Fig 2: NRF2 and CD44 showed a vice versa expression pattern in melanoma patients. (A) Histologic stain of representative, consecutive melanoma tissue sections for NRF2 and CD44. Positive staining shown in red (AEC), nuclei in blue (hematoxylin) color. Scale bars: 20× = 200 µm, 60× = 50 µm; (B) AEC and hematoxylin signals of the stained tissue array slides were analyzed with the software ImageJ and are shown as NRF2 and CD44 staining intensity normalized to the number of nuclei. Grey dots represent tissues of primary melanomas, red dots represent tissues of metastatic melanomas. The graph was quartered at the halves of the measured NRF2 and CD44 intensities to gain quarters 1–4 (Q1–4). Only tissues containing mostly tumor cells were used for the analysis. N = 49 (C) Correlation of NRF2 target gene ME1 with marker genes, expression levels derived from the TCGA human skin cutaneous melanoma dataset (linear correlation, confidence interval 95%, non‐zero test for slope)

Fig 3: Melanoma cell lines were modified by siRNA to activate or inactivate the NRF2/KEAP1 regulation system. (A) Immunofluorescence images of WM793b cells with siRNA treatments, nucleus staining by DAPI, ROS stimulation with 5 µg/ml of Pyocyanin for 6 h, Scale bar: 10 µm. (B) Quantification of color intensity of NRF2 secondary antibody fluorescence in nuclear area, indicated by DAPI staining (unpaired t‐test; ***p‐value < 0.001; ****p‐value < 0.0001). (C) Western blot of WM793b protein extract 48 h after siRNA transfection, NRF2 and KEAP1 as siRNA targets, ME1 and GSR as targets of NRF2 transcription factor, β‐Actin as loading control. (D) Quantification of Western blot shown in C (unpaired t‐test; **p‐value < 0.01; ***p‐value < 0.001; ****p‐value < 0.0001). (E) qPCR of WM793b cDNA 24 h after siRNA transfection, NRF2 and KEAP1 as siRNA targets, ME1, GSR, TXNRD1 and G6PD as targets of NRF2 transcription factor (unpaired t‐test; *p‐value < 0.05; **p‐value < 0.01; ***p‐value < 0.001; ****p‐value < 0.0001). (F) ROS quantification by DCFDA fluorescence assay over 2 h with 10 mM Hydrogen peroxide (H2O2), in WM793b 48 h after siRNA transfection

Fig 4: Exploiting alterations in ROS homeostasis and the glutathione/thioredoxin axis at cystine import. A, Schematic diagram: xCT: glutamine-cystine antiporter (SLC3A2 and SLC7A11); ME1: malic enzyme 1; GSH: representing glutathione antioxidant system; thioredoxin: representing thioredoxin antioxidant system. B,In vitro cell death in response to ACXT-3102 (Sigma-2/erastin) inhibition of xCT in ME1-null SS cell lines (SYO-1 WT and Fuji WT) and ME1-overexpressing cell lines (SYO-1 OE and Fuji OE). Data, mean ± SD (n = 3); ****, P < 0.0001. C, Tumor volume of SYO-1 WT ME1-null xenografts treated until day 12 with ACXT-3102 (n = 10) (ACXT-3102) or vehicle (n = 10). Data, mean ± SEM; ****, P < 0.0001. D, Tumor volume of SYO-1 ME1-OE xenografts treated until day 12 with ACXT-3102 (n = 10) (ACXT-3102) or vehicle (n = 10). Data, mean ± SEM; NS, P > 0.05.

Fig 5: ME1 absence in SS results in altered redox homeostasis and iron redox status. A, Metabolite levels (pmol/106 cells) of reduced glutathione (GSH; top) and oxidized glutathione (GSSG; bottom) in ME1-null SS cell lines (SYO-1 WT and Fuji WT) and ME1-overexpressing cell lines (SYO-1 OE and Fuji OE). Data, mean ± SD (n = 6); ***, P < 0.001. B, Average normalized isotopomer abundance from stable isotope tracing of U13C-glutamine in reduced glutathione (GSH; top) and oxidized glutathione (GSSG; bottom) over time in ME1-null SS cell line SYO-1 WT and ME1-overexpressing cell line SYO-1 OE. Isotopomer abundance was normalized to cell count. C, Average normalized isotopomer abundance from stable isotope tracing of U13C-glutamine in reduced glutathione (GSH; top) and oxidized glutathione (GSSG; bottom) over time in ME1-null SS cell line Fuji WT and ME1-overexpressing cell line Fuji OE. Isotopomer abundance was normalized to cell count. D,In vitro cell death response to BSO inhibition of gamma-glutamyl l in ME1-null SS cell lines (SYO-1 WT and Fuji WT), ME1 overexpression cell lines (SYO-1 OE and Fuji OE), ME1-expressing cell lines (HTB-93 WT), and ME1 KD cell lines (HTB-93 KD1). Data, mean ± SD (n = 3); ****, P < 0.0001; ***, P < 0.001; **, P < 0.01. E, Relative fluorescence measurements (RFU) of global thiol levels to treatment with BSO relative to no treatment (NT) in ME1-null SS cell line SYO-1 WT and ME1-overexpressing cell line SYO-1 OE. Data, mean ± SD (n = 3); **, P < 0.01; ****, P < 0.0001. F, Relative fluorescence measurements (RFU) of global thiol levels to treatment with BSO relative to NT in ME1-null SS cell line Fuji WT and ME1-overexpressing cell line Fuji OE. Data, mean ± SD (n = 3); ***, P < 0.001; ****, P < 0.0001. G,In vitro cell death response to D9 inhibition of thioredoxin reductase in ME1-null SS cell lines (SYO-1 WT, Fuji WT), ME1 overexpression cell lines (SYO-1 OE and Fuji OE), ME1-expressing cell line (HTB-93 WT), and ME1 KD cell line (HTB-93 KD1). Data, mean ± SD (n = 3); ****, P < 0.0001; **, P < 0.01. H, Relative fold change of H2O2 in ME1-null SS cell lines (SYO-1 WT and Fuji WT) and ME1 overexpression cell lines (SYO-1 OE and Fuji OE). Data, mean ± SD (n = 3); **, P < 0.01. I, Relative fold change of Fe2+ levels in ME1-null SS cell lines (SYO-1 WT and Fuji WT) and ME1 overexpression cell lines (SYO-1 OE and Fuji OE) at baseline. Data, mean ± SD (n = 3); **, P < 0.01; ***, P < 0.001.