Organizational Events

Subplate Neurons

The importance of the subplate neurons in cerebral organizational events was defined by an elegant series of studies both in experimental animals and the human brain. Cells destined to be subplate neurons are generated in the germinative zones and migrate both radially and tangentially to the primitive marginal zone at approximately 7 weeks of gestation before generation and migration of neurons of the cortical plate ( Fig. 7.1 and Table 7.2 ). Initially, these cells are part of the preplate that is split by approximately 10 weeks of gestation by the developing cortical plate into the subplate neurons below and the Cajal-Retzius neurons of the marginal zone above. (The cortical plate neurons give rise to layers II through VI of the cerebral cortex.) Through human fetal development, the subplate zone shows changes in laminar organization, proceeding from a monolayer in the early fetal period (preplate) to a bilaminar structure between 13 and 15 weeks of gestation and to a trilaminar appearance at 15 to 21 weeks of gestation. The trilaminar structure shows subtle sublamination with deep, intermediate, and superficial “floors” and persists until approximately 28 weeks of gestation when it gradually begins to lose its sublaminar organization to become the single subplate zone of the newborn brain. The subplate zone contains some of the earliest born neurons of the cerebral cortex, and during this developmental period these neurons are more mature than those of the overlying cortical plate. The subplate neurons rapidly exhibit morphological differentiation and express a variety of receptors for neurotransmitters (gamma-aminobutyric acid [GABA], excitatory amino acids), neuropeptides, and growth factors (nerve growth factor, neuropeptide gamma, somatostatin, calbindin).

The subplate neurons elaborate a dendritic arbor with spines, receive synaptic inputs from ascending afferents from thalamus and distant cortical sites, and extend axonal collaterals to overlying cerebral cortex and to other cortical and subcortical sites (thalamus, other cortical regions, corpus callosum). In addition to the subplate neurons , the subplate zone contains other tissue components, such as radial glial processes, radially and tangentially migrating neurons, early developing astrocytes, microglia, and oligodendrocyte precursors. Yet this zone is distinguished by an extensive extracellular space that is filled with hydrophilic extracellular matrix (ECM) and heterogeneous contingents of waiting cortical afferents and transient synapses. This hydrophilic feature underlies the visibility of the zone in T2-weighted magnetic resonance imaging (MRI) scans of the human fetus from approximately 18 to 26 weeks of gestation ( Fig. 7.2 ). During development, the changes in laminar structure discussed above occur concomitant with changes in distribution and characteristics of neuronal, glial, synaptic, and ECM markers within these lamina, consistent with a role for these changes in the navigation and establishment of cortical circuitry.

The subplate zone is also molecularly distinct , as determined by the identification of subplate-enriched genes by transcriptome profiling of different fetal layers. Gene expression profiling of the subplate zone in midgestation fetal human brains indicates that the human subplate is functionally enriched for synaptic plasticity and generally shows signs of more advanced maturity compared with the overlying cortical plate. Thus some of the molecular hallmarks of the subplate zone during early development primarily relate to cell maturity, and as subplate cells form, they extend axons and receive synaptic inputs earlier than the cortical plate. Importantly, the subplate is specifically rich in chondroitin sulfate proteoglycans, and the subplate transcriptome is enriched for genes involved in the production of ECM and proteoglycans. Chondroitin sulfate proteoglycans are known to interact with laminin, fibronectin, tenascin, and collagen, and their differential distribution supports a role in axonal pathfinding and cell migration.

