Elevated CREPT Expression Enhances the Progression of Salivary Gland Adenoid Cystic Carcinoma

The elevated expression of Cell cycle-Related and Expression-elevated Protein in Tumor (CREPT) is reported to promote the growth of several tumors by enhancing Wnt/β-catenin signaling and cell cycle. However, the relevance of CREPT to the malignancy of salivary gland adenoid cystic carcinoma (SACC) remains unclear. The samples from 51 SACC patients were exploited in this study. We found that SACC samples exhibited a noticeably robuster CREPT expression than the para-cancerous tissues. Statistical analysis suggested that CREPT expression was significantly correlated with the T classification of SACC. To upor down-regulating CREPT expression, the specific shRNA or full length of CREPT was delivered into SACC cell lines to examine the cell proliferation, migration, colony formation and implanted xenograft survival. Western blot assay and immunohistochemistry were applied to evaluate the expression of CREPT, cyclin D1, c-Myc and CDK4. Up-regulated CREPT in the low metastatic SACC line significantly promoted proliferation and colony formation, as well as cyclin D1, c-Myc and CDK4 expression. While knocking down of CREPT in the high metastatic SACC line remarkably reduced above effects. Furthermore, the SACC xenograft in mice confirmed that down-regulation of CREPT inhibited the in vivo tumor growth. Our study indicated that the elevated CREPT expression promoted the cell proliferation and tumor size of SACC by enhancing the expression of cyclin D1, c-Myc and CDK4, suggesting that CREPT contributed to SACC progression by stimulating cell proliferation, and might act as a potential target in future SACC therapy.


Introduction
Salivary adenoid cystic carcinoma (SACC) arises from the secretory epithelial cells of the salivary glands. It accounts for approximately 25% of malignant tumors in the major salivary glands and 50% in the minor glands [1][2][3][4][5] . SACC is characterized by the slow but relentless growth, the nerve and blood vessel invasion, and the lung metastasis with a high incidence 6) . However, the most unfortunate aspect of SACC is the poor prognosis. The overall survival rates at 5, 10 and 15 years after SACC diagnosis are 71%, 54% and 34%, respectively 4) . While the survival rate in the patients with SACC progression even decreases to 35% after 5 years, 15% after 10 years, and almost 0% after 15 years 6) , which is primarily attributed to the late distant lung metastasis 7) . Therefore, it is imperative to identify an associated gene to predict the prognosis of SACC to guide individualized treatment.
Since the first report at 2012, CREPT has been proven as an accelerator in multiple tumors 8) . Known as the regulator of nuclear pre-mRNA domain containing protein 1B (RPRD1B), CREPT fundamentally accelerates cell cycle by activating cyclin D1, cyclin E, CDK2, CDK4 and CKD6 expression via the direct binding to their promoters. Moreover, CREPT facilitates transcription by changing DNA configuration into loops, which promotes RNA polymerase recycle to inhibit transcription termination [8][9][10][11] . The phosphorylation of the Ser145 in a conserved motif of CREPT by Aurora B is suggested to be critical to accelerate the G2/ M transition by activating the transcription of Cyclin B1 12) . Further studies revealed that CREPT could also activated cyclin D1 and c-Myc by stabilizing β-catenin/TCF4 complex on their promoters 13) . Latest studies found that CREPT not only increased the acetylation and stabilization of β-catenin through the cooperation with p300, but also elevated the active histone acetylation and decreased the repressive histone at the promoters of Wnt target genes 14) . Additionally, the promotion on cell cycle by CREPT is also reported in peripheral T cells and keratinocyte 15) . Consistently, by targeting CREPT, microRNA-300 (miR-300) was capable of inhibiting the proliferation of liver cancer cells, as well as suppressing the Wnt/β-catenin signaling 16) . Thus, CREPT is implicated to promote cell cycle via Wnt signaling pathway.
There are the increasing amount of studies suggesting that the high level expression of CREPT could be considered as an indicator for the poor prognosis of colorectal cancer, non-small cell lung cancer and gastric cancer [17][18][19] . On the other hand, the silence of CREPT induces cell cycle arrest at G0/Gl phase, which results in the decreased Cyclin D1 and CDK4 expression, and increased p53 and p21 transcription 20) . Interestingly, the suppressed CREPT level reduced the migration of gastric cancer cells by altering the expression of extracellular matrices, such as E-cadherin, N-cadherin, vimentin and MMP1 20) . Therefore, CREPT is suggested to act as an oncogene in tumorogenesis. However, if the CREPT expression could fulfill the identical role in the prognosis of SACC is still required to be elucidated.

