TOB1 suppresses proliferation in K‐Ras wild‐type pancreatic cancer

Abstract TOB1 participates in various kinds of cancers. However, its role in pancreatic cancer has rarely been reported. In this study, we explored the expression and mechanisms of TOB1 in regulating the malignant phenotype of pancreatic cancer cells. TOB1 expression was determined by data mining and immunohistochemistry (IHC), and its localization was observed by immunofluorescence. CCK‐8 cell proliferation, colony formation, flow cytometric, transwell migration, and Western blot (WB) assays were used to examine how it impacts the malignant phenotype of pancreatic cancer. Furthermore, Foxa2 binding to TOB1 was tested by dual‐luciferase reporter assays, and RNA‐Seq was performed to identify signaling pathways. We found TOB1 was downregulated in pancreatic cancer tissues and was mainly located in the cytoplasm. TOB1 overexpression reduced the proliferation of K‐Ras wild‐type pancreatic cancer cells but made no difference to cell migration and invasion. Foxa2 overexpression significantly enhanced TOB1 promoter activity. Moreover, overexpressing TOB1 substantially enriched the calcium pathway in K‐Ras wild‐type pancreatic cancer cells. In conclusion, TOB1 may suppress the proliferation of K‐Ras wild‐type pancreatic cancer cells by regulating calcium pathway genes.


| INTRODUCTION
Pancreatic cancer ranks fourth among cancer-related deaths in the United States, 1 and will be second in 2030. 2 Over the past decades, pancreatic cancer has also been in the top 10 of all malignant tumors in China. 3 Due to a lack of early symptoms, over half patients had a 5-year survival of 3%. 1 Pancreaticoduodenectomy is a potentially curative treatment, but less than 20% of patients have the opportunity to undergo surgery. 4 Pancreatic cancer remains a challenging, noncurable disease. Therefore, there is a pressing need for identification of effective treatments based on deeper exploration of the potential mechanisms of this cancer to improve survival.
The regulation of cell proliferation is crucial to the activation of important cellular processes. Cell hyperproliferation is a significant biological feature of tumors. 5 In the early 1990s, the Tob/BTG antiproliferative (APRO) protein family was found, which could negatively control tissue growth and development. 6 TOB1 was first found to interact with ErbB-2 (also known as HER-2) in a human breast cell line. 7 Research has shown that TOB1 has vital roles in embryogenesis, activation of T cells, bone formation, and metabolism. 8 The roles of TOB1 in cancer became apparent when mice lacking TOB1 were observed to spontaneously develop tumors. 9 Subsequent studies reported that reduced TOB1 expression or inactivation by phosphorylation contributed to carcinogenesis and tumor growth. [10][11][12] Studies have indicated that regulation of nuclear localization of TOB1 also affects its antiproliferative activity, 13,14 but this conclusion is still unclear. Other reports revealed that TOB1 participated in regulating the proliferation, invasion, metastasis, and resistance of some cancers. [15][16][17] Furthermore, TOB1 may predict the prognosis of node-negative breast cancer. 18 Accumulating evidence has shown that TOB1 functions as a tumor suppressor. However, few research have examined the expression and biological role of TOB1 in pancreatic cancer. In this study, we found that TOB1 is mainly located in the cytoplasm. Low expression of TOB1 was detected in pancreatic cancer tissues and inversely correlated with tumor size. TOB1 overexpression prevented the cell transition from G 1 phase to S phase, and caused the proliferation of K-Ras wild-type pancreatic cancer cells was inhibited, at least in part through the calcium pathway. Furthermore, a luciferase reporter assay indicated that Foxa2, as an upstream transcription factor, regulates TOB1 expression.

| Clinical samples
We purchased two pancreatic cancer tissue microarrays from Shanghai Outdo Biotech: HPanA030PG02 contains 15 cases of pancreatic ductal adenocarcinoma (PDAC) and their adjacent noncancerous tissues (NCTs) with one point for each tissue; and HPan-Ade180Sur-02 contains 80 cases of matched cancer/paracancerous samples and 20 cases of unpaired cancer tissues with one point for each tissue. All patients in HPan-Ade180Sur-02 were followed up from 1 to 87 months after the operation, and all specimens involved were determined by HE staining.

