Up‐regulation of CDHR5 expression promotes malignant phenotype of pancreatic ductal adenocarcinoma

Abstract CDHR5 has been reported to play key roles in carcinogenesis of various cancers, but its roles in pancreatic cancer have not been reported. The present study was designed to investigate its clinical value in pancreatic ductal adenocarcinoma (PDAC). Tissue microarray‐based immunohistochemistry was performed to analyse the correlation between CDHR5 expression and clinical and pathological features of PDAC, as well as the CDHR5 expression during tumour progression. Cell function assays were performed to investigate CDHR5’s effects on PDAC cells. Moreover, qRT‐PCR was applied to investigate the expression of CDHR5 isoforms in PDAC cells. Expression of CDHR5 was higher on the membrane of PDAC cells. This high expression level was associated with shorter overall survival of PDAC patients and was identified as an independent prognostic factor for overall survival by multivariate Cox regression analysis. In addition, expression level of CDHR5 presented an increased trend in the occurrence and progression of PDAC. Cell experiment suggested that CDHR5 could notably promote invasion and migration of PDAC cells. Moreover, analysis of CDHR5 isoforms indicated CDHR5‐L was the major isoform expressed in PDAC cell lines. CDHR5 appears to be a promising and novel prognostic factor for PDAC, and its promotion in PDAC metastasis might be ascribed to the isoform CDHR5‐L.


| INTRODUC TI ON
Pancreatic ductal adenocarcinoma (PDAC) is the forth lethal malignancy, and its overall 5-year survival rate is only 3%-9%. [1][2][3] Surgical resection is currently the only potential curative option, but over 80% of PDAC patients have suffered an unresectable tumour at the time of diagnosis, in which most of the tumours have been locally advanced invading or even occurred metastasis. 4 Although lots of efforts have been invested in researching adjuvant and neoadjuvant therapy for pancreatic cancer, its indications and benefits are still controversial and need further studies. 5,6 Therefore, it is vital to search novel molecular targets for inhibition of tumour metastasis.
CDHR5, also known as MLPCDH, MUCDHL, MUPCDH and MU-PCDH, is a member of the cadherin superfamily and encodes the transmembrane protein CDHR5. CDHR5 contains four main functional domains, including the cadherin domain, the mucin-like domain, the proline-rich region and the PDZ binding motif. In normal physiological conditions, CDHR5 is generally distributed in the surface of epithelial cell of kidneys, lungs and other organs, playing an essential role in various cellular process including cell adhesion and branching morphogenesis of organs. 7 Studies on renal cell carcinoma, colorectal cancers and liver cancers revealed the low expression of CDHR5 on the surface of tumour cells and its correlation with poor prognosis. [8][9][10] However, the roles of CDHR5 in pancreatic cancer have not been reported.
In this study, tissue microarrays were used to investigate the relationship between CDHR5 expression and clinicopathological characteristics of PDAC patients. In addition, cellular and molecular studies have been applied to further explored the roles of CDHR5 in pancreatic cancer cells and the potential mechanisms.

| Immunohistochemistry staining and evaluation
The immunohistochemistry (IHC) staining was performed on the TMAs. Both TMAs were deparaffinized with xylene and rehydrated in a series of ethanol solutions decreasing in concentration. They were then undergoing antigen retrieve step in Colloidal Bismuth Subcitrate buffer solution pH 6, followed by treated with 3% H 2 O 2methanol for 10 minutes. After blocking the non-specific antigens with 5% goat serum, TMAs were incubated with polyclonal anti-MUPCDH antibody (ab121306, Abcam; 1:100) overnight at room temperature. At the end of the reaction, the TMAs were reacted with 3,3′-diaminobenzidine, tetrahydrochloride (DAB) and counterstained with haematoxylin.

| Patients and study design
HPan-Ade170Sur-01 was applied to analyse the expression of CDHR5 in PDAC and NAT, as well as the relationship between CDHR5 expression and patients' clinicopathological characteristics. To investigate the difference of CDHR5 expression between tumour and para-tumour tissues, patients were included according to the following criteria: with both PDAC tissue and matched NAT; without pre-operative chemotherapy or radiotherapy; pathological diagnosis as PDAC. To further study the relationship between CDHR5 expression and patients' clinicopathological characteristics, patients were included according to the following criteria: less than 75 years old; without pre-operative chemotherapy or radiotherapy; pathological diagnosis as PDAC; and without perioperative death.
BIC14011 was used to detect the potential changes of CDHR5 expression during tumour progression, and patients were included according to the following criteria: without pre-operative chemotherapy or radiotherapy; pathological diagnosis as pancreatitis, pancreas intraepithelial neoplasia or PDAC.
Both culture mediums were supplemented with 10% heat-inactivated foetal bovine serum (SH30084.03, HyClone), and all the cells were cultured at 37°C under a humidified 5% CO 2 atmosphere.
CDHR5 overexpression plasmids were constructed and used for the transfection with Lipofectamine™ 2000 reagent (Invitrogen).

| RNA extraction and quantitative real-time PCR
Total RNA was extracted from several PDAC cell lines by using the

| Transwell assay
Transwell assays were performed to test cell migration and invasion abilities. The 24-well transwell plates used in this assay contained a polycarbonate membrane with 8.0-μm pores. In the invasion assays, chambers were pre-coated with 10-fold diluted Corning Matrigel.
Cells transfected by siRNAs were seeded onto the upper surface of the transwell membrane at a density of 30 000 cells/100 μL/well.
After 48 hours, we gently cleared cell in the upper surface of chambers and stained cells migrating to the bottom with haematoxylin and eosin. Cell counting was performed in 5 randomly fields under light microscope.

