Fig 2: Phafin2 associates with Akt in mammalian cells. A. Flag-Phafin2 interacted only with HA-Akt, but not with HA-PDK1 or HA-PrKA. Similar levels of expression of HA-Akt, PDK1, PrKA, and Flag-Phafin2 were shown by immunoblot (HA = anti-HA antibody; F = anti-Flag antibody; C = Control antibody). B. Flag-tagged Phafin2 interacted with HA-tagged Akt1 (lane 1–3) and Akt2 (lane 4–6), but not with Akt3 (lane 7–9). Similar levels of three Akt isoforms and Phafin2 were expressed (lower panels). C. Endogenous Akt1 and Akt2 interacted with Phafin2 in HT1080 cells by co-immunoprecipitation assays. Expression of Akt isoforms and Phafin2 were shown (lower panels). D. The C-terminal Akt kinase domain is the binding domain for Phafin2 interaction in co-immunoprecipitation assays. Expression of the Flag-Phafin2 and HA-Akt subfragments were shown (lower panels). E. Structural features of Phafin2 used in this study are shown. Phafin2 consisted of N-terminal PH domain and C-terminal FYVE domain. Both PH domain and FYVE domain interacted with Akt in co-immunoprecipitation assays. Expression of Akt and Phafin2 subfragments were shown (lower panels). F. Interaction between recombinant Phafin2 and Akt was verified in GST pull-down assays. Recombinant active (a) or unactive (un) Akt was incubated with GST-Phafin2 beads and subsequently resolved onto SDS-PAGE and detected by immunoblotting using anti-Akt antibody (right panel). Levels of phosphorylation of active Akt (a) and unactive Akt (un) were shown (lower panels).

Fig 3: Identification of upstream transcription factors mediating the overall transcriptomic changes and insulin receptor expression. (A) Transcription factor (TF)-gene network predicting upstream transcriptional regulators of differentially expressed genes by hyperinsulinemia (BS, 200 vs. 0 nM). Names of the top 30 TFs are labeled. Genes that are up or down-regulated are shown in red or blue dots, respectively. TFs that are not differentially expressed are in grey rhombus. (B) Common TFs between the top 30 TF nodes from (A) and the TF binding sites for several key genes in insulin signaling pathways that are downregulated by hyperinsulinemia. Adjusted p values of differentially expressed TFs (200 vs. 0 nM, both BS and AS) are labeled next to the heatmap. (C) mRNA levels of the common TFs in (B) after siRNA knockdown. (D) The effects of each siRNA knockdown on Insr and Irs2 mRNA levels. (n = 4. *p < .05, 1-ANOVA followed by Dunnett multiple comparison test against siControl.) (E) Graphic summary of our current model. Hyperinsulinemia induced sustained phosphorylation of INSR and AKT, which resulted in the inhibition of FOXO1 leading to reduced Insr transcription. Downregulated INSR and post-receptor components resulted in reduced insulin signaling upon acute insulin stimulation. SIN3A, which was upregulated by hyperinsulinemia, represses Insr transcription

Fig 4: Further evidences supporting the involvement of mTOR in tumorigenesis upon mutations of MLL2.A. Mutual exclusivity of driver alterations of MLL2 and genes upstream and in the mTOR pathway. (Mutex p-value: 5.4×10-5) B. Loss-of-function mutations of MLL2 concomitant with miss-regulation of its related DE genes possess predictive survival value. HNSC tumors were separated in two groups: those bearing mutations of MLL2 and concomitant miss-regulation of related down-regulated genes (red curve), and those without mutations of MLL2 and no sign of down-regulation of the same genes. (A) Left panel. The levels of MLL2 of lysates of T24 cells infected with an irrelevant short hairpin RNA (shControl) or specific for MLL2 (shMLL2) were checked by real-time quantitative RT-PCR (qRT-PCR). Gene expression was normalized against an endogenous control and represented as RNA levels relative to those obtained in shControl-infected cells, which was set to 1. Right panel. The lysates were analysed by western blot with an anti-P-4E-BP1, 4E-BP1, P-AKT, AKT and Tubulin antibodies. (B) Knock-down of MLL2 increased T24 cells sensitivity to everolimus treatment. The proliferation of both shControl and shMLL2 cells treated with everolimus in the course of 3 days (three replicates in each point) is presented relative to the proliferation of shControl and shMLL2 untreated cells, respectively. The units in the abscissa represent a proliferation ‘fold change’.

Fig 5: Distribution of pan-Akt, pT308-Akt and pS473-Akt. (A) Top panel shows an example of an axon stained with a pan-Akt antibody. The distribution is uniform. (B) Graph showing the mean staining intensity of pan-Akt staining in axons (n = 42) in segments devoid of mitochondria and those populated by mitochondria. An example of an axon stained with anti-pan-Akt and containing labeled mitochondria is shown below the graph. (C) Example of pT308-Akt staining along an axon treated with NGF for 10 min. The top panel shows a phase contrast image of the axon. (D) Example of pS473-Akt staining along an axon treated with NGF for 10 min. The top panel shows a phase contrast image of the axon with overlaid pS473-Akt. The scale bar in (D) also applies to (A,C).