The extracellular space in the developing rat brain: its variation with changes in osmolarity of the fixative, method of fixation and maturation

Extracellular space parameters in the rat neocortex and subcortical white matter during postnatal development determined by diffusion analysis

Extracellular space volume fraction, tortuosity and nonspecific uptake of tetramethylammonium--three diffusion parameters of brain tissue--were measured in gray matter of the somatosensory neocortex and subcortical white matter of the rat during postnatal development. The three parameters were determined from concentration-time profiles of tetramethylammonium in postnatal days 2-120 in vivo. Tetramethylammonium concentration was measured with ion-selective microelectrodes positioned 130-200 microns from an iontophoretic source. Data were correlated with cytoarchitectonic structure and average thickness of the regions in 0-90-day-old rats using rapidly frozen tissue. Extracellular space volume fraction was largest in the newborn rats and diminished with age. In two-to three-day-old animals, volume fraction (mean +/- S.E.) was 0.36 +/- 0.04 in layers III and IV, 0.38 +/- 0.02 in layer V, 0.41 +/- 0.01 in layer VI and 0.46 +/- 0.01 in white matter. The earliest decrease in volume fraction was found in layers V and VI at postnatal days 6-7 followed by a decrease in layer III and IV at postnatal days 8-9 and in white matter at postnatal days 10-11. A further dramatic reduction in volume fraction occurred in all cortical layers and especially in the white matter between postnatal days 10 and 21. There was no further decrease in volume fraction between postnatal day 21 and adults (90-120 days old). The adult volume fraction values were: layer II, 0.19 +/- 0.002; III, 0.20 +/- 0.004; IV, 0.21 +/- 0.003; V, 0.22 +/- 0.003; VI, 0.23 +/- 0.007; white matter, 0.20 +/- 0.008. Values of tortuosity ranged between 1.51 and 1.65, nonspecific cellular uptake varied from 3.3 x 10(-3)/s to 6.3 x 10(-3)/s. The variations in each parameter were not statistically significant at any age. These data represent the first characterization of diffusion parameters in a developing brain. They confirm previous histological indications of a relatively large extracellular volume fraction during early postnatal development. The constancy of the tortuosity shows that diffusion of small molecules is no more hindered in the developing brain than in the adult. The large extracellular space volume fraction of the neonatal brain could significantly dilute ions, metabolites and neuroactive substances released from cells, relative to release in adults, and may be a factor in preventing anoxia, seizure and spreading depression in young animals. The diffusion characteristics could also play an important role in the developmental process itself.

The earliest ultrastructural development of the tangential vestibular nucleus in the chick embryo

The tangential nucleus is a primary vestibular nucleus located where the vestibular fibers enter the medulla. It is composed of neurons that migrate between 5 and 8 days in the chick embryo. Although primary vestibular fibers enter the medulla at 3 days, the first synapses are formed at 5 days on the processes of neuron precursors by longitudinally coursing fibers. Since the major components, or their precursors, are present at 3 days within the presumptive nucleus, we are interested in determining what cellular interactions occur among these structures following their entry and during the time leading up to synapse formation. At 2 days, prior to the appearance of VIIth and VIIIth nerve fibers in the medulla, the tangential nucleus anlage contained processes and endfeet of primitive epithelial cells, separated from each other by enlarged extracellular spaces. Longitudinal fibers first appeared within these spaces coincident with the appearance of root fibers, including some identified VIIth motor axons, associated with the primordial VII/VIIIth ganglia. By 3 days, some vestibular and VIIth nerve fibers could be identified by their ultrastructure and relative positions within the marginal zone and nerve roots. However, it was not until 4 days that the presumptive tangential nucleus acquired its orderly, characteristic organization. Although synapses were rare from 2 to 4 days, attachment plaques and coated pits were observed commonly between structures, especially between future synaptic structures. Thus, we confirm that synapse formation begins at 5 days. This represents the first detailed ultrastructural study of cranial sensory nerve ingrowth into the medulla.

