In a Series of Autoradiographic Experimentsemploying

Home , Cerebral hemisphere, Fascia dentata

SELECTIVE, ACUTE DESTRUCTION OF GRANULE-CELL PRECURSORS IN THE DENTATE GYRUS BY PROTON IRRADIATION

LIGHT AND ELECTRON MICROSCOPIC STUDY

ROBERT H. BROWNSON, PH.D.,* AND LARRY W. MCDONALD, M.D. From the Division of Medical Physics, Donner Laboratory, University of California, Berkeley, Calif. Undifferentiated, maturing cells in general are considered to be highly sensitive to ionizing radiation. From this conclusion, Altman 1 suggested that a population of undifferentiated granule cells of the mammalian nervous system, many of which are formed after birth in various regions of the brain (at least in the rat), would be selectively damaged in maturing animals exposed to ionizing radiation. Altman and Das 2,3 have shown, in a series of autoradiographic experiments employing tritiated thymidine, that labeled brain cells (neurons and neuroglia) underwent mi- gration, multiplication, and differentiation when traced in serial sections through the neonatal brain. They further pointed out that small neurons, designated as microneurons, form discrete granule-cell layers in a variety of brain regions in rodents, notably in the cerebellar cortex, the olfactory bulb, and the dentate fascia of the hippocampus. The dentate fascia, to which this study is directed, derives its name from the tooth-like configuration of granular and molecular layers and the hilus which is embraced by them within the hippocampus. One of the most striking features in this area is the closely packed multilayer of large oval or short pyramidal cells.4 These neurons in the rat brain are believed to be derived partly from cells multiplying at a rapid rate after birth in the ependymal and subependymal layers of the forebrain ventricles and, to a lesser extent, from cells multiplying regionally in the internal molecular layer of the dentate gyrus. From these 2 regions the newly formed cells apparently migrate into the granule-cell layer and become fully differentiated there.1"3 The time sequence for precursor granule cells (microneurons of Altman) to differentiate into adult gran-

