Ajuba Overexpression Promotes Breast Cancer Chemoresistance and Glucose Uptake through TAZ-GLUT3/Survivin Pathway

The LIM protein Ajuba has been implicated in the development of human cancers. To date, its expression pattern and biological significance in breast cancers (BC) have not been fully investigated. In the current study, we examined Ajuba protein levels in 93 invasive ductal carcinoma specimens by immunohistochemistry. The Ajuba expression level was elevated in breast cancer tissue compared with normal tissue. Ajuba overexpression is correlated with advanced tumor-node-metastasis (TNM) stage, positive node status, and adverse patient outcomes. The Ajuba protein level was also higher in BC cell lines compared to normal breast epithelial cell line MCF-10A. Ectopically expressed Ajuba in MCF-7 cells stimulated in vitro and in vivo cell growth, invasion, cell cycle progression, and decreased paclitaxel-induced apoptosis. RNA-sequencing (RNA-seq) followed by gene set enrichment analysis (GSEA) analysis showed that Ajuba overexpression regulated the Hippo signaling pathway. Ajuba overexpression also increased glucose uptake and increased expression of TAZ, GLUT3, and Survivin. TAZ knockdown abolished the role of Ajuba on GLUT3 and Survivin induction. The ChIP assay showed that TEAD4, a major TAZ binding transcription factor, could bind to the GLUT3 and Survivin promoter regions. In conclusion, our data demonstrated that elevated Ajuba expression is correlated with poor BC prognosis and regulated malignant behavior through TAZ-GLUT3/Survivin signaling in BC cells.


Introduction
Breast cancer (BC) is one of the most commonly diagnosed cancers and the second leading cause of cancer-related deaths in women [1][2][3][4]. In the past decades, the development of novel therapeutic treatments has reduced mortality and improved the survival of breast cancer [5]. However, due to chemotherapy resistance and lack of efficient treatment for highly aggressive subtypes such as TNBC, the prognosis of advanced breast cancer remains poor. Therefore, there is a compelling need for identifying novel biomarkers and therapeutic targets to improve the overall outcome of breast cancer patients [6][7][8][9].
Ajuba is a LIM family member (Ajuba, LIMD1, and WTIP) [10], which is characterized by tandem LIM protein in the C-termini. As an adaptor protein that could shuttle between the cytoplasm and the nucleus, Ajuba interacts with various proteins to form various complexes involving multiple signaling pathways [11][12][13]. Recently, several studies imply that Ajuba interacts with the Hippo pathway's core components, including LATS and WW45, and acts as a negative regulator of Hippo signaling [10,14].
It has been reported that Ajuba is involved in the development of various human cancers. Ajuba expression is increased in cervical cancer [15], esophageal squamous cell carcinoma (ESCC) [16], colorectal cancer (CRC) [17,18], gastric cancer [19], and pancreatic cancer [20]. Ajuba has also been implicated in a variety of oncogenic processes. Ajuba promotes cell proliferation in CRC and pancreatic cancer cells [17,21]. Ajuba promotes ESCC cell invasion by activating ERK1/2 [16]. Ajuba augments tumor metastasis [18] and inhibits apoptosis [17]. Mutations of Ajuba regulate drug sensitivity of head and neck squamous cell carcinoma (HNSCC) [22]. Although growing evidence suggests that Ajuba acts as an oncogene to promote tumorigenesis, several studies reported Ajuba could function as a tumor suppressor [23,24]. To date, the expression pattern and biological functions of Ajuba in human breast cancers have not been fully elucidated.

