Hairy Baskets Associated With Degenerative Purkinje Cell Changes in Essential Tremor

Transcription

1 J Neuropathol Exp Neurol Copyright Ó 2010 by the American Association of Neuropathologists, Inc. Vol. 69, No. 3 March 2010 pp. 262Y271 ORIGINAL ARTICLE Hairy Baskets Associated With Degenerative Purkinje Cell Changes in Essential Tremor Cordelia R. Erickson-Davis, BA, Phyllis L. Faust, MD, PhD, Jean-Paul G. Vonsattel, MD, Sachin Gupta, MD, Lawrence S. Honig, MD, PhD, and Elan D. Louis, MD, MSc Abstract Essential tremor (ET) is one of the most common neurologic diseases. Increased numbers of torpedoes and Purkinje cell (PC) loss have been documented in the brains of patients with ET. We recently observed a dense and tangled appearance (Bhairiness[) of the basket cell axonal plexuses that surround PC soma in Bielschowsky preparations of cerebellar cortex in ET brains. Here, we assessed basket cell Bhairiness[ in 37 ET (32 cerebellar ET; 5 Lewy body variant ET), 21 nondisease control, and 48 disease control brains using a semiquantitative scale. In 8 cerebellar ET cases (25%), there were high basket scores (rating = 3), whereas no Lewy body variant ET, 1 nondisease control (4.8%), and 2 diseased controls (4.2%) had high basket scores (p = 0.001). The hairy basket scores correlated with numbers of torpedoes (p G 0.001) and inversely with numbers of PCs (p = 0.06). Axonal plexus density obtained by image analysis of basket cell processes traced from digitized images was higher in ET than in nondiseased control cases (p = 0.016). Closely spaced sites of synaptic contact between basket cell processes and PCs were identified by electron microscopy in ET cases. These data indicate that structural changes are not restricted to PCs in ET, and that other neurons within their functional network may be involved in its pathogenesis. Key Words: Basket cells, Cerebellum, Essential tremor, Neurodegenerative, Pathology, Pathophysiology, Purkinje cells. INTRODUCTION Essential tremor (ET) is a chronic brain disease, the most recognizable feature of which is a 4- to 12-Hz action tremor (1, 2). It is one of the most common neurologic disorders present in 4.0% of individuals 40 years or older (3), and in as many as 21.7% of individuals aged 95 years and older From the GH Sergievsky Center (CED, SG, LSH, EDL); Department of Pathology and Cell Biology (PLF, JPGV); Taub Institute for Research on Alzheimer s Disease and the Aging Brain (JPGV, LSH, EDL); Department of Neurology, College of Physicians and Surgeons (LSH, EDL); and Department of Epidemiology, Mailman School of Public Health, Columbia University, New York, New York (EDL). Send correspondence and reprint requests to: Elan D. Louis, MD, MSc, Unit 198, Neurological Institute, 710 West 168th Street, New York, NY 10032; EDL2@columbia.edu Supported by Grant Nos. R01 NS42859, P50 AG08702, and P01 AG07232 from the National Institutes of Health (Bethesda, MD); the Parkinson s Disease Foundation (New York, NY); and the Arlene Bronstein Essential Tremor Research Fund (Columbia University). Online-only figures are available at (4). The tremor is usually progressive (5) and produces functional disabilities (6). Recent evidence also suggests an associated increased risk of mortality (7). A wealth of clinical data suggests that ET is a disorder of cerebellar dysfunction. Intention (i.e. Bcerebellar[) tremor of the hands occurs in 44% of ET cases (8). Abnormalities in tandem gait and balance have been repeatedly described in ET patients (9Y11), and ET patients with intention tremor may also have other cerebellar signs such as dysdiadochokinesia (12). Eye movement abnormalities that indicate cerebellar dysfunction have also been described in ET (13). Unilateral cerebellar stroke has been reported to abolish ipsilateral arm tremor in ET (14), and cerebellar outflow pathways are the targets of deep brain stimulation surgery, which is effective in treating ET (15, 16). Numerous neuroimaging studies using functional magnetic resonance imaging (17), positron emission tomography (18, 19), [ 1 H] magnetic resonance spectroscopic imaging (20, 21), diffusion-tensor imaging (22), and voxel-based morphometry (23) have provided evidence that cerebellar structure and function are abnormal in ET. Until recently, there had been few postmortem studies of ET (24). Studies that explore the pathological anatomy and improve our understanding of the pathophysiology of ET are critically important as there is no cure for ET and first-line medications, of which there are only two, are estimated to be ineffective in as many as 50% of patients (25). In each of 2 recent large postmortem series, the presence of degenerative changes in the cerebellum has become evident (24, 26Y28). To date, these changes consist of an increased number of torpedoes (5- to 6-fold greater than seen in age-matched control brains) and a mild, approximately 40%, loss of Purkinje cells (PCs) (27, 29). Whether the structural changes in ET are restricted to PCs or whether they involve other neurons in their functional network has not been investigated. Recently, we observed an unusual dense and tangled (we now term Bhairy[) appearance of the basket cell axonal plexus surrounding PC soma in Bielschowsky preparations of cerebellar cortical sections in ET cases. Basket cells are F-aminobutyric acidyergic inhibitory interneurons found in the molecular layer; they send out axonal collaterals to form a pericellular basket around PC perikarya. To our knowledge, there are only 2 previous reports describing a dense and tangled appearance of basket cell axonal plexus profiles occurring with cerebellar degeneration: a single case of Creutzfeldt-Jakob disease (30) and an early study examining J Neuropathol Exp Neurol Volume 69, Number 3, March 2010

