Inhibiting YAP expression suppresses pancreatic cancer progression by disrupting tumor-stromal interactions

We obtained 72 pancreatic cancer samples and 20 normal pancreatic tissues from the Department of Hepatobiliary Surgery, the First Affiliated Hospital of Xi’an Jiaotong University between 2010 and 2014 after receiving approval from the Ethical Committee of Xi’an Jiaotong University. The pathological TNM status was assessed according to the criteria of the sixth edition of the TNM classification of the American Joint Commission on Cancer (AJCC). The pathological factors were examined by two pathologists. The results are summarized in Table 1. Immunohistochemical staining was performed using a SABC kit (Maxim, Fuzhou, China) according to the manufacturer’s instructions. Briefly, the tissue sections were incubated with primary antibodies overnight at 4 °C and incubated with the appropriate biotinylated secondary antibody for 30 min at room temperature, followed by 30 min of incubation with streptavidin peroxidase (Dako LSAB+HRP kit). After rinsing, the results were visualized using DAB, and the slides were counterstained with hematoxylin. The staining results were scored by 2 pathologists blinded to the clinical data as described previously [22]. The YAP staining status was evaluated according both nucleus and cytoplasm expression. Depending on the percentage of positive cells and staining intensity, YAP staining was classified into four groups: negative (0), weak (1+), moderate (2+) and strong (3+). Specifically, the percentage of positive cells was divided into five grades (percentage scores): < 10% (0), 10–25% (1), 25–50% (2), 50–75% (3), and 75% (4). The intensity of staining was divided into four grades (intensity scores): no staining (0), light brown (1), brown (2), and dark brown (3). YAP staining positivity was determined by the formula: overall scores = percentage score × intensity score. The overall score of ≤3 was defined as negative (0), of > 3 and ≤ 6 as weak (1+); of > 6 and ≤ 9 as moderate (2+), and of > 9 as strong (3+).

Human pancreatic cancer cell lines (AsPC-1, BxPC-3, CFPAC-1, Panc-1, and SW-1990) were purchased from the Chinese Academy of Sciences Cell Bank of Type Culture Collection (CBTCCCAS, Shanghai, China). According to the instructions, all cell lines were cultured in the proper medium (HyClone, Logan, USA) supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin and 100 μg/ml streptomycin. The cultures were incubated at 37 °C in a humidified atmosphere containing 5% CO₂. Recombinant human TGF-β1 was purchased from PeproTech (Rocky Hill, USA). Detailed information regarding the antibodies used in this study is presented in Additional file 1: Table S1. All reagents were stored as recommended by the manufacturer.

Pdx1-Cre mice, LSL-Kras^G12D mice and Trp53^fl/fl mice were purchased from the Nanjing Biomedical Research Institute of Nanjing University, Nanjing, China. The breeding of LSL-Kras^G12D/+; Pdx1-Cre (KC) transgenic mice was achieved by crossing LSL-Kras^G12D mice with Pdx1-Cre mice. LSL-Kras^G12D/+; Trp53^fl/+; Pdx1-Cre (KPC) mice were obtained by firstly crossing Trp53^fl/fl mice with Pdx1-Cre mice to generate Trp53^fl/fl; Pdx1-Cre offspring. Trp53^fl/fl; Pdx1-Cre mice were then crossed with LSL-Kras^G12D mice to generate KPC animals. All mice were housed under pathogen-free conditions and with free access to water and food. All experimental protocols were approved by the Ethical Committee of the First Affiliated Hospital of Medical College, Xi’an Jiaotong University, Xi’an, China.

YAP shRNA (shYAP) and negative control shRNA (shNC) in eukaryotic GV248 lentiviral vectors were purchased from GeneChem Co, Ltd. (Shanghai, China). The target sequence for YAP shRNA was CACCAAGCTAGATAAAGAA, and the negative control sequence was TTCTCCGAACGTGTCACGT. Cells were seeded at 1 × 10⁵ cells/well into 6-well plates 24 h prior to transfection. Transfection was carried out using lentiviral particles (Pan-1 MOI = 10; BxPC-3 MOI = 20), polybrene (5 μg/ml) and ENi.S according to the manufacturer’s protocol. Then, 12 h post-transfection, virus-containing medium was replaced with complete medium, and 96 h post-transfection, all cells were selected with puromycin (Merck, USA) at a final concentration of 5 μg/ml (Panc-1) or 4 μg/ml (BxPC-3) for 10 days. Cells were then maintained in 2.5 μg/ml (Panc-1) or 2 μg/ml (BxPC-3) puromycin. For the generation of stable transfected cells, media was changed three times a week. After 3 weeks, puromycin-resistant colonies were isolated for further study. The stable YAP-suppressed PCs and the Control PCs were named sh-YAP and sh-NC, respectively. The effect of gene silencing was evaluated by qRT-PCR and western blot.

