Carboxypeptidase E-∆N Promotes Proliferation and Invasion of Pancreatic Cancer Cells via Upregulation of CXCR2 Gene Expression

Pancreatic cancer is one of the leading causes of cancer-related mortality worldwide. The molecular basis for the pathogenesis of this disease remains elusive. In this study, we have investigated the role of wild-type Carboxypeptidase E (CPE-WT) and a 40 kDa N-terminal truncated isoform, CPE-ΔN in promoting proliferation and invasion of Panc-1 cells, a pancreatic cancer cell line. Both CPE-WT and CPE-ΔN were expressed in Panc-1 and BXPC-3 pancreatic cancer cells. Immunocytochemical studies revealed that in CPE transfected Panc-1 cells, CPE-ΔN was found primarily in the nucleus, whereas CPE-WT was present exclusively in the cytoplasm as puncta, characteristic of secretory vesicles. Endogenous CPE-WT was secreted into the media. Overexpression of CPE-ΔN in Panc-1 cells resulted in enhancement of proliferation and invasion of these cells, as determined by MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) cell proliferation assay and Matrigel invasion assay, respectively. In contrast, the expression of CPE-WT protein at comparable levels to CPE-ΔN in Panc-1 cells resulted in promotion of proliferation but not invasion. Importantly, there was an upregulation of the expression of CXCR2 mRNA and protein in Panc-1 cells overexpressing CPE-ΔN, and these cells exhibited significant increase in proliferation in a CXCR2-dependent manner. Thus, CPE-ΔN may play an important role in promoting pancreatic cancer growth and malignancy through upregulating the expression of the metastasis-related gene, CXCR2.


Introduction
Pancreatic cancer (PC) is the 12th most common cancer in the world (wcrf.org) and the 7th leading cause of cancer-related death globally [1]. Most PCs are epithelial exocrine cancers with a high occurrence of malignancy. PCs are generally diagnosed in the advanced stages, when metastasis has already occurred, leading to poor prognosis and a high incidence of mortality [2]. The molecular and cellular mechanisms underlying the pathogenesis of PC malignancy are poorly understood. In the early low-grade PanIN-1 (Pancreatic Intraepithelial Neoplasia) stage, KRAS is mutated, oncogenic miRNAs are overexpressed, and associated stromal factors are activated. In the PanIN-2 intermediate stage, inactivating mutations in the p16/CDKN2A gene and overexpression of MUC1 are observed. In the late PanIN-3 stage, inactivating mutations in TP53, BRCA2, and SMAD4 genes were found. The tumor environment, especially tumor-stromal interactions, also contribute to the aggressive progression of the disease [1]. Identification of novel molecular factors and mechanisms involved in the development of PC will uncover diagnostic and prognostic biomarkers and therapeutic targets.

CPE Transcripts and Proteins Expressed in Human Pancreatic Cancer Cell Lines
The CPE mRNA expression pattern for two pancreatic cancer lines BXPC-3 and Panc-1 was examined by Northern blot. Two mRNA transcripts, a~2.4 kb CPE and~1.7 kb CPE transcript variant, were detected in the BXPC-3 cells ( Figure 1A). The 1.7 kb RNA represented a CPE ∆189-386 splice variant encoding a 40 kDa N-terminal truncated CPE protein (CPE-∆N) and the 2.4 kb, CPE-WT transcript [19] ( Figure 1B). In the Panc-1 cells, only a 2.4 kb CPE-WT transcript was detectable. To determine the protein expression of CPE, BXPC-3 and Panc-1 cell extracts were immunoprecipitated with CPE antibody 6135 followed by Western blot using BD monoclonal antibody against CPE. The pull-down experiments revealed expression of a major 40 kDa CPE-∆N protein and a minor 53 kDa CPE-WT protein in BXPC-3 ( Figure 1C

