Tumor Biol. DOI 10.1007/s13277-015-4353-7

ORIGINAL ARTICLE

Integrin β6 acts as an unfavorable prognostic indicator and promotes cellular malignant behaviors via ERK-ETS1 pathway in pancreatic ductal adenocarcinoma (PDAC) Zequn Li 1,2 & Pengfei Lin 1,2 & Chao Gao 1 & Cheng Peng 1 & Song Liu 3 & Huijie Gao 1 & Ben Wang 1 & Jiayong Wang 1 & Jun Niu 1 & Weibo Niu 1

Received: 9 August 2015 / Accepted: 30 October 2015 # International Society of Oncology and BioMarkers (ISOBM) 2015

Abstract Pancreatic ductal adenocarcinoma (PDAC) remains one of the most deadly cancers and is expected to become the second leading cause of cancer death by 2030. Despite extensive efforts to improve surgical treatment, limited progress has been made. Increasing evidence indicates that integrin β6 plays a crucial role in carcinoma invasion and metastasis. However, the expression and role of β6 in PDAC remain largely unknown. In the present study, we investigated the expression of β6 in PDAC and its potential value as a prognostic factor and therapeutic target. β6 upregulation was identified as an independent unfavorable prognostic indicator. Integrin β6 markedly promoted the proliferation and invasion of pancreatic carcinoma cells and induced ETS1 phosphorylation in an ERK-dependent manner, leading to the upregulation of matrix metalloprotease-9, which is essential for β6mediated invasiveness of pancreatic carcinoma cells. Accordingly, small interfering RNA-mediated silencing of integrin β6 markedly suppressed xenograft tumor growth in vivo. Taken together, our results suggest that integrin β6 plays important roles in the progression of pancreatic carcinoma and contributes to reduced survival times, and may serve as a novel therapeutic target for the treatment of PDAC. Zequn Li and Pengfei Lin contributed equally to this work. * Weibo Niu [email protected] 1

Department of General Surgery, QiLu Hospital of Shandong University, Jinan 250012, Shandong, China

2

Key Laboratory of Cardiovascular Remodeling and Function Research, Chinese Ministry of Education and Public Health, Jinan 250012, Shandong, China

3

Department of Thyroid & Breast Surgery, Affiliated Hospital of Binzhou Medical College, Binzhou 256603, Shandong, China

Keywords Integrin β6 . PDAC . Prognostic indicator . Invasiveness . ETS1

Introduction Pancreatic ductal adenocarcinoma (PDAC) is a highly malignant carcinoma with extremely poor prognosis and its incidence has increased recently in many countries [1, 2]. Currently, it is the fourth most common cause of cancer-related death and is expected to rise to the second by 2030 [3]. PDAC has the worst prognosis among all digestive malignancies, with a 5-year survival rate of less than 5 % [4]. Most patients are not eligible for curative surgical resection at diagnosis because of tumor infiltration into adjacent tissues and metastasis to distant organs in the early stages, which may account for the extremely poor prognosis of the disease. The early metastatic potential of PDAC also leads to limited therapeutic options [5, 6]. Thus, it is critical to explore the detailed mechanisms underlying the aggressive growth and metastatic properties of PDAC. Tumor invasion and metastasis are multistep processes involving a variety of cell surface receptors that mediate interactions with the extracellular matrix (ECM). The integrin family of adhesion molecules plays an important role in this process [7, 8]. One integrin in particular, alphavbeta6, is not detectable in fully differentiated epithelia but is strongly induced during wound healing, inflammation and carcinogenesis [7, 9]. Enhanced or de novo expression of the β6 subunit has been observed in several epithelial malignancies, such as gastric carcinoma, colorectal carcinoma, breast carcinoma, ovarian carcinoma, and oral squamous cell carcinoma [9–15]. There is increasing evidence to suggest that elevated expression of integrin β6 may be associated with tumor invasion and metastasis [8, 16, 17]. In a previous study from our group, β6-

