Pancreatic Fibroblasts Stimulate the Motility of Pancreatic Cancer Cells through IGF1/IGF1R Signaling under Hypoxia

Pancreatic ductal adenocarcinoma (PDAC) is characterized by its hypovascularity, with an extremely poor prognosis because of its highly invasive nature. PDAC proliferates with abundant stromal cells, suggesting that its invasive activity might be controlled by intercellular interactions between cancer cells and fibroblasts. Using four PDAC cell lines and two pancreas cancer-associated fibroblasts (CAFs), the expression of insulin-like growth factor-1 (IGF1) and IGF1 receptor (IGF1R) was evaluated by RT-PCR, FACScan, western blot, or ELISA. Correlation between IGF1R and the hypoxia marker carbonic anhydrase 9 (CA9) was examined by immunohistochemical staining of 120 pancreatic specimens. The effects of CAFs, IGF1, and IGF1R inhibitors on the motility of cancer cells were examined by wound-healing assay or invasion assay under normoxia (20% O2) and hypoxia (1% O2). IGF1R expression was significantly higher in RWP-1, MiaPaCa-2, and OCUP-AT cells than in Panc-1 cells. Hypoxia increased the expression level of IGF1R in RWP-1, MiaPaCa-2, and OCUP-AT cells. CA9 expression was correlated with IGF1R expression in pancreatic specimens. CAFs produced IGF1 under hypoxia, but PDAC cells did not. A conditioned medium from CAFs, which expressed αSMA, stimulated the migration and invasion ability of MiaPaCa-2, RWP-1, and OCUP-AT cells. The motility of all PDAC cells was greater under hypoxia than under normoxia. The motility-stimulating ability of CAFs was decreased by IGF1R inhibitors. These findings might suggest that pancreas CAFs stimulate the invasion activity of PDAC cells through paracrine IGF1/IGF1R signaling, especially under hypoxia. Therefore the targeting of IGF1R signaling might represent a promising therapeutic approach in IGF1R-dependent PDAC.


Introduction
Pancreatic ductal adenocarcinoma (PDAC) is one of the most lethal types of cancer, carrying an extremely poor prognosis because it is highly invasive and shows rapid progression [1][2][3].

Preparation of conditioned medium
We prepared conditioned medium (CM) from CAFs and pancreas cancer cells by seeding into 100 mm plastic dishes with 10 mL DMEM containing 10% FCS and then incubating for 3 days. To obtain CM, CAFs were washed with PBS and then incubated for an additional 3 days in 3 ml DMEM without FCS. Next, CM was collected from each dish. The supernatant was stored as CM at -20°C until use. All experiments were performed in medium contain 2% FCS. As a control, DMEM was used instead of CM.

Reverse-transcription polymerase chain reaction (RT-PCR)
Real-time RT-PCR was done on the ABI Prism 7000 (Applied Biosystems, Foster City, CA) using the commercially available gene expression assay for IGF1R(Hs00609566) and IGF1 (Hs01547656). Total cellular RNA was extracted from 4 pancreatic carcinoma cell lines with Trizol (Life Technologies, Gaithersburg, MD). As internal standard to normalize mRNA levels, amplification of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used. The threshold cycle (Ct) values were used for calculation of the relative expression ratios using the formula described by Pfaffl [28].

Flow cytometric analysis
The expression of IGF1R on PDAC cells was examined using a FACScan (BD LSR II; Becton Dickinson, San Diego, CA). Cells (2 x 10 6 cells/mL) were fixed with 2% paraformaldehyde and incubated in PBS with anti IGF1R antibody (ab16890, Abcam) or mouse IgG1-isotype control (ab91353, Abcam) for 30 min at 22°C. Cells were subsequently labeled with FITC-conjugated secondary antibody (1:500; ab96879, Abcam) for 30 minutes at 22°C. The percentage of positive cells were calculated and compared with isotype-matched control-stained cells.