2 J Neuropathol Exp Neurol Volume 69, Number 3, March 2010 brains of individuals diagnosed with Bsenile dementia[ (31). Because this pathologic finding may provide important mechanistic insights into the pathogenesis of ET, we assessed basket cell changes in ET, other neurodegenerative diseases, and nondiseased control cases. MATERIALS AND METHODS Brain Repository This study was conducted at the Essential Tremor Centralized Brain Repository at the New York Brain Bank, Columbia University (32). The 106 brains included 37 from ET cases; 31 (83.8%) of these were collected prospectively beginning in 2003, and 6 were archival. All 37 had been diagnosed by their treating neurologist, and all diagnoses were confirmed using Essential Tremor Centralized Brain Repository criteria (27). Essential tremor cases were divided into 2 groups based on the presence or absence of brainstem Lewy bodies on postmortem examination (27, 29, 33, 34). Essential tremor brains without Lewy bodies have an array of cerebellar abnormalities that include increased numbers of torpedoes and PC loss (27, 29). The Lewy bodyyfree brains have been referred to as cerebellar ET and brains with Lewy bodies as Lewy body variant of ET in previous publications (24, 32). Demographic and clinical information were collected on each brain; the severity of tremor on Archimedes spiral drawings was rated using a validated rating scale (0Y3) by a neurologist (E.D.L.) who was blinded to postmortem data (35). Diseased control brains (n = 49) included individuals diagnosed with Alzheimer disease (AD; n = 20), Parkinson disease (PD) or diffuse Lewy body disease (DLBD; n = 15) that were selected for matching to the age distribution of available cerebellar ET brains. Alzheimer disease brains were from individuals who had been enrolled as AD cases in the Alzheimer Disease Research Center or the Washington Heights Inwood Columbia Aging Project, where they were followed prospectively with serial neurologic assessments, including assessment of action tremor, and were free of ET, PD, DLBD, or progressive supranuclear palsy (PSP). Parkinson disease or DLBD brains at the New York Brain Bank had been patients at the Movement Disorder Division, Neurological Institute Hairy Baskets Associated With Essential Tremor of New York, or were outside referrals to the brain bank. We also used the brains of 12 PSP cases and of 1 case of clinically diagnosed adult-onset focal dystonia (torticollis) even if they were younger than the cerebellar ET cases because tissue availability was limited. Nondiseased control brains (n = 21) were matched to the age distribution of cerebellar ET brains and were from individuals who had been enrolled as normal elderly control subjects in the Alzheimer s Disease Research Center or the Washington Heights Inwood Columbia Aging Project; they were followed prospectively with serial neurologic assessments and were free of the diagnosis of AD, ET, PD, DLBD, or PSP, and had normal neuropathologic examinations. Tissue Processing All brains underwent complete neuropathologic assessment at the New York Brain Bank (26, 27, 32, 33, 36). Brains received ratings of neurofibrillary tangles using Braak and Braak staging (37, 38) and Consortium to Establish a Registry for Alzheimer s Disease (CERAD) ratings (39) for neuritic plaques. National Institute on AgingYReagan Institute criteria for AD (40) were also assigned. Postmortem interval (PMI) was the number of hours between death and placement of the brain in a cold room or on ice. A standard Ymm parasagittal tissue block from the neocerebellum was harvested from each brain as described (26, 27, 29, 33). Seven-micrometer-thick paraffin sections, impregnated using a modified Bielschowsky silver technique, were used for semiquantitative and quantitative analyses of basket cell processes and quantification of torpedoes. Sections were also stained with Luxol fast bluehematoxylin and eosin (LH&E) for quantification of torpedoes and PCs (26, 27, 33). For immunohistochemical analyses, sections were incubated with mouse monoclonal SMI-31 (Covance, Princeton, NJ), as previously described (41). In each brain, torpedoes in an entire LH&E section and an entire Bielschowsky preparation were counted, and PCs in five 100 LH&E fields were counted and averaged, as previously described (27). Semiquantitative Rating of Hairy Baskets A semiquantitative rating of the appearance of the basket cell plexus surrounding PC bodies throughout Bielschowsky TABLE 1. Demographic and Clinical Characteristics of Cases and Controls Cerebellar ET LBVET AD PD/DLBD PSP Dystonia Nondiseased Controls Diseased Controls All Controls N Age (years) 83.6 T T T T T 7.2* T T 7.3* 78.3 T 9.6* Female sex 20 (62.5%) 2 (40.0%) 12 (60.0%) 6 (40.0%) 5 (41.7%) 1 (100%) 8 (38.1%) 24 (50.0%) 32 (46.4%) PMI (hours) 4.7 T T T T T T T T 7.0* Brain weight (g) 1,211 T 138 1,184 T 131 1,097 T 119 1,128 T 272 1,164 T 110 1,632 1,284 T 182 1,135 T 192 1,180 T 200 CERAD score 0.9 T T T T T T T 1.3* 1.3 T 1.3 Braak AD score 2.0 T T T T T T T 2.4* 2.4 T 0.4 Values are either mean T SD or number (percentage). Diseased controls = AD + PD/DLBD + PSP + dystonia. All controls = diseased controls and nondiseased controls. *, p G 0.01;, pg 0.05;, pg compared with cerebellar ET (analysis of variance with Tukey post hoc comparison). AD, Alzheimer disease; DLBD, diffuse Lewy body disease; ET, essential tremor; LBVET, Lewy body variant of ET; PD, Parkinson disease; PMI, postmortem interval; PSP, progressive supranuclear palsy. Ó 2010 American Association of Neuropathologists, Inc. 263

