Item

Abstract

Segmentation or metamerization is a basic feature of almost all higher organisms. In recent years genes have been identified in Drosophila (Nüsslein-Volhard and Wieschaus, 1980) which, when mutated, alter segmentation in a predictable way. These observations have enabled us to construct molecularly feasible models.
Mutations affecting segment polarity can be explained under the assumption that segmentation arises from a repetition of three subunits (… P/SAP/SAP/S …). A segment border is formed whenever P and S cells are juxtaposed while the A-P confrontation is a precondition for the formation of limbs and wings. This threefold subdivision guarantees polar development within each segment. A loss of a gene for one of the three regions leads to a symmetrical pattern, e.g. after A-removal, to … S/P/S/P … with twice as many normal segment borders (corresponding to the mutation patch).
The pair rule mutations indicate an intermediate formation of double segments. An explanation of these mutations can be given under the assumption that in the wild-type double segments are determined by an iteration of four cell states - the two extreme positional values in two successive segments. If one of these cell states is lost due to a mutation, the three remaining states allow the development of half the number of normal segments. The model predicts specific denticle patterns which are in agreement with the experimental observation.
Gap mutations indicate a primary subdivision into a few cardinal regions (three internal and two marginal regions). I propose that the border between two cardinal regions organizes two double segments. Thus, segmentation seems to proceed under the control of a hierarchical series of pattern forming events: a primary positional information controls the formation of few cardinal regions, each of which forms in turn four double segments. Each segment forms three compartments (of which only two seem to contribute to the adult organism).