Minamata Disease)

Neuropathology 2010; 30, 471–479 doi:10.1111/j.1440-1789.2010.01119.x

The 50th Anniversary of Japanese Society of Neuropathology Memorial Symposium: Milestones in Neuropathology from Japan The pathology of methylmercury poisoning (Minamata disease)

Komyo Eto,1 Masumi Marumoto2 and Motohiro Takeya3 1Health and Nursing Facilities for the Aged, Jushindai, Shinwakai, 2Pathology Section, Department of Basic Medicine, National Institute for Minamata Disease, and 3Department of Cell Pathology, Graduate School of Medical Sciences, Kumamoto University, Kumamoto, Japan

Methylmercury (Me-Hg) poisoning (Minamata disease: imura et al.2,3 reported that large amounts of Me-Hg had MD) is one of the most severe types of disease caused by been generated by chemical processes of the Chisso Co. humans to humans in Japan. The disease is a special class of acetaldehyde plant in August 1951 and were later dumped food-borne methylmercury intoxication in humans as typi- directly into Minamata Bay (Fig. 1). fied by the outbreak that began in 1953 in Minamata and its The pathogenesis of chronic types of MD was at first vicinity in Kumamoto Prefecture, Japan. There are 450 considered to be due to brain damage by low-level persis- autopsy cases in Kumamoto and 30 autopsy cases in tent exposure to Me-Hg.4 However, it was later realized to Niigata Prefecture related to MD in Japan. Two hundred be the after-effects of high-level Me-Hg intake by the resi- and one cases in Kumamoto and 22 cases in Niigata showed dents around Minamata Bay between 1951 and 1968, pathological changes of MD. This report provides a brief because the mercury levels of fish abruptly dropped in 1968 research history and overview of the pathological changes (Fig. 2). of MD, and also presents representative cases of adult, Also, the pathogenesis of selective vulnerability within infantile and fetal forms of MD among the 450 MD-related the cerebral cortex was not clear for a long time. Eto et al.5,6

autopsy cases in Kumamoto Prefecture.neup_1119 471..479 demonstrated experimentally using common marmosets that edema in the white matter near the deep sulci may Key words: human autopsy cases, methylmercury poison- contribute to the selective damage of the cerebral cortex. ing, Minamata disease, pathogenesis, three types of According to new reports over the last decade, medical Minamata disease. studies appear to have resolved the MD problem.

INTRODUCTION RESEARCH HISTORY OF MD The results of the investigation of the causes of Minimata It was in 1953 that MD was first recognized by the medical disease (MD) by the first MD study group at Kumamoto profession as a mysterious neurological illness occurring in University School of Medicine have been widely acknowl- the Minamata Bay area of Kumamoto Prefecture, Kyushu, edged in Japan.1 In 1968, the Japanese government Japan. The earliest phase of investigation into this disorder officially recognized the disease was caused by human was a personal one; Hosokawa, then Physician-in-chief at ingestion of a large amount of methylmercury (Me-Hg)- the hospital run by the chemical plant later identified as the contaminated fish or shellfish from Minamata Bay and that source of the mercury pollution responsible for the illness, it injured mainly the nervous system. But it was long made clear the unique clinical features of the disorder unclear that the cause was the huge amount of Me-Hg through detailed observation of patients during the period dumped into Minamata Bay. New facts came to light only 1953 through 1956, and further suggested the likely after the political solution of MD problems in 1995. Nish- involvement of seafood from Minamata Bay in its etiology. This ground-breaking work of Hosokawa should have immediately become widely known but instead remained Correspondence: Komyo Eto, Director, Health and Nursing Facilities largely in the form of personal notes mainly due to sup- for the Aged, Jushindai, Shinwakai, 272 Ikura Kitakata, Tamana City, pression by his employer. Kumamoto 865-0041, Japan. Email: [email protected] Received 18 February 2010 and accepted 8 March 2010; published In 1956 when the outbreak was already in an endemic online 24 May 2010. stage, a systematic endeavor to clarify the nature of the