The functions of the subplate neurons now appear to be particularly far-reaching (see Table 7.2 ). Thus they provide a site for synaptic contact for axons ascending from thalamus and other cortical sites, termed waiting thalamocortical and corticocortical afferents because their neuronal targets in the cortical plate have not yet arrived or differentiated. These afferents presumably would undergo degeneration if they did not have the subplate neurons as transient targets. Moreover, the subplate neurons have been shown to establish a functional synaptic link between these waiting afferents and their cortical targets. This link could exert a trophic influence on the cortical neuronal targets by the release of neuropeptides or excitatory amino acid neurotransmitter by the subplate axon terminals. There are additional genes with subplate-restricted expression in the cortex that encode secreted proteins, including Serpini1 (which encodes neuroserpin) and neuronal pentraxin 1 (Nptx1) . These two proteins are neural specific, with proposed roles in synaptic function or maturation. Thus the subplate may additionally influence cortical circuit formation through a transient secretory function. A third function appears to be the guidance by subplate axons entering cerebral cortex of the ascending axons to their targets. Indeed, if the subplate neurons are eliminated, thalamocortical afferents destined for the overlying cortex fail to move superiorly into the cortex at the appropriate site and continue to grow aimlessly in the subcortical region. A fourth function of subplate neurons is involvement in cerebral cortical organization; for example, ocular dominance columns in visual cortex fail to develop if underlying subplate neurons are eliminated during development. Related to this role is the importance for subplate neurons in cortical synaptic development and function. A fifth function appears to be mediated by the descending axon collaterals from the subplate neurons; these collaterals appear to pioneer or guide the initial projections from cerebral cortex toward subcortical targets (e.g., thalamus, corpus callosum, and other cortical sites). Finally, transient synaptic transmissions from subplate neurons to migrating multipolar neocortical neurons lead to a multipolar-to-bipolar morphological transition as well as transition from slow multipolar migration to faster radial glial-guided locomotion. This function of subplate neurons to control radial migration of neocortical neurons is regulated, in part, by N -methyl- d -aspartate (NMDA) receptor–mediated Ca ²⁺ signaling.

Concomitant studies of subplate neurons of developing human cerebral cortex provide a crucial link with the experimental studies ( Fig. 7.3 ). The subplate neuron layer in human cortex reaches a peak between approximately 24 and 32 weeks of gestation. At this peak time, the width of the subplate zone is approximately four times that of the cortical plate. Programmed cell death (apoptosis) of this layer appears to begin generally late in the third trimester, and approximately 90% of subplate neurons have disappeared after approximately the sixth month of postnatal life. Slightly different time courses for peak development and regression of the subplate neurons exist for somatosensory and visual cortices. In the subplate dissolution stage, which occurs in humans at greater than 35 postconceptional weeks, subplate neurons decline in number and the volume of the subplate zone decreases. The reduction in volume reflects primarily a decrease in extracellular space and fewer axon bundles within the subplate zone. A distinct subplate zone is no longer identifiable by about 6 months post term in humans, but large neurons embedded in white matter are thought to be the remaining subplate cells, which are referred to as interstitial white matter neurons.

Lamination

Attainment of the proper alignment, orientation, and layering of cortical neurons, defined as lamination, occurs during and following neuronal migration (see Table 7.1 ). These events are among the earliest in cortical organization. The microcircuitry of the cerebral cortex underlying cognitive processing is dependent on precise interrelationships between variable numbers of excitatory pyramidal neurons and inhibitory nonpyramidal (granular) neurons in cortical modules. The cortical layers are specialized compartments that contain neurons with unique properties that underlie specific roles in neural circuitry. The normal cerebral cortex is composed of two main classes of neurons: (1) pyramidal neurons, which comprise 75% to 85% of total cortical neurons, are glutamatergic, and project outside the cortex; and (2) nonpyramidal neurons (interneurons), which comprise 15% to 25% of cortical neurons, are GABAergic, and project within the cortical layers.

