Hoxa1 is required for the retinoic acid-induced differentiation of embryonic stem cells into neurons.

The ability of embryonic stem (ES) cells to differentiate into different cell fates has been extensively evaluated, and several protocols exist for the generation of various types of cells from mouse and human ES cells. We used a differentiation protocol that involves embryoid body formation and all-trans-retinoic acid (RA, 5 microM) treatment (EB/5 microM RA) to test the ability of Hoxa1 null ES cells to adopt a neuronal fate. Hoxa1(-/-) ES cells, when treated in this EB/5 microM RA protocol, failed to differentiate along a neural lineage; Hoxa1(-/-) ES cells express severalfold lower levels of many neuronal differentiation markers, including nestin, beta-tubulin III, and MAP2, and conversely, higher levels of endodermal differentiation markers (i.e., Sox17, Col4a1) than wild type (Wt) cells. Reintroduction of exogenous Hoxa1, under the control of the metallothionein I promoter, into Hoxa1(-/-) ES cells restored their capacity to generate neurons. Moreover, overexpression of Sox17, a gene that regulates endodermal differentiation, in Wt ES cells resulted in endodermal differentiation and in a complete abolition of beta-tubulin III expression. Thus, Hoxa1 activity is essential for the neuronal differentiation of ES cells in the presence of all-trans-RA, and Hoxa1 may promote neural differentiation by inhibiting Sox17 expression. Pharmacological manipulation of Hoxa1 levels may provide a method for promoting neuronal differentiation for therapeutic uses. Furthermore, because mutations in the Hoxa1 gene can cause autism spectrum disorder in humans, these data also provide important mechanistic insights into the early developmental processes that may result in this disorder.