Materials and Methods
2.1. Specimens. The current protocol was reviewed and approved by the Institutional Review Board, Cancer Hospital of China Medical University. This study was carried out following the Declaration of Helsinki principle. Written informed consent was obtained from patients. Breast cancer specimen paraffin blocks were from the Pathology Department in the First Affiliated Hospital of China Medical University, containing specimens no longer required to be maintained.
2.2. Immunohistochemistry. Immunohistochemistry was performed according to protocols reported previously [25]. In brief, 4 μm paraffin sections were deparaffinized using xylene and treated with ethanol. Peroxidase blocking was performed using H 2 O 2 solution (concentration: 3% v/v). Antigen retrieval was performed using citrate buffer. After blocking with ready-to-serum and incubation with Ajuba primary antibody (1 : 100, Sigma, USA), the section was treated with the Elivision plus Kit from Maixin (Fuzhou, China) and immunostaining was developed using the DAB kit. Counterstaining was performed using hematoxylin.
The Ajuba plasmid and the corresponding negative pCMV6 empty vector were from Origene company and transfected into cells using Lipofectamine 3000. Ajuba specific siRNA was from Dharmacon and transfected using the Dharmafect1 reagent. All transfection procedures were conducted following the manufacturer's protocol.
2.6. RNA-Sequencing. The RNA-sequencing experiments were carried out by Novogene corporation (Beijing, China). The libraries were sequenced on an Illumina NovaSeq6000 platform. The sequence data was subjected to standard quality control (QC). Gene set enrichment analysis (GSEA) is performed using the software downloaded from the GSEA website (http://software.broadinstitute.org).

CCK-8 and Colony Formation
Assays. CCK-8 and colony formation assays were performed according to protocols reported previously [25,28]. For CCK-8, cells were firstly plated into 96-well plates (3000 cells/well). Ten μl of the CCK-8 reagent (Cell Counting Kit-8; Dojindo, Kumamoto, Japan) was added into each well. After incubation for 2 hours, absorbance was examined using a microplate reader (wavelength: 450 nm). For colony formation capacity, cells were seeded into 6 cm plates (concentration: 1000 cells/ plate) and cultured for about two weeks. The cells were then stained using Giemsa and counted under a microscope. The experiment was repeated three times.

2
BioMed Research International  2.8. Transwell Invasion Assay. The invasion assay was performed using 24-well transwell chambers (costar, 8 μm pore), and the chambers were coated with matrigel (BD bioscience) at 37°C for 4 hours. Cell mixture with no FBS was plated in the top chamber. Medium supplied with 10% (v/v) FBS was placed in the lower chamber. Subsequently, the plate was incubated at 37°C for 18-24 h. Invading cells at the bottom of the transwell were stained using 0.1% hematoxylin. The invading cell number was counted using a microscope.

Cell Cycle Transition and Apoptosis.
Cell cycle transition and cell apoptosis rates were investigated after cells were modified. The modified cells were treated with 0.5% trypsin and washed with PBS butter. For the cell cycle assay, cells were fixed in 1% paraformaldehyde and stained with 5 mg/ ml propidium iodide. The apoptosis assay was performed using the BD Annexin V/FITC kit (BD, USA) and detected using a flow cytometer.
2.10. Chromatin Immunoprecipitation (ChIP) Assay. The chromatin immunoprecipitation (ChIP) assay was performed using the Magna ChIP A/G Assay Kit (Millipore, CA, USA). Briefly, cells were crosslinked with 37% formaldehyde. The DNA/protein complexes were treated using TEAD4 and IgG (Cell signaling technology) antibodies and protein A/G magnetic beads. The precipitated chromatin complexes were purified and decrosslinked at 62°C for 2 h. The precipitated DNA fragments were quantified using PCR analysis. The primers for ChIP were listed as follows: 3′; BIRC5 position2 forward, 5′CTCTCACAGCCTTCTC TTGTCA 3′; and BIRC5 position2 reverse, 5′ CACCCC GAGGTACGATCAGT 3 ′ .
2.11. Glucose Uptake. The glucose uptake assay was performed as previously reported [19,25]. Briefly, cells were washed and resuspended in PBS with 2% FBS (FBS/PBS) and treated with the 2-NBDG (Thermo Fisher Scientific, USA) at a final concentration of 10 μM for 30 minutes. Then, these cells were washed and resuspended in 300 μl 2% FBS/PBS. The fluorescence intensity was measured using a flow cytometer.   (Figure 1(a)). In 93 cases examined, 51 (54.8%) cases showed high Ajuba expression (Figures 1(b)-1(d)). Ajuba was located in the cytoplasm with nuclear staining in some cases. High Ajuba expression is positively associated with advanced TNM stage (p = 0:0009) and lymph node metasta-sis (p = 0:0007) (Table 1). Importantly, Ajuba expression in triple-negative breast carcinoma (TNBC) was higher than that in non-triple-negative breast carcinoma (p = 0:0196) ( Table 1). Additionally, we found a significant association between Ajuba status and poor patient prognosis, which was analyzed using the Log-rank test (p = 0:0179, Figure 1(e)). Oncomine data was also analyzed. The Turashvili dataset of Oncomine indicated that Ajuba mRNA was higher in invasive lobular breast carcinoma than normal lobular breast cells (p = 0:009, Figure 2(a)). The Ma breast dataset showed that Ajuba expression was higher in ductal breast carcinoma in situ (p = 0:007, Figure 2(b)). The Karnoub breast dataset showed that Ajuba in invasive ductal breast carcinoma was higher compared with that of normal breasts (p = 0:031, Figure 2(c)). TCGA data also indicated that Ajuba mRNA was higher in mixed lobular and ductal carcinoma than normal breast tissues (p = 0:038, Figure 2(d)).    (Figure 3(a)). Ajuba overexpression and knockdown were performed in MCF-7 and BT549 cell lines, respectively. Western blot showed successful transfection of Ajuba plasmid into MCF-7 cells. Ajuba knockdown significantly downregulated the endogenous Ajuba level in the BT549 cell line. Transfection efficiency was also confirmed by RT-qPCR (Figure 3(b)).

