miR-6077 promotes cisplatin/pemetrexed resistance in lung adenocarcinoma via CDKN1A/cell cycle arrest and KEAP1/ferroptosis pathways

Abstract

Lung adenocarcinoma (LUAD) is one of the most common malignancies worldwide. Combination chemotherapy with cisplatin (CDDP) plus pemetrexed (PEM) remains the predominant therapeutic regimen; however, chemoresistance greatly limits its curative potential. Here, through CRISPR-Cas9 screening, we identified miR-6077 as a key driver of CDDP/PEM resistance in LUAD. Functional experiments verified that ectopic overexpression of miR-6077 desensitized LUAD cells to CDDP/PEM in both cell lines and patient-derived xenograft models. Through RNA sequencing in cells and single-cell sequencing of samples from patients with CDDP/PEM treatments, we observed CDDP/PEM-induced upregulation of CDKN1A and KEAP1, which in turn activated cell-cycle arrest and ferroptosis, respectively, thus leading to cell death. Through miRNA pull-down, we identified and validated that miR-6077 targets CDKN1A and KEAP1. Furthermore, we demonstrated that miR-6077 protects LUAD cells from cell death induced by CDDP/PEM via CDKN1A-CDK1-mediated cell-cycle arrest and KEAP1-NRF2-SLC7A11/NQO1-mediated ferroptosis, thus resulting in chemoresistance in multiple LUAD cells both in vitro and in vivo. Moreover, we found that GMDS-AS1 and LINC01128 sensitized LUAD cells to CDDP/PEM by sponging miR-6077. Collectively, these results imply the critical role of miR-6077 in LUAD's sensitivity to CDDP/PEM, thus providing a novel therapeutic strategy for overcoming chemoresistance in clinical practice.

Keywords: CRISPR-Cas9; MT: Non-coding RNAs; RNA-seq; cell-cycle arrest; chemoresistance; cisplatin; ferroptosis; homology-directed repair; lncRNA; miRNA; pemetrexed.

Graphical abstract

Figure 1

Genome-wide CRISPR-Cas9 screening identifies miR-6077 as a determinant of CDDP/PEM sensitivity in LUAD (A) Schematic outline of the CRISPR-Cas9 screening workflow in the A549 cell line. (B) Scatterplot showing the top hits in 7- (horizontal axis) or 14-day (vertical axis) CDDP/PEM-treated A549 cells, highlighting miR-6077. Red dots represent miRNAs whose depletion led to chemoresistance, while blue dots represent sensitivity. For presentation purpose, only miRNAs meeting the criteria of |logFC| > 2.7 and |logFC_{(day 14)}| > |logFC_{(day 7)}| are highlighted. (C) CRISPR-Cas9 results of top five miRNAs potentially conferring resistance to CDDP/PEM. (D) Cell viability in miRNA-transfected A549 cells treated with PBS (left) or CDDP (10 μM)/PEM (1 μM) (right) for 48 h. (E) Dose-toxicity curves showing the viability of A549 cells transfected with miR-NC1, miR-NC2, and miR-6077 upon CDDP or PEM treatment at the indicated concentrations for 48 h. (F) Viability of A549 cells transfected with miR-NC1, miR-NC2, and miR-6077 upon treatment with PEM and oxaliplatin or carboplatin at the indicated concentrations for 48 h. (G) Quantification of the colony formation ability of A549 and H358 cells transfected with miR-NC1, miR-NC2, and miR-6077 upon treatment with PBS or CDDP for 14 days. (H) Relative expression levels of miR-6077 in CDDP/PEM-sensitive or -resistant tumor tissues obtained from LUAD patients. (I) ROC curve of miR-6077 exhibiting its predicting value when assessing LUAD patients’ response to CDDP/PEM therapy. Data are presented as the mean ± SD, n = 3 independent repeats. Unpaired, two-tailed t test; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001; ns, not significant.

Figure 2

CDKN1A and KEAP1 are direct targets of miR-6077 (A) Venn plot showing the intersection of predicted targets of miR-6077. (B) The CRISPR-Cas 9 results of top eight predicted targets of miR-6077 whose depletion potentially confers resistance to CDDP/PEM. (C) Heatmap showing the relative expression levels of the eight predicted targets of miR-6077 in A549 cells transfected with miR-NC1, miR-NC2, and miR-6077 upon treatment with PBS or CDDP (10 μM)/PEM (1 μM) for 48 h. The results were obtained from RNA-seq. (D and E) Quantitative real-time PCR and western blotting assays showing the mRNA (D) and protein (E) levels of CDKN1A and KEAP1 in A549 cells transfected with miR-NC1, miR-NC2, and miR-6077 upon treatment with PBS or CDDP (10 μM)/PEM (1 μM) for 48 h. (F) Quantitative real-time PCR showing the relative expression levels of CDKN1A and KEAP1 in A549 cells upon treatment with PBS or CDDP (10 μM)/PEM (1 μM) for 48 h. (G) The predicted target sites of miR-6077 in the 3′ UTR of CDKN1A and KEAP1. Normal and mutant seed regions are highlighted and underlined. (H) Luciferase reporter plasmids containing wild- or mutant-type CDKN1A (left) and KEAP1 (right) were co-transfected into HEK293T cells with miR-NC1, miR-NC2, or miR-6077. Bar plots exhibit the luciferase activity of the transfected cells. (I) Sanger sequencing confirming the G333C inactivating mutation within the first Kelch domain (KLD) of KEAP1 in A549 cells. Data are presented as the mean ± SD, n = 3 independent repeats. Unpaired, two-tailed t test; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001; ns, not significant.