3 Erickson-Davis et al J Neuropathol Exp Neurol Volume 69, Number 3, March 2010 TABLE 2. Hairy Basket Categories by Diagnosis Cerebellar ET (n = 32) LBVET (n = 5) AD (n = 20) PD/DLBD (n = 15) PSP (n = 12) Dystonia (n = 1) Diseased Controls (n = 48) Nondiseased Controls (n = 21) All Controls (69) Hairy basket category 1 (low) 13 (40.6%) 2 (40.0%) 15 (75.0%) 10 (66.7%) 9 (75.0%) 1 (100%) 35 (72.9%) 14 (66.7%) 49 (71.0%) 2 (intermediate) 11 (34.4%) 3 (60.0%) 4 (20.0%) 4 (26.7%) 3 (25.0%) 0 (0.0%) 11 (22.9%) 6 (28.6%) 17 (24.6%) 3 (high) 8 (25.0%) 0 (0.0%) 1 (5.0%) 1 (6.7%) 0 (0.0%) 0 (0.0%) 2 (4.2%) 1 (4.8%) 3 (4.3%) P* Not applicable Hairy basket category: 1(low; hairy basket rating = 0Y1.5), 2 (intermediate; hairy basket rating = 2Y2.5), 3 (high; hairy basket rating = 3). Values are number of cases (percentage). *Compared with cerebellar ET (Mantel-Haenszel W 2 test). Only 1 case; p value not computed. AD, Alzheimer disease; DLBD, diffuse Lewy body disease; ET, essential tremor; LBVET, Lewy body variant of ET; PD, Parkinson disease; PSP, progressive supranuclear palsy. preparations was performed by a senior neuropathologist (P.L.F.) who was blinded to all clinical information. The following scale was used: 0 (few or no discernible processes); 1 (sparse number of processes); 2 (moderate number of processes); and 3 (dense tangle of processes). In some instances, the rater used intermediate values (0.5, 1.5, and 2.5). Hence, the hairy basket rating ranged included the values 0, 0.5, 1, 1.5, 2, 2.5, and 3. Quantitative Analysis of Basket Cell Plexus Quantification of the density of the basket cell axonal plexus formations was performed on a subsample of 10 brains (including 5 randomly selected cerebellar ET brains and 5 randomly selected nondiseased control brains) by a single trained physician who was blinded to clinical information. Digital images of the Bielschowsky neocerebellum preparations were obtained using a Zeiss Axioplan 2 microscope fit with an Axiocam HR digital camera (20 objective lens). For image acquisition, slides were oriented with the dentate nucleus situated on the right and the folia branches extending leftward. Folia were labeled numerically, and the 2 folia located most centrally in the slide were selected. Within each of these folia, 3 smaller subregions consisting of 1 to 2 distal peaks (gyri) and 1 to 2 distal troughs (sulci) were chosen such that there were 6 subregions photographed per brain. In each photograph, the basket cell processes surrounding each visible PC perikaryon were traced using Adobe Photoshop 5.0 (approximately 40 PCs/brain). Images of the basket cell tracings were then inverted and imported into Image J (NIH), where the integrated density was assessed (i.e. a function of the program whereby each pixel of the basket tracing was assigned a gray value using a variable gray scale and then summed). Ultrastructural Analysis Flat Eponate 12 (Ted Pella, Redding, CA) embedded vibratome sections of cerebellar cortical tissue were dissected into 3 4Ymm sections and re-embedded in the same resin. Semithin sections were then stained with toluidine blue and examined by light microscopy for PCs. Once PCs were identified on semithin section, the block was further trimmed around the PCs to the size required for ultrathin sectioning. Ultrathin sections were stained with uranyl acetate and lead citrate and then examined by transmission electron microscope (JEOL 1200EXII, Peabody, MA), as described (41). Statistical Analyses All analyses were performed in SPSS (version 16.0). As previously noted, a priori, the focus of our interest was cerebellar ET cases rather than Lewy body variant of ET cases. Demographic and clinical features of diagnostic groups (e.g. cerebellar ET vs controls) were compared using W 2 tests and analysis of variance (ANOVA) with Tukey post hoc comparisons (Table 1). The hairy basket rating included the values 0, 0.5, 1, 1.5, 2, 2.5, and 3. In nondiseased controls, the correlates of the hairy basket rating were assessed using Pearson correlation coefficients (e.g. hairy basket rating by age). To facilitate further analyses, hairy basket ratings were then collapsed into a smaller number of rating categories as follows: low (0 or 1, FIGURE 1. Semiquantitative rating of basket cell processes. (AYD) A rating scale (from 0 to 3) of the appearance of the basket cell plexus surrounding extant Purkinje cell bodies was determined in a blinded analysis of Bielschowsky preparations of neocerebellum sections. (A) Score = 0 (few, or no discernible processes); (B) Score = 1 (sparse number of processes); (C) score = 2 (moderate number of processes); (D) score = 3 (dense tangle of processes, i.e. hairy basket). 264 Ó 2010 American Association of Neuropathologists, Inc.

4 J Neuropathol Exp Neurol Volume 69, Number 3, March 2010 Hairy Baskets Associated With Essential Tremor FIGURE 2. Immunohistochemical analysis of basket cell processes. (A, B) Vibratome sections (100-Km-thick) immunostained with monoclonal antibody SMI-31 demonstrate prominence of phosphorylated neurofilaments in basket cell processes (long arrows). (A) A dense and tangled, or hairy, appearance of basket cell axonal plexus formation in a section from a cerebellar essential tremor case with a hairy basket rating of 3. (B) A control case with a rating of 0. Torpedoes (short arrows) are also seen. and also including any ratings of 0.5 and 1.5); intermediate (2 and also including any ratings of 2.5); and high (3). Diagnostic groups were compared with regard to hairy basket rating category (Mantel-Haenszel W 2 test; Table 2). For several analyses, diagnostic groups were combined (AD + PD/DLBD + PSP + dystonia = diseased controls; diseased FIGURE 3. Quantification of basket cell axonal plexus density. (AYD) Examples of tracings scored according to a semiquantitative scale are shown: (A) 0 (few, or no discernible processes); (B) 1 (sparse processes); (C) 2 (moderate numbers of processes); and (D) 3 (dense tangle of processes). (E) Basket cell plexus densities were higher in cerebellar essential tremor (ET) cases (closed circles; median value = 15.10) than in controls (open squares; median value = 6.98; p = 0.016). Ó 2010 American Association of Neuropathologists, Inc. 265

5 Erickson-Davis et al J Neuropathol Exp Neurol Volume 69, Number 3, March Ó 2010 American Association of Neuropathologists, Inc.