Subcellular Distribution of CPE and CPE-ΔN in Panc-1 Cells
Analysis of the cell lysate and concentrated culture media of Panc-1 cells showed the presence of 53 kDa CPE in the media but not CPE-ΔN. In contrast, the cell lysate showed only 40 kDa CPE-ΔN protein ( Figure 2A). These data indicated that endogenous CPE-WT protein was secreted, while the 40 kDa CPE-ΔN remained within the cells. To study the subcellular localization of CPE-WT and CPE-ΔN, we carried out immunocytochemistry on the Panc-1 cells stably transfected with V5 tagged CPE-WT and CPE-ΔN. The expression of the V5 tag was validated by Western blot, which showed V5 bands corresponding to the size of CPE-WT and CPE-ΔN in these transfected Panc-1 cells ( Figure 2B). Immunocytochemical analysis revealed that CPE-ΔN-V5 immunostaining overlapped with DAPI (4,6-Diamidino-2-phenylindole) staining, indicating its localization in the nucleus, as well as in the cytoplasm (Figure 2Ci). However, in CPE-WT-V5 transfected Panc-1 cells, V5 immunostaining was not present in the nucleus, but only in the cytoplasm, showing a punctate appearance indicative of the presence in secretory vesicles ( Figure 2Cii). As a negative control, cells transfected with the vector

Subcellular Distribution of CPE and CPE-∆N in Panc-1 Cells
Analysis of the cell lysate and concentrated culture media of Panc-1 cells showed the presence of 53 kDa CPE in the media but not CPE-∆N. In contrast, the cell lysate showed only 40 kDa CPE-∆N protein ( Figure 2A). These data indicated that endogenous CPE-WT protein was secreted, while the 40 kDa CPE-∆N remained within the cells. To study the subcellular localization of CPE-WT and CPE-∆N, we carried out immunocytochemistry on the Panc-1 cells stably transfected with V5 tagged CPE-WT and CPE-∆N. The expression of the V5 tag was validated by Western blot, which showed V5 bands corresponding to the size of CPE-WT and CPE-∆N in these transfected Panc-1 cells ( Figure 2B). Immunocytochemical analysis revealed that CPE-∆N-V5 immunostaining overlapped with DAPI (4,6-Diamidino-2-phenylindole) staining, indicating its localization in the nucleus, as well as in the cytoplasm (Figure 2Ci). However, in CPE-WT-V5 transfected Panc-1 cells, V5 immunostaining was not present in the nucleus, but only in the cytoplasm, showing a punctate appearance indicative of the presence in secretory vesicles ( Figure 2Cii). As a negative control, cells transfected with the vector alone showed no V5 immunostaining (Figure 2Ciii). These results indicated that CPE-∆N, but not CPE-WT was localized in the nucleus of the Panc-1 cells. alone showed no V5 immunostaining (Figure 2Ciii). These results indicated that CPE-ΔN, but not CPE-WT was localized in the nucleus of the Panc-1 cells.  Figure S2A,B (C). Representative confocal images of Panc-1 cells stably expressing CPE-WT-V5 and CPE-ΔN-V5 or vector alone (control). Anti-V5 (red) and nuclear staining with DAPI (blue) are shown. Note V5 tag staining in the cells expressing CPE-ΔN-V5 was localized in the nucleus overlapping with DAPI stain, as well as in the cytoplasm (top panels) while V5 staining in cells expressing CPE-WT-V5 was localized exclusively in the cytoplasm (middle panels). Panc-1 cells expressing vector alone show only background staining for V5 since this tag is not in the construct (lower panels). Inset shows high magnification image of a single cell. Scale bar = 20 μm.

