Immortalization up‐regulated protein promotes tumorigenesis and inhibits apoptosis of papillary thyroid cancer

Abstract The incidence of thyroid cancer is increasing in recent years worldwide, but the underlying mechanisms await further exploration. We utilized the bioinformatic analysis to discover that Immortalization up‐regulated protein (IMUP) could be a potential oncogene in the papillary thyroid cancer (PTC). We verified this finding in several databases and locally validated cohorts. Clinicopathological features analyses showed that high expression of IMUP is positively related to malignant clinicopathological features in PTC. Braf‐like PTC patients with higher IMUP expression had shorter disease‐free survival. The biological function of IMUP in PTC cell lines (KTC‐1 and TPC‐1) was investigated using small interfering RNA. Our results showed that silencing IMUP suppresses proliferation, migration and invasion while inducing apoptosis in PTC cell lines. Changes of the expression of apoptosis‐related molecules were identified by real‐time quantitative polymerase chain reaction and Western blotting. We also found that YAP1 and TAZ, the critical effectors in the Hippo pathway, were down‐regulated when the IMUP is silenced. Rescue experiments showed that overexpression of YAP1 reverses the tumour inhibitory effect caused by IMUP knockdown. Our study demonstrated that IMUP has an oncogenic function in PTC and might be a new target gene in the treatment of PTC.

actual increase of all TC cases. 3,4 Almost 50% of PTC patients appear with positive lymph nodes, and 68% of high-risk differentiated thyroid cancer patients could have persistent structural disease or recurrence after a standard regimen. 5,6 These observations motivated us to explore the underlying mechanisms of PTC further.
Previous studies have shown that the majority of mutations in PTC are associated with mitogen-activated protein kinase (MAPK) signalling pathway, such as BRAF and RAS genes. Novel low-frequency mutations such as EIF1AX, PPM1D and CHEK2 are discovered to reclassify TC into different molecular subtypes. 7 In our study, we found that Immortalization up-regulated protein (IMUP) could be associated with the progression and development of PTC.
IMUP is also known as Hepatocyte Growth Factor Activator Inhibitor Type 2-Related Small Protein (H2RSP) or C19orf33. This protein-coding gene locates on chromosome 19q13.2. Previous studies showed that IMUP is involved in SV40-mediated immortalization in human fibroblasts and is also related to tumorigenicity and cellular proliferation. 8,9 In the area of cancer research, IMUP was up-regulated in endometrial carcinoma and its higher expression was associated with the aggressive features of breast tumours. 10,11 However, no investigation has been done regarding the biological function of IMUP in PTC yet.
In the present study, we validated that IMUP is significantly higher in PTC tumour tissues compared with matched normal tissues, examined the relationship between PTC clinical characteristics and IMUP expression. We performed a serial of experiments to reveal the effect of IMUP silencing on PTC cell lines, as well as the relationship between the IMUP expression and the Hippo pathway markers such as YAP1. Our results implied that the IMUP could be a crucial oncogene and might be a potential target for the therapy of PTC in the future.

| Bioinformatics analysis
The mRNA expression data were downloaded from The Cancer Genome Atlas (TCGA) data portal (https://tcgad ata.nci.nih.gov/ tcga/) and Gene Expression Omnibus (GEO) database (https://www. ncbi.nlm.nih.gov/gds/). Transcriptome sequencing data of 502 PTC tissues with matched complete clinicopathological characteristics data and 58 non-tumour tissues were collected from TCGA. The GSE33630, GSE60542 and GSE35570 data sets were based on the GPL570 platform, Affymetrix Human Genome U133 Plus 2.0 Array, merged and processed by robust multi-array average (RMA) then normalized using the quantile normalization. The GSE50901 database was based on the GPL13607 platform, Agilent-028004 SurePrint G3 Human GE 8x60K Microarray, numeric data were generated and normalized by intensity-dependent global normalization (LOWESS). In total, 170 PTC tumour samples with associated clinicopathological features and 127 non-tumour samples were selected.
The volcano plot was generated using the Limma R and ggplot package to present the differential expression genes (DEGs) between cancerous tissues and normal tissues in corresponding databases.
Kaplan-Meier plots for disease-free survival (DFS) were obtained from the Gene Expression Profiling Interactive Analysis 2 (GEPIA2) (http://gepia2.cance r-pku.cn). Gene Set Enrichment Analysis (GSEA) was conducted using GSEA v3.0 software (http://www.broad insti tute.org/gsea), which analysed the differences in mRNA expression levels of biological annotation and pathways to discover the downstream signalling pathway of IMUP.