| Statistical analysis
Compared t test was applied to analyse the difference of CDHR5 expression between PDAC tissue and adjacent normal ductal tissues. Chi-square test and Fisher's exact test were used to evaluate significance of associations between CDHR5 expression and clinicopathological parameters. Survival analyses were conducted by using uni-and multivariate Cox proportional hazards and Kaplan-Meier analyses. To define the expression status of CDHR5, ROC curve and median method were used to determine the cut-off value. Our study was carried out under the guideline of medical ethics committee at Peking Union Medical College Hospital, and all the experiments above were repeated for three times, and results were presented as mean ± SEM values. The statistical analyses were performed with SPSS for mac, version 24. All statistical tests were two-sided, and the significance level was defined as P < .05.

| Immunohistochemical evaluation of CDHR5 in PDAC and adjacent normal tissues
Fifty-nine PDAC cases with both tumour and matched NAT samples were included in this analysis. Results of IHC staining showed that CDHR5 is mainly expressed on the membrane of PDAC cells, and expression was observed on the membrane of normal ductal cells ( Figure 1A). Based on the IHC evaluation standards mentioned before, it is suggested that the expression of CDHR5 was remarkedly increased on the membrane of PDAC cells than that on the matched NAT (P < .001, Figure 1B). In addition, we have also analysed the prognostic value of CDHR5 in PDAC patients by using TCGA dataset. Results also showed that CDHR5 expression was significantly higher in pancreatic tumours compared to that in adjacent normal tissues ( Figure S1).

| Association of CDHR5 expression with clinicopathological features
To study the potential interaction between CDHR5 expression and clinicopathological features and prognosis of PDAC patients, sixtynine cases were finally enrolled in terms of the inclusive criteria.
According to the optimal cut-off value of CDHR5 expression, 40 (58.0%) patients were classified into the low expression group and 29 (42.0%) into the high expression group. No significantly association was observed between CDHR5 expression and main clinicopathological parameters (Table 1).

| Prognostic value of CDHR5 expression in PDAC patients
Result of Kaplan-Meier analysis suggested that PDAC patients in the CDHR5 high expression group had significantly shorter overall survival time (OS) than those patients with low expression of CDHR5 (median OS: 8 months vs 33 months, P < .001, Figure 2). In the univariate analyses, significance level was defined as P < .15 to avoid omitting important risk factors. As a result, pathological grade, N stage and CDHR5 expression were included for further multivariate Cox regression analysis, and results indicated that CDHR5 expression and pathological grade were independent prognostic factors for OS of PDAC patients (P < .001 and P = .014, respectively; Table 2).
Results showed that CDHR5 expression experienced a significantly increase with the occurrence and progression of PDAC (P = .028, Table 3). According to the updated 2-tiered classification of neoplastic pre-cursor pancreatic lesions, PanIN1 and PanIN2 were categorized as low-grade PanIN (L-PanIN), and PanIN3 was classified into the highgrade PanIN (H-PanIN). 11 In this type of classification, chi-square test also showed a significant difference of CDHR5 expression between CP＆L-PanIN group and H-PanIN＆PDAC group (P = .019, Table 4).  Figure 4B). However, the migration and invasion assay showed that down-regulation of CDHR5 significantly suppressed the migration and invasion capability of AsPC-1 cells (Figure 4C-E). In addition, similar results were observed in the overexpression groups. It was showed that CDHR5 overexpression could not influence cell proliferation but remarkably promoted cell migration and invasion ( Figure 5).

| Expression of CDHR5 isoforms in PDAC
Expression of CDHR5 transmembrane isoforms, the CDHR5-L and CDHR5-M, was detected by qRT-PCR, respectively, and results showed that expression of CDHR5-L was extremely higher in AsPC-1 cells than in the other three PDAC cell lines, while CDHR5-M was about 10 times higher in AsPC-1 cells than in the other three PDAC cell lines ( Figure 6).

| D ISCUSS I ON
In the present study, we investigated the CDHR5 expression pattern and its clinical value in PDAC. Results showed that CDHR5 was extremely highly expressed in pancreatic tumour tissues and its high expression was associated with tumour metastasis. Moreover, the isoform CDHR5-L was identified as the functional protein dominating in the process. In cell experiments, Western blot and qRT-PCR were applied to screen the CDHR5 expression levels in multiple PDAC cell lines.
Results showed that the expression levels of CDHR5 in AsPC-1 cell lines were significantly higher than that in the other three cell lines. Cell functional assays showed that down-regulation of

| CON CLUS ION
In conclusion, our study revealed that CDHR5 expression was correlated with outcomes of PDAC patients and could be a promising prognostic factor for PDAC. As the isoform CDHR5-L might function in promoting PDAC metastasis, it provided a novel therapeutic target for PDAC management.

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

CO N S E NT FO R PU B LI C ATI O N
Not applicable.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data sets used and/or analysed during the current study are available from the corresponding author on reasonable request.

O RCI D
Junyi Gao https://orcid.org/0000-0002-7214-8448 F I G U R E 6 Expression of CDHR5 isoforms in PDAC cell lines. A, CDHR5-L relative expression in PDAC cell lines. B, CDHR5-M relative expression in PDAC cell lines