Cytosolic THUMPD1 promotes breast cancer cells invasion and metastasis via the AKT-GSK3-Snail pathway

Human THUMP domain-containing protein 1 (THUMPD1) is a specific adaptor protein that modulates tRNA acetylation through interaction with NAT10. Immunohistochemical analysis of 146 breast cancer specimens (82 triple-negative and 64 non-triple-negative cases) indicated THUMPD1 expression is higher in breast cancer tissues (60.9%, 89/146) than normal breast tissues (28.3%, 15/53; p < 0.001). Overall and cytosolic, but not nuclear, THUMPD1 expression in breast cancer correlated with advanced TNM stage (p = 0.003 and p < 0.001, respectively), lymph node metastasis (p = 0.001 and p < 0.001, respectively), and poor patient prognosis (p = 0.001 and p < 0.001, respectively). THUMPD1 interacted and co-localized with YAP, but did not affect Hippo pathway activity. THUMPD1 overexpression enhanced breast cancer cells invasion and migration in vivo and in vitro, possibly through activation of AKT, GSK3β and Snail, and inhibition of E-cadherin. Treatment with the AKT inhibitor, LY294002, reduced the effects of THUMPD1 overexpression in breast cancer cells. These results indicate that THUMPD1 promotes breast cancer cells invasion and migration via the AKT-GSK3β-Snail pathway.


INTRODUCTION
Breast cancer is the leading cause of cancer death in women [1,2]. Despite advances in tumor diagnostic and therapeutic strategies, patient outcomes are still poor due in part to cancer cell metastasis and heterogeneity [3,4]. Novel biomarkers are needed to more accurately identify high-risk patients and to predict disease prognosis [5]. Genes abnormally expressed during tumor progression and metastasis may be used as markers to provide prognostic information beyond standard clinical assessment [6,7].
Human THUMP domain-containing protein 1 (THUMPD1) is a specific adaptor protein that interacts with NAT10, a human acetyltransferase involved in histone and microtubule modification, and thus modulates tRNA acetylation [8]. Havugimana, et al. [9] suggested that THUMPD1 interacted with Yes-associated protein (YAP), a major transcriptional co-activator of the Hippo pathway, which plays crucial roles in tumor proliferation and invasion [10][11][12][13]. Our preliminary experiments showed that THUMPD1 expression was elevated in breast carcinoma samples as compared to normal breast tissues. However, THUMPD1 downstream signaling pathways in human breast cancer are as yet largely unknown. We hypothesized that THUMPD1 may promote malignant transformation through its interaction with YAP.
To test this hypothesis, we examined THUMPD1 expression and localization in breast cancer tissues and investigated associations between THUMPD1 subcellular localization and patient clinicopathological factors. We also addressed the effects of THUMPD1 on cancer cell migration and invasion, and analyzed potential downstream signaling pathways.

THUMPD1 expression and subcellular localization in breast carcinoma
We performed immunohistochemistry to assess THUMPD1 expression in 146 breast cancer and 53 paired noncancerous specimens. THUMPD1 presented
We then assessed THUMPD1 expression and subcellular localization in cell lines through western blotting and immunofluorescence. All breast cancer cell lines were positive for THUMPD1 expression, whereas the normal breast ductal cell line, MCF-10A, showed only weak expression (Figure 2A). Breast cancer cells exhibited both cytosolic and nuclear THUMPD1 expression, with only nuclear expression in MCF-10A cells ( Figure 2B). In MCF-7 cells transfected with the THUMPD1-myc expression plasmid, exogenous THUMPD1 was largely localized in the cytoplasm ( Figure 2C). THUMPD1 subcellular localization as shown by immunohistochemistry. In normal breast ductal cells, THUMPD1 expression was either absent (A) or weak (B). In carcinoma in situ (C) and IDC cells (D), THUMPD1 was observed in the nucleus and cytoplasm at moderate and high levels, respectively. In some IDC specimens, THUMPD1 was exclusively localized in the cytoplasm (E). THUMPD1 cytosolic and nuclear expression was higher in IDC (F) than in normal breast ductal cells. Magnification, ×200 and ×400. Kaplan-Meier analysis demonstrated that patient overall survival negatively correlated with overall (G) and cytosolic (H), but not nuclear (I), THUMPD1 expression.

