Extended Data Fig. 3: PFKL acts as a protein kinase and phosphorylates PLIN2 S159.
f, l, r, s, Immunoblotting analyses with the indicated antibodies were performed, presenting representative results from three independent experiments. a, b, The crystal structures of PFKP dimer (PDB: 4XYJ) (a) and PFKL dimer generated by homology modeling based on PFKP (b). c, Molecular dynamics simulations of the nonphosphorylated PFKL and the T331-phosphorylated PFKL dimers were performed. d, Coomassie blue staining of purified tag-free PFKL proteins. e, The activity of the WT PFKL (mock-phosphorylated control), WT PFKL (pT331), PFKL T331A and T331D mutants with respect to F6P titrations (ATP 0.5 mM) or ATP titrations (F6P 4 mM) is presented. Two preparations of protein were used. f, Flag-rPFKL proteins were expressed in Huh7 cells with depletion of endogenous PFKL (left). The media were collected for analysis of remaining glucose (middle) or lactate production. n = 3. The P values were labeled in front of the corresponding data points (right). g-k, Huh7 cells with depleted endogenous PFKL and reconstituted expression of Flag-rPFKL proteins were subjected to ECAR analysis by Seahorse; n = 8 (g), glycolytic metabolites analysis by LC-MS/MS; n = 3 (h), metabolic flux analysis by LC-MS/MS; n = 3 (i, k), glycogen level analysis; n = 3 (j). l, An in vitro kinase assay was performed by mixing Flag-PFKL with His-PLIN2 in the presence of [γ-32P]ATP. Autoradiography and immunoblotting analyses were performed. m, The binding ability of ATP to His-SUMO-PFKL proteins was measured by biolayer interferometry (Sartorius Octet) assay. Representative results from three independent experiments are presented. n, Enzyme kinetics plots of velocity relative to PLIN2 concentrations between PFKL WT and T331D. The Vmax and Km of PFKL in phosphorylating PLIN2 were calculated. n = 3. o, Stoichiometry of PLIN2 phosphorylation by PFKL T331D was presented. n = 3. p, An in vitro phosphorylation assay was performed by mixing His-PLIN2 with Flag-PFKL. LC-MS/MS analyses showed a tryptic fragment at m/z 805.85974 Da (-6.52 mmu/-8.09 ppm), which was matched with the +4 charged peptide 154-SVVSGp(S)INTVLGSRMMQL VSSGVENALTK-182 of PLIN2, suggesting that PLIN2 S159 was phosphorylated. The XCorr score was 2.8. q, Alignment of protein sequences spanning PLIN2 S159 in different species (top). The human PLIN2 S159 peptide sequence and the corresponding sequence of PLIN3 were aligned (bottom). r, WT Flag-rPLIN2 and Flag-rPLIN2 S159A were expressed in Huh7 or LM3 cells with depletion of endogenous of PLIN2 (top). PLIN2 depletion efficiency was quantified. n = 3 (bottom). s, Huh7 cells stably expressing shCtrl or shPLIN2 were treated with 40 min of glucose deprivation. t, IHC analyses of human HCC samples were performed with the indicated antibodies in the presence or absence of a blocking peptide for PLIN2 pS159. Representative images from three independent experiments are shown. u, Molecular docking analyses of F6P in the catalytic domain of nonphosphorylated PFKL (top) or T331-phosphorylated PFKL (bottom). e-k, n, o, r, Data are presented as means ± SD of three biologically independent replicates, analyzed by two-sided unpaired Student’s t-test (h, r), one-way ANOVA (f, g) or two-way ANOVA (i, k).