For fluorescent immunocytochemistry, the pancreatic cancer cells and pancreatic stellate cells were fixed for 20 min in 4% paraformaldehyde in PBS, and the endogenous peroxidase activity was quenched with 3% hydrogen peroxide. The specimens were permeabilized with 0.3% Triton X-100 in PBS for 15 min on ice, pre-blocked for 60 min with bovine serum albumin (BSA) at room temperature, and incubated with primary antibody overnight at 4 °C. Staining was detected with the corresponding fluorescein-conjugated secondary antibodies (Jackson ImmunoResearch). Slides were mounted and examined using a Zeiss Instruments confocal microscope.

Cells (1000 shYAP or shNC cells) were seeded into a 35-mm petri dish and allowed to adhere overnight. Cells were further cultured for 2 weeks to allow colonies to form. At the indicated time point, colonies were fixed with 4% paraformaldehyde, stained with 0.1% crystal violet solution, rinsed and then imaged. The colonies > 0.5 mm in diameter were counted using a microscope (Nikon Eclipse Ti-S, Japan) at a magnification of 400 × .

Cell viability was measured using an MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay (Biotime). The cells were seeded into 96-well plates at a density of 5000 cells per well and incubated overnight in 10% FBS medium. After incubation for 24, 48, and 72 h at 37 °C and 5% CO₂, the cell proliferation rate was determined. Briefly, 20 μl of MTT solution (5 mg/ml in distilled water) was added to each well, and the cells were incubated for 4 h at 37 °C, after which the medium was removed. Then, 150 μl of DMSO was added, and the optical density (OD) was measured at 490 nm on a multifunctional microplate reader (POLARstar OPTIMA; BMG, Offenburg, Germany).

Total RNA was extracted using a Fastgen1000 RNA isolation system (Fastgen, Shanghai, China) according to the manufacturer’s protocol. Total RNA was reverse-transcribed into cDNA using a Prime Script RT reagent kit (TaKaRa, Dalian, China). Real-time PCR was used to quantitatively examine the expression of YAP, CTGF, E-cadherin, N-cadherin and Vimentin at the mRNA level. Real-time PCR was conducted according to a previous report [23]. The PCR primer sequences for YAP, CTGF, E-cadherin, N-cadherin, Vimentin and β-actin are shown in Additional file 1: Table S2. The expression level of each target gene was determined using β-actin as the normalization control. Relative gene expression was calculated using the 2^-ΔΔCt method [24].

Cells were conditioned in serum-free medium for 48 h. The culture media were collected and centrifuged at 1500 rpm for 5 min to remove particles, and the supernatants were frozen at − 80 °C until use. The production of TGF-β1 in the supernatants of PSCs was assessed by ELISA using a commercially available ELISA kit (R&D Systems, USA) according to the manufacturer’s recommendations.

Transwell chambers (pore size, 8.0 μm; Millipore, Billerica, USA) were coated with Matrigel (BD Biosciences, Oxford, UK). PCs were cultured in 6-well plates in medium containing 1% FBS for 24 h before drug treatment. PCs (200 μl, 5 × 10⁴) suspended in DMEM containing 1% FBS were seeded in the top chamber, and 500 μl of medium containing 10% FBS was placed in the lower chamber as a chemoattractant. The Transwell chamber was incubated for 48 h. The invaded cells on the bottom surface of the filter were fixed with methanol and stained with crystal violet (Boster Biological Technology Ltd., Wuhan, China). Cell migration and invasion were determined by counting the stained cells under a light microscope in 10 randomly selected fields.

Total protein was extracted using RIPA Lysis Buffer (Beyotime, Guangzhou, China), and the protein concentration was determined using a BCA protein assay kit (Pierce, Rockford, USA) according to the manufacturer’s instructions. Then, a western blot assay was performed as previously described. The primary antibodies used are listed in Additional file 1: Table S1. The protein expression was visualized by enhanced chemiluminescence (Millipore, USA). Images were captured using a ChemiDoc XRS imaging system (Bio-Rad, USA), and Quantity One image software was used for the densitometry analysis of each band; β-actin was used as an internal loading control.

