Fig 2: Rescue of Insulin resistance upon NGF stimulation occurs mainly by IRS1 activation and TrkA/IRS1 binding. (a–f) Densitometric analyses of AKT (a, b; respectively), c-Fos staining (c, d; respectively) and Glucose uptake (e, f; respectively) in control (a, c, e) and insulin resistant (b, d, f) cholinergic neurons. (a-b) Levels of pAKT, total AKT, and ß-actin are reported (a) in control (CTR), NGF-treated (NGF), and NGF-treated neurons preincubated with the specific IRS inhibitor NT157 (NGF + NT157), and (b) in RI, RN, and RN following incubation with NT157 (RN + NT157). NGF increases AKT activation in control (NGF, 248.9 ± 14.5% of CTR; **p < 0.01; Fig. 4a) and insulin resistant (RN, 359.4 ± 68.9% of CTR; **p < 0.01; Fig. 4b) cholinergic neurons, but fails to modulate it in presence of NT157 in both control (NGF + NT157, 68.6 ± 9.3% of CTR; **p < 0.01; Fig. 4a) and insulin resistance conditions (RN + NT157, 93.1 ± 15.1% of CTR; **p < 0.01, Fig. 4b). (c-d) Quantitative analysis of the c-Fos positive nuclei/field (× 20 magnification) in (a) CTR, NGF, NGF-treated neurons in presence of a specific IRS inhibitor NT157 (NGF + NT157), and in (d) insulin-resistance (RI), insulin-resistance NGF-treated (RN), and insulin-resistance NGF-treated in presence of a specific IRS inhibitor NT157 (RN + NT157). NGF elevates the number of c-Fos expressing neurons in control (383.3 ± 74.1% of CTR; **p < 0.01; Fig. 4c), as well as in insulin resistant cholinergic neurons (210 ± 40.2% of CTR; **p < 0.01; Fig. 4d), but its effect is lost following IRS inhibition by NT157 both in control (96.1 ± 35.5% of CTR; NGF + NT157 vs NGF, **p < 0.01; Fig. 4c) and insulin resistant conditions (96 ± 35.7% of CTR; RN + NT157 vs RN, **p < 0.01; Fig. 4d). (e-f) Quantitative analysis of glucose uptake in (e) control untreated (CTR), NGF-treated (NGF), NGF-treated with NT157 (NGF + NT157), and in (f) insulin-resistant (RI), insulin-resistant NGF-treated (RN), and insulin-resistant NGF-treated neurons preincubated with NT157 (RN + NT157). NGF stimulated glucose uptake in control (183.2 ± 33% of CTR, *p < 0.05; Fig. 4e) and insulin resistant neurons (140.2 ± 12% of CTR, **p < 0.01; Fig. 4f), but lost its efficacy after preincubation with NT157 in control (35.3 ± 7.5% of CTR; NGF + NT157 vs NGF, **p < 0.01; Fig. 4e) and insulin resistant (11.3 ± 6.5% of CTR; RN + NT157 vs RN, **p < 0.01; Fig. 4f) conditions, where glucose uptake is lower than in RI condition (RN + NT157 vs RI, *p < 0.05). The fluorescence intensity was calculated as percentage of control (CTR) and reported as mean ± sem. CTR neurons were starved DIV10 cholinergic neurons in control conditions and starved DIV13 neurons in insulin resistant experiments. (g) Glut2 immunofluorescent labeling in cholinergic neurons (DIV10) treated with NGF (100 ng/ml, 30') with and without preincubation with NT157 (10 µM, 2 h). Specific anti-Glut2 antibodies were used and detected Glut2 immunoreactivity mainly in the cytosolic compartment of both the cell bodies (arrows) and neurites (arrowheads) in untreated neurons (CTR). NGF incubation (NGF) drives plasma membrane translocation of Glut2, both at the cell bodies (arrows) and neurites (arrowheads) in absence of NT157. While NT157 preincubation (NGF + NT157) hampers the effect of NGF on Glut2 translocation, Glut2 labeling being mainly cytosolic, as observed in CTR neurons. (h,k) Detection of TrkA-pIRS1Y608 (h) and TrkA-IRS1 (k) interactions by PLA assay in cholinergic neurons (DIV10) treated with NGF (100ng/ml, 30'), with and without preincubation with NT157 (10µM, 2 h). Neurons were fixed with PFA, blocked with normal donkey serum (10%, 1 h, RT) and incubated (ON, 4 °C) with a combination of mouse anti-Trk and rabbit anti-pIRS1Y608 or mouse anti-Trk and rabbit anti-total IRS1. Detection and amplification of single events of interaction by a single dot (red) was achieved by the PLA assay (PLA Duoset, Sigma). Nuclei were counterstained with DAPI. PLA staining was analyzed by an observer blind of the experimental group and representative picture were randomly chosen. (i-j) Quantification of the number (i) and intensity (j) of the TrkA-pIRS1Y608 PLA signal dots in primary neurons untreated (CTR, 13 ± 4 dots/neuron and 0.02 ± 0.002 intensity/dot), treated with NGF (NGF, 47.2 ± 7.3 dots/neuron, ***p < 0.001 and 0.04 ± 0.002 intensity/dot, **p < 0.01) or NGF following NT157 preincubation (NGF + NT157, 2.2 ± 0.65 dots/neuron; NGF + NT157 vs NGF, ***p < 0.001; and 0.008 ± 0.0002 intensity/dot, NGF + NT157 vs NGF, ***p < 0.001). (k-l) Quantification of the number (k) and intensity (l) of the TrkA-IRS1 PLA signal dots in primary neurons untreated (CTR, 50.6 ± 5.9 dots/neuron and 0.11 ± 0.015, intensity per dot), NGF treated (121.5 ± 16.4 dots/neuron, *p < 0.05; 0.59 ± 0.12 intensity/dot, *p < 0.05), or NGF treated following NT157 preincubation (NGF + NT157, 45.8 ± 9.6 dots/neuron NGF + NT157 vs NGF; *p < 0.05; 0.12 ± 0.02 intensity per dot, NGF + NT157 vs NGF *p < 0.05). The quantitative analysis of the PLA fluorescent signal was performed in five different fields, for a total of 20 to 30 neurons and was expressed as mean ± sem. Integrated intensity per dot is expressed in arbitrary unit (AU) and reported as mean ± sem. Scale bars: g = 50 µm; h,k = 25 µm. CTR neurons in a, c, e are DIV10 untreated neurons; CTR neurons in b, d, f are DIV13 untreated neurons

