Fig. 3.

Both Ftz-F1 and Ets21c are required for tumorigenesis. (A) Whereas control larvae underwent pupariation on day 6-7 AEL, most of the animals bearing ras^V12scrib¹ EAD tumors died as giant larvae, only rarely forming pseudopuparia. Interfering with Fos, Ftz-F1 or Ets21c^LONG function in ras^V12scrib¹ clones markedly improved pupation rate, whereas jun depletion had no effect. The timing of the larval-pupal transition was partly rescued upon ftz-f1 RNAi and loss of fos (kay³). The graph shows the cumulative percentage of pupae forming over time. All genotypes differed significantly from control, and except jun^RNAi, also from ras^V12scrib¹ (P<0.0001). (B) Reducing ftz-f1, but not jun or ets21c^LONG, significantly suppressed tumor invasiveness (P<0.001). Four grades of invasiveness were scored based on spreading of clonal GFP-positive cells into larval brains dissected on day 7 AEL. Results are the percentage of brains falling into each category. (C-F) Loss of fos or knockdown of ets21c^LONG in ras^V12scrib¹ tumors did not affect size of the GFP-labeled clones, whereas ftz-f1^RNAi slightly reduced the tumor burden. The EAD morphology was visualized by immunostaining against Fasciclin III (C-F). Overgrowing ras^V12scrib¹ cells failed to differentiate into photoreceptors, as shown by loss of ELAV staining (C′). Few ELAV-positive cells were detected in ras^V12scrib¹ftz-f1^RNAi and ras^V12scrib¹ets21c^{LONG RNAi} discs, which formed greatly disorganized ommatidial clusters (E′,F′), whereas many more ras^V12scrib¹kay³ cells differentiated (D′). All images show EAD dissected 6 days AEL, either as projections of multiple confocal sections (C-F) or as single sections (C′-F′). Scale bars: 100 µm (C-F) and 20 µm (C′-F′). (G) Activity of the ex::lacZ reporter is markedly lowered upon inhibition of ftz-f1 but not fos in ras^V12scrib¹ clones of EAD. All samples were stained for the same period of time. (H) Thirteen per cent of the ras^V12scrib¹ftz-f1^RNAi tumor-bearing animals eclosed as adults with enlarged, rough eyes.