1951.8 1932 1951 1952 1956 1968

Acetaldehyde producing system

The outbreak Acetylene Acetaldehyde of Minamata Methylmercury disease chloride Mercury-catalyst Ferric sulfide Manganese dioxide (Co-catalyst) (Co-catalyst)

Pollution of methylmercury chloride in Minamata Bay, Kumamoto Fig. 1 The co-catalyst was changed in Chisso Co.’s acetaldehyde-producing system

(Eto K and Kuwana T, 2002) from manganese dioxide to ferric sulﬁde in August, 1951.

such that intoxication was strongly implicated. A number

of possibilities including Mn, CS2, Cu, Zn, Tl, Se, As, and V were considered. In 1959, Takeuchi read the previous description of human alkylmercury poisoning made by Hunter and Russell.8 This led him to the notion that the neurological disorder seen around Minamata Bay must have been caused by alkylmercury compounds. In the meantime, he and his colleagues were able to demonstrate the feeding animals with fish or shellfish from Minamata Bay could produce a similar neurological disorder. This finding, which was consistent with the possibility of food- borne intoxication, was soon confirmed by Hosokawa and his collaborators. Investigation revealed that the chemical plant had been utilizing mercuric sulfate as the catalyst for acetaldehyde synthesis in sharply increasing amounts and dis- Fig. 2 After Chisso Co. stopped dumping wastewater into carding the waste catalyst into the effluent outlet directly Minamata Bay in 1868, the contents of mercury in the fish abruptly decreased. connected to the sea. It was strongly suggested that the inorganic mercury discharged from the plant was somehow responsible for the disease. However, there was disease was initiated. A five-member committee compris- a missing link between the organic and inorganic forms of ing Katsuki (internal medicine), Rokutanda (microbiol- mercury. ogy), Takeuchi (pathology), Kitamura (public health) and Soon afterwards, a second outbreak of Minamata Ozaki (pharmacology), was organized at Kumamoto Uni- disease took place between 1964 and 1965, in Niigata versity School of Medicine. By 1959,Takeuchi and his asso- approximately 250 km north of Tokyo. This outbreak was ciates had made autopsies on a total of 11 human victims the subject of detailed studies by Tsubaki and other and subjected local animals (cats, birds and fish) exhibiting researchers from Niigata University School of Medicine.9–11 motor function abnormalities through pathological exami- Mercuric catalyst for acetaldehyde synthesis was again nation.7 The pathological findings in the central nervous identified as the culprit. A difference from the Minamata system of affected humans and animals, characterized by outbreak was that a river (the Agano River) rather than atrophy and the absence of inflammatory changes, were the sea was polluted.

