Apolipoprotein E affects the central nervous system response to injury and the development of cerebral edema.код для вставкиСкачать
BRIEF COMMUNICATIONS Apolipoprotein E Affects the Central Nervous System Response to Injury and the Development of Cerebral Edema John R. Lynch, MD,1,2 Jose A. Pineda, MD,1,3 Duncan Morgan, BS,2 Lin Zhang, MD,4 David S. Warner, MD,1,4 Helen Benveniste, MD, PhD,4 and Daniel T. Laskowitz, MD1,2 Apolipoprotein E has been implicated in modifying neurological outcome after traumatic brain injury, although the mechanisms by which this occurs remain poorly defined. To investigate the role of endogenous apolipoprotein E following acute brain injury, noninvasive magnetic resonance imaging was performed on anesthetized mice following closed head injury. Effacement of the lateral ventricle was used as a radiographic surrogate for cerebral edema. At 24 hours following injury, apolipoprotein E-deficient animals had a greater degree of cerebral edema as compared to matched controls. In addition, the brains of apolipoprotein E-deficient animals had a significantly greater upregulation of tissue necrosis factor ␣ messenger ribonucleic acid as compared to controls as early as 1-hr post injury. Thus, modulation of the endogenous central nervous system inflammatory response may be one mechanism by which apolipoprotein E affects outcome following acute brain injury. Ann Neurol 2002;51:113–117 There are three common human isoforms of apolipoprotein E (apoE), designated apoE2, apoE3, and apoE4.1 Numerous clinical reports have demonstrated that apoE modifies neurological recovery following closed head injury in an isoform-specific manner.2–5 To model the effect of endogenous apoE on acute brain trauma, C57BL/6 mice and apoE-deficient mice matched for age, sex, and genetic background were subjected to closed head injury using a stereotactically guided pneumatic compression device to exert an acute acceleration/deceleration injury. This injury causes ce- From the 1Multidisciplinary Neuroprotection Laboratory, and Departments of 2Medicine (Neurology), 3Pediatrics, and 4Anesthesiology, Duke University Medical Center, Durham, NC. Received Apr 10, 2001, and in revised form Aug 15, 2001. Accepted for publication Sep 5, 2001. Published online Dec 3, 2001 Address correspondence to Dr Laskowitz, Department of Medicine (Neurology), Box 2900, Duke University Medical Center, Durham, NC 27710. E-mail: email@example.com rebral edema, which can increase intracranial hypertension and lead to herniation of brain tissue. The latter represents a significant source of neurological morbidity in the clinical setting following closed head injury. In order to obtain additional information regarding the effect of cerebral edema in producing anatomic distortions in situ, we utilized magnetic resonance imaging (MRI) as a noninvasive assessment measurement of edema. Astrocytic and microglial activation with the resultant secretion of inflammatory mediators is believed to promote breakdown of the blood brain barrier and subsequent development of cerebral edema. In particular, TNF␣ has been suggested to play an important role in this regard.7 This is of particular interest given recent in vitro and in vivo data suggesting that one role of apoE in the injured central nervous system may be to downregulate glial activation and the endogenous inflammatory response.8 –12,15 We measured the differential upregulation of TNF␣ messenger RNA (mRNA) after closed head injury in apoE-deficient mice compared to wild-type controls. Materials and Methods Animals All experiments were performed with the approval of the Duke University Animal Care and Use Committee. Briefly, 8 –16week-old male C57-BL/6 mice and apoE-deficient mice matched for gender and age that had been previously back bred for 10 generations to C57BL/6 background were obtained commercially from Jackson Laboratories (Bar Harbor, ME). Head Injury Model with Controlled Pneumatic Impact Device Mice were anesthetized with 4.3% isoflurane in oxygen at a FiO2 of 50% in an anesthesia induction box for 90 seconds. The trachea was intubated and the lungs were mechanically ventilated and anesthetized with 1.4% isoflurane in 50% O2 and 50% N2. Body temperature was maintained at 37°C using surface heating/cooling. Each mouse was positioned in a stereotactic device. The top of the skull was exposed to identify anatomical landmarks. Animals were subjected to controlled skull impact with a pneumatic impactor (Air-Power, High Point, NC) using a 2.0mm steel tip impounder at a controlled velocity (6.0 ⫾ 0.2m/sec) and vertical displacement (3.0mm). The animals were allowed to recover spontaneous ventilation prior to extubation. Following recovery, mice were allowed free access to food and water. Assessment of Functional Effects of apoE Following Head Injury Neurological severity score: Motor function was assessed at 1 and 24 hours following injury using a modified neuroseverity score that included beam walking on 3cm, 2cm, and 1cm beams, ability to exit from circle (30cm in diameter), presence of seeking behavior, hemiplegia, ability to walk straight, and presence of a startle reflex. This generated a score from 0 (normal) to 10 (moribund). This scale is weighted toward © 2001 Wiley-Liss, Inc. DOI 10.1002/ana.10098 113 motor function, and minor modifications of this scale have been described previously.6 Magnetic Resonance Imaging Anesthetized, spontaneously breathing mice were scanned using a 7.1, 15cm bore superconducting Oxford magnet with shielded gradient coils controlled by a Signa console (General Electric Medical Systems, Milwaukee, WI). We used a threedimensional diffusion-weighted spin-echo pulse sequence with the following acquisition parameters: (a) 2.4-hour scan time; (b) 20ms echo time; (c) 600ms repetition time; (d) diffusion gradients of 35 Gauss/cm applied in the slice direction; and (e) b value of 965mm2/sec. All images were acquired at a spatial resolution of 58m ⫻ 58m ⫻ 469m ⫽ 1.6 ⫻ 10⫺3mm3. The raw data were reconstructed by Fourier transform and displayed as magnitude images. Ventricular volumes were measured on the threedimensional diffusion data sets by manually outlining regions of interest in each hemisphere using National Institutes of Health Image 1.68. Wet-to-Dry Method of Assessing Brain Water Content Brains were removed from mice 24 hours after injury, dissected along the midline saggital plane, and immediately weighed. As previously described,17 to determine water content, the brains were then placed in a desiccating oven at 105°C for 48 hours and then reweighed. Measurement of Cytokine mRNA At 1 and 24 hours following injury, cohort animals were perfused with normal saline and decapitated. The left and right hemispheres were separated and placed in a 3cc volume of Trizol. Prompt homogenization was performed by rapidly passing the tissue through an 18g and 22g needle respectively until virtually no tissue debris was visible (mRNA was quantified by RNAse protection assay [RPA]). RNA isolation was performed using Trizol reagent manufactured by Gibco-BRL (Carlesbad, CA) and their recommended protocol was followed. RPA results were resolved using a 5% acrylamide gel, which was subsequently dried and placed on a phosphorimaging screen for overnight exposure. Numerical data were collected and analyzed using the Molecular Dynamics Storm威 phosphorimaging system and the ImageQuant威 (Sunnyvale, CA) software package provided. To normalize for any potential differences in loading conditions, mRNA was expressed as ratio of gene of interest to the housekeeping gene GAPDH. Statistical Analysis All results were analyzed using Student’s two-tailed t test. Values are expressed as mean ⫾ standard error of the mean. Results Prior to injury, there were no significant differences between C57BL/6 (n ⫽ 7) and apoE-deficient (n ⫽ 7) mice with respect to weight (32 ⫾ 1gm vs 31 ⫾ 1gm; p ⫽ 0.3) or functional neurological status as deter- 114 Annals of Neurology Vol 51 No 1 January 2002 mined by a blinded observer on a 10-point neurological scale.6 There was no asymmetry between left and right ventricular volumes in the two noninjured control groups. However, at 24 hours following injury, we found that both apoE-deficient and C57BLJ/6 mice had significant effacement of the lateral ventricle ipsilateral to the closed head injury side compared to the right, noninjured hemisphere (Fig 1). The apoE-deficient mice had an average decrease in left lateral ventricular volume of 30%, as compared to 15% in C57BL6/J mice ( p ⬍ 0.01). Figure 2 shows diffusion images at the level of dorsal hippocampus from 3 apoE-deficient and 3 C57BLJ/6 mice. All apoE-deficient mice show effacement of the left lateral ventricle ipsilateral to the injury site and a subtle midline shift. In contrast, the only one of the C57BLJ/6 mice shows clear ventricle effacement (see Fig 2, top). At 6-week follow-up, there was no longer any evidence of ventricular asymmetry in either group. Ventricular volumes were symmetric and similar to controls, suggesting a complete resolution of cerebral edema. To demonstrate that radiographic evidence of ventricular effacement is a valid surrogate for cerebral edema, we directly measured water content in the injured and control hemispheres. Using the standard wet-to-dry method, we found significantly increased water content in the injured hemisphere of the apoEdeficient animals as compared to the wild-type controls (80.12 ⫾ 1.22mg vs 78.35 ⫾ 1.02mg in apoEdeficient and wild-type controls, respectively; n ⫽ 12, p ⫽ 0.038. Data expressed as mean ⫾ standard error of the mean). However, there were no significant differences in the control nonlesioned hemisphere of apoE-deficient and wild-type animals. To determine whether animals received a uniform injury, a blinded observer using a 10-point neurological scale assessed neurological function after controlled head injury, in which 0 represents a normal exam, and increasing scores represent progressive motor deficit.6 At 1 hour following injury, there were no significant differences in neurological function between groups (median ⫽ 5 ⫾ 1 vs 5 ⫾ 2 in wild-type versus apoEdeficient animals, respectively; p ⫽ 0.5). Animals were reassessed at 24 hours and again there was no difference in motor function between groups (3 ⫾ 1 vs 3 ⫾ 1 in wild-type vs apoE-deficient animals, respectively; p ⫽ 0.9), although neurological function had improved from baseline in both groups. To determine whether TNF␣ was downregulated as a function of endogenous apoE in our model of controlled head injury, cohort animals were sacrificed and perfusion fixed at different time points following injury. Messenger RNA was extracted from brain homogenates, and TNF␣ mRNA levels were quantified in the injured hemisphere by RNAse protection assay Fig 1. Eight representative diffusion magnetic resonance images from a three-dimensional data set obtained 24 hours after closed head injury in a C57BLJ/6 mouse. The injury site is clearly seen as a high signal intensity area in the most superior part of the cortex (slices 1 and 2). More ventrally the effacement of the left lateral ventricle is clearly seen. We have demonstrated how measuring left and right lateral ventricle volume indirectly assessed the amount of edema. The apoE-deficient mice had an average decrease in left lateral ventricular volume of 30%, as compared to 15% in C57BL6/J mice (p ⬍ 0.01). Absolute ventricular volumes were 2.83 ⫾ 0.10mm3 at baseline. After injury the volumes were 2.38 ⫾ 0.17mm3 in the wild-type mice and 2.09 ⫾ 0.27mm3 in the apoE-deficient mice. Seven animals were used per group. Data expressed as mean ⫾ standard error of the mean. Fig 2. Diffusion images at the level of the dorsal hippocampus acquired 24 hours after closed head injury from 3 apoE-deficient and 3 C57BLJ/6 mice are shown. Left lateral effacement and a subtle midline shift (as indirectly indicated by the red line) can be appreciated in all 3 apoE-deficient mice. In contrast, only 1 (no. 9) of the 3 C57BLJ/6 mice shown demonstrates such changes. (RPA). We found that as early as 1 hour following injury, apoE-deficient mice had significantly greater TNF␣ mRNA levels compared to control mice (ratio TNF␣/GAPDH 0.0018 vs 0.0034; p ⫽ 0.006; Fig 3). Discussion Our observations suggest an increase in hemispheric edema in apoE-deficient animals as compared to wildtype controls at 24 hours following closed head injury. This is consistent with a prior report demonstrating that histological