Mitotic cycling of radial glial cells of the fetal murine cerebral wall: a combined autoradiographic and immunohistochemical study

183

Developmental Brain Research, 38 (1988) 183-190 Elsevier BRD 50675

Mitotic cycling of radial glial cells of the fetal murine cerebral wall: a combined autoradiographic and immunohistochemical study Jean-Paul Misson 1, Michael A. Edwards, Miyuki Yamamoto and Verne S. Caviness, Jr. IEunice Kennedy Shriver Center, Department of Developmental Anatomy and Pathology, Waltham, MA 02254 ( U.S.A.) and Department of Neurology, Massachusetts General Hospital, Harvard Medical School, Boston, MA 02114 ( U. S.A . ) (Accepted 14 July 1987)

Key words: Radial glial cell; Cerebral cortex; Lineage; Immunohistochemistry; Autoradiography

Radial glial cells of the embryonic murine cerebral wall are selectively labeled by staining with antibody RC1. In order to study the mitotic cycling of these cells, we combined RCI immunohistochemistry and autoradiographic analysis following [3H]thymidine injection at 1, 2, 6, 48 h prior to sacrifice. Many radial glial cells, i.e. RCl-positive cells, incorporate the DNA tracer and hence must be mitotically active. Other proliferative cells of the ventricular zone do not stain with RC1. With the transition from S to M phase, the nuclei of the radial glial cells participate in the interkinetic 'to-and-fro' nuclear translocation characteristic of the non-radial glial cells of the ventricular zone. The density of radioactive grains over nuclei of both RCl-positive and negative cells of the ventricular zone becomes similarly reduced in the 48 h following the [3H]thymidine incorporation. Thus, the subpopulation of radial gila with nuclei within the ventricular zone which have incorporated the DNA tracer does not appear to become arrested in a prolonged G1 phase. The resuits suggest that the ventricular zone includes at least two subpopulations of stem cells, neuronal and radial glial. Radial glial cells, i.e. RCl-positive cells, are inferred to serve initially as a progenitor population for new radial glial cells. Later in development, they probably become a source of other cells of astroglial lineage.

INTRODUCTION

Radial glial cells are a specialized cell class of the astroglial lineage which are present transiently in the developing mammalian central nervous system tT'2t. Their numbers increase dramatically during the neurogenetic phase of brain development and appear maximal during the period of neuronal migration it'21. Postmitotic neurons are believed to be guided in their migration by ascending the surfaces of radial glial fibers 15't6. These processes span the full width of the wall of the developing central nervous system, extending from somata in the ventricular zone to the pial surface, where their endfeet are continuous with the external limiting glial membrane. The precursor cell for radial glial cells has not been specifically identified. The hypothesis of His 8 holds

that the ventricular zone is composed of a mixed population of precursor cells, some committed to the neuronal and others to the glial lineage. Other hypotheses are tenable. For example, Schaper t9 suggested that the ventricular zone is composed of uncommitted generative cells, competent to give rise to cells of either glial or neuronal lineage. Recent studies have been interpreted to support the hypothesis of His. Levitt et a1.12'13 and Hockfield and McKay 9 have identified immunohistochemically cells in mitosis at the ventricular margin which already express glial cell antigens. These results suggest a commitment to the glial lineage at least as early as a terminal mitosis. These studies leave unanswered the question of whether some or all bipolar radial glial cells remain in a mitotically active precursor pool. The present analysis, in murine embryos, is de-

Correspondence: J.P. Misson, Eunice Kennedy Shriver Center, Department of Developmental Anatomy and Pathology, 200 Trapelo Road, Waltham, MA 02254, U.S.A. 0165-3806/88/$03.50 © 1988 Elsevier Science Publishers B.V. (Biomedical Division)