Overexpression of CPE-ΔN in Panc-1 Cells Increases Proliferation and Invasion
The effect of overexpression of 40 kDa CPE-ΔN in promoting proliferation and invasion of Panc-1 cells was investigated. Since the expression of exogenous CPE-WT protein was at a much higher level (at least 10 times higher) than CPE-ΔN, as shown in Figure 2B, in order to compare the efficacy of CPE-WT with CPE-ΔN, we transfected Panc-1 cells with CPE-WT and CPE-ΔN constructs at different plasmid concentrations, such that the expression levels of each of these proteins in the cell lysate were comparable. In the case of CPE-WT, more of the protein was present in the medium than in the cell lysate ( Figure 3A). Analysis using the MTT assay showed that Panc-1 cells overexpressing CPE-ΔN exhibited ~1.5-fold enhanced proliferation after five days compared to the control cells transfected with the vector alone. CPE-WT transfected cells demonstrated a ~2.2-fold increase in proliferation compared to empty vector control cells at day 5 ( Figure 3B). In the Matrigel invasion assay, Panc-1 cells overexpressing CPE-ΔN displayed significantly enhanced invasion (~2.3-fold)  Figure S2A,B (C) Representative confocal images of Panc-1 cells stably expressing CPE-WT-V5 and CPE-∆N-V5 or vector alone (control). Anti-V5 (red) and nuclear staining with DAPI (blue) are shown. Note V5 tag staining in the cells expressing CPE-∆N-V5 was localized in the nucleus overlapping with DAPI stain, as well as in the cytoplasm (top panels) while V5 staining in cells expressing CPE-WT-V5 was localized exclusively in the cytoplasm (middle panels). Panc-1 cells expressing vector alone show only background staining for V5 since this tag is not in the construct (lower panels). Inset shows high magnification image of a single cell. Scale bar = 20 µm.

Overexpression of CPE-∆N in Panc-1 Cells Increases Proliferation and Invasion
The effect of overexpression of 40 kDa CPE-∆N in promoting proliferation and invasion of Panc-1 cells was investigated. Since the expression of exogenous CPE-WT protein was at a much higher level (at least 10 times higher) than CPE-∆N, as shown in Figure 2B, in order to compare the efficacy of CPE-WT with CPE-∆N, we transfected Panc-1 cells with CPE-WT and CPE-∆N constructs at different plasmid concentrations, such that the expression levels of each of these proteins in the cell lysate were comparable. In the case of CPE-WT, more of the protein was present in the medium than in the cell lysate ( Figure 3A). Analysis using the MTT assay showed that Panc-1 cells overexpressing CPE-∆N exhibited~1.5-fold enhanced proliferation after five days compared to the control cells transfected with the vector alone. CPE-WT transfected cells demonstrated a~2.2-fold increase in proliferation compared to empty vector control cells at day 5 ( Figure 3B). In the Matrigel invasion assay, Panc-1 cells

CPE-ΔN Upregulates CXCR2 Expression in Panc-1 Cells
As CXCR2 expression was upregulated downstream of CPE-ΔN in the HCC cells [19], we speculated whether CXCR2, a chemokine receptor implicated in driving pancreatic cancer metastasis [17] could be regulated by CPE-ΔN in Panc-1 cells as a mechanism to promote the proliferation and invasion of these cells. We found that Panc-1 cells transfected with the CPE-ΔN construct for 48 h showed a ~2.2-fold increase in expression of CXCR2 mRNA ( Figure 4A) and a marked increase in the CXCR2 protein ( Figure 4B), while transfection of CPE-WT only marginally enhanced CXCR2 mRNA and protein levels (Supplementary Figure S4).

CPE-∆N Upregulates CXCR2 Expression in Panc-1 Cells
As CXCR2 expression was upregulated downstream of CPE-∆N in the HCC cells [19], we speculated whether CXCR2, a chemokine receptor implicated in driving pancreatic cancer metastasis [17] could be regulated by CPE-∆N in Panc-1 cells as a mechanism to promote the proliferation and invasion of these cells. We found that Panc-1 cells transfected with the CPE-∆N construct for 48 h showed a~2.2-fold increase in expression of CXCR2 mRNA ( Figure 4A) and a marked increase in the CXCR2 protein ( Figure 4B), while transfection of CPE-WT only marginally enhanced CXCR2 mRNA and protein levels (Supplementary Figure S4).
Next, we determined if the enhancement of proliferation by CPE-ΔN in Panc-1 cells was associated with the downstream expression of CXCR2. Indeed, on day 5 of the MTT assay, we observed a ~2.5-fold decrease in the proliferation of Panc-1 cells overexpressing CPE-ΔN following the suppression of CXCR2 levels by RNAi, when compared to similar cells treated with scrambled siRNA ( Figure 4C). These data suggest that CXCR2 mediates the tumorigenic functions of CPE-ΔN during pancreatic cancer progression.

