Brain Research, 148 (1978) 295-311 © Elsevier/North-HollandBiomedicalPress
295
RADIAL GLIA IN THE HUMAN FETAL CEREBRUM: A COMBINED GOLGI, IMMUNOFLUORESCENT AND ELECTRON MICROSCOPIC STUDY*
BEN H. CHOI and LOWELL W. LAPHAM Department of Pathology, Neuropathology Division, University of Rochester Medical Center, Rochester, N.Y. (U.S.A.)
(Accepted October 13th, 1977)
SUMMARY Golgi techniques, immunofluorescence for glial fibrillary acidic (GFA) protein, and electron microscopy (EM) were used to determine the nature of radial glia in the cerebrum of human fetuses ranging from 7 to 20 weeks of ovulation age. Successful Golgi impregnation of radial fibers was achieved in fetuses 12 weeks of age and older. These fibers spanned the entire thickness of the hemisphere. At the pial surface many of them branched and terminated in pyramidal end feet expansions. Indirect immunofluorescent preparations utilizing antiserum to GFA protein, a protein specific for astrocytes, demonstrated numerous radially oriented nearly parallel fluorescent fibres between the ventricular zone and pia mater. GFA protein-positive fibers were demonstrated in all fetal specimens examined with this technique (10 weeks of age and older). Along the outer border of the marginal zone they formed a horizontal GFA proteincontaining subpial membrane. By EM there were numerous linear electron lucent astrocytic processes containing 8-9 nm filaments and occasional glycogen granules at all levels of the cerebrum. They were interspersed among smaller and darker neuronal processes containing 20-25 nm neurotubules, and were demonstrable at all fetal ages between 7 and 18 weeks. They formed pericapillary investments and subpial terminal expansions closely abutting basal lamina of pia mater in every specimen examined. On the basis of these combined analyses, we conclude that radial glial fibers in early human fetal cerebrum represent processes of immature astrocytes. Although subsequently undergoing further maturation, radial glia already possess fundamental immunocytochemical and morphological characteristics indicative of astrocytic differentiation. A significant implication of our findings is that the development of astrocytes in the human fetal brain occurs much earlier than formerly believed.
* Presented in part at the 52nd Annual Meetingof the AmericanAssociationof Neuropathologists, San Francisco, California,June 11-13, 1976.
296 INTRODUCTION Since the introduction of the Golgi technique the existence of long, radially oriented fibers spanning the entire thickness of the developing fetal cerebrum has been recognized in many vertebrate species including man. These fibers are often referred to as radial glia, although they have also been designated by other names such as Golgi epithelial cells ~4, spongioblasts 16,19,'~6,97 or matrix cells 1~. Definitive determination of their nature, cell of origin and development has been beset by many difficulties, in considerable part related to the difficulty in distinguishing between the various cellular elements observed during early stages of brain development. Work in this area has also been heavily influenced by the concept that glial cells develop only after the majority of neurons have formed and migrated 11,13,~4. Recently there has been a renewal of interest in the functional role of radial fibers, following the proposal of Rakic 21-z3 that they constitute glial elements which, in addition to serving as the scaffolding for the fragile developing cerebrum, provide a guidance role for migrating neurons. In this report morphological details of radial fibers in early human fetal cerebrum are described. Because of the difficult problem of identifying cell types in fetal tissue, we have found it essential to utilize combined techniques, the results of which can be correlated in order to enhance the accuracy of identification. Based upon Golgi, immunofluorescent and electron microscopic techniques, we report evidence that radial fibers are astroglial fibers, and propose that they be designated as processes of immature astrocytes. MATERIAL AND METHODS Material consisted of full thickness blocks from the midconvexity of the primitive frontoparietal region of the cerebrum of 36 human fetuses. The fetuses were received as surgical pathology specimens following abortion by hysterotomy. Estimated ovulation ages* (based on crown-rump length 1) varied from 7 to 20 weeks. Procedures utilized were Golgi, immunofluorescent analysis for glial fibrillary acidic (GFA) protein and transmission electron microscopy (EM). Of these procedures, it was generally possible to carry out no more than two on a single specimen of brain because of the small size of the cerebrum. In some instances only one procedure could be accomplished. Identification of cell bodies and cell processes therefore was based upon correlation of the observations made by the use of these three procedures on multiple specimens at various fetal ages.