Interaction between THUMPD1 and YAP
Immunoprecipitation and immunofluorescence were performed using THUMPD1-overexpressing MCF-7 cells to investigate the interaction between THUMPD1 and YAP. Exogenous THUMPD1 interacted directly with YAP ( Figure 3A), and expression was co-localized in both the cytoplasm and nucleus ( Figure 3B).
We assessed YAP phosphorylation status and distribution in THUMPD1-overexpressing MCF-7 and MDA-MB-468 cells, and in THUMPD1-deficient MCF-7 and BT-549 cells. Neither overexpression nor depletion of THUMPD1 affected levels of YAP, p-YAP, or the upstream regulator, LATS and p-LATS1 ( Figure 3C). YAP and p-YAP subcellular distributions ( Figure 3D) were also unchanged. Therefore, THUMPD1, despite its interaction with YAP, may not influence Hippo signaling.

DISCUSSION
The present study showed that THUMPD1 is weakly expressed in normal breast cell nuclei, and strongly expressed in IDC cell nuclei. THUMPD1 is also expressed in the cytoplasm of IDC cells, and cytosolic THUMPD1 expression positively correlated with high TNM stage, lymph node metastasis, and poor patient prognosis. In cultured breast cancer cells, endogenous THUMPD1 also localized to both the cytoplasm and nucleus. Interestingly, exogenous THUMPD1 was found in the cytoplasm rather than the nucleus ( Figures 1I and 3D), although the reasons for this are unclear, and must be addressed in future studies. We also examined THUMPD1 expression in both triple-negative and non-triple-negative breast cancers, and found no correlation between THUMPD1 distribution or expression and breast cancer type.
A previous study suggested that THUMPD1 might interact with the transcriptional regulator, YAP, a potential oncogene and a member of the Hippo signaling pathway [9]. Consistent with this, our results showed that THUMPD1 interacted and co-localized with YAP in both the cytoplasm and nucleus. However, changes in THUMPD1 expression had no obvious effects on Hippo pathway activity or YAP subcellular distribution. Nucleocytoplasmic shuttling of YAP may be responsible for THUMPD1 translocation; however, this hypothesis needs further investigation [10].
Sharma, et al. demonstrated that THUMPD1 binds NAT10, which is a biomarker in several cancer types [8]. Ma, et al. indicated that NAT10 upregulation promotes hepatocellular carcinoma invasion by decreasing E-cadherin [34]. Zhang, et al. also confirmed that NAT10 enhanced colorectal cancer invasion and correlated with poor prognosis. We speculate that THUMPD1-NAT10 binding could accelerate cancer cell invasion cooperatively. Alternatively, THUMPD1 may promote tumor invasion as a NAT10 downstream factor.
In conclusion, we observed THUMPD1 overexpression in the cytoplasm of breast cancer cells, which positively correlated with high TNM stage, lymph node metastasis, and poor patient prognosis. Although THUMPD1 interacted with YAP, no effects on other Hippo signaling pathway members were identified. THUMPD1 promoted breast cancer cells invasion and migration, and downregulated E-cadherin via the AKT-GSK3β-Snail

Patients and clinical specimens
The study protocol was approved by the institutional review board of China Medical University. All participants provided written informed consent, and the study was conducted according to the principles expressed in the Declaration of Helsinki. Primary tumor specimens were obtained from 146 breast cancer patients, including 82 triple-negative (deficient in estrogen receptor, progesterone receptor, and Her2/neu expression) and 64 non-triple-negative tumors. All patients diagnosed with invasive ductal carcinoma (IDC) underwent complete surgical resection at the Affiliated Cancer Hospital of China Medical University between 2001 and 2003. Among the 146 IDC specimens, there were 43 cases of ductal carcinoma in situ (DCIS). Complete follow-up data were available for all 146 analyzed cases. Patient survival was defined as the time from the day of surgery to the end of the follow-up period or the day of death due to recurrence or metastasis. None of the patients had received radiotherapy or chemotherapy before undergoing surgical Oncotarget 13364 www.impactjournals.com/oncotarget resection, and all patients were treated with routine chemotherapy after surgery.