Ethics statements
This study was approved by the Ethical Committee of Chinese PLA (People's Liberation Army) General Hospital and conducted in accordance with the guidelines of the Declaration of Helsinki. All of the patients participated in this study have been informed of the nature and purpose of the research and agreed to use their medical records and tissue specimens for research by signing the written consent. The follow-up data were collected by directly interviewing with or calling the patients or their relatives. The animal experiments were approved by the Animal Research Committee at Beijing Laboratory Animal Research Center and carried out in accordance with the authorized Protocol (P2015015).

Samples collection
All of the 51 samples of salivary adenoid cystic carcinomas (SACCs) were collected from the patients in Chinese PLA general Hospital from 2006 to 2010. All the SACC samples were originated from the parotid, sub-lingual and sub-mandibular glands, and the para-cancerous tissues (PCT) were also collected as the controls. None of the patients received preoperative therapy before surgical resection. The SACC diagnosis was histopathologically confirmed and staged according to the 2009 UICC-TNM Classification of Malignant Tumors. The demographic and clinic-pathological information for each patient is listed in Table 1.

Immunohistochemistry
The preparation of paraffin-embedded sections, the reagents and procedure of immunohistochemistry followed the protocol previously described 21) . The anti-CREPT antibody (gifted by Pro. Zhijie Chang at the State Key Laboratory of Biomembrane and Membrane Biotechnology in Tsinghua University) was used as primary antibody in the dilution of 1:60 10, 13) . For the mouse xenografts, the primary antibody against Ki67 was diluted in 1:500 (Wuhan Boster Biological Technology, Ltd., Wuhan, China). The secondary antibody was the HRP-conjugated anti-rabbit/ mouse IgG (MXB Biotechnologies Inc. Fujian, China). The DAB substrate kit (MXB Biotechnologies Inc. Fujian, China) was applied for color development. Hematoxylin was used for counter-staining.

Cell culture
The SACC cell line, SACC-83 was established from a patient with SACC in the sub-lingual gland. The SACC-LM cell line was derived from the in vivo selection of the lung metastatic foci from SACC-83, which has a higher rate of lung metastasis compared to SACC-83 23,24) . Both cell lines were gifted by Dr. Sheng-Lin Li of Peking University School and Hospital of Stomatology, Beijing, China. Cells were cultured in RPMI 1640 medium (GIBCO Co., Carlsbad, CA, USA) containing 10% fetal bovine serum (FBS, GIBCO Co., Carlsbad, CA, USA) at 37ºC in 5% CO 2 atmosphere.

Stable transfection
SACC-83 was transfected with the lentivirus vector pLKO-puro encoding CREPT or lentivirus vector pLKO-puro (con). SACC-LM was transfected with the lentivirus vector pLKO-puro encoding shCREPT1/2 or sh-control (con) (Lentivirus vector pLKO-puro encoding CREPT, shCREPT1/2 or shcontrol is constructed by Pro. Zhijie Chang at the State Key Laboratory of Biomembrane and Membrane Biotechnology in Tsinghua University). The shRNA sequences have been described previously 21) . SACC-83 and SACC-LM grew in 24-well plates to 80% confluence, and then, medium containing CREPT, shCREPT1/2 or sh-control lentiviral particles supplied with Polybrene (8 μg/ml, sc-134220, Santa Cruz Ltd., CA, USA) was added to these cells, respectively. The medium was refreshed after 12h transfection, and the stable clones were selected by Puromycine dihydrochloride (1 μg/ml, sigma) after 72 h of transfection. The CREPT expression in the selected clones was assessed by Western blot as described above after 3 weeks of selection to confirm the successful transfection.

Cell proliferation assay
SACC cell lines were seeded at 5×10 3 cells/well with complete medium. Cell Counting Kit-8 (CCK-8) was used to assess cell proliferation rate as instructed in the manufacturer's protocol (Dojindo Ltd., Kyushu, Japan). The absorption peak at 450 nm was measured by microplate reader (Thermo Multiskan MK3 Co., MA, USA). In the mouse xenografts, the percentages of cell proliferation were determined by the numbers of Ki67 positive nuclei to the numbers of the total nuclei.

