The Notochord and Hypotheses about the Evolution of the Vertebral Column

In this last chapter in our journey through notochord and vertebral development and evolution, we addressed the perennial question of the role and fate of the notochord across the vertebrates. Each group of specialists has approached the notochord from the perspective of their own discipline. We tease apart those perspectives to conclude that the function of the notochord for vertebral column development is threefold and essential: (i) as a signaling center, and to provide (ii) structural and (iii) functional components at all stages of vertebral column development. Since 1895, one hypothesis — the arcualia hypothesis — has dominated the interpretation of the relationship between the notochord and vertebral centra development. According to this hypothesis, vertebral centra evolved from sclerotome-derived paired cartilaginous elements (arcualia) that abut dorsally and ventrally against the notochord sheath to form the bases of the neural and hemal arches and eventually the vertebral centra. In contrast, based on developmental studies, the now widely accepted autocentra hypothesis claims that a perichordal ring of fibrous tissue forms a second layer around the notochord and that skeletogenic material of the vertebral bodies forms within this perichordal tube. In about half of all extant vertebrate species (teleosts), vertebral centra (chordacentra) development is initiated in the notochord sheath and then proceeds to the second layer, the autocentrum. The formation of chordacentra and autocentra is regulated by the notochord, two processes clearly in conflict with the arcualia hypothesis. Chordacentra have not been identified in tetrapods. Rather, sclerotome-derived cells form an unsegmented layer around the notochord that subsequently separates into vertebral centra (autocentra). There is molecular evidence that defining the intervertebral spaces is the early step of vertebral body formation. Extension of the notochord sheath and proliferation of the notochord epithelium are visible characters in this process. After considering the available evidence, we conclude that, although chordacentra arise from the notochord, and although autocentra are sclerotomal, they are functionally equivalent and represent one developmental module. The notochord controls mineralization and patterning of both structures. We conclude that chordacentra and autocentra, and the process of intervertebral joint initiation, are homologous across the vertebrates. Therefore, vertebral columns are homologous across the vertebrates.