l ,',4

signed to determine if radial glial cells are a mitotically active population, capable of serving as a stem cell for more radial glial cells. It takes advantage of an immunocytochemical m a r k e r , monoclonal antibody RC1 (formerly designated 8C7), which stains radial glial cells from at least as early as E l 0 to the early postnatal period 3'a and is compatible with [3H]thymidine a u t o r a d i o g r a p h y ~a. MATERIALS AND METHODS Pregnant females of the C57B × C3H hybrid mouse strain received a single intraperitoneal injection of [3H]thymidine, 5 BC/g, on either the 13th or 15th gestational days ( E l 3 , E l 5 ) . The animals, one per age and survival time, were anesthetized and decapitated at 1,2, 6 or 48 h after the thymidine injection. E m b r y o s were r e m o v e d by hysterotomy and frozen in isopentane cooled on dry ice. A t each age and survival time, two embryos were processed for c o m b i n e d immunohistochemistry and autoradiography. Coronal brain sections, 10/~M thick, were cut in a cryostat a t - 1 5 °C, m o u n t e d on gelatin-coated slides, desiccated under vacuum for 1 h, and stored at - 7 0 °C. Following acetone t r e a t m e n t for 3 - 5 rain and washing in p h o s p h a t e - b u f f e r e d saline solution. the brain sections were incubated 15-20 h with the RC1 h y b r i d o m a supernatant at 4 °C. A f t e r washing with 1% normal goat serum in phosphate-buffered saline, the sections were incubated for 2 h at room t e m p e r a t u r e with goat anti-mouse IgM c o n j u g a t e d to horseradish peroxidase (Boehringer, M a n n h e i m / . Sections were reacted with a solution of 0.05% 3.3' diaminobenzidine ( D A B ) (Sigma) and 0 . 0 0 2 g H 2 0 > washed in buffer and postfixed for 1 - 3 min m 10% formalin. Sections processed for a u t o r a d i o g r a p h y were defatted in series of alcohol solutions, dried for 30 mm at 37 °C, and d i p p e d in NTB-2 K o d a k nuclear emul-

sion diluted 1:1 with distilled water. Following exposure for 4 weeks at 4 °C, the a u t o r a d i o g r a m s were d e v e l o p e d in K o d a k D19 followed by stop-bath and Kodak R a p i d Fixer. Some sections of each case were counterstained with Cresyl violet to allow delineation of layers in the fetal cortex. The relative incidence of cells labeled with both methods in the ventricutar zone was quantified in samples of dorsolateral cortex from some E l 3 and E l 5 cases with short survival time. Because counterstaining obscures immunoreactivity, counts were made from non-counterstained sections. The distribution of R C l - p o s i t i v e somata with heavv autoradiographic labeling (>15 grains ow'r their nuclei) was assessed in the inner half of the ventricular zone using a 10/) x oil immersion objective and Nomarksi optics. The density of d o u b l e - l a b e l e d cells in transition from S to M phase was calculated from charts of 12-17 fields m a d e with an eyepiece graticule (10 x 10 grid of squares with 9.7 l,m sides). In o r d e r to calculate the p r o p o r t i o n of mitotically cycling cells expressing radial glial antigen, the distribution of cells with heavy [3H]thymidine labeling was also charted for the same fields and quantified. Counts of the latter from small areas of counterstained sections on E13 and E l 5 with 1 h survival were within I 5 - 2 0 % of mean values obtained using Nomarski optics alone to discern cell boundaries. RESULTS Patterns of labeling with the two m e t h o d s were closely examined in the dorsolateral neocortex on E 13 and E 15. a p e r i o d when cells of the infragranular and midcortical layers are being g e n e r a t e d L-~. At these ages. there is no distinct subventricular zone in this region of the pallium, and. except for mesenchyreal cells of blood vessels and meninges, mitotic acnvity is confined to the ventricular zone 26'27. Accordmg to previous studies m'e°es, the average length of

Fig. 1. Distribution of autoradiographically labeled nuclei m RCl-immunostained sections through parietal ct~rtcx on El3 and El5 at short intervals after exposure to [3H]thymidine. In A, a transverse sector through the cortical wall on El3 1 h after rejection, shows a dense population of immunoreactive radial gila and heavy autoradiographic labeling of nuclei in the extcrnat portion of the ve ntricutar zone (VZe). At this magnification, double-labeled cells are easiest to discern among the few somata which have descended into the internal portion of the ventricular zone (VZi) at this short interval e.g. arrowhead|. In high magnification micrographs of the VZc on El3 and El5 (B and C, respectively) and of the VZi on El3 I D-F), arrows point to somata of double-labeled cells at 1 h (B-D) and at 2 h (E,F) after [3Hlthymidine exposure. Note RCl-positivc radial processes ascending and descending from these cells D i,~ an enlarged view from A. The ventricular margin is present at the bottom of panels A and D-F. Nomarski optics. Bars =: 25 am m A aml 10urn in B-F.

