- Email: [email protected]
Modifications of Oligodendroglial Cells in Spongiform Encephalopathies
EXPERIMENTAL NEUROLOGY ARTICLE NO.
154, 23–30 (1998)
EN986894
Modifications of Oligodendroglial Cells in Spongiform Encephalopathies Khalid Hamid El Hachimi,* Marie-Pierre Chaunu,† Paul Brown,‡ and Jean-Franc¸ois Foncin* *Ecole Pratique des Hautes Etudes and U.106 INSERM, Hoˆpital de la Salpeˆtrie`re, Bat. de Pe´diatrie, 47 Bd de l’Hoˆpital, 75651 Paris cedex 13, France; †Service de Neurologie, CHU de Reims, 51100 Reims, France; and ‡Laboratory of Central Nervous System Studies, NINCDS, National Institutes of Health, Bethedsa, Maryland Received February 6, 1998; accepted July 2, 1998
Although gray matter lesions involving neurones and astrocytes are prominent in human transmissible spongiform encephalopathies (TSE), white matter lesions have also been occasionally observed. Secondary (Wallerian) degeneration and direct myelin damage have been invoked, but the physiopathology of white matter involvement is still debated. We performed an immunohistochemistry study with anti-PrP antibodies of autopsy material of four patients with Creutzfeldt–Jakob disease (CJD), together with transmission electron microscopy (TEM) studies of conventionally processed biopsy specimens of the same patients. Light microscopy immunolabeling was observed as arrays adjacent to myelinic fibers and as a clumps adjacent to oligodendroglial nuclei; both cerebrum and cerebellum were involved. At the ultrastructural level, two types of intracellular inclusions were seen in the white matter. They were associated with dense lysosomes in oligodendroglial perikarya and in their processes. The inclusions were made of finely fibrillar, paracrystalline, amorphous, or densely osmophilic material. Thus, our findings may suggest that white matter involvement in spongiform encephalopathy is due to direct modifications of oligodendroglial cells associated with abnormal metabolism of PrP. r 1998 Academic Press Key Words: prions; immunohistochemistry; ultrastructure; Creutzfeldt–Jakob disease; white matter.
(5, 20, 22, 24, 27). Such cases have been designated as a ‘‘panencephalic’’ variant of CJD (20, 22, 25). Its pathogenesis has remained unclear. The amyloid plaques observed in some forms of TSE, as well as immunoreactive ‘‘synaptic’’ deposits and white matter deposits (4, 13), are mostly constituted by an abnormal variant of the PrP protein, a proteaseresistant fragment (PrPres ) of a protease-sensitive precursor protein (PrPsens ). The fine localization of white matter PrPres immunolabeled deposits has not yet been elucidated. Several experimental studies have suggested that many types of CNS cells (as well as other cell types) are able to produce PrPsens (2). PrP mRNA has been detected by in situ hybridization in neurones (12), but the possibility of in situ production of PrPsens by normal glial cells (astrocytes and oligodendroglial cells) has also been suggested by Moser et al. (21). The presence of PrPres immunoreactivity in corpus callosum white matter tracts of Syrian hamsters with experimental TSE, however, has been interpreted as a consequence of PrPres production by cortical neurones and secondary axonal transport (26). We wanted to study whether white matter involvement in spongiform encephalopathies is due to direct participation of oligodendroglial cells or to secondary phenomena; we addressed this question by means of light microscopy immunohistochemistry of autopsy CJD material and by TEM analysis of corresponding biopsy specimens.
INTRODUCTION
MATERIAL AND METHODS
Transmissible spongiform encephalopathies (TSE) comprise a group of slow degenerative diseases of the central nervous system (CNS) characterized by microspongiosis (spongiform change), gliosis, neuronal loss, and amyloid plaques (although one or more of these characteristics may be lacking). Attention has been focused mainly on the gray matter changes observed in these conditions. In some instances, however, histopathological involvement of white matter in the cerebral hemisphere and brain stem has been reported
A group of 20 autopsy CJD brains were retrospectively studied with PrP immunohistochemistry. Four cases, aged 45–69 years, were selected for further study based on the following twin criteria: the presence of PrP deposits in white matter and the existence of white matter in biopsy material embedded for TEM. As control, we used archivial material obtained from agematched controls affected with other conditions, e.g., Alzheimer’s disease, and dating back to the same epoch as did the CJD material. 23
0014-4886/98 $25.00 Copyright r 1998 by Academic Press All rights of reproduction in any form reserved.
24
EL HACHIMI ET AL.
