5 deficient mice: Distinct effects in Finnish variant late infantile NCL. Neurobiol Dis 2009;34:308Y19 26. Bible E, Gupta P, Hofmann SL, et al. Regional and cellular neuropathology in the palmitoyl protein thioesterase-1 null mutant mouse model of infantile neuronal ceroid lipofuscinosis. Neurobiol Dis 2004;16: 346Y59 27. Stoward PJ. Some comments on the mechanism of the Schiff reaction. J Histochem Cytochem 1966;14:681Y83. 28. Pontikis CC, Cella CV, Parihar N, et al. Late onset neurodegeneration in the Cln3j/j mouse model of juvenile neuronal ceroid lipofuscinosis is preceded by low level glial activation. Brain Res 2004;1023:231Y42 29. Paxinos G, Franklin BJ. The Mouse Brain in Stereotaxic Coordinates. 2nd ed. San Diego, CA: Academic Press, 2001 30. West MJ, Slomianka L, Gundersen HJ. Unbiased stereological estimation of the total number of neurons in the subdivisions of the rat hippocampus using the optical fractionator. Anat Rec 1991;231:482Y97 1245 Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited. Pressey et al J Neuropathol Exp Neurol Volume 69, Number 12, December 2010 31. Cooper JD, Messer A, Feng AK, et al. Apparent loss and hypertrophy of interneurons in a mouse model of neuronal ceroid lipofuscinosis: Evidence for partial response to insulin-like growth factor-1 treatment. J Neurosci 1999;19:2556Y67 32. Pontikis CC, Cotman SL, MacDonald ME, et al. Thalamocortical neuron loss and localized astrocytosis in the Cln3Deltaex7/8 knock-in mouse model of Batten disease. Neurobiol Dis 2005;20:823Y36 33. Gundersen HJ, Jensen EB. The efficiency of systematic sampling in stereology and its prediction. J Microsc 1987;147:229Y63 34. Raivich G, Bohatschek M, Kloss CU, et al. Neuroglial activation repertoire in the injured brain: Graded response, molecular mechanisms and cues to physiological function. Brain Res Brain Res Rev 1999;30: 77Y105 35. Seth P, Koul N. Astrocyte, the star avatar: Redefined. J Biosci 2008;33: 405Y21 36. Oswald MJ, Palmer DN, Kay GW, et al. Glial activation spreads from specific cerebral foci and precedes neurodegeneration in presymptomatic ovine neuronal ceroid lipofuscinosis (CLN6). Neurobiol Dis 2005;20:49Y63 37. Yoshikawa M, Uchida S, Ezaki J, et al. CLC-3 deficiency leads to phenotypes similar to human neuronal ceroid lipofuscinosis. Genes Cells 2002;7:597Y605 38. Stobrawa SM, Breiderhoff T, Takamori S, et al. Disruption of ClC-3, a chloride channel expressed on synaptic vesicles, leads to a loss of the hippocampus. Neuron 2001;29:185Y96 39. Siintola E, Partanen S, Stromme P, et al. Cathepsin D deficiency underlies congenital human neuronal ceroid-lipofuscinosis. Brain 2006;129: 1438Y45 40. Bronson RT, Lake BD, Cook S, et al. Motor neuron degeneration of mice is a model of neuronal ceroid lipofuscinosis (Batten’s disease). Ann Neurol 1993;33:381Y85 41. Bronson RT, Donahue LR, Johnson KR, et al. Neuronal ceroid lipofuscinosis (nclf), a new disorder of the mouse linked to chromosome 9. Am J Med Genet 1998;77:289Y97 42. Gao H, Boustany RM, Espinola JA, et al. Mutations in a novel 1246 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. CLN6-encoded transmembrane protein cause variant neuronal ceroid lipofuscinosis in man and mouse. Am J Hum Genet 2002;70:324Y35 Ranta S, Zhang Y, Ross B, et al. The neuronal ceroid lipofuscinoses in human EPMR and mnd mutant mice are associated with mutations in CLN8. Nat Genet 1999;23:233Y36 Saftig P, Hetman M, Schmahl L, et al. Mice deficient for the lysosomal proteinase cathepsin D exhibit progressive atrophy of the intestinal mucosa and profound destruction of lymphoid cells. EMBO J 1995;14:3599Y608 Koike M, Nakanishi H, Saftig P, et al. Cathepsin D deficiency induces lysosomal storage with ceroid lipofuscin in mouse CNS neurons. J Neurosci 2000;20:6898Y906 Steinfeld R, Reinhardt K, Schreiber K, et al. Cathepsin D deficiency is associated with a human neurodegenerative disorder. Am J Hum Genet 2006;78:988Y98 Cohen M, Bitner-Glindzicz M, Luxon L. The changing face of Usher syndrome: Clinical implications. Int J Audiol 2007;46:82Y93 Saihan Z, Webster AR, Luxon L, et al. Update on Usher syndrome. Curr Opin Neurol 2009;22:19Y27 Seeliger MW, Fischer MD, Pfister M. [Usher syndrome: Clinical features, diagnostic options, and therapeutic prospects]. Ophthalmologe 2009;106:505Y11 Berkovic SF, Carpenter S, Andermann F, et al. Kufs’ disease: A critical reappraisal. Brain 1988;111:27Y62 Goebel HH, Braak H, Seidel D, et al. Morphologic studies on adult neuronal-ceroid lipofuscinosis (NCL). Clin Neuropathol 1982;1:151Y62 Kyttala A, Lahtinen U, Braulke T, et al. Functional biology of the neuronal ceroid lipofuscinoses (NCL) proteins. Biochim Biophys Acta 2006; 1762:920Y33 Neagoe I, Stauber T, Fidzinski P, et al. The late endosomal ClC-6 mediates proton/chloride countertransport in heterologous plasma membrane expression. J Biol Chem 2010;285:21689Y97 Weinert S, Jabs S, Supanchart C, et al. Lysosomal pathology and osteopetrosis upon loss of H+-driven lysosomal Clj accumulation. Science 2010;328:1401Y3 Ó 2010 American Association of Neuropathologists, Inc. Copyright @ 2010 by the American Association of Neuropathologists, Inc. Unauthorized reproduction of this article is prohibited.