185

18(~

Fig. 2. Distribution of autoradiographic and immunocytochemical labeling on E13 and El5 at 6 or 48 h after [3H]thymidine exposure. A-C: dark-field photomicrographs showing autoradiographic labeling in sectors through dorsolateral cortex on E 13 at 6 h after in ~ection (A) and on El5 at 6 and 48 h (B and C, respectively), Heavily labeled nuclei are concentrated in the VZi in A and B and in the cortical plate (CP) in C. IZ, intermediate zone. D,E: high-magnification views of the VZi from the cases shown in A and C, respectively. The arrow in D points to a double-labeled soma at the ventricular margin. The arrow in E points to a similar RC l-positive radial glial cell which, compared to nuclei in the CP (see C), exhibits only a moderate grain density over its nucleus. Bars = 50 ~nl in A - C and 25 um in D,E.

p h a s e s o f t h e m i t o t i c cycle of cells o f t h e v e n t r i c u l a r

4 - 7 h f o r G1 p h a s e . A s u r v i v a l t i m e o f 1 h a f t e r t h v -

z o n e in t h e m u r i n e c e r e b r a l wall a r e a p p r o x i m a t e l y

m i d i n e i n j e c t i o n was c h o s e n to allow a u t o r a d i o g r a -

6 - 7 h f o r S p h a s e , 1 h for G 2 p h a s e , l h f o r M p h a s e ,

phic i d e n t i f i c a t i o n of t h e s u b p o p u i a t i o n o f cells in S

187 phase. The survival times of 2 and 6 h allow identification of groups of cells entering G2 and M phases, respectively. The longer survival time of 48 h, equivalent to at least two average cell cycles, was selected to permit identification of cells that do not re-enter S phase following M + G 1, i.e. cells which retain heavy [3H]thymidine labeling. In accord with the observations of Edwards et al. 3'4, somata of RCl-positive cells can be recognized at all depths of the ventricular germinative zone throughout the presently studied period. Where the plane of section is favorable, RCl-stained radial processes ascending and descending from the immunoreactive subpopulation of somata are readily discernible, as consistent with the morphology of radial glial cells. At a survival time of 1 h, thymidine-labeled cell nuclei are concentrated in the outer half of the ventricular zone (Fig. 1A), corresponding to the zone of DNA synthesis during the S phase 5'1s'25. Among this thymidine-labeled population are cells which are also densely stained immunohistochemically with RC1, that is, cells which are 'double-labeled' (Fig. 1B,C). A small number of double-labeled cells may be identified in the inner half of the ventricular zone, presumably corresponding to radial glial cells which have just entered the G2 phase (Fig. 1D). Two h after thymidine exposure, autoradiographically labeled nuclei remain abundant in the outer half of the ventricular zone. However, many heavily labeled nuclei have now shifted to an intermediate depth of the ventricular zone and some are distributed near the ventricular lumen (Fig. 1E,F). Doublelabeled cells with radial processes are readily detected both among the population of cells with nuclei in the outer ventricular zone, presumably in S phase, and in the inner ventricular zone, presumably in G2 phase. At a survival time of 6 h, most autoradiographicaily labeled nuclei have shifted to the inner part of the ventricular zone and become concentrated at the ventricular margin (Fig. 2A,B). Subcellular cytological resolution in frozen sections is not sufficient to allow identification of mitotic figures, but, from results of previous studies 5'7'18'25, it is presumed that most of the nuclei at the ventricular margin belong to cells in M phase. A few nuclei are still located in the outer half of the ventricular zone, corresponding to cells still in S or G2 phases. Double-labeled cell somata

TABLE I

Quantification of incidence of double-labeling Age

Time after [SH]thymidine exposure

Area sampled (mm 2)

Density of doublelabeled cells (number/ mm 2 + S.D.)