Golgi methods Golgi impregnation was carried out on 11 specimens, ranging in ovulation age from 10 to 20 weeks. Stensaas' modified rapid Golgi method 26, followed by rapid * In man ovulation age is generally about 2 weeks shorter than gestational age based on the last menstrual period.
297 embedding in low-viscosity nitrocellulose, and Van der Loos' modification of the Golgi-Cox procedure ~0, were used. All sections were cut at thicknesses of 100-200/zm, and selected sections were counterstained with methylene blue in order to provide a Nissl view of the cortical architecture. Piccolyte* was used as mounting medium.
Immunofluor escence Indirect immunofluorescence for G F A protein was carried out on 6 / , m cryostat sections according to the method of Coons and Kaplan 10. Nine specimens were utilized, ranging in age from 10 to 19 weeks ovulation age. Details of our procedure on fetal tissue including use of controls have been followed as described in previous publications ~,4, except for fixation. We have found anhydrous ether: absolute ethanol, 50:50, to be a more satisfactory fixative for this technique than acetone. Slides were examined with a Zeiss fluorescence microscope, using an HBO 200 W/4 superpressure mercury burner light source, BG 38 and BG 12 exciter filters and a 500 nm barrier filter. Photographs were taken with Kodak High Speed Ektachrome film.
Electron microscopy Eighteen specimens were subjected to EM analysis, varying in ovulation age from 7 to 18 weeks. Full thickness blocks (leptomeningeal surface to ventricle) were cut, rapidly placed in cold, buffered glutaraldehyde (4 ~ , pH 7.4) and fixed for 18-20 h at 4 °C. They were then rinsed in phosphate buffered saline and cut into thin strips, each containing pial surface at one end and ventricular surface at the opposite. The strips were fixed further in 1 ~ osmium tetroxide, dehydrated in ethanol and embedded in Epon. Before embedding, strips were cut into smaller pieces and each piece numbered and oriented for embedment so that plane of section was known and original anatomic site in the cerebral wall was recorded. One micrometer sections stained with toluidine blue were utilized for correlation and selection of areas for EM study. Thin sections were cut with a diamond knife, stained with uranyl acetate and lead citrate, and examined in a Philips EM 201 electron microscope. RESULTS
Golgi Preparations Successful Golgi impregnation of radial glial fibers was achieved in fetuses 12 weeks of ovulation age and older. Radial glia were identified by the presence of long, bipolar processes with a radial (i.e. vertical or somewhat oblique) orientation, spanning the hemisphere from the ventricular zone** to the pial surface (Fig. 1A). At the pial surface many of them branched and terminated in pyramidal end feet expansions (Fig. 1B). Some ended on vascular walls (Fig. 1A) at various levels of the hemisphere, whereas still others terminated in free arborizations. In 12-14-week fetuses cell bodies giving rise to radial fibers (Fig. 1A and C) * Ward's Natural Science Establishment, Inc., Rochester, N.Y. 14603, U.S.A. ** The terminology applied to zones of the developing cerebrum is that recommended in ref. 2.
298
Fig. 1. Composite camera lucida drawing (1A) and photomicrographs (1B-I E) of Golgi preparations from 12-week-old human fetal cerebrum. A: M, marginal zone; CP, cortical plate; IM, intermediate zone; S, subventricular zone; V, ventricular zone. Note the radial processes, some of which span the entire thickness of the cerebrum from ventricular zone to pial surface. Some processes terminate at the pial surface and some on vascular walls, while others arborize freely in the cerebrum. Golgi-Cox. B: characteristic subpial branching and terminal expansion of a radial process. Golgi-Cox. x 1615. C: bipolar glial cell in the subventricular zone giving rise to a long radial process directed toward pia. Golgi-Cox. x 1030. D: neurons in the lower portion of the cortical plate. Note beginning formation of basal dendrites. Rapid Golgi. x 923. E: bipolar neuron with short processes in the intermediate zone. Rapid Golgi. x 923.