Colony formation assay
1x10 3 cells were plated into a well of the 6-well plate with complete medium. After 2 weeks, the cells were fixed by methanol and stained by crystal violet to count the colony numbers.

Flow cytometry
The percentages of cells in G1, S and G2 phases were calibrated by flow cytometry. SACC cells were digested with trypsin, washed twice with phosphate buffer saline (PBS) and then, fixed with cold 70% ethanol at 4℃ overnight. The fixed SACC cells were incubated with RNase A for 1 h at 37℃, and stained with propidium iodide (PI) for 30 mins (Cell cycle and Apoptosis Analysis Kit, Beyotime Biotechnology Ltd., Dalian, China). The flow cytometer (BD FACS Calibur™) was utilized for cell cycle analysis as the manufacturer instructed (Becton, Dickinson and Company, USA).

SACC xenografts
All animal experiments were approved by the Animal Research Committee at Beijing Laboratory Animal Research Center and were carried out in accordance with established International Guiding Principles for Animal Research (Protocol NO. P2015015). SACC-LM-con and SACC-LM-shCREPT2 treated cells (5×10 6 cells /mouse) were subcutaneously injected into the right flank of to 5-week old female athymic nude mice. Twenty mice were assigned in each group. After 28 days, the SACC xenografts were harvested for immunohistochemistry assay.

Statistical analysis
The data were input into SPSS 17.0 software (IBM, NY, USA) for Mann-Whitney U test, the Student's t-test, the X 2 test and the Fisher's exact test. The statistical difference was though significant when p value is less than 0.05.

The implicated correlation of CREPT level to SACC malignancy
To verify if the CREPT expression is elevated in SACC as in other tumors, the CREPT expression in 51 paraffin-embedded SACC and para-cancerous samples were evaluated by immunohistochemical staining (Fig. 1A; Table 1). The immunohistochemical scores ranged from 0 to 10.8 in the SACC samples (median=5.294), but from 0 to 2.6 in the normal counterparts (median=0.710). The average CREPT expression in SACC samples was obviously higher than that in the normal counterparts, as the statistical significance confirmed (Fig. 1B). Thus, we further analyzed the relationship between clinicopathologic characteristics and CREPT expression levels in SACC individuals (Table 1). Although there was no significant correlation of CREPT levels to patients' age, gender or SACC stage, the elevated expression level of CREPT was markedly associated with the larger tumor size, hence a higher T classification. To further verify the correlation between CREPT level and malignancy of SACC, the CREPT expression was analyzed in the SACC-83 and the SACC-LM, respectively. Western blotting showed that the CREPT level in SACC-83 was significantly lower than that in SACC-LM, implicating the correlation of CREPT level to metastasis (Fig. 1C, D).

Up-regulating CREPT enhanced cell proliferation, colony formation and cell cycle in SACC
To confirm the oncogenic role of CREPT in SACC, the lower CREPT-expressing SACC cell line, SACC-83 was selected for the gainof-function assay. Stable transfection and expression of CREPT in SACC-83 cells were accomplished by infection of lentivirus carrying CREPT encoding sequence ( Fig. 2A). As expected, the capabilities of cell proliferation and colony formation were both significantly elevated in the CREPT transfected group (Fig. 2B, C). Moreover, the cell cycle assay demonstrated that the increased CREPT expression attracted more cells into S phase, suggesting the promotion of DNA synthesis by CREPT (Fig. 2D). Coincided with the cell cycle distribution, the expression of cyclin D1, c-myc and CDK4, which were associated with S phase, were also dramatically increased by CREPT transfection (Fig.  2E).

Down-regulated CREPT expression suppressed proliferation, colony formation and cell cycle in SACC
To perform the loss-of-function assay, the SACC-LM cell line, which possessed a high CREPT expression, was transfected with shCREPT1/2. Western blot analysis revealed that the protein levels of CREPT in shCREPT1/2 transfected cells were significantly lower than those in the control cells (Fig. 3A). The proliferation capability of shCREPT1/2-transfected SACC-LM cells decreases approximately 39% and 50%, respectively, after 6 days (Fig. 3B). Colony formation assay was also exhibited an approximately 3.1 and 3.78 folds decrease in the colony number of CREPT-shRNA1/2 transfected cells compared with the SACC-LM and shcon-transfected SACC-LM, respectively (Fig.  3C). Cell cycle assay disclosed that the suppression on CREPT by shCREPT transfection arrested cell cycle in G0-G1 phase, though the G2-M phase seemed no impact (Fig. 3D). Moreover, the obviously decreased ratio in S phase implicated that the suppression of CREPT was correlated with the reduced DNA synthesis (Fig. 3D). Consistent with the results in the gain-of-function assay, western blot assay showed that knocking down the endogenous CREPT expression by shCREPT1/2 transfection in SACC-LM also decreased the amounts of cyclin D1, c-Myc and CDK4 proteins noticeably (Fig. 3E), suggesting that the promoted cell cycle or DNA synthesis by CREPT through cyclin D1, c-Myc and CDK4 .