Two important discoveries soon followed. In 1961, The selective vulnerability within the cerebral cortex was Uchida and his associate at the Department of Biochem- clarified with the study of Me-Hg poisoning in common istry, Kumamoto University School of Medicine, succeeded marmosets by Eto et al. in 2001.5 The selective cortical in detecting a methylmercury compound (methylmercury degeneration occurred along the deep cerebral fissures or sulfide) in shellfish samples taken from Minamata Bay. In sulci (Figs 3,4). 1962, Irukayama and his colleagues at the Department of The following three cases reports involve an adult case, a Hygiene, Kumamoto University School of Medicine, iden- mild type of MD, a postnatal MD and a fetal MD among tified methylmercuric chloride in sludge from the acetalde- autopsy cases in Kumamoto Prefecture. There were five hyde plant and the bottom sediment of the effluent postnatal cases of MD, and all of them showed severe channel. He postulated that it was formed from mercuric neuronal damage with spongy change in the cerebral cortex. sulfate as a by-product in the reaction for acetaldehyde Five fetal cases of MD showed hypoplasia of the nervous synthesis. The causal links between the source and the system without spongy change in the cerebral cortex. disease thus became evident. It should be added that Hosokawa independently succeeded in detecting a meth- OVERVIEW OF THE PATHOLOGICAL ylmercuric compound in the effluent of the plant at about CHANGES OF MD the same time. This achievement was published by Eto et al. in 2001.12 The most prominent feature of MD, or Me-Hg poisoning in After 1995, the political problems related to MD were general, is marked organ selectivity.Thus, significant patho- resolved in Japan and new facts have been gradually logical changes are limited essentially to the nervous revealed. For example, Nishimura2 and Nishimura and system. According to the studies conducted by the study Okamoto3 reported that large amounts of Me-Hg were group of Kumamoto University,14 changes in other organs generated by the chemical processes of the Chisso Co. and tissues were generally slight and included erosive acetaldehyde plant and later dumped directly into inflammation in the digestive tracts (the duodenum in par- Minamata Bay.This occurred due to technological changes ticular), hypoplasia of the bone marrow, atrophy of the introduced in the production process. In August 1951, man- lymph node, fatty degeneration of the liver and kidney, and ganese dioxide, initially used as a reaction to maintain the the alteration of pancreatic islet cells. activity of Hg catalyst, was changed to ferric sulphide. In the liver and kidney, in which mercury levels at the Ferrous iron was reduced in the reaction and then oxidized time of death were invariably higher than those in the with nitric acid. In 1968, the plant stopped releasing waste- brain, pathological changes were rather slight not only in water into the bay. During 17 years of pollution, fish and mild types but also in severe types of MD. In the kidney, shellfish accumulated Me-Hg in their gills and intestinal abundant mercury deposits were demonstrated in the epi- tracts. The amount of Me-Hg in the aquatic biota rose thelial cells of proximal convoluted tubules, although there sharply in 1952, but dropped in 1968 (Fig. 2). were no noticeable pathological changes. In the liver, Minamata disease is divided into seven different clinical mercury deposits were detected in hepatocytes as well as types.4 The acute type is characterized by acute onset, Kupffer cells, but tissue damage was not substantial. severe neurological signs, and an onset–death interval of In contrast, Me-Hg-induced damage to the nervous shorter than 2 months. The subacute type also exhibits system can be devastating. However, it never affects the acute onset and severe neurological signs, but the onset– system evenly: as a rule, the damage was the severest in death interval is between 2 and 12 months. The prolonged- the cerebral and cerebellar cortices, in which some parts severe type has acute or subacute onset and severe were affected more severely than others. The brain stem neurological signs and symptoms, with an onset–death was affected to a lesser extent, and the spinal cord was least interval of longer than 12 months.The prolonged-mild type affected. On the other hand, the pathology of peripheral is characterized by mild neurological manifestations and nerves is unique in that it appears to be associated with an onset–death interval of longer than 12 months. The prolonged duration of the disease: the nerves are affected chronic type shows insidious onset and only vague neuro- only in cases other than those of acute and subacute types. logical signs. The fetal and postnatal types are both MD in The sensory nerves are damaged selectively with regenera- infants and children, caused by intrauterine and postnatal tion in prolonged cases. exposures to Me-Hg, respectively. In acute MD, two outstanding features were apparent. Case 1 (adult case)15 One was circulatory disturbance resulting from damage to Clinical history the blood–brain barrier by the Me-Hg compound. Brain edema was observed in the perivascular space, and was This patient was a 64-year-old fisherman who lived in accentuated in the boundary zones with perivascular space. Minamata City in the southern part of Minamata Bay,