OLIGODENDROGLIAL CELLS IN SPONGIFORM ENCEPHALOPATHIES
Biopsy Material Biopsies were performed for diagnostic purposes in the years 1965–1975 as requested by clinical neurologists with the informed consent of the next of kin. Specimens were obtained from the second frontal gyrus of the right hemisphere. Fixation was performed in the operating room a few seconds after the specimen had been taken with a cutting no. 4 curette. The specimens were fixed overnight at 4°C by immersion in 1% glutaraldehyde and 2% paraformaldehyde freshly prepared in 0.1 M phosphate buffer (Karnovsky’s solution). Specimens were then cut into 1-mm pieces, postfixed in 1% osmium tetroxyde for 2 h, dehydrated through a graded series of ethanol, and embedded in Araldite. Semithin sections from each block were performed and were stained with toluidine blue. Ultrathin sections from white matter blocks were counterstained with lead citrate and uranyl acetate and examined with a Siemens 102 electron microscope. Autopsy Material Brains obtained at autopsy were fixed in formaline and slices were embedded in paraffin or celloidin. Multiple blocks for paraffin embedding were taken from cerebellum, frontal and temporal cortex, hippocampus, thalamus, basal ganglia, and medulla oblongata. Blocks embedded in celloidin generally comprised a whole cerebral or cerebellar hemispheric slice. Sections were stained, respectively, with hematein-eosine and modified Masson’s trichrome, and with hemateine eosine, thionin blue (Nissl), Loyez ferric hematoxylin myelin stain, Van Gieson picro-fuchsin, and Mallory’s phophotungstic acid hematoxylin. Immunohistochemistry Seven-micrometer sections from paraffin-embedded blocks were treated according to standard immunostaining methods. We used two rabbit anti-PrP antisera at a dilution of 1/500 (both with 95% formic acid pretreatment for 10 mn). One antiserum was developed by one of us (P.B.), and the second was a gift of Professor Bugiani (6). Antibody reactivity was visualized by the Peroxydase-antiperoxydase (PAP) system (Amersham) and 3,38 diaminobenzidine (DAB) as chromogen. We also used as control antibodies raised against b/A4 amyloid (Dako) at a dilution of 1/400 (with 95% formic
25
acid pretreatment for 10 mn) and against Tau protein (ICN) at a dilution of 1/400. For immunostaining (counterstaining) of cerebellar white matter Purkinje axones, an antibody against Calbindin (CaBP28K) protein at a dilution of 1/30000 (Swant) was used. Specificity of the immunostaining was controlled by incubating sections with Tris buffer, or preimmune serum as a primary antibody, or by absorption of primary antibodies, as well as by applying the full procedures to other material without spongiform encephalopathy. RESULTS
Diagnosis The neuropathological diagnosis of spongiform encephalopathy was confirmed by light microscopy analysis of paraffin and celloidin sections of autopsy material and electron microscopy analysis of biopsy material. They showed spongiosis, neuronal loss and severe astrogliosis of grey matter, but neither severe involvement of white matter nor definite myelinic palor were observed. At the ultrastructural level, membranebound microspongiosis bubbles containing secondary vesicles (‘‘spongiform change’’) were abundant in the cortex. Immunocytochemistry Anti-PrP immunolabeling revealed several types of deposits in white matter (Figs. 1A–1D). Immunolabeling was observed in all the regions studied, namely cerebellar and hemispheric white matter including hippocampus, although not simultaneously in each case. We could distinguish morphologically two types of PrP deposits in white matter. One of them presented as small clumps or elongated deposits arranged parallel to, and in close topographic relationship with, white matter myelinated fibers, especially in cerebellar white matter (Figs. 1B, 1C, and 1E). The second one presented as a ‘‘spider’s web’’-like structure in white matter especially in the cerebral white matter. Immunolabeled clumps appeared to be intracellular; the corresponding cells displayed the morphological characteristics of white matter oligodendroglial cells as seen after formaline fixation and routine embedding (small cells with round or oval eccentric nuclei and clear cytoplasm) (Figs. 1A, 1D, and 1E). No significant glial reaction
FIG. 1. Light microscopy, sections were hematoxylin nuclear counterstain. (A) (3160) Parietal white matter, small patches of PrP immunoreactivity. Note deposit associated to oligodendroglial cell (arrow). (B) (3180) Cerebellum: double immunostaining; anti-PrP reaction was revealed by 5-bromo-4-chloro-3-indolyl-phosphate and 4-nitro blue tetrazolium chloride as chromogen producing a blue deposit; anti-CaBp28K reaction was revealed by 3,38 diaminobenzidine as chromogen, producing a brown deposit, used as counterstaining. PrP deposits are parallel to CaBP-positive fibers. Note PrP deposits in granular layer (GL). (C) (3140) White matter of 2nd parietal gyrus. Arrays of parallel PrP immunostaining material. (D) (3180)) PrP immunoreactivity located in cells with morphological characteristics of oligodendroglial cells. (E) (3160) arrays of parallel PrP immunostaining adjacent to oligodendroglial cell (arrow) associated to clumped immunostaining.
26
OLIGODENDROGLIAL CELLS IN SPONGIFORM ENCEPHALOPATHIES
around the PrP deposits was detected either at the light microscope or at the electron microscope levels. No immunolabeling with anti-PrP was observed in controls, conversely no immunolabeling with anti-b amyloid or anti-Tau was observed in the CJD specimens. Electron Microscopy Identification of oligodendroglial cells, as distinct from microglial cells or astrocytes, was established by morphological criteria (23). Their rounded or ovalshaped nucleus showed an electron density greater than that of astrocytes and contained a clumped heterochromatin; it was generally eccentric and surrounded by only a thin rim of cytoplasm, except at one pole of the cell where a larger cytoplasmic expansion was found. The cytoplasm was electron dense with rare glycogen granules; processes contained microtubules. The general appearance of TSE oligodendroglial cells was moderately abnormal, with a dilated endoplasmic reticulum and cytoplasm containing dense membranebound lysosomes in greater abundance and/or size than in controls, without any evidence of macrophagic activity. The most striking feature was the presence of abnormal inclusions in these cells and in their processes. Two types of inclusions could be distinguished according to their osmophilia: the less osmiophilic inclusions (type I) were made of amorphous and fibrillar material (Figs. 2A, 2B, and 3C); the more osmiophilic ones (type II) were made of compact material (Figs. 3A and 3B), resolved at high magnification as finely granular (Fig. 3B). Both materials sometimes coexisted in the same inclusion (mixed type) (Figs. 3A and 3B). The fibrillar material was either loose or arranged in a ‘‘fingerprint’’ structure (Fig. 3C). The fibrils were 5 to 10 nm in diameter. The fibrillarparacrystalline inclusions did not display any special arrangement, in contrast to osmiophilic inclusions, which appeared generally star-like. These inclusions were not membrane-bound. Their structure was quite distinct from the lysosome-like dark structures observed in controls. Myeline sheaths frequently showed vacuoles, disruption, and occasional phagocytosis of myelin debris; neurites, however, were generally normal. DISCUSSION