were located within the ventricular a n d subventricular zones. The cell configuration was bipolar a n d radial processes at this age were s m o o t h in outline. I n 18-20-week fetuses the cell bodies were primarily located within the subventricular zone, a n d to a lesser extent in the intermediate zone. A striking feature of radial glia in older fetuses was the presence of n u m e r o u s , unevenly spaced, bushy projections (Figs. 2A a n d 4B) over their entire surface, including b o t h soma a n d fibers. Stellate glial forms with distinct vascular a t t a c h m e n t s (Fig. 2B a n d C) were observed in the subventricular zone in 14-16-week fetuses, a n d in the intermediate zone a n d cortical plate as well in older fetuses. Since stellate cells were f o u n d in increasing n u m b e r s with advancing fetal age, a search was made for t r a n s i t i o n a l forms which might indicate that stellate cells were derived from radial glia. I n fetuses 18 weeks of
299
Fig. 2. Photomicrographs (A and C) and camera lucida drawing (B) of glia from 20-week-oldhuman fetal cerebrum. A: soma and radial glial process with numerous irregular bushy projections. The cell body is located in the subventricular zone. Rapid Golgi. x 1260.B: astroglia displaying processes extending to blood vessels and longer processes directed toward pia mater. These are located at approximately 350 #m from pial surface in the cortical plate. Rapid Golgi. C: well-differentiated stellate astroglia in the cortical plate with vascular endfeet (arrows). Rapid Golgi. x 1260.
age and older, cells exemplified in Fig. 2B were encountered within the cortical plate and marginal zone. Cells of this nature had stellate processes, some of which extended to a blood vessel wall, but in addition had a somewhat longer process projecting externally toward the pial surface. These features suggested possible transitional forms originating from radial glia. Migrating neurons in the subventricular and intermediate zones were bipolar with short processes (Fig. 1E). They were closely juxtaposed to radial glia. Within the outer portion of the cortical plate neurons were bipolar, as in the intermediate zone. At deeper levels of the cortical plate there were features of cytodifferentiation, with formation of basal dendrites (Fig. 1D), and, in older fetuses in this study, development of axons.
Immunofluorescence Indirect immunofluorescence demonstrated large numbers of radially oriented, nearly parallel fluorescent (GFA protein-positive) fibers (Fig. 3A, B and C) between the ventricular zone and pia mater. These fibres were found in all fetal specimens in which the procedure was carried out (10 weeks fetal age and older). There was a remarkable degree of similarity between the fluorescence picture and the Golgi images, again regardless of the age of the fetus. Thus, with both techniques, fibers tended to be smoother at earlier ages and irregular in outline in older specimens (Fig. 4A and B). Neuronal processes were non-fluorescent, although sections counterstained with
300 :i
Fig. 3. Photomicrograph of sections exposed to immunofluorescent procedure for GFA protein demonstrating roughly parallel fluorescent radial processes extending toward pia mater from ventricular surface (A, B and C) and horizontal fluorescent membrane at the pial surface (arrows) (D). A" 10-week-old human fetal cerebrum. V, ventricular zone. × 533. B: IM, intermediate zone. 12-week fetus, x 426. C: CP, cortical plate. M, marginal zone. 15-16 week fetus. × 426. D: methyl green counterstaio produces red fluorescence of nuclei, and are shown as white images in the photograph. PM, pia mater. 16-week-old fetus. • 533. E: control section incubated with normal rabbit serum in place of anti GFA protein serum showing no fluorescence of radial fibers or membrane at the pial surface (arrows). 16-week-old fetus. > 533. m e t h y l green for nuclear staining revealed close p r o x i m i t y o f small n e u r o n a l nuclei to fluorescent fibres, especially in the i n t e r m e d i a t e zone. R a d i a l fluorescent fibers t e r m i n a t e d in a fluorescent m e m b r a n e at the o u t e r surface o f the m a r g i n a l zone (Fig. 3D) directly beneath the p i a mater. N o structures in the leptomeninges or c h o r o i d plexus were positive for G F A protein. C o n t r o l sections in which the G F A p r o t e i n a n t i s e r u m was replaced with n o r m a l r a b b i t serum were consistently negative (Fig. 3E).
