Inhibition of MUC1‐C regulates metabolism by AKT pathway in esophageal squamous cell carcinoma

Abstract Esophageal squamous cell carcinoma (ESCC) is one of the most common digestive tumors worldwide. The Mucin 1 (MUC1) heterodimeric protein has been confirmed that is overexpressed in ESCC and induced adverse outcomes. However, the detailed mechanism(s) remained challenging. So, we investigated the relationship between MUC1‐C and metabolism in ESCC cells. In the results, TP53‐induced glycolysis and apoptosis regulator (TIGAR) was overexpressed and correlative with MUC1‐C positively in ESCC tissue. Targeting MUC1‐C inhibits AKT–mTORC–S6K1 signaling and blocks TIGAR translation. We found that the inhibitory effect of GO‐203 on TIGAR was mediated by inhibition of AKT–mTOR–S6K1 pathway. The findings also demonstrated that the suppressive effect of GO‐203 on TIGAR is related to the decrease of glutathione level, the increase of reactive oxygen species and the loss of mitochondrial transmembrane membrane potential. In xenograft tissues, GO‐203 inhibited the growth of ESCC cells and lead to the low expression of transmembrane C‐terminal subunit (MUC1‐C) and TIGAR. This evidence supports the contention that MUC1‐C is significant for metabolism in ESCC and indicated that MUC1‐C is a potential target for the treatment of ESCC.


| INTRODUCTION
Esophageal squamous cell carcinoma (ESCC) is one of the most common digestive tumors in the world, especially in developing countries (Torre et al., 2015). Surgical resection is the standard treatment for ESCC, but, it is not that effective for patients who want to survive for a long time (Mariette, Piessen, & Triboulet, 2007;Mizoguchi et al., 2014). The biological characteristics of ESCC are growing malignantly and metastasizing early, which lead to poor prognosis. Therefore, to explore the molecular pathological mechanism of ESCC is of great significance in diagnosis and treatment.
Mucin 1 (MUC1) is a transmembrane heterodimer glycoprotein which is aberrantly overexpressed in ESCC (Ye et al., 2011). And it had been reported that the overexpression is related to the characteristics of a tumor (Nath & Mukherjee, 2014). MUC1 is processed by autocleavage into two subunits, the extracellular N-terminal subunit and the transmembrane C-terminal subunit (MUC1-C). These two subunits form a stable heterodimeric complex at the cell membrane (Kufe, 2009). MUC1-C consists of an intrinsically disordered 72 amino acid cytoplasmic domain (MUC1-CD), which is phosphorylated by different kinases and interacts with various effectors associated with transformation (D.W. Kufe, 2013;Huang et al., 2005). Importantly, MUC1-C interacts with diverse effectors such as β-catenin (Yamamoto, Bharti, Li, & Kufe, 1997), IKK β (Ahmad et al., 2007), and NFkB p65 , which have been connected to transcription. MUC1-C inhibits mitochondrial transmembrane potential (MMP) loss induced by stress through the incorporation of the mitochondrial outer membrane (Ren et al., 2006).
Proliferation and metastasis of the tumor are accompanied by changes in metabolism and secondary microenvironment. TP53induced glycolysis and apoptosis regulator (TIGAR) is considered as a metabolic index. Previous studies have shown that high levels of TIGAR expression are closely associated with poor clinical outcomes in patients with multiple types of cancer including chronic lymphocytic leukemia (Hong et al., 2016), invasive breast cancer (Won et al., 2012), Stages II and III colorectal cancer (Alkhayal et al., 2016), nasopharyngeal carcinoma (Wong et al., 2015;Zhao et al., 2016), and non-small-cell lung cancer (Shen et al., 2018).
The phosphoinositide 3-kinase (PI3K), AKT (also known as protein kinase B, PKB) and mammalian target of rapamycin (mTOR), dominate various indications of cancer including cell cycle, survival, metabolism, motility, and genomic instability. PI3K-AKT signaling pathway regulates the translation in the upstream of rapamycin complex 1 (mTORC1) target. mTORC1 energizes 40S ribosomal protein S6 kinases (S6Ks) that promote cap-dependent translation and thus upgrade the eIF4A RNA helicase activity (Sonenberg & Hinnebusch, 2009). S6Ks lead to degradation of suppressor programmed cell death protein 4 (PDCD4), which is an eIF4A inhibitor in cancer cells (Dorrello et al., 2006). eIF4A sets up translation by unraveling highly organized 5′ untranslated regions in messenger RNA (mRNA), just like the encoding cyclin D1 and MYC (Rogers, Komar, & Merrick, 2002). Tumor cells can induce eIF4A RNA helicase to regulate intracellular transcriptional and translation through the AKT-mTORC pathway.
Our previous study has demonstrated that the MUCI-C is overexpressed in ESCC tissue with an important function in hallmarks of ESCC such as proliferation, metastasis, and antiapoptosis (Xin et al., 2018). And another report has described that effect of GO-203 is related to downregulation of the AKT-mTOR pathway and inhibition of cap-dependent translation of TIGAR protein (Ahmad et al., 2017). However, the detailed mechanism of MUC1-C in ESCC remains unclear. In the study, we had explored the effects of MUC1-C in metabolism and the mechanism in human ESCC cells. (c) the incisal edge was proved to be as normal esophageal tissue by pathology. This study was approved by the Ethics Committee of Shandong Provincial Hospital Affiliated to Shandong University in accordance with the ethical guidelines of the Declaration of Helsinki.
All participants had signed written informed consent.
Reactivity was quantified with horseradish peroxidase-conjugated secondary antibodies and enhanced chemiluminescence detection system (Amersham Imager 600; General Electric, Fairfield, CT).
Reactive oxygen was measured by flow cytometry.
The specific experimental steps are operated according to the instructions by GSH content determination kit (Solarbio, Beijing, China). The absorbance was measured at 412 nm and calculation based on the standard curve of GSH.