White matter has been generally considered to be unaffected in CJD, although its involvement has been occasionally reported (20, 22, 25, 27) as ‘‘myelinic palor’’ or myelin degeneration associated with phagocytosis
27
and gliosis. White matter involvement has been considered to be most frequent in Japanese cases of CJD (20). Moderate degeneration of cerebral white matter or of fibers tracts in the spinal cord has been reported in most cases of Kuru (9) and was a feature of the original case of GSS (10). The question is whether, when present, white matter involvement is a primary feature (20) or a secondary one (5, 22). The fine localization of PrP deposits might be a clue towards the solution of this problem through the identification of the primary lesion, inasmuch PrPres is widely held to be directly involved in TSE pathogenesis (4, 13). PrP deposits are potentially detectable in most, if not all, cases of TSE, if sensitive immunohistochemistry procedures are used (3, 14, 15). Various types of PrP deposits have been described; probably all regions of the CNS may host one or another type of deposit. White matter deposits seem to be an uncommon feature of ‘‘European’’ CJD, as they were present in only 4 of our 20 cases. These PrPres deposits were generally not correlated with white matter neuropathological lesions as defined by classic histological methods. The deposits did not display the specific tinctorial and physicochemical properties of amyloid fibrils: they were neither birefringent nor congophilic. White matter deposits have been regarded (26) as produced by neurones, implying that the products are axonally transported; other explanations are possible, because glial cells have also been shown to be able to produce PrP (6, 21). The fine localization of PrP, however, cannot be completely determined with light microscope methods. The fact that immunolabeling of PrPres at the ultrastructural level could not be achieved with postembedding methods (the only ones applicable to archivial material), either by ourselves or by previous authors, makes impossible direct TEM determination of immunoreactive sites; as a consequence, we are restricted to comparison with standard TEM images. Electron microscopy of white matter, however, has attracted little interest in human TSE (8, 10). A few ultrastructural abnormalities have been described, such as myelin sheath disruption and intramyelin vacuolization (19), but these were for years dismissed as artefacts due to fixation or embedding procedures. The ultrastructural pathology of animals with experimental TSE, which provides optimal conditions of fixation and embedding, has been extensively studied, particularly by Liberski et al. (16, 17). These studies have shown that, in contrast with the predominantly gray matter pathology produced by most CJD strains, widespread myelin and axonal pathology are produced by the Fujisaki CJD and
FIG. 2. Electron microscopy. (A) (320,000) White matter oligodendrocyte shows fibrillar-paracrystalline (type I) inclusion (I). Note the enlargement of rough and smouth ergastoplasmic saccules. (B) (369,500) High magnification of the same oligodendroglial process with inclusion constituted by fibrillar and amorphous material. The inclusion is not membrane bound.
FIG. 3. Electron microscopy. (A) (318,000) Two white matter oligodendroglial cells showing two inclusions (type II and mixed). One inclusion contains only dense osmiophil material (II) the second one is made of more or less osmiophilic materials (mixed inclusion) (M). Note the presence of type II inclusion in a cell process (arrow). (B) (334,000) mixed inclusion in cell process, the dense, osmiophilic material is granular (arrow). (C) (337,500) Type I inclusion in cell process; note the paracrystaline, often fingerprint-like, texture of the material (arrows).
28
OLIGODENDROGLIAL CELLS IN SPONGIFORM ENCEPHALOPATHIES
263K scrapie strains. White matter lesions included myelin sheath damage or vacuolation, phagocytosis of myelin debris, and neuroaxonal dystrophy; myelin damage has been considered to be related to tumor necrosis factor; no significant alteration of oligodendroglial cells has been reported (18). On the other hand, four findings concur with our conclusion that oligodendroglial cells are primarily affected in our material. First, they show slight, but consistent general cytoplasmic changes. Second, they contain fibrous inclusions; although comparison of this material with Scrapie Associated Fibrils (SAF) is tempting, it cannot directly be done, the morphology of SAF being largely determined by the extraction procedure. Third, these inclusions are not seen in controls, which display only the classical lipofuscin-like inclusions. Fourth, there is a general concordance between the presence of these inclusions and the sites of PrP immunoreactivity; although the fine localization of the latter could not be determined (at the light microscopy level), their localization in the vicinity of oligodendroglia-like nuclei and the absence of any significant microglial reaction around these deposits (in contrast with Kuru plaques or with AD senile plaques, which are often associated with microglial cells or astrocytes (1, 7)) is consistent with an intracellular, and more precisely oligodendroglial, localization of white matter PrP deposits. Negatively, the absence of significant axone damage speaks again of primitive neuronal involvement in TSE white matter changes. Our interpretation of the abnormal inclusions seen in oligodendroglial cells as intracellular PrPres deposits is in consequence founded on four arguments, none of which would be convincing if taken alone, but which, taken together, amount to a prima facie case. Although myelin lesions were minimal in our samples, we feel that our conclusions can be extrapolated to the rarer instances with obvious white matter damage. In conclusion, we found modifications in oligodendrocytes, but not in axones, in the white matter of CJD brains. This result may indicate the primary nature of white matter involvement that occurs in some instances of that condition.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13. 14.
ACKNOWLEDGMENTS We are grateful to Dr. O. Bugiani for the generous gift of anti-PrP antisera. We thank Dr. P. Liberski for review of the manuscript. D. Lecren is thanked for skillful iconographic assistance. This work has been supported in part by l’Action Concerte´e Coordonne´e du Ministe`re de la Recherche, by La Fondation de France (grant to K. H. E.), and aided by European Union Biomed 2 Concerted Action ‘‘The human prion diseases’’ (Project Leader: H. Budka).
15.
16.
17.
REFERENCES 1.
Barcikowska, M., P. P. Liberski, J. W. Boellaard, P. Brown, D. C. Gajdusek, and H. Budka. 1993. Microglia is a component of the
18.