Electron microscopy I n every specimen e x a m i n e d (ages varying f r o m 7 to 18 weeks) there were n u m e r o u s linear processes which were distinct f r o m n e u r o n a l processes on the basis o f ultrastructure. Their characteristics were the following: They were m o d e r a t e l y electron-lucent (Figs. 4C, 5A a n d 6). F i l a m e n t s 8-9 n m in d i a m e t e r were present (Figs. 4C, 5 a n d 6), often in bundles, a n d scattered glycogen granules were usually
301
!
Fig. 4. Glial processes in the intermediate zone demonstrated by immunofluorescent, Golgi and electron microscopic techniques. A: fluorescent radial processes, single and in clusters. Note surface irregularities. Indirect immunofluorescence for GFA protein. 17-week-old human fetal cerebrum. × 1008. B" rapid Golgi technique showing shaft of a radial glial process. Note lateral projections. 20-week-old human fetal cerebrum. × 800. C: electron micrograph of glial process (GP). It is relatively electron-lucent, larger than adjacent neuritic processes (large arrow) and contains 8-9 nm glial filaments (gf) and occasional glycogen granules (gly). One of the neuritic processes shows 20-25 nm neurotubules (nt). M, Mitochondrion. 17-week-old human fetal cerebrum, x 43,264.
encountered (Figs. 4C and 5). The caliber of these processes was generally larger than those of surrounding, neurotubule-containing neurites (Figs. 4C, 5A and C and 6). Short lamellar expansions or varicose enlargements (Figs. 4C and 5C) were often noted, occasionally containing numerous vesicles of varying sizes (Fig. 5C). Longitudinally oriented microtubules, scattered mitochondria, smooth endoplasmic reticulum and ribosomes were also observed, but less frequently. O f these features, the most useful in terms of recognizing this group of processes were 8-9 nm filaments, glycogen granules and a relatively electron-lucent matrix. This constellation of morphologic features is distinctive of astrocytes and processes of this nature were therefore interpreted as astrocytic. Processes meeting these criteria were numerous in the subventricular zone, and only slightly less abundant in the intermediate zone. They were present within the cortical plate, though less conspicuous as they tended to be obscured by the densely packed neurons of the pallium. On entering the marginal zone they often branched in an irregular manner, and many exhibited localized expansions. At the pial surface they formed fusiform expansions which were closely applied to the basal lamina of the pia mater (Fig. 7). Each terminal expansion was tightly adherent to its
302
Fig. 5. Electron micrographs of glial processes. A : glial process (GP) in the intermediate zone. Note glycogen granules (gly), glial filaments (gf) and a few microtubules in the large electron-lucent glial process. Cross-sections of smaller neuritic processes (np) containing neurotubules are also shown in the photograph. 17-week-old human fetal cerebrum, x 30,703. B: glial process (GP) in the cortical plate. The glial process is closely apposed to cell membrane of a neuron and contains glycogen granules (gly) and smooth endoplasmic reticulum. 9-week-old human fetal cerebrum, x 26,800. C: glial process (GP) showing varicosity filled with vesicles (V). Glycogen granules (gly) are numerous. Note also adjacent neuritic process with 20 25 nm neurotubules (nt). l l-week-old human fetal cerebrum, x 18,090. neighbor. D e s m o s o m e - l i k e contacts were occasionally seen between a d j a c e n t terminal processes. Processes with u l t r a s t r u c t u r a l characteristics similar to the a b o v e but with fewer filaments were also f o u n d a d j a c e n t to capillaries in all o f the specimens examined (Fig. 8A a n d B). They were closely a p p l i e d to the basal l a m i n a o f capillaries and frequently f o r m e d pericapillary investments. G l y c o g e n granules were p a r t i c u l a r l y prom i n e n t within these pericapillary processes. A close a p p o s i t i o n o f astrocytic processes to the s o m a a n d processes o f migrating n e u r o n s was often observed (Figs. 9 a n d 10) in different zones o f the developing c e r e b r u m . Sometimes n e u r o n a l cell bodies were flanked by extensions o f astrocytes (Fig. 10), a n d there were often localized thickenings o f a p p o s e d m e m b r a n e s consistent with d e s m o s o m e s . Cell bodies giving rise to a s t r o c y t i c processes were difficult to identify with certainty in o u r y o u n g e r fetuses, even t h o u g h the latter were distinctive a n d easily recognized. N e i t h e r nuclear n o r c y t o p l a s m i c characteristics were helpful. I n all fetuses 16 weeks o f age a n d older, however, cells giving rise to astrocytic processes were
303
Fig. 6. Electron micrograph demonstrating relatively large, electron-lucent glial process (GP) in the subventricular zone. Note scattered 8-9 nm glial filaments (gf), mitochondria, ribosomes and smooth endoplasmic reticulum. Adjacent neuritic processes (np) are more electron-dense and contain 20-25 nm neurotubules. 11-week-old human fetal cerebrum, x 6,450.