MMP was measured using the Mitochondrial Membrane Potential
Assay Kit with JC-1 (Beyotime) as described (Lu et al., 2015). The images were examined by confocal microscopy (Molecular Devices).

| Immunohistochemistry
The expression of MUC1-C and TIGAR was performed by the streptavidin-peroxidase method. Patient tissues and xenograft tumor tissues were fixed with formalin and sectioned after paraffin embedded.
Antigen was retrieved after dewaxing and hydration and then incubated with hydrogen peroxide. Tissue sections were incubated with rabbit anti-MUC1-C and anti-TIGAR antibodies (1:100; Abcam) overnight at 4°C . Then marked these sections by a secondary biotinylated antibody (ZSGB Biotech, Beijing, China) and color development utilizes the diaminobenzidine (DAB) method. The images were examined using a fluorescence microscope (Olympus BX51; Olympus, Japan) and evaluated by two independent pathologists.

| Apoptosis analysis
Cells were prepared according to the P-phycoerythrin (PE) Annexin V apoptosis detection kit with 7-amino-actinomycin D (D7-AAD; Becton Dickinson, New Jersey). After incubated using PE and 7-AAD for 15 min, the cells were determined by flow cytometry.

| ESCC tumor xenograft in mice
Six-to eight-week-old female BALB/c nu/nu mice were injected with 1 × 10 7 ESCC ECA109 cells subcutaneously in the flank. When tumors reached~150 mm 3 , the mice were pair matched into three groups of five mice each, and treated with (a) 15 mg/kg CP-2 administered intraperitoneally daily, (b) 15 mg/kg GO-203 administered intraperitoneally each day, (c) phosphate-buffered saline (PBS; control vehicle). Recording tumors size every 4 days and weighed them after 20 days. Tumor volumes were calculated by using the formula V = L x W 2 /2, where L and W are the largest and smallest diameters, respectively. The protocol was approved by the Animal Ethics Committee of Shandong Provincial Hospital Affiliated to Shandong University.

| Statistical analysis
SPSS 20.0 (IBM, Armonk, NY) was used to create databases for data analysis. The quantitative data were expressed as a mean ± standard deviation. For two groups, two-tailed Student's t test was performed, and for three or more than three groups, one-way analysis of variance was used. The relationship between MUC1-C and TIGAR expression in patient tumors and xenograft tumor was calculated by correlation analysis. A value of p < 0.05 was selected before the experiments to indicate statistical significance.

| TIGAR overexpresses in ESCC tissue and targeting MUC1-C inhibits TIGAR translation in ESCC cells
The MUC1-C expression was obviously inhibited by the treatment of immunohistochemistry. Obvious staining was shown in the cytoplasm of tumor tissue but weak in normal esophageal tissue (Figure 1b; left), and the correlation of MUC1-C and TIGAR was positive (Figure 1b; right). Targeting MUC1-C has been reported to inhibit TIGAR at the protein level in hematologic malignancies (Yin, Kosugi, & Kufe, 2012;Yin, Kufe, Avigan, & Kufe, 2014) and colon cancer (Ahmad et al., 2017). In addition, the treatment of ECA109 and KYSE150 cells with GO-203 lead to a low expression of TIGAR levels ( Figure 1c). As mentioned, there was no significant change in TIGAR mRNA levels between the two treatment cells (Figure 1d). These results suggest that TIGAR is closely related to the ESCC and it is regulated effectively by MUC1-C. which manifest TIGAR is the downstream target protein of the PI3K-AKT pathway. mTORC inhibitor rapamycin was used to detect TIGAR levels to confirm its potential role. Indeed, the treatment of ECA109 cells with rapamycin inhibited TIGAR levels (Figure 2f). These findings demonstrated that the MUC1-C regulates TIGAR protein translation by the PI3K-AKT-S6K1 pathway.

