In the present project three groups at the Karolinska Institute join forces in order to: 1) Identify molecular players capable of inducing a midbrain dopaminergic (DA) phenotype. We previously found that soluble signal(s) derived from glial cells in the developing ventral midbrain (VM) induce a midbrain DA phenotype in Nurr1-expressing stem/progenitor cells. To identify the DA inductive signal we are undertaking several approaches: a) Purify the inductive activity from VM glia conditioned media; b) perform Affymetrix Genechip experiments to compare glial cells from inductive and non-inductive regions; c) determine the expression pattern of soluble developmental signals in the VM at the time of birth of DA neurons and d) test putative DA inductive molecules in our induction assay (Nurr1-expressing neural stem cells). 2) Induce a midbrain DA phenotype in human embryonic stem (ES) cells. This includes the characterization of new and previously isolated hES cells (KA 08 and 09) and their differentiation in neural stem cells and midbrain DA neurons. The strategy includes the induction or the overexpression of Nurr1 in ES cells and their exposure to the DA inductive signal. Preliminary experiments indicate that VM glia conditioned media can induce mES cells to adopt a DA phenotype. 3) Develop an ES cell replacement approach for DA neurons in MPTP and 6-hydroxydopamine rodent models of Parkinson's disease. Two alternative approaches will be examined: the induction of DA neurons from human ES cells in vitro, prior to grafting, and their induction in vivo, after grafting in the striatum or substantia nigra. The phenotypes adopted by the ES cells, host responses, graft integration, survival and function, including appropriate projections, synapses, dopamine content and release, and behavioral recovery of the animal, will be evaluated.