<i>bmp4</i> regulates ML patterning upstream of lateral <i>wnt5</i>.

(A-C) FISH detecting slit or wnt5 following 28 days of homeostatic control or bmp4 or nog1;nog2 RNAi. Right panels show enlargements of boxed regions. Scale bars represent 300 μm. (A) bmp4 RNAi caused reduction of slit in the anterior and elimination in the posterior (arrows). (B) Inhibition of bmp4 caused medial expansion of wnt5 (arrows). (C) RNAi of nog1;nog2 reduced the size of the wnt5 lateral expression domain. (B-C) Graphs show wnt5 expression domain width (dorsal side) normalized to body width. *p(D) Eyes assessed by live imaging (top) or opsin FISH (bottom) after 21 days of homeostatic RNAi to inhibit bmp and/or wnt5 and scored for lateral (green arrows) or medial ectopic eyes (red arrows). For single-gene RNAi, dsRNA for the targeted gene was mixed with an equal amount of control dsRNA so that the total amounts of each targeted dsRNA delivered were equal between single- and double-RNAi conditions. 100% of bmp(RNAi) animals had ectopic medial eyes and 92% wnt5(RNAi) animals had ectopic lateral eyes. By contrast, 100% of bmp4;wnt5(RNAi) animals had at least two lateral ectopic eyes, and of these 38% also formed a single medial ectopic eye (right panels) while 62% only formed ectopic lateral eyes (left panels). Co-inhibition of wnt5 reduced the penetrance of the medial ectopic eye phenotype due to bmp4 inhibition (p = 0.0027, 2-tailed Fisher’s exact test).