Supported by Grant NBOI289 from the National Institutes of Health, U. S. Public Health Service, and the Atomic Energy Commission. Presented at a meeting of the American Association of Anatomists, San Francsco, Calif., Apr. 6-8, I966. Accepted for publication Dec. 22, I966. * National Institutes of Health Special Postdoctoral Fellow 2FII NBr362 NSRB. Present address: Department of Anatomy, Medical College of Virginia, Richmond, Va. 927 928 BROWNSON AND MC DONALD Vol. 50, No. 5 ule cells in the dentate fascia of the hippocampus in rat brain extends well beyond the usual maturation time for neurons. During this prolonged period the precursor cells have been shown actively to incorporate thymidine, presumably during the synthetic stage of DNA replication. It would not be presumptuous, therefore, to expect that this particular cell would have a similarly prolonged period of increased radiosensitivity. In order to understand better the role played by granule cells in the cell turnover and maturation of the rodent hippocampus, we must first establish certain fine-structure details concerning the precursor granule cells (microneurons) both in normal and pathologic reactions. For this purpose we have selectively sought to induce acute degenerative reactions among the precursor granule cells during differentiation and/or multiplication by means of ionizing radiation. The absorbed tissue dose and postirradiation time sequence are based on a previous study which showed an intense selective destruction of cells in this area as early as 6 hr. after exposing young rat brain to 5000 R or less of X-irradiation.5 METHODS Six i8-day-old male rats were anesthetized and given a single 5-sec. exposure to 4.5 krad of cyclotron-accelerated 55-Mev protons limited to the head. The beam was directed to one cerebral hemisphere, using the plateau of the Bragg curve. Two irradiated animals and I nonirradiated animal were sacrificed at I.5, 3, and 6.5 hr. post exposure. Each animal was anesthetized by an intraperitoneal injection of pentobarbital, 0.047 mg./gm. of body weight, given a mixture of O.I5 ml. of I :iooo heparin and i% sodium nitrite by intracardiac injection, and sacrificed by perfusion- fixation with 3% redistilled glutaraldehyde buffered with o.i M sodium cacodylate. Tissue was removed from the dentate fascia of the dorsal hippocampi of irradiated, nonirradiated (shielded), and control hemispheres with the aid of a low-power dis- secting microscope. The tissues for electron microscopy were postfixed in osmium tetroxide and embedded in Epon0. The tissues for light microscopy were postfixed in neutral buffered formalin and embedded in paraffin. It is recognized that a dose of 4.5 krad is excessive and a lower dose would probably produce the desired effect among actively differentiating, radiosensitive cells such as those in this study. However, in order to be reasonably optimistic about locating the small precursor granule cells undergoing degeneration, we attempted to involve maximum numbers and thereby increase the odds for a successful recovery in ultrathin sections of the dentate fascia area. RESULTS Light microscopy on thick-sectioned, Epon-embedded blocks and con- ventionally sectioned paraffin-embedded blocks revealed that the granule cells of the dentate gyrus were undergoing morphologic changes as early as 3 hr. post exposure to 4.5 krad proton irradiation (Fig. i and 2). In ultrathin sections examined by electron microscopy, cellular alterations were noted at I.5 hr. (Fig. 3-5). These changes were found to be pro- gressive, terminating in what appeared to be irreversible pyknosis and May I967 GRANULE CELL RADIOSENSITIVITY 929 cytoplasmic degeneration in 3-6.5 hr. post exposure (Fig. 6 and 7). The pyknotic cells were concentrated chiefly at the inferior or internal border of the granule cells at the polymorph junction. There also appeared to be a number of large, irregularly shaped, darkly stained cells present in this area by i.5 hr. post exposure. Such cells had large round nuclei and generous, polymorphic cytoplasm, rich in ribosomes and concentric ar- rays of rough-surfaced endoplasmic reticulum. Although similar cells were observed in controls, they were far fewer in number. By 3 hr. following irradiation, one of the earliest cell changes involved the nucleus, which appeared to shrink and increase in density, beginning at the nuclear membrane. The cytoplasm, in contrast to the nucleus, became dis- organized and increasingly electron-lucid. Densely stained tissue fragments and enlarged clear spaces were observed among the degenerating cells (Fig. 6). At 6.5 hr. post exposure the intensity of cellular necrosis appeared to have reached a plateau. Large round osmiophilic globules, resembling pyknotic nuclei in variable dense masses of protoplasm, were abundant. Granule cells bordering the necrotic zone were frequently im- plicated in partial cellular changes notably involving liquefaction of the cytoplasm and decrease in nuclear density (Fig. 7). Granule cells distal to the internal necrotic layer of cells bordering the hilus were conspicu- ously normal in appearance. In addition to cell necrosis, a fair number of degenerating cells or fragments appeared to be undergoing phagocytosis as early as I.5 hr. post exposure. In each instance the cell or its remnant was seen to be engulfed by a watery-like cytoplasm. The nucleus of the engulfing cell was generally oval and moderate in chromatin; and when fibrils were present in the cytoplasm, as they frequently were, a positive identification of astrocyte was made (Fig. 4 and 5). The granule cells in the dentate gyrus of the nonirradiated, shielded hemisphere and control brains appeared entirely normal. The absence of necrotic cells in the dentate gyrus of the shielded half of the brain indicated that there was no effective cross scatter of secondary ionizing irradiation or transfer of secondary effects from the irradiated hemisphere. DISCUSSION The hippocampal area has long been known for its low tolerance to a variety of noxious insults such as asphyxia, encephalitis, and whole- head ionizing irradiation.>8 More recently Schoenbrun, Campeau, and Adey9 have demonstrated changes in motor behavior and brain electrical activity in cats 30-60 min. following either unilateral or bilateral focal- beam (8 mm.) exposure of the dorsal hippocampus to as little as 400- 500 R of 250 kvp X-rays. Hicks and Montgomery10 first called attention to the acute localized destruction of the granule cells in the dentate 930 BROWNSON AND MC DONALD Vol. 50, No. 5 gyrus of mice and rats after I50, 250, and 400 kvp X-rays, 2.5 Mev electrons, and 3 Mev X-rays acutely following high dose and high-dose- rate studies. This phenomenon of acute granule cell destruction was again observed by Brownson, Suter, and Diller 5in young rat brain exposed to 500-IO,OOO R X-rays. However, in contrast to the findings of Hicks and Montgomery,10 it was noted that the granule-cell destruction was limited entirely to a narrow margin of cells bordering the hilus of the dentate fascia. Although the degenerating cells were incorrectly identified at that time as neuroglial elements, we now believe that the differing accounts in degree of destruction in these 2 reports of granule-cell radio- necrosis6"10 were probably related to differences in dose, dose rates, and age which were of a different magnitude in the earlier study.'0 A survey of the literature relating to the morphology of the hippocampus and, in particular, the dentate fascia revealed surprisingly few papers on normal or abnormal ultrastructure, particularly in comparison to the large number of electrophysiological or behavioral studies involving this area of the brain. Studies by