6 J Neuropathol Exp Neurol Volume 69, Number 3, March 2010 Hairy Baskets Associated With Essential Tremor FIGURE 5. Ultrastructural morphology of basket cellypurkinje cell (P) contacts in essential tremor cases. (AYC) Multiple, closely spaced sites of synaptic contact (arrows) between a neurofilament-rich basket fiber and a Purkinje cell soma. Post-synaptic densities are apparent in (A) and (B) (arrows). The basket fiber in (C) shows several vesicle-lined contacts with the Purkinje cell (arrows). (D) Numerous processes converging on the initial segment (axon hillock) of a Purkinje cell (P-axh); at least 1 of these is a neurofilament-rich basket cell axon (large arrow). (d ) In a higher magnification of (D), the Purkinje cell contact shows 2 distinct clusters of vesicles (arrows). Magnifications: (A, B, d ), 40,000; (C) 50,000; (D) 20,000. controls + nondiseased controls = all controls; Table 2). Because of the modest sample size, basket cell plexus density was compared across diagnostic groups using the Mann- Whitney test. In an unadjusted logistic regression model, diagnosis (cerebellar ET vs all controls) was the outcome variable. In this model, the hairy basket rating category was the independent categorical variable (high, intermediate, and low, as previously described). In adjusted regression models, covariates that were associated with either the outcome or independent variable or for which a priori evidence was considerable that the variable may be a confounder were included. The correlates of the hairy basket rating in cerebellar ET cases and nondiseased controls and in cerebellar ET cases alone were determined using Pearson correlation coefficient when both variables were normally distributed; otherwise, Spearman correlation coefficient was used. The degeneration index was defined as the number of torpedoes on Bielschowsky section divided by the mean number of PCs per 100 LH&E-stained high-power field. High values indicated large numbers of torpedoes and/or low PC counts (i.e. PC loss) (42). RESULTS ET, Nondiseased Controls, and Diseased Controls Cerebellar ET cases were similar to Lewy body variant of ET, AD, and PD/DLBLD cases and nondiseased controls with respect to age and PMI. However, when diseased controls (AD + PD/DLBD + PSP + dystonia) were combined and all controls were combined (diseased controls + nondiseased controls), cerebellar ET cases were older by an average approximately 5 years, and mean PMI was shorter (Table 1). All groups were similar with respect to gender and brain weight (Table 1). Consortium to Establish a Registry for Alzheimer s Disease scores and Braak AD scores were higher in AD cases than cerebellar ET cases (Table 1). Braak AD scores were higher in cerebellar ET cases than nondiseased controls (Table 1). Basket Cell Plexus Alterations in Cerebellar ET Bielschowsky sections of cerebellar neocortex revealed variable appearances of basket cell axon collaterals that surround the PC perikarya. Figure 1 illustrates examples of the semiquantitative scores of the basket cell plexus morphologies. BHairy baskets[ (rating = 3, the severe end of the spectrum) were defined as a dense and tangled axonal plexus (Fig. 1D). These hairy baskets generally surrounded viableappearing PCs, rather than PCs that had previously died (i.e. Bempty baskets[). In each of the 8 ET cases with hairy basket scores of 3, 20 hairy baskets were assessed (160 hairy baskets). Purkinje cell somata were visible in 141 of these (88.1%). Immunohistochemistry highlighted the neurofilamentrich basket cell axonal processes that comprise the hairy baskets and PC axonal torpedoes (Fig. 2). We also assessed whether the hairy basket rating was similar in different neocerebellar regions within the same brain. For this analysis, 1 Bielschowsky section was examined from 3 separate neocerebellar blocks in 7 brains with hairy basket ratings ranging from 0 to 3. The agreement between ratings was high (Block 1 vs Block 2, Pearson r = 0.91, p = 0.005; Block 1 vs Block 3, Pearson r = 0.95, p = 0.001; and Block 2 vs Block 3, Pearson r =0.96,pG 0.001). This indicated that the data FIGURE 4. Ultrastructure of neurofilament-rich basket cell processes in cerebellar essential tremor (ET) and control cases. (A) The dense plexus of basket cell processes (arrows) adjacent to the Purkinje cell soma (P) in a cerebellar ET case contains neurofilaments and mitochondria. (B) Only 1 to 2 layer(s) of basket cell processes (arrows) are seen adjacent to a Purkinje cell soma (P) in a control case. (C) Higher magnification of basket cell axonal plexus in (A) shows many layers of the basket cell formation (arrows). (D) A dense nest of basket cell processes (arrows) surrounds another Purkinje cell in an ET case. (E) Intertwining of basket cell processes around one another in an ET case. Some processes are longitudinal to the plane of section (long arrows); others have been cut transversely (short arrows). Magnifications: (A, B) 10,000; (C) 20,000; (D) 25,000; (E) 30,000. Ó 2010 American Association of Neuropathologists, Inc. 267

