MTCH2 Deficiency Promotes E2F4/TFRC-Mediated Ferroptosis and Sensitizes Colorectal Cancer Liver Metastasis to Sorafenib

Abstract

Ferroptosis is a specific type of lipid peroxide-mediated cell death which is crucial in tumor suppression. While the mitochondrial carrier homolog 2 (MTCH2) is implicated in lipid homeostasis and mitochondrial metabolism, its role in ferroptosis and colorectal cancer (CRC) remains uncharacterized. Here, MTCH2 is identified as a crucial regulator of ferroptosis in CRC progression. Clinically, high expression of MTCH2 in CRC tissues predicts poor prognosis. Functionally, loss of MTCH2 inhibits azoxymethane (AOM)/dextran sodium sulfate (DSS)-induced colorectal tumorigenesis in MTCH2^cKO mice and leads to accumulation of ferrous ion and enhances ferroptosis of CRC in vitro and in vivo. Mechanistically, MTCH2 deficiency promotes the proteasome-dependent ubiquitination of E2F4 and attenuates transcriptional inhibition of transferrin receptor (TFRC) by E2F4, ultimately facilitating TFRC-mediated ferroptosis in CRC cells. Moreover, MTCH2 depletion combined with sorafenib treatment synergistically triggers ferroptosis, suppresses liver metastasis, and effectively eradicates tumors in liver metastasis foci. Taken together, This study reveals the mechanism of MTCH2 deficiency-induced ferroptosis to inhibit the progression of CRC and supports a potential therapeutic strategy targeting the MTCH2/E2F4/TFRC signaling axis in CRC patients with liver metastasis.

Keywords: MTCH2; colorectal cancer; ferroptosis; liver metastasis; sorafenib.

Figure 1

Upregulation of MTCH2 in CRC is associated with poor prognosis. A) Representative immunohistochemistry images with low, medium, and high MTCH2 expression scores in CRC and normal tissues (n = 172). Scale bars, 300 µm and 50 µm. B,C) Comparison of MTCH2 IHC expression scores in colon adenocarcinoma (COAD; B) or rectum adenocarcinoma (READ; C) tissues with normal tissues. D,E) Quantification of qRT‐PCR assay of MTCH2 mRNA levels in nine paired CRC samples (D) and Western blot of MTCH2 protein levels in four paired CRC samples (E). F,G) Analysis of MTCH2 expression in CRC tissues from the TCGA (F) and GEO (G) databases. H) Relative mRNA expression levels of MTCH2 in a normal human colon mucosal epithelial cell line (NCM460) and five human CRC cell lines. I) Western blot of MTCH2 expression and quantification of relative protein expression levels (bar graph) in the same cell lines. J,K) Kaplan–Meier curves of overall survival based on MTCH2 IHC staining in COAD (J) and READ (K) samples. Statistics were performed using an unpaired Student's t‐test. Data are presented as the means ± SD of at least three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 2

MTCH2 mediates the proliferation, migration, and invasion of CRC cells. A) Western blot assay showing the efficiency of CRISPR/Cas9‐mediated knockout of MTCH2 by transfection with sgMTCH2#1 and #2 in RKO and SW620 cells. B–D) CCK‐8 (B), clonogenic assays (C), and Transwell assays (D) were performed to assess the cell proliferative and metastatic ability of MTCH2 knockout CRC cells. Bar graphs display quantification of colonies (C) and migratory cells (D), respectively. Scale bar, 100 µm. E–G) CCK‐8 (E), clonogenic assays (F), and Transwell assays (G) were performed to evaluate cell growth and metastatic ability in MTCH2‐overexpressing cells. Bar graphs display quantification of colonies (F) and migratory cells (G), respectively. Scale bar, 100 µm. H,I) Western blot analysis of MTCH2 levels (H) and CCK‐8 viability assay (I) of HCT116 cells transfected with sgMTCH2#1 and Flag‐MTCH2 rescue. J) Images of resected mouse tumors after subcutaneous injection with RKO cells stably transfected with sgMTCH2#1 (n = 5 per group). Red circles indicate no tumor formation was observed. K) Tumor growth curves of mice injected with wild‐type or MTCH2 knockout RKO cells. L) Mean weight (±SD) of the resected tumors in the two groups. M) Representative histological images of xenografted tumors with hematoxylin and eosin (H&E) and IHC staining for MTCH2 and Ki67. Scale bar, 100 µm. Two‐tailed Student's t‐test and one‐way ANOVA with Tukey's multiple comparisons test were used for statistical analysis. The results shown indicate the means ± SD of at least three independent replicates. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant.