Fig. 3 Preferential sites in the nervous system for pathological changes in mild, chronic types of MD. The degree of damage to the brain is highest in the meshed region followed by the stippled and hatched region in order of decreasing severity. The sensory nerves prone to pathological changes are marked with crosses. Adapted from Eto.13 which was found to be polluted with mercury from the the Sugamo myelin and Suzuki’s axon staining methods. nearby Chisso Co. Onset of disease was marked by numb- The Sugamo myelin stain was modified for use on frozen ness of the feet and disturbance in speech in the Spring of sections from Kultschiky’s method. Inorganic mercury was 1959. The patient was treated at Minamata City Hospital detected by photo-emulsion. for pulmonary tuberculosis during the period from May The gyri of both hemispheres were atrophic and the 1965 until July 1968. Neurological examination in October sulci were widened.This was particularly remarkable in the 1968 and December 1969 revealed slight constriction of calcarine cortex and pre- and postcentral gyri. The surface visual fields on the temporal side, muscle rigidity, increased of the calcarine cortex showed moderate atrophy, with tendon reflexes, tremor of the fingers, dysgraphia, and adia- widening of the calcarine fissure on the coronal section. dochokinesis. Other clinical findings included labyrinthine Gennari’s band on the calcarine cortex was stained pale deafness, hyperesthesia, and hypalgesia as well as dysesthe- with the KB staining method. sia in the hands and regions below the knees, elevated Light microscopic observation revealed depopulation of blood pressure of 170–192 mmHg, a mask-like face, and neurons in calcarine, temporal, pre- and postcentral corti- dyskinesia.The patient died of massive hemorrhage from a ces, and the deep white matter showed diffuse degenera- gastroduodenal ulcer in January 1970. tion, as is commonly observed in cases of chronic MD. The calcarine cortex showed severe neuronal loss of whole layers. There was moderate loss of granule cells under the Autopsy pathology Purkinje cell layer in the cerebellar hemispheres (Fig. 5). Autopsy materials from the cerebrum, cerebellum, brain Mercury granules were detected in Bergman’s glial cells stem, spinal cord, and peripheral nerves were embedded in and the granule cell layer using a photo-emulsion his- paraffin and stained with HE, and with KB and Bodian tochemical method for inorganic mercury. Degeneration of staining methods. Frozen sections were made from periph- the fasciculus gracilis (Goll’s tract) in the spinal cord was eral nerves including ventral and dorsal root nerve fibers, noted, but ganglion cells in the spinal ganglion were rela- sciatic nerve, radial nerve and sural nerve, and stained by tively well preserved. Sensory nerves, such as dorsal roots

Fig. 5 Slight or moderate loss of granule cells beneath the Purkinje cell layer associated with scars in the cerebellum of Case 1 (HE stain).

Biopsy findings As the patient was not initially recognized as having MD, a sural nerve biopsy was performed on December 9, 1969, about 1 month before his death. The biopsy of the sural nerve showed a decrease in the number of myelinated nerve fibers and increase in small axons with attendant proliferation of fibroblasts and Schwann’s cells. Electron microscopic observation of the sural nerve included irregular Schwann’s cells, and appearance of fibroblasts with an increase of collagen fibers. Regressive changes were characterized by degeneration resulting in swollen myelin, wavy degeneration of myelin with extremely thin and electron-dense axons, incomplete regeneration including abnormally small axons and incomplete myelination and absence of myelin.

Case 2 (infantile case)16 Clinical history The patient was a 23 year-old woman, born on November 8, 1950.The onset of Minamata disease was on June 8, 1956, when she was 5 years and 7 months old, and she died after Fig. 4 Differential distribution of central nervous system lesions a total course of 18 years. She came from a family with in different types of MD of prolonged duration.Top, adult; middle, many MD patients.Around June 8, 1956, salivation became 14 infantile; bottom, fetal type of MD. Adapted from Takeuchi. striking. On June 15, motions of the upper limbs, especially those of the fingers, became jerky. On June 18, tremors of and sural nerves, were disintegrated, showing Büngner’s the fingers and a disturbance in gait appeared. On June 20, bands and a loss of nerve fibers with increase of collagen her speech became inarticulate and she was admitted to fibers. Myelinated nerve fibers of the ventral root were well the Chisso Co. hospital. On July 3, she became entirely preserved by myelin staining, but myelin sheath destruc- unable to walk and showed tremors in the neck. Aphasia tion was seen in the dorsal root.Axon staining showed that appeared on July 30. Her condition progressively wors- axons of ventral root nerve fibers were well preserved, but ened, and she became manic following the onset of dysph- the dorsal nerve fibers showed a band-like increase in the agia and somnipathy. On August 30 she was transferred small nerve fibers with associated proliferation of fibro- to the Department of Pediatrics, Kumamoto University blasts and Schwann’s cells. Hospital, Kumamoto.