| Cell lines and culture
The human pancreatic cancer cell lines BxPC-3 and Patu8988-t were cultured with RPMI-1640 medium (HyClone) containing 10% fetal bovine serum (FBS, Gibco) at 37°C in 5% CO 2 . All cell lines were authenticated by the China Center for Type Culture Collection (CCTCC) using short tandem repeat (STR) analysis.

| Oncomine database analysis
Oncomine (http://www.oncom ine.org) was utilized to analyze the mRNA expression difference of TOB1 in pancreatic cancer and normal pancreatic tissues. The thresholds were set as follows: gene: TOB1; analysis type: cancer vs normal analysis; data type: mRNA; sample type: clinical specimen; P value: .05; fold change: 2; gene rank: all. The original expression data were collected to draw a box plot.

| Lentivirus transfection
All lentiviruses we used were purchased from Shanghai Genechem. First, cells were transduced at a multiplicity of infection (MOI) value of 10, 20, 50, or 100 under different transfection conditions for the preliminary experiments. The results showed that cells in the transfected group with MOI values of 10 and 20 did not show green fluorescence, while those in the group with values of 50 and 100 had stronger fluorescence ( Figure S1). We chose an MOI value of 50 for the formal experiment. BxPC-3 and Patu8988-t cells were seeded at 1.5 × 10 5 | 1505 cells/well and 0.5 × 10 5 cells/well into 6-well plates, respectively. Cells were transduced at an MOI value of 50. Viruses and cells were incubated in enhanced infection solution with a final volume of 1 mL. Twenty-four hours later, viruses were removed, cells were rinsed twice with phosphate buffered saline (PBS), and fresh RPMI-1640 medium supplemented with 10% FBS was added. Puromycin (2.5 µg/mL for BxPC-3 and 5 µg/mL for Patu8988-t) was added at 72 hours posttransfection, and cells were cultured under this condition for a month. Cells of the negative control group were established under the same conditions. The efficiency of transfection was verified using a fluorescence microscope and Western blot (WB) analysis. After verification, puromycin was reduced to half of the selection dose for long-term culture.

| Immunofluorescence
Cells were seeded at 5 × 10 4 cells/well into a Millicell EZ SLIDE (Millipore). The next day, cells were washed twice with PBS, each time for 2 minutes, before being fixed with 4% paraformaldehyde (PFA) for 30 minutes, followed by washing with PBS for 15 minutes. Then, the cells were treated with 0.3% Triton X-100 PBS for 30 minutes, washed with PBS for 15 minutes, and incubated with 5% BSA for 30 minutes. Anti-TOB1 antibody was diluted 1:50 at 4°C overnight. After the cells were washed with PBS three times for 5 minutes each, in the dark conditions, secondary antibodies conjugated with Alexa Fluor 594 goat anti-rabbit IgG (Invitrogen) were loaded into wells at room temperature (RT) for 60 minutes. Cells were washed as before, followed by DAPI (Beyotime) for 3 minutes. Prior to the end of the immunostaining process, PBS was used to wash the cells again. Immunofluorescence images were harvested with a laser scanning confocal microscope (Lecia).

| Cell proliferation assay
The Cell Counting Kit-8 assay (Dojindo) was used to measure the proliferation of different cell lines according to the manufacturer's instructions. 3000 BxPC-3 cells and 2000 Patu8988-t cells were plated in 100-μL medium in 96-well plates in quintuplicate. Cells were incubated in normal culture medium containing 10% CCK-8 for 2 hours at 37℃ in 5% CO 2 . The optical density values were determined by reading the absorbance at 450 nm.

| Colony formation assay
Patu8988-t and BxPC-3 were seeded in 6-well plates at a density of 500/well and 1000/well, respectively. About 7-10 days later, after fixing with 4% PFA for 30 minutes, we stained cells with 0.05% crystal violet for 20 minutes at RT. Finally, the number of colonies was counted, with a colony of at least 50 cells counted as one colony.

| Cell cycle analysis
Single-cell suspensions were collected. After washing twice with cold PBS, we fixed cells with cold 75% ethanol and stored them at −20°C for over 24 hours. Cells were washed with cold PBS twice, then 1 × 10 6 cells were resuspended in 1-mL PI/RNase staining solution (BD Biosciences), and finally analyzed by the FACSCalibur flow cytometer (BD Biosciences).