Preparation of fetal rat brains for light and electron microscopy

To study cellular shapes, growth patterns, and fine structure during early stages of CNS development in rat embryos, preparative procedures were evaluated and modified to meet two criteria: 1) Coronal semithin sections should reveal undeformed telencephalic hemispheres that were symmetrically expanded on both sides of midline structures and were surrounded by contiguous mesenchyme. 2) In electron micrographs, cells should have intact, undistorted surface membranes, evenly distributed nucleoplasm and well preserved cytoplasmic organelles. To meet these criteria, 378 fetuses with a gestational age of 11-20 days (E11-E20) were used to test and modify procedures for anesthesia, embryo removal and handling, dissection, fixation, dehydration, and embedding of the embryonic CNS. Most specimens were in an early stage of development (E11-E13), which, in case of the neopallial wall, is the preneural period. The tests produced methods that met the above criteria and identified the most common artifacts and their causes. Deformities of the cerebral hemispheres and separations between the brain and its coverings were usually caused by trauma during embryo removal and during handling before fixation. Changes in cellular volumes, especially swelling during fixation and dehydration, were the most important causes of histological artifacts. The procedures and methods that consistently produced the best light and electron microscopic preservation of the E11-E13 rat CNS are described. Fixation was best when the brains were treated with glutaraldehyde and s-collidine buffer, followed by osmium tetroxide in s-collidine buffer. A surprisingly beneficial effect of sodium chloride in the dehydrating alcohol was noted.

A morphometric analysis of extracellular space in the developing spinal cord of the chick embryo

The developmental changes in the amount and distribution of the expanded extracellular space (ECS) (i.e. wider than 100 nm) were analyzed in the cervical spinal cord of chick embryos between stage 9 and 29, using electron micrograph montages, which cover one half of the cross-sectional area of the cord. The percentage of the ECS expansion to the whole cross-sectional area of the cord was 11.0% at stage 9, 7.7% at stage 11, 7.8% at stage 15, and 9.7% at stage 17. It decreased markedly to 3.0% at stage 22 and 1.3% at stage 29. The highest percentage at stage 9 may reflect the dynamic structural changes associated with neural groove closure which takes place around this time. The marked decrease after stage 22 is associated with the rapid overall growth of the cord. Until stage 19, the ECS expansions were mostly elongated and arranged radially with respect to the central canal. The ECS became scarce and arranged randomly thereafter. Throughout the stages examined, especially between stages 17 and 19, percentage was higher in the outer half of the cord than in the inner half. The outer glial limiting membrane was not established by stage 29. Between stages 17 and 22, the percentage was higher in the dorsal region than in the ventral region. This appears to be associated with the regional difference in neuronal maturation. The first blood vessels penetrated the ventromedial portion of the cord around stage 22, where the ECS expansions were relatively scarce. The successive rapid decrease in the amount of ECS expansions can be correlated to the development of vascularization.

Observations of vascularization in the spinal cord of mouse embryos, with special reference to development of boundary membranes and perivascular spaces

The early stages of vascularizations of the spinal cord of the mouse were studied by graphic reconstruction techniques and electron microscopy. Vascular sprouts arise from the perineural vascular plexus (PNVP) to invade the cord of 10-day embryos. These enter the cord most frequently via the lateral surface between the dorsal root and the ventral root; less frequently, they enter via the ventral and/or dorsal surfaces and anastomose with sprouts that have entered via the lateral surface. During the development of intramedullary blood vessels there are essential changes both in the basal laminae covering the neural parenchyma of the cord and in the relationship between the neural tissue and vascular walls. The basal laminae of the developing spinal cord were classified into three categories. The first is the perineural, external, or primary neural, basal lamina (PNBL), which is the earliest of the three in formation and covers the entire external surface of the cord. The second one is the internal, or secondary neural, basal lamina (INBL), which invests the internal surface of the neural tissue facing the walls of invading blood vessels. The third type is the perivascular basal lamina (PVBL), which surrounds the vascular wall. Blood vessels enter the spinal cord by penetrating the PNBL. Since the PVBL and INBL are absent or incomplete in early stages of vascularization, the neural tissue is in direct contact with intramedullary blood vessels. However, following their development, boundary membranes are formed, separating the neural tissue from neighboring vessels, a situation characteristic of capillaries in the mature CNS. Perivascular spaces are seen along the course of developing vessels and secondarily become continuous with the extramedullary connective tissue space. They are neither artifact nor intramedullary extensions of extramedullary connective tissue space along invading sprouts. The boundary membranes are formed by connection of membrane plaques or by fusion of the INBL and PVBL.

A fibronectin-like molecule is present in the developing cat cerebral cortex and is correlated with subplate neurons