Physical examinations disclosed the presence of tonic paralysis which rendered the activities of daily living (standing and walking) impossible. Disorders of visual acuity, hearing disturbance, aphasia and disturbance of consciousness were present. Tendon reflexes were markedly increased, accompanied by ankle clonus, tremors of the four limbs and pathological reflexes. Neck rigidity and Kernig’s sign were also present. There were no striking abnormalities in the eye grounds. On June 5, 1957, she suffered her first seizure of convulsions, followed by similar attacks about 10 times a day. She occasionally assumed a posture with her four limbs stretched or with the knee and hip joints flexed at right angles. She also occasionally kicked and struggled with her lower limbs. Her dementia advanced. The tonicity and spasticity of her four extremities became aggravated, and the motor and mental functions were entirely lost. On July 29, 1959, she was transferred to the Minamata City Hospital. When she received food and liquid directly into her mouth, she was able to swallow. When an excessive amount of food was given, she refused it by closing her mouth. She occasionally had general convulsions. On May 22, 1974, tracheotomy was performed against aspiration. Oral alimentation became impossible, and she was placed on a naso-gastric tube for alimentation of synthetic formula. She showed apallic syndrome. Infections of the urethra and respiratory disturbances occurred repeatedly until she died on August 25, 1974.

Fig. 6 A horizontal section of the right hemisphere showing Autopsy pathology gliosis in the white matter corresponding to the affected areas of Case 2 (Holzer stain). From Eto and Takeuchi.16 The brain weighed 775 g and the atrophy degree was 37% compared to a control (brain weight, 1234 1 17.9 g). The lesions involved a wide area of the cerebral hemisphere, tracts occurred secondarily and predominantly in the and the calcarine cortex, pre-and postcentral gyri were lateral column. There were no remarkable changes in the severely damaged (Fig. 6). The white matter of the cere- neurons of the anterior and posterior horns, apart from brum displayed secondary degeneration in accordance occasional atrophy. In the spinal ganglia, there was rela- with the intense damage of the cerebral cortex. The pyra- tively slight satellitosis following loss of ganglion cells, com- midal tracts from the precentral gyri and internal sagittal pared with the situation in the brain cortex. The dorsal strata, consisting of corticofugal fibers passing from the roots were predominantly damaged with regeneration. occipital lobe to the superior colliculi and the lateral gen- iculate bodies, were involved. They showed little or no Case 3 (fetal case)17 myelin staining. The fibers of the corpus callosum desig- Clinical history nated as the tapetum were less strikingly involved. The lesion of the cerebellum was severe. The neurons in the The patient was a 29-year-old woman, born in 1957, who dentate nucleus were relatively well preserved compared died in 1987 in Minamata. Four of her eight siblings (four to those in the cerebellar cortex. In this case, changes in the males and four females) and her parents were diagnosed dendrites of Purkinje cells and torpedoes were prominent. with MD. Her mother’s hair contained 101 ppm of total Stellate cells were found in the molecular layer as the mercury in 1959. The mother died of rectal cancer in 1972 report of a Hunter-Russell’s case.8 No loss of neurons was at 55 years of age. identified in the nuclei of the basal ganglion or brain stem, This patient’s birth weight was 3000 g. As a baby, she but the cell bodies of the neurons were frequently atrophic. was fed mainly her mother’s milk mixed with formula. She Systemic damage of both the Goll’s tracts and pyramidal sucked poorly,her development was slow,and her neck was