damage and long-term behavioral sequelae were more severe in apoE-deficient mice after head in- 116 Annals of Neurology Vol 51 No 1 January 2002 jury.16 The two groups of mice in this study received the same initial insult using the pneumatic compression device, but the apoE-deficient mice suffered greater injuries compared to C57BLJ/6 as evaluated by MRI. One potential mechanism by which this may occur is that apoE downregulates glial activation and subsequent secretion of inflammatory cytokines such as TNF␣. This function of apoE is consistent with prior in vitro and in vivo data.8 –12,15 We did observe a more robust increase in the levels of TNF␣ mRNA in the brains of apoEdeficient animals, which more than likely reflects an elevation of TNF␣ in the brain after closed head injury. following acute brain injury, as well as designing novel therapeutic strategies in this clinical setting. This work was supported by the National Institutes of Health (1K08NS01949, T32GM08600, R01NS37235), a Novartis Pilot Grant, and two Paul Beeson Physician Faculty Awards to D. Laskowitz and H. Benveniste. Imaging was performed at Center for in vivo Microscopy supported by funds from the National Institutes of Health Research Grants NCRR P41 RR05959 (Principle Investigator Dr. G. Allan Johnson). References Fig 3. Baseline brain TNF␣ messenger RNA levels were not significantly different in apoE-deficient vs wild-type mice. At 1 hour following closed head injury, TNF␣ messenger RNA had greater expression in the brains of apoE-deficient mice as compared to matched wild-type controls (**p ⬍ 0.01). To control for loading conditions, quantification of genes of interest were represented as a ratio to the housekeeping gene GAPDH. Six animals were used per time point. Data represented as mean ⫾ standard error of the mean. Although the role of TNF␣ in the injured central nervous system has not been fully elucidated, it is upregulated by neurons and astrocytes following human traumatic brain injury and appears to play an integral role in the breakdown of the blood brain barrier and subsequent development of cerebral edema in humans.13 In addition, inhibition of TNF␣ in a rodent model of closed head injury resulted in decrease of blood brain barrier breakdown and the subsequent development of cerebral edema.14 In this study we demonstrate that apoE-deficient mice exhibit enhancement of TNF␣ mRNA synthesis following closed head injury relative to wild-type controls. These results are consistent with prior observations that apoE suppresses microglial and astrocytic activation in vitro,8,9,11,12,15 and the inflammatory cytokines, including TNF␣, are more robustly upregulated after injury in apoE-deficient mice in vivo.10 This excess production of TNF␣ might promote more severe blood brain barrier breakdown and thus more extensive edema. Our results suggest that the presence of endogenous apoE might be protective in closed head injury by modulating glial activation, TNF␣ production and thus decrease the amount of cerebral edema in the injured brain. The results of this study may serve as a framework for understanding clinically relevant isoform-specific differences in neurological outcome 1. Weisgraber KH. Apolipoprotein E: structure-function relationships. Adv Protein Chem 1994;45:249 –302. 2. Sorbi S, Nacmias N, Piacentini S, et al. ApoE as a prognostic factor for post-traumatic coma. Nature Med 1995;1:852. 3. Teasdale, GM, Nicoll, JA, Murray, G, et al. Association of apolipoprotein E polymorphism with outcome after head injury. Lancet 1997;350:1069 –1071. 4. Friedman, G, Froom, P, Sazbon, L, et al. Apolipoprotein E-epsilon 4 genotype predicts a poor outcome in survivors of traumatic brain injury. Neurology 1999;52:244 –248. 5. Jordan, BD, Relkin NR, Raydin LD, et al. Apolipoprotein E epsilon-4 associated with chronic traumatic brain injury in boxing. JAMA 1997;278:136 –140. 6. Chen Y, Constantini S, Trembovler V, et al. An experimental model of closed head injury in mice: pathophysiology, histopathology, and cognitive deficits. J Neurotrauma 1996;13:557–568. 7. Ramilo O, Saez-Llorens X, Mertsola J, et al. Tumor necrosis factor alpha/cachectin and interleukin 1 beta initiate meningeal inflammation. J Exp Med 1990;172:497–507. 8. Laskowitz D, Goel S, Bennett ER, et al. Apolipoprotein E suppresses glial cell secretion of TNF␣. J Neuroimmunology 1997; 76:70 –74. 9. Laskowitz DT, Thekdi A., Thekdi S, et al. Downregulation of glial activation by apolipoprotein E and apoE-mimetic peptides. Exp Neurol 2001;167:74 – 85. 10. Lynch JR, Morgan D, Mance J, et al. Apolipoprotein E modulates glial activation and the endogenous central nervous system response. J Neuroimmunology 2001;114:107–113. 11. Barger SW, Harmon AD. Microglial activation by Alzheimer amyloid precursor protein and modulation by apolipoprotein E Nature 1997;388:878 – 881. 12. Hu J, Ladu MJ, Van Eldik LJ. Apolipoprotein E attenuates beta-amyloid-induced astrocyte activation. J Neurochem 1998; 71:1626 –1634. 13. Ott L, McClain CJ, Gillespie M, Toung B. Cytokines and metabolic dysfunction after severe head injury. J Neurotrauma 1994;11:447– 472. 14. Shohami E, Bass R, Wallach D, et al. Inhibition of tumor necrosis factor alpha (TNFalpha) activity in rat brain is associated with cerebroprotection after closed head injury. J Cereb Blood Flow Metab 1996;16:378 –384. 15. Laskowitz DT, Matthew WD, Bennett ER, et al. Endogenous apolipoprotein E suppresses LPS-stimulated microglial nitricoxide production. Neuroreport 1998;9:615– 618. 16. Chen Y, Lomnitski L, Michaelson DM, Shohami E. Motor and cognitive deficits in apolipoprotein-E deficient mice after closed head injury. Neuroscience 1997;80:1255–1262. 17. Pineda JA, Aono M, Sheng H, et al. Extracellular superoxide dismutase overexpression improves behavioral outcome from closed head injury in the mouse. J Neurotrauma 2001;18: 625– 634. Lynch et al: Apolipoprotein E Affects the Central Nervous System 117 Novel Heteroplasmic mtDNA Mutation in a Family with Heterogeneous Clinical Presentations P. Corona, MSc,1 E. Lamantea, MSc,1 M. Greco, PhD,1 F. Carrara, BSc,1 A. Agostino, MSc,1 D. Guidetti, MD,2 M. T. Dotti, MD,3 C. Mariotti, MD,1 and M. Zeviani, MD, PhD1 The protean manifestations of a novel maternally inherited point mutation of the mitochondrial genome are reported. The proband showed isolated, spastic paraparesis. A brother, who had suffered from a multisystem progressive disorder, ultimately died of cardiomyopathy. Another brother is healthy. The proband’s mother showed truncal ataxia, dysarthria, severe hearing loss, mental regression, ptosis, ophthalmoparesis, distal cyclones, and diabetes mellitus. A muscle biopsy performed in the proband failed to show the morphological abnormalities typical of mitochondrial disorders; the activities of respiratory chain complexes were normal. However, complex I and IV activities were low in the muscle homogenate of the affected mother and brother. Sequence analysis of mtDNA showed a heteroplasmic mutation of the tRNAIle gene (G4284A). The mutation load was approximately 55%, 80%, and 90% in the muscle mtDNA of the proband, his mother, and his affected brother, respectively. Mutation was undetected in the healthy brother, as well as in 100 control samples. Several cybrid clones containing homoplasmic mutant mtDNA from the proband showed significant reductions of complex IV activity and maximum oxygen consumption rate, compared with homoplasmic wild-type clones derived from the same subject. Ann Neurol 2002;51:118 –122 Mutations of mitochondrial DNA (mtDNA) are responsible for a wide spectrum of syndromes, often characterized by a combination of symptoms that affect muscle, brain, and peripheral nerves, and occasionally the heart.1,2 In several cases, syndromes are rather well From the 1Division of Biochemistry and Genetics, National Neurological Institute C. Besta, Milan; 2Division of Neurology, Public Health Hospital Santa Maria Nuova, Reggio Emilia; 3Neurometabolic Unit, Institute of Neurological Sciences, University of Siena, Siena; Italy. Received Jul 24, 2001, and in revised form Sep 7, 2001. Accepted for publication Sep 10, 2001. Published online Dec 28, 2001 Address correspondence to Dr Zeviani, Divisione di Biochimica e Genetica, Istituto Nazionale Neurologico Carlo Besta, via Celoria 11, 20133 Milano, Italy. E-mail: firstname.lastname@example.org. 118 © 2001 Wiley-Liss, Inc. DOI 10.1002/ana.10059 defined. In other cases, overlap presentations or the unusual combination of symptoms, slow progression, and qualitative and quantitative variations in different members of the same family can complicate the diagnosis.2,3 Additional difficulty can be attributable to the absence or scarcity of diagnostic clues that are considered typical of mitochondrial encephalomyopathies, such as morphological abnormalities in the muscle biopsy.4 The family reported exemplifies these diagnostic problems. Case Report The family tree is reported in Figure 1. The proband, subject III-3, is a 35-year-old man with an 8-year history of progressive spastic paraparesis. He was reported to have bilateral genu valgum and what was characterized as a “clumsy walk” since childhood. A brain magnetic resonance image (MRI) performed when he was 31 years old disclosed the presence of a poroencephalic cavitation in the medial part of the left frontal lobe, associated with moderate atrophy of the anterior segment of the corpus callosum. The lesion was attributed to an old, possibly perinatal, infarction in the territory of the left callosomarginal artery. Nevertheless, the patient’s psychomotor development was reported to be normal. He obtained a master’s degree in life sciences and, in spite of walking problems, practiced several sports during adolescence. The clinical examination, carried out when he was 35 years of age, disclosed the presence of a “pure” spastic paraparesis with bilateral ankle clonus and extensor plantar reflexes. No other neurological or systemic abnormalities were found at physical examination. An electrocardiogram (EKG) and echocardiogram were both normal. Electromyography showed a myopathic pattern; the latencies of visual and Fig 1. Family tree. The proband is indicated by an arrow. Black symbols indicate the three affected subjects reported in this study. Asterisk indicates subjects reported to be affected by a neurologic disorder which could not be investigated further (see text). The presence of hearing loss or diabetes mellitus in other members of the pedigree is also indicated. Table. Respiratory Chain Activities in Muscle Homogenate NADH:CoQ1 reductase/CS (complex I) Succinate dehydrogenase/CS (SDH) Succinate:CoQ1 reductase/CS (complex II) DBH2:cyt. c reductase/CS (complex III) Cytochrome c oxidase/CS (complex IV) ATPase/CS (complex V) Citrate synthase (CS)a a Pt II-3 Pt III-2 Pt III-3 Control Range 9.5 16.6 27.7 91 65 97 148 10.9 15.7 23.3 124 5.5 82 104 21.4 17.8 34.3 188 202 165 120 17–33 10–20 18–35 70–150 80–180 100–200 80–210 Expressed as nanomoles/min/mg. CoQ1 ⫽ coenzyme Q1; DBH2 ⫽ decyl-ubiquinol; cyt. c ⫽ cytochrome c. brainstem-evoked potentials were moderately delayed. Blood lactate and pyruvate levels were both normal at rest and after standard exercise. A muscle biopsy was morphologically normal, as were the activities of the mitochondrial respiratory chain (RC) complexes in both muscle homogenate (Table) and cultured fibroblasts (not shown). An older brother of the proband, subject III-2, had died at 27 years of age of heart failure as a result of rapidly progressive dilating cardiomyopathy. This patient had suffered from petit mal epilepsy during infancy and had a single episode of generalized tonic–clonic seizures when he was 17; myoclonic jerks of the distal segments of upper limbs, eyelids, and facial muscles developed from age 16 on. Additional neurological symptoms ensued during the subsequent years, including severe truncal ataxia, moderate dysarthria and dysmetria, bilateral sensorineural hearing loss, markedly slowed saccades, and change in mood and personality, which evolved into overt mental deterioration. Bilateral macular degeneration was noted when he was 23 years old, and an episode of transitory hemiparesis on the right side was reported at age 26 years. Computed tomography (CT) scan and MRI of the brain displayed marked cortical and subcortical atrophy. The electroencephalogram showed the presence of diffuse slow waves with no paroxysms. Visual and brainstem-evoked potentials were reduced with