Fig 3: Knockdown of ErbB2 decreases phosphorylation of Akt and CNC migration.(A) Western blot of non-injected (NI) and injected embryo extracts probed for flag. Single-cell embryos were injected with 800 pg DN-ErbB2-Flag mRNA with or without 25 ng of ErbB2 morpholino (MO ErbB2). Translation of DN-ErbB2-flag was hindered in embryos co-injected with MO ErbB2. GAPDH is shown as a loading control. One-embryo equivalents are loaded per lane. (B) Western blot of CNC extracts probed for phospho-Akt, total Akt and GAPDH. Eight-cell embryos were either injected with 200 pg RFP mRNA alone (N = 4, n = 32) or with 3.1 ng MO Erb2 (N = 3, n = 43) into CNC progenitor cells. CNC cells were then dissected from embryos once they reached stage 17–18. Injection of MO ErbB2 into CNC cells precursors dramatically reduced phosphorylation of Akt without altering the levels of control proteins. Ten CNC explants are loaded in each lane. (C-E) Early tailbud stage embryos labeled for Twist and Sox10 CNC markers using in situ hybridization. Embryos were injected as in part (B) along with a set injected with 300 pg rat ErbB2-flag and 1.6 ng MO ErbB2 (E; N = 2, n = 20). Embryos injected with RFP mRNA and MO ErbB2 showed defects in CNC migration (D) compared to RFP-injected controls (C). Migration was partially rescued by co-injecting the morpholino with rat ErbB2-flag mRNA (E). (F) CNC migration was measured in each branchial arch and normalized to head size from embryos in part (C-E). Co-injection of MO ErbB2 with RFP mRNA significantly reduced CNC migration in the mandibular (M), hyoid (H), 3rd and 4th branchial arches. Co-injection of MO ErbB2 with rat ErbB2 mRNA partially rescued migration. Error bars are one standard deviation to the mean. One-tailed, Student’s t-tests were performed to determine statistical significance. ** p<0.01, *** p<0.001. N, number of experiments; n, number of embryos.

Fig 4: Examples on identification and de-convolution of active regulators.(A–B) identification of the MAPK/Erk activator fbxo43 (erp1). (A) Western blot of stage 8 embryos injected with 12 pools (01–12) derived from plate #01 and 7 pools (01–07) from plate #02 and probed with anti-phospho-Erk (pErk) antibody. The arrow indicates increased Erk phosphorylation upon injection of mRNAs derived from plate #02, pool 07. (B) De-convolution of the above pool. Embryos injected with single RNAs were collected at stage 8 and uninjected lysate from stage 8 and stage 10.5 were used as negative and positive control of Erk phosphorylation respectively. The arrow indicates the active clone of TEgg009F05, identified in plate #2, column 08, row G. This clone was confirmed as the X. tropicalis fbxo43 (erp1) gene. (C–D) Identification of PI3K/Akt inhibitor prkaca. (C) Western blot of stage 10.5 embryos injected with 24 pools (01–12) derived from plate #09 and #10 and probed with anti-phospho-Akt (pAkt) antibody. The arrow indicates decreased Akt phosphorylation upon injection of mRNAs derived from plate #10, pool 02. (D) De-convolution of the above pool. mRNA synthesis, injection, and Western blot were performed as in (B) except that stage 10.5 embryos were used. The arrow indicates the active clone of TEgg046d13 is identified in plate #09, column 02, row E. This clone was later identified as encoding the X. tropicalis prkaca gene.

Fig 5: Akt inactivation mediates apoptosis in IM-9 cells(A) IM-9 and IM-9/ Bcl-2 cells were treated with BetA for different periods of time. Treated cells were lysed for detecting Akt phosphorylation (p-Akt), total Akt (t-Akt), Bax oligomerization, conformational change and mitochondrial translocation by Western blotting, with ß-Actin serving as a loading control. Bax conformational change was detected as described before (Hu et al., 2012b). (B) IM-9 and IM-9/ Bcl-2 cells were transfected with the constitutively active Akt1 for 48 h, and then treated with BetA for 48 h. treated cells were lysed for Western blotting. (C) Cells were treated with BetA and/or 25 mM LY294002 for 72 h, and then cells were lysed and assayed for individual protein levels by Western blot. (D) Cells were treated with BetA for 48 h. Top row, treated cells were collected for detection of cell viability. Graphs showing results of quantitative analyses (n=3, mean ± S.D. *,P<0.05). Down, Apoptotic cell death was quantitatively detected by a cell death ELISA kit as described in Materials and methods. Graphs showing results of quantitative analyses (n=3, mean ± S.D. *,P<0.05). Representative results of three experiments with consistent results are shown.