© 2010 Japanese Society of Neuropathology Neuropathology of Minamata disease 477 not fixed at 6 months of age. She developed her first con- vulsive seizure at 3 years, when she was taken to a private hospital. There, she was diagnosed as “Kibyo” (a strange disease), a term used in earlier phases of the MD outbreak. She suffered repeated convulsions. At age eight, EEG at sleep showed diffuse and persistent slow waves with high voltage. Somatic and mental developments were retarded. She salivated copiously, never learned to speak, and was bedridden. Neurological examination revealed the presence of spastic quadriparesis, primitive and pathological reflexes, increased deep-tendon reflexes, and ankle clonus. Choreic and athetotic movements were observed episodically. There were external strabismus and abnormal dentition. Finally she died of bronchopneumonia at 29 years of age. The content of total mercury in her hair was 61.9 ppm in 1959 at two years of age, and 5.4 ppm 15 years later when she was 17 years old.

Autopsy pathology The body weighed 23 kg and measured 143 cm in height. The brain weighed 920 g. Grossly, the brain exhibited marked diffuse atrophy of both the cerebral cortex and white matter, thinning of the corpus callosum, and status marmoratus of the thalamus. Microscopically, there was atrophy and a slight decrease in the number of neurons with gliosis in the calcarine, Fig. 7 Electron micrograph of a neuron in the calcarine cortex of postcentral, and precentral cortices in the cerebrum. Cal- Case 3. Nerve cells are shrunken and darkly stained with an increase of nuclear chromatin. Free ribosomes are seen through- cification was present in the globus pallidus and neurons out the cytoplasm, with some being aggregated in places. Disap- decreased in number in the basal ganglia. Granule cells in pearance of rough endoplasmic reticula (Nissl bodies) and the cerebellum were relatively well-preserved as revealed well-preserved mitochondria are noted. From Eto et al.17 by HE stain, whereas slight but distinct pathological changes in the apex of the folia, so-called apical scar for- were markedly decreased in number. Aggregates of rough mation, were observed with gliosis in the granule cell layer ER (Nissl bodies) were not observed in the cytoplasm of beneath the Purkinje cell layer. neurons, but a few swollen cisternae were noted.The Golgi Histochemical analysis revealed mercury deposits in the apparatus was occasionally found, and its cisternae were brain, kidney and liver. In the brain, deposits were found in usually swollen. Lipofuscin was also observed in the cyto- neurons and other cells in the cerebral cortices, basal plasm. Mitochondria were well-preserved (Fig. 7). In addi- ganglia, ependymal cells, epithelial cells of the choroid tion, autophagosomes were increased in number. They plexus, and the nuclei of the cerebellum and brain stem. localized widely in perikaryon occasionally with grouping, They were found diffusely in granule cells in the cerebellar and engulfed some pieces of cytoplasm or membranous cortex. Ventral nerve roots of the spinal cord were intact, structures in large or small vacuoles. Membrane-bound but connective tissues increased in the endoneurium of globular dense bodies of 0.3–1.8 mm in diameter were small bundles of dorsal nerve roots. Segmental demyelina- found in the cerebrum. One or several of these structures tion in the dorsal nerve fibers was revealed by a teasing were observed in both perikarya and dendrites of the method. neuron. In the cerebellum, Purkinje cells were atrophic with high electron density. Nuclei of Purkinje cells were Electron microscopic findings shrunken with aggregation of chromatin, and the nuclear In the cerebrum, nerve cells were shrunken and darkly membrane was occasionally indistinct. Many autophago- stained with an increase of nuclear chromatin. Free ribosomes which were seen in cerebral neurons were also somes were present diffusely with focal aggregation in the found in the perikarya of Purkinje cells. The Golgi appara- cytoplasm of neurons. Rough endoscopic reticula (ER) tus showed enlargement of the cisternae. Membrane-