severely delayed latencies; nerve conduction velocity was also moderately reduced. Additional symptoms included diabetes mellitus type 2 since age 18 years, the presence of a single lipoma on the back of the neck, and a hypogonadal appearance. Endocrine tests suggested a hypogonadotropic hypogonadism attributable to hypothalamic dysfunction. At age 25, rapidly progressive heart failure developed due to hypertrophic-dilating cardiomyopathy, leading to death at age 27 years. An EKG performed at age 25 showed the presence of a complete left bundle branch block and ischemic abnormalities. The echocardiogram showed a reduced ejection fraction with hypomotility of a dilated and hypertrophic left ventricle. Blood lactate at rest was 3.5 mg% (normal values ⬍2 mg%). A muscle biopsy failed to show ragged-red fibers. Histochemical analysis was not performed. However, biochemical assays, carried out years later in our laboratory, disclosed a dramatic decrease in complex IV (COX) activity and a less profound decrease of complex I and complex V activities (Table). The mother of the proband, subject II-3, is a 64year-old woman with a 10-year history of profound bilateral hearing loss, type 2 diabetes mellitus, and a complex neurological syndrome, including truncal ataxia, mild dysarthria, myoclonic jerks of the hands and forearms, proximal muscle weakness, severe ophthalmoparesis, and mild mental deterioration. Blood lactate at rest was normal (0.9 mg%). Both the EKG and the echocardiogram showed no abnormalities. MRI of the brain disclosed cortical and subcortical diffuse atrophy. A muscle biopsy showed that the homogenate disclosed a combined reduction of the activities of complex I and complex IV (Table). A younger brother of the proband (subject III-4) is a 33-year-old, apparently healthy, man. Other members of the maternal lineage of this family were reported to be affected. In particular, subject I-2, the proband’s maternal grandmother, had adult diabetes mellitus and profound hearing loss. The latter was the major complaint of the older maternal aunt of the proband (II-1), whose only daughter (III-1) was reported to be affected by muscular dystrophy. A maternal uncle (III-5) had died at 8 years as a result of “encephalopathy.” Methods Morphological and Biochemical Analyses Morphological examination of skeletal muscle and biochemical assays of the individual RC and citrate synthase (CS) on muscle homogenate and digitonin-treated fibroblasts were carried out as described.5 The maximum oxygen consumption rate was measured in intact cells by polarography6 in the presence of 25 M dinitrophenol (DNP). Analysis of mtDNA Southern blot analysis of linearized mtDNA, single-strand conformation polymorphism analysis of the 22 mitochon- Corona et al: Novel Mutation in tRNAIle mtDNA Gene 119 Fig 2. (A) Sequence analysis of muscle mtDNA from patient II-3. Asterisk indicates the mutant A at position 4284; note the smaller underneath peak corresponding to the wild-type G. (B) Proposed tRNA Ile cloverleaf-like secondary structure. Arrow indicates the 4284G. (C) Tsp509I–restriction fragment length polymorphism (RFLP) analysis of muscle mtDNA of a control muscle (C) and of affected subjects (III-3, II-3, III-2), as well as lymphocyte mtDNA of the proband’s healthy brother (III-4). (D) Biochemical activities in cybrid clones. Scattergrams of COX-to-CS activity ratios (left panel) and maximum oxygen consumption rate (right panel) in 143B.206-derived cybrid clones containing 100% mutant (m) and 100% wild-type (wt) mtDNA, both derived from the proband. Horizontal bars indicate the mean values. drial tRNA genes, and automated sequence analysis were carried out as described.7,8 A 150-bp fragment encompassing nucleotides (nt) 4230 – 4380 of mtDNA9 was polymerase chain reaction (PCR)-amplified, using suitable primers. The 4284G3 A transition creates a novel Tsp509I restriction site (AATT). A second Tsp509I site is present at nt4369. Thus, Tsp509I cuts the wild-type (wt) fragment into two fragments of 140 and 10 bp, while the mutant fragment is cut into three fragments of 86, 54, and 10 bp. The fragments were separated by 8% acrylamide–tris-borate-EDTA gel electrophoresis and visualized under ultraviolet (UV) light by ethidium bromide staining. The proportion of mutant versus total mtDNA was calculated by densitometry. Fibroblast and Cybrid Cell Cultures Transmitochondrial cybrids were obtained by polyethylene glycol fusion of 143B.206-derived ° cells with probandderived enucleated fibroblasts, followed by selection in a uridine-free medium.6,10 Results Respiratory Chain Complexes in Muscle Homogenate The Table reports the values of the specific activities of the RC complexes normalized to the specific activity of CS, measured in muscle homogenates. No abnormality was detected in the proband’s sample, while combined reduction of different mtDNA-dependent RC complexes was present in the muscle homogenate of his brother and mother. Identification and Quantitative Analysis of a 4284G3 A Transition The biochemical results suggested the presence of an mtDNA mutation, likely involving a tRNA gene. Southern blot analysis of muscle mtDNA excluded the presence of large-scale rearrangements. SSCP and nu- 120 Annals of Neurology Vol 51 No 1 January 2002 cleotide sequence analysis failed to show mutations in all the 22 mtDNA tRNA genes with the exception of a G3 A heteroplasmic transition at nucleotide position 4284 (Fig 2A), in the tRNAIle gene (see Fig 2B). Tsp509I–RFLP analysis (see Fig 2C) showed that the relative mutant load was approximately 50%, 80% and 90% in the muscle samples of the proband, his mother and affected brother, respectively. No mutation was detected in the blood lymphocyte DNA and urinary epithelial cells of the healthy brother of the proband, as well as in 100 blood DNA samples from healthy, unrelated Italian subjects. The 4284G3 A mutation load was also measured in the proband’s hair follicles (⬃50%), urinary mucosal cells (⬃40%), skin fibroblasts (⬃30%), buccal smear (⬃20%), and blood lymphocytes (⬃20%), and was ⬃20% in the urinary cells and ⬃30% in blood lymphocytes of the mother. Biochemical Assays in Cybrids A highly significant mean reduction of COX-to-CS activity ratio was demonstrated in a group of 10 homoplasmic mutant clones (20 ⫾ 7 SD) compared with 10 homoplasmic wt clones (33.8 ⫾ 7.6, unpaired twotailed Student’s t test p ⫽ 0.0005) (see Fig 2D, left panel). Likewise, the maximum oxygen consumption rate was significantly reduced in a group of eight 100% mutant clones (3.74 fmol O2/cell ⫾ 1.22 SD) compared with eight 100% wt clones (6.29 fmol O2/cell ⫾ 1.89 SD, Student’s t test p ⫽ 0.0062) (see Fig. 2D, right panel). Discussion The 4284G3 A is the tenth mutation in the mtDNA tRNAIle gene to be associated with disease (see www. gen.emory.edu/mitomap.html). Four mutations have been reported in patients affected by chronic progressive external ophthalmoplegia,11–14 while five additional mutations were found in individuals affected by cardiomyopathy of variable severity.15–19 Although the pathogenetic mechanisms of some of these mutations have not been characterized completely, the tRNAIle gene can well be considered a major mutational hotspot in mtDNA disorders. The 4284G3 A transition affects the boundary between the DHU and the anticodon stems of the putative tRNAIle cloverleaf, a region of uncertain functional significance, and involves a poorly conserved nucleotide. Thus, its functional consequences cannot be deduced by an obvious deleterious effect on the structure or function of tRNAIle or by evolutionistic considerations. However, several lines of evidence support the pathogenicity of the 4284G3 A transition: 1. It was heteroplasmic and, as often observed in pathogenic mutations of mtDNA, the mutant load was higher in a postmitotic tissue (ie, muscle), than in rapid turnover tissues. 2. The degree of heteroplasmy in muscle was concordant with the clinical severity, which was relatively low in the proband, who was affected by isolated spastic paraparesis, higher in his mother, affected by a late-onset multisystem neurological disorder, and even higher in the proband’s brother, who suffered from juvenile-onset rapidly progressive encephalocardiomyopathy. 3. The mutation segregated with the disease, which was undetectable in the blood lymphocytes of a cohort of 100 control subjects, and in the blood and urinary epithelium of a normal family member (subject III-4), although the presence of the mutation cannot be excluded in other tissues of this subject. 4. The mutation load was correlated with impairment of respiratory chain activities in muscle homogenates. In addition, statistically significant differences in the activities of COX and maximum oxygen consumption rate were obtained in two populations of cybrid clones carrying homoplasmic mutant and homoplasmic wt mitochondrial genomes, respectively. As frequently observed in mitochondrial disorders, the threshold at which the mutant load leads to an overt biochemical impairment appears to be higher than the threshold at which clinical symptoms are produced, as the RC activities were within the normal range in the muscle sample from a clinically affected patient, ie, the proband. In addition, some of the 100% mutant cybrid clones displayed biochemical activities within the normal range, suggesting a relatively mild deleterious effect of the mutation on the RC biochemistry, as measured by our standard assays. 5. The complex clinical and laboratory features in the mother and affected brother of the proband were indeed suggestive of a mitochondrial cause. The onset of rapidly progressive heart failure in subject III-2 confirms the frequent association between mutations of tRNAIle gene and cardiomyopathy. Patient III-3, the proband, was affected by isolated spastic paraparesis. In this patient, morphological examination of the muscle biopsy was completely uninformative. The possibility of a mitochondrial cause was considered only for the presence of a suggestive family history. A causative link between the poroencephalic lesion shown by MRI and the mtDNA mutation found in this patient remains an interesting, but unproved, hypothesis. Although pyramidal signs, including spastic paraparesis, have been reported in mtDNA-related disease, they are usually accompanied by other neurological and extraneurological symptoms. We cannot exclude that additional clinical features will present in the future in our patient, but it is tempting to speculate that his monosymptomatic, relatively benign, clinical course can be correlated with the lower mutation load found in this subject, compared with the mutation load detected in his more severely affected relatives. This work was supported by Fondazione Telethon-Italy (1180) (to M.Z.), by Ricerca Finalizzata Ministero Sanitá. (ICS 030.3/ RF98.37), and by Fondazione Pierfranco e Luisa Mariani. We are indebted to Ms B Geehan for revising the manuscript. We thank Dr Thomas Klopstock for critical discussion and Dr Marina Mora for technical advice. References 1. Hirano M, Davidson M, DiMauro S. Mitochondria and the heart. Curr Opin Cardiol 2001;16:201–210. 2. Smeitink J, van den Heuvel L, DiMauro S. The genetics and pathology of oxidative phosphorylation. Nature Rev Genet 2001;2:342–352. 3. DiMauro S, Andreu AL. Mutations in mtDNA: are we scraping the bottom of the barrel? Brain Pathol 2000;10:431– 441. 4. Zeviani M, Tiranti V, Piantadosi C. Mitochondrial disorders. Medicine 1998;77:59 –72. 5. Tiranti V, Munaro M, Sandonà D, et al. Nuclear DNA origin of cytochrome c oxidase deficiency in Leigh’s syndrome: genetic evidence based on patient’s derived rho° transformants. Hum Mol Genet 1995;4:2017–2023. 6. Mariotti C, Tiranti V, Carrara F, et al. Defective respiratory capacity and mitochondrial protein synthesis in transformant cybrids harboring the tRNA leu (UUR) mutation associated with maternally inherited myopathy and cardiomyopathy. J Clin Invest 1994;93:1102–1107. 7. Zeviani M, Moraes CT, DiMauro S, et al. Deletions of mitochondrial DNA in Kearns-Sayre syndrome. Neurology 1988;38: 1339 –1346. Corona et al: Novel Mutation in tRNAIle mtDNA Gene 121 8. Tiranti V, Carrara F, Confalonieri P, et al. A novel mutation (8342G3 A) in the mitochondrial tRNA(Lys) gene associated with progressive external ophthalmoplegia and myoclonus. Neuromusc Disord 1999;9:66 –71. 9. Anderson S, Bankier AT, Barrell BG, et al. Sequence and organization of the human mitochondrial genome. Nature 1981; 290:457– 465. 