Fig. 7. Electron micrograph at piad surface demonstrating expanded glial processes (GP) tightly abutting the basal lamina (arrows) of pia mater (PM). Note scattered glycogen granules (gly), mitochondria, smooth endoplasmic reticulum~andlribosomes. The glial processes are also tightly apposed to each other. 9-week-old human fetal~erebrum, x 13,950.
Fig. 8. Electron micrographs demonstrating attachments of glial processes to capillaries. A : electronlucent glial processes (GP) forming investment surrounding capillary. Arrows point to glycogen granules 10-week-eld human fetal cerebrum. < 6723. B: glial process (GP) outlined in black lying adjacent to capillary. I I-week-old human fetal cerebrum. :~ 13,071.
Fig. 9. Electron micrograph depicting glial process (GP) outlined in black closely apposed to membrane of process of a migrating neuron in the cortical plate. 7-week-old h u m a n fetal cerebrum. × 18,900.
305
Fig. 10. Electron micrograph of migrating neuron in the intermediate zone flanked by glial processes (GP) containing glycogen granules (gly) and vesicles (v). The process of the neuron contains a growth cone (go). nt, neurotubule. 10-week-old human fetal cerebrum. × 15,700.
306
Fig. 11. Electron micrograph of cell body and process of astrocyte in the subventricular zone. Note abundant glial filaments (gf) in the cytoplasm. Rough endoplasmic reticulum, mitochondria (M) and ribosomes are also present. 16-week-old human fetal cerebrum. ~', 12,150.
307
Fig. 12. Electron micrograph of astrocyte in the subventricular zone. The cytoplasm is packed with glial filaments (gf). Golgi apparatus (G), rough endoplasmic reticulum (RER), mitochondria (M) and dense bodies are also present. Thick arrow points to neurotubules in adjacent neuritic process. 18-week-old human fetal cerebrum, x 20,350.
308 readily identified (Figs. 11 and 12) and were abundant. The cell bodies were found principally in the subventricular zone toward the ventricular surface. The nucleus was usually round or ovoid, but occasionally lobulated. Cell boundaries were irregular, cytoplasm was electron-lucent and abundant, and 8-9 nm filaments were numerous. Glycogen granules were present but generally more sparse than in the processes. In addition, there were scattered mitochondria, stacks of rough endoplasmic reticulum, ribosomes and a prominent Golgi apparatus. DISCUSSION In this study we have shown that in Golgi preparations radial glia are demonstrable in human fetal cerebrum by 12 weeks ovulation age, the earliest age at which Golgi impregnation was successful. In immunofluorescent preparations utilizing antiserum to GFA protein, fluorescent fibers of similar radial orientation were numerous in all zones of the cerebrum as early as the 10th week (the age of the youngest specimen examined with this technique). On the basis of the specificity of GFA protein for astrocytesS,~, 12, and the similarity of the fibers demonstrated by the Golgi and immunofluorescent procedures, we infer that radial fibers are processes of astrocytes. EM analysis revealed unmistakable evidence of processes of astrocytic nature in the cerebrum of the human fetus. They were found at all levels of the cerebrum and at all fetal ages examined by EM (7-18 weeks). Their electron-lucence and the presence of 8-9 nm filaments and glycogen granules distinctly set them apart from neurotubule-containing neurites 7,19,2°,~9. By 7 weeks they had already formed pericapillary investments. Equally significant, the terminal expansions of astrocytic processes formed a subpial layer adherent to the basal lamina of the pia mater analogous to the subpial membrane positive for GFA protein by immunofluorescence. The fine structural observations thus correlate with and complement the results of Goigi impregnation and immunofluorescence analysis for GFA protein. The soma of the cells giving rise to astrocytic processes were replete with 8-9 nm filaments as well as the usual organelles characteristic of more mature cells. Cells of this nature meet the ultrastructural criteria of astrocytes, and can hardly be confused with any other type of cell in the normal central nervous system. We recognize that it