7 Erickson-Davis et al J Neuropathol Exp Neurol Volume 69, Number 3, March 2010 from a single block were likely to be broadly representative of the entire neocerebellum. The hairy basket rating was not correlated with age (Pearson r = 0.24, p = 0.31), PMI (Pearson r = j0.04, p = 0.87), brain weight (Pearson r = j0.20, p = 0.41), CERAD score (Pearson r = j0.09, p = 0.71), or Braak AD score (Pearson r = 0.05, p = 0.83) in nondiseased controls. It also did not differ by gender in nondiseased controls (1.5 T 0.9 in males vs 1.1 T 0.8 in females, p = 0.34). Hairy basket rating category was higher in cerebellar ET cases than in AD, PSP, dystonia, PD/DLBD (marginal difference), and nondiseased controls (Table 2). When control groups were combined, the differences with cerebellar ET were more marked. In 8 of 32 ET cases (25.0%), there was a hairy basket rating of 3, whereas only 4.2% of diseased controls and 4.8% of nondiseased controls had a score of 3 (Table 2). In an unadjusted logistic regression model in which diagnosis (cerebellar ET vs all controls) was the outcome variable, brains in the high (rating = 3) hairy basket rating category were 10 times more likely to be cerebellar ET than controls (odds ratio [OR] = 10.05; 95% confidence interval [CI] = 2.33Y43.32; p = 0.002). Brains in the intermediate (rating = 2) basket rating category were 2 to 3 times more likely to be cerebellar ET than controls (OR = 2.44; 95% CI = 0.92Y6.46; p = 0.07). In a logistic regression model that adjusted for age, gender, PMI, brain weight, CERAD score, and Braak AD score, these values were as follows: OR = (95% CI = 1.88Y191.56; p = 0.01) and OR = 3.52 (95% CI = 0.82Y15.02; p = 0.089). To provide a quantitative measure of these basket cell plexus alterations, the axonal plexuses surrounding extant PCs in 5 cerebellar ET cases and 5 nondiseased controls were traced from Bielschowsky preparation images (Fig. 3AYD). When the densities of the tracings were determined, the average basket cell plexus in cerebellar ET cases was approximately twice the density of the average plexus in nondiseased controls (mean = T 3.56 [median = 15.10] vs 6.98 T 2.20 [median = 6.68]; Mann-Whitney z = 2.40; p = 0.016; Fig. 3E). Basket cell plexus density was highly correlated with the hairy basket rating (Pearson r = 0.73; p = 0.018). Basket cell plexus density was also strongly associated with the number of torpedoes (Spearman r = 0.79; p = 0.007), marginally inversely with the number of PCs (Spearman r = j0.25; p = 0.49), and with the degeneration index (Spearman r = 0.76; p = 0.01). Ultrastructural Analysis of the Basket Cell Plexus We performed ultrastructural studies to compare basket cell processes in ET and control cases. In an ET case with a hairy basket rating of 3, there were several layers of neurofilament-rich basket cell processes adjacent to the PC soma that intertwined at various angles (Fig. 4A, CYE). In contrast, in a nondiseased control with a hairy basket rating of 0, only 1 or 2 basket cell processes having a somewhat thinner diameter were found adjacent to the PC soma (Fig. 4B). The identity of these neurofilament-rich processes as basket cell fibers was further confirmed in the ET case by the identification of synapses with the PC soma and its initial axon segment. In several instances, contacts with the PC somata were marked by multiple, closely spaced synaptic sites within 1 basket cell process (Fig. 5AYC; arrows). A basket fiber axon was also identified making contact with the axon hillock of the PC. In the basket axon, there were multiple synaptic sites marked by vesicle accumulations beneath the axonal membrane (Fig. 5D, d ). These findings further corroborate and extend our light microscopic findings of basket cell plexus morphology. Correlates of the Hairy Basket Rating In cerebellar ET cases and nondiseased controls, the hairy basket rating was robustly correlated with number of torpedoes (Spearman r = 0.49, p G 0.001; Fig. 6A), inversely with the number of PCs (Spearman r = j0.27; p = 0.06; Fig. 6B), and inversely with the degeneration index (Spearman r = 0.48; p = 0.001; Fig. 6C), indicating that these basket cell changes were associated with torpedo formation and PC loss. Within cerebellar ET cases, the hairy basket rating was associated with age (Pearson r = 0.35; p = 0.048) and marginally with disease duration (disease duration was known in 19 cases and Pearson r = 0.35; p = 0.19). Hairy basket rating was associated with number of torpedoes (Spearman r = 0.49; p = 0.005) but not with number of PCs (Pearson r = 0.01; p = 0.98); however, it was associated with the degeneration index (Spearman r = 0.45; p = 0.01). Hairy FIGURE 6. Correlations of hairy basket scores in cerebellar essential tremor (ET) and nondiseased control cases. (AYC) Hairy basket score correlated with number of torpedoes (A) (Spearman r = 0.49; p G 0.001), inversely with the number of Purkinje cells (B) (Spearman r = j0.27; p = 0.06), and degeneration index (C) (Spearman r = 0.48; p = 0.001). 268 Ó 2010 American Association of Neuropathologists, Inc.

