During amniote embryogenesis, the segmented pattern characteristic of the vertebral column appears early during development through the sequential formation of multipotent structures called somites. Somites differentiate subsequently into dermomyotome (giving rise later to skin and skeletal muscles) and sclerotome (giving rise to vertebral bone structures and cartilage). In addition, sclerotomes subdivide following their rostro-caudal intrasegmental boundary into an axon growth-permissive region (anterior half) and an axon growth-repulsive region (posterior half). This binary system instructs motor and sensory axon navigation, as well as neural crest cell migration, to ensure that the peripheral nervous system develops without obstruction by the future cartilage and bones of the vertebral column. Repellent cues are expressed in posterior half-sclerotomes in order to exclude navigating axons from “no-go” areas and restrict their growth to specific exit points of the future vertebral column. Interestingly, similar repellent cues (e.g. Eph/Ephrins) are expressed in the adult central nervous system (CNS) and have been shown to control connectivity and plasticity throughout life. Following brain or spinal cord injury, these repellent molecules are upregulated by reactive astrocytes accumulating at the lesion site, and may impede axon regeneration in this region. In this dissertation, I am presenting the results of a differential gene expression analysis of anterior and posterior half-sclerotomes, based on RNA-sequencing data and using the chick embryo as a model organism. This study led to the identification of molecules, previously uncharacterized in this system, that may play a role in adhesive and mechanical properties of somites and in axon guidance and fasciculation. I focused on the functional analysis of one molecule of the posterior half-sclerotome, the extracellular matrix protein Fibulin-2. To look at its role in the segmentation of spinal axons, I used ectopic misexpression in a subset of segments based on somite electroporation. The width of spinal nerve bundle growth was restricted by Fibulin-2 overexpression in posterior and anterior half-sclerotomes, suggesting a role in sharpening/controlling the path of spinal axon growth. In addition, I showed that this could occur via an interaction with the axon growth repellent Semaphorin 3A. Then I looked at the expression of Fibulin-2 in two models of CNS injury: mouse cerebral cortical stab injury and rat dorsal crush spinal cord injury. In both cases, I observed an increase in Fibulin-2 protein level compared to control. I also used primary cultures of rat cortical astrocytes to show that the expression of Fibulin-2 after inflammatory cytokine-induced activation is increased. Finally, I studied a candidate axon growth repellent previously identified in the laboratory. I explored the hypothesis that this repellent molecule is an O-glycosylated, spliced variant form of a known protein. To characterize this repellent molecule, I used RNA-sequencing data from chick embryonic somites and 2D gel electrophoresis of an astrocytic cell line protein extract. Together, these results suggested that the developing vertebral column and the adult CNS share molecular features to control axon growth and plasticity. This type of study could lead to the characterization of molecular systems that regulate axon growth, and to the identification of novel therapeutic targets in brain or spinal cord injury.