Pathological changes caused by Me-Hg occur predominantly in selective areas of the cerebrum, including the calcarine region, the post- and precentral gyri and the temporal transverse gyrus.19 These are localized near the deep sulci, comprising the calcarine fissure, central sulci (Roland’s fissure) and Sylvian’s fissure (Figs 3,4). Ischemia may be a result of the compression of arteries by edema of the adjacent tissues. Studies of acute Me-Hg poisoning in marmosets revealed edema in the white matter of occipital lobes. In acute cases of Me-Hg poisoning, neuron loss with gliosis was found in all layers of the cortex.The second and third layers of cortices are damaged in moderate or mild cases of poisoning. As a result of the location of the pathological changes, there were bilateral concentric constriction of the visual fields and impairment of visual acuity. Simi- larly, other effects became noticeable, including sensory disturbance due to damage of the sensory center (postcentral gyrus), motor disturbance due to damage of the motor center (precentral gyrus) and hearing impairment.13 Takeu- chi and Eto4 have summarized all MD-related autopsy cases in Kumamoto Prefecture from 1956 to 1995. It was difficult to clarify the pathogenesis of chronic MD. Nishimura3 and Nishimura and Okamoto4 found the true causes of MD. Examinations were made on formalin- preserved specimens, obtained in 1956 and since kept in the Second Department of Pathology of Kumamoto Uni- Fig. 8 Electron micrograph of molecular layer of the cerebellum versity.The contents of mercury in fish and shellfish caught showing well-developed parallel fibers of irregular thickness, in Minamata Bay in 1956 showed remarkable levels. Purkinje cell spines forming synapses (center) with large parallel Total mercury levels showed 51.6 ppm in the muscle and fibers and enlarged presynaptic terminals of Case 3. From Eto et al.17 109.6 ppm in the liver of Pagrus major (bream), and 38.6 ppm in the muscle and 200.0 ppm in the liver of Phyn- 4 bound dense bodies were observed in the cytoplasm of copelates oxyhynchus (sharpnose tigerfish). Purkinje cells. Granule cells in the cerebellum were focally After Chisso Co. stopped dumping wastewater into the atrophic with high electron density. Others were clear with Bay in 1968, the contents of mercury in the fish and shell- an edematous perikaryon. A few free ribosomes were fish abruptly decreased. Then the pathogenesis of chronic found in each of the atrophic granule cells, but they were type of MD was thought to be the after-effects of the rare in swollen granule cells. Parallel fibers were mixed in high-level Me-Hg intake by the residents around the molecular layer. Parallel fibers were well-preserved, Minamata Bay. but their size was not uniform. The spines of Purkinje cells Sensory disturbance was the most important sign and showed high electron density.These spines formed synaptic symptom of MD, not only in human autopsy cases, but also 6 contacts to the big parallel fibers. The terminals of presyn- with the experimental Me-Hg poisoning in marmosets, apses were enlarged and contained large mitochondria or rats, mice, and swine. The cause of sensory disturbance of synaptic vesicles (Fig. 8). MD was considered to be damage to both the central sensory center (postcentral gyri) and peripheral sensory nerves. DISCUSSION A report from the Second Department of Pathology, ACKNOWLEDGMENTS Kumamoto University School of Medicine in 1959, indi- cated that organic mercury was the most probable cause of The authors thank the late Dr Tadao Takeuchi, Professor MD.18 One week later, Hosokawa et al. initiated an experi- Emeritus, Kumamoto University, and members of the ment in order to assess the toxicity of industrial wastewater Second Department of Pathology at the Kumamoto Uni- from the acetaldehyde plant but the results were not pub- versity School of Medicine for their cooperation with the lished until 2001.12 autopsies. The authors also thank Dr Cheng-Mei Shaw,