10. Tiranti V, Munaro M, Sandonà D, et al. Nuclear DNA origin of cytochrome c oxidase deficiency in Leigh’s syndrome: genetic evidence based on patient’s derived rho0 transformants. Hum Mol Genet 1995;4:2017–2023. 11. Chinnery PF, Johnson MA, Taylor RW, et al. A novel mitochondrial tRNA isoleucine gene mutation causing chronic progressive external ophthalmoplegia. Neurology 1997;49:1166–1168. 12. Silvestri G, Servidei S, Rana M, et al. A novel mitochondrial DNA point mutation in the tRNA(Ile) gene is associated with progressive external ophtalmoplegia. Biochem Biophys Res Commun 1996;220:623– 627. 13. Taylor RW, Chinnery PF, Bates MJ, et al. A novel mitochondrial DNA point mutation in the tRNA(Ile) gene: studies in a patient presenting with chronic progressive external ophthalmoplegia and multiple sclerosis. Biochem Biophys Res Commun 1998;243:47–51. 14. Franceschina L, Salani S, Bordoni A, et al. A novel mitochondrial tRNA(Ile) point mutation in chronic progressive external ophthalmoplegia. J Neurol 1998;245:755–758. 15. Taniike M, Fukushima H, Yanagihara I, et al. Mitochondrial tRNA Ile mutation in fatal cardiomyopathy. Biochem Biophys Res Commun 1992;186:47–53. 16. Merante F, Myint T, Tein I, et al. An additional mitochondrial tRNA(Ile) point mutation (A-to-G at nucleotide 4295) causing hypertrophic cardiomyopathy. Hum Mutat 1996;8:216 –222. 17. Casali C, Santorelli FM, D’Amati G, et al. A novel mtDNA point mutation in maternally inherited cardiomyopathy. Biochem Biophys Res Commun 1995;213:588 –593. 18. Tanaka M, Ino H, Ohno K, et al. Mitochondrial mutation in fatal infantile cardiomyopathy. Lancet 1990;336:1452. 19. Santorelli FM, Mak SC, Vazquez-Acevedo M, et al. A novel mitochondrial DNA point mutation associated with mitochondrial encephalocardiomyopathy. Biochem Biophys Res Commun 1995;216:835– 840. Increase in Hand Muscle Strength of Stroke Patients after Somatosensory Stimulation Adriana B. Conforto, MD, Alain Kaelin-Lang, MD, and Leonardo G. Cohen, MD It has been proposed that somatosensory input in the form of peripheral nerve stimulation can influence functional measures of motor performance. We studied the effects of median nerve stimulation on pinch muscle strength (a function mediated predominantly by median nerve innervated muscles) in the affected hand of chronic stroke patients. A 2-hour period of median nerve stimulation elicited an increase in pinch strength that outlasted the stimulation period. The improvement in muscle strength correlated with stimulus intensity and was identified in the absence of motor training. These results suggest that somatosensory stimulation may be a promising adjuvant to rehabilitation of the motor deficits in stroke patients. Ann Neurol 2002;51:122–125 Somatosensory input can modulate reorganization in sensorimotor cortical and subcortical structures.1– 4 It has been suggested that somatosensory input in the form of peripheral nerve stimulation (PNS) could contribute to improved motor function after stroke.5 Here, we studied the effects of median nerve stimulation (MNS) on pinch strength in the affected hand of chronic stroke patients in a randomized, crossover design. Patients and Methods Eight patients (7 men, 1 woman; mean age 65 years; range 38 – 81 years) with hemiparesis caused by ischemic stroke participated in this study. The protocol was approved by the NINDS Investigational Review Board, and all subjects gave written informed consent. Lesions identified on computerized tomography or magnetic resonance imaging were located in the left internal capsule (3), left pons (1), right in- From the Human Cortical Physiology Section, NINDS, National Institutes of Health, Bethesda, MD. Received Jun 29, 2001, and in revised form Sep 13. Accepted for publication Sep 14, 2001. Published online Dec 3, 2001 Address correspondence to Dr Cohen and Dr Kaelin-Lang, Human Cortical Physiology Section, National Institutes of Health, Building 10, Room 5N23, 10 Center Drive, MSC 1430, Bethesda, MD 20892-1430. E-mail: email@example.com 122 © 2001 Wiley-Liss, Inc. DOI 10.1002/ana.10070 Fig 1. Representation of hand areas where patients reported paresthesias during the median nerve stimulation session. ternal capsule and basal ganglia (2), right frontoparietal cortex, basal ganglia and internal capsule (1), and right pons (1). The average time after stroke was 5 years and 6 months (range 14 months to 7 years). Previously, we performed a pilot study and found that a 2-hour period of median nerve stimulation resulted in an average increase of 2.55 ⫾ 0.9 Newtons (N) (mean ⫾ standard error [SE]) in pinch strength in 5 chronic stroke patients (unpublished observations). Based on these preliminary results, we planned the present study in a randomized, prospective, crossover design. Each patient participated in two different sessions separated by at least 24 hours: 2-hour MNS and 2-hour control stimulation (CS). Patients were comfortably seated and were instructed to remain at rest. Background EMG activity recorded from surface electrodes was continuously monitored during the experiments. In both sessions, silver-silver surface chloride electrodes (diameter 10mm) were optimally placed to stimulate the median nerve at the wrist in the affected arm. Trains of electrical stimulation were delivered at 1Hz (Grass stimulator S8800 with stimulus isolation unit [SIU], Grass Instrument Division, Astro-Med Inc., West Warwick, RI). Each train consisted of five single pulses of 1ms duration delivered at 10Hz.5 These stimulus parameters are thought to preferentially activate large cutaneous and proprioceptive sensory fibers.6 In the MNS session, stimulus intensity was gradually increased until the patients reported strong paresthesias in the median nerve territory in the absence of pain (Fig 1). Average stimulus intensity in the MNS session was 47 ⫾ 8.3% above the minimal stimulus intensity required to elicit paresthesias (mean ⫾ SE; range 22– 83%). This intensity of stimulation usually elicited compound muscle action potentials smaller than 100V from the abductor pollicis brevis muscle in the absence of visible finger movements. In the CS session, stimulus intensity was kept immediately below that required to elicit paresthesias. None of the patients reported paresthesias in the absence of electrical stimulation. In the neurological exam, Patient 1 had slightly decreased joint position sense, and Patient 8 had tactile hypoesthesia in the affected hand. Joint position sense, vibration, tactile, and pain sensation were normal in the affected hand of the other 6 patients. Maximal key pinch strength7 of the affected hand was measured according to a protocol that exhibits good validity and test-retest reliability7,8 (Jamar dynamometer, Sammons Preston, Inc., Bolingbrook, IL). Patients held the arm of the dynamometer between the lateral aspect of the middle phalanx of the index finger and the thumb pad and were instructed to squeeze as hard as they could. Muscle strength measurements of five consecutive trials were averaged. One investigator recorded and analyzed muscle strength measurements blind to the intervention type while another investigator administered the interventions. The order of the sessions was randomized. In the first session, 5 patients received CS and 3 patients received MNS. Two-tailed, paired t tests were used to compare muscle strength measurements before and after interventions (data normally distributed according to Shapiro-Wilk test, significance level set to 0.05). Patient 8 was excluded from this main comparison because of poor compliance during the CS session. Linear regression analysis was used to investigate the relation between changes in muscle strength and stimulus intensity in the MNS sessions (n ⫽ 8). Results Muscle strength measurements recorded before the CS and MNS sessions were comparable: 54.40 ⫾ 10.63N Fig 2. Absolute changes in muscle strength (Newtons) in the control stimulation session and the median nerve stimulation session (n ⫽ 7). Error bars represent standard errors of the mean. Conforto et al: Somatosensory Stimulation of Stroke Patients 123 Fig 3. Relation between stimulus intensity (intensity of median nerve stimulation [MNS] relative to the threshold to elicit paresthesias in each individual) and changes in pinch muscle strength (pinch muscle strength after median nerve stimulation relative to prestimulation levels) in the MNS session (n ⫽ 8). (mean ⫾ SE; range 22.56 –103.59N) and 49.72 ⫾ 8.09N (mean ⫾ SE; range 24.53–91.43N), respectively ( p ⫽ 0.22), indicating similar baseline conditions in both sessions. Stimulation resulted in an increase in pinch strength of 2.41 ⫾ 0.74N ( p ⫽ 0.017) in the MNS session and in a nonsignificant decrease of 1.07 ⫾ 2.4N ( p ⫽ 0.67) in the CS session (Fig 2). In an “intent to treat analysis,” in the control intervention there were no significant changes in strength with a paired t test ( p ⫽ 0.62), whereas in the median nerve stimulation session there was a significant increase of 2.57N ( p ⫽ 0.006). Following the MNS session, 2 patients spontaneously reported that they could “write better” and “hold objects and play cards more accurately”; this perception lasted for approximately 24 hours. No patients reported any changes after the CS session. The increased pinch muscle strength identified in the MNS session correlated well with the intensity of MNS (r ⫽ 0.729, p ⫽ 0.04, Fig 3). Discussion The main result of this study is that a 2-hour period of median nerve stimulation increased pinch muscle strength in the affected hand of stroke patients. This effect was clearly identifiable in 6 of 8 individuals and outlasted the stimulation period. The magnitude of the improvement in pinch strength was similar in this randomized crossover study and in the open-label pilot experiments that we had previously performed. Additionally, increase in muscle strength was proportional to the intensity of sensory stimulation, suggesting a causal relationship. These results are consistent with emerging 124 Annals of Neurology Vol 51 No 1 January 2002 evidence that somatosensory input can modify motor function.9,10 Prior experiments demonstrated motor improvement in the hand of stroke patients after a combination of motor training and manipulation of sensory input (Muellbacher and colleagues, unpublished observations) or electrical muscular stimulation over a period of several weeks or months.11,12 Hand motor deficits play an important role in stroke disability.13 In some patients, rehabilitative strategies that are based on motor practice can be difficult or impossible to implement because of severe muscle weakness. Our findings in a small sample of stroke patients indicate that a single session of predominantly sensory nerve stimulation can improve pinch muscle strength in the absence of practice. This type of intervention may be a promising adjuvant to enhance neurorehabilitative strategies when hand weakness makes motor training difficult or impossible. Supported by the National Institutes of Neurological Diseases and Stroke Intramural Program. We thank Mark Hallett and Carolyn Wu for helpful discussion and Devera G. Schoenberg for skillful editing. References 1. Asanuma H. Functional role of sensory inputs to the motor cortex. Prog Neurobiol 1981;16:241–262. 2. Kaas JH. Plasticity of sensory and motor maps in adult mammals. Annu Rev Neurosci 1991;14:137–167. 3. Recanzone GH, Allard TT, Jenkins WM, Merzenich MM. Re ceptive-field changes induced by peripheral nerve stimulation in SI of adult cats. J Neurophysiol 1990;63:1213–1225. 4. Fox K, Glazewski S, Schulze S. Plasticity and stability of somatosensory maps in thalamus and cortex. Curr Opin Neurobiol 2000;10:494 – 497. 5. Ridding MC, Brouwer B, Miles TS, et al. Changes in muscle responses to stimulation of the motor cortex induced by peripheral nerve stimulation in human subjects. Exp Brain Res 2000; 131:135–143. 6. Panizza M, Nilsson J, Roth BJ, et al. Relevance of stimulus duration for activation of motor and sensory fibers: implications for the study of H-reflexes and magnetic stimulation. Electroencephalogr Clin Neurophysiol 1992;85:22–29. 7. Mathiowetz V, Kashman N, Volland G, et al. Grip and pinch strength: normative data for adults. Arch Phys Med Rehabil 1985;66:69 –74. 8. Mathiowetz V, Weber K, Volland G, Kashman N. Reliability and validity of grip and pinch strength evaluations. J Hand Surg [Am] 1984;9:222–226. 