29
prion protein amyloid plaque in the Gerstmann-Stra¨usslerScheinker syndrome. Acta Neuropathol. 85: 623–627. Bendheim, P. E., H. R. Brown, R. D. Rudelli, L. J. Scala, N. L. Goller, G. Y. Wen, R. J. Kascsak, N. R. Cashman, and D. C. Bolton. 1992. Nearly ubiquitous tissue distribution of the scrapie agent precursor protein. Neurology 42: 149–156. Bruce, M. E., P. A. McBride, and C. A. Farquhar. 1989. Precise targeting of the pathology of the sialoglycoprotein, PrP, and vacuolar degeneration in mouse Scrapie. Neurosci. Lett. 102: 1–6. Budka, H., A. Aguzzi, P. Brown, J. M. Brucher, O. Bugiani, F. Gullota, M. Haltia, J-J. Hauw, J. W. Ironside, K. Jellinger, H. A. Krezschmar, P. L. Lantos, C. Masullo, W. Schlote, J. Tateishi, and R. O. Weller. 1995. Neuropathological diagnostic criteria for Creutzfeldt-Jakob disease (CJD) and other human spongiform encephalopathies (Prion diseases). Brain Pathol. 5: 459–466. Bugiani, O., F. Tagliavini, G. Giaccone, and R. Boeri. 1986. Creutzfeldt-Jakob disease: Astrocytosis and spongiform changes of the white matter. In Unconventional Virus Diseases of the Central Nervous System (L. A. Court, D. Dormont, P. Brown, and D. T. Kinsbury, Eds.), pp. 172–183. Diedrich, J. F., P. E. Bendheim, Y. S. Kim, R. I. Carp, and A. T. Haase. 1991. Scrapie-associated prion protein accumulates in astrocytes during Scrapie infection. Proc. Natl. Acad. Sci. USA 88: 375–379. El Hachimi, K. H., and J-F. Foncin. 1994. Do microglial cells phagocyte the b/A4-amyloid senile plaque core of Alzheimer disease? C. R. Acad. Sc. Se´rie III 317: 445–451. Foncin, J-F., J. Gaches, and J. Le Beau. 1964. Ence´phalite spongiforme (apparente´e a` la maladie de Creutzfeldt-Jakob). Biopsie e´tudie´e au microscope e´lectronique, confirmation autopsique. Rev. Neurol. 111: 507–515. Gajdusek, D. C., and C. J. Gibbs. 1978. Unconventional viruses causing the spongiform virus encephalopathies. A fruitless search for the coat and core. In Viruses and Environment. (E. Kurkas and K. Maramovosch, Eds.), pp. 70–98, Academic Press, New York. ¨ ber eine Gerstmann, J., E. Stra¨ussler, and I. Scheinker. 1936. U eigenartige heredita¨r-familia¨re Erkrankung des Zentralnervensystems. Zugleich ein Beitrag zur Frage des vorzeitigen localen Alterns. Z. Ges. Neurol. Psychiatr. 154: 736–762. Gonatas, N. C., R. D. Terry, and M. Weiss. 1965. Electron microscopy study in two cases of Jakob-Creutzfeldt disease. J. Neuropathol. Exp. Neurol. 24: 575–598. Harris, D. A., P. Lele, and W. D. Snider. 1993. Localization of the mRNA for a chicken prion protein by in situ hybridization. Proc. Natl. Acad. Sci. USA 90: 4309–4313. Ironside, J. W. 1996. Review: Creutzfeldt-Jakob disease. Brain Pathol 6: 379–388. Jeffrey, M., C. M. Goodsir, M. E. Bruce, P. A. McBride, J. R. Scott, and W. G. Halliday. 1992. Infection specific prion protein (PrP) accumulates on neuronal plasmalemma in Scrapie infected mice. Neurosci. Lett. 147: 106–109. Kitamoto, T., K. Doh-Ura, T. Muramoto, M. Miyazono, and J. Tateishi. 1992. The primary structure of the prion protein influences the distribution of abnormal prion protein in the central nervous system. Am. J. Pathol. 141: 271–277. Liberski, P. P., R. Yanagihara, C. J. Gibbs, and D. C. Gajdusek. 1989. White matter ultrastructural pathology of experimental Creutzfeldt-Jakob disease in mice. Acta Neuropathol. 79: 1–9. Liberski, P. P., R. Yanagihara, D. M. Asher, C. J. Gibbs, and D. C. Gajdusek. 1990. Reevaluation of the ultrastructural pathology of experimental Creutzfeldt-Jakob disease. Brain 113: 121–137. Liberski, P. P., R. Yanagihara, V. R. Nerurkar, and D. C. Gajdusek. 1993. Tumor Necrosis ~Factor produces Creutzfeldt-
30
EL HACHIMI ET AL.
Jakob disease-like lesions in vivo. Neurodegeneration 2: 215– 225. 19. Masters, C. L., and D. C. Gajdusek. 1982. The spectrum of Creutzfeldt-Jakob disease and the virus-induced subacute spongiform encephalopathies. In Recent Advances in Neuropathology (W. T. Smith, and J. B. Cavanagh, Eds.), pp. 139–163, Churchill Livingstone, Edinbourg. 20. Mizutani, T., A. Okumara, M. Oda, and H. Shiraki. 1982. Panencephalic type of Creutzfeldt-Jakob disease: Primary involvement of white matter. J. Neurol. Neurosurg. Psychiatr. 44: 103–115. 21. Moser, M., R. J. Collelo, U. Pott, B. Oesch. 1995. Developmental expression of the prion protein gene in glial cells. Neuron 14: 509–517. 22. Park, T. S., G. M. Kleiman, and E. P. Richardson. 1980. Creutzfeldt-Jakob disease with extensive degeneration of white matter. Acta Neuropathol. 52: 239–242.
23. 24.
25.
26.
27.
Peters, A., S. L. Palay, and H. de F. Webster. 1976. The Fine Structure of Nervous System, Saunders, Philadelphia. Power, J. M., Y. Liu, L. Hair, R. J. Kascsack, L. D. Lewis, and L. A. Levy. 1991. Concomitant Creutzfeldt-Jakob disease and Alzheimer’s disease. Acta Neuropathol. 83: 95–98. Su, M., S. Wang, H. Chiang, J. Wang, Y. Tan, J. Fuh, and H. Liu. 1993. Creutzfeldt-Jakob disease: report of case of the pancenphalopathic type and experience of a teaching hospital in Taiwan. Eur. Neurol. 33: 241–24. Taraboulos, A., K. Jendroska, D. Serban, S. L. Yang, S. J. DeArmond, and S. B. Prusiner. 1992. Regional mapping of prions proteins in brain. Proc. Natl. Acad. Sci. USA 89: 7620– 7624. Vallat, J. M., M. Dumas, N. Corvisier, M. J. Lebutet, A. Loubet, P. Dumas, and F. Cathala. 1983. Familial Creutzfeldt-Jakob disease with extensive degeneration of white matter. Ultrastructural study of peripheral nerves. J. Neurol. Sci. 61: 261–275.
