The Emergence of Modern Neuroanatomy and Developmental Neurobiology

It had been generally assumed, mainly from studies on the innervation of limbs, that the patterns of connections formed during development are, from the beginning, rather precise. However, during the 1970s it became clear that this is often not the case and that there is normally a period during which the initial pattern of connections is progressively refined. Perhaps the most striking evidence for this phenomenon was the demonstration that in monkeys there is, at first, considerable overlap in the projection of the inputs from the two eyes (by way of the lateral geniculate nucleus) to layer IV of the visual cortex. Later, the inputs from the two eyes become progressively separated until finally the characteristic “eye dominance columns” are established (Rakic 1977xPrenatal development of the visual system in rhesus monkey. Rakic, P. Phil. Trans. R. Soc. Lond. [Biol.]. 1977; 278: 245–260CrossRef | PubMedSee all ReferencesRakic 1977). Several other examples of this type of connectional refinement were reported over the next decade, leading to the general conclusion that “pruning” and “sharpening” are general (if not universal) features in the development of connections in the vertebrate nervous system. The fact that the refining process could be prevented by functionally blocking activity in the projecting neurons (for example, by prolonged treatment with tetrodotoxin throughout the refinement period) suggested that although activity plays no discernible role in the initial patterning of connections, their later refinement is an activity-dependent process (Fawcett et al. 1984xActivity and the control of ganglion cell death in the rat retina. Fawcett, J.W., O'Leary, D.D., and Cowan, W.M. Proc. Natl. Acad. Sci. USA. 1984; 81: 5589–5593CrossRef | PubMedSee all ReferencesFawcett et al. 1984). Exactly how activity shapes neuronal connectivity was not evident at the time, although selective visual deprivation experiments indicated that in some way patterned stimulation and competition between the inputs from the two eyes was involved (LeVay et al. 1980xThe development of ocular dominance columns in normal and visually deprived monkeys. LeVay, S., Wiesel, T.N., and Hubel, DH. J. Comp. Neurol. 1980; 191: 1–51CrossRef | PubMedSee all ReferencesLeVay et al. 1980). It would not be until the 1990s that this issue began to be understood in molecular terms (Shatz 1997xSee all ReferencesShatz 1997).In 1970, RedfernxNeuromuscular transmission in new-born rats. Redfern, P.A. J. Physiol. 1970; 147: 291–320See all References1970, Redfern had shown that initially muscle fibers are innervated by a number of different axons and that over a period of several days all but one of the innervating axons is eliminated. That polyneuronal innervation of this kind was a fairly widespread phenomenon was later demonstrated for autonomic ganglion cells by Lichtman 1977xThe reorganization of synaptic connections in the rat submandibular ganglion during postnatal development. Lichtman, J.W. J. Physiol. 1977; 273: 155–177PubMedSee all ReferencesLichtman 1977; see alsoLichtman and Purves 1980xThe elimination of redundant preganglionic innervation to hamster sympathetic ganglion cells in early post-natal life. Lichtman, J.W. and Purves, D. J. Physiol. 1980; 301: 213–228PubMedSee all ReferencesLichtman and Purves 1980 and for the climbing fibers of the cerebellum by Crepel et al. 1976xEvidence for a multiple innervation of Purkinje cells by climbing fibers in the immature rat cerebellum. Crepel, F., Mariani, J., and Delhaye-Bouchaud, N. J. Neurobiol. 1976; 7: 567–578CrossRef | PubMedSee all ReferencesCrepel et al. 1976. It is still not clear why the terminals of one axon persist in these situations while the others degenerate or are withdrawn, but Lichtman's work made it clear that it is not because the postsynaptic cell is unable to sustain a large number of synapses: at the same time that multiple innervation is being eliminated in the submandibular ganglion, each surviving axon can significantly increase the number of synapses it forms on the related postganglionic cell.Except in rather special cases like the cerebellar climbing fibers, it was difficult to demonstrate the selective elimination of axon terminals in the central nervous system. However, it was clear (as Cajal had initially proposed) that in many systems appreciably more connections are formed than persist in the mature nervous system. One of the clearest demonstrations of this was provided in 1977 by Innocenti and his colleagues (Innocenti et al. 1977xExuberant projection into the corpus callosum from the visual cortex of newborn cats. Innocenti, G.M., Fiore, L., and Caminiti, R. Neurosci. Lett. 1977; 4: 237–242CrossRef | PubMed | Scopus (152)See all ReferencesInnocenti et al. 1977). In mature animals, callosal connections between the two cerebral hemispheres are sharply limited to certain cortical fields and not others (e.g., the visual cortex); however, at an earlier stage in their development, callosal connections are found throughout the entire hemisphere (including the visual cortex). At the time, it was not clear whether the refinement of the callosal projection was brought about by the selective death of the neurons in certain cortical areas, or by the elimination of callosal axon collaterals given off by neurons that project elsewhere within the ipsilateral hemisphere. This issue could only be resolved by the use of the double-labeling procedures described above. These clearly established that in those areas of the cortex that in adults lack callosal connections, there are, at earlier stages, neurons that send collaterals to the contralateral cortex. Most (and possibly all) of these cells survive beyond the period when the callosal projection is refined, but the collaterals they send into the corpus callosum are selectively eliminated (O'Leary et al. 1981xEvidence that the early postnatal restriction of the cells of origin of the callosal projection is due to the elimination of axonal collaterals rather than to the death of neurons. O'Leary, D.D.M., Stanfield, B.B., and Cowan, W.M. Brain Res. 1981; 227: 607–617CrossRef | PubMed | Scopus (37)See all ReferencesO'Leary et al. 1981). Later work by O'Leary and his colleagues was to show that the selective elimination of axon collaterals occurs in other cortical projection systems, and has led to the notion that from the point of view of their subcortical connections, all cortical areas are initially equivalent. Only later do the regional differences become apparent as different axon collateral projections are selectively eliminated (O'Leary and Stanfield 1989xSelective elimination of axons extended by developing cortical neurons is dependent on region of locale. Experiments utilizing fetal cortical transplants. O'Leary, D.D.M. and Stanfield, B.B. J. Neurosci. 1989; 9: 2230–2246PubMedSee all ReferencesO'Leary and Stanfield 1989).It is hoped that the foregoing account of some of the discoveries made during the 1970s will serve to illustrate how the advances made during that decade built on the work of the previous 70 or more years and, at the same time, provided the basis for much of the research of the ensuing 18 years. Certainly, from the point of view of developmental studies, the decade of the 1970s was a transitional period between the era of purely descriptive studies to the modern era, in which molecular and cellular approaches are beginning to elucidate the underlying mechanisms that shape and regulate the development of the nervous system. In retrospect, one might argue that this transition would not have occurred as rapidly or as effectively were it not for the new neuroanatomical methods that were developed during the same period.‡To whom correspondence should be addressed.