9. Fraser C, Power M, Hobday D, et al. Driving plasticity in adult human motor cortex improves functional performance after cerebral injury [abstract]. In: Proceedings of the 15th International Congress of Clinical Neurophysiology, May 2001; Buenos Aires, Argentina. 10. Struppler A, Jakob C, Muller-Barna P, et al. Eine neue Methode zur Fruhrehabilitation zentralbedingter Lahmungen von Arm und Hand mittels Magnetstimulation. Z EEG EMG 1996;27:151–157. 11. Cauraugh J, Light K, Kim S, et al. Chronic motor dysfunction after stroke: recovering wrist and finger extension by electromyography-triggered neuromuscular stimulation. Stroke 2000;31:1360 –1364. 12. Dimitrijevic MM, Stokic DS, Wawro AW, Wun CC. Modification of motor control of wrist extension by mesh-glove electrical afferent stimulation in stroke patients. Arch Phys Med Rehabil 1996;77:252–258. 13. Whitall J, McCombe Waller S, Silver KH, Macko RF. Repetitive bilateral arm training with rhythmic auditory cueing improves motor function in chronic hemiparetic stroke. Stroke 2000;31:2390 –2395. Selective Hippocampal Neuron Loss in Dementia with Lewy Bodies Antony J. Harding, PhD, Bronwyn Lakay, BSc(Hons), and Glenda M. Halliday, PhD Hippocampal volume and neuron number were measured using stereological techniques in pathologically confirmed dementia with Lewy bodies (n ⴝ 8), Parkinson’s disease only (n ⴝ 4), and controls (n ⴝ 9). We, and others, have previously shown considerable cell loss in the CA1 and subiculum subregions in Alzheimer’s disease. In contrast, these regions were spared in dementia with Lewy bodies where a selective loss of lower presubiculum pyramidal neurons was found. These findings suggest a selective loss of frontally projecting hippocampal neurons in dementia with Lewy bodies versus those projecting to temporal lobe regions in Alzheimer’s disease. Ann Neurol 2002;51:125–128 The hippocampus is a complex structure with multiple anatomical and functional compartments. Degeneration of the CA1 and subiculum subregions underlies the dementia found in Alzheimer’s disease (AD),1–3 with gross atrophy of the hippocampus a marker of this disease.4 – 8 At end stage, similar hippocampal atrophy is found in dementia with Lewy bodies (DLB),9 but the CA1 and subiculum are spared.2,10 The hippocampal atrophy in DLB correlates with atrophy and Lewy body formation in the frontal lobes, as well as with the severity of Lewy neurite formation in the CA2/3 subregions of the hippocampus.9 This suggests that the subregions of the hippocampus affected in DLB differ significantly from those affected in AD. The present study quantifies the density and number of neurons in the different hippocampal subdivisions (CA1, CA2, CA3, CA4, dentate gyrus, subiculum, and presubiculum) comparing cases with DLB to age- and sex-matched controls, nondemented cases of Parkinson’s disease (PD) only, and previous data published on cases with AD.1–3 From the Prince of Wales Medical Research Institute and University of New South Wales, Sydney, Australia. Received Jun 8, 2001, and in revised form Sep 20. Accepted for publication Sep 20, 2001. Published online Dec 3, 2001 Address correspondence to Dr Harding, Prince of Wales Medical Research Institute, Barker Street, Randwick, Sydney, NSW 2031, Australia. E-mail: A.Harding@unsw.edu.au © 2001 Wiley-Liss, Inc. DOI 10.1002/ana.10071 125 Table. Demographics, Quantitative Data and Results of Statistical Analyses Details Number Age PMD (hr) Volume (mm3) Hippocampus Grey matter White matter CA1 CA2–3 CA4 Subiculum Presubiculum Dentate gyrus Neuron number (⫻106) CA1 CA2–3 CA4 Subiculum Presubiculum Cloudsa Pyramidala Dentate gyrus Hippocampus Controls PD only DLB ANOVA p values 9 73 ⫾ 3 14 ⫾ 3 4 76 ⫾ 3 18 ⫾ 8 8 76 ⫾ 2 23 ⫾ 7 4,967 ⫾ 273 1,772 ⫾ 101 3,195 ⫾ 238 649 ⫾ 56 123 ⫾ 13 168 ⫾ 15 465 ⫾ 55 310 ⫾ 16 56 ⫾ 4 4,050 ⫾ 313 1,768 ⫾ 231 2,282 ⫾ 336 522 ⫾ 30 135 ⫾ 14 198 ⫾ 22 548 ⫾ 51 276 ⫾ 30 58 ⫾ 3 3,796 ⴞ 259 1,603 ⫾ 70 2,193 ⴞ 282 499 ⫾ 30 132 ⫾ 14 171 ⫾ 10 516 ⫾ 41 234 ⴞ 10 51 ⫾ 5 0.014 0.32 0.027 0.071 0.82 0.42 0.56 0.0077 0.52 6.1 ⫾ 0.5 2.2 ⫾ 0.2 1.1 ⫾ 0.1 4.7 ⫾ 0.3 10.7 ⫾ 0.7 100 ⫾ 11 100 ⫾ 7 13.2 ⫾ 1.1 37.8 ⫾ 1.0 5.9 ⫾ 0.6 2.4 ⫾ 0.5 1.3 ⫾ 0.2 4.9 ⫾ 0.7 9.8 ⫾ 1.5 100 ⫾ 8 88 ⫾ 19 15.8 ⫾ 1.1 37.1 ⫾ 1.5 6.1 ⫾ 0.3 2.1 ⫾ 0.2 1.1 ⫾ 0.1 4.3 ⫾ 0.6 6.8 ⴞ 0.5 100 ⫾ 11 50 ⴞ 6 13.9 ⫾ 1.4 34.2 ⫾ 1.8 0.96 0.77 0.69 0.74 0.0031 0.61 0.0008 0.44 0.18 0.63 0.48 Values in bold are significantly different from control values, p ⬍ 0.05. a Values given as a proportion of control values ⫾ standard error of the mean. Materials and Methods Cases were selected using standardized clinicopathological procedures from our regional brain donor program (approved by Human Ethics Committees) as previously detailed.9 Cases were excluded if they had significant neuropathology (eg, cerebral infarction, head injury, or hepatic encephalopathy) other than those related to Parkinson’s disease (PD) or AD. Four cases without dementia had PD only, 8 cases had DLB,11,12 and 9 age- and sex-matched controls were selected. Neocortical neurofibrillary tangles were absent in all cases and all PD only and 2 DLB cases had no neuritic plaque. Three DLB cases reached Consortium to Establish a Registry for Alzheimer’s Disease (CERAD) criteria for probable AD and the remaining three reached CERAD criteria for definite AD. There were no significant differences in age or postmortem delay between the groups (Table). Brains were fixed in 15% buffered formalin for 2 weeks, the anteroposterior dimensions of each hemisphere recorded, and the cerebrum embedded in agar, prior to being cut into 3mm thick coronal slices. All tissue blocks containing the right hippocampus were dissected from the coronal brain slices, cryoprotected and 50m-thick sections cut from the more posterior end of each hippocampal block using a freezing microtome. Sections were mounted onto glass slides, stained with cresyl violet, and coverslipped using DPX mountant. All tissue analyses were performed blinded to classification. Quantitation was performed using well-established published methods.3,9,13 Briefly, the boundaries of the hippocampal subdivisions were delineated (dentate gyrus, CA1, 126 Annals of Neurology Vol 51 No 1 January 2002 CA2-3, CA4, subiculum, and presubiculum) according to the criteria of Duvernoy14 and Amaral and Insausti.15 The cross-sectional areas of the hippocampal subdivisions in each section were determined by tracing their boundaries using a microfiche reader at 19 to 45⫻ magnification, and applying point counting techniques. The volume of each hippocampal subdivision was determined by multiplying the sum of the cross-sectional areas by the distance between the sections using Cavalieri’s principle. The optical disector technique, using the full section thickness (50m) as the disector height, was used to estimate neuronal number for each hippocampal region, as previously described in detail.3,13 Between cases, the number of sections quantified varied from 9 to 12, and the number of disector frames per subregion varied from 14 to 136. The number of nucleolated neurons within the inclusion boundaries of the frame varied between 93 and 276 for the dentate gyrus; 97 and 627 for the CA1; 84 and 293 for the CA2-3; 102 and 243 for the CA4; 87 and 287 for the subiculum; and 93 and 218 for the presubiculum. Repeated measures of the neuronal number in multiple sections from multiple cases always gave similar results, even between different investigators. Neuron density (coefficient of error range 0.035– 0.054) was determined from the total number of neurons within the sample volume. Regional neuronal number was estimated by multiplying the density and volume. Statistical analysis was performed on Statview 5.0 program (Abacus Concepts, Berkeley, CA). Group differences were analyzed using analysis of variance (ANOVA) with post-hoc Fisher’s least significant difference test. Means ⫾ standard Fig. (A) Schematic diagram demonstrating the hippocampus subdivisions: CA1, CA2–3, CA4, subiculum (SUB), presubiculum (PS), and dentate gyrus (thick black line around CA4). WM ⫽ white matter. (B–D) Photomicrographs of 50m-thick frozen cresyl violet-stained sections of the lower pyramidal layers of presubiculum in (B) a control case, (C) a nondemented case with PD only, and (D) a case with dementia with Lewy bodies (DLB). Note the obvious decrease in density of the lower pyramidal neurons in the D compared with compared to B and C. Scale in D is equivalent for B and C. errors are given for all variables and a p value less than 0.05 accepted as statistically significant. Results Gross hippocampal atrophy was confined to the white matter (see Table; ANOVA ⫽ 4.4, p ⫽ 0.027) rather than the gray matter (ANOVA ⫽ 1.2, p ⫽ 0.32). However, one hippocampal subregion, the presubiculum, was selectively atrophic in DLB compared with controls and cases with PD only (ANOVA ⫽ 6.5, p ⫽ 0.008). Cellular analysis revealed that the total estimated number of presubicular neurons was significantly decreased in DLB (ANOVA ⫽ 8.1, p ⫽ 0.003). A separate analysis of the upper clouds of small pyramidal neurons versus the lower larger pyramidal cell layer of the presubiculum, revealed no significant differences in the number of neurons contained within the upper clouds. There was, however, a significant 50% decrease in the number of neurons in the lower pyramidal layer of the presubiculum in DLB (Fig; ANOVA ⫽ 11, p ⫽ 0.0008). Neuron loss in the presubiculum was similar between cases, regardless of the presence or absence of neuritic plaque. There was no other significant neuropathology within the presubiculum, nor was there any significant loss of neurons in any of the other hippocampal subdivisions in any group (see Table). Discussion In vivo atrophy of the hippocampus in DLB is less than that observed for AD,16,17 possibly because of the selective subregional cell loss observed in the present study. In DLB, neuronal loss was confined to the pre- subiculum and Lewy neurites concentrate in the CA2-3 subregion. The presubiculum has not been previously evaluated in DLB and, together with the CA2-3 subregion, makes up about 25% of hippocampal gray matter volume (see Table). By contrast, in AD the hippocampal cell loss and neuritic pathology concentrates in the CA1 and subicular subregions,1–3 which together account for over 60% of hippocampal gray matter volume (see Table). Only neuronal densities have been previously evaluated in the hippocampus of cases with PD and no dementia.10 However, consistent with the findings of this and other volumetric studies,9,18 hippocampal neuron number and volume were not reduced in cases with PD only. This data confirms that hippocampal pyramidal cell loss occurs only in neurodegenerative disorders with clinical dementia syndromes, and that the selective pattern of hippocampal cell loss may contribute to the different clinical features observed in these dementia syndromes. Different hippocampal subregions have significantly different projections and functions.14,15 The presubiculum differs from other hippocampal regions in that it has considerable reciprocal connections with the dorsolateral prefrontal cortex,19 whereas the CA1 and subiculum have major reciprocal connections with the entorhinal cortex and innervate the cingulate cortex through the thalamus.14 The direct prefrontalhippocampal connections are thought to coordinate working memory tasks,19 whereas the thalamic relays are important for memory consolidation and retrieval.14 The present study suggests that DLB cases would have a disruption to working memory because of the Harding et al: Hippocampal Neuron Loss in Lewy Body Dementia 127 considerable pyramidal cell loss in the direct hippocampal output to the dorsolateral prefrontal cortex. Our previous work showing strong correlations between Lewy body formation in the frontal lobe and frontal and hippocampal atrophy in DLB9 also support this thesis, and further suggest that frontal afferent connections with the hippocampus may also be disrupted by Lewy body formation. It would appear that the direct connections between the frontal lobe and hippocampus are significantly affected in DLB. Loss of this pathway would significantly affect memory circuits in a different manner to the loss of the temporal lobe connections found in AD.20 We acknowledge support from Parkinson’s New South Wales, Australian Brain Foundation, National Health and Medical Research Council, Clive and Vera Ramaciotti Foundation, and the Ian Potter Foundation. We thank Heidi Cartwright for the figure work and Dr. Jasmine Henderson for her comments on the manuscript. References 1. West MJ, Coleman PD, Flood DG, Troncoso JC. Differences in the pattern of hippocampal neuronal loss in normal ageing and Alzheimer’s disease. Lancet 1994;344:769 –772. 