| RNA extraction and real-time quantitative polymerase chain reaction (qRT-PCR)
RNA from the patients' tumour specimens and cell lines was extracted using TRIzol reagent (Thermo Fisher Scientific, Waltham, USA) in compliance with the manufacturer's protocol. The quality (A260/A280 ratios) and quantity of the extracted RNA were assessed using spectrophotometry NanoDrop 1000 (Thermo Fisher Scientific). Reverse transcription reaction was performed using ReverTra Ace qPCR RT Kit (Toyobo, Osaka, Japan) following the manufacturer's protocol (20 μl reaction; 1000 ng of total RNA; step 1 16°C for 5 minutes, step 2 42°C for 30 minutes, step 3 98°C for 5 minutes). cDNA was stored at −20°C. The PCR analysis was performed using the Applied Biosystems 7500 Real-Time PCR system (Applied Biosystems, Thermo Fisher Scientific, Inc) and the SYBR Premix Ex Taq II kit (RR820A, TaKaRa

| Cell transfection
Thyroid cancer cells were seeded into 6-well plates and cultured  for 2-4 hours. The absorbance at 450 nm was quantitated by spectrophotometer on four consecutive days. As for the colony formation assay, the same number of cells was seeded into 6-well plates and incubated for 7 days in the atmosphere as above. Then, the cells were fixed with 4% paraformaldehyde for 30 minutes and stained with 0.1% crystal violet solution for 30 minutes at 37°C. Images were captured by a digital camera.

| Cell migration and invasion
Transwell chambers (#3422, Corning, NY, USA) were used in the cell migration assays. TC cells (3.5 × 10 4 cells/well) were transferred into the upper chamber in the serum-free medium after transfection. The lower chamber contained 0.6 ml medium supplemented with 10% FBS. Cells were incubated for 22 hours before the low chamber cells being fixed with 4% paraformaldehyde and stained with 0.4% crystal violet solution. Cells that were not able to migrate to the lower chamber were carefully removed. Five stochastic fields of view were captured by the digital camera under the microscope at ×20 magnification for analysis. The invasion ability was examined using the Matrigel invasion chamber (#354480; Corning Biocat, NY, USA) with the same procedure as described. Scratch wound experiments were conducted in 24-well plates; 2.0 × 10 5 cells were incubated with serum-free culture for 48 hours after wounded with a plastic tip.
The gaps and cells were imaged before and after scratching, and images were captured by a digital camera under the microscope at ×5 magnification. The wound healing area was calculated using Image J software (NIH, MD, USA). The migrating rate (%) = (wound area at 0 h-wound area at 48 h)/wound area at 0 h × 100%.   was used for comparisons of data with more than two group. The receiver operating characteristic (ROC) curve was applied to evaluate the diagnostic efficacy of IMUP in PTC, and P < 0.05 was considered to indicate a statistically significant difference.

| IMUP is up-regulated in the PTC tumour
Among all DEGs, the gene IMUP drew our attention. In microarray analysis, we found that the IMUP expression level is significantly higher in PTC tumour tissues than non-tumour tissues ( Figure 1D: Three integrated GEO dataset, P < 0.0001; Figure 1E: GSE50901, P < 0.0001).

| The association between IMUP expression and clinical features of PTC patients
To explore the association between IMUP expression and clinicopathological features of PTC patients, we divided PTC tumour samples into the high and low-expression group according to respective IMUP expression median value in the validated and TCGA cohorts.
In our validated cohort, IMUP expression was correlated with tumour size (P = 0.023) and lymph node metastasis (LNM) (P = 0.015) ( Table 1). In the TCGA cohort, our results showed that histological type (P < 0.001), T stage (P = 0.004), LNM (P < 0.001) and disease stage (P = 0.006) have a significant association with IMUP expression (Table 2). Then, we evaluated IMUP expression in PTC patients with different tumour stages, subtypes, and molecular classification in the TCGA cohort. As illustrated in Figure 2A, most patients with higher disease stages had higher expression of IMUP IMUP was significantly higher in the classical and columnar variant subtypes than in the follicular subtype ( Figure 2B). Moreover, the IMUP mRNA levels in patients with LNM were significantly higher than those without LNM (P < 0.0001, Figure 2C). According to a previous report, PTC could be mainly classified into BRAF-like and RAS-like subtypes based on the  7 In our study, we discovered that IMUP expression was higher in the BRAF mutation group ( Figure 2D, P < 0.0001) and RAS wild-type group ( Figure 2E, P < 0.0001) than in the respective counterpart group. Based on the classification mentioned above, we found that IMUP expression is significantly higher in the BRAF-like group than the RAS-like group (P < 0.0001, Figure 2F).
To further explore the prognostic value of IMUP in the BRAFlike group, we performed the survival analyses using Kaplan-Meier curves from the GEPIA2. Higher IMUP expression based on median value was associated with a higher risk of relapse or death in the BRAF-like subgroup (hazard ratio [HR] = 2.0, P = 0.085, Figure 2G).
The result was more significant when the classification was based on the quartile value (HR = 3.7, P = 0.035, Figure 2H). Thus, IMUP overexpression is correlated with more aggressive clinicopathological features and may predict a worse prognosis of PTC patients.  (Table 3). These data suggested that high level of IMUP expression is an independent predictive factor for LNM.

