ACTL6A protects gastric cancer cells against ferroptosis through induction of glutathione synthesis

Abstract

Gastric cancer (GC), one of the most common malignant tumors in the world, exhibits a rapid metastasis rate and causes high mortality. Diagnostic markers and potential therapeutic targets for GCs are urgently needed. Here we show that Actin-like protein 6 A (ACTL6A), encoding an SWI/SNF subunit, is highly expressed in GCs. ACTL6A is found to be critical for regulating the glutathione (GSH) metabolism pathway because it upregulates γ-glutamyl-cysteine ligase catalytic subunit (GCLC) expression, thereby reducing reactive oxygen species (ROS) levels and inhibiting ferroptosis, a regulated form of cell death driven by the accumulation of lipid-based ROS. Mechanistic studies show that ACTL6A upregulates GCLC as a cotranscription factor with Nuclear factor (erythroid-derived 2)-like 2 (NRF2) and that the hydrophobic region of ACTL6A plays an important role. Our data highlight the oncogenic role of ACTL6A in GCs and indicate that inhibition of ACTL6A or GCLC could be a potential treatment strategy for GCs.

Fig. 1. ACTL6A is overexpressed in GC and promotes GC progress.

a–c ACTL6A expression levels in GC and normal tissue from three GEO datasets, GSE13911(a), GSE27342 (b), and GSE13861 (c). The data are presented as the means ± SD (standard deviation). d Waterfall plot of the relative ACTL6A mRNA levels from 21 paired samples of GC and normal tissue measured using qRT-PCR. The log2 T/N ACTL6A mRNA levels are presented. Each bar chart represents one case, presented as the means ± SD. e Relative cell growth rate of SNU638 cells treated with ACTL6A shRNA and scrambled shRNA. The data are presented as the means ± SEM, n = 3 biologically independent experiments. Western blot analysis of ACTL6A protein level is shown. f Relative cell growth rate of SNU638 cells treated with flag-ACTL6A and vector. The data are presented as the means ± SEM, n = 3 biologically independent experiments. Western blot analysis of ACTL6A protein level is shown. g Wound-healing assay of SNU638 cells treated with ACTL6A shRNA or scrambled shRNA. Scale bar, 200 μm. The data are presented as the means ± SD, n = 3 biologically independent experiments. h Transwell migration assay of SNU638 and SNU668 cells treated with ACTL6A shRNA or scrambled shRNA. The scale, 100 μm. The data are presented as the means ± SD, n = 3 biologically independent experiments. i Patient-derived organoids (PDOs) were treated with ACTL6A shRNA or scrambled for 9 days. Representative pictures at day 0, day 3, and day 9. Scale bar, 25 μm. The size of organoids on the 9th day were calculated, presented as the means ± SD, n = 8 for each group. j–l Xenograft experiment. SNU638 cells (1 × 10⁶) treated with ACTL6A shRNA or scrambled were subcutaneously injected into nude mice (n = 5 for each group). Images (j), tumor volumes (k) and tumor weight (l) are shown. The data are presented as the means ± SD. P values were determined by unpaired two-tailed T test for panels a–c, g–i, l and two-way ANOVA followed by Tukey test for panels e, f, k.

Fig. 2. ACTL6A reprograms GSH metabolism to maintain GC malignant progression.

