Fig. 6.
Function of projection neurons in phototaxis. a Larvae of all genotypes steer away from light. +/shiTS: n = 10, p = 2.2 × 10−4; PVL09 > shiTS: n = 10, p = 0.0094; PVL09/+: n = 10, p = 9.0 × 10−5; Pdf-LaNs > shiTS: n = 10, p = 6.5 × 10−6; Pdf-LaNs /+: n = 10, p = 8.0 × 10−5; allLaNs > shiTS: n = 10, p = 1.8 × 10−4; allLaNs /+: n = 10, p = 1.7 × 10−5. Larvae with LaNs silenced steer less compared to controls. One-way ANOVA: p = 8.7 × 10−7. Dunnett’s test: PVL09 > shiTS vs +/shiTS: p = < 0.001; PVL09 > shiTS vs PVL09/+: p = 0.1825; Pdf-LaNs > shiTS vs +/shiTS: p = 0.9821; Pdf-LaNs > shiTS vs Pdf-LaNs/+: p = 0.9973; allLaNs > shiTS vs +/shiTS: p = 0.0418; allLaNs > shiTS vs allLaNs/+: p = 0.026. b LaNs are necessary for biasing the first head sweep direction. +/shiTS: n = 2130, p = 7.3 × 10−8; PVL09 > shiTS: n = 1740, p = 0.0104; PVL09/+: n = 3230, p = 5.9 × 10−13; Pdf-LaNs > shiTS: n = 2263, p = 5.9 × 10−13; Pdf-LaNs/+: n = 1582, p = 5.7 × 10−9; allLaNs > shiTS: n = 1841, p = 0.0622; allLaNs /+: n = 1211, p = 3.7 × 10−10. PVL09 > shiTS vs +/shiTS: p = 0.1372; PVL09 > shiTS vs PVL09/+: p = 0.0498; Pdf-LaNs > shiTS vs +/shiTS: p = 0.2675; Pdf-LaNs > shiTS vs Pdf-LaNs/+: p = 0.8683; allLaNs > shiTS vs +/shiTS: p = 0.0498; allLaNs > shiTS vs allLaNs/+: p = 8.9 × 10−4. c Larvae lacking functional LaNs don’t bias their turn direction. +/shiTS: n = 702, p = 3.3 × 10−12; PVL09 > shiTS: n = 843, p = 0.0084; PVL09/+: n = 1366, p = 3.3 × 10−12; Pdf-LaNs > shiTS: n = 900, p = 6.4 × 10−11; Pdf-LaNs/+: n = 606, p = 1.2 × 10−9; allLaNs > shiTS: n = 915, p = 0.0551; allLaNs /+: n = 515, p = 1.4 × 10−4. Larvae lacking functional PVL09 show a decreased probability to bias the turn direction compared to controls. PVL09 > shiTS vs +/shiTS: p = 0.0027; PVL09 > shiTS vs PVL09/+: p = 0.0474; Pdf-LaNs > shiTS vs +/shiTS: p = 0.5239; Pdf-LaNs > shiTS vs Pdf-LaNs/+: p = 0.5892; allLaNs > shiTS vs +/shiTS: p = 4.5 × 10−4; allLaNs > shiTS vs allLaNs/+: p = 0.0788. d Larvae with silenced PVL09 or LaNs accept first head sweeps during both phases with same probability. +/shiTS: n = 4260, p = 1.5 × 10−4; PVL09 > shiTS: n = 3463, p = 0.9368; PVL09/+: n = 4604, p = 3.0 × 10−5; Pdf-LaNs > shiTS: n = 4263, p = 2.7 × 10−5; Pdf-LaNs/+: n = 4107, p = 1.9 × 10−4; allLaNs > shiTS: n = 2788, p = 0.7371; allLaNs/+: n = 4420, p = 6.2 × 10−8. Loss of this bias is due to an increased acceptance rate during intensity increase. PVL09 > shiTS vs +/shiTS (intensity decrease): p = 0.1477; PVL09 > shiTS vs +/shiTS (intensity increase): p = 0.0041; PVL09 > shiTS vs PVL09/+ (intensity decrease): p = 0.9226; PVL09 > shiTS vs PVL09/+ (intensity increase): p = 2.7 × 10−7; Pdf-LaNs > shiTS vs +/shiTS (intensity decrease): p = 0.4548; Pdf-LaNs > shiTS vs +/shiTS (intensity increase): p = 0.2428; Pdf-LaNs > shiTS vs Pdf-LaNs/+ (intensity decrease): p = 0.1063; Pdf-LaNs > shiTS vs Pdf-LaNs/+ (intensity increase): p = 2.7 × 10−7; allLaNs > shiTS vs +/shiTS (intensity decrease): p = 1; allLaNs > shiTS vs +/shiTS (intensity increase): p = 5.8 × 10−7; allLaNs > shiTS vs allLaNs/+ (intensity decrease): p = 0.1477; allLaNs > shiTS vs allLaNs/+ (intensity increase): p = 1.5 × 10−5. e Larvae make greater turns when intensity increases. +/shiTS: n = 10, p = 3.8 × 10−8; PVL09 > shiTS: n = 10, p = 1.2 × 10−4; PVL09/+: n = 10, p = 0.0022; Pdf-LaNs > shiTS: n = 10, p = 1.4 × 10−4; Pdf-LaNs/+: n = 10, p = 1.5 × 10−4; allLaNs > shiTS: n = 10, p = 0.0015; allLaNs/+: n = 10, p = 1.2 × 10−9. Larvae with LaNs silenced show a decreased turn size bias compared to controls. One-way ANOVA of turn magnitude bias: p = 5.4 × 10−7. Dunnett’s test: PVL09 > shiTS vs +/shiTS: p = 0.0024; PVL09 > shiTS vs PVL09/+: p = 0.7852; Pdf-LaNs > shiTS vs +/shiTS: p = 0.5408; Pdf-LaNs > shiTS vs Pdf-LaNs/+: p = 0.6846; allLaNs > shiTS vs +/shiTS: p = 0.0016; allLaNs > shiTS vs allLaNs/+: p = <0.001. The data are mean ± SEM. a, e One sample t-test was used. b, c Exact binomial and b–d Fisher’s exact test were applied. a–e Benjamini Hochberg procedure was used. *p < 0.05, **p < 0.01, ***p < 0.001, n = not significant