Percentage of [~H]thymidine-labeled cells with RC1 label

El3

1h 2h 6h 1h

0.108 0.144 0.081 0.037

590 (± 160) 1410(+ 340) 1930 (__+580) 430(+ 191)

28 32 28 32

El5

are found at each site, but they are more prevalent near the ventricular margin. Some of these cells exhibit at least a short radial process (Fig. 2D). Quantitative analyses (Table I) established that between 1 and 6 h survival at El3, progressively more [3H]thymidine-labeled cell nuclei appear in the lower half of the ventricular zone. Among this population of mitotically active cells, approximately one third are estimated to stain with RC1. As the cycle progresses and more cells enter in G2 and M phases, the same ratio of about 30% cells remain identifiable as double-labeled in the inner portion of the ventricular zone. A similar proportion of double labeled cells is also estimated from counts of an El5 case with a 1-h survival. Following a 48-h survival time, thymidine-labeled cells are distributed at all levels of the cerebral wall. In the E15 cases, virtually all of the heavily labeled cells lie external to the ventricular zone in the cortical plate, and many cells with moderate grain densities over their nuclei occupy the intermediate zone (Fig. 2C). In an El7 case studied, exposed to [3H]thymidine on E15, the heavily labeled nuclei distribute primarily to the intermediate zone. These cells, RC1 negative, correspond to migratory postmitotic neurons 1~2.By contrast, virtually all nuclei in the ventricular zone exhibit grain densities one-fourth or less than that of the heavily labeled neurons (Fig. 2D), indicative of a dilution of the radioactive label by two or more rounds of D N A synthesis and division. As the grain density of autoradiographic labeling of RCl-positive radial glia appears to be similar to that of RCl-negative cells (Fig. 2E), both populations are inferred to have undergone a similar sequence of mitotic cycling over the two-day period.

DISCUSSION These findings establish that many fully differentiated radial glial cells marked by RC1 immunostaining undergo D N A synthesis and therefore constitute a mitotically active pool of stern cells. The cells observed in mitosis at the ventricular margin of the primate neocortex, marked by an antibody to the glial fibrillary acidic protein ( G F A P ) in the study of Levitt et al. 12'~, and also of the rat spinal cord, marked by antibody R401 in the study of Hockfield and McKay ~), probably correspond to cells of the radial glial lineage which were similarly engaged in repeated cell divisions within the ventricular zone. The present observations, taken together with these previous studies, confirm the hypothesis of His s to the extent that they identify in the ventricular zone a generative radial glial cell population, RCl-posifive, which from at least as early as E 9 - 1 0 is distinguishable from a larger proliferative pool of RCl-negative cells. Presumably neurons are the progeny of the RCl-negative population. However, the present analysis does not establish that only neurons arise from the pool of RCl-negative cells. Some of glial lineage could also arise from an unlabelled population of uncommitted cells. Thus the possible existence of an indifferent neural tube cell in the sense of Schaper [~;is not excluded. Radial glial cells are shown in the present study to participate in the 'to-and-fro' interkinetic nuclear migration characteristic of the overall generative celt population of the ventricular pseudostratified epithelium 5,1s'25. Thus, in S phase their nuclei are in the outer portion of the ventricular zone, and in the transition to G2 and M phases, they shift toward the ventricular margin, apparently transloeating within the descending process attached at the ventricular surface. Quantitative analysis of [3H]thymidine labeled cells in the lower half of the ventrieular zone indicates that about one third of the cells leaving S phase on E l 3 and El5 are RCl-positive. Judging from the constant fraction of RCl-positive cells at various intervals after thymidine labeling, the rate of radial glial cell nuclear migration with progression through the generative cycle appears to be similar to that of the nuclei of the RCl-negative cells. Whether all radial glia are mitotically active could not be determined with the pulse labeling method

used in the present study. However. roughly onethird to one-half of the total cell population in the ventricular zone appear by qualitative inspection to be RCl-positive ~, Thus. the estimate that 30% of the cycling cells express radial glial antigen indicates that at least a large fraction of radial gi~;~ tire mitotically active in the period studied. Moreover. the entire subpopulation of radial glial cells ol the murine cerebral wall which have mcorporated the D N A tracer appears to remain mitotically active at the ages studied, El3 and El5. We make this inlerence from the observation that 48 h after the tracer miection, the density of autoradiographic labelin~ m the ventricular zone decreases in parallel fashion in both R C i positive and negative cell populations. None of the double-labeled radial glial cell nuclei show a dense concentration of silver grams as would be expected it they had remained in G1 phase rather than continuing another cycle of D N A synthesis In the fetal rhesus monkey neocortex, a subpopulation of radial glial cells has been inferred by Schmechel and Rakic 22 to enter the O 1 state fol an interval prolonged over at least several weeks during the period of peak neuronal migratum. This conclusion was based on observations o! mitotically arrested cells m the subventricular ~one and of the presence of radial glial somata at the ~ame site m Golgi impregnations. Whether a similar population of mitotically stable radial glial cells occupies the subventricular zone of the murine cerebral wall in the late fetal period has not been established in the present study. The existence of cells transitional in form between the classic radial glial cell and immature astroc~tes has led to the conclusion that radial glia become transformed into astrocyte,~al~~'t. II has not been established whether such a cell shape transformation is or is not contingent upon a last round of D N A synthesis and mitosis. This transformation must be rare at E13-15 in the murine cerebral cortex 3'415a (Misson, unpublishedJ. It does occur later in development as the number of radial glia cells declines in the cerebral wall. Thus it is reasonable to infer that the mitotically cycling radial glial cells detected in the earlier period of E13-15 in the present study are the source of new radial glial cells. Consistent with this conclusion, the radial glial population is inferred to increase dramatically in this period considering both their apparent