a Gene categories significantly (P ≤ 0.05) enriched for genes deregulated in KEGG pathways, owing to knockdown of ACTL6A in SNU638 cells. b Venn diagram illustrating the overlap among the top 35 enriched pathways of RNA array and two GC databases. c An enrichment analysis of gene sets available from RNA array and two GC databases revealed that ACTL6A expression is positively correlated with Glutathione Metabolism. d Schematic of GSH/GSSG cycle that affects oxidative stress and NADPH/NADP+ ratio. e Measurement of relative GSH/GSSG ratio in SNU638 cells treated with ACTL6A shRNA and scrambled shRNA. The data are presented as the means ± SD, n = 3 biologically independent experiments. f Measurement of relative NADP + /NADPH ratio in SNU638 cells treated with ACTL6A shRNA and scrambled shRNA. The data are presented as the means ± SD, n = 3 biologically independent experiments. g Relative DCFH-DA fluorescence intensity measured by flow cytometry, SNU638 cells are treated with ACTL6A shRNA and scrambled shRNA and cultured with or without 50 μM H₂O₂ for 24 h. Data are presented as the means ± SD, n = 3 biologically independent experiments. h, i Relative cell growth rate of SNU638 cells treated with ACTL6A shRNA or scrambled shRNA and cultured with or without 100 μM NAC. The data are presented as the means ± SEM, n = 3 biologically independent experiments. j PDOs were treated with ACTL6A shRNA or scrambled shRNA and cultured with or without 100 μM NAC for 9 days. Representative pictures at day 0, day 3, and day 9 after being treated with NAC were shown. Scale bars represent 25 μm. The size of organoids on the 9th day was calculated, presented as the means ± SD, n = 8 for each group. k–n SNU638 cells (1 × 10⁶) treated with ACTL6A shRNA or scrambled shRNA were transplanted to nude mice (n = 6 for each group). Mice were treated with or without NAC (k). Images (l), tumor volumes (m) and tumor weight (n) are shown. The data are presented as the means ± SD. P values were determined by unpaired two-tailed T test for panels e–g, j, n, and two-way ANOVA followed by Tukey test for panels h, i, m.

Fig. 3. ACTL6A inhibits ferroptotic cell death.

a Viability of SNU638 cells treated with ACTL6A shRNA or scrambled shRNA after 24 h cultured with or without 1 μM Fer-1, 2 μM necrostatin-1, 5 μM ZVAD-FMK. The data are presented as the means ± SD, n = 3 biologically independent experiments. b–f Viability of SNU638 cells treated with ACTL6A shRNA or scrambled shRNA after 24 h cultured with or without varying concentrations of BSO (b), erastin (c), H₂O₂ (d), doxorubicin (e), and etoposide (f). The data are presented as the means ± SD, n = 3 biologically independent experiments. g Relative C11-BODIPY fluorescence measured by flow cytometry of SNU638 cells treated with ACTL6A shRNA or scrambled shRNA and cultured with either erastin (10 μM), Fer-1 (1 μM) or both for 24 h. The percentages of lipid peroxidation are presented as the means ± SD, n = 3 biologically independent experiments. h PDOs were treated with ACTL6A shRNA or scrambled shRNA and cultured with or without 5 μM Fer-1 for 9 days. Representative pictures at day 0, day 3, and day 9 after being treated with Fer-1 were shown. Scale bars represent 25μm. The size of organoids on the 9th day were calculated. Data are presented as the means ± SD, n = 8 for each group. i Immunohistochemical (IHC) staining for ACTL6A, Ki-67 and 4-HNE in xenograft tissues derived from ACTL6A-knockdown SNU638 cells in mice treated with or without NAC. Scale bars represent 50 μm. P values were determined by unpaired two-tailed T test for panels a, g, h.

Fig. 4. ACTL6A impacts GSH de novo synthesis mainly by upregulating γ-glutamyl-cysteine synthesis.

a Schematic metabolic map of [U-¹³C] glucose-labeled GSH de novo synthesis. Glu, glutamate; Cys, cysteine; Ser, serine; Gly, glycine. γ-GC, γ-glutamyl-cysteine; GSH, glutathione. b–d Incorporation of carbon atoms from [U-¹³C] glucose into glutamate (b), serine (c), and glycine (d) in scrambled and ACTL6A-KD SNU638 cells. Data are presented as the means ± SD, n = 3 biologically independent experiments. e Fractional contribution of carbon atoms from [U-¹³C] glucose into the γ-GC (m + 0 to m + 10) isotopomers in scrambled and ACTL6A-KD SNU638 cells. Data are presented as the means ± SD, n = 3 biologically independent experiments. f Fractional contribution of carbon atoms from [U-¹³C glucose into the GSH (m + 0 to m + 10) isotopomers in scrambled and ACTL6A-KD SNU638 cells. Data are presented as the means ± SD, n = 3 biologically independent experiments. g Schematic metabolic map of [U-¹³C glutamine]-labeled GSH de novo synthesis. Gln glutamine, Glu glutamate, Cys cysteine; Gly, glycine, γ-GC γ-glutamyl-cysteine, GSH glutathione. h Incorporation of carbon atoms from [U-¹³C] glutamine into glutamate in scrambled and ACTL6A-KD SNU638 cells. Data are presented as the means ± SD, n = 3 biologically independent experiments. i Fractional contribution of carbon atoms from [U-¹³C] glutamine into the γ-GC (m + 0 to m + 10) isotopomers in scrambled and ACTL6A-KD SNU638 cells. Data are presented as the means ± SD, n = 3 biologically independent experiments. j Fractional contribution of carbon atoms from [U-¹³C] glutamine into the GSH (m + 0 to m + 10) isotopomers in scrambled and ACTL6A-KD SNU638 cells. Data are presented as the means ± SD, n = 3 biologically independent experiments. P values were determined by unpaired two-tailed T test for panels b–f, h–j.