| Targeting MUC1-C enhances ROS and inhibits GSH levels
As mentioned earlier, the disruption of redox equilibrium was blocked by MUC1-C, and the silencing of MUC1-C was related to the increase of ROS production (Yin et al., 2003). The GO-203 treatment of ECA109 cells increases ROS production significantly compared with CP-2 (Figure 3a,b). GSH can affect the metabolic process of cells by activating a variety of enzymes. GO-203 shows a significant inhibitory effect on the GSH level of ECA109 and KYSE150 cells (Figure 3c). In addition, GO-203 reduced GSH levels and induced ROS production return to the original level after the addition of N-acetyl-L-cysteine (NAC). According to these results, inhibition of MUC1-C leads to the production of ROS and consumption of GSH. In addition, both of these effects of GO-203 were reversed by NAC.

| Targeting MUC1-C inhibits mitochondrial membrane potential and promotes apoptosis in ESCC cells
Redox balance is pivotal in metabolism, and its disruption can lead to loss of mitochondrial membrane potential loss and cell apoptosis (Yin et al., 2011). In the result, green fluorescence was increased in ECA109 and KYSE150 cells treated with GO-203, which confirmed the reduction of mitochondrial membrane potential (Figure 4a,b). In the flow result, we find the apoptosis of ECA109 and KYSE150 cells that treated with GO-203 increased significantly, but not obvious in control CP-2 groups (Figure 5a, B).

| GO-203 induces suppression of ESCC tumors in mice
Tumor xenografts of nude mice were established to study the GO-203 function in vivo. The average volumes of tumors in the GO-203-treated mice were repressed obviously. Contrary to the antitumor activity of GO-203, CP-2 treatment has no effect on ECA109 tumor growth (Figure 6a

| DISCUSSION
ESCC is one of the cancers with high incidence and high mortality.
The main cause of poor prognosis is that ESCC metastasizes at the early stage (Martin, Herbert, & Hocevar, 2010). Gene therapy has F I G U R E 3 Inhibition of MUC1-C disrupts the redox balance. (a-b), ECA109 and KYSE150 cells were treated each day with 5 μM GO-203 or CP-2 for 3 days, and 5 mM NAC was also added in GO-203 treated cells. The ROS generation was measured using DCFH-DA by flow cytometry, and mean fluorescence intensity was the symbol of ROS production. (c) Cells were prepared as previously mentioned in (a-b). The cells were analyzed for GSH levels. **p < 0.01; ***p < 0.001. MUC1-C: transmembrane C-terminal subunit; ROS: reactive oxygen species [Color figure can be viewed at wileyonlinelibrary.com] GONGSUN ET AL.