The neocortex begins to transform from an undifferentiated cortical plate to a highly specialized structure at around 30 gestational weeks ( Fig. 7.4 ). At this age, the cortical plate becomes composed of six layers in which each layer is characterized by a specific composite of pyramidal and nonpyramidal neurons. Dramatic changes in lamination, laminar thickness, and pyramidal and nonpyramidal cell differentiation and density in the last half of gestation are consistent with neuroimaging findings of marked increases in cortical thickness and surface area over this same time period (see later). Pyramidal neurons are known to originate from radial progenitors (radial glial cells) in the subventricular zone (SVZ); these early precursors produce neurons destined particularly for the deeper cortical layers and reach the cortex by radial migration before the second half of gestation (see Chapter 6 ). The early differentiation of pyramidal neurons in layer V is consistent with their early origin and migration. Over the second half of gestation and into infancy, there is a striking increase in the overall thickness of the cortex and in layers I–III, V, and VI. The increase in thickness in layers I–III in particular over the last half of gestation likely reflects their expansion by late migrating (GABAergic) interneurons, given that the SVZ continues to actively generate mainly GABAergic neurons beyond midgestation (see Chapters 5 and 6 ).

Gyral Development

Gyrification is the process whereby folding patterns of sulci and gyri develop on the surface of the brain. Several of these patterns are asymmetrical between the right and left side of the cerebral hemispheres, and several distinguish the human brain from that of other species. The gyrus is a ridge of cerebral cortex, whereas the sulcus is a depression or furrow on either side of the ridge. Gyrification results in a dramatic increase in the cortical surface area within the limited and rigid confines of the skull, and thus, in the volume of cortical gray matter . As discussed later, a striking increase in cerebral cortical volume in the human infant from approximately 28 to 40 weeks after conception has been shown by quantitative MRI measurements of cortical gray matter volumes. Thus a fourfold increase in cerebral cortical gray matter volume could be documented ( Fig. 7.5 ). The changes in cortical gyral development and cortical surface area that accompany and presumably are caused by the increase in cortical volume can be seen by MRI ( Fig. 7.6 ). The formation of gyri and sulci in the brain also allows for compact wiring that promotes and enhances efficient neural processing.

At the time of neural tube closure in the embryonic period, the surface of the brain is smooth, that is, lissencephalic; as development continues, gyri and sulci begin to take shape on the fetal brain surface ( Fig. 7.7 ). The pattern of gyrification follows orderly and defined sequences, such that the brain can be dated to a particular gestational week by its developmental stage of gyrification. Information about these sequences were garnered from 507 brains and serial sections of 207 brains from infants from 10 to 44 gestational weeks of age in the National Perinatal Collaborative Project. The sequential developmental changes of the individual fissures, sulci, and gyri of the cerebral hemispheres throughout the gestational period were tabulated. The period of greatest development of brain gyrification is during the third trimester of pregnancy , a period in which the brain undergoes considerable growth. At midgestation, the brain is lissencephalic, except for the presence of the Sylvian fissure and central sulcus; between midgestation and birth, all primary, secondary, and tertiary gyri are formed, an explosive period in gyrification.

The gyrification index (GI) , defined as the ratio between the lengths of coronal outlines for the brain including and excluding the sulcal regions, is a quantitative approach to measure gyrification; brains with higher degrees of cortical folding give higher GI values. This measure was used to quantify the developmental trajectory of gyrification in humans with the major finding that gyrification increases dramatically during the third trimester. Though the GI continues to increase beyond birth during infancy, childhood is marked with a decline in GI that continues into adulthood. Despite the stereotypical development of gyrification, it should be emphasized that there are structural variations among individuals , including in gyrification. In a study of 200 healthy young adults, leftward gyrification asymmetries were observed in multiple regions of the brain, including in middle and inferior temporal cortex and lateral paracentral regions. Rightward asymmetries included medial frontal, parietal, and occipital regions. Asymmetry was further examined using MRI analysis of cortical thickness and surface area in a large worldwide cohort of >17,000 individuals. These data revealed widespread asymmetries at the hemispheric and regional levels, with a general thicker cortex but smaller surface area in the left hemisphere compared with the right. Asymmetries were found in regions involved in lateralized functions, including language and visual processing (for example, inferior frontal gyrus and entorhinal cortex). In addition, variability in asymmetry of the medial temporal regions was related to sex, with some regions (e.g., parahippocampal gyrus) being more leftward in males and other regions (e.g., entorhinal cortex) being more rightward in females.

