Targeting LINC00152 activates cAMP/Ca2+/ferroptosis axis and overcomes tamoxifen resistance in ER+ breast cancer

Abstract

Tamoxifen has been the mainstay therapy to treat early, locally advanced, and metastatic estrogen receptor-positive (ER+) breast cancer, constituting around 75% of all cases. However, emergence of resistance is common, necessitating the identification of novel therapeutic targets. Here, we demonstrated that long-noncoding RNA LINC00152 confers tamoxifen resistance via blocking tamoxifen-induced ferroptosis, an iron-mediated cell death. Mechanistically, inhibiting LINC00152 reduces the mRNA stability of phosphodiesterase 4D (PDE4D), leading to activation of cAMP/PKA/CREB axis and increased expression of TRPC1 Ca²⁺ channel. This causes cytosolic Ca²⁺ overload and generation of reactive oxygen species (ROS) that is, on one hand, accompanied by downregulation of FTH1, a member of the iron sequestration unit, thus increasing intracellular Fe²⁺ levels; and on the other hand, inhibition of the peroxidase activity upon reduced GPX4 and xCT levels. These ultimately induce lipid peroxidation and ferroptotic cell death in combination with tamoxifen. Overexpressing PDE4D rescues LINC00152 inhibition-mediated tamoxifen sensitization by de-activating the cAMP/Ca²⁺/ferroptosis axis. Importantly, high LINC00152 expression is significantly correlated with high PDE4D/low ferroptosis and worse survival in multiple cohorts of tamoxifen- or tamoxifen-containing endocrine therapy-treated ER+ breast cancer patients. Overall, we identified LINC00152 inhibition as a novel mechanism of ferroptosis induction and tamoxifen sensitization, thereby revealing LINC00152 and its effectors as actionable therapeutic targets to improve clinical outcome in refractory ER+ breast cancer.

Keywords: ER+ breast cancer; cAMP; calcium; ferroptosis; lncRNA; tamoxifen resistance.

Figure 1.. LINC00152 is upregulated in tamoxifen resistance and its higher expression correlates with poor prognosis.

A Heatmap of the z-scores of the FPKM values of top 25 most abundant, upregulated, and top 25, least abundant, downregulated lncRNAs in MCF-7 tamoxifen resistant (TAMR) cells compared to parental (PAR) cells in three biological replicates (0, 1, 2). The names of 25 genes are depicted on the heatmap. B qRT-PCR analysis of LINC00152 in both MCF-7 and T47D parental and TamR cells. C LINC00152 levels in tamoxifen sensitive vs. resistant tumors at baseline from GSE87411. D, E The disease-free survival in tamoxifen-treated ER+ breast cancer patients with tumors expressing low vs. high LINC00152 expression from GSE6532 (D) and GSE58644 (E). F The overall survival in patients with tumors expressing low vs. high LINC00152 expression from GSE42568. G Representative in situ hybridization (ISH) staining of LINC00152 in low vs. high expressers among endocrine therapy-treated ER+ breast carcinoma specimens from the Hacettepe cohort. Scale bar = 20 μm. The numbers of Hacettepe patients having low vs high LINC00152 and with recurrence vs. no recurrence are provided at the bottom of the panel. H The disease-free survival in Hacettepe patients with tumors expressing low vs. high LINC00152 expression as quantified by ISH staining. Data are presented as mean values ± standard deviation (SD). P-values for the bar graphs and boxplots were calculated with the unpaired, two-tailed Student’s t test. Significance for the Kaplan Meier survival graphs was calculated with Log-rank test. Chi-square test was used for G. * P < 0.05. Source data for this figure are provided as a Source Data file.

Figure 2.. Inhibiting LINC00152 overcomes tamoxifen resistance by inducing ferroptosis.

