A decade of mammalian retinal stem cell research.

Ten years have now passed since the discovery of quiescent neural stem cells within the mammalian retina. Beside the fascinating aspect of stem cell biology in basic science, these cells have also offered hope for the treatment of incurable retinal diseases. The field has thus rapidly evolved, fluctuating between major advances and recurring doubts. In this review, we will retrace the efforts of scientists during this last decade to characterize these cells and to use them in regenerative medicine. We will also highlight advances made in animal models capable of stem cell-mediated retinal regeneration.

Noxp20 and Noxp70, two new markers of early neuronal differentiation, detected in teratocarcinoma-derived neuroectodermic precursor cells.

The murine 1C11 cell line, derived from F9 pluripotent teratocarcinoma cells, exhibits features of a bipotential neuronal precursor as it converts into serotonergic or catecholaminergic neurons under appropriate induction. In order to point out molecular markers expressed in this early neuroectodermic commitment, we used a cDNA subtractive hybridization method. The 105 different isolated cDNAs represented 75 known genes, expressed sequence tags (EST) or genomic fragments. A majority of known proteins encoded by these sequences are involved in cellular mobility or migration. We characterized two sequences showing identities with ESTs and we called them Noxp20 and Noxp70. The Noxp20 transcript encodes a putative protein with a predicted caspase recruitment domain and the Noxp70 transcript encodes a putative protein displaying a Zn-finger domain. Consistent with their roles in neuronal cell development, in situ hybridization showed that Noxp20 and Noxp70 are over-expressed in brain. At embryonic days 12 and 15, Noxp20 is strongly expressed in the ventricular and intermediate zones of the brain and of the spinal cord. At embryonic day 15, Noxp70 was found to be strongly expressed in the ventricular zone around the telencephalic ventricle, and to a lower extent in the thalamus and hypothalamus. At post-natal day 10, Noxp20 mRNA was detected in the dentate gyrus, the hippocampus, the cerebellum and the olfactory bulb.

Stepwise control of osteogenic differentiation by 5-HT(2B) receptor signaling: nitric oxide production and phospholipase A2 activation.

During development, antagonists of 5-HT(2) receptor subtypes cause morphological defects of mesodermal and neural crest derivatives including the craniofacial skeleton. We used an inducible mesoblastic cell line, C1, able to fully convert into osteocytes within 12 days, to assess the involvement of 5-HT(2) receptors during osteogenic differentiation. On day 5 of the osteogenic program, immediately before matrix mineralization, the cells selectively implement 5-HT(2B) receptors (5-HT(2B)R) which remain functional until terminal differentiation. In 5-HT-depleted medium, the receptor exhibits a constitutive activity leading to basal nitric oxide (NO) release and phospholipase A2 (PLA2)-dependent arachidonic acid (AA) production. Blockade of this intrinsic activity affects the efficiency of mineralization by decreasing calcium incorporation within the matrix by 40%. Optimal bone matrix mineralization involves both NO and PLA2 signaling pathways. Moreover, between day 5 and day 10, at the beginning of mineral deposition, the 5-HT(2B)R promotes prostaglandin E2 production through AA-dependent cyclooxygenase (COX) activation. From day 10 onwards, when C1 osteoblasts undergo conversion into osteocyte-like cells, COX activity is quenched. Altogether these observations indicate that the 5-HT(2B)R contributes in an autocrine manner to osteogenic differentiation and highlight a switch in the downstream targets of the receptor at the terminal stage of the program. Finally, in addition to its autocrine function, the 5-HT(2B)R responds to 5-HT by increasing NO production and AA release. These findings raise concern regarding the use of 5-HT(2B)R-related drugs that may interfere with bone metabolism in physiological or pathological situations.

New cellular models for tracking the odontoblast phenotype.

Odontoblasts and osteoblasts differ functionally and histologically. Because of their close relationship, mesenchymal cells derived from teeth and bone are difficult to distinguish ex vivo. Indeed, the main non-collagenous components of the odontoblastic extracellular matrix, dentin sialoprotein (DSP) or dentin matrix protein 1 (DMP1), have also been detected in osteoblasts. The need to develop cellular models of odontoblast differentiation and to identify markers specific for the odontoblast lineage, has led us to establish clonal cell lines from tooth germs of day 18 mouse embryos transgenic for an adenovirus-SV40 recombinant plasmid. In this study, we analyzed the phenotypes of three independent clones by RT-PCR and Western blot. These clones synthesised DSP, DMP1 and other extracellular matrix proteins typical of the odontoblast and are therefore likely to be derived from the pulp. Transcripts encoding a set of homeobox proteins involved in craniofacial development, such as Pax9, Msx1, Cbfa1, Dlx2 and 5 were also expressed albeit at a different level. These features of the pulpal clones are shared by the C1 mesodermal cells that are capable of differentiating along osteogenic, chondrogenic or adipogenic lineages In contrast, transcripts for two LIM-domain homeobox family genes (Lhx6 and Lhx7) were only detected in the dental clones. Since these genes are preferentially expressed in the mesenchyme of the developing tooth, this suggests that our transgenic-derived cell lines retain intrinsic properties of odontoblastic cells. They may help to characterise genes specifying the odontoblast phenotype and the signalling pathways underlying odontoblast differentiation.