Suppression of CREPT reduced SACC xenografts' volume and weight in vivo
To verify the oncogenic role of CREPT in vivo, the shCREPT2-transfected SACC-LM cells were injected into nude mice to generate SACC  xenografts. After 28 days of implantation, the volumes of shCERPT2 treated SACC-LM xenografts were obviously smaller than those in shcon treated SACC-LM xenografts (Fig. 4A). Consistent with the reduced cell proliferation by repressing CREPT, the average weight of shCERPT2-transfected SACC-LM xenografts was found to be 0.2215 g, which is almost 7.22-fold less than the average weight of control xenografts (Fig. 4B). The growing curvature revealed that the shcon treated SACC xenografts started the rapid growth from day 15, while the shCERPT2 treated SACC xenografts showed no evident increase in growth rate during the entire period (Fig. 4C). The histological and statistical assay of cell proliferation indicated that Ki67 was less activated in the shCREPT2 treatment SACC xenografts (Fig. 4D, E), which attributed the reduced volume and weight of SACC xenografts to the suppressed cell proliferation by the down-regulated CREPT.

Discussion
In this study, we reported that CREPT plays an oncogenic role in the tumorigenesis of SACC by enhancing cell proliferation, colony formation and the expression of cyclin D1, c-Myc and CDK4. Consistently, knocking down CREPT expression significantly reduced not only the cell proliferation, colony formation and the expression of cyclin D1, c-Myc and CDK4 in SACC cell lines, but also the size and weight of SACC xenografts. Surprisingly, when CREPT was over-expressed, the increased ratio of SACC in S-phase provided a new clue for the enhancement of CREPT on cell proliferation, and SACC volume and weight. Since cyclin D1 plays an oncogenic role in multiple cancers by controlling the G1/S transition [25][26][27] , CREPT was suggested to promote SACC cells into S phase and DNA synthesis by activating cyclin D1. This speculation was supported by the latest finds that the over-expression of CREPT made the colorectal cancer more sensitive to fluoroura-cil, an uracil analog inhibiting DNA synthesis and thus inducing apoptosis 28) . Additionally, c-myc was also regarded as an oncogene which elevation markedly increased the metastasis of tumors [29][30][31][32] . Therefore, it indicated that CREPT played the oncogenic role upstream to the cyclin D1 and c-myc, which suggested that CREPT might act as a potential target for the future SACC therapy.
Initially, we hypothesized that CREPT might work as the standard for the diagnosis of SACC. However, the statistical results of CREPT expression in the clinical samples denied this hypothesis because there were 3 cases with negative CREPT expression in all 51 cases. Although the CREPT expression in the metastasized SACC was significantly higher than the local SACC, the correlation between CREPT expression and the SACC metastasis was rejected because CREPT seemed no contribution to the lymph-node and distant metastasis in SACC patients. Thus, the only correlation between CREPT expression and the malignancy of SACC was implicated to be the promotion on SACC proliferation. This hypothesis was first verified by the higher level of CREPT in SACC-LM. Second, CERPT was definitely correlated to SACC proliferation. Suppression on CREPT slowed down not only the SACC proliferation and cell cycle ex vivo, but also the volume and weight of SACC in vivo. Third, the up-regulated CREPT expression was closely associated with the poor differentiation and clinical stage of many malignant tumors 8,11,18) . Our previous study also reported that the CERPT expression was hard to serve as a hallmark of oral squamous cell carcinoma (OSCC), it was strongly associated with the malignancy of OSCC by promoting proliferation 21) .
In summary, our study demonstrates that CREPT expression enhanced the progress of SACC ex vivo and in vivo by up-regulating the expression of cyclin D1, c-Myc and CDK4, which promoted the clinical malignancy of SACC. Therefore, CREPT was supposed to work as a potential indicator for SACC prognosis and a target for anticancer medicine.