CPE-∆N Induces Proliferation of Panc-1 Cells in a CXCR2-Dependent Manner
Next, we determined if the enhancement of proliferation by CPE-∆N in Panc-1 cells was associated with the downstream expression of CXCR2. Indeed, on day 5 of the MTT assay, we observed a~2.5-fold decrease in the proliferation of Panc-1 cells overexpressing CPE-∆N following the suppression of CXCR2 levels by RNAi, when compared to similar cells treated with scrambled siRNA ( Figure 4C). These data suggest that CXCR2 mediates the tumorigenic functions of CPE-∆N during pancreatic cancer progression.

Discussion
Most pancreatic cancers are diagnosed at advanced disease stages when the tumor has metastasized, leading to poor prognosis [23]. Elucidating molecules that contribute to the growth and metastasis of these cancers will facilitate the finding of therapeutic agents that can disrupt this process. In this study, we have demonstrated that CPE, especially a 40 kDa CPE-∆N isoform was important in promoting proliferation and invasion of pancreatic cancer cells. Northern blot analysis detected two CPE transcripts of 2.4 kb and 1.7 kb in size encoding CPE-WT and a 40 kDa N-terminal truncated CPE variant, CPE-∆N, respectively, in BXPC-3 pancreatic cancer cells, but only the 2.4 kb transcript was detectable in Panc-1, likely due to the low abundance or rapid turnover of the 1.7 kb transcript. Nevertheless, Western blot analysis showed expression of 40 kDa CPE-∆N protein in Panc-1 cells, as well as some CPE-WT in the secretion medium. BXPC-3 expressed both CPE-WT and 40 kDa CPE-∆N protein. Pull-down experiments indicated that 40 kDa CPE-∆N protein was recognized by two different antibodies. Analysis of the cell lysate and secretion media of Panc-1 cells suggested a low level of expression of CPE-WT protein in these cells relative to the large amounts of CPE mRNA present in the cells. Subcellular localization studies showed endogenous and transfected 40 kDa CPE-∆N in the nucleus as revealed by Western blot and immunocytochemistry. The nuclear localization of CPE-∆N in the HCC cells was consistent with its role in regulating the expression of several metastasis-related genes in these cells [19].
Gain-of-function studies determined the efficacy of CPE-WT versus CPE-∆N in promoting proliferation and invasion. A previous study showed that knockdown of endogenous CPE levels in BXPC-3 suppressed the proliferation, migration, and tumor-forming ability of these cells while showing improved sensitivity to the chemotherapeutic drug, cisplatin. Increased expression of CPE was found in tumor tissues from four pancreatic cancer patients. The study identified NF-κB as the probable downstream target through which CPE exerted its effects on pancreatic cancer [22]. Since the CPE siRNA used in this study knocks down both WT and the splice variant, it was difficult to attribute the functional deficits to either CPE-WT and/or CPE-∆N. Hence, we performed gain-of-function studies in Panc-1 cells, which expressed relatively low amounts of CPE compared to BXPC-3 (Figure 1), in order to distinguish the effects of CPE-WT and CPE-∆N in promoting PC growth and invasion. Our data indicated that the overexpression of approximately equivalent amounts of CPE-WT and CPE-∆N protein in Panc-1 cells, both significantly enhanced the proliferation of these cancer cells. However, while Panc-1 cells transfected with CPE-∆N exhibited a significant increase in invasion, cells transfected with CPE-WT expressed at a similar or even a higher level to CPE-∆N, if the secreted amounts were considered, and showed no increase in invasion compared to the control cells. Thus, CPE-∆N is an important player in promoting growth and invasion of pancreatic cancer cells.
Chemokines have been reported to contribute to the progression and metastasis of pancreatic cancer, especially pancreatic ductal adenocarcinoma (PDAC) [24]. Angiogenic CXC chemokines acting through the CXCR2 receptor are known to interact with myeloid cells in the tumor microenvironment and influence the cancer outcome [17,18]. CXCR2 signaling also acts downstream of the KRAS mutation or possibly in a feed-forward loop to drive the development of PDAC [16]. Besides pancreatic cancer, CXCR2 promotes tumorigenesis and metastasis in lung, breast, colon, and skin cancers [18,25]. Our recent study showed that CPE-∆N can enhance CXCR2 mRNA levels in HCC cells [19], which was the rationale for investigating if CXCR2 mediated the downstream effects of CPE-∆N on Panc-1 cells. We demonstrated that Panc-1 cells transiently transfected with CPE-∆N significantly enhanced expression of CXCR2 compared to control cells, and that CPE-∆N induced an increase in proliferation of Panc-1 cells, which was mediated by CXCR2. Hence, CPE-∆N may promote growth and invasion of PC cells, via CXCR2, and possibly in conjunction with other metastatic genes such as MMP3 and CCL12, found to be upregulated by CPE-∆N in HCC cells [19]. More studies, especially using in vivo models, are warranted to examine the mechanism of how CPE-∆N regulates CXCR2.