189 constant density and the large expansion of the cortical surface area 4"6. A t the time of S and G2 phases, radial glial cells have a long ascending process which is clearly delineated by staining with RC1. In our material it is not clear whether these cells withdraw, or otherwise eliminate, their apical process during M phase. Previous electron microscopic analyses of cortical neuroepithelial cells indicate such a loss of radial processes as they round up prior to m e t a p h a s e 7'24. H o w e v e r , the presence of a minority radial glial population which retains radial processes during mitotic division could have been missed in the limited cell sample examined. In the studies of Levitt et al. t2'13 and Hockfield and M c K a y 9, no processes were detected on metaphase somata expressing radial glial antigens, but serial section analyses were not performed. This issue of whether radial glia undergo a radical shape transformation upon entry into and exit from the mitotic phase of the cycle is particularly intriguing at later stages of neurogenesis when the radial glial fibers span broad zones of migratory and postmigrato-

REFERENCES

ry neurons. D e p e n d i n g on whether the radial glial process is maintained during division, one or both daughter cells must e l a b o r a t e a long process to the pia during the transition from M to G1 phases. Observation that radial glial fibers are organized in fascicles 3A'6A1 raises the possibility that the elongating radial glial processes might themselves be guided in their ascent along other radial glial fibers. In o r d e r to characterize morphological transformations during mitotic cycling, high resolution microscopy of immunostained radial glia with detailed reconstructions should be performed.

ACKNOWLEDGEMENTS W e thank A n n e Boiteau and Julie W h i t c o m b for technical assistance. L - P . M . is a F r a n k Boas Scholar at H a r v a r d University s u p p o r t e d by N A T O G r a n t 27B85BE and the F o n d a t i o n Princesse Marie-Christine. S u p p o r t e d also by N I H Grants HD21018 and EY06080 ( M . A . E . ) , H D 04147 ( M . Y . ) and NS 12005 (V.S.C.).

cell classes in the developing mammalian nervous system, J. Neurosci., 5 (1985) 3310-3328.

1 Angevine, J.B. and Sidman, R.L., Autoradiographic study of cell migration during histogenesis of the cerebral cortex, Nature (London), 192 (1961) 766-768. 2 Caviness, V.S., Jr., Neocortical histogenesis in normal and reeler mice: a developmental study based upon [3H]thymidine autoradiography, Dev. Brain Res., 4 (1982) 293-302. 3 Edwards, M.A., Yamamoto, M, Schwarting, G. and Caviness, V.S., Jr., Development of radial glia in the mouse; an immunohistochemical study with a cell-class specific monoclonal antibody, Soc. Neurosci. Abstr., 12 (1986) 182. 4 Edwards, M.A., Yamamoto, M., Schwarting, G. and Carlhess, V.S., Jr., Organization of radial glia and related cells in the developing murine CNS: an analysis based upon a new monoclonal antibody marker, Neuroscience, submitted. 5 Fujita, S., Mitotic pattern and histogenesis of the central nervous system, Nature (London), 185 (1960) 702-703. 6 Gadisseux, J.F. and Evrard, P., Glial neuronal relationship in the developing central nervous system: a histochemical elcctron microscope study of radial glial cell particulate glycogen in normal and reeler mice and the human fetus, Dev. Neurosci., 7(1985) 12-32. 7 Hinds. J.W. and Ruffett, T.L., Cell proliferation in the neural tube: an electron microscopic and Golgi analysis in the mouse cerebral vesicle. Z. Zellforsch., 115 (1971) 226-264. 8 His, W., Die Neuroblasten und deren Entstehung im Embryonal Marke, Abh. Math. Phys. CI. Kgl. Sach. Ges. Wiss., 15 (1889) 313-372. 9 Hockfield, S. and McKay, R.D.G., Identification of major