| 12023
become a new direction of cancer treatment. It is of great significance to search for the genes associated with ESCC and to clarify the mechanism of action in the treatment of ESCC.
MUC1 has been confirmed to be overexpressed in ESCC (Jing et al., 2015) and associated with its self-renewal and tumorigenicity (Shi, Chen, Yang, & Liu, 2015). MUC1 mainly relies on the MUC1-C subunit to function after translation. MUC1-C interacts with multiple kinases and effectors owing to its special structure MUC1-CD, and then participate in transformation (D.W. Kufe, 2013;Kufe, 2009). The MUC1-C transforming function is dependent on homodimer formation, which is mediated by a CQC motif present in the cytoplasmic domain of MUC1-C (Leng et al., 2007). MUC1-C transmits signals as signal proteins in cells, but this function can be inhibited by GO-203 (Raina et al., 2009). The inhibitor has been reported in inhibiting cell proliferation in vitro effectively of breast cancer   (Takahashi et al., 2015), and certain hematologic malignancy (Yin et al., 2010). We had demonstrated that MUC1-C is linked to the growth, metastasis, and other characteristics of ESCC cells (Xin et al., 2018). The specificity of GO-203 on ESCC cells is notable, because of the binding to the CQC motif of MUC1-C and inhibition of its function. In contrast, the CQC motif in CP-2 has been changed to AQA motif, so CP-2 does not inhibit MUC1-C functioned in ESCC cells.
TIGAR was originally thought to be the transcriptional target of p53 (Bensaad et al., 2006). As a major tumor suppressor protein, the function of p53 is to inhibit the growth or survival of tumor cells by inducing apoptosis or senescence. What is more, growing studies have reported that p53 can also regulate metabolism (Vousden & Ryan, 2009), and TIGAR facilitates the activation of p53. The activity of TIGAR and other genes by p53 promotes an antioxidant response (Budanov, Sablina, Feinstein, Koonin, & Chumakov, 2004;Cosentino, Grieco & Costanzo, 2014), which helps cells to survive transient or low levels of stress, and has lately been attested to be significant in impeding malignant progression (Li et al., 2012). We testified that MUC1-C and TIGAR are overexpressed in ESCC tissue and there is a positive correlation between them. The previous study had reported inhibition of MUC1-C in hematological malignancies and colon cancer (Ahmad et al., 2017) downregulates TIGAR at the protein level (Yin et al., 2012(Yin et al., , 2014. GO-203 also reduced TIGAR protein expression in the treated ESCC cells, but there was no change in mRNA level. We verified that MUC1-Cregulate TIGAR at the translation level but not at transcriptional level. In recent years, research of PI3K-AKT-mTOR signaling pathway has got diverse achievements. This signaling pathway now is considered as a pivotal regulator in the activities of the cell cycle, cellular proliferation, growth, survival, protein synthesis, and glucose metabolism (Engelman, Luo, & Cantley, 2006;Vanhaesebroeck, Stephens, & Hawkins, 2012). AKT controls translation by activation of mTORC1 that results in the phosphorylation of p70S6 Kinase (Ma & Blenis, 2009). In turn, S6K phosphorylates and thereby induces the degradation of PDCD4, an inhibitor of eIF4A RNA helicase activity that regulates translation of proteins (Dorrello et al., 2006). This study confirms that the targeting MUC1-C by GO-203 inhibits phosphorylation of AKT and S6K in ESCC cells. PDCD4 is a downstream target of S6K1, which enhances protein synthesis to promote cell growth and survival. GO-203 promotes PDCD4 expression, which is an inhibitor of elF4A helicase. According to these outcomes, GO-203 regulates AKT-S6K1-elF4A signaling in ESCC cells. In our study, it shows that the treatment of ESCC cells with PI3K inhibitor LY294002 and AKT inhibitor GSK690693 both downregulated TIGAR proteins. Rapamycin, that is an inhibitor to mTORC1 also downregulated TIGAR protein level. These findings thus confirm that the MUC1-C regulated TIGAR translation by AKT-S6K1-elF4A pathway.
TIGAR also reduces the content of ROS in cells by increasing the production of NADPH (Yin et al., 2012). NADPH can convert GSH oxide into reduced glutathione. With the GO-203 treatment of ESCC cells, GSH decreased and ROS increased, which suggest that inhibition of MUC1-C could deplete NADPH. Changes in redox equilibrium caused by GO-203 significantly decreased the level of GSH. After the addition of NAC into GO-203, it was reversed that GSH recovered and ROS decreased to the original level. The electron transport chain present in the mitochondrial inner membrane generates ROS. The previous research has confirmed that the MUC1-C translocates to the mitochondrial outer membrane during the stress response, thus attenuating the reducing of mitochondrial membrane potential (Ren et al., 2004(Ren et al., , 2006. Our findings have shown that inhibition of MUC1-C leads to mitochondrial membrane potential loss. The inhibition of MUC1-C induced disruption redox imbalance by   This study supports that MUC1-C is crucial in supporting the growth, survival, and metabolism of ESCC cells. Importantly, MUC1-C is basically disordered and thereby has the capacity to serve as a substrate for multiple kinases and as a binding partner for diverse effectors of gene transcription. Such as, (a) the MUC1-C cytoplasmic domain is phosphorylated by GSK3b and binds directly to β-catenin, linking MUC1-C to the WNT pathway (Yin et al., 2017); (b) MUC1-C is also known to bind IKKβ (Ahmad et al., 2007) and Rel A p65  and then participates constitutive NFκB activation; (c) MUC1-C activates the BMI1 gene through MYC protein-dependent mechanism, which affects the epigenetics of cells (David, Hamilton, & Palena, 2016;Hiraki et al., 2017); (d) MUC1-C could promote immune evasion and protect cells from damage through reducing the expression of PD-L1 Raina et al., 2015). These reports provide evidence that MUC1-C represents a potential target for the treatment of ESCC although further mechanisms are still needed to be explored. And MUC1-C inhibitor may become a new agent for ESCC therapy.
In conclusions based on our study, MUC1-C affects tumor cell metabolism and alters cell apoptosis by regulating GSH levels, ROS production, and changes in mitochondrial membrane potential. All these effects are achieved through the AKT-mTORC-S6K1 signaling pathway in conjunction with TIGAR. Furthermore, targeting MUC1-C function inhibits the cellular proliferation and survival of ESCC cells in vivo and represses the expression of TIGAR in xenograft tissues.
Hence, the MUC1-C and TIGAR may become potential targets for the treatment of ESCC cancer, and the combination therapy of the two genes may be more effective for ESCC patients.

CONFLICTS OF INTEREST
The authors have declared that there are no conflicts of interest.