Blackstad"1 and Blackstad and Dahl 12 were perhaps the first comprehensive coverage of the quantitative analysis of normal granule-cell ultrastructure. In these reports the authors noted 2 features pertinent to this investigation-namely, that there were virtually no oligodendroglia present among granule cells and that a very high percentage of the granule-cell soma was invested in astroglial cytoplasm. The necrotic cells observed in this study were specifically found at the granule cell-polymorph junction, an area which has previously been shown to contain immature and undifferentiated cells."3 It has also been shown that the polymorph area contains at least 2 other cell types in addition to cell processes-that is, basket cells and modified pyramidal cells.13 However, there is no evidence indicating that these basket cells or pyramidal cells normally cluster at the granule-cell border or that they are selectively radiosensitive immature cells. For these reasons, therefore, it appears unlikely that they would contribute significantly to this reaction. It also seems unlikely that immature precursor oligodendroglia as granule-cell satellites would have been involved, primarily since they are not normally found as satellites to granule cells.'2 The comparatively small numbers of microglia among neuroglia 14"15 of the CNS and their apparent acute resistance to moderate doses of ionizing radiation 16 would seem to minimize their likelihood of major involvement. Although the astroglia have been shown to contribute heavily to inter-granule-cell neuropil,1""12 the astrocyte has been shown to become progressively more active and hypertrophic as a result of moderate exposure to ionizing radiation.'S'8 Furthermore, it was noted that some of the astrocytes observed in this study were actively engulfing debris-an acute phenom- May 1967 GRANULE CELL RADIOSENSITIVITY 93I enon which we have previously reported in rat brain following a-particle irradiation,19 as have others more recently in various types of tissue reactions.20,21 Finally, it is highly unlikely that any concentrated numbers of exogenous mononuclear cells would have infiltrated this area and subse- quently undergone degeneration during this time period.22 Although this study provides no direct proof that the TH3-labeled undifferentiated rat brain granule cells (microneurons) observed in dentate gyrus by others ' are necessarily identical with the necrotic cells shown here, it appears highly suggestive that they may be, at least in part, related. And if indeed they are identical, it then follows that a selective destruction of highly radiosensitive precursor granule cells (microneurons) would effect the time and course of development of the dentate gyrus of the hippocampus. Such a loss or replacement failure of the normal histogenesis of granule cells in the dentate gyrus could influence the normal function of the hippocampus per se. SUMMARY At I.5 hr. post exposure to 4.5 krad of 55 Mev protons, darkly stained pleomorphic cells have been observed massed along the internal layer of granule cells at the hilus of the dentate gyrus of the hippocampus. Occa- sionally this type of cell was seen to be partially or completely engulfed by a clear cytoplasm frequently identified as astrocyte. By 3 hr. post exposure the internal layer of granule cells had entered into a marked degenerative phase. Degenerating dark pleomorphic cells were intermixed among granule cells, some of which had undergone nuclear margination and cytoplasmic disorganization. Pyknotic nuclei, cell fragments, and enlarged intracellular and extracellular spaces were prominent. Fre- quently astrocytes were observed engulfing degenerating cells or cell fragments. Through the 6.5-hr. postexposure period there appeared to be some reduction in severity of degeneration and over-all containment of reaction along the internal layer of granule cells. Throughout the period under investigation, the irradiated granule cells distal to the polymorph layer appeared morphologically normal. From these findings, it is suggested that the cells demonstrated to undergo early radionecrosis when exposed to ionizing radiation are principally immature, precursor granule cells (microneurons) which normally differentiate into adult neurons and lose their radiosensitivity as they migrate toward the molecular layer. REFERENCES i. ALTMAN, J. The selective destruction of microneurons by low-dose irradiation. (Personal communication) 2. ALTMAN, J., and DAS, G. D. Autoradiographic and histological evidence of 932 BROWNSON AND MC DONALD Vol. so, No.5 postnatal hippocampal neurogenesis in rats. J Comp Neurol Z24:3I9-336, I965. 3. ALTMAN, J., and DAS, G. D. Post-natal origin of microneurons in the rat brain. Nature (London) 207:953-956, I965. 4. CROSBY, E. C., HUMPHREY, T., and LAUER, E. W. Correlative Anatomy of the Nervous System. Macmillan, New York, I962. S. BRowNsoN, R. H., SUTER, D. B., and DILLER, D. A. Acute brain damage induced by low dosage x-irradiation. Neurology z3:I8I-I9I, I963. 6. SPIELMEYER, W. Die Pathogenese des epileptischen Krampfes. Z Neurol Psy- chiat I09:50I-520, I927. 7. MEYrER, A., and BECK, E. The hippocampal formation in temporal lobe epilepsy. Proc Roy Soc Med 48:457-462, 1955. 8. GANGLOPP, H., and HAEY, T. J. Effects of x-irradiation upon spontaneous and evoked brain electrical activity in cats. Radiat Res Z2:694-704, I960. 9. SCHOENBRUN, R. L., CAMPEAU, E., and ADEY, W. R. "Electroencephalo- graphic and Behavioral Effects from X-irradiation of the Hippocampal System." In Response of the Nervous System to Ionizing Radiation, Haley, T. J., and Snider, R. S., eds. Little, Boston, I964, PP. 59i-620. i0. HICKS, S. P., and MONTGOMERY, P. O'B. Effects of acute radiation on the adult mammalian central nervous system. Proc Soc Exp Biol Med 8o:iS-i8, I952. II. BLACKSTAD, T. W. A note on the electron microscopy of the fascia dentata. Acta Morph Neeri Scand 3:395-404, I96I. 12. BLACKSTAD, T. W., and DAHL, H. A. Quantitative evaluation of structures in contact with neuronal somata. An electron microscopic study on the fascia dentata of the rat. Acta Morph Neerl Scand 4:329-343, i962. I3. RAm6N Y CAJAL, S. Histologie du Systame Nerveux de l'Homme et des Vertibrds. Maloine, Paris, I909, 993 PP. 14. GLEES, P. Neuroglia Morphology and Function. Thomas, Springfield, Ill. I955, IIZI PP. IS. BROWNSON, R. H. Perineuronal satellite cells in the motor cortex of aging brains. J Neuropath Exp Neurol z4:424-432, 1955. I6. MAXWELL, D. S., and KRUGER, L. "Electron Microscopy of Radiation-Induced Laminar Lesions in the Cerebral Cortex of the Rat." In Response of the Nervous System to Ionizing Radiation, HALEY, T. J., and SNIDER, R. S., Eds. Little, Boston, I964, PP. 54-83. 17. KLATZO, I., MIQUEL, J., TOBIAs, C., and HAYMAKER, W. Effects of alpha particle radiation on the rat brain, including vascular permeability and gly- cogen studies. J Neuropath Exp Neurol 20:459-483, I96I. i8. BROWNSoN, R. H., SUE, D. B., OIVER, J. L., INGERSOLL, E. H., and BURT, D. H. "Histochemical and Histological Changes Induced in Rat Brain by X-irradiation." In Response of the Nervous System to Ionizing Radiation, HALEY, T. J., and SNIDER, R. S., Eds. Little, Boston, I964, PP. 307-33S. 19. BROWNSON, R. H., and McDoNALD, L. W. "Alterations in the Fine Structure of Rat Brain Following Heavy-Particle Radiation." In Proceedings of the Fifth International Congress of Neuropathology. International Congress Series IOO, Excerpta Medica Foundation, New York, I966, PP. 780-783. 20. LAMPERT, P., and CRESSMAN, M. Fine structural changes of myelin sheaths after axonal degeneration in the spinal cord. (Abst.) J Neuropath Exp Neurol 26:I56, i967. May 1967 GRANULE CELL RADIOSENSITIVITY 933