Western blot analysis
For examining the effect of hypoxia on IGF1R expression, pancreas cancer cells were lysed 48 hours after incubation under normoxic or hypoxic conditions. Total protein (50 μg) was separated on polyacrylamide gels and transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA) with subsequent antibody (anti-IGF1R antibody; ab16890, Abcam, or anti-β-actin antibody; Cell Signaling Tec, Danvers, CO.), incubation. The bands were detected using an enhanced chemiluminescence system (Wako). Densitometry quantification was performed using ImageQuant software (Molecular Dynamics, Sunnyvale, CA) on a LAS 4000-mini Image Reader (GE Healthcare UK Ltd, Little Chalfont, UK).
Enzyme-Linked Immunosorbent Assay (ELISA) IGF1 in conditioned medium was quantified using IGF1 ELISA kit (R&D Systems). The cell number was counted when the CM was collected, and ELISA data was normalized by 1x10 6 cells in all cell lines.

Immunohistochemical staining of primary pancreatic tumors
A total of 120 patients who had undergone resection of a primary pancreatic tumor were included. The pathologic diagnoses were made according to the UICC Classification of Malignant Tumors [29]. The study protocol conformed to the ethical guidelines of the Declaration of Helsinki (1975). Sections of paraffin-embedded tissue were prepared. Immunohistochemical staining for IGF1R and carbonic anhydrase 9 (CA9) was performed using the avidin-biotinperoxidase complex method using primary monoclonal antibody against, IGF1R (Abcam) and CA9 (clone; M75, 1:1000, Novus Biologicals).

Immunohistochemical determination
All slides were examined by two of the authors who were blinded to clinical data. Scores for IGF1R were given for the staining intensity and the percentage of positive cells as follows: score of 0, no staining is observed, or is observed in less than 10% of the tumor cells; score of 1+, weak staining is detected in 10% or more of the tumor cells; score of 2+, moderate staining is observed in 10% or more of the tumor cells; and score of 3+, strong staining is observed in 10% or more of the tumor cells. CA-9 immunostaining was scored as 0 for 0%, 1+ for 1-20%, 2+ for 21-50%, and 3+ for >50%, Scores of 0 and 1+ were considered to be negative.

Wound healing assay
Pancreas cancer cells were plated on 96-well plates (Essen ImageLock, Essen Instruments, Birmingham, UK) and a single wound per well was scratched with wound scratcher (Wound Maker, Essen BioScience, MI, USA). Compounds and appropriate controls were added after wound scratching and wound confluence was monitored with Incucyte Live-Cell Imaging System and software (Essen Instruments). Each pancreatic carcinoma cell lines were resuspended to a final concentration of 1.0x10 5 cells/ml in DMEM with 2% FBS. One hundred microliters of cancer cell suspension and 100 μl of DMEM with 2% FBS with IGF1 (10 ng/ml), and CM in the absence or presence of IGF1R antibody (2 μg/ml) and PPP (0.25 μM) were added. Wound closure was observed every 3 hours for 24 h. Wound closure was determined as a percentage of wound confluence. The mean of 4 fields was calculated as the sample value.

Invasion assay
The in vitro invasiveness was measured by two-chamber matrigel invasion assay, as previously reported [30]. We used the chemotaxis chambers with a 8 μm-pore membrane filter (Kubota, Osaka, Japan) coated with 50 μg of matrigel in a 24-well culture plate. Pancreas cancer cells (2x10 3 cells/chamber) were seeded in upper chambers, and CM in the absence or presence of IGF1R antibody (2 μg/ml) and PPP (0.25 μM) were added. After 24 h incubation, cancer cells that invaded through a filter coated with matrigel to the lower surface of the membrane were manually counted under a microscope.