Professor Emeritus, University of Washington, Seattle, 10. Tsubaki T, Irukayama K, eds. Minamata Disease: Meth- Washington and Dr Hajime Nishimura for their comments ylmercury Poisoning in Minamata and Niigata, Japan. on the pathogenesis of MD. Tokyo: Kodansha Ltd, 1977. 11. Eto K, Takahashi H, Kakita A et al. Pathological and REFERENCES biochemical studies of 30 Niigata autopsy cases related 1. Kumamoto University Study Group for Minamata to Minamata disease (in Japanese). Jpn J Hyg 2007; 62: Disease. Historical perspective of study on Minamata 70–88. disease. In: Kutsuna M, ed. Minamata Disease. Kuma- 12. Eto K, Yasutake A, Nakano A et al. Reappraisal of moto: Kumamoto Shuhan Publishing Co, 1968; 1–4. historic 1959 cat experiment in Minamata by Chisso 2. Nishimura H. The resolution of the questions of occur- Factory. Tohoku J Exp Med 2001; 194: 197–203. rences in Minamata disease (in Japanese). Gendai 13. Eto K. Pathology of Minamata disease. Toxicol Pathol Kagaku (Chemistry Today) 1998; 323: 60–66. 1997; 6: 614–623. 3. Nishimura H, Okamoto T. The Science of Minamata 14. Takeuchi T. Pathology of Minamata disease. In: Disease (in Japanese). Tokyo: Nihon Hyoronsha Co., Kutsuna M, ed. Minamata Disease. Kumamoto: Kuma- Ltd., 2001. moto Shuhan Publishing Co, 1968; 141–256. 4. Takeuchi T, Eto K. The Pathology of Minamata 15. Eto K, Tokunaga H, Nagashima K et al. An autopsy Disease. A Tragic Story of Water Pollution (ed. collabo- case on Minamata disease (Methylmercury poisoning) ration by Nakayama H, Sumiyoshi A). Fukuoka: – pathological viewpoints of peripheral nerves. Toxicol Kyushu University Press Inc., 1999. Pathol 2001; 6: 714–722. 5. Eto K, Yasutake A, Kuwana T et al. Mercury poisoning 16. Takeuchi T, Eto N, Eto K. Neuropathology of child- in common marmosets – a study of selective vulner- hood cases of methylmercury poisoning (Minamata ability within the cerebral cortex. Toxicol Pathol 2001; disease) with prolonged symptoms, with particular ref- 5: 565–573. erence to the decortication syndrome. Neurotoxicology 6. Eto K, Yasutake Y, Korogi Y et al. Methylmercury 1979; 1: 1–20. poisoning in common marmosets – MRI ﬁndings and 17. Eto K, Oyanagi S, Itai Y et al. A fetal type of Minamata peripheral nerve lesions. Toxicol Pathol 2001; 6: 723– disease–An autopsy case report with special reference 734. to the nervous system. Mol Chem Neuropathol 1992; 7. Takeuchi T, Morikawa N, Matsumoto H et al. A patho- 16: 171–186. logical study of Minamata disease in Japan. Acta Neu- 18. Takeuchi T, Kambara T, Morikawa N. Pathologic roathol 1962; 2: 40–57. observations of the Minamata disease. Acta Pathol Jpn 8. Hunter D, Russell DS. Focal cerebral and cerebellar 1959; 9 (Suppl):769–783. atrophy in a human subject due to organic mercury 19. Eto K, Takeuchi T. A pathological study of prolonged compounds. J Neurol Neurosurg Psychiatry 1954; 17: cases of Minamata disease. Acta Pathol Jpn 1978; 28: 235–241. 565–584. 9. Tsubaki T. Outbreak of intoxication by organic compounds in Niigata Prefecture. Jpn J Med Sci Biol 1967; 6: 132–133.