2I. LAATSCH, R. H., and COWAN, W. M. Degeneration of commisural afferents to the dentate gyrus of the rat, studied with the electron microscope. (Abst.) Anat Rec I54:371, I966. 22. BROWNSON, R. H. Proton-irradiation and early mononuclear cell response in brain. (Abst.) Radiat Res 27:5I0, i966. The authors wish to express their indebtedness to Dr. Webb Haymaker, Ames Research Center, NASA, Moffett Field, Calif., and Dr. Cornelius Tobias, Donner Laboratory, Uni- versity of California, Berkeley, Calif., for their advice and criticism in the preparation of this manuscript. An expression of gratitude is due Mrs. Ethel Lovings, Department of Anat- omy, Medical College of Virginia, Richmond, Va., Mr. Michael Donovan and the 88-in. cyclotron crew, Lawrence Radiation Laboratory, University of California, Berkeley, Calif., for their technical assistance. ADDENDUM An additional report on the normal ultrastructure of the cells and processes in the subdivisions of rat dentate gyrus, with special reference to synaptic organization (LAATSCH, R. H., and COWAN, W. M. J Comp Neurol 128:359-396, I966), came to our attention after submission of this manuscript. The failure of these authors to describe the dark cell or granule-cell precursor may be attributed to the fact that this cell type is relatively scarce in adult rats weighing between i8o and 300 gin., which were used in their study.