| Down-regulation of IMUP suppresses proliferation of PTC cell lines
To specify the function of IMUP, we started our in vitro study by examining the relative expression level of IMUP compared with GAPDH in different PTC cell lines. Results showed that KTC-1, TPC-1 and BCPAP cell lines express higher IMUP than normal thyroid cells (HTORI3) ( Figure 3A). We then used two small interfering RNA sequences (SiRNA1 and SiRNA2) to silence the target gene in our selected cell lines with the highest IMUP expression (KTC-1 and TPC-1); results showed that the relative mRNA and protein expression of IMUP was effectively down-regulated ( Figure 3B,C). It was clearly shown in our CCK-8 assay that the proliferation capacity of IMUP down-regulated cell lines was suppressed ( Figure 3D,E), the same as the result of colony formation where the number of IMUPknockdown colonies was significantly lesser than the normal ones ( Figure 3F). The above results demonstrated that IMUP could enhance the proliferation of PTC cells.

| Down-regulation of IMUP suppresses migration and invasion of PTC cell lines
To validate the hypothesis that the expression of IMUP would be associated with the metastasis of PTC cell lines, we conducted further experiments. We found that the migratory capacity of down-regulated IMUP PTC cells was inhibited compared with the control groups ( Figure 4A); similar results were observed in the invasion experiments ( Figure 4B). The scratch tests showed   19 The results of Western blotting showed that the expression of YAP1 and TAZ was significantly decreased; the protein expression of BAX, Cleaved-Caspase9 increased while BCL-XL and BCL-2 were decreased ( Figure 5D).

| Down-regulation of IMUP promotes apoptosis and decreases the YAP1/TAZ expression of PTC cell lines
These results suggested that IMUP in PTC cells may promote tumour malignant phenotype by affecting the Hippo-YAP1/TAZ pathway.

| The oncogenic role of IMUP is partly dependent on YAP1
According to reports, YAP1 is a key effector of the Hippo pathway and a cancer-promoting gene in PTC. [20][21][22] As the expression of YAP1 was down-regulated when the IMUP was silenced, rescue experiments were employed to explore whether YAP1 is involved in the effect of IMUP on PTC cell progression. The YAP1 was overexpressed in the TPC-1 and KTC-1 using plasmid (pcDNA3.1-YAP1), and the transfection efficiency was confirmed at RNA and protein levels ( Figure 6A,B). The CCK8 assay and colony formation assay showed that the up-regulation of YAP1 rescues IMUP silencing-induced inhibition of cell proliferation ( Figure 6C,D). The Transwell migration, invasion and wound healing assays indicated that overexpression of YAP1 partially counteracts the IMUP silencing-mediated effects on migration and invasion in PTC cell lines ( Figure 6E-G). The apoptosis rate of IMUP-knockdown cells was decreased when the expression of YAP1 was up-regulated ( Figure 6H). These results revealed that IMUP promotes PTC cell proliferation, migration and invasion while inhibiting apoptosis via YAP1.  dering tumour proliferation, metastasis and anti-apoptosis. 29 The activation of the Hippo pathway could be initiated by cell density sensing, DNA damage along with other signalling molecules. 30,31 YAP1 is one of the most important effectors downstream of the Hippo signalling pathway and its aberrant expression that contributes to tumour progression indicates poor outcomes in various cancers. [32][33][34] It has been reported that the HIPPO pathway is relatively inactive in PTC compared with that in normal thyroid tissues. 35 Some also reported that YAP1 is overexpressed and serves as an oncogene that correlates with poor prognosis of PTC patients. 22 YAP1). E-G, Transwell and wound healing assays showed that exogenous YAP1 expression reversed the suppression of migration and invasion caused by IMUP silencing. H, Apoptosis assay showed that overexpression of YAP1 rescues the apoptosis-promoting effect caused by IMUP silencing. Student's t test was used for statistical analyses; *P < 0.05, **P < 0.01, ***P < 0.001 that IMUP might promote tumorigenesis and progression of PTC In summary, we found that the expression of IMUP is up-regu-

ACK N OWLED G EM ENTS
This study was funded by the Natural Science Foundation of Zhejiang Province (LGF18H160031) and the Science And Technology Project of Wenzhou (Y20190204; Y20170740).

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.