The number, types, and distribution of distinct classes of axons and glia in four cerebral commissures of the adult rhesus monkey (Macaca mulatta) were determined using electron microscopic and immunocytochemical methods. The two neocortical commissures, the corpus callosum, and the anterior commissure contain small but cytologically distinct archicortical components: the hippocampal commissure, which lies ventral to the splenium of the corpus callosum, and the basal telencephalic commissure, which forms a small crescent at the anterior margin of the anterior commissure. Each archicortical pathway is delineated from the adjacent neocortical commissure by a glial capsule. The glia cells that form this border are immunoreactive with antisera directed against glial fibrillary acidic protein (GFAP) and issue long processes that form numerous desmosomal junctions with one another. Braids of these glial processes envelop axonal fascicles within the archicortical commissures. In contrast, the GFAP-positive cells of the corpus callosum and anterior commissure are randomly distributed cells with relatively short stellate processes that do not form boundaries around axon fascicles. Quantitative electron microscopic analysis reveals that approximately 60 million axons connect the two cerebral hemispheres: the corpus callosum contains 56.0 million +/- 3.8 million axons (n = 8), the anterior commissure contains 3.15 million +/- 0.24 million axons (n = 8), the hippocampal commissure has 237,000 axons +/- 31,000 (n = 6), and the basal telencephalic commissure has 193,000 axons +/- 28,000 (n = 5). The number of axons is not directly proportional to the cross-sectional area in any of the commissures because of variation in axonal composition. On the basis of an estimate of approximately 3 billion neurons in the monkey cortex (Shariff, '53), we estimate that between 2 and 3% of all cortical neurons project to the opposite cerebral hemisphere. Subregions of the corpus callosum as well as each of the other commissures consist of characteristic subsets of five classes of axons and contain different proportions of myelinated to unmyelinated fibers. The largest myelinated axons and the smallest proportion of unmyelinated axons (approximately 6%) are found in regions of the corpus callosum that carry projections from primary sensory cortices, whereas the smallest myelinated axons and largest proportion of unmyelinated axons (approximately 30%) are found in regions of the corpus callosum that carry projections from association cortices. Axon composition in the anterior commissure is uniform and resembles that of callosal sectors that contain association projections.(ABSTRACT TRUNCATED AT 400 WORDS)

We have studied the cytological and quantitative aspects of axon addition and elimination in the corpus callosum of the developing rhesus monkey. Electron microscopic analysis reveals that during fetal development the number of callosal axons increases from 4 million at embryonic day 65 (E65) to 188 million at birth (E 165). Thus, the number of callosal axons in newborn monkeys exceeds the number present in the adult (an average of 56 million; LaMantia and Rakic, 1990a) by at least 3.5 times. Although there is some variability among the 11 fetal and newborn monkeys examined, there appears to be a progressive increase in the total number of callosal axons from midgestation through birth. The presence and numbers of growth cones from E65 through birth suggests that axon addition occurs exclusively during this period. There is no ultrastructural or quantitative indication of postnatal axon addition. After birth, about 70% of the axons in the callosum are eliminated in 2 phases. During the first phase, which includes the first 3 postnatal weeks, approximately 80 million axons are lost at an estimated rate of 4.4 million/d or 50/sec. During the second phase, which continues for the following 3 months, an additional 50 million axons are eliminated at a rate of 0.5 million/d or 5/sec until the adult value is reached. A discontinuous distribution of different classes of axons along the anterior-posterior axis of the tract reminiscent of the pattern seen in the adult is detectable before the onset of the first phase of axon elimination. Since the basic topography and terminal field patterns of callosal projections are well established before birth in all regions of the monkey cortex examined so far (Goldman-Rakic et al., 1983; Killackey and Chalupa, 1986; Dehay et al., 1988; Schwartz and Goldman-Rakic, 1990), we conclude that the massive postnatal elimination of callosal axons described here is unlikely to play a significant role in the development of discretely patterned callosal projection zones or their columnar terminations. The coincidence of axon elimination and the increase in synaptic density throughout the primate cerebral cortex during the first 6 postnatal months (Rakic et al., 1986), however, suggests that supernumerary axons may be lost during a process that results in the local proliferation of synapses from a subset of initial interhemispheric projections.

We have undertaken a quantitative analysis of the mouse olfactory bulb to address several major questions concerning the development of neural circuitry in the postnatal mammalian brain. These are: (1) To what degree are new elements and circuits added during maturation? (2) How long do such processes go on? and (3) Does postnatal development involve a net addition of circuits and their constituent elements, or is there elimination of some portion of an initial surfeit? Using male mice of known age, weight, and length, we measured the overall size of the bulb, the numbers of processing units (glomeruli) within the bulb, the extent and complexity of postsynaptic dendrites within the glomeruli, and the number of synapses in different regions of the bulb. Between birth and the time mice reach sexual maturity at 6-7 weeks of age, the bulb increases in size by a factor of 8, the number of glomeruli by a factor of 4-5, the length of mitral cell dendritic branches by a factor of 11, and the number of glomerular and extraglomerular synapses by factors of 90 and 170, respectively. Each of these parameters increases steadily from birth, in concert with the enlargement of the olfactory mucosa, the overall growth of the brain, and indeed, of the entire animal. We found no evidence of an initial surfeit of processing units, dendritic branches, or synapses. Further elaboration of neural circuitry by each of these measures is also apparent from the time of sexual maturity until the animals reach their full adult size at about 10-12 weeks of age. The developmental strategy in this part of the mouse brain evidently involves prolonged construction that persists until the growth of the body is complete. This ongoing elaboration of neural circuitry in the postnatal mammalian brain may be relevant to understanding a number of unexplained developmental phenomena, including critical periods, the ability of the juvenile brain to recover from injuries that would cause severe and permanent deficits in older animals, and the special ability of the maturing brain to encode large amounts of new information.

We have monitored the pattern of identified glomeruli in the olfactory bulbs of newborn, juvenile, and adult mice over intervals of several hours to several weeks. Our purpose was to assess the development and stability of these complex units in the mammalian brain. Glomeruli can be observed by vital fluorescent staining and laser-scanning confocal microscopy without causing acute or long-term damage to brain tissue. Repeated observation of bulbs in the same animals between birth and 3 weeks of age showed that this region of the brain develops by progressive addition of these units to the original population. This increment occurs by the genesis of smaller new glomeruli between larger existing ones; no elimination of glomeruli was observed during this process. Finally, no addition (or loss) of glomeruli occurred in adult animals over a 2 week interval; once established, the number, size, and pattern of glomeruli are evidently stable.