[ Illustrations follow] 934 BROWNSON AND MC DONALD Vol. 50, No. 5

LEGENDS FOR FIGURES FIG. i. Epon-embedded, thick-sectioned fascia dentata. A. Control, showing granule cells (G), pyramidal cells (P), molecular layer (M), and polymorph layer (Pm). B. At i.S hr. post exposure. C. At 3 hr. post exposure. Note aggregation of darkly stained cells (arrows) at border of granule cells and polymorph layer. D. At 6.5 hr. post exposure. Osmium tetroxide stain. X I90. May I967 GRANULE CELL RADIOSENSITIVITY 935

ID 936 BROWNSON AND MC DONALD Vol. 50, No. 5

FIG. 2. Fascia dentata from opposite hemispheres of the same brain. A. Shielded hemisphere, showing normal structure; molecular layer (M), granule-cell layer (G), and polymorph layer (Pm). Pleomorphic cells (P1) are among granule cells of internal layers. B. Nonshielded, 4.5-krad proton-irradiated hemisphere 6.5 hr. post exposure. There is compact localized arrangement of pyknotic cells and clear spaces at junction of granule cells with polymorph layer. Note loss of pleomorphic cells seen in shielded hemisphere. Periodic acid-Schiff and hematoxylin stain. X 505. May I967 GRANULE CELL RADIOSENSITIVITY 937

r r:J' M~N--f ,'A,=

$.~-~::' ....:.S*' :.:'j

2A Uh.. AO . sM. 10 An, a

2B 938 BROWNSON AND MC DONALD Vol. 5o, No. 5

FIG. 3. Granule cells in dentate gyrus of nonirradiated control, showing granule- cell nucleus (GRn) and cytoplasm, and astrocyte nucleus (ASn). Enlarged clear areas are either astroglial processes or extracellular spaces. Uranyl acetate and lead citrate stain. X 5400. FIG. 4. Polymorph layer of dentate gyrus i.5 hr. post proton irradiation, showing astrocyte nucleus (ASn) and watery cytoplasm (Cyt) containing fibrils (Fi), rough-surfaced endoplasmic reticulum, and numerous vesicles (Ve); and pleomorphic cell nucleus (Pln) and densely stained cytoplasm containing mito- chondria. Latter cell is partially engulfed by astroglial cytoplasm. Uranyl acetate and lead citrate stain. X i8,000. May I967 GRANULE CELL RADIOSENSITIVITY 939

4 940 BROWNSON AND MC DONALD Vol. 50, No. 5

FIG. 5. Polymorph layer of dentate gyrus, i.5 hr. post proton irradiation, showing astrocyte nucleus (ASn) and watery cytoplasm containing fibrils (Fi), mito- chondria, and vesicles. Pleomorphic cell nucleus (Pln) is surrounded by narrow band of densely stained cytoplasm undergoing degeneration and partially engulfed by astroglial cytoplasm. Uranyl acetate and lead citrate stain. X I5,400. FIG. 6. Granule cells (GR) and pleomorphic cell (P1) at border of polymorph layer 3 hr. post proton irradiation. Note various stages of degeneration from increasing nuclear and cytoplasmic density to irreversible cell necrosis with accumulations of densely stained cellular debris. Mature granule cells (GR) and myelinated axons appear to be reasonably intact. Uranyl acetate and lead citrate stain. X 5000. May I967 GRANULE CELL RADIOSENSITIVITY 94I

6 942 BROWNSON AND MC DONALD Vol. 50, No. 5

.j.y ".9.A 7

I j1- :. d

., A.. *

FIG. 7. Granule cells at border of polymorph layer 6.5 hr. post proton irradiation. Granule-cell nucleus (GRn) appears intact but lightly stained. Portions of cytoplasm are nearly electron lucid (arrow) while other areas of cell appear normal. Round, densely stained sphere (arrow) resembles late stage of pyknosis. Uranyl acetate and lead citrate stain. X I3,000.