Ajuba Regulates Proliferation and Invasion in Breast
CCK8, colony formation, and invasion assays were performed to examine the biological roles of Ajuba. CCK-8 assays demonstrated that Ajuba overexpression increased while Ajuba knockdown decreased the cell growth rate (Figure 4(a)). The colony formation assay also showed that Ajuba overexpression increased the colony number while Ajuba knockdown decreased the colony number (Figure 4(b)).

BioMed Research International
Because of the positive association between high Ajuba level and nodal status in clinical samples, we examined the change of invasion after Ajuba overexpression and knockdown. As shown in Figure 4(c), Ajuba depletion inhibited invasion while Ajuba overexpression increased invading ability in breast cancer cells.

Ajuba Promotes Cell Cycle Progression and Glucose
Uptake in Breast Cancer Cells. Cell cycle analysis showed that Ajuba overexpression in MCF-7 cells increased the percentage of S phase cells and downregulated the percentage of G1 phase cells. Ajuba knockdown showed the opposite effects on BT549 cell line ( Figure 5(a)). Since glucose metabolism is pivotal in the energy production of cancer cells, we examined the possible change in glucose uptake, the critical steps during glucose metabolism. The 2-NBDG glucose uptake assay demonstrated that Ajuba knockdown inhibited the glucose uptake level while Ajuba overexpression upregulated glucose uptake ( Figure 5(b)).

Ajuba Regulates Paclitaxel Resistance in Breast Cancer
Cells. Next, we explored the role of Ajuba in chemosensitivity. MCF-7 and BT549 cells were treated with paclitaxel (10 nM for MCF-7; 25 nM for BT549, 24 hours). CCK-8 assays indicated that Ajuba knockdown upregulated the level of inhibition induced by paclitaxel while Ajuba overexpression showed the opposite effect (Figure 6(a)).
Annexin V/PI staining indicated that Ajuba knockdown upregulated the percentage of paclitaxel-induced apoptosis while Ajuba overexpression decreased the level of apoptosis, suggesting that Ajuba could induce resistance to chemotherapeutic drugs (Figure 6(b)). Ajuba could also slightly reduce apoptosis in BC cells without paclitaxel treatment.

Ajuba Positively Regulates TAZ-GLUT3/Survivin
Signaling. To elucidate the potential mechanisms of Ajuba in BC, RNA-sequencing was performed to profile the global mRNA change induced by Ajuba. GSEA revealed enrichment for Hippo signaling-related genes (Figure 7(a)). We also screened several potential proteins related to glucose metabolism and chemosensitivity using western blotting. Our results showed that Ajuba increased GLUT3 and Survivin expression at both mRNA and protein levels (Figure 7(b)). Moreover, we found that Ajuba enhanced TAZ protein expression in MCF-7 cells. Ajuba depletion downregulated TAZ protein in BT549 cells (Figure 7(b)).
TAZ, the Hippo pathway's core component, has been reported to be a transcriptional regulator of many cancerrelated genes. Analysis of the published Chip-Seq dataset indicated that both GLUT3/SLC2A3 and Survivin/BIRC5 are downstream targets of the Hippo signaling pathway. To further confirm their association, we used TAZ siRNA in MCF7 cells cotransfected with the Ajuba plasmid. RT-qPCR showed that TAZ knockdown significantly suppressed mRNA expression of GLUT3 and Survivin. TAZ depletion also ameliorated the effect of Ajuba overexpression on GLUT3/Survivin (Figure 7(c)).
TAZ has been reported as a transcription coactivator of TEAD4, which can bind promoter regions of Hippo target genes through the TEA domain. We examined if TEAD4 regulated GLUT3 through its promoter. JASPAR was used to predict the potential binding sites and weight matrix (Figure 7(d)). The chromatin immunoprecipitation assay (ChIP) demonstrated that TEAD4 could interact with GLUT3 and Survivin promoter regions (Figure 7(d)). The above result indicated that Ajuba could regulate GLUT3/ Survivin through TAZ in BC. 10 BioMed Research International subcutaneously. As shown in Figure 8, the in vivo growth rate and tumor sizes of Ajuba overexpressing MCF-7 cells were much larger than that of control cells.