| Western blot analysis
The protocol was reported in a previous study. 19 The antibodies we used were as follows: anti-E-cadherin (at 1: For verification of equal protein loading, polyclonal rabbit anti-GAPDH (at 1:6000, 10494-1-AP, Proteintech Technology) was used as a loading control. ImageJ software was used to analyze and quantify each band and relative band intensities.

| Luciferase reporter assay
Fragments of the TOB1 (NM_001243877) promoter were cloned into the pGL3-Basic Vector to generate promoter-reporter constructs. All constructs were verified by sequencing.
Cells were seeded in 24-well plates and transiently transfected with plasmids containing firefly luciferase reporters and recombinant promoter-reporter constructs. Reporter constructs containing the full-length TOB1 promoter region and other truncated fragments were transiently transfected into Foxa2overexpressing cells (transiently transfected with plasmids) and the respective control cells. The luciferase activity was measured after incubation for 48 hours using the Dual-Luciferase Reporter Assay Kit (Promega). The transfection efficiency was normalized with Renilla luciferase activity. The promoter sequence of TOB1 was listed in Supporting Information.

| RNA extraction and RNA-Sequencing (RNA-Seq)
Total RNA was extracted from four cell lines, BxPC-3-LV-TOB1, BxPC-3-LV-NC, Patu8988t-LV-TOB1, and Patu8988t-LV-NC, using TRIzol ® Reagent (Invitrogen) according to the manufacturer's instructions. NanoPhotometer spectrophotometer and Agilent 2100 Bioanalyzer were applied to evaluate the quality of the RNA samples. The RNA-Seq service was provided by Beijing Novogene. What tools used in this work contained Illumina HiSeq™ 4000 system (Illumina), the FastQC tool and the HISAT2 tool.

| Statistical analyses
Each experiment should be repeated independently at least three times. SPSS statistical software (Version 21.0) was used to conduct statistical analyses. Images were obtained using GraphPad Prism software (Version 7.0). The counting data are represented by frequency or percentage, and the measurement data are expressed as the mean ± standard deviation (SD). The rank sum test was used to reveal correlation between TOB1 expression and clinicopathologic features. The overall survival (OS) curve was plotted by Kaplan-Meier and compared with the log-rank test. The differences between the two groups were assessed using Student's t test. P < .05 was accepted as indicative of significant differences.

| TOB1 expression is downregulated in pancreatic cancer
To clarify the expression of TOB1 in pancreatic cancer, we explored the mRNA and protein expression of TOB1 in pancreatic samples via the Oncomine database and immunohistochemistry (IHC). Compared with normal pancreatic tissues, pancreatic cancer tissues showed significantly reduced TOB1 mRNA ( Figure 1A); the detailed data are shown in Table S1. By IHC, TOB1 was shown to be mainly expressed in the cytoplasm ( Figure 1B), and decreased TOB1 protein levels were observed in PDAC tissues ( Figure 1C).
To evaluate the clinical significance of TOB1 in pancreatic cancer, we analyzed the association between TOB1 expression and patients' clinicopathological characteristics (Table 1). Notably, TOB1 protein expression was significantly correlated with tumor size (Figure 1D), suggesting that TOB1 may play a role in negatively regulating the growth or proliferation of pancreatic cancer cells.
We next analyzed the association between TOB1 protein expression levels and survival time in 97 pancreatic cancer patients. Univariate analysis demonstrated that N stage and TNM stage were associated with the OS of pancreatic cancer patients in the Kaplan-Meier analysis (Table S2). Patients with negative TOB1 expression had worse outcomes than patients with positive TOB1 expression ( Figure 1E), although this finding was not significant.
We also analyzed the data in The Cancer Genome Atlas-Pancreatic adenocarcinoma (TCGA-PAAD), and found that TOB1 was higher in cancer tissues (t = −2.208, P = .029, Table S3). After dividing TOB1 mRNA expression into high and low according to the median, we analyzed the association between TOB1 mRNA expression levels and survival time in 174 pancreatic cancer patients (four patients were excluded because of no survival data), which showed that patients with high TOB1 mRNA expression had worse outcomes than patients with low TOB1 mRNA expression (P = .002, Figure S2), but its mRNA expression level had no relationship with clinicopathological characteristics of PAAD tissues (Table S4).

