Fig 2: DrICE is required for trafficking of tracheal size determinants Serpentine and Uif. a–h Mutation of the basolateral polarity protein Yurt (yrt65A) causes abnormal intracellular accumulation of the tracheal size determinants Uif (b, arrowhead) and Serp (f, arrowhead), an effect that is suppressed by DrICE17 (c, g). White dashed lines outline basal cell surface; blue dashed lines outline apical surface. Scale bar a–h in h, 5 μm. i Relative number of cytoplasmic punctae quantified, where the number of punctae in each of two ROIs in a maximum projection of yrt65A (N = 5 embryos), yrt65A/+ (N = 5 embryos), and yrt65A, DrICE17 (N = 6 embryos) was counted, then divided by the area of the given ROI. Resulting number of puncta per area is reported relative to yrt65A/+. j The DrICE17 mutation suppresses tracheal overelongation caused by a mutation in the basolateral polarity gene yrt. Quantification of tracheal DT length at stage 16 for yurt65A (N = 6) and yurt65A, DrICE17 (N = 7) and DrICE17 (N = 5) mutants reveals that yurt65A, DrICE17 double mutants are shorter than yurt65A mutants. Tracheal lengths normalized to body length and reported relative to WT (w1118). For all graphs, each data point is represented as a point. k–o Mutation of the Hippo Network component Yki (ykiB5) reduces the levels of Serp (green) in the tracheal anterior relative to the lumenal chitin binding protein Gasp, marked by staining with the monoclonal antibody 2A12 (arrowhead, m). This reduction is suppressed by DrICE17 (n). k Quantification of the mean intensity of Serp divided by the mean intensity of 2A12 in a maximum projection of N = 4 embryos for WT (w1118), N = 4 embryos for ykiB5; DrICE17, N = 7 embryos for ykiB5, N = 3 embryos for DrICE17; Numbers reported as relative to w1118. Scale bar l–o in o, 12.5 μm. For all graphs, error bars represent S.E.M. and each data point is represented as a point. Source data are provided as a source data file. p < 0.05; **p < 0.005; ***p < 0.0005 Student’s t-test

Fig 3: DrICE governs tracheal size downstream of Yorkie without triggering apoptosis. a–l Compared to WT (w1118) (a), the dorsal trunks (DTs) of ykiB5 mutant embryos (b), and those overexpressing DrICE in the tracheal system (btl > DrICE) (c) are elongated, while DTs in DrICE17 are too short (d). DrICE17 is epistatic to ykiB5 since ykiB5; DrICE17 double mutants do not have long trachea (e). Loss of the DrICE inhibitor Diap1/th can cause tracheal overelongation (f) or missing DT segments when homozygous (i) or heterozygous (l). Expression of the pro-apoptotic genes btl>Grim (g) or btl>Reaper (h) causes segment loss dependent on DrICE17 dosage (j, k). Red arrows in g mark remnants of the tracheal system, while yellow arrowheads in h–l mark missing dorsal trunk segments. Scale bar for a–l in a, 25 μm. m Loss of Diap1 (thJ5C8) decreases the number of embryos with intact dorsal trunks (N = 17 embryos over five experiments) and the number of tracheal cells (N = 6 embryos total: 3 broken, 3 intact), indicating apoptosis. The loss of intact trachea and overall cell numbers can be suppressed by DrICE17 (N = 20 over three experiments and N = 4, respectively). Neither cell death nor dorsal trunk breaks are present in ykiB5 embryos (N = 4). n DT length in single and double mutant combinations of HN and DrICE mutants. DT length normalized to WT (w1118) (N = 10) for HN and DrICE mutants, and to btl-Gal4, UAS-GFP/+ for overexpressed DrICE (btl-Gal4 UAS-GFP/+; UAS-DrICE/+) (N ≥ 5 for all except N = 4 for thJ5C8;DrICE17). o Western blot of stage 16 embryos using the α-DrICECST13085 antibody, which recognizes full-length DrICE. A representative blot is shown in Supplementary Fig. 1i. The 47 kDa DrICE full-length band was quantified relative to total protein from at least three experiments. DrICE protein levels in DrICE17 homozygotes are not different than WT, consistent with DrICE17 being a dominant negative allele that causes more severe tracheal phenotypes than DrICEΔ118 (n, o). (N = 4 blots). All source data are provided in the source data file. For all graphs, each data point is indicated as a point. Error bars, S.E.M. *p < 0.05; **p < 0.005; ***p < 0.0005 Student’s t-test

Fig 4: A pool of DrICE partially co-localizes with Clathrin. a–j Staining with the α-DrICESK31 antibody against full-length DrICE (a–i) reveals broad cytoplasmic signal in WT trachea (a) that is reduced in trachea zygotically homozygous for the null allele DrICEΔ1 (g) and elevated in trachea overexpressing DrICE (i). Staining with α-DrICECST9478 (b–j), which was raised against a peptide containing the DrICE sequence that is cleaved during activation, reveals a more restricted punctate pattern that is enriched at the tracheal apical surface in WT trachea (dashed blue lines), and particularly at earlier developmental stages (Supplemental Fig. 1b, e). α-DrICECST9478 signal is reduced in DrICEΔ1 trachea (h), elevated in trachea overexpressing DrICE (j). DrICECST9478 signal in h results from the maternal contribution of DrICE that is also observed by western blot in Fig. 1o (representative gel image in Supplementary Fig. 1l) and is absent in late larval tissue (Supplementary Fig. 2d–g). k Partial schematic of the endocytic system showing the roles of endocytic markers. l–o Localization of DrICECST9478 signal with markers of the endocytic system in WT (w1118) trachea. α-DrICECST9478 signal shows partial colocalization with Clathrin (l) but not and Rab5 (m), with Rab11 (n) or Rab7 (o). Co-localization was quantified in two ROIs in each of N ≥ 3 z-slices for N ≥ 4 embryos. Source data are provided as a source data file. Blue dashed lines mark the apical cell surface, white dashed lines mark the basal cell surface. Scale bar for a–j in j, 5 μm; for l–o in o, 5 μm