The subplate is a transient zone of the developing cerebral cortex through which postmitotic neurons migrate and growing axons elongate en route to their adult positions within the cortical plate. To learn more about the cellular interactions that occur in this zone, we have examined whether fibronectins (FNs), a family of molecules known to promote migration and elongation in other systems, are present during the fetal and postnatal development of the cat's cerebral cortex. Three different anti-FN antisera recognized a single broad band with an apparent molecular mass of 200-250 kD in antigen-transfer analyses (reducing conditions) of plasma-depleted (perfused) whole fetal brain or synaptosome preparations, indicating that FNs are present at these ages. This band can be detected as early as 1 mo before birth at embryonic day 39. Immunohistochemical examination of the developing cerebral cortex from animals between embryonic day 46 and postnatal day 7 using any of the three antisera revealed that FN-like immunoreactivity is restricted to the subplate and the marginal zones, and is not found in the cortical plate. As these zones mature into their adult counterparts (the white matter and layer 1 of the cerebral cortex), immunostaining gradually disappears and is not detectable by postnatal day 70. Previous studies have shown that the subplate and marginal zones contain a special, transient population of neurons (Chun, J. J. M., M. J. Nakamura, and C. J. Shatz. 1987. Nature (Lond.). 325:617-620). The FN-like immunostaining in the subplate and marginal zone is closely associated with these neurons, and some of the immunostaining delineates them. Moreover, the postnatal disappearance of FN-like immunostaining from the subplate is correlated spatially and temporally with the disappearance of the subplate neurons. When subplate neurons are killed by neurotoxins, FN-like immunostaining is depleted in the lesioned area. These observations show that an FN-like molecule is present transiently in the subplate of the developing cerebral cortex and, further, is spatially and temporally correlated with the transient subplate neurons. The presence of FNs within this zone, but not in the cortical plate, suggests that the extracellular milieu of the subplate mediates a unique set of interactions required for the development of the cerebral cortex.

Differential adhesivity of neuroepithelial cells and pioneering circumferential axons

To determine the initial growth pattern of pioneering axons and investigate the factors that may influence their guidance, the lateral margin of a stage 16+ chick brachial spinal cord was examined in serial thin sections. The specimen was prepared with hypertonic fixative which partially shrank the tissue and increased extracellular space. The retention of surface contact after shrinkage was used as an index of the relative adhesiveness between cells in situ. Six axons and growth cones were found within the reconstructed tissue; five were oriented dorsoventrally and one apparent motor neuron growth cone was oriented radially. The five circumferential axons originated from presumptive interneurons distributed in a dispersed pattern along the neural tube lateral wall. Four terminated with growth cones, and each extended a short distance (less than 30 microns) ventrally along the outer margin. No contact was found between these nonfasciculating axons or growth cones. Thus, the earliest intracentral axons appear to grow dorsoventrally from the outset with no appreciable wandering. Morphological features that may indicate their mechanism of guidance, including preformed cellular guides, extracellular channels, contact with basal lamina, and intercellular junctions were not found. The preferential retention of surface contact between adjacent endfeet, as well as between pioneering circumferential axons and neuroepithelial cells, suggests that these particular surfaces are mutually adherent. These findings are consistent with a proposed dorsal-to-ventral adhesive gradient mechanism of circumferential axonal guidance.

Morphological effects of chronic haloperidol administration on the postnatal development of the striatum

The aim of this paper is to describe the morphological changes induced in the striatum after the administration of haloperidol during the first postnatal month, a period in which a lack of tolerance to treatment with neuroleptics has been reported. At the end of the treatment several morphological parameters were evaluated including neuron size and density and the synaptic profile areas of cross-sectioned dendrites and axon terminals. The results evidenced a loss of the smallest dendritic profiles without the rest of the parameters examined being affected. This response differs from the one observed in the adult rat striatum that does develop tolerance to haloperidol. It seems to more closely correspond to the changes found in the prefrontal cortex, a region that does not develop tolerance after chronic treatment with neuroleptics.

Postnatal development of the feline lateral cervical nucleus: I. A quantitative light and electron microscopic study

With the aim of obtaining some basic information for future developmental studies, the lateral cervical nucleus (LCN) was investigated in 32 kittens of different ages by electron microscopic and stereologic methods. Corresponding light microscopic measurements of neuronal and nuclear profiles and of the total LCN volume were also performed. The total LCN volume increased sixfold between the ages of 12 hours and 120 days, the most rapid increase occurring during the first month. The neuronal size was fairly constant up to the age of 9 days, whereafter it showed greater variation. The mean profile area increased rapidly during the second week and then more slowly. The relative volume of boutons increased significantly between birth and the age of 34 days and then decreased slightly up to 120 days postnatally. The total bouton volume showed a steady increase, which was most pronounced between the ages of 9 and 34 days. The relative dendritic volume decreased during the 120 days of observation, whereas the total volume of dendrites increased up to the age of 92 days and then decreased. The total volume of glial cells increased during the 120-day observation period, as did both the relative and total volumes of myelinated axons. The changes in the relative volumes of mitochondria in boutons and dendrites were very similar, with increases that were most marked between the ages of 9 and 34 days and between 92 and 120 days.