2. Lippa CF, Smith TW, Swearer JM. Alzheimer’s disease and Lewy body disease: a comparative clinicopathological study. Ann Neurol 1994;35:81– 88. 3. Harding AJ, Wong A, Svoboda M, et al. Chronic alcohol consumption does not cause hippocampal neuron loss in humans. Hippocampus 1997;7:78 – 87. 4. Laakso M, Partanen K, Riekkinen P, et al. Hippocampal volumes in Alzheimer’s disease, Parkinson’s disease with and without dementia, and in vascular dementia: an MRI study. Neurology 1996;46:679 – 681. 5. Nagy Z, Jobst K, Esini M, et al. Hippocampal pathology reflects memory deficit and brain imaging measurements in Alzheimer’s disease: clinicopathologic correlations using three sets of pathologic diagnostic criteria. Dementia 1996;7:76 – 81. 6. O’Brien JT, Paling S, Barber R, et al. Progressive brain atrophy on serial MRI in dementia with Lewy bodies, AD, and vascular dementia. Neurology 2001;56:1386 –1388. 7. Bobinski M, Wegiel J, Wisniewski HM, et al. Atrophy of hippocampal formation subdivisions with stage and duration of Alzheimer’s disease. Dementia 1995;6:205–210. 8. Lehtovirta M, Laakso MP, Soininen H, et al. Volumes of hippocampus, amygdala and frontal lobe in Alzheimer patients with different apolipoprotein E genotypes. Neuroscience 1995; 67:65–72. 9. Cordato NJ, Halliday GM, Harding AJ, et al. Regional brain atrophy in progressive supranuclear palsy and Lewy body disease. Ann Neurol 2000;47:718 –728. 10. Churchyard A, Lees AJ. The relationship between dementia and direct involvement of the hippocampus and amygdala in Parkinson’s disease. Neurology 1997;49:1570 –1576. 11. Harding A, Halliday G. Simplified neuropathological diagnosis of dementia with Lewy bodies. Neuropathol Appl Neurobiol 1998;24:195–201. 12. McKeith IG, Galasko D, Kosaka K, et al. Consensus guidelines for the clinical and pathologic diagnosis of dementia with Lewy bodies (DLB): report of the consortium on DLB international workshop. Neurology 1996;47:1113–1124. 128 Annals of Neurology Vol 51 No 1 January 2002 13. Harding AJ, Halliday GM, Kril JJ. Variation in hippocampal neuron number with age and brain volume. Cereb Cortex 1998; 8:710 –718. 14. Duvernoy HM. The human hippocampus. Functional anatomy, vascularization and serial sections with MRI. Berlin: Springer-Verlag, 1998. 15. Amaral D, Insausti R. Hippocampal formation. In: Paxinos G, ed. The human nervous system. Melbourne: Academic, 1990: 712–756. 16. Barber R, Ballard C, McKeith IG, et al. MRI volumetric study of dementia with Lewy bodies: a comparison with AD and vascular dementia. Neurology 2000;54:1304 –1309. 17. Barber R, McKeith IG, Ballard C, et al. A comparison of medial and lateral temporal lobe atrophy in dementia with Lewy bodies and Alzheimer’s disease: magnetic resonance imaging volumetric study. Dement Geriatr Cogn Disord 2001;12: 198 –205. 18. Double KL, Halliday GM, McRitchie DA, et al. Regional brain atrophy in idiopathic Parkinson’s disease and diffuse Lewy body disease. Dementia 1996;7:304 –313. 19. Goldman-Rakic P, Selemon L, Schwartz M. Dual pathways connecting the dorsolateral prefrontal cortex with the hippocampal formation and parahippocampal cortex in the rhesus monkey. Neuroscience 1984;12:719 –743. 20. Braak H, Braak E, Yilmazer D, et al. Pattern of brain destruction in Parkinson’s and Alzheimer’s disease. J Neural Trans 1996;103:455– 490. A Novel, Blood-Based Diagnostic Assay for Limb Girdle Muscular Dystrophy 2B and Miyoshi Myopathy Mengfatt Ho, DPhil,1 Eduard Gallardo, PhD,2 Diane McKenna-Yasek, RN, BSN,1 Noemi De Luna, PhD,2 Isabel Illa, MD,2 and Robert H. Brown, Jr., MD, DPhil1 Limb girdle muscular dystrophy 2B and Miyoshi myopathy were recently found to be allelic disorders arising from defects in the dysferlin gene. We have developed a new diagnostic assay for limb girdle muscular dystrophy 2B and Miyoshi myopathy, which screens for dysferlin expression in blood using a commercially available monoclonal antibody. Unlike current methods that require muscle biopsy for immunodiagnosis, the new method is simple and entails a significantly less invasive procedure for tissue sampling. Moreover, it overcomes some of the problems associated with the handling and storage of muscle specimens. In our analysis of 12 patients with limb girdle muscular dystrophy 2B or Miyoshi myopathy, the findings obtained using the new assay are fully consistent with the results from muscle immunodiagnosis. Ann Neurol 2002;51:129 –133 Limb girdle muscular dystrophy type 2B (LGMD 2B) and Miyoshi myopathy (MM) have been considered distinct clinical entities because different muscle groups are involved at onset. LGMD 2B begins with predominantly proximal muscle weakness, whereas initial weakness in MM patients invariably affects the distal musculature. However, both disorders are characterized by autosomal recessive inheritance, adult onset, and marked elevations of the muscle enzyme creatine kinase.1 LGMD 2B and MM were recently shown to arise from defects in a novel gene that encodes dysferlin.2,3 Remarkably, the same mutation in the dysferlin gene From the 1Day Laboratory for Neuromuscular Research, Massachusetts General Hospital, Harvard Medical School, Charlestown, Massachusetts; 2Department of Neurology, Neuromuscular Diseases Section, Hospital Santa Creu i St Pau, Universitat Autonoma de Barcelona, Barcelona, Spain Received Aug 16, 2001, and in revised form Aug 16, 2001. Accepted for publication Aug 16, 2001. Published online Dec 28, 2001 Address correspondence to Dr Brown, Day Laboratory for Neuromuscular Research, Massachusetts General Hospital, Harvard Medical School, Bldg. 114, 16th Street, Rm 3125, Charlestown, MA 02129. can cause different clinical presentation, even among members of the same family.4 – 6 In addition, an anterior distal myopathy was recently linked to dysferlin mutation.7 Thus, there is considerable clinical heterogeneity in dysferlinopathy. The dysferlin gene is large, comprising 55 exons that span a genomic region of ⬎150kb.8 It encodes the 237kDa dysferlin protein, composed of 2,080 amino acids. The function of dysferlin is unknown, but its homology to a Caenorhabitis elegans protein, fer-1, suggests that it might function in calcium-mediated membrane fusion or trafficking.9 This hypothesis is supported by recent findings that dysferlin is membrane associated.10 –12 An accurate diagnosis of dysferlinopathy requires a combination of clinical evaluation, protein studies (immunoblot or immunohistochemical analysis), or direct gene analysis. Although the latter approach provides the most definitive diagnosis, it is costly, time-consuming, and labor intensive because of the large size of the dysferlin gene. Moreover, defects in the dysferlin gene are predominantly single nucleotide changes with no evidence of recurrent mutations, gross rearrangements, or mutational hotspots to aid detection.3,8,10 For these reasons, DNA-based diagnosis is difficult as an initial screening strategy to distinguish dysferlinopathies from the other forms of muscular dystrophy. By contrast, screening for defective protein expression by immunoblot analysis has proved to be a reliable and rapid means for differential diagnosis in muscular dystrophies.13 However, current methods require muscle biopsy samples. To avoid this painful and invasive procedure, we sought to develop a new protein-based diagnosis using nonmuscle tissues and a less invasive sampling technique. We have previously shown that dysferlin is expressed in multiple tissues,3,11 we therefore investigated its expression in peripheral blood cells. In the present study, we report that dysferlin is expressed specifically in monocytes and we describe a novel bloodbased diagnostic assay for LGMD 2B and MM. Patients and Methods Isolation of Peripheral Blood Mononuclear Cells and Immunoblot Analysis Peripheral blood mononuclear cells (PBMC) were isolated from whole blood of patients and healthy controls by FicollHypaque gradient centrifugation according to the manufacturer’s instructions (Amersham, Buckinghamshire, UK). PBMC were washed twice in phosphate-buffered saline (PBS) and lysed in 10 volumes of protein extraction buffer, M-PER (Pierce, Rockford, IL). Approximately 20g of protein was separated on a 4 to 15% gradient sodium dodecyl sulfate-polyacrylamide gel electrophoresis gel. Immunoblotting was performed according to standard methods, using primary anti-dysferlin monoclonal antibodies (NCL-Hamlet from Novacastra, UK) at 1:300 dilution. Immunoreactive bands were detected with the enhanced chemiluminescence system (Amersham). © 2001 Wiley-Liss, Inc. DOI 10.1002/ana.10080 129 Fig 1. Dysferlin expression in peripheral blood mononuclear cells (PBMC). (A) Northern blot analysis of dysferlin gene expression in peripheral blood cells from a healthy individual, with a probe corresponding to nucleotides 5364 –5732 of the dysferlin cDNA. A 7.5kb band corresponding to the dysferlin transcript is detected in peripheral blood cells and skeletal muscle. Additional transcripts of ⬃4.5kb and 2.0kb are also present in peripheral blood. (B) Western blot analysis of dysferlin expression in multiple tissues, using the NCL-Hamlet monoclonal antibody. A prominent ⬃230kDa protein is detected in PBMC, but not in white blood cells (WBC), of two unrelated, healthy individuals. The positive controls include skeletal muscle and brain tissues from mouse and human. (C to E) Immunocytochemical analysis of dysferlin expression in CD14⫺ cells (C) and CD14⫹ cells (D). (E) Western blot analysis of dysferlin expression in CD14⫹ cells (lane 1) and CD14⫺ cells (lane 2). Immunoblotting of the 42 kDa actin antigen served as a positive control in both preparations (lanes 1, 2, bottom panel). Separation of PBMC into CD14⫹ and CD14⫺ cell populations PBMC (approximately 107 cells) were mixed with 20l of CD14-coated microbeads (Milteny Biotec, Germany) and incubated at 6 to 12°C for 30 minutes. Unbound microbeads were removed by washing cells in excess PBS buffer followed by centrifugation at 300g for 10 minutes. The cell pellet was resuspended in PBS buffer to a concentration of 2 ⫻ 108 130 Annals of Neurology Vol 51 No 1 January 2002 cells/ml before separation on a MACS apparatus according to manufacturer’s instructions (Milteny Biotec, Germany). Immunocytochemistry PBMC were spun onto microscopic slides using a Cytospin (Shandon, UK) centrifuge. Acetone-fixed preparations were preincubated in PBS buffer containing 0.5% bovine serum albumin and 5% normal goat serum. The sections were in- cubated with NCL-Hamlet (Novacastra, UK), followed by horseradish peroxidase-labeled secondary antibodies (Jackson Immunoresearch, PA) and visualized by diaminobenzidine (Vector Laboratories, Burlingame, CA). Results Detection of Dysferlin Expression in Peripheral Blood Mononuclear Cells As a first step to develop a nonmuscle diagnostic assay for dysferlinopathies, we examined dysferlin gene expression in blood cells of healthy individuals by northern blot analysis. A weak but distinct band of ⬃7.5kb, corresponding to the dysferlin transcript, was detected in total RNA isolated from blood (Fig 1A). Smaller transcripts of ⬃4kb and 2.0kb were also detected in blood cells, suggesting the presence of tissue-specific spliced variants. To determine whether the dysferlin protein is expressed, a Western blot of total white blood cells (WBC), peripheral blood mononuclear cells (PBMC), brain, and skeletal muscle was screened with an antidysferlin monoclonal antibody (NCL-Hamlet). Despite the lower levels of dysferlin gene expression in blood compared with muscle, a prominent ⬃230kDa band corresponding to the dysferlin protein was detected in PBMC, skeletal muscle, and brain but was absent in total WBC (Fig 1B). As PBMC consist primarily of lymphocytes and monocytes, this result suggests that dysferlin is expressed in these cell types. Immunocytochemical Analysis of Dysferlin Expression in PBMC To define further the cell type that expresses dysferlin, we separated PBMC into CD14⫹ and CD14⫺ cells, using magnetic beads coated with CD14 antibodies. This approach distinguishes the two main cell types in PBMC, because CD14 is a specific marker for monocytes. As shown in Figure 1C and D, immunocytochemical analysis showed dysferlin staining only in CD14⫹ cells, but not in CD14⫺ cells. This result was