Though the mechanistic basis of cortical folding is complex, leading theories have come to dominate the field based on experimental and computer modeling. The first theory involves differential tangential expansion and points to differential growth rates of adjacent cortical regions as the primary drivers of folding. In cortical development, the outer cortical layer expands faster than an inner layer, resulting in mechanistic instability that is resolved through convolutions on the outer surface. The second hypothesis is the axonal tension hypothesis that posits that cortical folding involves hydrostatic pressure and variations in axonal tension between adjacent cortical regions. In this model, neighboring regions are pulled together by connecting axons, creating the folding patterns. Though the second hypothesis is supported by evidence showing that axons grow in a state of tension, other evidence refutes this growth as a primary driving force of gyration. In addition to physical factors, a third notion involves complex gene expression patterns across regions of the developing cortex that have been shown to play a role in cortical folding.

Neurite Outgrowth

Neurite outgrowth refers to the elaboration of dendritic and axonal ramifications of neurons (see Table 7.1 ). While beginning in the first trimester, it becomes a dominant organizational event in the second half of pregnancy, the neonatal period, and infancy. The most significant early studies of neurite outgrowth in human cerebral cortex were made in 1939 by Conel, whose Golgi-Cox preparations of cerebral cortex from birth to 2 years of age demonstrated progressive enrichment of the dendritic and axonal plexus, with much smaller increases in size and no proportionate increases in the number of individual neurons. The studies of Mrzljak and co-workers showed similar events in frontal cortex before birth (see Fig. 7.1 ). Accompanying the elaboration of dendritic and axonal ramifications are the appearance of synaptic elements, the development of neurofibrils, and an increase in size of endoplasmic reticulum within the cytoplasm of cells. The biochemical correlates of these changes are increasing cerebral content of RNA and protein relative to DNA. Immunocytochemical studies document parallel expression of a variety of neurotrophins, neurotransmitters (NMDA, alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid/ainite, GABA, and glycine receptors), surface glycoconjugates, and cytoskeletal components. The maturational changes occur relatively rapidly in the hippocampus, whereas they occur over a more protracted period in the supralimbic region; the latter, of course, is of great significance, because it is the locus of the major association areas. Dendritic development in the human occurs earlier in thalamus and brainstem than in cerebral cortical regions. These findings were amplified by studies by Purpura, Huttenlocher, Marin-Padilla, Rakic, and other investigators, who used electron microscopic and immunocytochemical methods as well as Golgi techniques.

Perhaps the most striking demonstration of the remarkable increase in dendritic development in the third trimester of human gestation is apparent in the Golgi-Cox studies of human cortical development from 15 to 35 weeks gestation (see Fig. 7.8 ). There is remarkable apical and basilar dendritic development, especially after 24 weeks gestation. This development depends on afferent inputs and correlates with development of neurophysiological activity. The rapidity and complexity of these structural and physiological changes in cortex in the third trimester likely play a crucial role in the vulnerability of these changes to adverse events in premature infants (see later).

One study of developing human brain demonstrated the strikingly active axonal development in the cerebrum over the last trimester of gestation and in the early postnatal period ( Fig. 7.9 ). Thus immunostaining with GAP-43, a protein expressed on growing axons, shows exuberant expression in the cerebral white matter to approximately the subplate region at 20 weeks, to the cortical plate at 27 weeks, within the cortex at 37 weeks, and into the first year of life. This differential pattern may reflect, at 20 weeks, growth of axons from thalamus to subplate neurons and, at 27 weeks, from subplate neurons to cerebral cortex. At 37 weeks, the increase in cerebral cortical expression of GAP-43 may reflect the increase in cortical penetration of thalamic ascending fibers no longer waiting at the subplate layer, in corticocortical fibers, and in descending cortical fibers, initially pioneered by subplate axons (see earlier). These findings are consistent with more recent studies with different techniques by Kostovic and Jovanov-Milosevic. Although more data are needed on these issues, it is clear that the last trimester of human gestation is a period of rapid axonal development .

