Directed differentiation and purification of mesencephalic dopaminergic (mesDA) neurons from stem cells are crucial issues for realizing safe and efficient cell transplantation therapies for Parkinson's disease. Although recent studies have identified the factors that regulate mesDA neuron development, the mechanisms underlying mesDA neuron specification are not fully understood. Recently, it has been suggested that mesencephalic floor plate (FP) cells acquire neural progenitor characteristics to generate mesDA neurons. Here, we directly examined this in a fate mapping experiment using fluorescence-activated cell sorting (FACS) with an FP cell-specific surface marker, and demonstrate that mesencephalic FP cells have neurogenic activity and generate mesDA neurons in vitro. By contrast, sorted caudal FP cells have no neurogenic potential, as previously thought. Analysis of dreher mutant mice carrying a mutation in the Lmx1a locus and transgenic mice ectopically expressing Otx2 in caudal FP cells demonstrated that Otx2 determines anterior identity that confers neurogenic activity to FP cells and specifies a mesDA fate, at least in part through the induction of Lmx1a. We further show that FACS can isolate mesDA progenitors, a suitable transplantation material, from embryonic stem cell-derived neural cells. Our data provide insights into the mechanisms of specification and generation of mesDA neurons, and illustrate a useful cell replacement approach for Parkinson's disease.

Mesencephalic dopaminergic (mesDA) neurons control movement and behavior, and their loss causes severe neurological disorders, such as Parkinson's disease. Recent studies have revealed that mesDA neurons originate from mesencephalic floor plate (FP) cells, which had been thought of as non-neurogenic organizer cells regulating regional patterning and axonal projections. Otx2 and its FP-specific downstream factor Lmx1a have been shown to be sufficient to confer neurogenic activity on FP cells and determine a mesDA fate. However, the mechanism underlying how these factors control mesDA development and how FP cells and mesDA neurons are coordinately specified are still largely unknown. In the present study, we obtained evidence that Lmx1a and Lmx1b cooperate with Foxa2 to specify mesDA neuron identity by gain-of-function approaches using transgenic mice. Lmx1a/b appeared to select a mesDA fate by suppressing red nucleus fate in the context of Foxa2-positive progenitors, at least in part, through repressing the Sim1-Lhx1 and Ngn1 pathways that inhibit proper mesDA differentiation. We also found that, in the mesencephalon, FP cell fate is primarily determined by Foxa2 with a supportive action of Lmx1a/b through repressing Nkx6.1, which inhibits FP cell differentiation. Thus, FP and mesDA identities are determined by distinct specification pathways, both of which are controlled by the same combination of transcription factors, Lmx1a/b and Foxa2, and, as a consequence, mesDA neurons are generated from mesencephalic FP cells.

Purkinje cells are the sole output neurons of the cerebellar cortex and their dysfunction causes severe ataxia. We found that Purkinje cells could be robustly generated from mouse embryonic stem (ES) cells by recapitulating the self-inductive signaling microenvironments of the isthmic organizer. The cell-surface marker Neph3 enabled us to carry out timed prospective selection of Purkinje cell progenitors, which generated morphologically characteristic neurons with highly arborized dendrites that expressed mature Purkinje cell-specific markers such as the glutamate receptor subunit GluRδ2. Similar to mature Purkinje cells, these neurons also showed characteristic spontaneous and repeated action potentials and their postsynaptic excitatory potentials were generated exclusively through nonNMDA glutamate receptors. Fetal transplantation of precursors isolated by fluorescence-activated cell sorting showed orthotopic integration of the grafted neurons into the Purkinje cell layer with their axons extending to the deep cerebellar nuclei and dendrites receiving climbing and parallel fibers. This selective preparation of bona fide Purkinje cells should aid future investigation of this important neuron.

GABAergic Purkinje cells (PCs) provide the primary output from the cerebellar cortex, which controls movement and posture. Although the mechanisms of PC differentiation have been well studied, the precise origin and initial specification mechanism of PCs remain to be clarified. Here, we identified a cerebellar and spinal cord GABAergic progenitor-selective cell surface marker, Neph3, which is a direct downstream target gene of Ptf1a, an essential regulator of GABAergic neuron development. Using FACS, Neph3(+) GABAergic progenitors were sorted from the embryonic cerebellum, and the cell fate of this population was mapped by culturing in vitro. We found that most of the Neph3(+) populations sorted from the mouse E12.5 cerebellum were fated to differentiate into PCs while the remaining small fraction of Neph3(+) cells were progenitors for Pax2(+) interneurons, which are likely to be deep cerebellar nuclei GABAergic neurons. These results were confirmed by short-term in vivo lineage-tracing experiments using transgenic mice expressing Neph3 promoter-driven GFP. In addition, we identified E-cadherin as a marker selectively expressed by a dorsally localized subset of cerebellar Neph3(+) cells. Sorting experiments revealed that the Neph3(+) E-cadherin(high) population in the embryonic cerebellum defined PC progenitors while progenitors for Pax2(+) interneurons were enriched in the Neph3(+) E-cadherin(low) population. Taken together, our results identify two spatially demarcated subregions that generate distinct cerebellar GABAergic subtypes and reveal the origin of PCs in the ventricular zone of the cerebellar primordium.

No abstract