Fig 2: Tankyrases Regulate AMPK Activation through ribosylating LKB1. a Identification of the tankyrase inhibitor JW55 as an AMPK activator. A drug screen for regulators of AMPK was done by treating U2OS cells with a set of 406 drugs (the small-molecule Informer-Set contains 481 drugs, but we can only get 406 drugs from commercial sources) and metformin/AICAR. Both JW55 and JW-55 were the same compounds from two different commercial sources. b Tankyrase inhibitors specifically induce AMPK activation. HEK293A cells were treated with indicated inhibitors and proteins were detected by western blotting. c Double-knockdown of TNKS1/2 induces AMPK activation. HEK293A cells were transduced with the indicated shRNA and subjected to western blotting. d Depletion of TNKS1/2 suppresses expression of AMPK downstream glycolytic genes. Relative expression levels of PDK1, FAS SREBP-1c, and TNKS1/2 were detected by qPCR. e Interactions between endogenous LKB1 and TNKS1. Immunoprecipitation/western blotting were performed with the indicated antibodies. f LKB1 is ribosylated in vivo. Lysates from HEK293A cells were subjected to immunoprecipitation/western blotting assays. g LKB1 is ribosylated by TNKS1 in vitro. In vitro ribosylation was assessed by using recombinant TNKS1/LKB1 (E.coli purified GST-LKB1 and enzymatic active form of TNKS1, amino acids 1000–1328 with GST-tag). h E130 and E138 on LKB1 are key residues for TNKS1-induced ribosylation of LKB1. In vitro ribosylation was assessed by using the indicated proteins. i Structure view of LKB1/STRAD/MO25 complex. The LKB1-E130 residue is located close to the interface between LKB1 and STRAD. j Structure view of LKB1. The LKB1-E138 residue is located close to the ATP-binding pocket of LKB1. Statistical significance was determined by a two-tailed, unpaired Student’s t-test. *P < 0.05; **P < 0.01; ***P < 0.001

Fig 3: Molecular analysis of LKB1 reexpression in human PCa cells. (A) Western blot analysis for LKB1 and MO25 protein expression of input and flag immunoprecipitates in DU145 cell line with ectopic expression of Flag-tagged LKB1WT and LKB1K78I. (B) Picture of SDS-PAGE gel with separation of immunoprecipitated proteins. Blue squares indicate areas selected for subsequent LC-MS–based identification and quantitation. (C) Representative analysis of relative MO25α and MO25β mRNA abundance by qRT-PCR in DU145 cells (n = 3). (D) Representative analysis of MO25α gene expression by qRT-PCR in DU145 cells transduced with scramble shRNA (shCtrl) or two independent shRNA (shMO25α 02; shMO25α 04; n = 3). (E and F) Effect of MO25α gene silencing on cellular growth (E; n = 3; independent experiments) and anchorage independent growth (F; n = 5; independent experiments; data are normalized to control; shCtrl, dashed line). (G) Representative Western blot analysis using AMPK motif antibody in control and LKB1WT- and LKB1K78I-expressing cells (n = 3). (H) Representative Western blot analysis of LKB1, p-AMPK, p-ACC, p-Raptor, Raptor, and β-actin protein expression in control and LKB1WT- and LKB1K78I-expressing cells (n = 5). (I) Analysis of LKB1 gene expression by qRT-PCR in DU145 cells with ectopic expression of LKB1WT, LKB1K78I, and LKB1D194A or transduced with mock vector (pBABE; n = 3; independent experiments). Data are normalized to control (pBABE; dashed line). (J) Effect of LKB1WT, LKB1K78I, and LKB1D194A expression on cell growth (n = 3; independent experiments). (K) Western blot analysis using AMPK motif antibody in control and LKB1WT-, LKB1K78I-, and LKB1D194A-expressing DU145 cells. (L) Representative Western blot analysis for LKB1, p-AMPK, p-ACC, p-Raptor, Raptor, and HSP90 protein expression in control and LKB1WT-, LKB1K78I-, and LKB1D194A-expressing DU145 cells (n = 5). (M) Western blot analysis of LKB1 protein expression in harvested xenograft tumors (n = 5 samples per condition). (N) Effect of LKB1 reconstitution in DU145 cells with ectopic expression LKB1WT, the kinase-defective mutants (LKB1K78I, LKB1D194A) or infected with mock vector (pBABE) on tumor weight (left). Representative picture of harvested tumors (right) in xenotransplantation experiments (n = 20 injections per condition). (O) Quantification (left) and representative image (right) of Ki67 staining in harvested xenograft tumors in control and LKB1WT-, LKB1K78I-, and LKB1D194A-expressing DU145 cells. Statistic test: one-sample t test (I); one-tailed Mann–Whitney test (N and O). *, P < 0.05; ****, P < 0.0001. Error bars represent SD (C, E, and F) and SEM (I, J, N, and O).