might be argued that the glial elements we have described are 'spongioblastic', 'gtioblastic' or 'astroblastic' forms rather than astrocytes. The presence of GFA protein and large numbers of 8-9 nm filaments, however, indicate that these cells, even though immature, already possess fundamental immunocytochemical and morphologic characteristics by which adult astrocytes are identified. The maturational features which we have observed during the fetal age span between 7 and 20 weeks and which occur subsequently are appropriately viewed as phenomena of cytodifferentiation. They are analogous to the changes that neurons undergo as they mature from bipolar cells to cells with well-defined dendrites and axons and the configuration of mature neurons. To invoke the concept that the astrocytic forms we have described are 'blastic' or 'matrix' precursor forms can no longer be justified, even though the cells are undoubtedly migrating, changing form, taking on new
309 functions and probably still dividings throughout this phase of development. Our data of course provide no indication as to the cell of origin of astrocytes, only that their generation is well underway by 7 weeks. Our findings also do not exclude the possibility that more primitive forms are present during this 7-20-week period which are continually giving rise to new astrocytes. It seems probable, in fact, that the population of astrocytes steadily increases during this time, though we did not examine this aspect. Earlier stages of histogenesis thus remain to be delineated. Finally, it should be emphasized that in none of our specimens of fetal cerebrum were cells observed that bore features characteristic of oligodendrocytes. Identification of cells in early development of the brain presents a difficult challenge, and criteria for identifying cell types have only gradually emergedT,16,19,20, 26-29. MaginilS in 1888 appears to have been the first to mention radial glia in fetal cerebrum, but he was unable to identify cell type more specifically. Ram6n y Cajal 2a described the morphology of radial fibers in developing vertebrate brain (including man) in considerable detail, and identified the cells from which these fibers originate as 'Golgi epithelial cells'. He speculated that they develop into neuroglial cells. In more recent literature, Stensaas28,z7 has identified fibers in rabbit cerebrum with the configuration of radial glia as spongioblasts, and has suggested that they give rise to at least a portion of the ultimate adult population of astrocytes. Peters and Feldman 19, in a combined Golgi and EM study, reported the existence of vertically oriented, long radial processes in the cerebral hemispheres of late rat fetuses, and considered them to be the processes of spongioblasts or primitive epithelial cells. In many respects, the Golgi images and EM features of the spongioblasts they described were similar to those of the radial glia we have observed in human fetal cerebrum. Whether or not such processes in the late rat fetus also possess specific immunocytochemical characteristics of astrocytic differentiation is not known at the present time. The concept of guidance of migrating neurons by radial glia was proposed and elaborated by Rakic 2a-za. Sidman and Rakic 25 demonstrated radial glia in the cerebrum of a 16-week-old human fetus, and suggested that in man they also serve in the role of guidance of neurons during fetal development. One of the chief drawbacks in the past to the concept of glial guidance of neuronal migration has been the notion that most gliogenesis in the cerebrum of the human fetus occurs only after the majority of neurons have formed and have assumed their final position in the cortical plate11,lz, 14. In man, gliogenesis has been thought to take place mostly from the 18th to 20th week onward. Kostovi6 et al.17, however, have examined human fetuses of varying ages with Golgi and Nissl techniques, and to a limited extent by EM. Although they relied primarily upon electron-lucency in the EM identification of astrocytes, they concluded that immature astrocytes were present at least as early as 11 weeks of fetal age. Sturrock za also reported early proliferation and maturation of fibrous astrocytes in optic nerves of human fetuses ranging from 8 to 18 weeks postconception age. t n our own previously reported series of tissue culture studiesa,S,9, we have repeatedly observed the early appearance and active proliferation of astrocytes in the ou'tgrov~th of'explahtS bf Cerebrum from human fetuses ranging in age from 7 to 20