Figure 3

MTCH2 depletion induces ferroptosis in CRC cell lines. A) Gene set enrichment analysis (GSEA) of ferroptosis signaling based on differentially expressed genes related to MTCH2 expression in CRC data from The Cancer Genome Atlas. B) CCK‐8 viability assays in sgRNA control and sgMTCH2 RKO and SW620 cells treated with Fer‐1 (1 µm), Nec‐1 (2 µm), or Z‐VAD‐FMK (20 µm). C–E) Levels of Fe²⁺ (C), ROS (D), and MDA (E) in sgRNA control and sgMTCH2 RKO and SW620 cells treated with Fer‐1 (1 µm). F,G) CCK‐8 viability assays of MTCH2‐knockout (F) or MTCH2‐overexpressing (G) CRC cells treated with erastin. H–J) Levels of Fe²⁺ (H), ROS (I), and MDA (J) in sgRNA control or sgMTCH2 HCT116 cells treated with erastin. K–M) Levels of Fe²⁺ (K), ROS (L), and MDA (M) in MTCH2‐overexpressing HCT116 cells treated with erastin. N) Confocal microscopy images of HCT116 and SW620 cells showing the effect of sgMTCH2#1 transfection with or without erastin treatment (20 µm) on cellular Fe²⁺ levels detected using FerroOrange. Bar graphs represent the quantification of mean fluorescent intensity. Scale bar, 20 µm. O) Transmission electron microscopy images of mitochondrial morphology in MTCH2 knockout HCT116 cells with the treatment of erastin. Scale bars, 1 µm and 500 nm. Two‐way ANOVA with Tukey's multiple comparisons test was used for statistical analysis. The results shown indicate the means ± SD of at least three independent replicates. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant. “sgMT” refers to “sgMTCH2”.

Figure 4

Knockout of MTCH2 promotes ferroptosis through TFRC. A) Heat map of normalized mRNA levels of 13 ferroptosis‐associated genes detected in control or sgMTCH2 HCT116 cells using qRT‐PCR. B) A volcano plot illustrating the fold changes and P values of the 13 ferroptosis‐associated genes. C) Levels of TFRC, SLC7A11, GPX4, and MTCH2 were detected and quantified (in the bar graph) in HCT116 cells with loss of MTCH2. D,E) qRT‐PCR (D) and Western blot (E) analyses of TFRC expression in HCT116 and SW620 cells transfected with indicated concentrations of Flag‐MTCH2 plasmid or control plasmid. F) Representative immunofluorescence images of TFRC in HCT116 and SW620 cells with MTCH2 knockout. Scale bar, 20 µm. G–K) The levels of TFRC (G) and cell viability (H), as well as levels of Fe²⁺ (I), ROS (J), and MDA (K), were measured in MTCH2 knockout CRC cell lines transfected with siTFRC. L) Representative IHC images of MTCH2 and TFRC staining of colorectal tumor samples from CRC patients. The graph depicts Pearson correlation analysis based on IHC scores (n = 30). Scale bars, 500 and 100 µm. The statistical analysis was calculated by two‐way ANOVA for multiple comparisons, one‐way ANOVA with Tukey's honest difference post hoc test, or Pearson's correlation test. Data present means ± SD from three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant.