Discussion
It has been recently reported that Ajuba played critical roles in a variety of human tumors. However, the involvement of Ajuba in breast cancer remains unclear. Here, we investigated the expression and tumorigenic function of Ajuba in breast cancer. Ajuba overexpression was found in 51/93 human breast cancer specimens and positively associated with TNM stage, lymph node metastasis, and poor prognosis. Interestingly, Ajuba expression was significantly enhanced in triple-negative breast cancer specimens and cell lines. The results demonstrated that Ajuba could be a promising biomarker for diagnosing aggressive breast cancer and a potential therapeutic target.
The LIM protein Ajuba acts as a scaffold participating in a diverse array of cellular processes, including cell adhesion [29,30], mitosis [31], and apoptosis [17]. Involvement of Ajuba was reported in the oncogenic processes. We confirmed that ectopically expressed Ajuba in MCF-7 cells stimulated in vivo and in vitro proliferation, invasion, cell cycle progression, and reduced apoptosis. Our data suggest that Ajuba plays an important role in the malignant biological behavior of BC.
Our results further revealed that increased Ajuba upregulated glucose uptake, making it a positive regulator of glucose metabolism. By the fact that Ajuba increased glucose uptake, our data further demonstrated that Ajuba upregulated GLUT3 expression. GLUT family proteins, which mediate glucose transport across membranes, were reported to be elevated in human cancers [19,32]. It has been reported that GLUT3 mediated the growth and survival of breast cancer cells [33,34]. Survivin/BIRC5 is a member of the inhibitor of apoptosis (IAP) family. Survivin localizes in the mitochondria and inhibits apoptosis. Reports indicate that Survivin protects cancer cells from drug-induced cell death [35]. Many studies indicated that aberrant expression of Survivin is associated with poor prognosis and drug/radiation resistance in breast cancers [36][37][38]. Strategies targeting survivin to treat breast cancer have got promising initial results [39]. Thus, Ajuba might regulate proliferation and chemosensitivity through GLUT3 and Survivin. We also found that Ajuba upregulated TAZ protein expression. It has been reported that Ajuba could sequester the Hippo kinase complex and sustain cell proliferation by limiting YAP inhibition [14]. We also checked YAP protein levels in the current study but did not found significant change after Ajuba transfection or knockdown. TAZ has been reported to control genes regulating glucose metabolism and apoptosis [40]. TAZ is a transcription coactivator which does not directly control gene transcription. Instead, it interacts with transcription factor TEAD3/4, which directly interacts with the promoter region of target genes [41]. TAZ forms a complex with TEAD4 transcription factor to activate downstream gene transcription [42]. Using TAZ siRNA, we showed that TAZ mediated the upregulating effect of Ajuba on GLUT3/Survivin, which was also supported by ChIP results showing that TEAD4 could bind to the GLUT3/Survivin promoter region. Together, our results demonstrated a strong link among Ajuba, TAZ, GLUT3, and Survivin.
Taken together, the current study indicated that Ajuba overexpression promoted proliferation, invasion, chemoresistance, and glucose uptake in breast cancer cells and correlated with poor patient prognosis. The present study also linked its oncogenic role of Ajuba with TAZ-GLUT3/Survivin signaling, indicating the therapeutic possibility by targeting Ajuba in BC.

Data Availability
The data that support the findings of this study are available on request from the corresponding author.

Conflicts of Interest
We declare that we have no conflict of interest.