The relationship between neurite outgrowth in cortex and development of functional capacity can be illustrated in human visual cortex during the third trimester ( Fig. 7.10 ). Most impressive are the appearance and elaboration of basilar dendrites and the tangential spread of apical dendrites. This dendritic development is accompanied by the appearance of dendritic spines or, in other words, sites of synaptic contact (see the subsequent section on synaptic development). These anatomical expressions of differentiation are paralleled by the neurophysiological expression of maturation of the visual evoked potential (see Fig. 7.10 ). Such detailed relationships between dendritic structural development and specific details of neurophysiological development have been studied in depth in developing animals.

The progress of dendritic differentiation depends on the establishment of afferent input and presumably synaptic activity . In certain developing neural systems, the importance of receiving and making proper connections has been emphasized as highly critical for further organization. At least part of the influence of these connections is mediated by the functional activity generated through them (see earlier discussion of subplate neurons). This role of functional activity has implications for the effects of a variety of environmental stimuli on the postnatal progress of organizational development. Neuronal activity initiates its effects on dendritic development by inducing calcium influx, through both activation of glutamate receptors, principally the NMDA receptor but likely also GluR2-deficient alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors, and by opening of voltage-dependent calcium channels. The calcium-mediated effects include both a direct effect on the actin and microtubular components of the cytoskeleton and on several adhesion molecules and major indirect effects by activating multiple signaling pathways that target nuclear transcription factors and thereby many genes involved in dendritic development. Studies of developing human cerebral cortex show transient exuberant expression of calcium-permeable glutamate receptors in cortical neurons during this perinatal period.

Axonal differentiation depends on both internal properties and on responses to environmental cues. The central axon is a smooth, thin process of variable length that extends from the polarized neuronal cell body and propagates action potentials, electrical signals or impulses that travel the length of an axon to convey information or generate a response. The neuronal cytoskeleton of the axon is composed of three different types of filaments: actin microfilaments, microtubules, and intermediate filaments . A mechanism for transport down the axon involves a system of both anterograde transport from the cell body to the terminal and retrograde transport from the terminal to the cell body. Once produced in the cell body, membrane-bound organelles, including mitochondria and secretory vesicles, are transported down the length of the axon using microtubules and the motor molecule kinesin, which is thought to walk molecules along the microtubule. The growth cone is formed at the tip of the growing axon and is responsible for sensing environmental cues, determining the direction of growth, and guiding the growing axon in the proper direction. The temporal and spatial movement of the axon to its appropriate target is critical to the proper development of neuronal circuits . This movement is controlled by a number of different environmental cues that direct the axons via interactions with the growth cone receptor. Signals include a number of molecules categorized into three families of ligands—cell adhesion molecules, ECM molecules, and ephrins—each of which acts through receptor-mediated interactions on the cell surface of the growth cone. Other signaling factors include the netrins, slits, and semaphorins. Though many of the signals are attractive guidance cues, some are inhibitory or repellent. Glia are important sources of many of these signals, because glial functions include axonal growth and guidance (see Table 7.5 ).

Advanced diffusion-based MRI imaging known as tractography provides further insight into the changes in fiber tract development in the living premature infant (see later). Intracerebral connectivity emerges in a sequential manner, starting from dorsal posterior brain areas, continuing in an anteroventral direction, and ending in inferior temporal and inferior frontal lobes ( Fig. 7.11 ). This observed order follows the same order of normal gyrification and myelination. Corticocortical association fibers have been found in the cortical plate as early as 23 to 25 gestational weeks. Decreases in ipsilateral corticocortical connections observed after 22 gestational weeks may be due to axonal loss from failure of final targeting or from pruning after initial exuberant overconnectivity.