Figure 3

miR-6077 protects LUAD cells from cell-cycle arrest and ferroptosis induced by CDDP/PEM treatment (A) Volcano plot (left) and bar plot (right) showing the differentially expressed genes in A549 cells treated with PBS or CDDP (10 μM)/PEM (1 μM) for 48 h and corresponding Gene Ontology (GO) functional enrichment analyses, respectively. (B) The uniform manifold approximation and projection (UMAP) plots visualizing the cell-type clusters of the 70,971 high-quality cells based on the expression of known marker genes. (C) The sample origin of the 70,971 cells. (D) Gene set enrichment analysis showing the activation of cell-cycle- and ferroptosis-related signaling pathways when analyzing the differentially expressed genes between the cell populations derived from patients receiving or not receiving neoadjuvant CDDP/PEM treatment. (E) Cell-cycle analyses of A549 cells treated with PBS or CDDP (10 μM)/PEM (1 μM) for 48 h. (F) Cell-cycle analyses of A549 cells transfected with miR-NC1, miR-NC2, and miR-6077 upon treatment with PBS or CDDP (10 μM)/PEM (1 μM) for 48 h. (G and H) Lipid peroxidation (G) and relative glutathione levels (H) in A549 cells treated with PBS or CDDP (10 μM)/PEM (1 μM) for 48 h. (I) Viability of A549 cells treated as indicated for 48 h. (J) Relative expression levels of PTGS2 in A549 cells treated with PBS or CDDP (10 μM)/PEM (1 μM) for 48 h. (K) Transmission electron microscopy images of H1299 cells subjected to indicated treatments (white arrowheads indicate mitochondria). (L and M) Lipid peroxidation (L) and relative glutathione levels (M) in H358 transfected with miR-NC1, miR-NC2, and miR-6077 upon treatment with PBS or CDDP (20 μM)/PEM (2 μM) for 48 h. Data are presented as the mean ± SD, n = 3 independent repeats. Unpaired, two-tailed t test; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001; ns, not significant.

Figure 4

miR-6077 protects LUAD cells from CDDP/PEM-induced cell-cycle arrest by directly targeting CDKN1A (A) Quantitative real-time PCR assay confirming the ectopic overexpression of CDKN1A in A549 cells. (B) Dose-toxicity curves showing the viability of A549 cells (with or without CDKN1A overexpression) transfected with miR-NC1, miR-NC2, and miR-6077 upon CDDP/PEM treatment at the indicated concentrations for 48 h. (C–E) Colony formation ability (C), cell-cycle proportion (D), and protein levels of CDKN1A and its downstream cell-cycle regulators (E) in A549 cells (with or without CDKN1A overexpression) transfected with miR-NC1, miR-NC2, and miR-6077 upon treatment with PBS or CDDP (10 μM)/PEM (1 μM) for 14 days (C) or 48 h (D and E). (F) Dose-toxicity curves showing the viability of A549 cells (with or without CDKN1A overexpression) transfected with miR-NC1, miR-NC2, and miR-6077 upon treatment with PEM and oxaliplatin or carboplatin at the indicated concentrations for 48 h. Data are presented as the mean ± SD, n = 3 independent repeats. Unpaired, two-tailed t test; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001; ns, not significant.

Figure 5

miR-6077 protects LUAD cells from CDDP/PEM-induced ferroptosis by directly targeting KEAP1 (A) Quantitative real-time PCR assay confirming the ectopic overexpression of KEAP1 in H358 cells. (B) Dose-toxicity curves showing the viability of H358 cells with or without KEAP1 overexpression transfected with miR-NC1, miR-NC2, and miR-6077 upon CDDP/PEM treatment at the indicated concentrations for 48 h. (C–F) Colony formation ability (C), lipid peroxidation (D), relative glutathione levels (E), and protein levels of KEAP1 and its downstream ferroptosis regulators (F) in H358 cells (with or without KEAP1 overexpression) transfected with miR-NC1, miR-NC2, and miR-6077 upon treatment with PBS or CDDP (20 μM)/PEM (2 μM) for 14 days (C) or 48 h (D–F). (G) Schematic outline of the CRISPR-Cas9-mediated homology-directed repair (HDR) in A549 cells. The mutated KEAP1 was replaced by a wild-type sequence. Agarose gel electrophoresis confirming the successful HDR in A549 cells, and the introduced repair template containing homology arms was digested by ApaI enzyme as expected. (H and I) Lipid peroxidation (H) and relative glutathione levels (I) in A549-HDR cells transfected with miR-NC1, miR-NC2, and miR-6077 upon treatment with PBS or CDDP (20 μM)/PEM (2 μM) for 48 h. Data are presented as the mean ± SD, n = 3 independent repeats. Unpaired, two-tailed t test; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001; ns, not significant.