| TOB1 overexpression reduces proliferation of human pancreatic cancer cells
Given that TOB1 is involved in regulating the biological behaviors of various tumors, we examined whether TOB1 acts as a tumor suppressor in pancreatic cancer cells by utilizing a gain-of-function approach. BxPC-3-LV-TOB1 and Patu8988t-LV-TOB1 stable transfectants were established using lentivirus containing TOB1 cDNA. TOB1 stable transfectants were selected and confirmed by WB. As shown in Figure 2A, TOB1 was highly expressed in TOB1 stable transfectants compared with the corresponding NC and parental cells. In addition, the fluorescence intensity of TOB1-overexpressing cells was stronger than that in the corresponding NC and parental cells, and TOB1 protein was mainly expressed in the cytoplasm ( Figure S3 Figure 2C). Interestingly, these results occurred in BxPC-3 cells (K-Ras wild-type) but not in Patu8988t cells (K-Ras mutated). Metastasis is one of the toughest challenges in PDAC, and epithelial-to-mesenchymal transformation (EMT) is critical to the metastasis cascade of epithelial-derived solid tumor. We also assessed whether TOB1 affects migration-or EMT-related invasion. A transwell migration assay found that overexpression of TOB1 had no significant effect on cell migration in these two cell lines ( Figure 3A). The expression of several EMT-related markers, such as E-cadherin, N-cadherin, beta-catenin, and Vimentin, was evaluated by WB. As shown in Figure 3B, E-cadherin was slightly increased and N-cadherin, beta-catenin, and Vimentin were slightly decreased after overexpressing TOB1, but there was no statistically significant difference. Taken together, these results indicate that the mechanism of TOB1 against pancreatic cancer might be an antiproliferative effect rather than inhibition of invasion or metastasis in BxPC-3 cells in vitro.

| TOB1 mediates the antiproliferative effect through alterations in the cell cycle
Cell cycle analysis showed that overexpression of TOB1 in Bxpc-3 cells significantly blocked the transition from G 1 to S phase compared with negative control cells, but there was no significant difference in Patu8988t cells ( Figure 4A). Based on the above results, we studied the effect of TOB1 expression on cell cycle modulators. We evaluated alterations in CDK2, CDK4, and cyclin D1 levels in TOB1overexpressing cells compared to the negative control cells by WB. Figure 4B shows that cyclin D1 expression was slightly decreased in BxPC-3-LV-TOB1 cells, and CDK2 and CDK4 were strongly downregulated after overexpression of TOB1 in Bxpc-3 cells. Thus, the antiproliferative effects of TOB1 are likely mediated by regulating cell cycle progression through CDK2 and CDK4.

TOB1 targets the calcium pathway genes
After observation of the reversal of the malignant phenotype of pancreatic cancer cells by overexpressing TOB1, we explored the underlying molecular mechanisms. Specifically, the mechanistic research involved analysis of the upstream transcription factors and the downstream signaling pathways.
We first used TFBIND software to search for transcription factor binding sites of TOB1 (data not shown). Among transcription factors with a score greater than 0.90, Foxa2, which showed decreased expression and related to the malignant phenotype of pancreatic cancer cells, was selected for further analysis. [20][21][22] As shown in Figure 5A, a luciferase reporter assay revealed that overexpression of Foxa2 significantly enhanced TOB1 promoter activity, indicating that Foxa2 may act upstream of TOB1 and regulate its mRNA expression.
To identify related genes that are regulated by TOB1 in pancreatic cancer, we performed RNA-Seq analyses of controls or transfectants. The data revealed that 795 differential mRNAs were detected in the BxPC-3-LV-TOB1 vs BxPC-3-LV-NC groups ( Figure 5B). Through KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway analysis, 56 pathways were found to be involved in the BxPC-3-LV-TOB1 vs BxPC-3-LV-NC groups (Table S5), and it is worth noting that the calcium pathway showed significant enrichment. However, few differential genes and pathways were found in the Patu8988t-LV-TOB1 vs Patu8988t-LV-NC groups (Table S6). From these data, we found that after overexpression of TOB1, more genes and pathways are involved in the biological behavior of K-Ras wild-type cells than K-Ras mutated cells, consistent with the results of cytological phenotype.
To assess the expression of some typical differentially expressed genes involved in the calcium pathway in pancreatic cancer tissues, we performed an analysis via the Oncomine database and determined that some of these genes, such as CACNA1B, ADRB3, P2RX1, CAMKIIA, and NTSR1, had lower expression in the cancer tissues than normal tissues ( Figure 5C), while these genes were elevated after TOB1 overexpression ( Figure 5D). These findings indirectly suggest that TOB1 is likely to function through these calcium pathway genes.
In summary, Foxa2, as a transcription factor of TOB1, regulates its mRNA expression, while TOB1 participates in the anti-pancreatic cancer process by regulating the expression of calcium pathway genes.

