Fig 2: Dysregulation of the oxidative stress response causes learning impairment in 330°C -4518°C Aß flies. (A) Both XBP1t and XBP1s mRNA levels were reduced in the brain of 330°C -4518°C Aß flies, but the ratio of XBP1s/XBP1t remained similar, N = 3 respectively. ****p <0.0001 in XBP1t, #### p<0.0001 in XBP1s. (B) The level of protein ubiquitination was not changed in the brain of 330°C -4518°C Aß flies. The quantitative data is shown on the right, N = 7 respectively. (C) The proteasome activity was not changed in brain of 330°C -4518°C Aß flies, N = 3 respectively. (D) The amount of Ref2P was increased in the brain of 330°C -4518°C Aß flies. The quantitative data is shown on the right, N = 12 respectively. *p< 0.05. (E) The GABARAP II/I ratio was not changed in the brain of 330°C -4518°C Aß flies. The quantitative data is shown on the right, N = 15 respectively. (F) Rapamycin treatment for 4 weeks could not improve the learning impairment in 330°C -4518°C Aß flies. (Ctrl (3-45) vs. Aß42 (3-45): p<0.0001 respectively; Aß42 (3-45) vs. Aß42+Rapamycin (3-45): p > 0.9999, N = 6 respectively). **** p < 0.0001. (G) Vitamin E, melatonin (M) or lipoic acid (L) treatment for 4 weeks improved the learning impairment in 330°C -4518°C Aß flies. (Left: Ctrl (3-45) vs. Aß42 (3-45): p < 0.0001; Aß42 (3-45) vs. Aß42+Vit. E (3-45): p = 0.0002, respectively, n = 6; Right: Ctrl (3-45) vs. Aß42 (3-45): p < 0.0001; Aß42 (3-45) vs. Aß42+L (3-45): p < 0.0001; Aß42 (3-45) vs. Aß42+M (3-45): p < 0.0001, N = 6 respectively). *** p < 0.001, **** p < 0.0001. (H) Vitamin E, melatonin (M) or lipoic acid (L) treatment for 4 weeks decreased the levels of ROS in 330°C -4518°C Aß flies. The quantitative data is shown on the right (Right Top: Ctrl (3-45) vs. Aß42 (3-45): p = 0.0270; Aß42 (3-45) vs. Aß42+Vit. E (3-45): p = 0.0438; Right bottom: Ctrl (3-45) vs. Aß42 (3-45): p = 0.0389; Aß42 (3-45) vs. Aß42+L (3-45): p = 0.0201; Aß42 (3-45) vs. Aß42+M (3-45): p = 0.0391, N = 3 respectively). * p < 0.05. ns represents non-significant. All Ctrl represents Elav-Gal4+Gal80TS. White bar indicates 50 µm.

Fig 3: H/R condition disturbs Atg12–Atg5, LC3/GABARAP conjugation systems, and their related regulators in PHTs. (A) Western blots of indicated molecules in normoxia (Nor)- and H/R-treated PHTs. (B) Quantification of the intensity of bands shown in (A). Data are expressed as mean ± SEM and analyzed by a Student’s t-test (n = 3). *: p < 0.05, n.s.: not statistically significant.

Fig 4: Endogenous Kenny protein accumulates in Atg8a and Atg7 mutant flies. a, b Full body lysates from wild-type (WT) and kenny mutant (key 1) flies a or Atg8a and Atg7 mutant flies b were subjected to SDS-PAGE and immunoblotting for Kenny. Tubulin was used as loading control. c Quantification of the quantity of Kenny protein normalized to tubulin. Bar chart shows means ± s.d. Statistical significance was determined using one-way ANOVA, *P < 0.05, ***P < 0.001. d–l Confocal images from adult brains d–f, fat bodies g–i and midguts j–l from WT d, g, j, Atg8a e, h, k and Atg7 f, i, l mutant flies stained for Kenny (green) and DNA (blue). Arrowheads show some Kenny aggregates. Scale bars are 20 µm

Fig 5: Utility of transgenic reporter lines in RNAi based genetic screen. (A) Germaria of nosP GFP-Ref(2)P nos 3'UTR and nosP mCherry-Atg8a nos 3'UTR transgenic flies in background of either no knockdown (control) or knockdown of Atg5 by RNAi in the germline. Scale bar 10 µm. (B) Interleaved scatter graph showing the increase in GFP-Ref(2)P punctae upon Atg5 knockdown. Error bars represent SD in red and the mean is blue. n = 20 for both the data sets. ****p < 0.0001. (C) Interleaved scatter graph showing the decrease in mCherry-Atg8a punctae upon Atg5 knockdown. Error bars represent SD in red and the mean is blue. n = 15 for both the data sets. ***p < 0.001.