Synaptic Development

Synapses are the principal sites for communication between presynaptic and postsynaptic neurons via chemical messengers termed neurotransmitters (see Table 7.1 ). In the human brain, about 10 ¹⁵ synaptic contacts interconnect the 10 ¹⁰ to 10 ¹¹ neurons. Synaptic formation differs appreciably among brain regions in the human brain. In the brainstem, the number of dendritic spines, the sites of synaptic contacts, in the medullary reticular formation reaches a peak at 34 to 36 weeks of gestation and declines rapidly after birth (see later discussion of disorders of organizational events). In the cerebrum, synapses are observed initially on neurons of the preplate at 4 to 5 gestational weeks and by 10 weeks on neurons in subplate and marginal zones. In the hippocampus, synapses are abundant as early as 15 and 16.5 weeks of gestation ( Table 7.3 ). The earliest synapses in the cerebral cortex are observed around 18 postconceptional weeks and are located on prospective layer V neurons (pyramidal neurons). Shortly after the pronounced ingrowth of thalamocortical axons to the cortical plate around 21 postconceptional weeks, dendritic spines begin to appear on immature pyramidal neurons and interneurons between 24 and 27 weeks.

With Golgi preparations, Purpura and co-workers defined the subsequent progression of dendritic spine development in the human cortex from the fifth month of gestation ( Figs. 7.12 and 7.13 ). Initially, dendrites appear as thick processes with only a few fine spicules. As development progresses, a great number and variety of dendritic spines appear. In visual cortex, synaptogenesis is fastest between 2 and 4 months after term, a time also critical for the development of function in visual cortex, and maximal synaptic density is attained at 8 months ( Fig. 7.14 ). Synapse elimination or pruning then begins, and by age 11 months, approximately 40% of synapses have been lost. In frontal cortex, the time course of synaptic formation and of elimination differs somewhat from that in visual cortex; maximal synaptic density is reached at approximately 15 to 24 months, and synapse elimination, although reaching the same loss of 40%, is more gradual. In the prefrontal cortex, synapse elimination extends into midadolescence. Elegant studies in the monkey exhibit more uniformity in the temporal features of synaptogenesis among cortical regions, but the basic principles are formation of earliest synapses in the marginal and subplate zones, an increase in synapses in cortical plate to a peak in excess of the adult number, and a subsequent period of synaptic elimination. Synaptic function results in the development of such neurophysiological measures as cortical evoked responses (see Fig. 7.10 ). Function is dependent on the action of neurotransmitters. These substances are stored in vesicles and are released from presynaptic nerve terminals at the so-called active zone, a restricted area of the cell membrane situated exactly opposite to the postsynaptic neurotransmitter reception apparatus. Communication between neurons is characterized by spatial and temporal control of soluble N -ethylmaleimide-sensitive factor attachment protein receptor (SNARE) complexes at the presynaptic site. These complexes contain a large family of membrane associated proteins that are required for the regulated membrane fusion that occurs in synaptic vesicle fusions. Action potential-elicited Ca ²⁺ influx at the presynaptic release site triggers dynamic changes in the SNARE complex at the intersection of the synaptic vesicle and plasma membrane to drive fusion of the membranes and release of the neurotransmitters into the active zone or synaptic cleft. Once the neurotransmitter has been released, the synaptic vessel material, including the SNARE complexes, is recycled.

At the postsynaptic site, released neurotransmitters bind to specific receptors and directly or indirectly control the opening of ion channels and flow of charged ions into or out of the neuron. This leads to a shift of electric charge across the postsynaptic membrane, a change in electric polarization, and ultimately a postsynaptic potential that is either excitatory or inhibitory, increasing or decreasing, respectively, the likelihood of a postsynaptic action potential.

