Extended Data Fig. 9: Behavioral controls of nebu and ALTA KD in glia, and glucose imaging for nebu and InR (related to Fig. 7).
From: Glycolysis-derived alanine from glia fuels neuronal mitochondria for memory in Drosophila
a. Without induction of nebu RNAi, memory after spaced training (n = 10, F2,27 = 1.31, P = 0.29) was normal. b. Behavioral results were confirmed with a second RNAi against Nebu. LTM was impaired (n = 12, F2,33 = 7.47, P = 0.002), but not memory after single-cycle (n = 15, F2,42 = 2.40, P = 0.10; cold shock: n = 15, F2,42 = 1.15, P = 0.33), or massed training (n = 10, F2,27 = 0.96, P = 0.40). Without induction, LTM was normal (n = 10, F2,27 = 0.60, P = 0.55). c. Glucose concentration measurements corresponding to data in Fig. 7d (see also Supplementary Video 3). Validamycin A (4 mM, dashed line) was applied to measure the sensor floor level, providing an internal reference. d. Mild decrease in both nebu and InR expression in cortex glia impaired memory after spaced training, whereas the separate mild induction of either RNAi did not (n = 22 for all datasets except plain blue and purple bars (n = 21), F4,103 = 6.03, P = 0.00021). e. ALAT KD in ensheathing glia did not affect LTM (n = 12, F2,33 = 0.76, P = 0.48). f. Without induction of ALAT RNAi in astrocyte-like glia, LTM was normal (n = 12, F2,33 = 0.23, P = 0.80). g. Immunohistochemistry of ALAT-HA; Alrm-GAL4 > mCD8::GFP brain (dashed line delimits MB lobes). Scale bar: 20 µm. h. Behavioral experiments with a second RNAi against ALAT in astrocyte-like glia. LTM was impaired (n = 18, F2,51 = 5.76, P = 0.006), but not memory after single-cycle (n = 11, F2,30 = 0.38, P = 0.69; cold shock: n = 11, F2,30 = 0.43, P = 0.66), or massed training (n = 18, F2,51 = 1.96, P = 0.15). Without RNAi induction, LTM was normal (n = 18, F2,51 = 0.53, P = 0.59). All data are presented as mean ± SEM. Asterisks (*P < 0.05; **P < 0.01; ***P < 0.001; ns: not significant, P > 0.05) illustrate the least significant pairwise comparison following 1-way ANOVA.
