Ecdysoneless Overexpression Drives Mammary Tumorigenesis through Upregulation of C-MYC and Glucose Metabolism

Abstract

Ecdysoneless (ECD) protein is essential for embryogenesis, cell-cycle progression, and cellular stress mitigation with an emerging role in mRNA biogenesis. We have previously shown that ECD protein as well as its mRNA are overexpressed in breast cancer and ECD overexpression predicts shorter survival in patients with breast cancer. However, the genetic evidence for an oncogenic role of ECD has not been established. Here, we generated transgenic mice with mammary epithelium-targeted overexpression of an inducible human ECD transgene (ECDTg). Significantly, ECDTg mice develop mammary hyperplasia, preneoplastic lesions, and heterogeneous tumors with occasional lung metastasis. ECDTg tumors exhibit epithelial to mesenchymal transition and cancer stem cell characteristics. Organoid cultures of ECDTg tumors showed ECD dependency for in vitro oncogenic phenotype and in vivo growth when implanted in mice. RNA sequencing (RNA-seq) analysis of ECDTg tumors showed a c-MYC signature, and alterations in ECD levels regulated c-MYC mRNA and protein levels as well as glucose metabolism. ECD knockdown-induced decrease in glucose uptake was rescued by overexpression of mouse ECD as well as c-MYC. Publicly available expression data analyses showed a significant correlation of ECD and c-MYC overexpression in breast cancer, and ECD and c-MYC coexpression exhibits worse survival in patients with breast cancer. Taken together, we establish a novel role of overexpressed ECD as an oncogenesis driver in the mouse mammary gland through upregulation of c-MYC-mediated glucose metabolism.

Implications: We demonstrate ECD overexpression in the mammary gland of mice led to the development of a tumor progression model through upregulation of c-MYC signaling and glucose metabolism.

Figure 1.

ECD overexpression promotes mammary gland hyperplasia, heterogeneous tumors, and preneoplastic lesions. A, Representative whole-mount staining images of littermate control and mouse mammary gland in ECDTg mouse at 5 to 6 months of age. Scale bar, 400 μm; inset 1,000 μm. B, H&E staining of ECDTg tumors. Scale bar, 100 μm. C, Western blot for ECD expression (tumor numbers in Supplementary Table S1). D, Age-matched control and ECDTg mammary gland at 15 to 25 months, inset 1,000 μm. E, H&E staining of ECDTg mammary gland. Scale bar, 400 μm; and inset 100 μm. F, Western blot for ECD protein. HSC-70, used as a loading control. MG, mammary gland.

Figure 2.

Characterization of ECDTg tumors by IHC. A–F,ECDTg tumor subtypes were immunostained with indicated antibodies. Scale bar, 50 μm (magnification 400×); inset scale bar, 10 μm (magnification 1,000×).

Figure 3.

Organoid formation and tumor growth of ECDTg tumor-derived cells. A, Images of organoids upon doxycycline treatment. B, Western blot of lysates from organoids, β-actin used as a loading control. C, Three independent tumors ±doxycycline, insets show high magnification. Scale bar, 400 μm. D and E, Organoid number and size after 4 days of doxycycline treatment (N = 3 times and 4 wells per condition). **, P < 0.01; ***, P < 0.001. Data represents mean ± SEM with two tailed unpaired t test, and nested t test. F, Western blot shows ECD expression; densitometry in respect to without doxycycline in each tumor normalized with β-actin, shown below. G, qRT-PCR using human ECD specific primers. H, T3 organoids were injected orthotopically into 5 NSG mice. The tumor growth is plotted as tumor volume over days. I, Images of the isolated tumors. 2 of the mice (1F and 4F) died during experiment due to unknown reason. J, tumor fragments of 2 mm³ size from T3 were transplanted in NSG mice. After 10 days, the tumor volume was measured and mice were distributed into two groups for with and without doxycycline treatment, and growth was monitored for the next 12 days. Mean ± SEM of tumors is calculated by mixed model of ANOVA analysis. ***, P < 0.001. K, Images of the tumors harvested after the dissection. L, qRT-PCR using human and mouse ECD specific primers. M, IHC of tumor sections from doxycycline-treated mice 3F, stained with indicated antibodies. DOX, doxycycline. Lt, left; Rt, right.

Figure 4.

RNA-seq analyses comparison of ECDTg tumors and control mammary glands. A, Principal component analysis (PCA) analysis of RNA-seq data shows clustering of control mammary glands and ECDTg tumor datasets. First Principal component (PC1), 51.2% viability; and Second Principal component (PC2), 16.9% viability. B, Heatmap of the top 50 differentially expressed genes among different biologic replicates of control mammary glands and tumors. Upregulated genes in red, downregulated genes in green. C, Box plot shows the enrichment scores obtained using single sample Gene Set Enrichment Analysis of the c-MYC signature genes. D, MYC-regulated metabolic genes in ECDTg tumors. RNA-seq followed by cluster comparison analyses of 4 tumors. 82 up- and downregulated genes are shown. Red (upregulated) green (downregulated). E, PCA shows clustering of ECDTg tumors (n = 4) based on tumor type. PC1 represent 45.5% viability and PC2 represent 34.2% viability. F, Heatmap shows heterogeneity among the tumors. G, Heatmap depicting EMT signature upregulated in spindle cell carcinoma (Tumor#T3). H, Heatmap displaying papillary carcinoma signature genes upregulation in papillary carcinoma (T4).

