Figure 3.3
Proneural clusters, whose existence was deduced by Stern in 1954 [4095] and confirmed histologically in 1989-91 [912, 3637, 3982].

a-f. Stern's reasoning, which was based on the nonautonomous formation of an ectopic MC by wild-type tissue in achaete¹ (ac¹) mosaics. a. Heminotum (minus humerus) of a wild-type fly, with all bristles omitted except the PDC (posterior dorsocentral) MC. b. Homozygous ac¹ flies typically lack a PDC. Absence of shading indicates yellow^null (y^null, yellow cuticle), which was a linked marker for ac¹ [636, 1442], as was singed¹ (sn¹, gnarled bristles). c, d. Gynandromorphs arising from X-chromosome loss in y^null ac¹ sn¹/+ + + heterozygotes (redrawn from [4095]). Both heminota lack a PDC, evidently because tissue there is mutant. In some cases (e.g., d), a PDC-like MC arises in nearby heterozygous tissue [3613, 4095, 4096]. e, f. Illustrations based on Stern's hypothesis for bristle displacement [4095]. Squares show imaginary details for PDC areas in a and d. Dashed line marks the normal PDC site. A 'proneural cluster' (dark shading) of cells (small hexagons) is 'competent' to make a PDC. The central cell becomes the PDC (e) unless its ac¹ genotype renders it unable (f), whereupon a cell in the remnant substitutes -- thus displacing the PDC. All SOPs (sensory organ precursors), Stern argued, must inhibit their neighbors from making bristles.

g-j. Drawing of a real cluster on a right heminotum at ~25, 20, 10, and 0 hrs. before pupariation, as revealed by antibodies to Scute [912, 3982]. Ovals are nuclei (diameters ≈ 4 µm). The actual dorsocentral cluster (1) has an irregular shape, (2) grows, and (3) yields two MCs whose SOPs arise at the cluster's edge. g. Some cells express more Scute (dark shading). h. The SOP for the PDC (black) arises in a high-Scute area and acquires a 'halo' of low-Scute cells. i. As the cluster grows, the high-Scute area shifts, and the ADC (no halo yet) arises. The reason for this shift is not known. j. In this cluster, non-SOP cells stop expressing Scute before SOPs, which do so later (before dividing). Further growth widens the gap between the ADC and PDC. As for why the ADC and PDC typically form on the medial (dorsal) side of their common cluster, the answer appears to be that a diffusible bristle-promoting signal (Dpp) encounters this cluster on its medial face and then wanes (cf. Fig. 6.14) [3373, 4368].