target cells. During the normal development of an embryo, axons need to be guided to their appropriate targets through the combination of actions of attractive and repulsive axon guidance molecules, such as netrins, semaphorins, slits, and ephrins. Many of these guidance molecules arise from glia (astrocytes and oligodendrocytes), which act as guideposts, and intermediate target cells that steer a growing axon to its appropriate target (Chotard and Salecker, 2004). Each of these molecules also has at least one complementary receptor on the axon. When the guidance molecule and receptor interact, the receptor transmits a signal to the growing axon to either keep growing or avoid the area. These groups of molecules act in complex ways to guide developing axons. Axon guidance relies on the interplay of many different guidance molecules and receptors. Furthermore, the concentration gradients of the molecules also significantly influence the effects of the molecules on steering the axon in a specific direction.
The complexity of this mechanism is also underscored by the example of diffusible netrin molecules that, depending on the receptor on the axon with which they interact, can act as either an attractant molecule (Keino-Masu et al., 1996) or a repulsive molecule (Leonardo et al., 1997). Much information has been garnered about how these molecules affect axonal targeting in the developing nervous system; however, studies are under way to determine whether injured axons in the adult CNS are able to reexpress their receptors for these guidance molecules and whether the axonal targets can once again express their guidance cues (Koeberle and Bahr, 2004). Studies to date demonstrate that the expression patterns of many guidance molecules and receptors are the same during nervous system development and after an injury; but some are very different, and these differences could have important consequences on the correct targeting of a growing axon. For instance, the level of expression of a specific class of ephrins (ephrin-Bs) appears to be decreased in the brain, which could limit reinnervation by regenerating axons (Hindges et al., 2002). To overcome this, methods are being developed to examine the effectiveness of using gene therapy strategies and scaffolds (discussed below) to express different combinations of guidance molecules. These guidance molecules could be used as physical conduits that promote regrowth (Dobkin and Havton, 2004).
Gene therapy is another treatment strategy that has great potential to provide the injured spinal cord with the specific gene products—proteins—that it needs to promote functional recovery. Gene therapy is not a current treatment for spinal cord injuries but is being studied with animal models of spinal cord injury. The concept is to transfer into the spinal cord a gene encoding a therapeutic protein, such as a growth factor or an axon guid-