| DISCUSSION
During previous decades, growing evidence has emphasized the rising importance of TOB1 in diversified human cancers, but how TOB1 works in pancreatic cancer has not been fully expounded. Therefore, we tried to explore whether altering TOB1 expression could affect the development of pancreatic cancer and elucidate its potential mechanism.
Originally, Yoshida 9 discovered that TOB1 knockout mice spontaneously formed tumors and suggested that the TOB1 gene played an important role in the formation of malignant tumors. Hereafter, abnormal TOB1 expression has been reported in many human malignant tumors, including breast cancer, 11 lung cancer, 10 papillary thyroid cancer, 12 skin squamous cell carcinoma, 23 and gastric cancer. 24 Studies also showed that the TOB1 expression level was correlated with the clinicopathological characteristics of tumors and could be used as a prognostic marker for some tumors. Among breast cancer patients without lymph node metastasis, patients with lower TOB1 expression had a worse prognosis. 18 In patients Error bars indicate the mean ± SD. NS, no significance. *P < .05; ****P < .0001 with intestinal gastric cancer, low TOB1 and high phosphorylation of TOB1 both predicted poor prognosis. 25 No studies have been conducted on the expression of TOB1 in pancreatic cancer tissues.
We discussed TOB1 expression and its clinical significance in pancreatic cancer tissues. Compared with normal pancreatic tissues, pancreatic cancer tissues showed substantially decreased TOB1 mRNA and protein levels. In addition, by analyzing the correlation between TOB1 protein and the clinicopathological features of pancreatic cancer patients, we found that larger cancer tissues had a lower TOB1 protein level. Although there was no significant difference, patients with negative TOB1 expression had worse outcomes. This result may be explained by the limited number of cases, and increased sample sizes are needed to confirm the prognostic value of this molecule in pancreatic cancer patients. But, we got an opposite result in TCGA-PAAD analysis. As we all know, ductal adenocarcinoma is an invasive mucin-producing and duct-forming epithelial neoplasm with an intense stromal desmoplastic response, often presents low neoplastic cellularity, only 5%-20%. 26 Prior genome sequencing studies employed techniques to purify tumor samples, which makes that samples with low neoplastic cellularity have been underrepresented, even though low cellularity cancers comprise the majority of surgically resected PDAC. 27 Besides, there are only four normal control cases in TCGA-PAAD. Given this, we considered that the analysis result of TCGA-PAAD may be not very reliable. To sum up, we concluded that TOB1 expression is downregulated in pancreatic cancer and may act as a tumor suppressor.
As a candidate tumor suppressor gene, TOB1 is also involved in regulating proliferation, apoptosis, invasion, migration, and radiochemotherapeutic sensitivity of various tumor cells. 15,17,18,[28][29][30] We demonstrated that Error bars indicate the mean ± SD. NS, no significance overexpression of TOB1 could inhibit proliferation and colony formation. The cell cycle analysis suggested that TOB1 overexpression led to cell aggregation in G 1 phase and cell depletion in S phase by decreasing the expression of CDK2 and CDK4. A strong relationship between TOB1 and cyclin D1 has been reported in the literature, 31-33 but we did not observe a correlation between TOB1 expression and cyclin D1. No effects on cell migration or expression of EMT-related markers were found in our study. All the above results demonstrated that the antitumor effect of TOB1 in K-Ras wild-type pancreatic cancer cells is due to its antiproliferative effect. Abnormal activation of K-Ras is present in more than 90% of PDAC, 34 which represents a molecular typing of pancreatic cancer. One study found that TOB1 was involved in the regulation of chemotherapeutic sensitivity in HER2-positive breast cancers, 18 suggesting that TOB1 plays different roles in different molecular subtypes of cancer. From our results, we found that TOB1 may function as a tumor suppressor gene in K-Ras wild-type pancreatic cancer and dysregulated TOB1 may partially interpret the underlying mechanism of K-Ras wild-type pancreatic cancer carcinogenesis, but what roles it plays in K-Ras mutant pancreatic cancer cells deserve further study.