Tumor Biol.

mediated gelatinase B secretion was shown to be important for the progression of colon cancer, and transfection of the antisense β6 gene remarkably inhibited the multiplication of colon cancer cells in vitro and their capability of tumor formation in vivo [17, 18]. Bates et al. confirmed that high β6 expression in human colon carcinoma is a prognostic indicator of poor survival. We also found that positive β6 expression is linked to significantly reduced survival times in gastric carcinoma [15, 19]. Recently, we identified a direct link between extracellular signal-regulated kinase (ERK2) and β6 integrin in colon cancer cells, and blockage of this direct binding significantly impaired cytosolic ERK/MAP kinase activity, cell density-dependent expression of β6, and β6-mediated matrix metalloprotease (MMP)-9 secretion [12, 18, 20]. Our previous research has already studied the effect of αvβ6 gene silencing by RNA interference in PDAC cells [21]. However, little information is available on its role and related mechanisms in PDAC. In the present study, we investigated the expression of β6 in PDAC tissues by immunohistochemistry (IHC) in relation to various clinicopathologic parameters and patient survival to determine the value of β6 as a prognostic predictor. The effects of β6 on cell proliferation and invasion and the underlying mechanisms were investigated to determine the role of β6 in regulating the malignant behaviors of cells and its potential as a therapeutic target. β6 upregulation was identified as an independent unfavorable prognostic indicator in PDAC. Integrin β6 markedly promoted the proliferation and invasion of pancreatic carcinoma cells via the ERK-ETS1 signaling pathway, and small interfering RNA (siRNA)-mediated β6 silencing markedly suppressed xenograft tumor growth in vivo. Taken together, these data indicate that β6 may serve as a promising target for the treatment of PDAC.

Materials and methods Antibodies and reagents Mouse anti-human monoclonal antibody (mAb) against β6 (2G2) was a kind gift from Biogen Idec, Cambridge, MA, USA. The anti-β6 mAb E7P6 and the functional-blocking mAb 10D5 were purchased from Chemicon International, Temecula, CA, USA. Mouse immunoglobulin IgG1 and IgG2a were obtained from Dako, Glostrup, Denmark. Antibodies against ETS-1 and ETS-1 (phospho-Thr38) were purchased from Assay Biotechnology, Sunnyvale, CA, USA. Antibodies against phosphorylated and total ERK1/2 (p-ERK1/2 and tERK1/2) were from Cell Signaling Technology, Danvers, MA, USA. These antibodies are well characterized and have been used in many previous studies [10, 15–19]. The MEK inhibitor PD98059 was from Promega, and primaquine was from Sigma Chemical Co. (St. Louis, MO, USA).

Patients and tissue specimens Formalin-fixed, paraffin-embedded tumor specimens were obtained from 78 consecutive PDAC patients who underwent surgical resection between 2002 and 2010 at the Department of General Surgery, Qilu Hospital of Shandong University. The patients consisted of 45 men and 33 women ranging in age from 29 to 80 years, with a mean age of 59.6 years. The surgical procedures were as follows: 48 cases were treated by conventional pancreatoduodenectomy, 14 were treated by pylorus-preserving pancreatoduodenectomy, and 16 were treated by distal pancreatectomy. The size of resected tumors ranged between 1 and 8.4 cm (median, 3.3 cm). Lymph node dissection was performed in all cases, and lymph node metastasis was positive in 48 cases. None of the patients had received any sort of chemotherapy or radiotherapy prior to surgery. Ten normal autopsy pancreatic tissues were used as controls. The H&E-stained slides of each case were obtained from the pathology archives and reviewed by two trained pathologists to confirm the histological diagnosis and pathological grade without knowledge of the patients’ outcome. The clinical records of each patient were also reviewed. The following data were collected: the age and gender of patients, tumor location, tumor size, histological type, TNM stage, and presence of lymph node and distant metastasis. The tumors included in this study were histologically classified using a grading system proposed by the WHO and staged according to the International Union Against Cancer (UICC) TNM classification system (sixth edition, 2002). In total, 17 cases were well differentiated, 36 were moderately differentiated, and 25 were poorly differentiated. Twelve patients were classified as stage I, 56 as stage II, four as stage III, and six as stage IV. After surgery, the patients were followed up in the surgical outpatient clinic. Follow-up data until June 2008 provided information concerning survival time and performance status at the last visit. Written informed consent was obtained from all patients and the study was approved by the Institutional Ethics Committee of Shandong University. Immunohistochemistry Immunohistochemical staining on the whole series was performed on 5-μm sections using a standard streptavidin–biotin immunoperoxidase method. Briefly, the slides were deparaffinized and rehydrated, and endogenous peroxidase activity was blocked with 3 % hydrogen peroxide solution. Antigen retrieval was performed by incubation of the slides in a 0.4 % pepsin solution at 37 °C for 15 min. After rinsing in phosphate-buffered saline (PBS), 10 % bovine serum was applied for 10 min to block nonspecific binding. Sections were then incubated with primary antibody 2G2 (1.67 μg/ml; Biogen Idec, Cambridge, MA, USA; dilution: 1:500) overnight at 4 °C. For the negative controls, the primary antibody

