Our laboratory has developed a protocol for tooth organ engineering, based on biomimetic and the use of dissociated embryonic dental cells. Searching for mesenchymal cell sources avoiding the use of embryonic cells still remains a major challenge. We tested the odontogenic potential of several cell lines, dental or not, embryonic or adult. These cells were re-associated with a competent intact dental epithelium. We searched for tooth formation in vitro and/or after implantation in adult mice and studied different biological and technical parameters. For this purpose we analyzed the effects of the age, of a pre-culture step, and of the cellular heterogeneity on the odontogenetic potential of mesenchymal cells. To test the heterogeneity, we compared the patterns of expression of cell surface markers in cultured and implanted re-association with those observed during tooth development..Our results show that the absence of odontogenetic potential in the different cell lines that were tested, in part depends on the age of cells and that it is lost when mesenchymal cells are cultured in monolayer before their re-association. This could be explained by phenotypical changes as shown by testing several genes involved in tooth development, and tracing cell surface markers expression. Changes were observed, wich could be related to a cell selection in vitro, leading to variations in cellular heterogeneity. Indeed, pulpal cellular heterogeneity shows specific patterns of expression, that are space-time defined during tooth development, and is controlled by epithelial-mesenchymal interactions.