Normal pancreatic tissues (1.0–1.5 g) obtained from patients undergoing a pancreatic partial resection for benign pancreatic conditions at the First Affiliated Hospital of Xi’an Jiaotong University were immediately collected in sterile ice-cold Hanks balanced salt solution (HBSS) containing 100 U/ml penicillin and 100 μg/ml streptomycin (Gibco). The histological diagnostic assessment of specimens was confirmed by pathologists. Human pancreatic stellate cells (PSCs) were isolated using a density gradient method as previously described. Isolated PSCs were maintained at 37 °C with 5% CO₂ in DMEM/F12 (HyClone, Logan, USA) medium supplemented with 10% heat-inactivated fetal bovine serum (FBS) (HyClone), 100 U/ml penicillin and 100 μg/ml streptomycin. PSCs were identified by oil red staining of intracellular fat droplets and immunofluorescence of α-smooth muscle actin (α-SMA). Cells cultured in the above medium conditions for 24 h were used in additional experiments.

After pancreatic cancer cells were cultured in media supplemented with 10% FBS and grown to 50% confluence, the media was changed to contain 1% FBS, 100 U/ml penicillin and 100 μg/ml streptomycin. Two days later, cancer cell conditioned medium was collected, centrifuged and filtered prior to incubation with isolated PSCs as previously described [3], and the PSCs were incubated with the conditioned medium for up to 2 days. For the direct PC-PSC co-culture model, PCs and PSCs were proportionally mixed (cell proportion, 2:1) and seeded into 6-well plates.

Mice were housed and maintained under specific pathogen-free conditions in facilities approved by the Animal Care and Use Committee guidelines of the Xi’an Jiaotong University, Shaanxi, China. Investigations were conducted in accordance with ethical standards, the Declaration of Helsinki and national and international guidelines and were approved by the authors’ institutional review board. The mice were used according to institutional guidelines when they were 6 to 8 weeks of age. Cells were resuspended in a 1:1 (v/v) mixture of culture medium and Matrigel (BD Biosciences, San Jose, CA, USA), and 1 × 10⁶ BxPC-3-shNC cells, 1 × 10⁶ BxPC-3-shYAP cells, 0.8 × 10⁶ BxPC-3-shNC cells + 2 × 10⁵ PSCs, or 0.8 × 10⁶ BxPC-3-shYAP cells + 2 × 10⁵ PSCs were injected s.c. into the right flank of nude mice. A total of 5 mice per group were used. After 8 weeks, the animals were sacrificed, and the subcutaneous tumors were isolated. Tumors were fixed in formalin as soon as possible and embedded in paraffin. Tumor volume was calculated as (length/2) × (width^2). Tumor samples were analyzed using H&E staining. Representative images were taken of each tumor using a light microscope at × 400 magnification.

To determine whether YAP is overexpressed at the protein level in human PDAC tissues, we examined YAP expression in five human pancreatic cancer tissues and the corresponding normal pancreatic tissue via western blot. The results show that the levels of YAP protein in pancreatic cancer tissues were significantly elevated compared with normal pancreatic tissues (Fig. 1A). To further confirm these results, pancreatic tissue sections from 72 patients identified as PDAC and 20 cases of normal pancreatic specimens were analyzed using immunohistochemistry (IHC). Intense IHC staining of YAP was detected in the cytoplasm and nucleus of cancer cells, but cells exhibited a predominantly nuclear localization pattern, whereas rare staining events were observed in normal pancreatic tissues (Fig. 1B). The statistics for YAP expression levels in different pancreatic tissue groups are shown in Table 1. Figure 1 shows representative pictures of negative (0; Fig. 1Bc), weak (1+; Fig. 1B d), moderate (2+; Fig. 1B e) and strong (3+; Fig. 1Bf) YAP staining in pancreatic cancer. As shown in Fig. 1B, no (Fig. 1Ba) or moderate (Fig. 1Bb) YAP immunoreactivity was observed in normal pancreatic tissues. YAP expression was significantly increased in PDAC compared to normal pancreatic tissues (P < 0.001; Fig. 1C). Notably, the χ2 analysis revealed that histologic markers of aggressive disease, including tumor–node–metastasis (TNM) stage (P = 0.010; Fig. 1D and E), and pM stage (P = 0.038) were significantly associated with YAP expression levels (Table 1). LSL-Kras^G12D/+; Pdx1-Cre (KC) and LSL-Kras^G12D/+; Trp53^fl/+; Pdx1-Cre (KPC) mice, in which Pdx1 induces the expression of mutant Kras alone or together with mutant Trp53 in murine pancreatic epithelium, fully recapitulate the pathogenesis of human PDAC and are generally regarded as two of the best genetically engineered mouse models (GEMMs) for human PDAC. Next, we detected the YAP expression in KC and KPC mice, and we found that YAP protein abundance was also markedly greater in pancreatic tissue from KC mice that had early and late pancreatic intraepithelial neoplasia (PanIN) or KPC mice that had fully established PDAC compared with wild-type mice (Fig. 1F). These findings indicated that YAP plays critical roles in PDAC development and progression and may be a valuable biomarker for this disease.