Figure 6

GMDS-AS1 and LINC01128 sensitize LUAD to CDDP/PEM and function as sponges for miR-6077 in LUAD (A) Flowchart displaying the screening process of lncRNAs potentially interacting with miR-6077. (B) Bar plot showing that the overexpression of selected lncRNAs leads to altered sensitivity to CDDP (20 μM)/PEM (2 μM) treatment in H1299. (C) Relative expression levels of selected lncRNAs in H1299 and H1299-CDDP/PEM-resistant cells. (D and E) Dose-toxicity curves showing the viability of H358 (D) and H1299 (E) cells (with or without GMDS-AS1/LINC01128 overexpression) transfected with miR-NC1, miR-NC2, and miR-6077 upon CDDP/PEM treatment at the indicated concentrations for 48 h. Blue asterisks represent the significance of comparison between GMDS-AS1-overexpressing H1299 transfected with miR-NC and with miR-6077, while red asterisks represent that of LINC01128-overexpressing H1299. (F) The predicted target sites of miR-6077 in GMDS-AS1 and LINC01128. (G) Luciferase reporter plasmids containing wild- or mutant-type GMDS-AS1 (left) and LINC01128 (right) were co-transfected into HEK293T cells with miR-NC1, miR-NC2, or miR-6077. Bar plots exhibit the luciferase activity of the transfected cells. (H) RNA pull-down followed by western blotting showing a possible interaction between Ago2 and GMDS-AS1 or LINC01128. (I) RIP assay of the enrichment of GMDS-AS1, LINC01128, miR-6077, CDKN1A, and KEAP1 transcripts on Ago2 relative to IgG in H1299 cells. (J) RIP assay showing that overexpression of GMDS-AS1 or LINC01128 disrupts the interactions between miR-6077 and CDKN1A or KEAP1. Data are presented as the mean ± SD, n = 3 independent repeats. (K and L) Quantitative real-time PCR assay confirming the ectopic overexpression of GMDS-AS1 and LINC01128 in H358 (K) and H1299 (L) cells. Unpaired, two-tailed t test; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001; ns, not significant.

Figure 7

Overexpression of GMDS-AS1 and LINC01128 augments the magnitude of CDDP/PEM-caused G2/M arrest and lipid peroxidation by inhibiting miR-6077, and inhibiting miR-6077 results in chemosensitivity to CDDP/PEM in LUAD cells (A–D) Cell-cycle analyses (A), lipid peroxidation (B), glutathione levels (C), and protein levels of CDKN1A/KEAP1 and their downstream molecules (D) in GMDS-AS1/LINC01128-overexpressing H358 cells transfected with miR-NC1, miR-NC2, and miR-6077 upon treatment with PBS or CDDP (20 μM)/PEM (2 μM) for 48 h. (E and F) Relative expression levels of miR-6077 (E) and CDKN1A and KEAP1 (F) in H1299 and H1299-CDDP/PEM-resistant cells. (G) Western blotting assay showing the protein levels of CDKN1A and KEAP1 in H1299-CDDP/PEM-resistant cells transfected with miR-NC1, miR-NC2 inhibitor, and miR-6077 inhibitor upon CDDP (20 μM)/PEM (2 μM) treatment for 48 h. (H) Single-cell sequencing data displaying the sample origin (top) and expression levels of CDKN1A (middle) and KEAP1 (bottom) of tumor cells derived from patients receiving or not receiving neoadjuvant CDDP/PEM treatment. (I–L) Dose-toxicity curves (I), cell-cycle analyses (J), lipid peroxidation levels (K), and relative glutathione levels (L) in H1299-CDDP/PEM-resistant cells transfected with miR-NC1, miR-NC2, and miR-6077 upon PBS or CDDP (20 μM)/PEM (2 μM) treatment for 48 h; ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001; ns, not significant.

Figure 8

Ectopic overexpression of miR-6077 in H1299 cells confers resistance to CDDP/PEM in vivo (A–C) Growth curves (A), image (B), and weights (C) of the xenograft tumors of sacrificed nude mice treated as indicated in each group (n = 8). H1299 cells with or without CDKN1A/KEAP1 overexpression were used to form the xenograft tumor subcutaneously. (D–F) Tumor cells derived from three LUAD patients were subcutaneously injected into NSG mice to form the PDX models, and the animals were treated with CDDP/PEM and miR-agomir as indicated for each group (n = 6). Growth image (D), curves (E), and weights (F) of the PDX tumors are shown. (G) A schematic diagram indicating the mechanism by which miR-6077 confers LUAD chemoresistance by mitigating CDDP/PEM-induced G2/M arrest and ferroptosis by specifically targeting CDKN1A and KEAP1, respectively. CDDP plus PEM is the predominant therapeutic regimen for LUAD, whereas chemoresistance greatly limits its curative potential. We found that miR-6077 promotes chemoresistance by targeting CDKN1A-mediated cell-cycle arrest and KEAP1-mediated ferroptosis. Our discovery provides a novel therapeutic strategy for clinical practice.