The chronology of lesion repair in the developing rat brain: biological significance of the pre-existing extracellular space

We observed the histological peculiarities of the repair process in a destructive lesion of the developing rat brain during neurogenesis. Degeneration was induced selectively in certain cells of the proliferating phase in the rat fetal neopallium on embryonic day 16 by transplacental administration of ethylnitrosourea. Successive elimination of necrotic cells and the restoration process were observed. The repair process was divided into the following steps: elimination of individually affected cells by phagocytes in the pre-existing extracellular space; successive restoration of the disintegrated area by cells which differentiated from remaining matrix cells. No reactive gliosis, fibrosis, abnormal vascularization or infiltration of granulocytes and lymphocytes was observed at any time. The thinned neopallium on postnatal day 21 revealed only a small number and abnormal distribution of the cortical neurons. It may be assumed that the fetal brain owes its unique repair features to the presence of a vast extracellular space under normal conditions. In this pre-existing extracellular space, every kind of cell seems to exist separately without the intracellular adhesions characteristic of the adult brain. When degeneration occurs in certain cells the phagocytes would be able to eliminate the degenerate cells completely in this space without having to break intercellular adhesions. As a result, after the completion of cell elimination, the injured brain is restored to its original state with no cell reaction, giving the appearance of a small brain with normal-looking histological architecture, save only for the sparseness of cells.

Extracellular matrix during laminar pattern formation of neocortex in normal and reeler mutant mice

The spatial and temporal distribution of extracellular matrix, which occupied the large extracellular spaces in the developing cerebral cortex, was studied during pre- and perinatal ontogenesis of normal and reeler mutant mice. Colloidal iron-staining material was localized principally in the marginal zone and subplate of normal mice, whereas in reeler mutants, most of the material was found in the outer layers of the cortex. Patterns of extracellular matrix localization in both genotypes followed the laminar pattern formation of cerebral cortex architecture. Histochemical ultrastructural visualization of this extracellular matrix and its susceptibility to enzymatic treatment suggested that the major components are glycosaminoglycans. Their possible role in relation to afferent axon targeting is discussed.

Early development of the brain and spinal cord in dysraphic mice: a transmission electron microscopic study

The hindbrain and spinal cord were studied by transmission electron microscopy with and without lanthanum nitrate treatment in nine-day embryos of the loop-tail (Lp) mutant mouse. Homozygous (Lp/Lp) individuals exhibit dysraphism from the hindbrain caudally throughout the embryo; in +/+ and Lp/+ individuals, the brains and spinal cords are normal. In contrast, the ventricular cells in the abnormal hindbrain and spinal cord showed increased amounts of intercellular space in an area intermediate between the luminal border and basal zone, and a flattening which occurs variably in their luminal surfaces. A most striking difference occurred in the frequency of gap junctional vesicles, circular structures bounded by a double membrane and containing ribosome-like material. A quantitative analysis of the distribution of these organelles revealed that they are more numerous in the dysraphic hindbrain and lumbosacral spinal cord of the abnormal animals than in comparable regions of the normal; however, in the cervicothoracic spinal cord, the frequency of these vesicles is similar. In specimens treated with lanthanum, the tracer freely penetrated the luminal junctional complexes in both normal and abnormal animals, but was not present between the membranes of the gap junctional vesicles. The developmental significance of gap junctional vesicles frequently located in juxtaluminal regions of mitotic cells, but also found deeply in dividing cells, is not known; they may relate to cell-to-cell attachment and/or communication. In any event, further study of them may prove valuable in understanding normal and abnormal development of the neural tube.

The role of the blood-brain barrier in perfusion fixation of the brain for electron microscopy

Regions of the brain vascularized by capillaries of the blood-brain barrier (BBB) type require a different fixative from regions which have capillaries of the endocrine type. Fixative with isotonic buffer gives excellent ultrastructural preservation in the BBB regions, but cause severe shrinkage of cells in the endocrine regions. This is evidently due to the difference in the permeability of the capillary walls to solutes in the fixative. In the BBB regions in less permeable capillaries do not allow outflow of osmotically active particles to a harmful extent, whereas in the endocrine regions osomotic imbalances are created between the intra- and extracellular compartments. The diffusion rate of the fixative and the final volume of the fixed brain depend on the balance between the intravascular and intersitial hydrostatic and oncotic pressures across the capillary wall during the perfusion fixation, as those pressures regulate the amount of perfusate that will enter the parenchyma. Generally, as high a perfusion pressure as possible is recommended to obtain effective wash-out of blood and rapid diffusion of fixative into the tissue. Addition of macromolecules (2% PVP, mol. wt. 40,000) into the fixative slightly improved the ultrastructural preservation in the BBB regions of the central nervous system.