Tumor Biol.

was replaced with an identical concentration of mouse immunoglobulin IgG1 (Dako). The following day, after washing in PBS, biotinylated anti-mouse IgG (1:200; Dako) was applied to the slides for 30 min. After rinsing again in PBS, sections were treated with horseradish peroxidase-labeled streptavidin complex for 15 min. To visualize the peroxidase reaction, a 0.05 % solution of diaminobenzidine containing 0.01 % hydrogen peroxide in a 0.05 M Tris buffer (pH 7.6) was applied. Then the slides were counterstained with hematoxylin, dehydrated, cleared, and mounted. Paraffin-embedded sections of SW-480/β6 cell line pellets were used as the positive control for β6 [22]. Immunostaining evaluation The stained slides were evaluated by two independent observers in a blind fashion without knowledge of the clinicopathologic data. The concordance ratio was >95 %. Differences of opinion were resolved by discussion to reach a consensus. Brown staining of the cell membrane or cytoplasm indicated positivity for β6 in tumor cells. The staining results were scored using previously described parameters and semiquantitative criteria [23, 24]. In brief, the intensity of staining was classified into four-point levels as follows: 0, no staining; 1, weak staining; 2, moderate staining; 3, strong staining. In addition, the percentage of positively stained tumor cells was scored as follows: 0, 0 %; 1, <20 %; 2, 20–50 %; 3, >50 %. Both the intensity and the proportion of positive cells were considered to give a semiquantitative estimate of the expression level of β6 in each section. A combined score of 0–6 was derived by adding the intensity and proportion scores. The staining class was defined according to the final score as follows: 0–1, negative; 2, weakly positive (1+); 3–4, moderately positive (2+); and 5–6, strongly positive (3+). To perform statistical evaluations, the following subdivision was made: the staining class was considered negative/weak (0/1+) versus moderate/strong (2+/3+) positive. Cell lines and cell culture The human PDAC cell lines PANC-1, Capan-2, and CFPAC-1 were obtained from the American Type Culture Collection (ATCC; Rockville, MD, USA). The PANC-1 and Capan-2 cells were maintained in Dulbecco’s Modified Eagle’s Medium (DMEM, Invitrogen, Waltham, MA, USA) supplemented with 10 % fetal bovine serum (FBS, Sigma Chemical Co.). CFPAC-1 cells were maintained in Iscove’s Modified Dulbecco’s Medium (IMDM) with 10 % FBS. All cells were incubated in a humidified (37 °C, 5 % CO2) incubator and passaged upon reaching 80 % confluence. Both PANC-1 and Capan-2 cells were transfected with a β6 gene construct in

