Neuronal Aggregations

The CNS of vertebrates shows three basic patterns of neuronal aggregations. The first is the scattered distribution of cell somas with relatively diffuse boundaries, as in the telencephalon of most fish and some amphibians. Although random in appearance, these cell aggregates are quite organized and functionally distinct. Brain areas can be identified by their connections to other areas and by their neurotransmitters.

The second pattern is that of nuclei. A nucleus is a cell cluster with clear anatomical boundaries. The structure of a nucleus sometimes includes the alignment of cell somas in layers resembling an onion. For example, the dorsolateral geniculate nucleus of the thalamus in monkeys receives visual input from the retina and has six layers of cells. Nuclei are formed by nerve cells that differentiate in the area surrounding the ventricles and migrate toward their final location according to a so-called outside-in pattern: The outside layer forms first and the innermost layer last.

The third pattern involves cortices. A cortex is an external (i.e., peripherally located) laminar sheet, usually including several distinct layers of cell somas. The dorsal optic tectum of some salamanders and the mammalian neocortex are organized in this manner. Such cortical structures are usually formed according to a so-called inside-out pattern. The innermost cell layer is established first, and more peripheral layers are added by subsequent cell migrations.

Neural development is characterized by the early proliferation of a large number of neurons and neuronal connections. Axons not only grow toward a specific target but also reach transient destinations, often by the development of collateral axonal branches. As the organism matures, there is a selective death of neurons (called apoptosis) and many of the connections previously established are eliminated. A major cause of apoptosis is the failure to establish contact with the appropriate target region. For example, if a normal target (such as an eye or a limb) is surgically removed, then many of the neurons that would normally innervate that target also die. The prevailing view is that more neurons and connections than needed are produced, but only functionally successful ones are preserved. Neural targets produce a variety of proteins that prevent neuronal death. One such protein is known as nerve growth factor (Levi-Montalcini, 1987). These proteins, known as neurotrophins, are absorbed by axon terminals and transported toward the neuron’s soma where they promote cell survival. Growing axons compete for access to a limited amount of neurotrophins in a sort of “neural Darwinism.”

Cell differentiation, migration, and death are all potential determinants of the amount of tissue allocated to a particular area. Homologous nuclei and cortical structures vary considerably even among closely related species. For example, the neocortex of primates is quite extensive compared to that of insectivores or bats. These differences must be achieved by species differences in developmental processes such as those described in this chapter. In vertebrates, brain size scales orderly to body size. Similarly, neural aggregates also scale orderly to brain size (Finlay & Darlington, 1995). In mammals, the paleocortex, septum, hippocampus, and cerebellum, among others, increase their size in correlation with the size of the entire brain (see Chapter 5 for vertebrate brain areas). The increase is steeper for the cerebellum, diencephalon, and striatum (slope value ranging between 0.34 and 0.36) than it is for limbic structures such as the septum and hippocampus (slopes between 0.28 and 0.30), and steepest for the neocortex (slope of 0.45).

These structures develop in a fixed order across mammals, but the moment in which neurogenic cells differentiate into neurons, and with some exceptions no longer reproduce, varies across species. This timing process determines the relative size of the aggregate in the adult. In species in which neurogenic precursors continue to divide for a longer period before yielding true neurons, the terminal size of the structure is relatively larger. It appears that the high-rate growth of the neocortex in primates is achieved by delaying the point in time at which precursor cells differentiate into neurons. A comparison of the onset and duration of terminal neurogenesis for a wide range of brain structures, on a common scale, for rats and rhesus monkeys shows substantial overlap in developmental timing, except for some layers of the neocortex (layers II to V). In primates, these layers are established relatively late, which allows for the production of a large population of precursor cells that differentiate into neurons (Finlay & Darlington, 1995). Such timing effects can lead to differences in the relative size of homologous structures across species.

Notice an apparent paradox in brain evolution. On the one hand, sometimes a particular structure is relatively enlarged in species that rely heavily on a behavior controlled by that structure. For example, the hippocampus (a structure that contributes to spatial learning) is relatively enlarged in food-storing birds (which rely on spatial learning to retrieve previously stored food), relative to that of nonstoring species. On the other hand, the size of a particular structure, including the hippocampus, depends on the size of the brain. Any increase in the size of a particular structure is, therefore, restricted by the size of the brain. These restrictions are referred to as developmental constraints. Does this mean that for a food-storing species to increase the size of its hippocampus, the entire brain must also increase? An affirmative answer to such a question would imply an inefficient trade-off since structures that may not be needed would have to increase in size. However, this is not necessarily the case. There is enough variation in the size of specific structures to allow for what would be significant increases in its size in any given species, without violating the constraint that relates the structure’s size to the size of the brain. A homologous structure may differ in size among two closely related species by as much as a 2.5 factor (a significant difference) and still be within the constraint imposed by the species’brain size (Finlay & Darlington, 1995). Surely, an increase in the size of one structure would lead to readjustments, even if minor, in the size of other parts of the brain.