8 J Neuropathol Exp Neurol Volume 69, Number 3, March 2010 basket rating did not differ by family history (presence vs absence of Q1 first-degree relative with ET), age of tremor onset (Pearson r = j0.21; p = 0.44), severity (0Y3) of tremor on Archimedes spiral (Pearson r = j0.08; p = 0.79), or presence versus absence of head tremor (data not shown). Hairy Baskets Associated With Essential Tremor DISCUSSION Recent postmortem studies have documented degenerative changes in the cerebellum of ET cases, as manifested by increased numbers of torpedoes and loss of PCs (24, 27Y29, 42). These postmortem data complement a rich history of clinical, physiologic, and imaging studies that have long implicated cerebellar and cerebellothalamacortical circuitry abnormalities in ET (8Y15, 19Y21, 43). Although previous pathologic studies focused on degenerative changes in PCs, the relationship between ET and cerebellar interneurons has remained unexplored. We examined archival and prospectively acquired cerebellar tissue of ET cases and both diseased and nondiseased controls and found that the axonal plexuses of basket cells surrounding PCs in cerebellar ET brains are significantly Bhairier[ as compared with controls. Basket cells are F-aminobutyric acidyergic inhibitory interneurons that receive input from parallel fibers and, to a limited extent, from climbing and mossy fibers. In humans, up to 50 axon collaterals from neighboring basket cells descend from the molecular layer and combine to form a complex basket structure around the PC soma to which the basket cells entire axonal output is devoted (44). Reductions in cerebellar basket cell axonal plexuses have been reported in several disease states, including hypothyroidism (44), ataxia telangiectasia (45), and Menkes kinky hair disease (46). Preservation of basket cell plexuses in the setting of PC death (i.e. Bempty baskets[) has been reported in Creutzfeldt-Jakob disease (47) and spinocerebellar ataxia Type 1 (48), but a dense and tangled appearance of the axonal plexus profile has not, to our knowledge, been documented in the ataxia literature. Curiously, an early 20th century study describing PC abnormalities in 16 brains of individuals diagnosed with Bsenile dementia[ noted increased basket cell fiber Btangles and masses[ alongside PC soma showing signs of degeneration in 3 cases (31). Another more recent study qualitatively examined basket cells in Creutzfeldt-Jakob disease and noted empty baskets with abnormally dense phosphorylated neurofilament positivity in a single case (31). Here, we documented that individuals with a diagnosis of cerebellar ET were more than 10 times as likely as controls (both diseased and nondiseased) to have a hairy basket rating of 3 (the marked or severe end of the spectrum). In addition, the median axonal plexus density in cerebellar ET was approximately 2 to 3 times higher than in controls. BHairiness[ of basket cell axonal plexuses in these independent semiquantitative and quantitative measures was positively associated with number of torpedoes and the degeneration index, and inversely with number of PCs, suggesting that hairy baskets and torpedoes may be concomitant features of cerebellar degeneration in ET. The mechanism by which this increased Bhairiness[ occurs is unknown. One possible explanation is that the increased profiles observed in our study represent an accumulation of converging basket cell processes recruited from neighboring PCs that have been damaged or died. Although there has been little investigation of such a phenomenon in the human cerebellum, selective preservation and reorganization of basket cell axonal processes has been demonstrated in basket cells in the CA1 and CA3 region of the hippocampus (49, 50). These cells form baskets around hippocampal pyramidal cells and function as local circuit inhibitory F-aminobutyric acidyergic interneurons, analogous to the relationship between cerebellar basket cells and PCs. These disease-resistant hippocampal basket cells undergo extensive reorganization in the setting of pyramidal cell death (51). In a few cases, basket cell plexuses displayed increased density adjacent to areas with severe pyramidal cell loss and no basket formations, suggesting that these basket cell processes might be converging on and reorganizing around remaining pyramidal cells (51). The molecular mechanisms that underlie the degenerative changes observed in the ET cerebellum are not clear. The array of cellular changes within the complex neuronal network in cerebellar cortex have yet to be fully catalogued in ET; thus, further work with human postmortem tissue is needed. In addition, genes have not been identified for ET, and no transgenic mice currently exist that provide an animal model for ET. The most widely used animal model for ET is based on the administration of neurotoxins (e.g. harmaline and ibogaine) to rodents and other mammals. These toxins induce synchronous firing of inferior olivary neurons at an 8- to 12-Hz frequency and sustained glutamatergic stimulation of PCs via their climbing fiber afferents (52Y54), triggering a massive increase in intracellular calcium and an excitotoxicmediated PC degeneration and cell death (55Y57). However, this toxin-based animal model produces an acute, reversible action tremor. Most animals shake for a matter of hours, and there is more severe destruction of cerebellar cortex in readily identifiable linear bands. Although these latter clinical and pathologic features clearly differ from those in ET, this model demonstrates the functional significance of the olivocerebellar system and a potential role for excitotoxic-mediated PC death in the generation of this type of tremor. It is intriguing that harmane (a tremor-producing, indole-alkaloid structurally similar to harmaline) has been found in the blood of some ET patients at levels up to 50% higher than age-matched controls (58). Although additional studies are needed to establish whether this exposure is of etiologic importance in ET, we postulate that the PC loss documented in ET (whether excitotoxic or not) leads to a reorganization of the cerebellar interneuron network, with sprouting and/or accumulation of neighboring basket cell processes on surviving PC bodies. Ultrastructural analysis revealed the presence of multiple, closely spaced sites of synaptic contact between the dense, multilayered basket cell processes and PC soma in ET. The neurophysiologic abnormality in ET is not clear, but is thought to be the result of decreased cerebellar PC inhibitory output. Our observations raise the possibility that the increased density of basket cell processes identified in a significant number of ET cases contribute to this inhibitory effect on PC output. Recent studies have demonstrated that climbing fiber responses in Ó 2010 American Association of Neuropathologists, Inc. 269