154, 23–30 (1998)
EN986894
Modifications of Oligodendroglial Cells in Spongiform Encephalopathies Khalid Hamid El Hachimi,* Marie-Pierre Chaunu,† Paul Brown,‡ and Jean-Franc¸ois Foncin* *Ecole Pratique des Hautes Etudes and U.106 INSERM, Hoˆpital de la Salpeˆtrie`re, Bat. de Pe´diatrie, 47 Bd de l’Hoˆpital, 75651 Paris cedex 13, France; †Service de Neurologie, CHU de Reims, 51100 Reims, France; and ‡Laboratory of Central Nervous System Studies, NINCDS, National Institutes of Health, Bethedsa, Maryland Received February 6, 1998; accepted July 2, 1998
Although gray matter lesions involving neurones and astrocytes are prominent in human transmissible spongiform encephalopathies (TSE), white matter lesions have also been occasionally observed. Secondary (Wallerian) degeneration and direct myelin damage have been invoked, but the physiopathology of white matter involvement is still debated. We performed an immunohistochemistry study with anti-PrP antibodies of autopsy material of four patients with Creutzfeldt–Jakob disease (CJD), together with transmission electron microscopy (TEM) studies of conventionally processed biopsy specimens of the same patients. Light microscopy immunolabeling was observed as arrays adjacent to myelinic fibers and as a clumps adjacent to oligodendroglial nuclei; both cerebrum and cerebellum were involved. At the ultrastructural level, two types of intracellular inclusions were seen in the white matter. They were associated with dense lysosomes in oligodendroglial perikarya and in their processes. The inclusions were made of finely fibrillar, paracrystalline, amorphous, or densely osmophilic material. Thus, our findings may suggest that white matter involvement in spongiform encephalopathy is due to direct modifications of oligodendroglial cells associated with abnormal metabolism of PrP. r 1998 Academic Press Key Words: prions; immunohistochemistry; ultrastructure; Creutzfeldt–Jakob disease; white matter.
(5, 20, 22, 24, 27). Such cases have been designated as a ‘‘panencephalic’’ variant of CJD (20, 22, 25). Its pathogenesis has remained unclear. The amyloid plaques observed in some forms of TSE, as well as immunoreactive ‘‘synaptic’’ deposits and white matter deposits (4, 13), are mostly constituted by an abnormal variant of the PrP protein, a proteaseresistant fragment (PrPres ) of a protease-sensitive precursor protein (PrPsens ). The fine localization of white matter PrPres immunolabeled deposits has not yet been elucidated. Several experimental studies have suggested that many types of CNS cells (as well as other cell types) are able to produce PrPsens (2). PrP mRNA has been detected by in situ hybridization in neurones (12), but the possibility of in situ production of PrPsens by normal glial cells (astrocytes and oligodendroglial cells) has also been suggested by Moser et al. (21). The presence of PrPres immunoreactivity in corpus callosum white matter tracts of Syrian hamsters with experimental TSE, however, has been interpreted as a consequence of PrPres production by cortical neurones and secondary axonal transport (26). We wanted to study whether white matter involvement in spongiform encephalopathies is due to direct participation of oligodendroglial cells or to secondary phenomena; we addressed this question by means of light microscopy immunohistochemistry of autopsy CJD material and by TEM analysis of corresponding biopsy specimens.
INTRODUCTION
MATERIAL AND METHODS
Transmissible spongiform encephalopathies (TSE) comprise a group of slow degenerative diseases of the central nervous system (CNS) characterized by microspongiosis (spongiform change), gliosis, neuronal loss, and amyloid plaques (although one or more of these characteristics may be lacking). Attention has been focused mainly on the gray matter changes observed in these conditions. In some instances, however, histopathological involvement of white matter in the cerebral hemisphere and brain stem has been reported
A group of 20 autopsy CJD brains were retrospectively studied with PrP immunohistochemistry. Four cases, aged 45–69 years, were selected for further study based on the following twin criteria: the presence of PrP deposits in white matter and the existence of white matter in biopsy material embedded for TEM. As control, we used archivial material obtained from agematched controls affected with other conditions, e.g., Alzheimer’s disease, and dating back to the same epoch as did the CJD material. 23
0014-4886/98 $25.00 Copyright r 1998 by Academic Press All rights of reproduction in any form reserved.
24
EL HACHIMI ET AL.