Immunohistochemistry
All tissue specimens were fixed in neutral formaldehyde, embedded in paraffin, and sectioned (thickness, 4 μm). The streptavidin-peroxidase immunohistochemical method was used to improve staining. Tissue sections were incubated at 4°C overnight with THUMPD1 mouse monoclonal antibody (1:50 dilution; Santa Cruz Biotechnology, Inc., Dallas, TX, USA); phosphate-buffered saline was used as a blank control. Sections were then incubated with biotin-labeled secondary antibodies (Ultrasensitive; MaiXin, Fuzhou, China) at 37°C for 30 min, followed by diaminobenzidine for coloration.

Plasmid construction and cell transfection
The pCMV6-DDK-Myc empty vector and the pCMV6-DDK-Myc-THUMPD1 vector were purchased from OriGene (Rockville, MD, USA). THUMPD1-siRNA (sc-93083) and control siRNA (sc-37007) were purchased from Santa Cruz Biotechnology. Cells were transfected using the Lipofectamine 3000 kit (Invitrogen) according to the manufacturer's instructions.

Matrigel invasion assay
Cell invasion assays were performed using 24-well Transwell chambers with 8-μm pores (Costar, Cambridge, MA, USA). Inserts were coated with 20μl Matrigel (1:3 dilution; BD Bioscience, San Jose, CA, USA). Cells were trypsinized 48 h after transfection, resuspended at 3 × 10 5 cells in 100 μL of serum-free medium, and transferred to the upper chambers of Transwell plates. Fetal bovine serum (10%) was added to the lower chambers as chemoattractant. After 18 h incubation, cells that passed through the filter were fixed with 4% paraformaldehyde, stained with hematoxylin, and counted under a microscope in 10 randomly selected fields at ×40 magnification.

Wound healing assay
Wounds were created in cell monolayers at < 90% confluence 48 h after transfection, using a 200-μl pipette tip. Cell migration into the wound was observed at different time points. Wound areas were measured using Image J software, and representative images were taken. www.impactjournals.com/oncotarget Each experimental condition was analyzed in duplicate, and three independent experiments were performed.

Transplantation of tumor cells into nude mice
Animals used in this study were treated according to the National Institutes of Health guide for the care and use of laboratory animals (NIH Publications No. 8023, revised 1978). Four-week-old female BALB/c nude mice were purchased from Slac (Shanghai, China) and kept in a laminar-flow cabinet under specific pathogenfree conditions for two weeks before use. Each mouse was then inoculated intravenously (tail vein) with 2 × 10 6 THUMPD1-transfected MCF-7 tumor cells in 0.2 mL sterile PBS. Six weeks after inoculation, mice were euthanized and examined for tumor growth and dissemination. Tumors, hearts, livers, lungs, and kidneys were dissected, fixed in 4% formaldehyde (Sigma) and embedded in paraffin. Serial 6-µm-thick sections were cut, stained with hematoxylin and eosin, and examined by microscopy.

Statistical analyses
All statistical analyses were performed using SPSS 22.0 software (SPSS Statistics, Inc., Chicago, IL, USA). Immunohistochemistry results were analyzed via chisquare and Spearman's rank correlation tests. Kaplan-Meier survival analysis results were compared using the log-rank test. The Cox regression model was used to test prognostic values. All clinicopathological parameters were included in the Cox regression model and tested by univariate and multivariate analysis according to the enter method. Differences between groups were compared using Student t-tests, and p < 0.05 was considered statistically significant.