Morphogen propagation and action: towards molecular models.

Theoretical views on morphogen gradients have altered dramatically with the massive arrival of molecular data regarding the establishment of graded concentrations in the embryo, and the finely tuned reading by cells of these concentration levels. I review these new perspectives, and analyze in detail two models, one pertaining to the propagation of activin in Xenopus embryos, the other to the interpretation of retinoic acid levels into transcription patterns by nuclear receptors. The unifying threads that seem to emerge are the combinatorial uses of receptor subtypes, cooperativity and autocatalysis (positive feedback) to achieve specificity and reliability.