Statistical analysis
Data were analyzed using Student's t test. The correlation between the expression level of IGF1R and CA-9 was analyzed by Spearman's rank correlation analysis using SPSS 13.0 (SPSS Inc, Chicago). A P value less than 0.05 was considered statistically significant.
Association between expression of IGF1R and CA9 in 120 pancreatic specimens Fig 2A and 2B shows representative picture of CA9 staining and IGF1R staining of PDAC cells. Of 120 patients with PDAC, 41 (34%) were positive for CA9 expression and 54 (45%) were positive for IGF1R overexpression. A significant correlation was found between IGF1R and CA9 immunoreactivity in PDAC cells (p < 0.01, rs = 0.649; Spearman's rank sum test) (Fig 2C). Effect of conditioned medium from pancreas cancer-associated fibroblasts on motility of PDAC cells Effect of IGF1 on motility of PDAC cells IGF1 significantly increased the number of migrating RWP-1, MiaPaCa-2, OCUP-AT, and Panc-1 cells under hypoxia. Under normoxia, the migratory activity of RWP-1, OCUP-AT, and MiaPaCa-2 cells was significantly increased by IGF1, but that of Panc-1 was not (Fig 3B,  3C and 3D).

Effect of signaling inhibitors on migration-stimulating activity of CM from pancreas cancer-associated fibroblasts
The migration-stimulating ability of CM was inhibited by the IGF1R inhibitor, PPP, in RWP-1, MiaPaCa-2, OCUP-AT, and Panc-1 cells under hypoxia. In contrast, PPP significantly decreased the migration-stimulating activity of CM only in MiaPaCa-2 cells under normoxia (Fig 3B, 3C and 3D). Under hypoxia, IGF1R-neutralizing antibody decreased the migrationstimulating activity of CM from fibroblasts in RWP-1, MiaPaCa-2, OCUP-AT, and Panc-1 cells. Under normoxia, the migration-stimulating activity of fibroblasts was decreased by anti-IGF1R antibody only in MiaPaCa-2 cells.  activity of CM was decreased in the presence of IGF1R-neutralizing antibody or PPP. CM from pCaF-1 significantly stimulated the invasion ability of all 4 pancreas cancer cell lines. IGF1R inhibitor, IGF1R-neutralizing antibody and PPP, significantly inhibited invasion-stimulating activity of CM from fibroblasts (Fig 4B).

