Tissue polarity in an insect segment: denticle patterns resemble spontaneously forming fibroblast patterns

The insect integument displays planar tissue polarity in the uniform posterior orientation of denticles and bristles. How do cell polarities become uniformly oriented in the plane of the epidermal sheet? We have already shown that it is possible to disturb uniform denticle orientation in abdominal segments of Dysdercus (Nubler- Jung, 1987). Here I report that abnormally oriented denticles tend to form small arrays with uniform orientation. Adjacent arrays with divergent orientations realize a small repertory of characteristic pattern elements. We obtain these pattern elements by orthogonal transformation of pattern elements that form spontaneously in confluent fibroblast cultures, which rely on autonomous cell behaviour, and which later simplify into patterns predicted by specific boundary conditions (Elsdale & Wasoff, 1976); the only additional parameter required is planar cell polarity. The abnormal patterns in Dysdercus may thus also form spontaneously and may also rely on autonomous cell behaviour. The normal pattern is predicted by the parallel segment boundaries. I propose that the characteristic pattern elements in the larval epidermis may arise because elongated epidermal cells tend to arrange in parallel arrays and to orient in the same direction. The normal posterior orientation of cell polarities may result from orienting cues provided by the anterior and by the posterior intersegmental regions.