confirmed by Western blot analysis (see Fig 1E). Taken together, these findings indicate that dysferlin is expressed primarily in monocytes. As monocytes constitute only 3 to 7% of total WBC, it is therefore not surprising that we were unable to detect dysferlin expression in total WBC (see Fig 1A). Diagnostic Applications of Dysferlin Expression in PBMC To evaluate the accuracy of a blood-based diagnostic assay for dysferlinopathies, we compared dysferlin expression in PBMC and skeletal muscle in 12 patients with LGMD 2B or MM. In each case, dysferlin deficiency in skeletal muscle was confirmed by Western Fig 2. Diagnostic screening of dysferlin expression in peripheral blood mononuclear cells (PBMC) by Western blot analysis. Lanes 1, 2, 3, 7, 8 contain PBMC from normal healthy controls; lanes 4, 5, 6 (RB 2329, RB 3533, and RB 2079, respectively) are from Miyoshi myopathy (MM) patients (see Table 1 for details of patient genotypes). The effect of delayed processing of blood samples on the stability of the dysferlin protein can be seen by comparing samples that were left for 48 hours (lane 1), 36 hours (lane 2), 24 hours (lanes 3, 7), and 6 hours (lane 8) before being processed for immunoblot analysis. blotting. Conversely, immunoblot analyses of PBMC showed dysferlin reactivity in controls (all 6 tested) but was absent in all 12 patients tested (Fig 2, Table), indicating there is an excellent correlation between dysferlin expression in PBMC and skeletal muscle. We also investigated the effect of delayed processing of blood samples on the stability of the dysferlin protein. Samples that were processed within 36 hours showed no evidence of dysferlin degradation while minor degradation can be seen in samples left for 48 hours (see Fig 2). A lower band of ⬃55kDa was detected in control samples that were left for ⱖ24 hours but not in 6-hour-old samples (see Fig 2), suggesting that this is a partially degraded dysferlin product, as it is present only in control, but not in patient, samples. Thus, only the full-length dysferlin protein should be used as a diagnostic indicator for dysferlin deficiency. Discussion We report the novel finding that dysferlin is expressed in monocytes. More importantly, we showed that dysferlin expression in monocytes correlates with its expression in skeletal muscle (see Table). These findings have led us to develop a new protein-based diagnostic assay for LGMD 2B and MM patients, without the need for a muscle biopsy. The new blood-based diagnostic assay offers several advantages over current methods of muscle immunodiagnosis. This simple and rapid technique does not require specialized equipment or training. In addition, it overcomes some of the inherent problems associated with the handling and storage of muscle specimens. Ho et al: Limb Girdle Muscular Dystrophy 2B and Miyoshi Myopathy 131 Table. Evaluation of Dysferlin Deficiency by Immunodiagnosis in Peripheral Blood Mononuclear Cells and Skeletal Muscle Patient RB 2329 RB 1859 RB 2079 II 1101 II 1102 II 1103 II 1104 II 1303 II 1304 II 1305 II 1306 RB 3533 Clinical Diagnosis MM MM DACM MM MM MM MM LGMD LGMD LGMD LGMD MM 2B 2B 2B 2B Mutation Dysferlin Expression in PBMC Dysferlin Expression in Skeletal Muscle Immunocytochemical Analysis of PBMC E1883X, 6319 ⫹ 1G Deletion AG, at 6071 Deletion G at 5966 n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. n.d. (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) n.d. (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) (⫺) n.d. For positive controls, all 6 PBMC samples obtained from healthy individuals showed positive dysferlin reactivity by Western blot analysis. MM ⫽ Miyoshi myopathy; LGMD 2B ⫽ limb girdle muscular dystrophy; DACM ⫽ distal anterior compartment myopathy (ref 7); X ⫽ stop codon; n.d. ⫽ not determined; PBMC ⫽ peripheral blood mononuclear cells; (⫺) ⫽ absence of dysferlin staining. The most significant improvement offered by the new assay is the considerably less invasive tissue sampling method compared to muscle biopsy. Thus, it should become easier to obtain samples from patients to confirm diagnosis or to guide further testing. Moreover, collection of patient materials will no longer require the operating theater, which will be a significant costsavings benefit. In addition to diagnostic applications, dysferlin expression in monocytes provides a new paradigm to study the biological function of this protein in a nonmuscle cell type that is readily accessible and amenable to in-vitro functional assays. Also, for experimental therapies intended to increase the levels of dysferlin, we could now track efficacy by monitoring dysferlin levels in blood, rather than skeletal muscle. LGMD 2B is not uncommon, accounting for 1 to 25% of limb girdle dystrophies.14,15 Unlike MM patients, LGMD 2B patients are not readily diagnosed because there is considerable genetic heterogeneity in LGMD, and precise DNA diagnostic studies are costly and difficult. Thus, the new assay will be particularly useful as a rapid screen to identify the dysferlin-deficient subgroup within LGMDs. More complex DNA analysis can then be performed selectively in this subgroup, saving both time and costs. To date, all reports of dysferlin deficiency are associated with defects in the dysferlin gene.8,10,11 However, it is unknown whether other forms of muscular dystrophies could cause a secondary reduction in dysferlin levels. Therefore, the new assay will be a useful tool to address this question. Identification of patients with reduced dysferlin expression but without defects in the dysferlin gene could lead to the discovery of proteins that normally associate with dysferlin and help elucidate the pathophysiological pathway involved in dysferlinopathies. 132 Annals of Neurology Vol 51 No 1 January 2002 The authors are very grateful to all the patients who have participated in this study. We thank Drs K. Bejaoui and B. Hosler for the critical reading of this manuscript. This work is supported in part by the Muscular Dystrophy Association (USA), the C. B. Day Investment Company, and Fondo de Investigacı̂on Sanitaria (FIS 99/019-01; FIS 01/0979). References 1. Bushby KM. Making sense of the limb-girdle muscular dystrophies. Brain 1999;122:1403–1420. 2. Bashir R, Britton S, Strachan T, et al. A gene related to Caenorhabditis elegans spermatogenesis factor fer-1 is mutated in limb-girdle muscular dystrophy type 2B. Nat Genet 1998;20: 37– 42. 3. Liu J, Aoki M, Illa I, et al. Dysferlin, a novel skeletal muscle gene, is mutated in Miyoshi myopathy and limb girdle muscular dystrophy. Nat Genet 1998;20:31–36. 4. Illarioshkin SN, Ivanova-Smolenskaya IA, Greenberg CR, et al. Identical dysferlin mutation in limb-girdle muscular dystrophy type 2B and distal myopathy. Neurology 2000;55: 1931–1933. 5. Weiler T, Bashir R, Anderson LV, et al. Identical mutation in patients with limb girdle muscular dystrophy type 2B or Miyoshi myopathy suggests a role for modifier gene(s). Hum Mol Genet 1999;8:871– 877. 6. Weiler T, Greenberg CR, Nylen E, et al. Limb-girdle muscular dystrophy and Miyoshi myopathy in an aboriginal Canadian kindred map to LGMD2B and segregate with the same haplotype. Am J Hum Genet 1996;59:872– 878. 7. Illa I, Serrano-Munuera C, Gallardo E, et al. Distal anterior compartment myopathy: a dysferlin mutation causing a new muscular dystrophy phenotype. Ann Neurol 2001;49: 130 –134. 8. Aoki M, Liu J, Richard I, et al. Genomic organization of the dysferlin gene and novel mutations in Miyoshi myopathy. Neurology 2001;57:271–278. 9. Achanzar W.E, Ward S. A nematode gene required for sperm vesicle fusion. J Cell Sci 1997;110:1073–1081. 10. Anderson LV, Davison K, Moss JA, et al. Dysferlin is a plasma membrane protein and is expressed early in human development. Hum Mol Genet 1999;8:855– 861. 11. Matsuda C, Aoki M, Hayashi YK, et al. Dysferlin is a surface membrane-associated protein that is absent in Miyoshi myopathy. Neurology 1999;53:1119 –1122. 12. Selcen D, Stilling G, Engel AG. The earliest pathologic alterations in dysferlinopathy. Neurology 2001;56:1472–1481. 13. Anderson LV, Davison K. Multiplex Western blotting system for the analysis of muscular dystrophy proteins. Am J Pathol 1999;154:1017–1022. 14. Passos-Bueno MR, Vainzof M, Moreira ES, et al. Seven autosomal recessive limb-girdle muscular dystrophies in the Brazilian population: from LGMD2A to LGMD2G. Am J Med Genet 1999;82:392–398. 15. Fanin M, Pegoraro E, Matsuda-Asada C, et al. Calpain-3 and dysferlin protein screening in patients with limb-girdle dystrophy and myopathy. Neurology 2001;56:660 – 665. Association Studies of Multiple Candidate Genes for Parkinson’s Disease using Single Nucleotide Polymorphisms Yoshio Momose, MD,1,2 Miho Murata, MD, PhD,2 Kazuhiro Kobayashi, PhD,1 Masaji Tachikawa, MSc,1 Yuko Nakabayashi, BSc,1 Ichiro Kanazawa, MD, PhD,2 and Tatsushi Toda MD, PhD1 We studied 20 single nucleotide polymorphisms in 18 candidate genes for association with Parkinson’s disease. We found that homozygosity for the V66M polymorphism of the brain-derived neurotrophic factor (BDNF) gene occurs more frequently in patients with Parkinson’s disease than in unaffected controls (2 ⴝ 5.46) and confirmed an association with the S18Y polymorphism of the UCH-L1 gene. Our results provide genetic evidence supporting a role for BDNF in the pathogenesis of Parkinson’s disease. Ann Neurol 2002;51:133–136 Single nucleotide polymorphisms (SNPs) are singlebase differences in DNA sequences that can be ob- From the 1Division of Functional Genomics, Department of PostGenomics and Diseases, Osaka University Graduate School of Medicine, Osaka, Japan; 2Department of Neurology, Graduate School of Medicine, University of Tokyo, Tokyo, Japan. Received Jun 18, 2001, and in revised form Sep 24, 2001. Accepted for publication Sep 24, 2001. Published online Dec 28, 2001 Address correspondence to: Dr Toda, Division of Functional Genomics, Department of Post-Genomics and Diseases, Osaka University Graduate School of Medicine, 2-2-B9 Yamadaoka, Suita, Osaka 565-0871, Japan. E-mail: firstname.lastname@example.org served between individuals in the population. As such, they represent high-potential polymorphic markers useful in the study of susceptibility genes for multifactorial diseases such as Parkinson’s disease (PD) or those related to drug responsiveness.1 PD, one of the most common human neurodegenerative diseases, is characterized by resting tremor, cogwheel rigidity, bradykinesia, and impaired postural reflexes. This disease affects 1 to 2% of persons older than 65 years of age. Recent studies have indicated the presence of genetic factors to the pathogenesis of PD. For example (1) there are large families with a history of PD, in which the existence of heredity is suggested; (2) approximately 10% of patients with PD have a positive family history2; (3) a recent large-scale survey in Iceland showed that the risk ratio for PD was 6.7 for siblings, 3.2 for offspring, and 2.7 for nephews and nieces of patients with PD3; and (4) a twin study using [18F]-dopa PET showed that the concordance rate for PD, including subclinical cases, is approximately three times higher in monozygotic twins (55%) than in dizygotic twins (18%).4 Causal genes for Mendelianinherited PD have been reported, including ␣-synuclein (4q21–23, autosomal dominant [AD]),5 parkin (6q25.2–27, autosomal recessive [AR]),6 UCH-L1 (4p14, AD),7 PARK3 (2p13, AD),8 PARK4 (4p15, AD),9 PARK6 (1p35–36, AR),10 PARK7 (1p36, AR).11 Thus far, 84 association studies on 14 genes have examined relationships between PD and polymorphisms12; however, SNPs that influence PD as strongly as APOE-ε4 influences Alzheimer’s disease have not been identified. To identify susceptibility genes for PD and obtain information helpful for establishing tailor-made medicines, we initiated large-scale association studies using SNPs in various candidate genes and further surveyed the correlation between SNPs and susceptibility to various side effects. Patients and Methods A total of 232 patients (114 men and 118 women) with PD and 249 normal controls (127 men and 122 women) were examined in this study. The mean age of onset was 54.1 ⫾ 10.5 (mean ⫾ SD) years [range:24 – 80]; 18 patients showed early onset of PD (⬍40 years of age), and 13 patients had a positive family history of PD. Patients who carried parkin mutations were excluded. All patients and controls were of Japanese ancestry. Patients were evaluated by certified neurologists specializing in PD. The diagnosis