To determine the effect of YAP on PDAC growth, we analyzed its expression in PDAC cell lines. YAP was highly expressed in AsPC-1, BxPC-3 and Panc-1 cells but expressed at relatively low levels in CFPAC-1 and SW-1990 cells (Fig. 2a). Immunofluorescence showed that YAP was located in the nucleus and the cytoplasm of the BxPC-3 and Panc-1 cells (Fig. 2b). Accordingly, we defined BxPC-3 and Panc-1 cells as samples with differential expression for further experiments. The lentiviral vector YAP-shRNA was used to suppress YAP expression in two pancreatic cancer cell lines, BxPC-3 and Panc-1. From qRT-PCR and western blot results, we found that the YAP gene was significantly knocked down after YAP-shRNA transfection (Fig. 2c and d). At various time points (24, 48, and 72 h), the proliferation rate of BxPC-3-shYAP and Panc-1-shYAP cells were determined with an MTT assay. The results show that YAP knockdown inhibited the proliferation of both pancreatic cancer cell lines (Fig. 2e and f). Next, we detected the effect of YAP on the clone formation capability of Panc-1 and BxPC-3 cells, and the colony formation ability of both cell lines was clearly decreased after knockdown of YAP expression (Fig. 2g and h).

To elucidate the role of YAP in the invasive ability of pancreatic cancer cells, we knocked down YAP expression with the lentiviral vector YAP-shRNA in BxPC-3 cells and Panc-1 cells. Using Transwell chamber assays with Matrigel, a significant decrease in the invasion of shYAP cells was observed compared to shNC cells (Fig. 3a and b). Furthermore, western blot verified that knockdown of YAP resulted in a marked decrease in the expression of N-cadherin and Vimentin and a significant increase in E-cadherin (Fig. 3c and e), consistent with reversion to an epithelial phenotype. These observations were confirmed at the mRNA level using real-time PCR (Fig. 3d and f). Together, these data suggested that YAP might participate in the invasion process of PDAC by regulating EMT phenotypes.

Several studies have indicated that CTGF is a target of YAP. Consistent with our results, real-time PCR and western blot showed that CTGF was down-regulated after YAP knockdown (Fig. 4a and b). The IHC results also showed a positive correlation between YAP and CTGF staining in pancreatic cancer (Fig. 4c). We noted that both pancreatic cancer cells and pancreatic satellite cells stained positive for YAP, and YAP could be found in both the nucleus and the cytoplasm (Fig. 4c). YAP expression was positively correlated with α-SMA expression, a marker for activated stellate cells, and this correlation was also present in the KPC PDAC tissues (Fig. 4c). Next, we further investigated the effects of YAP on the tumor-stroma interactions between PCs and PSCs, and PC-PSC co-culture models were designed. The BxPC-3 has the highest YAP expression and this cell line was derived from the primary pancreatic cancer. So, we choose BxPc-3 for the further investigation of the interaction among YAP, PCs and PSCs. For the indirect co-culture model (Fig. 4d), PC supernatant was prepared after BxPC-3-shNC or BxPC-3-shYAP cells were cultured in conditioned medium (CM) for 24 h. Then, the conditioned medium mixture was added to the serum-free starvation-synchronized PSCs. After 48 h, PSC activation was determined by examining the α-SMA level using immunofluorescence labeling and western blot. As shown in Fig. 4g and h, the PSC activation level was reduced after incubation with BxPC-3-shYAP-CM compared with BxPC-3-shNC-CM, as revealed by α-SMA expression. Increased TGF-β1 synthesis and secretion is a hallmark of activated PSCs. The immunoblotting and ELISA results confirmed the inhibitory effect of BxPC-3-shYAP-CM on PSC activation (Fig. 4f and g). To investigate the effects of YAP on the activation of PSCs in a PC-PSC co-culture system (Fig. 4e), immunofluorescence was used to visualize the CTGF and PSC activation in the co-culture system. As shown in Fig. 4i, the immunofluorescence results indicate that the PSC activation level was reduced in indirect co-culture conditions after YAP knockdown in BxPC-3 cells, as revealed by α-SMA staining. Surprisingly, CTGF was mainly present in the PCs and was reduced after YAP knockdown in BxPC-3 cells, as revealed by CTGF staining with α-SMA that marked PSCs. Taken together, these data indicated that down-regulation of YAP expression in PCs inhibited PSC activation in a co-culture system, and this may be associated with a decrease in CTGF.