Figure 5

MTCH2 decreases TFRC levels by repressing E2F4 degradation via inhibition of its ubiquitination. A) Venn diagram of the intersection of transcription factors of TFRC predicted to interact with MTCH2 at online databases (Genecard, PROMO, and NCBI). B,C) qRT‐PCR (B) and Western blot (C) analysis of TFRC mRNA and protein expression in SW480 and SW620 cells transfected with the Myc‐vector or Myc‐E2F4 plasmid. D–G) The mRNA (D,E) and protein levels (F,G) of E2F4 and MTCH2 in MTCH2‐knockout HCT116 cells and MTCH2‐overexpressing SW480 cells. H) Western blot analysis showing rescue of TFRC levels in MTCH2‐overexpressing CRC cells treated with or without E2F4 inhibitor (E2F4i) (40 µm). I) Co‐immunoprecipitation (IP) assay for the endogenous interaction between MTCH2 and E2F4 in HCT116 cells. J) Western blot of Co‐IP assays of exogenous MTCH2 and E2F4 using an anti‐Myc antibody to pull down Myc‐E2F4 (left) or anti‐Flag antibody to pull down Flag‐MTCH2 (right) from lysates of HCT116 cells transfected as indicated. Blots were probed using the indicated antibodies. K) Representative pictures of co‐localization with MTCH2 (green) and E2F4 (red) fluorescent proteins were obtained by confocal microscopy. The fluorescence intensity graph (bottom) shows the level of co‐localization, and each line indicates the corresponding fluorescence. Scale bars, 10 and 5 µm. L) Cycloheximide (Chx) chase assays and quantification of E2F4 protein levels in MTCH2‐overexpressing HCT116 (upper panels) and sgMTCH2 SW620 cells (lower panels) for the indicated times. M) Western blot of E2F4 in lysates of sgMTCH2 HCT116 and SW620 cells treated with MG132 (10 µm). N,O) MTCH2 overexpression inhibited endogenous (M) and exogenous (N) ubiquitination of E2F4. An endogenous ubiquitination assay was performed in HCT116 cells. Exogenous ubiquitination assay was conducted following IP assay with an anti‐Myc antibody in HEK293T cells transfected with Myc‐E2F4, HA‐Ub, Flag‐vector, and Flag‐MTCH2 plasmids for 24 h and subsequently treated with or without MG132 (10 µm) for 8 h. The blot was probed with the indicated antibodies. P) The K48‐ and K63‐linked ubiquitination patterns on E2F4 were examined using HCT116 cells transfected with indicated constructs, including HA‐Ub or its mutants containing only one lysine at either K48 or K63. Two‐tailed Student's t‐test and one‐way ANOVA with Tukey's honest difference post hoc tests were used for statistical analysis. The results shown indicate the means ± SD of at least three independent replicates. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant.

Figure 6

MTCH2 enhances E2F4‐modulated repression of TFRC transcription. A) Distributions of TFRC and E2F4 in cytoplasmic and nuclear fractions of MTCH2‐overexpressing HCT116 cells (left panels) and sgMTCH2 SW620 cells (right panels). β‐actin and histone H3 were used as cytoplasmic and nuclear markers, respectively. B) Two predicted binding sites for E2F4 in the TFRC promoter. C) Chromatin immunoprecipitation assay of the E2F4 binding sites in the promoter region of TFRC in HCT116 cells. D) Schematic representation of wild‐type (PGL3‐WT) and mutated (PGL3‐Mut1, PGL3‐Mut2) TFRC promoter luciferase reporter constructs. E,F) Luciferase reporter assays showing the influence of Myc‐E2F4 expression (E) and MTCH2 overexpression (F) on wild‐type (TFRCwt) and mutated (TFRCmut1 and TFRCmut2) promoter expression in HCT116 cells. The relative luciferase activity was determined by calculating the ratio of firefly luciferase activity to Renilla luciferase activity. G–K) Cell viability (G), EdU assays (H), the levels of Fe²⁺ (I), ROS (J), and MDA (K) of HCT116 cells treated with an E2F4i (40 µm) and Fer‐1 (1 µm). Scale bar, 100 µm. L–P) Cell viability (L), EdU assays (M), the levels of Fe²⁺ (N), ROS (O), and MDA (P) of E2F4‐overexpressing HCT116 cells treated with or without erastin (20 µm). Scale bar, 100 µm. Q) Western blot analysis showing the rescue of TFRC levels in sgMTCH2 CRC cells transfected with Myc‐E2F4 plasmid. R–U) Cell viability (R), the levels of Fe²⁺ (S), ROS (T), and MDA (U) of sgMTCH2 HCT116 cells transfected with Myc‐E2F4 plasmid. V) CCK‐8 viability assay of MTCH2‐overexpressing HCT116 and SW480 cells treated with E2F4i (40 µm). Two‐way ANOVA was used for statistical analysis. The results shown indicate the means ± SD of at least three independent replicates. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant.