of idiopathic PD was based on the presence of two or more of the cardinal features of PD (tremor, rigidity, bradykinesia, and postural instability) and a lack of other signs or clinical history that might indicate alternative diagnoses or causes. We were careful to exclude secondary parkinsonism. Informed consent was obtained from each patient, and approval for the study was obtained from the University Ethical Committees. © 2001 Wiley-Liss, Inc. DOI 10.1002/ana.10079 133 Table 1. Results of Association Studies Gene Name Dopamine-related Nerve growth factorrelated Metabolism or transportation of toxins Familial parkinsonism a SNP Namea GenBank Accession No. SNP Positionb PD : Control (Allele) 2 (df ⫽ 1) p DBH AADC COMT DRD3 BDNF DBHu1 AADCd5 COMTd7 DRD3u1 BDNFu1 G910T (A304S) A649G (M217V) G214T (A72S) A25G (S9G) G196A (V66M) G/T ⫽ 447/5 : 472/6 A/G ⫽ 26/438 : 32/464 G/T ⫽ 424/38 : 444/40 A/G ⫽ 319/145 : 339/155 G/A ⫽ 249/213 : 282/190 0.01 0.30 0.00 0.00 3.26 0.926 0.582 0.982 0.966 0.071 CNTF NGFB CYP2C9 CNTFu1 NGFBu5 CYP2C9*3 A545G (H182R) T116C (V39A) A1075C (I359L) A/G ⫽ 450/12 : 486/10 T/C ⫽ 69/391 : 96/396 A/C ⫽ 449/15 : 473/13 0.36 3.38 0.26 0.548 0.066 0.611 T163A (L55M) A1286C (E429A) C559T (P187S) C3435T (I1145I) C53A (S18Y) T/A ⫽ 427/29 : 443/45 A/C ⫽ 369/95 : 399/91 C/T ⫽ 270/188 : 300/192 C/T ⫽ 285/179 : 290/204 C/A ⫽ 261/199 : 244/252 2.67 0.55 0.40 0.74 5.45 0.102 0.458 0.525 0.391 0.020 PON1 MTHFR DIA4 ABCB1 UCH-L1 Z70723 AF105983 J03934 M29445 AF076269 SNP names follow nomenclature of the referenced SNP databases. A of the initial translation site ATG is designated ⫹1. Amino acid changes are indicated in parentheses. b Genomic DNA was extracted from whole blood using standard methods. Samples were genotyped by allele-specific oligonucleotide (ASO) hybridization. Briefly, each polymerase chain reaction (PCR) product containing a SNP locus was transferred to a nylon membrane (Biodyne, Pall, Port Washington, NY), chemically denatured, and fixed by ultraviolet light. 32P-labeled allele-specific probes were hybridized to PCR products on the membranes, and washed at 38 to 52°C in 6⫻ SSC. Chi-square analysis was used to evaluate the associations and Yates correction was carried out to avoid type I error, when there was a variable of ⬍10. Differences at p ⬍ 0.05 were considered significant. Most SNPs were obtained from the SNP database maintained by the Center for Genome Research at the Whitehead Institute for Biomedical Research, Cambridge, Massachusetts (http://www-genome.wi.mit.edu/cvar_snps) and by the Human Cytochrome P450 (CYP) Allele Nomenclature Committee (http://www.imm.ki.se/CYPalleles/default.htm). Results We selected 18 genes with potential involvement in PD on the basis of their actual or potential functions in the following categories: 1. Dopamine metabolism, receptors, and transportation: Aromatic L-amino acid decarboxylase (AADC); catechol-O-methyltransferase (COMT); dopamine ␤-hydroxylase (DBH); dopamine receptor D3 (DRD3); dopamine receptor D5 (DRD5); and solute carrier family 6, member 3 ⫽ dopamine transporter (SLC6A3 ⫽ DAT1) 2. Nerve growth factors and their receptors: Brainderived neurotrophic factor (BDNF); ciliary neurotrophic factor (CNTF); nerve growth factor, ␤-polypeptide (NGFB); and neurotrophic tyrosine kinase, receptor, type1 (NTRK1) 134 Annals of Neurology Vol 51 No 1 January 2002 3. Metabolism or transportation of toxins: Cytochrome P450 1A2 (CYP1A2); cytochrome P450 2E1 (CYP2E1); cytochrome P450 2C9 (CYP2C9); PON1 (paraoxonase 1); 5,10methylenetetrahydrofolate reductase (MTHFR); diaphorase 4 (DIA4); and ATP-binding cassette, subfamily B, member 1 ⫽ multidrugresistance 1 (ABCB1). 4. Familial parkinsonism: Ubiquitin C-terminal hydrolase L1 (UCH-L1) We genotyped 20 SNPs in the 18 candidate genes. All SNPs studied are coding SNPs that result in amino acid changes, except for the ABCB1 SNP, which affects mRNA stability. Table 1 shows the results of our study. Seven of 20 SNPs (AADCd4, COMTu4, SLC6A3d13, DRD5u15, NTRK1u7, CYP2E1*2, CYP1A2*2) showed only one allele in our samples (data not shown). Of the 20 SNPs tested, two showed association with PD. We found that homozygosity for the V66M polymorphism of the BDNF gene is more frequent in PD patients than in unaffected controls (2 ⫽ 5.46, p ⫽ 0.019). In addition, we observed that homozygosity for the S18Y polymorphism of the UCH-L1 gene is less frequent in PD patients (2 ⫽ 5.41, p ⫽ 0.020) (Table 2). We examined the association between each SNP and susceptibility to adverse L-dopa-associated effects such as dyskinesia, wearing-off-type motor fluctuation, and neuroleptic malignant syndrome, but we found no significant associations (data not shown). Discussion Our analysis shows that homozygosity for the V66M polymorphism in the BDNF gene (BDNFu1) is more Table 2. Frequency of the BDNF Gene Polymorphism (G196A, V66M) and UCH-L1 Gene Polymorphism (C53A, S18Y) Allele A BDNF PD Control UCH-L1 PD Control G Genotype Total AA AG or GG Total 213 249 462 190 282 472 (2 ⫽ 3.26, df ⫽ 1, p ⫽ 0.071) 48 30 183 206 231 236 199 261 460 252 244 496 2 ( ⫽ 5.45, df ⫽ 1, p ⫽ 0.020) 40 65 (2 ⫽ 5.46, df ⫽ 1, p ⫽ 0.019) 190 183 230 248 (2 ⫽ 5.41, df ⫽ 1, p ⫽ 0.020) PD ⫽ Parkinson’s disease. frequent in PD patients. The neurotrophin BDNF aids the survival, growth, and maintenance of neurons in the central and peripheral nervous systems. It protects dopaminergic neurons from stresses such as deprivation of serum or MPTP-induced damage.13 Studies have shown that BDNF concentrations are reduced in the nigrostriatal region of the brain in PD as compared with controls.14 The V66M polymorphism lies within the proBDNF sequence, which is cleaved posttranslationally. Although how this polymorphism might influence susceptibility to PD remains unclear, some possible scenarios have been suggested. First, the polymorphism may affect production of mature, biologically active BDNF by altering proBDNF processing. Lower levels of mature BDNF may reduce protection of dopaminergic neurons from environmental toxins such as pesticides. Second, Mowla and colleagues15 recently reported that some proBDNF is released extracellularly and is biologically active, as demonstrated by its ability to mediate receptor activation. Thus, the V66M polymorphism has the potential to affect the activity of extracellular proBDNF activity as well. Finally, other SNPs that associate with a more powerful influence on the onset of PD might exist in linkage disequilibrium with BDNFu1. UCH-L1 represents 1 to 2% of total soluble brain protein and is thought to hydrolyze polymeric ubiquitin and ubiquitin conjugates to produce monomeric ubiquitin. Immunoreactivity for this protein is found in Lewy bodies. A missense mutation in the UCH-L1 gene has been reported in relation to PD.7 In that paper, Leroy and colleagues showed that this mutation, I93M, causes a partial loss of catalytic thiol protease activity of this enzyme, which may lead to aberrations in the proteolytic pathway and aggregation of proteins. Some association studies demonstrated that carriers of the S18Y polymorphism have a significantly lower risk of PD.16 –18 Our results are in accordance with these studies. The amino acid sequence of UCH-L1 is highly conserved among rat, mouse, and humans, with approximately 95% sequence identity. Isoleucine (I) at position 93 is conserved in all three species, and mutation of this residue to methionine (M) is causative in familial PD in a German family. The amino acid at position 18 is serine (S) or tyrosine (Y) in humans and alanine (A) in rat and mouse. The amino acid change at this position may exert a milder functional influence than the change at position 93. Another study has suggested that the S18Y polymorphism in UCH-L1 does not influence the risk of the development of PD in the Australian population.19 Further investigation is needed to clarify the potential role of this polymorphism in PD. Another coding SNP of UCH-L1, M124L, was recognized in a French family with PD.20 Although it did not co-segregate and was considered not to be pathogenic in this family, it would be of interest to know whether this SNP has an association with sporadic PD. Considering the significance of the ubiquitin-protease system in neurodegenerative diseases, the role of UCH-L1 is of great interest. Recent research has established that SNPs influence the variability in patients’ responses to drugs. Although we found no significant results in our current stratified analyses, we consider it important to establish tailormade treatments by utilizing SNPs for patients with PD. We are increasing the number of patients and SNPs in our studies to test for association. In summary, we report the first genetic association between BDNFu1 and the onset of PD. This finding provides the evidence supporting a role for BDNF in the pathogenesis of PD and information that may prove useful for future gene therapy approaches using neurotrophic factors. This work was supported by research grants from the Ministry of Health, Labor, and Welfare (H12-Genome-025) and the Ministry Momose et al: Association Studies for PD by SNPs 135 of Education, Culture, Sports, Science, and Technology (12204035) in Japan. References 1. Gray IC, Campbell DA, Spurr NK. Single nucleotide polymorphisms as tools in human genetics. Hum Mol Genet 2000;9: 2403–2408. 2. Elbaz A, Grigoletto F, Baldereschi M, et al. Familial aggregation of Parkinson’s disease. Neurology 1999;52:1876 –1882. 3. Sveinbjörnsdóttir S, Hicks AA, Jónsson T, et al. Familial aggregation of Parkinson’s disease in Iceland. N Engl J Med 2000; 343:1765–1770. 4. Piccini P, Burn DJ, Ceravolo R, et al. The role of inheritance in sporadic Parkinson’s disease: evidence from a longitudinal study of dopaminergic function in twins. Ann Neurol 1999;45: 577–582. 5. Polymeropoulos MH, Lavedan C, Leroy E, et al. Mutation in the ␣-synuclein gene identified in families with Parkinson’s disease. Science 1997;276:2045–2047. 6. Kitada T, Asakawa S, Hattori N, et al. Mutations in the parkin gene cause autosomal recessive juvenile parkinsonism. Nature 1998;392:605– 608. 7. Leroy E, Boyer R, Auburger G, et al. The ubiquitin pathway in Parkinson’s disease. Nature 1998;395:451– 452. 8. Gasser T, Müller-Myhsok B, Wszolek ZK, et al. A susceptibility locus for Parkinson’s disease maps to chromosome 2p13. Nat Genet 1998;18:262–265. 9. Farrer M, Gwinn-Hardy K, Muenter M, et al. A chromosome 4p haplotype segregating with Parkinson’s disease and postural tremor. Hum Mol Genet 1999;8:81– 85. 10. Valente EM, Bentivoglio AR, Dixon PH, et al. Localization of a novel locus for autosomal recessive early-onset Parkinsonism, PARK6, on human chromosome 1p35–p36. Am J Hum Genet 2001;68:895–900. 136 Annals of Neurology Vol 51 No 1 January 2002 11. van Duijn CM, Dekker MCJ, Bonifati V, et al. PARK7, a novel locus for autosomal recessive early-onset parkinsonism, on chromosome 1p36. Am J Hum Genet 2001;69:629 – 634. 12. Tan EK, Khajavi M, Thornby JI, et al. Variability and validity of polymorphism association studies in Parkinson’s disease. Neurology 2000;55:533–538. 13. Hyman C, Hofer M, Barde YA, et al. BDNF is a neurotrophic factor for dopaminergic neurons of the substantia nigra. Nature 1991;350:230 –232. 14. Mogi M, Togari A, Kondo T, et al. Brain-derived growth factor and nerve growth factor concentrations are decreased in the substantia nigra in Parkinson’s disease. Neurosci Lett 1999; 270:45– 48. 15. Mowla SJ, Farhadi HF, Pareek S, et al. Biosynthesis and posttranslational processing of the precursor to brain-derived neurotrophic factor. J Biol Chem 2001;276:12660 –12666. 16. Maraganore DM, Farrer MJ, Hardy JA, et al. Case-control study of the ubiquitin carboxy-terminal hydrolase L1 gene in Parkinson’s disease. Neurology 1999;53:1858 –1860. 17. Wintermeyer P, Krüger R, Kuhn W, et al. Mutation analysis and association studies of the UCHL1 gene in German Parkinson’s disease patients. Neuroreport 2000;11:2079 –2082. 18. Zhang J, Hattori N, Leroy E, et al. Association between a polymorphism of ubiquitin carboxy-terminal hydrolase L1 (UCHL1) gene and sporadic Parkinson’s disease. Parkinsonism Relat Disord 2000;6:195–197. 19. Mellick GD, Silburn PA. The ubiquitin carboxy-terminal hydrolase-L1 gene S18Y polymorphism does not confer protection against idiopathic Parkinson’s disease. Neurosci Lett 2000; 293:127–130. 20. Farrer M, Destée A, Becquet E, et al. Linkage exclusion in French families with probable Parkinson’s disease. Mov Disord 2000;15:1075–1083.