antisense orientation, and stable transfectants were selected continuously in puromycin as previously described [20]. RNA isolation, RT-PCR, and real-time quantitative PCR Total cellular RNA was extracted using the Trizol reagent (Sigma), and cDNAwas synthesized according to the manufacturer’s instructions (Promega, Fitchburg, WI, USA). Equal amounts of cDNA were subjected to PCR or real-time quantitative PCR analysis. The sequences of the primers were as follows: β6 integrin forward: 5′-AGGATAGTTCTGTTTCCTGC-3′, and reverse: 5′-ATCATAGGAATATTTGGAGG-3′; β-actin forward: 5′-GAGACCTTCAACACCCCAGCC-3′, and reverse: 5′A ATG TC AC G CA C G ATTTC C C -3′ ; GA PDH : 5′AACGGATTTGGTCGTATTGGG-3′, and reverse 5′CCTGGAAGATGGTGATGGGAT-3′; MMP-2 forward: 5′TGATCTTGACCAGAATACCATCGA-3′, and reverse: 5′GGCTTGCGAGGGAAGAAGTT-3′; MMP-3 forward: 5′CGGTTCCGCCTGTCTCAAG-3′ and reverse: 5′CGCCAAAAGTGCCTGTCTT-3′; MMP-9 forward: 5′ACCTCGAACTTTGACAGCGAC-3′, and reverse: 5′GAGGAATGATCTAAGCCCAGC-3′. RT-PCR was performed using 2× Taq MasterMix (CWBIO, Beijing, China). Real-time PCR was performed using UltraSYBR Mixture (CWBIO). Relative levels of gene expression were determined with GAPDH as a control. Western blotting The TCM from each sample was collected and concentrated 50-fold using a 10-kDa cut-off Microcon filter (Amicon, Beverly, MA, USA). Protein concentrations were determined using the BCA protein assay kit (Sigma). Aliquots containing 30-μg protein were separated by 10 % SDS/PAGE and then transferred to nitrocellulose membranes (Amersham Pharmacia Biotech, Amersham, UK). The membranes were incubated with mouse monoclonal anti-MMP-9 antibody (MAB911, 1 μg/ml, R&D Systems, Minneapolis, MN, USA) overnight at 4 °C followed by horseradish peroxidaseconjugated secondary antibody. Immunoreactive blots were visualized using the ECL chemiluminescence detection system (Amersham Pharmacia Biotech) according to the manufacturer’s instructions. Cell proliferation assay A total of 3,000 cells were seeded in 96-well plates with triplicate wells and cultured for the indicated times. Cell viability was evaluated using the CCK8 (Beyotime, Haimen, China) assay according to the manufacturer’s instructions. The absorbance was determined at 450 nm. Each time point was repeated in three wells and the experiment was independently performed at least three times.

Tumor Biol.

In vitro invasion assay

Xenograft tumor model

The invasive capability of pancreatic carcinoma cells was determined using Transwell cell culture chambers (8-μm pore size polycarbonate membrane, Costar, Cambridge, MA, USA). Briefly, the membrane was coated with Matrigel™ (BD Biosciences, Bedford, MA, USA) diluted at a 1:3 ratio with medium. Cells (1×105) with different treatments in 100 μl of serum-free medium were seeded into the upper chamber, while 500 μl of complete medium containing 10 % FBS was placed in the lower chamber as a chemoattractant. After incubation at 37 °C for 12 h, the filters were fixed in 4 % paraformaldehyde for 10 min and stained with crystal violet for 20 min. The cells on the upper side of the filter were wiped off gently by a cotton swab. Cells that had invaded through the membrane to the lower filter surface were counted in five random microscopic fields.

