The notochord in the mammalian adult: A paradox

Abstract

This issue of Arthritis & Rheumatism contains a report by Erwin et al describing the authors’ investigation of a population of early embryonic cells from the notochord—by analyzing not embryos, but intervertebral discs from adult dogs (1). How is this possible? The fine chord defining the body axis in the early embryo of endoskeletal animals (vertebrates) is the basis of an evolutionary explosion, giving rise to everything from ascidians and amphioxus (2) to fish and humans (3). Its power lies in the specific mix of mesenchymal stem cells that ultimately are the origin of the central skeleton, including the spine and intervertebral bodies. Should there be a source of chordal stem cells, those cells might approach the ideal of embryonic stem cells residual in an adult organism. But is there a notochord cell within an adult spine? And how might one discriminate such a cell within the spine tissue cells? Current opinion holds that a certain subpopulation of intervertebral disc cells are remainders of the embryonic notochord and are therefore termed notochord cells (4). In the embryo, those cells are well defined by location in the notochord, by morphology (1,5,6), and by gene expression pattern (3,4). In the adult, morphology and location are the only criteria used to discriminate notochord cells from chondrocytes of the nucleus pulposus in the intervertebral disc. Since the notochord cells are thought to possess a buoyant density different from that of chondrocytes, density gradient centrifugation is applied to separate the 2 populations (7). As described in the report by Erwin et al (1), the cytoplasm of notochord cells is “foamy” (leading to the term “physaliphorous cells”), and their nucleus is amorphous and extended in size. These features may contribute to reduced buoyant density. An extensive endoplasmic reticulum (ER) is the origin of the physaliphorous morphology of that subpopulation in the embryonic notochord (5), and pioneering studies of the ER have involved investigations of its function in the embryonic notochord (8). Most adult mammals do not possess disc cells with similar morphology, but instead possess only fibroblastic or chondrocytic cells. However, a physaliphorous cytoplasm is expressed by a population of disc cells in some domestic animals, including dogs, sheep, and goats (4,9). Interestingly, Jun seems to prevent notochord cell apoptosis, and induced Jun deletion (in the Cre/lox system) causes scoliosis in the axial skeleton (10). In mongrel dogs, the presence of notochord cells is associated with resistance to agerelated disc degeneration, and the cells are able to promote in vitro proteoglycan production by chondrocytes (11). At least part of this capability is attributable to the tissue hormone connective tissue growth factor, as identified also in the study by Erwin and colleagues (1). But are these cells residual notochord cells, or successors, or cells that opportunistically and simply express the endoplasmic morphology of heavily active cells involved in the maintenance of disc tissue? The implications for preventive measures or curative approaches against disc degeneration are obvious. The critical issue is the definition of the term “notochord cells.” The biochemical findings presented by Erwin et al describe those cells as supportive for chondrocytes. Beyond the key hormonal function, the elaborate ER reflects intense metabolic activities, but which ones? Embryonic notochord cells could be identified via the expression of embryogenetic gene families such as Fox and Sox (3), but the current literature includes no information on Fox or Sox expression in those adult notochordal cells. Quite obviously, gene expression profiling would be beneficial. There is a major difference in perspectives: on the one side, there could be a supportive cell type enabling adult intervertebral discs to sustain the Juergen A. Mollenhauer, PhD, DSc: Naturwissenschaftliches und Medizinisches Institut an der Universitaet Tuebingen, Reutlingen, Germany, and Rush University Medical Center, Chicago, Illinois. Address correspondence and reprint requests to Juergen A. Mollenhauer, PhD, DSc, NMI, Markwiesenstr. 55, D-72770, Reutlingen, Germany. E-mail: juergen.mollenhauer@nmi.de. Submitted for publication June 19, 2006; accepted in revised form August 24, 2006.