Fig 4: Taurine activates the AMPK/NRF2 pathway by promoting the formation of the LKB1-STRAD-MO25 kinase complex (A) Quantitative abundance of taurine binding with LKB1-STRAD-MO25 complex subunits was measured using an LC-MS/MS-based trace-level detection method, n = 3. (B-C) Analysis of the binding of Taurine to purified LKB1 and MO25 WT proteins using surface plasmon resonance assay (SPR). (D) Plots of RMSD estimate averaged over all 10 trials versus simulation time for taurine docked with LKB1-STRAD-MO25 complex were calculated using Binding-pose metadynamics. A docking pose with a low PoseScore and high PersScore suggests a stable and reliable binding mode. (E) 3D diagram showing the most stable binding mode of taurine and LKB1-STRAD-MO25 (IFD2). (F) The LKB1 (Glu165 Arg301) and MO25 (Arg194 Leu197) sites were mutated to Ala, and quantitative abundance of taurine binding with LKB1-STRAD-MO25 complex subunits was measured using an LC-MS/MS-based trace-level detection method, n = 3. (G) The difference in the binding ability of LKB1 to MO25 in WT-NPCs and LKB1&MO25 point mutant NPCs (Mut-NPCs) was detected by co-immunoprecipitation after A1TP-HX-EVs treatment. Statistical analysis was performed using a two‐tailed unpaired Student's t‐test. n = 3, ∗∗P < 0.01. (H) WT-NPCs and Mut-NPCs were induced with TBHP, and then treated with A1TP-HX-EVs for 24 h. Cell lysates were immunoblotted with antibodies against IVDD markers and ferroptosis-related proteins. (I) Cell lysates were immunoblotted with antibodies against AMPK signaling pathway-related proteins. (J) Representative images of mitochondrial morphology of TBHP-induced WT-NPCs and Mut-NPCs using TEM after 24 h of treatment as indicated. n = 3. Scale bar, 500 nm. (K-L) Representative oxygen consumption traces of WT-NPCs and Mut-NPCs induced with TBHP and then treated with A1TP-HX-EVs for 24 h. Maximal respiration of NPCs were quantified. All data are expressed as the mean ± SD. One‐way ANOVA with Tukey's multiple comparison tests were used for statistical analysis. n = 3, ∗P < 0.05. ∗∗P < 0.01. ns, not significant.

Fig 5: AMPK activation upon eIF4F inhibition in melanoma bypasses the canonical AMPK activator, LKB1. A Western blot analysis of A375 and MelJuso (NRAS-mutant) melanoma cells treated with increasing concentrations of the eIF4A inhibitor Rocaglamide A (RocA) for 20 h. Both cell lines exhibited dose-dependent increases in phosphorylated AMPK, accompanied by decreased levels of LKB1 and its co-factor MO25. B Western blot analysis of G361 (BRAFV600E-mutant, LKB1-null) melanoma cells treated with increasing concentrations of RocA for 20 h. AMPK activation occurred in a dose-dependent manner despite the absence of functional LKB1, confirming LKB1-independent AMPK activation. C Western blot analysis of A375, MelJuso, and G361 cells treated with increasing concentrations of RocA for 20 h. The increased phosphorylation of a direct AMPK substrate, Acetyl-CoA Carboxylase, Ser79 (p-ACC), demonstrates enhanced downstream catalytic activity of AMPK in response to eIF4F inhibition. D Western blot analysis of A375, MelJuso, and G361 cells treated with increasing concentrations of the eIF4E-eIF4G disruptor 4E1RCat for 20 h. All three cell lines exhibited elevated p-AMPK and p-ACC levels regardless of LKB1 status or driver mutation. E Western blot analysis of A375 and MelJuso cells transfected with two different LKB1-targeting siRNAs for 48 h, combined with 100 nM RocA treatment for the last 20 h. LKB1 knockdown did not prevent RocA-induced AMPK activation, confirming the LKB1-independent mechanism. Non-targeting control siRNA (si-NT) was used as a control. The upper index letters refer to the corresponding loading control detected on the same membrane