310 weeks. O u r present study with the c o m b i n e d use o f Golgi, immunofluorescent and E M techniques on tissue blocks has d o c u m e n t e d that astrocytes are present in the form o f radial glia, pericapillary investments a n d stellate forms during the first trimester o f fetal life. In o u r o p i n i o n it is highly likely, t h o u g h this remains to be proved, t h a t astrocytes also develop much earlier in the c e r e b r u m o f o t h e r vertebrates than f o r m e r l y believed. A l t h o u g h the full implications o f this early astrocytic d e v e l o p m e n t can only be speculated u p o n at this time, further investigation o f this p h e n o m e n o n constitutes an i m p o r t a n t challenge in the field o f d e v e l o p m e n t a l neurobiology. C o n c e p t s dealing n o t only with the mechanisms involved in neuronal m i g r a t i o n a n d f o r m a t i o n o f the cortical plate but also with the m o d e o f pathogenesis o f those d e v e l o p m e n t a l disorders in which neuronal m i g r a t i o n a n d cerebral cortical architectonics are disturbed will need to i n c o r p o r a t e the knowledge that in m a n cells with features o f i m m a t u r e astrocytes are conspicuous at least as early as the seventh week of fetal life. ACKNOWLEDGEMENTS This w o r k was s u p p o r t e d in p a r t by Research G r a n t 5 R01 H D 07078 f r o m the N a t i o n a l Institute o f Child H e a l t h a n d H u m a n D e v e l o p m e n t ; a n d in p a r t by Center G r a n t ES01247 f r o m the N a t i o n a l I n s t i t u t e o f E n v i r o n m e n t a l H e a l t h Sciences. W e t h a n k Mrs. M a r t h a K u m l e r a n d Ms. M a r y l o u Peck for technical assistance, a n d Mrs. L i n d a C r a n d a l l for secretarial help.
REFERENCES 1 Altman, P. L. and Dittmer, D. S., Biology Data Book, Fed. Amer. Soc. Exp. Biol., Washington, D.C., 1964, pp. 82 84. 2 Angevine, Jr., J. B., Bodian, D., Coulombre, A. J., Edds, M. V., Hamburger, V., Jacobson, M., Lyser, K. M., Prestige, M. C., Sidman, R. L., Varon, S. and Weiss, P. A., Embryonic vertebrate central nervous system: revised terminology, Anat. Rec., 166 (1970) 257-262. 3 Antantius, D. S., Choi, B. H. and Lapham, L. W., Immunofluorescence staining of astrocytes in vitro using antiserum to glial fibrillary acidic protein, Brain Research, 89 (1975) 363-367. 4 Antanitus, D. S., Choi, B. H. and Lapham, L. W., The demonstration of glial fibrillary acidic protein in the cerebrum of the human fetus by indirect immunofluorescence, Brain Research, 103 (1976) 613-616. 5 Bignami, A. and Dahl, D., Differentiation of astrocytes in the cerebellar cortex and the pyramidal tracts of the newborn rat. An immunofluorescence study with antibodies to a protein specific to astrocytes, Brain Research, 49 (1973) 393402. 6 Bignami, A. and Dahl, D., Astrocyte-specific protein and radial gila in the cerebral cortex of newborn rat, Nature (Lond.), 252 (1974) 55-56. 7 Caley, D. W. and Maxwell, D. S., An electron microscopic study of the neuroglia during postnatal development of the rat cerebrum, J. comp. Neurol., 133 (1968) 45-70. 8 Choi, B. H. and Laphan, L. W., Autoradiographic studies of migratingneurons and astrocytes of human fetal cerebral cortex in vitro, Exp. molec. Pathol., 21 11974) 204-217. 9 Choi, B. H. and Lapham, L. W., Interactions of neurons and astrocytes during growth and development of human fetal brain in vitro, Exp. molec. Pathol., 24 (1976) 110-125. 10 Coons, A. H. and Kaplan, M. H., Localization of antigen in tissue cells II. Improvements in a method for the detection of antigen by means of fluorescent antibody, J. exp. Med., 91 (1950) 1-13. 11 Dobbing, J. and Sands, J., Quantitative growth and development of human brain, Arch. Dis. Child., 48 (1973) 757-767.