The development of dendritic spines and spine remodeling involves many molecules and signaling pathways. The two principal themes are modulation of the following: (1) ion channels, especially calcium-permeable channels, by neurotransmitters, especially glutamate, and (2) cell surface receptors by a variety of ligands. The intracellular events lead ultimately to effects on actin-binding proteins and the actin cytoskeleton, with resulting changes in spine shape, size, and motility.

The development of synaptic specificity begins once a neuron has identified its correct synaptic partner. The initial axodendritic contact is transformed into a functional synapse by the recruitment of presynaptic and postsynaptic components. The factors that stimulate synaptic formation and development in developing brain include initially activity-independent events (i.e., molecular mechanisms involved in targeting), followed by activity-dependent events occurring after the development of receptors on target neurons and the generation of electrical activity (see previous discussion on neurite outgrowth). Many molecules are known to be involved in synaptic generation and maintenance. However, a complete understanding of these processes has not been acquired, despite advances in technologies such as single-cell transcriptomics and gain- and loss-of-function approaches such as CRISPR. Cell adhesion molecules have been described that may initiate the formation of synapses, hold together the pre- and postsynaptic sides of a synapse, and coordinate the precise alignment of pre- and postsynaptic sites. The means by which these cell adhesion molecules function, however, is not yet clear. Given the precise coordination of synaptogenesis, synaptic specification, synaptic function, and synaptic elimination (see below), the consequences on developing neural function of abnormalities in these processes could be enormous.

Cell Death and Selective Elimination of Neuronal Processes and Synapses

Cell death and selective elimination of neuronal processes and synapses, or regressive events in brain development, are now recognized to be highly critical (see Table 7.1 ). Results of studies in a variety of developing neuronal systems showed that after formation of neuronal collections by the progressive processes of proliferation and migration, cell death occurs. Although variable in degree among neuronal regions, typically about half of the neurons in a given collection die before final maturation (see earlier). This process of cell death is initiated and sustained by the expression of specific genes and their transcription products that actively kill the neuron. Critical in the final phases of the sequence to cell death is the activation of a family of cysteine proteases known as caspases. The term programmed cell death has been used to emphasize that this is an active developmental process, although more commonly the term apoptosis, a Greek-rooted word referring to the naturally occurring seasonal loss or falling of flowers, is used to refer to this developmentally determined cell death.

The factors that activate this death system appear to relate to competition of neurons for limited amounts of trophic factors, generated by the target, afferent input, or associated glia. This loss of neurons appears to serve two major functions in development: quantitative adjustments (numerical matching) of interconnecting populations of neurons and elimination of projections that are aberrant or otherwise incorrect (refinement of synaptic connections or error correction). Failure of cell death or overactivation of this process clearly could have major deleterious implications for brain development and subsequent function.

Although apoptosis is the major form of cell death in the developing nervous system, other forms are also noted to occur, specifically pyroptosis. Pyroptosis, “fiery death,” is an inflammatory form of cell death mediated by the activation of inflammasomes, innate immune complexes whose assembly is driven by exogenous pathogens or endogenous cellular damage, the latter of which is proposed to play a role in developmental cell death. In the context of development, pyroptosis is hypothesized to be activated by DNA damage and thus contribute to the elimination of genetically compromised neurons.

Neural organization is refined further by a second regressive event, selective elimination of neuronal processes and synapses . This event primarily causes the removal of terminal axonal branches and their synapses, although even larger-scale elimination of a total pathway also occurs. Vivid demonstrations of synapse elimination (pruning) are apparent in developing brainstem and cortex of the human infant (see earlier). Neuronal activity plays a central role in synaptic pruning and model systems differentiate between pruning resulting from spontaneous neural activity and sensory experience-driven neural activity, the latter of which involves activity deprivation preceding synaptic elimination. A role for immune molecules as well as microglia and astrocytes in synaptic pruning is an active area of research that bridges both normal development and pathology. Although mechanisms involved in microglial synapsis engulfment are not fully understood, evidence supports a role for neuronal-microglial signaling via synaptic expression of complement proteins and chemokines, specifically fractalkine, and microglial expression of receptors that interact with these factors and mediate the phagocytosis. Like microglia, astrocytes engulf synaptic material but via mechanisms involving astrocyte-specific phagocytic receptors MEGF10 or MERTK. Astrocytes mediate synaptic elimination in additional ways, although the mechanisms involved are not fully understood. Finally, microglia and astrocytes may act together to promote synaptic elimination. One example of this glial cross-talk involves astrocytic cytokine interleukin 33 binding to its receptor, interleukin 1 receptor-like 1, expressed specifically on microglia. This interaction stimulates engulfment and pruning of synapsis.

