Embryonic stem cells, due to their self-renewal and pluripotency properties, can be used to repair damaged tissues and as an unlimited source of differentiated cells. Although stem cells represent an important opportunity for cell based therapy and small molecules screening (in the context of drug or target discovery) many drawbacks are still preventing their widespread use. One of the most significant limitations is related to the complexity, as well as the reliability, of current protocols driving stem cells into any homogeneously differentiated cellular population. In this respect there is a strong demand for molecular agents promoting differentiation and thereby enabling robust, efficient and safe production of differentiated cells. In order to identify novel molecules that enhance early stages of differentiation, we developed an image based high content screening (HCS) approach using human embryonic stem cells (hESC). In our approach, we took advantage of custom image mining software specifically adapted for the selection of stem cell differentiation agents and the rejection of false positive hits. As a proof of concept ~3500 small molecules originating from commercial libraries were screened and a number of molecules of interests were identified. These molecules show stem cell differentiation properties comparable to the phenotypic signature obtained with the reference compound retinoic acid.
Keywords: Human embryonic stem cells, image-based screening, phenotypic classification, high content screening, cell based assay, phenotype, cell differentiation, Image Analysis, Z Factor Determination, retinoic acid