Pattern formation in biology: a comparison of models and experiments

How are the complex structures of a higher organism generated in such a reproducible way? Models of biological pattern formation are given in the form of nonlinear partial differential equations that describe production and decay rates as well as the diffusion of substances involved in pattern formation. As shown by comparison between expected and observed regulatory behaviour, these models describe many experimental observations in detail. According to these models, the following processes play a key role. (i) Primary pattern formation results from short ranging autocatalysis and long ranging inhibition. Monotonically graded distributions of substances can be generated that can be used by cells to develop appropriate to their position within the organism. Periodic or stripe-like distributions can be generated in the same way by different diffusion rates and life times of the substances involved. (ii) Cells obtain a stable state of differentiation by direct or indirect autoregulation of genes accompanied by a mutual competition among alternative genes. In this way, only one of several alternative genes can remain active within a particular cell. Which of the genes becomes activated can be under the control of a gradient generated by the mechanism mentioned above. (iii) By mutual long range stabilization of cell states, a controlled neighbourhood of structures can be achieved. Segmentation such as seen in insects is proposed to result by a cyclic mutual activation of such locally self-stabilizing cell states. (iv) Boundaries between regions generated by these mechanisms can obtain organizing properties for the finer subdivision of an organism. Substructures such as eyes, legs or wings are proposed to be initiated around the intersection of two borders. This mechanism accounts for the pair-wise initiation of these structures at the correct positions and with the correct handedness.