Fig. 5. ACTL6A inhibits GC cell ferroptosis via regulating GCLC.

a Immunoblotting analysis of ACTL6A and GCLC protein levels in xenograft tissues derived from ACTL6A-knockdown SNU638 cells in mice. b Relative cell growth rate of SNU638 cells treated with ACTL6A shRNA or scrambled shRNA, and transfected with flag-GCLC or pcDNA3.1 vector. The Data are presented as the means ± SEM, n = 3 biologically independent experiments. Immunoblotting analysis of the indicated proteins is shown. c Relative DCFH-DA fluorescence measured by flow cytometry of cells treated with ACTL6A shRNA or scrambled shRNA, and cultured with either H₂O₂ (50 μM) or both of H₂O₂ (50 μM) and NAC (100 μM) for 24 h. The data are presented as the means ± SD, n = 3 biologically independent experiments. d Relative C11-BODIPY fluorescence measured by flow cytometry of cells treated with ACTL6A shRNA or scrambled shRNA, and cultured with either erastin (10 μM), or both of erastin (10 μM) and Fer-1 (1 μM) for 24 h. The data are presented as the means ± SD, n = 3 biologically independent experiments. P values were determined by two-way ANOVA followed by Tukey test for panels b, and unpaired two-tailed T test for panels c, d.

Fig. 6. ACTL6A transcriptionally regulates GCLC dependent on NRF2.

a Trace from ACTL6A ChIP-seq in SNU638 cells showing a binding peak upstream of the transcriptional start of GCLC. IgG was used as a control. b Predicting binding site of NRF2 in the peak from ChIP-seq result. NRF2-binding motif is from the website: http://jaspar.genereg.net/. c mRNA expression levels of GCLC in SNU638 cells treated with ACTL6A shRNA or scrambled shRNA, together with or without shNRF2. Data are presented as the means ± SD, n = 3 biologically independent experiments. d, e ChIP assay was performed in SNU638 cells using anti-ACTL6A (d) or anti-NRF2 (e) antibodies, followed by RT-qPCR with primers recognizing the predicting binding site of NRF2 in the transcriptional start of GCLC. The fold expression of ChIP-enriched mRNAs relative to the input was calculated, presented as the means ± SD, n = 3 biologically independent experiments. IgG was used as a control. f ChIP assay was performed in SNU638 cells treated with NRF2 shRNA or scrambled shRNA using anti-ACTL6A or anti-IgG antibodies, followed by RT-qPCR with primers recognizing the predicting binding site of NRF2 in the transcriptional start of GCLC. The fold expression of ChIP-enriched mRNAs relative to the input was calculated. The data are presented as the means ± SD, n = 3 biologically independent experiments. g ChIP assay was performed in SNU638 cells treated with ACTL6A shRNA or scrambled shRNA using anti-NRF2 or anti-IgG antibodies, followed by RT-qPCR with primers recognizing the predicting binding site of NRF2 in the transcriptional start of GCLC. The fold expression of ChIP-enriched mRNAs relative to the input was calculated. The data are presented as the means ± SD, n = 3 biologically independent experiments. h Immunoblot analysis of the ACTL6A and NRF2 from anti-NRF2 immunoprecipitates (IP) obtained from SNU638 cells. IgG serves as a control. i. Immunoblot analysis of the NRF2 and ACTL6A from anti-ACTL6A immunoprecipitates (IP) obtained from SNU638 cells. IgG serves as a control. P values were determined by unpaired two-tailed T test for panels c–g.