The observations that cell death and elimination of neuronal processes and synapses occur during the organizational period of development have implications for the frequent demonstration that the plasticity of developing brain decreases as this period is completed. It is likely that the regressive events described in this section are modified when the brain is injured and that neuronal processes and synapses destined for elimination can be retained if needed to preserve function. In addition, new projections may develop in response to injury during the period in which the brain has the capacity to perform organizational events. In favor of one or both of these predictions is the demonstration in both human infants and experimental models that, after neonatal cerebral lesions, ipsilateral corticospinal tract projections can be demonstrated and presumably can ameliorate the functional deficit. The demonstration of an ipsilateral corticospinal projection until early childhood in humans suggests that retention of a normally occurring ipsilateral corticospinal tract, which otherwise is eliminated during development, is the crucial event in this form of plasticity. The possible additional role of assumption of motor functions by ipsilateral cerebrum adjacent to the lesion was suggested by other studies.

Glial Proliferation and Differentiation

Astrocytes, oligodendrocytes, and microglia are the major glial cells of the central nervous system (CNS). Glial proliferation and differentiation are of major importance in the developing brain; glial cells clearly outnumber neurons in the CNS. In fact, in the human cerebral cortex, glial cells outnumber neurons by approximately 1.25 to 1 and are almost the exclusive cell type in white matter. Glial lineage, proliferation, and differentiation have been the topic of intense investigation in experimental systems, and sophisticated molecular biology technology, especially single-cell transcriptomics, has greatly advanced the field. Similarly, detailed molecular data are also emerging from studies of the human brain and are delineating the molecular heterogeneity of these cell types. General observations about glial lineage and differentiation are described in Table 7.4 . In general, astrocytes are generated primarily before oligodendrocytes. The progenitors of both astrocytes and oligodendrocytes initially are cells of the SVZ and probably radial glia (see Chapters 5 and 6 ). Radial glial progenitors may give rise to a glial-restricted progenitor that then generates astrocytes or oligodendrocytes. Proliferation of glia, unlike that of neurons, also may occur locally, during and after migration. Recent data suggest that some astrocytes retain “stemness” or progenitor properties that enable them under certain conditions to reenter the cell cycle and differentiate into other neural cell types. Microglia are mesoderm-derived cells and enter into the developing brain as early as 4.5 gestational weeks, prior to astrogliosis, oligodendrogenesis, and neurogenesis. Microglia are generally identified as ramified or amoeboid in morphology, with these characteristics associated with resting or active states. Microglia have the ability to self-renew through proliferation.

Organizational Events Studied In Vivo

Investigation of organizational events in living infants has been based principally on the study of premature infants by advanced MRI methodologies ( Table 7.6 ). Several principal MRI methods have been used and include three structural measures; that is, volumetric MRI, diffusion tensor MRI, and surface-based cerebral cortical measures, as well as functional MRI (fMRI), both task-related (e.g., response to specific sensory input) and resting state (RS) (see Table 7.6 ). Many excellent reviews of the application of these measures are available and are discussed in other relevant chapters in this book (see especially Chapters 13 and 20 ). This section focuses on studies relevant to normal development in the fetal and neonatal periods . Because of the principal application of these measures to preterm infants thus far, the great preponderance of data involves brain development primarily over the period from 28 to 40 postconceptional weeks. These findings are described briefly next.