10 Kauffman, S.L., Lengthening of the generation cycle during embryonic differentiation of the mouse neural tube, Exp. CellRes., 49 (1968) 420-424, 11 Levitt, P. and Rakic, P., Immunoperoxidase localization of glial fibrillary acidic protein in radial glial cells and astrocytes of developing rhesus monkey brain, J. Comp. Neurol., 193 (1980) 815-840. 12 Levitt, P., Cooper, M.L. and Rakic, P., Coexistence of neuronal and glial precursor cells in the cerebral ventricular zone of the fetal monkey: an ultrastructural immunoperoxidase analysis, J. Neurosci., 1 (1981) 27-39. 13 Levitt, P., Cooper, M.L. and Rakic, P., Early divergence and changing proportions of neuronal and glial precursor cells in the primate ventricular zone, Dev. Biol., 96 (1983) 472-484. 14 Misson, J.-P., Edwards, M.A., Yamamoto, M. and Caviness, V.S., Jr., Origin of radial glial cells in mouse cerebral cortex: a study using combined immunohistochemical and [3H]thymidine labeling, Soc. Neurosci. Abstr., 12 (1986) 1585. 15 Pinto-Lord, M.C., Evrard, P. and Caviness, V.S., Jr, Obstructed neuronal migration along radial fibers in the neocortex of reeler mouse: a Golgi-EM analysis, Dev. Brain Res., 4 (1982) 379-393. 15a Pixley S.R.K., de Vellis, J., Transition between immature radial glia and mature astrocytes studied with a monoclonal antibody to vimentin, Dev. Brain Res., 15 (1984) 201-209. 16 Rakic, P., Mode of cell migration to the superficial layers of the fetal monkey neocortex, J. Comp. Neurol., 145 (1972) 61-84.

I ~)() 17 Ramon v Cajal, S., tfistologie du Systeme Nerveux de I'Homme et des Vertebres. Reprinted by Consejo Superior de lnvestigaciones Cientificas, Paris, 1955, Vol. I]. 18 Sauer, F.C., Mitosis in the neural tube, J. Cornp. Neurol.. 62 (1935) 377-405. 19 Schaper, A., The earliest differentiation in the central nervous system of vertebrates, Science, 5 (1897) 430-431. 20 Schmahl, W., Developmental gradient of cell cycle in the telencephalic roof of the fetal NMRI-mouse, Anat. Embryol., 167 (1983) 355-364. 21 Schmechel, D.E. and Rakic, P,, A Golgi study of radial glia] cells in developing monkey telencephalon: morphogenesis and transformation into astrocytes. Anat. Embryol., 156 (1979) 115-152. 22 Schmechel, D.E. and Rakic, P., Arrested proliferation of radia[ glial cells during midgestation in rhesus monkey, Nature (London), 277 (1979) 303-305. 23 Schultze, B. and Korr, H., Cell kinetic studies of different cell types in the developing and adult brain of the rat and

the mouse: a review. Cell Tissue Kinet 14 t 1981) 309-325. 24 Seymour, R.M. and Berry, M.. Scanning and transmission electron microscope studies of interkinetic nuclear m~gration in the cerebral vesicles of the Fat~ J. Comp. Neurol.. I60 (1975) 105-126. 25 Sidman. R.I_.. Miale. [.L. and Feder, N.. Cell proliferation and m~gration in the primitive ependymal zone: an auroradiographic study of histogenesis m the nervous system. Exp. Neurol., 1 (19591 322-333. 26 Smart. I.H.M.. Proliferative characteristics of the ependyreal layer during the early development of the mouse neocortex: a pilot study based on recording the number, location and plane of cleavage of mitotic figures. J. Anat.. 116 (1973 109-129. 27 Smart. |.H.M and McSherry G,M.. ¢,irowth pattern in thc lateral wall of the mouse telencephalon. Histological changes during and subsequent to the period of isocortical neuron production, J. Anat. 134 (1982] 415-442.