EMT (epithelial-mesenchymal transition) is thought to be the underlying mechanism that promotes metastatic dissemination and chemotherapeutic resistance associated with PC [26,27]. EMT occurs as an early event in a pre-metastatic lesion or PanIN during PC progression [27,28]. It is characterized by the downregulation of E-cadherin and miR200 levels, and activation of TGF-b signaling to increase the expression of transcriptional factors that regulate EMT such as ZEB1, Snail and Slug [29,30]. A recent study demonstrated that CPE-∆N enhanced the migration and invasion of human OS cells while promoting their epithelial-mesenchymal transition by activation of the Wnt/ β-catenin pathway [31]. Stable CPE-∆N expressing OS cell lines were used to demonstrate that CPE-∆N significantly downregulated E-cadherin expression while upregulated vimentin and transcription factors snail and slug, thus facilitating the EMT process in OS cells. It would be interesting to determine if CPE-∆N supports the EMT process in PDAC development, based on our data that it augments the invasive property of PC cells. Nearly 94% of PDAC have KRAS mutations, which drive the initiation of PanIN formation and subsequent development to malignant disease [32,33]. Besides PC, KRAS mutations are predominant in lung and colorectal cancers [34]. A few reports suggest that CPE/ CPE-∆N could be involved in promoting these cancers. One study investigated if CPE-∆N protein expression is valuable in early identification of recurring and metastatic lung adenocarcinoma. Out of the 95 patient samples that were analyzed for protein expression of CPE-WT and CPE-∆N using Western Blot and IHC, it was observed that while WT-CPE was expressed at similar levels in tumor tissue and peri-carcinoma tissues, CPE-∆N was prominently expressed only in the tumor tissue. Multivariate Cox regression analysis revealed that CPE-∆N expression was associated with lymph node and distant metastasis. Patients having CPE-∆N expression exhibited higher 3-year postoperative recurrence and metastasis rates than patients lacking CPE-∆N expression [20]. Enhanced CPE expression was documented in colorectal cancer (CRC) cell lines and patient tumor samples. Using two CRC cell lines, it was shown that overexpression of full-length CPE enhanced the proliferation and colony-forming ability of these cells, by increasing the percentage of cells in the S-phase of cell cycle, as determined by flow cytometry. In concurrence, CPE down-regulated the expression of CDK inhibitors p21 and p27 and increased cyclin D1 expression [8]. Hence, CPE-∆N and CPE-WT contribute to tumor progression and invasion, possibly by distinct mechanisms depending on the cancer type.
Considering the importance of CPE-∆N in regulating the growth and invasion of PC cells, it could be an effective therapeutic strategy to target CPE-∆N. However, there are some challenges. RNAi mediated suppression of CPE is not specific to CPE-∆N, and can affect CPE-WT as well. Designing specific inhibitors for CPE-∆N requires further understanding of how the synthesis of CPE-∆N is regulated at the transcriptional and epigenetic levels. Moreover, the mechanism of regulation of target genes by CPE-∆N remains unclear. Future research centered on addressing these questions will provide clues to achieve specific targeting of CPE-∆N. We propose that CPE-∆N inhibition could be used as a combination therapy to anti-KRAS therapy against PC. In conclusion, 40 kDa CPE-∆N has a powerful effect on promoting proliferation and invasion of Panc-1 cells, partly through the upregulation of the CXCR2 chemokine receptor, known to drive PDAC progression and metastasis. Thus, 40 kDa CPE-∆N is an excellent therapeutic target for the treatment of invasive cancer of the pancreas.