311 12 Eng, L. F., Vanderhaeghen, J. J., Bignami, A. and Gerstl, B., An acidic protein isolated from fibrous astrocytes, Brain Research, 28 (1971) 351-354. 13 Fujita, H., Electron microscopic studies on the histogenesis of nerve cells and neuroglia in the domestic fowl, Acta anat. nippon, 38 (1963) 95-108. 14 Fujita, S., and Takeoka, O., Morphologic and quantitative studies on cellular proliferation and differentiation in the central nervous system of the human embryo, Acta pathoL japon., 11 (1961) 256. 15 Hattori, T. and Fujita, S., Scanning electron microscopic studies on morphology of matrix cells, and on development and migration of neuroblasts in human and chick embryos, J. Elect. Microsc., 23 (1974) 269-276. 16 Kershman, J., The medulloblast and the medulloblastoma. A study of human embryos, Arch. Neurol. Psychiat., 40 (1938) 937-966. 17 Kostovi~, J., Krmpoti~-Nemani6, J. and Kelovi6, Z., The early development of glia in human neocortex, Yugoslav. Acad. Sci., 371 (1975) 155-160. 18 Magini, G., Sur la n6vroglie et les cellules nerveuses c6r6brales chez les foetus, Arch. ital. BioL, 9 (1888) 59-60. 19 Peters, A. and Feldman, M., The cortical plate and molecular layer of the late rat fetus, Z. Anat. Entwickl.-Gesch., 141 (1973) 3-37. 20 Privat, A., Postnatal gliogenesis in the mammalian brain, Int. Rev. Cytol., 40 (1975) 281-323. 21 Rakic, P., Guidance of neurons migrating to the fetal monkey neocortex, Brain Research, 33 (1971) 471-476. 22 Rakic, P., Mode of cell migration to the superficial layers of fetal monkey neocortex, J. comp. Neurol., 145 (1972) 61-84. 23 Rakic, P., Timing of major ontogenetic events in the visual cortex of the rhesus monkey. In N. A. Buchwald and M. A. B. Brazier (Eds.), Brain Mechanisms in Mental Retardation, Academic Press, New York, 1975, pp. 3-39. 24 Ram6n y Cajal, S., Histog6n6se de l'6corce c~r~brale, In Histologie du Syst~me Nerveux del'Homme et des Vertdbr~s, Vol. 11., Consejo Superior de Investigaciones Cientificas, Instituto Ram6n y Cajal, Madrid, 1955, pp. 847-861. 25 Sidman, R. L. and Rakic, P., Neuronal migration, with special reference to developing human brain: a review, Brain Research, 62 (1973) 1-35. 26 Stensaas, L. J., The development of hippocampal and dorsolateral pallial regions of the cerebral hemisphere in fetal rabbits, I. Fifteen millimeter stage, spongioblast morphology, J. comp. Neurol., 129 (1967) 59-70. 27 Stensaas, L. J. and Stensaas, S. S., An electron microscope study of cells in the matrix and intermediate laminae of the cerebral hemisphere of the 45 mm rabbit embryo, Z. Zellforsch., 91 (1968) 341-365. 28 Sturrock, R. R., A light and electron microscopic study of proliferation and maturation of fibrous astrocytes in the optic nerve of the human embryo, J. Anat. (Lond.), 119 (1975) 223-234. 29 Vaughn, J. E. and Peters, A., A third neuroglial cell type, J. comp. Neurol., 133 (1968) 269-288. 30 Van der Loos, H., Une combinaison de deux vieilles methodes histologiques pour de syst6me nerveux central, Msehr. Psychiat. Neurol., 132 (1956) 330-334.