BALB/C nude mice (4–6 weeks old) were obtained from the Chinese Academy of Sciences (Shanghai, China) and maintained in laminar-flow cabinets under specific pathogen-free conditions. To evaluate tumor growth in vivo, 1×107 Panc-1 cells were injected subcutaneously into the flanks of nude mice (n=10). Ten days after subcutaneous inoculation, mice were divided randomly into two groups (six mice/group) and treated by intratumoral injection of β6 siRNA or nontargeting siRNA complexed with transfection reagent in vivo jet PEI (Polyplus-transfection Inc., New York, USA) as described previously [25]. The intratumoral injection was carried out every other day for 12 days. The tumor size was measured and tumor volume was calculated as follows: Longer diameter×Shorter diameter2 ×0.5. The tumor volume data are presented as the mean±SD (n=6). Two days after the last siRNA injection, the mice were sacrificed and the tumors were weighed. All of the animal studies were conducted using a protocol approved by the Institutional Animal Care and Use Committee at the School of Medicine, Shandong University (Jinan, Shandong, China).

MMP activity assay A total of 2×105 cells were cultured in a six-well cell culture plate and then treated with 10D5, IgG2a, PD98059, siRNA, or plasmid for 24 h. The levels of secreted MMP-9 in the culture supernatant were collected and subjected to enzyme-linked immunosorbent assay (ELISA) following the manufacturer’s guidelines (R&D). Samples were assayed in triplicate and calibrated against a standard curve.

Fig. 1 Immunohistochemical staining of β6 in pancreatic ductal adenocarcinoma tissues. a Positive staining was observed in the cytoplasm or cell membrane of tumor cells (original magnification ×400). b Normal pancreatic epithelium displayed no staining for β6. c More intense staining of β6 at the invading edges of the tumor. d More intense staining of β6 in the cells infiltrating the stroma (original magnification ×200)

Statistical analysis The association between β6 expression and clinicopathologic parameters was analyzed by the chi-square test and Fisher’s exact test. The overall survival curves were

Tumor Biol.

constructed using the Kaplan–Meier method and statistical differences were calculated by the log-rank test. Univariate and multivariate survival analyses were performed using the Cox proportional hazards models to identify independent prognostic factors. Quantitative data were presented as the mean±SD. The statistical significance of differences was determined by Student’s two-tailed paired t test in two groups and one-way ANOVA in multiple groups. All statistical analyses were performed using SPSS 16.0 software package (SPSS Inc., Chicago, IL, USA) and a P value< 0.05 was considered statistically significant.

Table 1 Association of αvβ6 expression and clinicopathologic parameters αvβ6 expression Parameters

n

(0/1+)

(2+/3+)

No. of patients

78

43

35

42 36

22 21

20 15

P value

Age ≤60 >60 Gender Female

33

18

15

Male

45

25

20

0.5982

0.9294

Results

Tumor location Head

62

32

30

16

11

5

0.2192

Expression of β6 in human PDAC tissues

Body–tail Tumor Size ≤3 cm >3 cm Grade Well Moderate

32 46

22 21

10 25

0.0436*

17 36

11 22

6 14

25

10

15

β6 staining was mainly detected in the cytoplasm or membrane of tumor cells by IHC (Fig. 1a). The normal pancreatic epithelium generally displayed no staining for β6 (Fig. 1b). Of 78 carcinoma specimens, 42 (53.8 %) were positive for β6 expression, seven (9.0 %) were weakly positive (1+), 16 (20.5 %) were moderately positive (2+), and 19 (24.3 %) were strongly positive (3+). A characteristic staining pattern was observed in carcinomas, with more intense staining at the invading edges of the tumor (Fig. 1c) and the cells infiltrating into the stroma (Fig. 1d), and relatively weaker staining intensity detected centrally within the tumor masses and nests. Statistical analysis showed that moderate/strong (2+/3+) staining scores of β6 were significantly correlated with larger tumor size (P=0.0436), more advanced TNM stage (P= 0.0072) and lymph node metastasis (P=0.0368); however, other parameters were not significantly associated with β6 staining scores. Table 1 summarizes the association of β6 expression with various clinicopathologic features of the patients.