Figure 5.

ECDTg tumors exhibit upregulation of c-MYC. A, qRT-PCR shows increased c-Myc, mRNA in tumors. Bar graph shows fold change of mRNAs (controls, n = 4; and tumors, n = 4). *, P < 0.05. B, Western blot with indicated antibodies. HSC-70, used as a loading control. C, IHC of indicated tumors and c-MYC staining. D and E, Representative IHC images and qRT-PCR mRNA expression analysis of 4 independent ECDTg transplanted tumors from doxycycline-untreated or doxycycline-treated mice. F, Western blot of lysates of organoids. HSC-70 used as a loading control. DOX, doxycycline.

Figure 6.

c-MYC mRNA levels and protein stability upon alterations in ECD levels. Western blot of ECD and c-MYC protein, HSC-70 used as a loading control (A–C) and mRNA by qRT-PCR (D–F) in indicated cell lines. G and H, Time-dependent mRNA expression of c-MYC upon induction of ECD. mRNA quantitation data represents mean ± SEM with two-tailed unpaired t test. n = 3; *, P < 0.05; **, P < 0.01; ***, P < 0.001. Schematic display of the qPCR primer pairs (marked arrows) used to measure pre-mRNA and mature mRNA species (I). Bar graphs show overexpression of ECD resulted in upregulation of c-MYC mRNA (J and K). Pre-mRNA to mRNA ratio in ECD-overexpressing cells is shown in bar graphs as fold change in comparison with controls. 18S was used for normalization (L and M). The stability of c-MYC mRNA was analyzed in doxycycline-inducible ECD-overexpressing indicated cells in the presence or absence of doxycycline for 72 hours. The cells were treated with actinomycin D (5 μg/mL) for indicated time points and qRT-PCR was performed after RNA isolation. GAPDH was used for normalization. Half-life of c-MYC mRNA in MCF10A (N) and 76NTERT (O). Data represents mean± SE of three independent experiments. The stability of c-MYC protein was analyzed in doxycycline-inducible ECD-overexpressing MCF10A cells in the presence or absence of doxycycline for 72 hours. The cells were treated with cycloheximide (50 μmol/L) for indicated time points and lysates were collected. P, Western blot with indicated antibodies. P and R, Half-life of c-MYC protein in ECD overexpressing MCF10A cells. Q and S, c-MYC protein stability was analyzed after 48 hours of treatment of SUM-159 with control or ECD siRNA, followed by cycloheximide (CHX) treatment. Q, Western blot shows c-MYC protein in ECD KD cells. S, Exponential decay plot shows half-life of c-MYC upon ECD KD. Densitometry of c-MYC after normalizing to their respective loading controls, GAPDH and β-actin in comparison with no CHX treatment are indicated on top in (P) and (Q). NS, not significant; DOX, doxycycline; h, hours; mnts, minutes.

Figure 7.

Metabolomics analysis of ECDTg tumor cells. A, Organoids from ECDTg T1, with or without doxycycline for 4 days, followed by analysis of glycolytic metabolites. Bar diagram shows the fold decrease in the levels of glycolytic metabolites. *, P < 0.05; **, P < 0.01; ***, P < 0.001. B, Glucose uptake in three ECDTg tumor organoids (in triplicates). The values were normalized with respective to cell counts and depicted as compared with doxycycline-untreated organoids. Quantification of results from four replicates is shown as a bar graph. *, P < 0.05; **, P < 0.02; ***, P < 0.002. Data represents as mean ± SD and two-tailed unpaired test with Welch correction. C, Western blot of lysates from organoids used for experiments in (A) and (B) HSC-70 used as a loading control. Densitometry of ECD in respect to without doxycycline in each tumor, after normalizing with loading control. D, Glucose uptake in control and ECD siRNA-treated cells. E, Western blot of ECD protein, β-actin used as a loading control. F and G, Glycolytic metabolites in hMECs upon doxycycline-inducible ECD upregulation. Bar diagram shows the fold increase in the levels of glucose and lactate in ECD-overexpressing cells compared with control cells. H and I, Glucose uptake in ECD-overexpressing cells, normalized with respect to cell counts and depicted in comparison with control cells. J and K, Western blot shows the levels of ECD level in cells used for experiments in (H) and (I). L, Schematic depicts the intermediate products and enzymes of glycolysis. Red-marked are c-MYC–regulated glycolytic enzymes. PM, plasma membrane. mRNA levels of indicated genes in cells upon ECDKD (M) or ECD overexpression (N and O). P, Glucose uptake rescue in SUM-159 cells by overexpression of HA-c-MYC or mECD-GFP in ECD KD cells. In each case *, P < 0.05; **, P < 0.01; ***, P < 0.001. Data represents as mean ± SD of four replicates and two-tailed unpaired tests with Welch correction. Q, Western blot shows expression of indicated proteins. β-actin used as a loading control. DOX, doxycycline; ctrl, control.