9 Erickson-Davis et al J Neuropathol Exp Neurol Volume 69, Number 3, March 2010 PCs are modulated through cerebellar interneuron networks that include the basket cell (59). Basket cell output is also inhibited by the axon collaterals of PCs and excited by both parallel fibers and collaterals of climbing fibers (60). Cerebellar interneurons are also coupled by dendro-somatic, dendrodendritic, and somato-somatic gap junctions, which may lead to synchronization of firing in groups of organized neural networks (61). Specialized axo-axonic septate junctions electrically couple basket fiber processes in the axon hillock region (62). Thus, the disease process in ET may cause an imbalance in the interactions between cerebellar afferents, PCs and their axon collaterals, and the cerebellar interneuron network that includes basket cells. Additional studies on other cerebellar neurons and their interactions in ET are clearly needed to understand the basis for tremor generation in this disorder. Despite being the largest series of ET brains reported to date, the sample size was modest. Despite this limitation, we were able to detect significant case-control differences, indicating that the sample was sufficient. This study had several strengths. This was the first investigation of a relatively novel pathologic finding in a large series of ET brains. We examined basket cell profiles both semiquantitatively and quantitatively and, furthermore, compared ET brains against both nondiseased control brains and a group of diseased controls with a variety of neurodegenerative pathologies. In summary, the degenerative changes in PCs that characterize cerebellar ET are accompanied, in some cases, by a dense and tangled, or Bhairy,[ appearance of basket cell axonal plexus formations. This relatively novel finding was not observed to any degree in either diseased controls or normal control brains. This finding provides initial evidence that structural changes in ET are not restricted to the PC but also involve their functional network. The presence of hairy baskets is a heretofore uninvestigated phenomenon, and further exploration is warranted to understand its relationship to PC degeneration and its role in the pathogenesis of ET. ACKNOWLEDGMENT The authors thank Hong Yi (Emory University) for assistance with preparation of tissue for ultrastructural studies. REFERENCES 1. Benito-Leon J, Louis ED. Essential tremor: Emerging views of a common disorder. Nat Clin Pract Neurol 2006;2:666Y78; quiz 662p following Louis ED. Essential tremor. Lancet Neurol 2005;4:100Y10 3. Dogu O, Sevim S, Camdeviren H, et al. Prevalence of essential tremor: Door-to-door neurologic exams in Mersin Province, Turkey. Neurology 2003;61:1804Y6 4. Louis ED, Thawani SP, Andrews HF. Prevalence of essential tremor in a multiethnic, community-based study in northern Manhattan, New York, NY. Neuroepidemiology 2009;32:208Y14 5. Critchley M. Observations on essential (heredofamilial) tremor. Brain 1949;72:113Y39 6. Busenbark KL, Nash J, Nash S, et al. Is essential tremor benign? Neurology 1991;41:1982Y83 7. Louis ED, Benito-Leon J, Ottman R, et al. A population-based study of mortality in essential tremor. Neurology 2007;69:1982Y89 8. Louis ED, Frucht SJ, Rios E. Intention tremor in essential tremor: Prevalence and association with disease duration. Mov Disord 2009;24:626Y27 9. Singer C, Sanchez-Ramos J, Weiner WJ. Gait abnormality in essential tremor. Mov Disord 1994;9:193Y Stolze H, Petersen G, Raethjen J, et al. The gait disorder of advanced essential tremor. Brain 2001;124:2278Y Parisi SL, Heroux ME, Culham EG, et al. Functional mobility and postural control in essential tremor. Arch Phys Med Rehabil 2006;87: 1357Y Koster B, Deuschl G, Lauk M, et al. Essential tremor and cerebellar dysfunction: Abnormal ballistic movements. J Neurol Neurosurg Psychiatry 2002;73:400Y5 13. Helmchen C, Hagenow A, Miesner J, et al. Eye movement abnormalities in essential tremor may indicate cerebellar dysfunction. Brain 2003;126: 1319Y Dupuis MJ, Delwaide PJ, Boucquey D, et al. Homolateral disappearance of essential tremor after cerebellar stroke. Mov Disord 1989;4:183Y Schuurman PR, Bosch DA, Bossuyt PM, et al. A comparison of continuous thalamic stimulation and thalamotomy for suppression of severe tremor. N Engl J Med 2000;342:461Y Benabid AL, Pollak P, Seigneuret E, et al. Chronic VIM thalamic stimulation in Parkinson s disease, essential tremor and extra-pyramidal dyskinesias. Acta Neurochir Suppl (Wien) 1993;58:39Y Bucher SF, Seelos KC, Dodel RC, et al. Activation mapping in essential tremor with functional magnetic resonance imaging. Ann Neurol 1997; 41:32Y Colebatch JG, Findley LJ, Frackowiak RS, et al. Preliminary report: Activation of the cerebellum in essential tremor. Lancet 1990;336: 1028Y Jenkins IH, Bain PG, Colebatch JG, et al. A positron emission tomography study of essential tremor: evidence for overactivity of cerebellar connections. Ann Neurol 1993;34:82Y Louis ED, Shungu DC, Chan S, et al. Metabolic abnormality in the cerebellum in patients with essential tremor: A proton magnetic resonance spectroscopic imaging study. Neurosci Lett 2002;333:17Y Pagan FL, Butman JA, Dambrosia JM, et al. Evaluation of essential tremor with multi-voxel magnetic resonance spectroscopy. Neurology 2003;60:1344Y Shin DH, Han BS, Kim HS, et al. Diffusion tensor imaging in patients with essential tremor. AJNR Am J Neuroradiol 2008;29:151Y Cerasa A, Messina D, Nicoletti G, et al. Cerebellar atrophy in essential tremor using an automated segmentation method. AJNR Am J Neuroradiol 2009;30:1240Y Louis ED, Vonsattel JP. The emerging neuropathology of essential tremor. Mov Disord 2007;23:174Y Zesiewicz TA, Elble R, Louis ED, et al. Practice parameter: therapies for essential tremor: Report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology 2005;64:2008Y Louis ED, Vonsattel JP, Honig LS, et al. Essential tremor associated with pathologic changes in the cerebellum. Arch Neurol 2006;63: 1189Y Louis ED, Faust PL, Vonsattel JP, et al. Neuropathological changes in essential tremor: 33 cases compared with 21 controls. Brain 2007;130: 3297Y Shill HA, Adler CH, Sabbagh MN, et al. Pathologic findings in prospectively ascertained essential tremor subjects. Neurology 2008;70: 1452Y Axelrad JE, Louis ED, Honig LS, et al. Reduced Purkinje cell number in essential tremor: A postmortem study. Arch Neurol 2008;65:101Y7 30. Yokoo H, Isoda K, Yamanouchi H, et al. Cerebellar basket cells of Creutzfeldt-Jakob disease: Immunohistochemical and ultrastructural study. Pathol Int 2000;50:291Y Uyematsu S. A study of some peculiar changes found in the oxons and dendrites of the Purkinje cells. Arch Neurol Psych 1921;5:231Y Louis ED, Borden S, Moskowitz CB. Essential tremor centralized brain repository: Diagnostic validity and clinical characteristics of a highly selected group of essential tremor cases. Mov Disord 2005;20:1361Y Louis ED, Vonsattel JP, Honig LS, et al. Neuropathologic findings in essential tremor. Neurology 2006;66:1756Y Louis ED, Faust PL, Vonsattel JP, et al. Torpedoes in Parkinson s disease, Alzheimer s disease, essential tremor, and control brains. Mov Disord 2009;24:1600Y5 35. Louis ED, Ottman R, Ford B, et al. The Washington HeightsYInwood Genetic Study of Essential Tremor: Methodologic issues in essentialtremor research. Neuroepidemiology 1997;16:124Y33 Ó 2010 American Association of Neuropathologists, Inc.