OLIGODENDROGLIAL CELLS IN SPONGIFORM ENCEPHALOPATHIES
Biopsy Material Biopsies were performed for diagnostic purposes in the years 1965–1975 as requested by clinical neurologists with the informed consent of the next of kin. Specimens were obtained from the second frontal gyrus of the right hemisphere. Fixation was performed in the operating room a few seconds after the specimen had been taken with a cutting no. 4 curette. The specimens were fixed overnight at 4°C by immersion in 1% glutaraldehyde and 2% paraformaldehyde freshly prepared in 0.1 M phosphate buffer (Karnovsky’s solution). Specimens were then cut into 1-mm pieces, postfixed in 1% osmium tetroxyde for 2 h, dehydrated through a graded series of ethanol, and embedded in Araldite. Semithin sections from each block were performed and were stained with toluidine blue. Ultrathin sections from white matter blocks were counterstained with lead citrate and uranyl acetate and examined with a Siemens 102 electron microscope. Autopsy Material Brains obtained at autopsy were fixed in formaline and slices were embedded in paraffin or celloidin. Multiple blocks for paraffin embedding were taken from cerebellum, frontal and temporal cortex, hippocampus, thalamus, basal ganglia, and medulla oblongata. Blocks embedded in celloidin generally comprised a whole cerebral or cerebellar hemispheric slice. Sections were stained, respectively, with hematein-eosine and modified Masson’s trichrome, and with hemateine eosine, thionin blue (Nissl), Loyez ferric hematoxylin myelin stain, Van Gieson picro-fuchsin, and Mallory’s phophotungstic acid hematoxylin. Immunohistochemistry Seven-micrometer sections from paraffin-embedded blocks were treated according to standard immunostaining methods. We used two rabbit anti-PrP antisera at a dilution of 1/500 (both with 95% formic acid pretreatment for 10 mn). One antiserum was developed by one of us (P.B.), and the second was a gift of Professor Bugiani (6). Antibody reactivity was visualized by the Peroxydase-antiperoxydase (PAP) system (Amersham) and 3,38 diaminobenzidine (DAB) as chromogen. We also used as control antibodies raised against b/A4 amyloid (Dako) at a dilution of 1/400 (with 95% formic
25
acid pretreatment for 10 mn) and against Tau protein (ICN) at a dilution of 1/400. For immunostaining (counterstaining) of cerebellar white matter Purkinje axones, an antibody against Calbindin (CaBP28K) protein at a dilution of 1/30000 (Swant) was used. Specificity of the immunostaining was controlled by incubating sections with Tris buffer, or preimmune serum as a primary antibody, or by absorption of primary antibodies, as well as by applying the full procedures to other material without spongiform encephalopathy. RESULTS
Diagnosis The neuropathological diagnosis of spongiform encephalopathy was confirmed by light microscopy analysis of paraffin and celloidin sections of autopsy material and electron microscopy analysis of biopsy material. They showed spongiosis, neuronal loss and severe astrogliosis of grey matter, but neither severe involvement of white matter nor definite myelinic palor were observed. At the ultrastructural level, membranebound microspongiosis bubbles containing secondary vesicles (‘‘spongiform change’’) were abundant in the cortex. Immunocytochemistry Anti-PrP immunolabeling revealed several types of deposits in white matter (Figs. 1A–1D). Immunolabeling was observed in all the regions studied, namely cerebellar and hemispheric white matter including hippocampus, although not simultaneously in each case. We could distinguish morphologically two types of PrP deposits in white matter. One of them presented as small clumps or elongated deposits arranged parallel to, and in close topographic relationship with, white matter myelinated fibers, especially in cerebellar white matter (Figs. 1B, 1C, and 1E). The second one presented as a ‘‘spider’s web’’-like structure in white matter especially in the cerebral white matter. Immunolabeled clumps appeared to be intracellular; the corresponding cells displayed the morphological characteristics of white matter oligodendroglial cells as seen after formaline fixation and routine embedding (small cells with round or oval eccentric nuclei and clear cytoplasm) (Figs. 1A, 1D, and 1E). No significant glial reaction
FIG. 1. Light microscopy, sections were hematoxylin nuclear counterstain. (A) (3160) Parietal white matter, small patches of PrP immunoreactivity. Note deposit associated to oligodendroglial cell (arrow). (B) (3180) Cerebellum: double immunostaining; anti-PrP reaction was revealed by 5-bromo-4-chloro-3-indolyl-phosphate and 4-nitro blue tetrazolium chloride as chromogen producing a blue deposit; anti-CaBp28K reaction was revealed by 3,38 diaminobenzidine as chromogen, producing a brown deposit, used as counterstaining. PrP deposits are parallel to CaBP-positive fibers. Note PrP deposits in granular layer (GL). (C) (3140) White matter of 2nd parietal gyrus. Arrays of parallel PrP immunostaining material. (D) (3180)) PrP immunoreactivity located in cells with morphological characteristics of oligodendroglial cells. (E) (3160) arrays of parallel PrP immunostaining adjacent to oligodendroglial cell (arrow) associated to clumped immunostaining.
26
OLIGODENDROGLIAL CELLS IN SPONGIFORM ENCEPHALOPATHIES
around the PrP deposits was detected either at the light microscope or at the electron microscope levels. No immunolabeling with anti-PrP was observed in controls, conversely no immunolabeling with anti-b amyloid or anti-Tau was observed in the CJD specimens. Electron Microscopy Identification of oligodendroglial cells, as distinct from microglial cells or astrocytes, was established by morphological criteria (23). Their rounded or ovalshaped nucleus showed an electron density greater than that of astrocytes and contained a clumped heterochromatin; it was generally eccentric and surrounded by only a thin rim of cytoplasm, except at one pole of the cell where a larger cytoplasmic expansion was found. The cytoplasm was electron dense with rare glycogen granules; processes contained microtubules. The general appearance of TSE oligodendroglial cells was moderately abnormal, with a dilated endoplasmic reticulum and cytoplasm containing dense membranebound lysosomes in greater abundance and/or size than in controls, without any evidence of macrophagic activity. The most striking feature was the presence of abnormal inclusions in these cells and in their processes. Two types of inclusions could be distinguished according to their osmophilia: the less osmiophilic inclusions (type I) were made of amorphous and fibrillar material (Figs. 2A, 2B, and 3C); the more osmiophilic ones (type II) were made of compact material (Figs. 3A and 3B), resolved at high magnification as finely granular (Fig. 3B). Both materials sometimes coexisted in the same inclusion (mixed type) (Figs. 3A and 3B). The fibrillar material was either loose or arranged in a ‘‘fingerprint’’ structure (Fig. 3C). The fibrils were 5 to 10 nm in diameter. The fibrillarparacrystalline inclusions did not display any special arrangement, in contrast to osmiophilic inclusions, which appeared generally star-like. These inclusions were not membrane-bound. Their structure was quite distinct from the lysosome-like dark structures observed in controls. Myeline sheaths frequently showed vacuoles, disruption, and occasional phagocytosis of myelin debris; neurites, however, were generally normal. DISCUSSION