A The nuclear vs. cytoplasmic localization of LINC00152 in MCF-7 TamR cells. GAPDH, MALAT1 and DANCR were used as controls for nuclear and cytoplasmic RNAs. B LINC00152 knockdown efficacy in shLINC00152-expressing MCF-7 TamR cells. C Percent cell viability of MCF-7 TamR cells expressing shLINC00152 upon treatment with tamoxifen. D Percent viability inhibition in MCF-7 TamR cells transfected with siLINC00152 and treated with tamoxifen. E Viability assay in LINC00152-overexpressing T47D cells under tamoxifen treatment. F GSEA analysis showing the gene sets enriched in low LINC00152-expressing ER+ breast cancer patients treated with endocrine therapy from GSE87411. G Pearson correlation analysis of LINC00152 expression with ferroptosis score in ER+ patients from GSE58644. H, I Levels of ferroptosis score before and after endocrine therapy in sensitive (H) and resistant (I) ER+ patients from GSE87411. J Percentage cell viability inhibition in MCF-7 parental cells treated with tamoxifen with or without 0.2 or 0.3 μM of the ferroptosis inhibitor, liproxstatin-1. K, L Dose-response curves in MCF-7 (K) and T47D (L) parental vs. TamR cells treated with increasing doses of RSL3. M Percentage cell viability inhibition in MCF-7 TamR cells treated with tamoxifen alone or in combination with RSL3. N Relative intracellular Fe²⁺ levels in MCF-7 TamR shCtrl vs. shLINC00152 cells treated with 7.5 μM tamoxifen for 1.5 hrs. O Lipid peroxidation assay in T47D TamR cells transfected with siLINC00152 (siLINC) in combination with tamoxifen, showing fold change of lipid peroxidation. Cumene was used as a positive control. P Western blot analysis of the negative regulators of ferroptosis, xCT, GPX4 and FTH1 in MCF-7 TamR cells expressing shLINC00152 and treated with tamoxifen. Actin is used as a loading control in all Western blots unless stated otherwise. Q Percentage cell viability of shLINC00152-expressing MCF-7 TamR cells treated with tamoxifen with or without 0.1 or 0.2 μM of liproxstatin-1 (Lipro). Data are presented as mean values ± standard deviation (SD). P-values for F were calculated using the GSEA software (Broad Institute) while others were calculated with the unpaired, two-tailed Student’s t test. * P < 0.05, ** P < 0.01, n.s. not significant. Source data for this figure are provided as a Source Data file. Source data for this figure are provided as a Source Data file.

Figure 3.. LINC00152 inhibition increases cytosolic Ca²⁺ by activating cAMP signaling.

A The Pearson correlation analysis of calcium signaling with ferroptosis score in endocrine therapy sensitive ER+ patients after therapy from GE87411. B, C Levels of calcium signaling score before and after endocrine therapy in sensitive (B) and resistant (C) ER+ patients from GSE87411. D The levels of cytosolic Ca²⁺ in MCF-7 TamR cells transfected with two different siRNAs against LINC00152 and treated with tamoxifen. E qRT-PCR analysis of TRPC1 mRNA in MCF-7 TamR shCtrl and shLINC00152 cells treated with tamoxifen. F Western Blot analysis of PDE4D, p-CREB (S133), p-PKA (Thr197) and TRPC1 in MCF-7 TamR shCtrl and shLINC00152 cells treated with tamoxifen. G The Pearson correlation analysis of calcium signaling score with cAMP score in endocrine therapy sensitive ER+ patients after therapy from GSE124647. H cAMP levels in MCF-7 TamR shCtrl or shLINC00152 cells treated with tamoxifen. I The Pearson correlation analysis of cAMP score and ferroptosis score in endocrine therapy sensitive ER+ patients after therapy from GSE87411. Data are presented as mean values ± standard deviation (SD). P-values were calculated with the unpaired, two-tailed Student’s t test. * P < 0.05, ** P < 0.01, n.s. not significant. Source data for this figure are provided as a Source Data file.

Figure 4.. LINC00152 inhibits ferroptosis by regulating PDE4D and downstream cAMP/Ca²⁺ signaling in tamoxifen resistant cells.