10 J Neuropathol Exp Neurol Volume 69, Number 3, March Louis ED, Honig LS, Vonsattel JP, et al. Essential tremor associated with focal nonnigral Lewy bodies: A clinicopathologic study. Arch Neurol 2005;62:1004Y7 37. Braak H, Braak E. Diagnostic criteria for neuropathologic assessment of Alzheimer s disease. Neurobiol Aging 1997;18:S85Y8 38. Braak H, Alafuzoff I, Arzberger T, et al. Staging of Alzheimer diseasey associated neurofibrillary pathology using paraffin sections and immunocytochemistry. Acta Neuropathol 2006;112:389Y Mirra SS. The CERAD neuropathology protocol and consensus recommendations for the postmortem diagnosis of Alzheimer s disease: A commentary. Neurobiol Aging 1997;18:S91Y4 40. The National Institute on Aging and Reagan Institute Working Group on Diagnostic Criteria for the Neuropathological Assessment of Alzheimer_s Disease 11: Consensus Recommendations for the Postmortem Diagnosis of Alzheimer_s disease. Neurobiol Aging 1997;18:S1Y2 41. Louis ED, Yi H, Erickson-Davis C, et al. Structural study of Purkinje cell axonal torpedoes in essential tremor. Neurosci Lett 2009;450:287Y Louis ED, Vonsattel JPG, Honig LS, et al. Older onset essential tremor: More rapid progression and more degenerative pathology. Mov Disord 2009;24:1606Y Deuschl G, Wenzelburger R, Loffler K, et al. Essential tremor and cerebellar dysfunction clinical and kinematic analysis of intention tremor. Brain 2000;123:1568Y Leclerc N, Gravel C, Plioplys A, et al. Basket cell development in the normal and hypothyroid rat cerebellar cortex studied with a monoclonal anti-neurofilament antibody. Can J Biochem Cell Biol 1985;63:564Y Perry TL, Kish SJ, Hinton D, et al. Neurochemical abnormalities in a patient with ataxia-telangiectasia. Neurology 1984;34:187Y Iwata M, Hirano A, French JH. Degeneration of the cerebellar system in X-chromosomeYlinked copper malabsorption. Ann Neurol 1979;5: 542Y Tiller-Borcich JK, Urich H. Abnormal arborizations of Purkinje cell dendrites in Creutzfeldt-Jakob disease: A manifestation of neuronal plasticity? J Neurol Neurosurg Psychiatry 1986;49:581Y Robitaille Y, Schut L, Kish SJ. Structural and immunocytochemical features of olivopontocerebellar atrophy caused by the spinocerebellar ataxia type 1 (SCA-1) mutation define a unique phenotype. Acta Neuropathol 1995;90:572Y81 Hairy Baskets Associated With Essential Tremor 49. de Lanerolle NC, Kim JH, Robbins RJ, et al. Hippocampal interneuron loss and plasticity in human temporal lobe epilepsy. Brain Res 1989;495: 387Y Buckmaster PS, Jongen-Relo AL. Highly specific neuron loss preserves lateral inhibitory circuits in the dentate gyrus of kainite-induced epileptic rats. J Neurosci 1999;19:9519Y Arellano JI, Munoz A, Ballesteros-Yanez I, et al. Histopathology and reorganization of chandelier cells in the human epileptic sclerotic hippocampus. Brain 2004;127:45Y De Montigny C, Lamarre Y. Rhythmic activity induced by harmaline in the olivocerebello-bulbar system of the cat. Brain Res 1973;53:81Y Llinás R, Volkind RA. The olivo-cerebellar system: Functional properties as revealed by bulbar system of the cat. Brain Res 1973;63: 430Y Llinás R, Mühlethaler M. Electrophysiology of guinea-pig cerebellar nuclear cells in the in vitro brain stem-cerebellar preparation. J Physiol 1988;404:241Y Trump BF, Berezesky IK. Calcium-mediated cell injury and cell death. FASEB J 1995;9:219Y Slemmer JE, De Zeeuw CI, Weber JT. Don t get too excited: Mechanisms of glutamate-mediated Purkinje cell death. Prog Brain Res 2005; 148:367Y Weber JT, De Zeeuw CI, Linden DJ, et al. Long-term depression of climbing fiber-evoked calcium transients in Purkinje cell dendrites. Proc Natl Acad Sci U S A 2003;100:2878Y Louis ED, Jiang W, Pellegrino KM, et al. Elevated blood harmane (1-methyl-9H-pyrido[3,4-b]indole) concentrations in essential tremor. Neurotoxicology 2008;29:294Y Barmack NH, Yakhnitsa V. Functions of interneurons in mouse cerebellum. J Neurosci 2008;28:1140Y O Donoghue DL, King JS, Bishop GA. Physiological and anatomical studies of the interactions between Purkinje cells and basket cells in the cat s cerebellar cortex: Evidence for a unitary relationship. J Neurosci 1989;9:2141Y Sotelo C, Llinás R. Specialized membrane junctions between neurons in the vertebrate cerebellar cortex. J Cell Biol 1972;53:271Y Gobel S. Axo-axonic septate junctions in the basket formations of the cat cerebellar cortex. J Cell Biol 1971;51:328Y33 Ó 2010 American Association of Neuropathologists, Inc. 271