White matter has been generally considered to be unaffected in CJD, although its involvement has been occasionally reported (20, 22, 25, 27) as ‘‘myelinic palor’’ or myelin degeneration associated with phagocytosis
27
and gliosis. White matter involvement has been considered to be most frequent in Japanese cases of CJD (20). Moderate degeneration of cerebral white matter or of fibers tracts in the spinal cord has been reported in most cases of Kuru (9) and was a feature of the original case of GSS (10). The question is whether, when present, white matter involvement is a primary feature (20) or a secondary one (5, 22). The fine localization of PrP deposits might be a clue towards the solution of this problem through the identification of the primary lesion, inasmuch PrPres is widely held to be directly involved in TSE pathogenesis (4, 13). PrP deposits are potentially detectable in most, if not all, cases of TSE, if sensitive immunohistochemistry procedures are used (3, 14, 15). Various types of PrP deposits have been described; probably all regions of the CNS may host one or another type of deposit. White matter deposits seem to be an uncommon feature of ‘‘European’’ CJD, as they were present in only 4 of our 20 cases. These PrPres deposits were generally not correlated with white matter neuropathological lesions as defined by classic histological methods. The deposits did not display the specific tinctorial and physicochemical properties of amyloid fibrils: they were neither birefringent nor congophilic. White matter deposits have been regarded (26) as produced by neurones, implying that the products are axonally transported; other explanations are possible, because glial cells have also been shown to be able to produce PrP (6, 21). The fine localization of PrP, however, cannot be completely determined with light microscope methods. The fact that immunolabeling of PrPres at the ultrastructural level could not be achieved with postembedding methods (the only ones applicable to archivial material), either by ourselves or by previous authors, makes impossible direct TEM determination of immunoreactive sites; as a consequence, we are restricted to comparison with standard TEM images. Electron microscopy of white matter, however, has attracted little interest in human TSE (8, 10). A few ultrastructural abnormalities have been described, such as myelin sheath disruption and intramyelin vacuolization (19), but these were for years dismissed as artefacts due to fixation or embedding procedures. The ultrastructural pathology of animals with experimental TSE, which provides optimal conditions of fixation and embedding, has been extensively studied, particularly by Liberski et al. (16, 17). These studies have shown that, in contrast with the predominantly gray matter pathology produced by most CJD strains, widespread myelin and axonal pathology are produced by the Fujisaki CJD and
FIG. 2. Electron microscopy. (A) (320,000) White matter oligodendrocyte shows fibrillar-paracrystalline (type I) inclusion (I). Note the enlargement of rough and smouth ergastoplasmic saccules. (B) (369,500) High magnification of the same oligodendroglial process with inclusion constituted by fibrillar and amorphous material. The inclusion is not membrane bound.
FIG. 3. Electron microscopy. (A) (318,000) Two white matter oligodendroglial cells showing two inclusions (type II and mixed). One inclusion contains only dense osmiophil material (II) the second one is made of more or less osmiophilic materials (mixed inclusion) (M). Note the presence of type II inclusion in a cell process (arrow). (B) (334,000) mixed inclusion in cell process, the dense, osmiophilic material is granular (arrow). (C) (337,500) Type I inclusion in cell process; note the paracrystaline, often fingerprint-like, texture of the material (arrows).
28
OLIGODENDROGLIAL CELLS IN SPONGIFORM ENCEPHALOPATHIES
263K scrapie strains. White matter lesions included myelin sheath damage or vacuolation, phagocytosis of myelin debris, and neuroaxonal dystrophy; myelin damage has been considered to be related to tumor necrosis factor; no significant alteration of oligodendroglial cells has been reported (18). On the other hand, four findings concur with our conclusion that oligodendroglial cells are primarily affected in our material. First, they show slight, but consistent general cytoplasmic changes. Second, they contain fibrous inclusions; although comparison of this material with Scrapie Associated Fibrils (SAF) is tempting, it cannot directly be done, the morphology of SAF being largely determined by the extraction procedure. Third, these inclusions are not seen in controls, which display only the classical lipofuscin-like inclusions. Fourth, there is a general concordance between the presence of these inclusions and the sites of PrP immunoreactivity; although the fine localization of the latter could not be determined (at the light microscopy level), their localization in the vicinity of oligodendroglia-like nuclei and the absence of any significant microglial reaction around these deposits (in contrast with Kuru plaques or with AD senile plaques, which are often associated with microglial cells or astrocytes (1, 7)) is consistent with an intracellular, and more precisely oligodendroglial, localization of white matter PrP deposits. Negatively, the absence of significant axone damage speaks again of primitive neuronal involvement in TSE white matter changes. Our interpretation of the abnormal inclusions seen in oligodendroglial cells as intracellular PrPres deposits is in consequence founded on four arguments, none of which would be convincing if taken alone, but which, taken together, amount to a prima facie case. Although myelin lesions were minimal in our samples, we feel that our conclusions can be extrapolated to the rarer instances with obvious white matter damage. In conclusion, we found modifications in oligodendrocytes, but not in axones, in the white matter of CJD brains. This result may indicate the primary nature of white matter involvement that occurs in some instances of that condition.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13. 14.
ACKNOWLEDGMENTS We are grateful to Dr. O. Bugiani for the generous gift of anti-PrP antisera. We thank Dr. P. Liberski for review of the manuscript. D. Lecren is thanked for skillful iconographic assistance. This work has been supported in part by l’Action Concerte´e Coordonne´e du Ministe`re de la Recherche, by La Fondation de France (grant to K. H. E.), and aided by European Union Biomed 2 Concerted Action ‘‘The human prion diseases’’ (Project Leader: H. Budka).
15.
16.
17.
REFERENCES 1.
Barcikowska, M., P. P. Liberski, J. W. Boellaard, P. Brown, D. C. Gajdusek, and H. Budka. 1993. Microglia is a component of the
18.