A qRT-PCR analysis of LINC00152 and PDE4D in the shLINC00152-expressing MCF-7 TamR cells. B Western blot analysis of PDE4D in the shLINC00152-expressing MCF-7 TamR cells. C Western blot analysis of p-CREB (S133), p-PKA (Thr197), TRPC1 and ferroptosis inhibitor GPX4 in T47D TamR cells treated with tamoxifen in combination with GebR-7b. D Levels of ferroptosis score in low vs. high PDE4D-expressing ER+ patients from GSE4922. E, F Percentage cell viability inhibition in MCF-7 and T47D TamR cells co-transfected with siLINC00152 and PDE4D ORF and treated with tamoxifen. G Percentage of viability inhibition in shLINC00152-expressing MCF-7 TamR cells stably overexpressing PDE4D and treated with tamoxifen. H Western blot analysis of p-CREB (S133), p-PKA (Thr197), TRPC1 and GPX4 in shLINC00152-expressing MCF-7 TamR cells stably overexpressing PDE4D and treated with tamoxifen. I Relative ROS levels in MCF-7 TamR cells expressing shLINC00152 and treated with tamoxifen. J Percentage viability inhibition in T47D TamR cells treated with tamoxifen and GebR-7b with or without 0.1 or 0.2 μM of liproxstatin-1 (Lipro). Data are presented as mean values ± standard deviation (SD). P-values were calculated with the unpaired, two-tailed Student’s t test. * P< 0.05, ** P < 0.01. Source data for this figure are provided as a Source Data file.

Figure 5.. LINC00152 stabilizes PDE4D mRNA by interacting with its 3’UTR.

A The Pearson correlation analysis of LINC00152 and PDE4D mRNA in patients from GSE6532. B PDE4D protein expression determined by IHC in low vs. high LINC00152-expressing patients from the Hacettepe cohort. C Representative images of LINC00152 ISH, PDE4D IHC and H&E staining in patients from the Hacettepe cohort. Scale bar = 20 μm. D, E qRT-PCR analysis of PDE4D in MCF-7 TamR (D) and T47D TamR (E) cells transfected with siLINC00152 for 48 hours followed by treatment with 5 μg/ml of Actinomycin D. F Percentage fold enrichment of PDE4D in LINC00152 pull down in MCF-7 TamR cells. G Interaction between LINC00152 and PDE4D 3’UTR predicted by the IntaRNA tool. The visualization of the interaction was obtained using RILogo. H, I Luciferase assay in parental MCF7 (H) and T47D (I) cells co-transfected with LINC00152 overexpression vector and PDE4D 3’UTR, demonstrating binding to PDE4D 3’UTR. Data are presented as mean values ± standard deviation (SD). P-values were calculated with the unpaired, two-tailed Student’s t test. * P < 0.05, ** P < 0.01. Source data for this figure are provided as a Source Data file.

Figure 6.. LINC00152 inhibition overcomes tamoxifen resistance in vivo.

A PDE4D mRNA expression in MCF-7 parental (PAR) vs. TamR xenograft tumors. B Tumor volumes of MCF-7 TamR xenografts expressing shLINC00152 and treated with tamoxifen (3 mg/kg, daily). C, D Tumor weights (C) and tumor images (D) from mice in B. E qRT-PCR analysis of LINC00152 in tumors from B. F IHC staining of PDE4D in tumors from B. Scale bar = 100 μm. Data for the bar graphs and box plots are represented as mean values ±SD, while data for the tumor volume graph are represented as mean values ± standard error of the mean (SEM). P-values for the bar graphs and box plots were calculated with the unpaired, two-tailed Student’s t test. Significance for the tumor volume graph was calculated with two-way ANOVA. Source data for this figure are provided as a Source Data file.

Figure 7.. Schematic summary of LINC00152/PDE4D-driven tamoxifen resistance.

A LINC00152 is overexpressed in tamoxifen resistance, binds to PDE4D 3’UTR, and increases its mRNA stability. This results in reduced cAMP levels, and de-activation of PKA/CREB, leading to decreased expression of TRPC1, reduction of cytosolic Ca²⁺ and blockage of reactive oxygen species (ROS) generation. Furthermore, the negative regulators of ferroptosis, FTH1, GPX4 and xCT are upregulated, and contribute to reduction of intracellular Fe²⁺, and inhibition of lipid peroxidation that ultimately block ferroptosis-mediated cell death and mediate tamoxifen resistance. B Targeting LINC00152 in combination with tamoxifen leads to reduced mRNA stability of PDE4D due to loss of binding to its 3’UTR, thereby activating cAMP/PKA/CREB, similar to targeting PDE4D with GebR-7b in combination with tamoxifen. This further increases TRPC1, elevates cytosolic Ca²⁺, generates reactive oxygen species (ROS) and increases Fe²⁺, accompanied by reduced levels of FTH1, GPX4 and xCT, leading to lipid peroxidation, ferroptotic cell death and tamoxifen sensitization.