Figure 7

Conditional knockout of MTCH2 inhibits AOM/DSS‐induced CRC in mice. A) Schematic diagram of the construction of intestine‐specific MTCH2 knockout (MTCH2^cKO) mice. B) Workflow of azoxymethane (AOM)/dextran sodium sulfate (DSS) induction of colitis‐associated CRC model mice. C) Representative images of colorectal tumors (red arrowheads) in control MTCH2^fl/fl mice and MTCH2^cKO mice. Scale bar, 1 cm. D–F) Tumor number (D) and tumor burden (E) per colon, and colon length (F) in the two groups (n = 5 per group). G) H&E and IHC staining of Ki67 in colorectal tissues from MTCH2^fl/fl and MTCH2^cKO mice. Quantification of Ki67 expression is shown in the bar graph. Scale bars, 2 mm and 200 µm. H–K) Levels of Fe²⁺ (H), ROS (I), MDA (J), and 4‐HNE (K) in the two groups. Scale bar, 200 µm. L,M) Representative image of MTCH2^fl/fl and MTCH2^cKO mouse tumor organoids exposed to Fer‐1 (5 µm) (L) or Erastin (4 µm) (M). Scale bar, 100 µm. N) Western blot of TFRC, E2F4, and MTCH2 in colorectal tumors of the two groups of mice (n = 3 per group). O) Representative immunofluorescence images of MTCH2, E2F4, and TFRC in colorectal tumors of the two groups. Green indicates MTCH2, white indicates E2F4, and red indicates TFRC. Scale bars, 40 µm (left column), 20 µm (right column). P) Fresh tumor samples obtained from groups of MTCH2^fl/fl and MTCH2^cKO mice, respectively, were pooled and subjected to single‐cell RNA sequencing. T‐distributed stochastic neighbor embedding (t‐SNE) visualization of unsupervised segregation of infiltrating cells into 21 clusters. Q) Dot plot showing the expression of cluster marker genes. The color scale represents the average gene expression level. Circle size represents the percentage of cells expressing the gene in each cluster. R,S) Proportions of the six cell clusters in MTCH2^fl/fl and MTCH2^cKO mice presented as tSNE plots (R) and a bar chart (S). T) Volcano plot of differentially expressed genes (DEGs) in CRC tumor cells between MTCH2^fl/fl and MTCH2^cKO mice. U,V) Gene ontology (U) and pathway enrichment analyses (V) for the DEGs from tumor cells of mice. Two‐tailed Student's t‐test was used for statistical analysis. The results shown indicate the means ± SD of at least three independent replicates. *P < 0.05, **P < 0.01, ***P < 0.001.

Figure 8

Loss of MTCH2 improves the response to sorafenib on the metastasis in vivo. A–C) Cell viability (A) and levels of ROS (B) and MDA (C) were measured in HCT116 and SW620 cells incubated with sorafenib (10 µm) and Fer‐1 (1 µm). D,E) CCK‐8 assay assessment of the half‐maximal inhibitory concentration (IC₅₀) of sorafenib in MTCH2‐knockout (D) and MTCH2‐overexpressing (E) CRC cells. F,G) EdU (F) and Transwell (G) assays were performed in sgMTCH2 HCT116 cells incubated with sorafenib (10 µm). Scale bar, 100 µm. H,I) The proliferative (H) and migration (I) capacities were evaluated in MTCH2‐overexpressing HCT116 cells incubated with sorafenib (10 µm). Scale bar, 100 µm. J) The body weights of the four groups of mice with indicated treatment were recorded regularly (n = 5 in each group). K,L) Representative bioluminescence imaging of mice (K) with indicated treatment and quantification of the radiance intensity (L) at 4 weeks post‐implantation with HCT116 cells. M) Representative photographs of metastatic foci of mouse livers and images of H&E‐stained sections of metastatic lesions from each group. Arrows indicate liver metastasis foci. Dashed lines encircle metastatic foci on the liver surface. N,O) Quantification of metastatic liver nodules (N) and the ratio of liver weight to body weight (O). P) Representative images of H&E‐stained liver metastatic foci and IHC staining of MTCH2, 4‐HNE, Ki67, E‐cadherin (E‐cad), TUNEL, and CD31 in liver sections from each group. Scale bars, 500 and 100 µm. The statistical analysis was calculated by two‐way ANOVA for multiple comparisons and one‐way ANOVA with Tukey's honest difference post hoc test. Data present means ± SD from three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant.

Figure 9

Schematic model of the potential mechanisms by which MTCH2 deficiency enhances ferroptosis and facilitates the effect of sorafenib on liver metastasis of CRC via the E2F4/TFRC axis. This graphical abstract was drawn using the Figdraw tool (www.figdraw.com).