PSC activation has been recognized as a major driving force in the tumor microenvironment that promotes pancreatic cancer progression, and TGF-β1 plays an important role in tumor-stroma interactions. Therefore, we continued to elucidate whether TGF-β1 could reverse the YAP inhibition of pancreatic cancer invasion and EMT phenotypes. BxPC-3 cells were cultured for 48 h with or without TGF-β1. The results showed that the invasion ability (Fig. 5a and b) and the mesenchymal-related gene (N-cadherin and Vimentin) expression in BxPC-3-shNC cells were significantly increased after treatment with 10 ng/ml TGF-β1. However, simultaneous knockdown of YAP using shRNA abolished TGF-β1-induced cell invasion (Fig. 5a and b) and EMT (Fig. 5d). Studies have indicated that YAP can bind TGF-β1-activated SMAD complexes to control SMAD localization and activity in a variety of cell types, including mammary epithelial cells [25] and breast cancer cells [26]. The results showed that simultaneous knockdown of YAP using shRNA did not affect the p-SMAD2 level (Fig. 5f). However, the immunofluorescence results showed that SMAD2 nuclear location was significantly decreased when YAP was knocked down (Fig. 5c), and it also abolished TGF-β1-induced SMAD2 nuclear localization (Fig. 5c). We also found that YAP was significantly up-regulated after treatment with TGF-β1 in a time-dependent and dose-dependent manner (Fig. 5e and g), and this up-regulation was reversed after treatment with the TGF-β1 receptor inhibitor SB431542 (Fig. 5h). Taken together, our observations indicate that YAP is a critical mediator of TGF-β1-induced tumorigenic events, including EMT and cell invasion.

Based on the promising in vitro findings reported above, we next sought to test the role of YAP in pancreatic cancer cells on tumor progression and PSC activation in vivo. We established a subcutaneous pancreatic cancer xenograft model in nude mice through injection of PCs alone or PCs plus PSCs. The tumor volume was monitored, as shown in Fig. 6a, and we noted that BxPC-3-shYAP cells exhibited a notable reduction in tumor growth rates compared to BxPC-3-shNC cells. Moreover, co-injection of BxPC-3-shNC cells and PSCs resulted in a significant increase in tumor growth rates compared to BxPC-3-shNC cells alone. There was a significant induction of tumor growth when PSCs were co-injected with BxPC-3-shYAP cells in which the YAP expression was repressed (Fig. 6a and b). At the same time, we observed different histologic structures, especially in the stromal component, in the tumor tissues. The results showed that there were more stromal components in the BxPC-3-shNC+PSC co-injection group compared to the BxPC-3-shYAP+PSC co-injection group. Consistent with the in vitro studies, the IHC results showed that the proliferation of cancer cells was inhibited in the BxPC-3-shYAP group, as there was less and weaker expression of PCNA in the BxPC-3-shYAP group compared with the BxPC-3-shNC group (Fig. 6c). Moreover, the tumor tissues from mice in the BxPC-3-shYAP+PSC co-injection group exhibited lower staining levels of the stromal marker α-SMA compared with the BxPC-3-shNC+PSC co-injection group (Fig. 6c). Taken together, our in vivo findings indicate that YAP is critical in tumor growth and the desmoplastic reaction.