Some studies have suggested that the cytoplasmic localization of TOB1 is required for its antiproliferative effect. 14 In contrast, another study reported that the nuclear localization of TOB1 had antiproliferative activity. 13 Prior studies have also noted the importance of phosphorylation of TOB1. 10,12 Our tissue and cell studies demonstrated that in pancreatic cancer, TOB1 protein was mainly located in the cytoplasm, indicating that its antiproliferative effect may be independent of nuclear localization.
Several reports have shown that TOB1 can function through a variety of pathways, including Smad 4/β-catenin, 28 PI3K/PTEN, 15 and MAPK/ERK. 29 Similarly, in our analysis of pathways, some typical pathways were included, such as the MAPK and TNF signaling pathways. Beyond that, the most obvious finding to emerge from pathway analysis is that TOB1 likely works through regulating the calcium pathway genes. Ca 2+ is a ubiquitous cell signaling molecule for various physiological and pathological processes. 35,36 In the 1970s, researchers discovered that cell proliferation may be regulated by calcium pathways, 37 and some key steps in cell cycle progression are dependent on calcium signaling, especially the early transition into the G 1 , G 1 /S, and G 2 /M phases. 38 Prior studies have also noted the importance of calcium signaling and tumors. [39][40][41][42][43][44][45][46][47] Davis 48 found that the concentration of Ca 2+ in the cytoplasm was temporarily increased when using epidermal growth factor (EGF) and hypoxia to induce EMT. Conversely, EGF and hypoxia-induced EMT were reduced when decreasing the calcium in breast cancer cells. 48 Giorgi 49 proposed that TP53 mutation could inhibit cell apoptosis through the calcium signaling pathway. Tumors driven by H-Ras were reported to be maintained by calcium signaling pathways. 50 Wang 51 established that the Wnt-Ca 2+ pathway was a key pathway that promoted the development and progression of tumors with K-Ras mutations and proposed that blocking the combination of K-Ras and calmodulin could be a F I G U R E 5 TOB1 regulates the expression of calcium pathway genes in pancreatic cancer. A, Luciferase reporter assay. 1, Promoter-NC + TF-NC; 2, Promoter-NC + Foxa2; 3, TOB1 Promoter + TF-NC; 4, TOB1 Promoter + Foxa2. B, All cluster heatmaps of differential genes. C, mRNA expression of some typical differentially expressed genes involved in the calcium pathway in pancreatic cancer tissues. The number in cells represents the number of analyses meeting the thresholds (P-value < .05, fold change = all, gene rank = all; data type: mRNA). Gene rank determines cell color. Specifically, a stronger red or blue color indicates a more pronounced upregulation or downregulation of the gene. D, Some typical differentially expressed genes involved in the calcium pathway after overexpressing TOB1. B-LV-NC, BxPC-3-LV-NC; B-LV-TOB1, BxPC-3-LV-TOB1; P-LV-NC, Patu8988t-LV-NC; P-LV-TOB1, Patu8988t-LV-TOB1. **** P < .0001 potential treatment strategy for cancer with K-Ras mutations. It is important to determine whether the TOB1 calcium pathways could be used as therapeutic targets in the further study.
In summary, our study demonstrates that TOB1 expression is downregulated in human pancreatic cancer. Overexpression of TOB1 can partially reverse the malignant phenotype of K-Ras wild-type pancreatic cancer cells by regulating calcium pathway genes.