A two-cilia model for vertebrate left-right axis specification.

The mechanisms underlying vertebrate left-right (L-R) axis specification have attracted much interest among developmental biologists (for review, see Burdine and Schier 2000; Capdevila et al. 2000). Of particular interest from a conceptual standpoint is the question of how L-R positional information first originates within the context of a bilaterally symmetric embryo. To achieve the consistent L-R handedness characteristic of the vertebrate body plan, the L-R axis must be reliably oriented with respect to the other two embryonic axes, anteroposterior (A-P) and dorsoventral (D-V; Brown and Wolpert 1990). Understanding how embryos achieve this goal without reference to any external cues represents a formidable challenge for embryologists and theoretical biologists alike. The first breakthrough in understanding the molecular basis of handed asymmetry was the discovery of a cascade of signals produced asymmetrically in the early embryo that was sufficient to direct the orientation of later L-R morphogenetic events (Levin et al. 1995). This, together with subsequent reports identifying additional genes expressed asymmetrically in the early embryo, opened the door to ongoing work directed at understanding how L-R positional information is propagated in the embryo as well as how this information is ultimately translated into morphological manifestations of L-R asymmetry (for review, see Burdine and Schier 2000; Capdevila et al. 2000). Perhaps most important, however, were the general patterns of asymmetric gene expression first reported by Levin et al. (1995)—initial tightly localized domains of asymmetric gene expression at the chick node (ActRIIa, Shh, and then Nodal medially) followed by broad domains of asymmetric gene expression throughout the lateral plate (Nodal in the left lateral plate mesoderm). Significantly, although some of the specific molecular players appear to vary across different species, the node has consistently been the location where the earliest molecular asymmetries are centered, drawing attention to this region as the most likely site for the initial symmetry breaking event responsible for specifying the orientation of the L-R axis (Vogan and Tabin 1999; Capdevila et al. 2000). The precise manner in which asymmetric gene expression might first be established, however, remained a mystery until a second breakthrough discovery documenting the existence of motile monocilia and directional fluid flow at the mouse node (Nonaka et al. 1998).