Discussion
PDAC is clinically characterized as hypovascular tumors, and these cells are continuously exposed to hypoxia. Intraoperatively Koong et al. measured tumor oxygenation in cases of PDAC, and revealed that a significant proportion of these cells were hypoxic [17]. Thus, PDAC cells can survive in hypoxia, as evidenced by the fact that PDAC cell lines were grown under hypoxia (1% O 2 ) in vitro in this study. Hypoxia increased the expression level of IGF1R in PDAC cells in comparison with that under normoxia. CA9, a hypoxia-associated endogenous protein, is considered as a cellular biomarker of hypoxic regions in solid tumors [31,32]. It has been reported that IGF1 is upregulated in PDAC tissues, but not in surrounding noncancerous tissues [33]. IGF1R is phosphorylated solely in PDAC tissues. In this histological study, the expression of CA9 was closely correlated with IGF1R expression. Using immunohistochemical investigation, we previously reported that PDAC patients with IGF1R overexpression had poor survival rates [22], suggesting that IGF1R signaling might be correlated with tumor aggressiveness in this cancer. These findings suggest that a hypoxic tumor microenvironment might affect the aggressiveness characteristics of PDAC cells via the IGF1R signaling system.
Fibroblasts have the ability to produce certain cytokines that influence neighboring cells, including malignant cells [34]. PDAC often has abundant stroma, which suggests that interactions between PDAC cells and stromal cells may play a critical role in the progression of this cancer [7,35]. There is increasing evidence that activated fibroblasts play a pivotal role in the development of PDAC [36,37]. In the cancer microenvironment, fibroblasts are transformed ("activated") from their quiescent phenotype into myofibroblast-like cells which express αSMA [36,38,39]. In the present study, two human pancreatic CAF cell lines, pCaF-1 and pCaF-2, were established from PDAC specimens, and αSMA expression was evaluated in both these cell lines. These findings suggest that both pCaF-1 and pCaF-2 are activated fibroblasts. To clarify whether hypoxia affects cancer cell-stromal cell interaction with regard to the migratory ability of pancreatic cells, using pCaF1 and pCaF2 cell lines, we examined the effect of CM from pancreas CAFs. We observed that CM from pancreas CAFs stimulated the motility of MiaPaCa-2, RWP-1, and OCUP-AT cells under normoxia, and the migration-stimulating activity of fibroblasts on PDAC cells was greater under hypoxia than under normoxia. These findings suggest that hypoxia markedly enhances the migration-stimulating activity of pancreas CAFs on PDAC cells. We then analyzed the mechanisms responsible for this migrationstimulating activity.
Previous studies revealed that increased production of several growth factors in human stromal cells under hypoxia plays a key role in the invasiveness of cancer cells [18,34,40]. Pancreas CAFs produced IGF1, and its level increased under hypoxia. The invasion-stimulating ability of CM was decreased by the IGF1R inhibitor PPP and anti-IGF1R-neutralizing antibody, but not by PDGF-neutralizing antibody and TGFβ-neutralizing antibody (data not shown). IGF1 is produced from CAFs, but not from PDAC cells. These findings suggest that IGF1R signaling in PDAC cells under hypoxia may respond to IGF1 in a paracrine manner. Altogether, hypoxia increased the IGF1 expression level of pancreas CAFs, suggesting that hypoxia stimulates not only IGF1R in PDAC cells but also the IGF1 expression level of pancreas CAFs, which could result in the synergistic stimulation of migratory activity in PDAC cells in paracrine IGF1/ IGF1R signaling. To our knowledge, this is the first study to find an association for IGF1/ IGF1R signaling in PDAC cell-stromal cell interaction in a hypoxic cancer microenvironment.
The migration-stimulating activity of conditioned medium from fibroblasts in PDAC cells was partly inhibited by neutralizing IGF1R antibody. Not only IGF1 but also other factor(s) might affect on the hypoxia-induced migration in PDAC cells. In future studies, it will be necessary to examine other factors from fibroblasts which can affect on the migration of PDAC cells in hypoxia.
In this study, PPP and IGF1R antibody suppressed the motility of PDAC cells. Consequently, multiple phase 1-2 clinical trials testing IGF1R inhibitors in a diverse number of epithelial cancers, including PDAC, are currently ongoing. IGF1R inhibitors are available and some are undergoing phase trials [41,42]. These findings suggest that IGF1R-targeted therapy may be useful with regard to its inhibitory effects on the motility of pancreatic tumor cells. Targeting of the pancreatic tumor hypoxic microenvironment may, therefore, represent a promising therapeutic approach in PDAC.
A high incidence of somatic K-ras mutations is found in some types of cancer including PDAC [43,44]. It has been reported that K-ras mutation is predictive of a very poor response to signal inhibitors which inhibit the downstream effector pathways of K-ras, such as the mitogen-activated protein kinase (MAPK) and phosphoinositide 3-kinase (PI3K) signaling pathways in colorectal cancer [45][46][47][48]. Although all four PDAC cell lines used in this study have K-ras mutation at codon 12 [49], the efficacy of IGF1R inhibitors against pancreatic carcinoma is independent of K-ras mutation status, and thus IGF1R inhibitors might be clinically useful in PDAC patients irrespective of K-ras status in PDAC patients. We observed a significantly higher expression level of IGF1R in RWP-1, MiaPaCa-2, and OCUP-AT cells than in Panc-1 cells, in which migratory activity was not increased by CM from pancreas fibroblasts. These findings indicate that aberrant IGF1R signaling may be involved in PDAC progression, and the expression level of IGF1R in PDAC may represent a predictive biomarker for response to IGF1R inhibitors. Because IGF1R represents a key molecule in the hypoxic tumor microenvironment, IGF1R inhibitors appear therapeutically promising in PDAC characterized by IGF1R overexpression.
In conclusion, we identified a motility-stimulating role of pancreas CAFs in hypoxic PDAC cells through the paracrine IGF1/IGF1R pathway. Targeting IGF1R signaling may represent a promising therapy against the progression of pancreatic carcinoma.