Fig 2: The stoichiometry of GNAI3 palmitoylation versus oleoylation depends on the levels of the lipids the cells are exposed to.a, b GNAI2 and GNAI3 are S-acylated with C15:0- and C17:0-azide by multiple acyltransferase enzymes in a redundant fashion. GNAI2 and GNAI3 acylation are only mildly reduced upon siRNA-mediated knockdown of ZDHHC7 (a), and still acylated upon combined knockdown of ZDHHC3 and ZDHHC7 (b). Metabolic labeling with 100 µM azido fatty acids was performed as in Fig. 1. Representative of three biological replicates. c Overexpression of V5-ZDHHC7 does not shift the balance in S-acylation of GNAI2 and GNAI3 proteins with C15:0- versus C17:0-azide. Metabolic labeling with 100 µM azido fatty acids was performed as in Fig. 1. Representative of four biological replicates. d V5-ZDHHC7 is modified by both C15:0- and C17:0-azide. Mutation of the active-site cysteine (C160A) blunts this acylation. Metabolic labeling with 100 µM azido fatty acids was performed as in Fig. 1. Representative of two biological replicates. e Exposure of cells to either C16:0 or C18:0 in the medium increases the fraction of GNAI protein that is either palmitoylated or oleoylated, respectively. Endogenous S-acylation of GNAI3-GFP was analyzed as in Fig. 2c in cells treated with different unlabeled fatty acids (100 µM for 3 h). f Exposure of cells to different fatty acids in the medium (100 µM for 3 h) shifts the intracellular pool of acyl-CoAs toward the corresponding fatty acid. Acyl-CoA levels were normalized to the total phosphatidyl choline (PC) content. Data are presented as mean ± SD of four biological replicates and the p-values were calculated by two-way ANOVA followed by Tukey’s post hoc test.

Fig 3: Exposure of cells to C18:0 shifts GNAI proteins out of detergent-resistant membranes (DRMs).a S-acylation of GNAI proteins is required for their membrane localization. Unlike wild-type (WT) GNAI proteins, GNAI mutants lacking S-acylation do not localize to the plasma membrane, detected by immunostaining (green: anti-V5, blue: DAPI, scale bar 25 µm). Representative of three biological replicates. b Exposure of cells to C18:0 or C16:0 does not alter the localization of endogenous GNAI2 to the cell membrane, detected by immunostaining (green: anti-GNAI2, blue: DAPI, scale bar 25 µm). Representative of two biological replicates. c, d Endogenous GNAI2 and GNAI3 are depleted from DRMs upon exposure of cells to C18:0 but not C16:0. DRMs are isolated by the protein flotation assay in discontinuous OptiPrep gradients after solubilization with 1% Triton X-100 (c). As markers of DRMs, flotillin-1 and caveolin-1 were used. Quantification of the relative amount of each protein in DRMs normalized to total cell content of that protein (sum of all fractions) is shown in (d). Representative of three biological replicates. e Schematic diagram summarizing the findings presented here on GNAI S-acylation.

Fig 4: LAMTOR1 and HRAS as examples of differential S-acylation on multiple acylation sites.A, a scheme illustrating the N-terminal sequence of LAMTOR1 with its lipid modifications. B, APE assay detects acylation sites on the endogenous LAMTOR1 in HeLa and HEK293E cells. A representative image of three biological replicates. C, APE assay detects that both Cys3 and Cys4 are acylated in LAMTOR1-HA. HeLa cells were transiently transfected with overexpression constructs and lysed for APE assay 24 h post transfection. WT LAMTOR1 was compared to acylation mutants upon detection with anti-HA antibody (single mutants C3S and C4S and the double mutant C3S/C4S). The endogenous GNAI2 with its single acylation is shown as a positive control of the assay and tubulin as a loading control. A representative of three independent experiments. D, both Cys3 and Cys4 are acylated by either C15-az or C17-az in LAMTOR-HA. Metabolic labeling with azido fatty acids (conditions as in Fig. 1A) in HeLa cells was performed 24 h after transfection. The purified labeled proteins were eluted for a Western blot and the WT LAMTOR1 was compared to the acylation mutants (detected with anti-HA antibody). The efficiency of GNAI2 and TFR1 labeling serves as a positive control of the assay. A representative of two independent experiments. E, a scheme illustrating the C-terminal sequence of HRAS with its lipid modifications. F, APE assay fails to show acylation sites on the endogenous HRAS in HeLa and HEK293E cells, likely due to a faint antibody signal. A representative image of three biological replicates. G, both Cys181 and Cys184 are acylated in V5-HRAS, detected via APE assay. Transiently transfected HeLa cells were lysed for APE assay 24 h post transfection. WT HRAS was compared to acylation mutants upon detection with anti-V5 antibody (single mutants C181A and C184A and the double mutant C181A/C184A). The endogenous GNAI2 with its single acylation is shown as a positive control of the assay and tubulin as a loading control. A representative of three independent experiments. H, both Cys181 and Cys184 are acylated by both C15-az and C17-az in V5-HRAS. Metabolic labeling with azido fatty acids (conditions as in Fig. 1A) in HeLa cells was performed 24 h after transfection. The WT V5-HRAS was compared to the acylation mutants (detected with anti-V5 antibody). The efficiency of GNAI2 and TFR1 labeling serves as a positive control of the assay. A representative of three independent experiments. APE, acyl-PEG exchange; TFR, transferrin receptor.

Fig 5: Validation of preferential S-acylation of proteins with one fatty acid.A, metabolic labeling with azido fatty acids in HeLa cells was performed under the conditions described in Figure 1A. After the click-based pulldown on alkyne agarose, labeled proteins were eluted and probed on Western blots. The efficiency of GNAI2 labeling serves as a positive control of the assay. B, the intensity of bands in A was quantified, the values were background corrected and normalized to the respective GNAI2 band. The ratio of the efficiency of C17-az and C15-az labeling is shown for each tested protein. The significance was tested by one sample t test (hypothetical mean 1), when significant, p-value is indicated, n = 2 to 5 (weak antibodies could not be detected in some samples), mean ± SD. C, metabolic labeling with azido fatty acids in HeLa cells was performed in the presence of increasing concentrations of the other unlabeled fatty acid. Cells were incubated with 100 µM C15-az and C17-az alone or together with 10, 100, and 1000 µM of C18:0 or C16:0, respectively. D, quantification of the competition experiments as shown in C.