Fig. 7. The HR domain of ACTL6A is essential in GCLC regulating and ferroptosis.

a A schematic drawing of ACTL6A domain deletion constructs. b Relative cell growth rate of SNU638 cells treated with ACTL6A shRNA or scrambled shRNA, and transfected with flag-ACTL6A-WT, constructs of domain deletion from (a) or pcDNA3.1 vector. The data are presented as the means ± SEM, n = 3 biologically independent experiments. c Immunoblotting analysis of the indicated proteins in cells from b. d mRNA levels of ACTL6A and GCLC in SNU638 cells treated with ACTL6A shRNA or scrambled shRNA, and transfected with ACTL6A-WT, increasing doses of ACTL6A-ΔHR or pcDNA3.1 vector. Data are presented as the means ± SD, n = 3 biologically independent experiments. e Immunoblot analysis of the indicated proteins from M2 beads immunoprecipitates (IP) and whole cell lysates (input) obtained from SNU638 cells transfected with ACTL6A-WT, ACTL6A-ΔHR, or pcDNA3.1 vector. f ChIP assay was performed in SNU638 cells treated with ACTL6A shRNA or scrambled shRNA, and transfected with ACTL6A-WT, ACTL6A-ΔHR, or pcDNA3.1 vector using anti-NRF2 or anti-IgG antibodies, followed by RT-qPCR with primers recognizing the predicting binding site of NRF2 in the transcriptional start of GCLC. The fold expression of ChIP-enriched mRNAs relative to the input was calculated, presented as the means ± SD, n = 3 biologically independent experiments. g Relative DCFH-DA fluorescence measured by flow cytometry of cells treated with scrambled or ACTL6A shRNA, and transfected with ACTL6A-WT or ACTL6A-ΔHR, and cultured with either H₂O₂ (50 μM) or both of H₂O₂ (50 μM) and NAC (100 μM) for 24 h. The Data are presented as the means ± SD, n = 3 biologically independent experiments. h Relative C11-BODIPY fluorescence measured by flow cytometry of cells treated with scrambled or ACTL6A shRNA, and transfected with ACTL6A-WT or ACTL6A-ΔHR, and cultured with either erastin (10 μM), or both of erastin (10 μM) and Fer-1 (1 μM) for 24 h. The data are presented as the means ± SD, n = 3 biologically independent experiments. P values were determined by unpaired two-tailed T test for panels d, f–h and two-way ANOVA followed by Tukey test for panels b.

Fig. 8. Clinical relevance of the ACTL6A-GCLC-GSH metabolism axis in ferroptosis of GC cells.

a Immunoblotting analysis of ACTL6A and GCLC protein levels in PDX tumors of 4 cases. b. The mice were treated with BSO (750 mg/kg/day) for two weeks. c–f Tumor volumes of PDX tumors, including case1 (c), case2 (d), case3 (e), and case4 (f). The data are presented as the means ± SD, n = 6 for each group in case 1 and 3, n = 4 for each group in case 2 and 4. g–j Representative fluorescent images of PDX frozen sections stained with DCFH-DA (g) and C11-BODIPY (h). Scale bars represent 200 μm. Relative ROS levels (i) and lipid peroxidation levels (j) were counted by image J and presented as bar graphs. The data are presented as the means ± SD, n = 5 for each group. BODIPY GFP represents lipid peroxidation; BODIPY tomato represents the original color of BODIPY. k Schematic metabolic map of [U-¹⁵N]-labeled GSH de novo synthesis. Gln glutamine, Glu glutamate, Cys cysteine, Gly glycine, γ-GC γ-glutamyl-cysteine, GSH glutathione. l–n. Relative intracellular pool levels of glutamine, glutamate, and GSH in case1 (ACTL6A high) PDX tumors (l). Incorporation of nitrogen atoms from [U-¹⁵N] glutamine into glutamine (gln), glutamate (glu), and GSH (m). Relative GSH/Gln ratios (n). Data are presented as the means ± SD, n = 3 for each group. o Kaplan–Meier curves of the overall survival of GC patients with different expression levels of ACTL6A and GCLC, and log-rank analysis was used to test for significance. p Representative images showing the correlation of ACTL6A and GCLC staining in human GC tissue microarray samples. Scale bars represent 50 μm. P values were determined by two-way ANOVA followed by Tukey test for panels c–f, unpaired two-tailed T test for panels i, j, l–n, h, and the log-rank test for panel o.

Fig. 9. Diagram of ACTL6A-mediated GSH synthesis and inhibits ferroptosis.

ACTL6A co-works with NRF2 on regulating GCLC in transcriptional level, and then promotes GSH synthesis and inhibits ferroptosis of GC cells.