29
prion protein amyloid plaque in the Gerstmann-Stra¨usslerScheinker syndrome. Acta Neuropathol. 85: 623–627. Bendheim, P. E., H. R. Brown, R. D. Rudelli, L. J. Scala, N. L. Goller, G. Y. Wen, R. J. Kascsak, N. R. Cashman, and D. C. Bolton. 1992. Nearly ubiquitous tissue distribution of the scrapie agent precursor protein. Neurology 42: 149–156. Bruce, M. E., P. A. McBride, and C. A. Farquhar. 1989. Precise targeting of the pathology of the sialoglycoprotein, PrP, and vacuolar degeneration in mouse Scrapie. Neurosci. Lett. 102: 1–6. Budka, H., A. Aguzzi, P. Brown, J. M. Brucher, O. Bugiani, F. Gullota, M. Haltia, J-J. Hauw, J. W. Ironside, K. Jellinger, H. A. Krezschmar, P. L. Lantos, C. Masullo, W. Schlote, J. Tateishi, and R. O. Weller. 1995. Neuropathological diagnostic criteria for Creutzfeldt-Jakob disease (CJD) and other human spongiform encephalopathies (Prion diseases). Brain Pathol. 5: 459–466. Bugiani, O., F. Tagliavini, G. Giaccone, and R. Boeri. 1986. Creutzfeldt-Jakob disease: Astrocytosis and spongiform changes of the white matter. In Unconventional Virus Diseases of the Central Nervous System (L. A. Court, D. Dormont, P. Brown, and D. T. Kinsbury, Eds.), pp. 172–183. Diedrich, J. F., P. E. Bendheim, Y. S. Kim, R. I. Carp, and A. T. Haase. 1991. Scrapie-associated prion protein accumulates in astrocytes during Scrapie infection. Proc. Natl. Acad. Sci. USA 88: 375–379. El Hachimi, K. H., and J-F. Foncin. 1994. Do microglial cells phagocyte the b/A4-amyloid senile plaque core of Alzheimer disease? C. R. Acad. Sc. Se´rie III 317: 445–451. Foncin, J-F., J. Gaches, and J. Le Beau. 1964. Ence´phalite spongiforme (apparente´e a` la maladie de Creutzfeldt-Jakob). Biopsie e´tudie´e au microscope e´lectronique, confirmation autopsique. Rev. Neurol. 111: 507–515. Gajdusek, D. C., and C. J. Gibbs. 1978. Unconventional viruses causing the spongiform virus encephalopathies. A fruitless search for the coat and core. In Viruses and Environment. (E. Kurkas and K. Maramovosch, Eds.), pp. 70–98, Academic Press, New York. ¨ ber eine Gerstmann, J., E. Stra¨ussler, and I. Scheinker. 1936. U eigenartige heredita¨r-familia¨re Erkrankung des Zentralnervensystems. Zugleich ein Beitrag zur Frage des vorzeitigen localen Alterns. Z. Ges. Neurol. Psychiatr. 154: 736–762. Gonatas, N. C., R. D. Terry, and M. Weiss. 1965. Electron microscopy study in two cases of Jakob-Creutzfeldt disease. J. Neuropathol. Exp. Neurol. 24: 575–598. Harris, D. A., P. Lele, and W. D. Snider. 1993. Localization of the mRNA for a chicken prion protein by in situ hybridization. Proc. Natl. Acad. Sci. USA 90: 4309–4313. Ironside, J. W. 1996. Review: Creutzfeldt-Jakob disease. Brain Pathol 6: 379–388. Jeffrey, M., C. M. Goodsir, M. E. Bruce, P. A. McBride, J. R. Scott, and W. G. Halliday. 1992. Infection specific prion protein (PrP) accumulates on neuronal plasmalemma in Scrapie infected mice. Neurosci. Lett. 147: 106–109. Kitamoto, T., K. Doh-Ura, T. Muramoto, M. Miyazono, and J. Tateishi. 1992. The primary structure of the prion protein influences the distribution of abnormal prion protein in the central nervous system. Am. J. Pathol. 141: 271–277. Liberski, P. P., R. Yanagihara, C. J. Gibbs, and D. C. Gajdusek. 1989. White matter ultrastructural pathology of experimental Creutzfeldt-Jakob disease in mice. Acta Neuropathol. 79: 1–9. Liberski, P. P., R. Yanagihara, D. M. Asher, C. J. Gibbs, and D. C. Gajdusek. 1990. Reevaluation of the ultrastructural pathology of experimental Creutzfeldt-Jakob disease. Brain 113: 121–137. Liberski, P. P., R. Yanagihara, V. R. Nerurkar, and D. C. Gajdusek. 1993. Tumor Necrosis ~Factor produces Creutzfeldt-
30
EL HACHIMI ET AL.
Jakob disease-like lesions in vivo. Neurodegeneration 2: 215– 225. 19. Masters, C. L., and D. C. Gajdusek. 1982. The spectrum of Creutzfeldt-Jakob disease and the virus-induced subacute spongiform encephalopathies. In Recent Advances in Neuropathology (W. T. Smith, and J. B. Cavanagh, Eds.), pp. 139–163, Churchill Livingstone, Edinbourg. 20. Mizutani, T., A. Okumara, M. Oda, and H. Shiraki. 1982. Panencephalic type of Creutzfeldt-Jakob disease: Primary involvement of white matter. J. Neurol. Neurosurg. Psychiatr. 44: 103–115. 21. Moser, M., R. J. Collelo, U. Pott, B. Oesch. 1995. Developmental expression of the prion protein gene in glial cells. Neuron 14: 509–517. 22. Park, T. S., G. M. Kleiman, and E. P. Richardson. 1980. Creutzfeldt-Jakob disease with extensive degeneration of white matter. Acta Neuropathol. 52: 239–242.
23. 24.
25.
26.
27.
Peters, A., S. L. Palay, and H. de F. Webster. 1976. The Fine Structure of Nervous System, Saunders, Philadelphia. Power, J. M., Y. Liu, L. Hair, R. J. Kascsack, L. D. Lewis, and L. A. Levy. 1991. Concomitant Creutzfeldt-Jakob disease and Alzheimer’s disease. Acta Neuropathol. 83: 95–98. Su, M., S. Wang, H. Chiang, J. Wang, Y. Tan, J. Fuh, and H. Liu. 1993. Creutzfeldt-Jakob disease: report of case of the pancenphalopathic type and experience of a teaching hospital in Taiwan. Eur. Neurol. 33: 241–24. Taraboulos, A., K. Jendroska, D. Serban, S. L. Yang, S. J. DeArmond, and S. B. Prusiner. 1992. Regional mapping of prions proteins in brain. Proc. Natl. Acad. Sci. USA 89: 7620– 7624. Vallat, J. M., M. Dumas, N. Corvisier, M. J. Lebutet, A. Loubet, P. Dumas, and F. Cathala. 1983. Familial Creutzfeldt-Jakob disease with extensive degeneration of white matter. Ultrastructural study of peripheral nerves. J. Neurol. Sci. 61: 261–275.