Cell Lines
Human pancreatic cell lines Panc-1 and BXPC-3 were obtained from ATCC (Manassas, VA, USA). The cells were cultured in a DMEM media (Millipore Sigma, Burlington, MA, USA), supplemented with 10% fetal bovine serum (Thermo Fisher Scientific, Waltham, MA, USA) at 37 • C in a humidified 5% CO 2 incubator.

Transfection of Cells with CPE-∆N and CPE-WT Constructs
Plasmid constructs containing CPE-WT or CPE-∆N sequences were generated with a pcDNA3.1 vector (Invitrogen, Carlsbad, CA, USA), and a V5-Tag was inserted into the C-terminal of both clones [19]. Transfection was performed using a Lipofectamine

Immunoprecipitation of CPE/CPE-∆N
Cell extract (~1 mg protein) prepared in TNE buffer (Tris 50 mM, NaCl 150 mM, EDTA 5 mM, pH 7.4) containing 1% NP-40 and 1X protease inhibitor cocktail, were incubated with rabbit polyclonal anti-CPE antibody 6135 (7 µg, generated in our laboratory) or control rabbit IgG (7 µg, Invitrogen) overnight at 4 • C, followed by incubation with Protein A/G PLUS-Agarose beads (Santa Cruz Biotechnology, Dallas, TX, USA) for 2 h. After thorough washing with lysis buffer to remove the non-specific binding, the protein-antibody complex was eluted from the beads by boiling in SDS (sodium dodecyl sulfate) protein gel loading solution (Quality Biological, Gaithersburg, MD, USA). Equal volumes of eluate were analyzed by Western blot using anti-CPE mouse monoclonal antibody (BD Biosciences, Franklin Lakes, NJ, USA).

MTT Cell Proliferation Assay
Panc-1 cells were seeded in a 96-well plate at a density of 2000 cells/well. An MTT (3-(4,5dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay was performed from days 1-4 or at 5 days, as reported previously [35]. Briefly, 25 µL of MTT reagent (5 mg/mL) (Sigma-Aldrich) was added to each well and incubated for 4 h at 37 • C in a CO 2 incubator. The supernatant was removed after incubation, and 150 µL of DMSO was added to each well. Five minutes later, the absorbance was measured at 490 nm or 450 nm in a microplate reader (BioTek, Winooski, VT, USA).

Invasion Assay
A cell invasion assay was performed in a 24-well Corning Matrigel invasion chamber (Corning, NY, USA) with 8 µm pores. Briefly, cell suspension containing 1 × 10 5 cells/ ml in serum-free media was added to the top chamber and serum-supplemented media to the lower chamber. After 24 h, cells that failed to invade through the pores were carefully removed using a cotton swab. The cells on the lower surface of the membrane were fixed with 100% methanol and stained with 1% crystal violet solution for 10 min. After removing the excess stain with water, images from 5 different fields within the well were captured, and the cells were counted.

Statistical Analysis
Data represent the mean of at least triplicate values (n) from independent experiments (N) as indicated in the figure legends. Significance was determined by a Student's t-test and p values are denoted as * p < 0.05, ** p < 0.01, *** p < 0.001. Error bars denote standard deviation (SD) or standard error of the mean (SEM). Acknowledgments: We thank Vincent Schram (NICHD Microscopy Core Facility) for his assistance in the confocal microscopy.

Conflicts of Interest:
The authors declare no conflict of interest.