The relatively new discipline of evo-devo relies on the fact that, because early developmental events determine the ground plan for further development, small alterations in the genetic programs underlying early development can lead to drastic changes in phenotype. Developmental neuroscience and evolutionary biology are thus complementary approaches. By studying normal development and the plastic mechanisms used by the brain to respond to injury, we can determine the constraints under which brain evolution operates when faced with alterations in peripheral innervation.
18.104.22.168 Brain Evolution Results from Minor Variations on a Basic Vertebrate Plan
Just as constructing a building requires first laying the foundation, constructing any vertebrate requires a foundation or Bauplan, incorporating bilateral symmetry, paired appendages, a dorsal nerve cord encased in a spinal column, and so forth. This requirement ensures that, at early stages, vertebrate embryos of any type will have common features. The prediction, and the finding, is that, the more common the feature, the earlier it appeared in evolution. The morphological and resulting functional differences that distinguish one vertebrate from another arise later in development, and represent fairly small deviations from the pre-existing plan. For example, a two-chambered heart becomes a three- and then a four-chambered heart over evolutionary time, but the heart itself is universal. Similarly, the vertebrate brain consists of hindbrain, midbrain, and forebrain, but the relative size and organization of each brain division and the functional regions within them can be modified during development according to the evolutionary history of each species (Pallas, 2001b; Krubitzer and Kaas, 2005).
What sorts of beneficial evolutionary modifications have been made to developmental programs? It is difficult to imagine how changes could even be tolerated, much less produce adaptive circuits, in a highly interconnected network such as the nervous system. It would seem that a change in one brain region or a change in one morphogenetic signal would either be ignored at the next level or would create a mismatch in the network and severely disrupt function. As we know from studies of recovery from brain trauma, however, the brain has a remarkable capacity to rewire itself in an adaptive fashion.
22.214.171.124 Developmental Mechanisms can Accommodate New Neurons and Trigger Matching Changes in Connected Populations
Both neurons and synapses are massively overproduced during development, and the final circuitry is pruned by processes such as programmed cell death and collateral elimination (see Finlay and Pallas, 1989, for a review of earlier literature). This pruning is directed in large part by activity-dependent processes (Hebb, 1949; Schneider, 1973; Innocenti et al., 1977; Pilar et al., 1980; O'Leary and Stanfield, 1986), ensuring that the final wiring pattern is appropriate to the environment and experiences of each individual. Extra neurons and connections produced during development can be stabilized by providing additional target space (Hollyday and Hamburger, 1976). Thus the natural plasticity built into the nervous system ensures that developmental (or evolutionary) errors can be absorbed, and even exploited. Any such acquired connections gained within an animal's lifetime would of course not be heritable, but relative excesses of target neurons produced through gene duplication, modifications of the cell cycle, or cell death of input neurons would automatically integrate into a circuit. Throughout vertebrate evolution there has been a consistent trend toward increasing the number of neurons, producing local variation in size of nuclei or layers and an increasing modularization (Finlay and Slattery, 1983; Caviness et al., 1995; Kornack and Rakic, 1998; Kornack,
2000; Ohki et al., 2005). Because of inherent plasticity in the inputs, matching changes in the pathway are not required. Environmental change is the primary source of selection pressure over generations, and a nervous system that responds on a developmental timescale to the environment provides a huge selective advantage, if the cost in terms of unintended connections is low. More importantly to this discussion, exuberant connections and extra neurons at any level of a sensory pathway are possible sources of variation in neural circuits between individuals (Sperry, 1963), and individual variations in connectivity could be selected for if they provide a survival advantage. Thus a mutation-induced increase or decrease in a neuronal pool could influence both upstream and downstream neuronal survival and collateral elimination. In this way, a change in one component of the pathway need not require a simultaneous and matching change in all components. Hebbian mechanisms would instruct the pathway to process the new inputs sensibly. In the following, I will illustrate how this can occur in developmental time, and propose how it might have occurred over evolutionary time.
Was this article helpful?