Poor TNM stage I

0.1769

12

10

2

II 56 III/IV 10 Lymph node metastasis Negative 30 Positive 48 Distant metastasis Negative 72 Positive 6

36 2

20 8

0.0072*

21 22

9 26

0.0368*

41 2

31 4

0.2639

*P<0.05 was considered statistically significant

Prognostic significance of β6 expression Survival data were analyzed by considering only diseaserelated death as an event, censoring deaths unrelated to disease and the patients who were still alive at the end of follow-up. One patient who died in the immediate period after surgery (0.8 months) from complications was excluded from the survival analysis. At the time of data accrual, 15 patients were alive and 62 had died of disease, of whom six had died of causes unrelated to the primary disease. The overall median survival time of these 62 patients was 11 months (minimum 3 months and maximum 35 months). The median time of follow-up was 13 months, ranging from 3 to 58 months.

Fig. 2 Kaplan-Meier survival curves of patients with pancreatic ductal adenocarcinoma. The patients were grouped according to β6 staining scores. The log-rank P value is given

Tumor Biol. Table 2 Univariate Cox proportional hazards analysis for survival of PDAC patients

Variable

Hazard ratio

95 % confidence interval

P value

Age (>60/≤60) Gender (male/female)

1.068 1.168

0.621–1.835 0.685–1.993

0.813 0.568

αvβ6 expression (++,+++/−,+) Tumor location (body-tail/head)

2.222 0.636

1.282–3.853 0.320–1.265

0.004* 0.197

Tumor size (>3 cm/≤3 cm)

1.570

0.913–2.701

0.103

Grade (moderate, poor/well) TNM stage (III–IV/I–II) Lymph node metastasis (+/−) Distant metastasis (+/−)

1.716 3.724 1.867 2.578

0.855–3.442 1.739–7.971 1.045–3.334 1.070–6.211

0.128 0.001* 0.035* 0.035*

*P<0.05 was considered statistically significant

The log-rank test indicated that patients with negative/ weak (0/1+) staining scores for β6 (n = 43) had longer survival times than those with moderate/strong (2+/3+) scores (n=34) (P=0.003). The survival estimates showed a striking difference in median survival times between these two groups, with a median survival of 17.0 months in the negative/weak staining group [95 % confidence interval (CI), 12.075–21.925] and 9.0 months in the moderate/strong staining group (95 % CI, 6.236– 11.764). This was illustrated in the Kaplan–Meier survival curves (Fig. 2). Univariate analysis with the Cox proportional hazards model identified high β6 expression (P= 0.004), advanced TNM stage (P=0.001), lymph node metastasis (P =0.035), and distant metastasis (P=0.035) as statistically significant risk factors affecting the outcome of patients with PDAC (Table 2). To obtain a more precise estimate, multivariate analysis was performed. The results indicated that only β6 expression [hazard ratio (HR) 2.141, P =0.014] and TNM stage (HR 4.320, P=0.027) retained prognostic significance for survival (Table 3). These results confirmed that increased β6 expression represented an independent unfavorable prognostic indicator for PDAC patients.

Table 3 Multivariate Cox proportional hazards analysis for survival of PDAC patients Variables

P

Hazard ratio 95.0 % CI Lower Upper

αvβ6 expression (++,+++/−,+) 0.014* 2.141 TNM stage (III–IV/ I–II) 0.027* 4.320 Lymph node metastasis (+/−) 0.079 1.232 Distant metastasis 0.062 1.833 *P<0.05 was considered statistically significant

1.165 1.182 0.853 1.020

3.933 15.789 2.785 5.533

β6 markedly promoted proliferation and invasion of pancreatic carcinoma cells To explore the role of integrin β6 in pancreatic carcinoma, three pancreatic carcinoma cell lines, PANC-1, Capan-2, and CFPAC-1, were analyzed for the presence of β6 at both mRNA and protein levels. RT-PCR analysis showed high transcript levels of β6 in PANC-1 and Capan-2 cells, whereas low mRNA expression of β6 was found in CFPAC-1 cells. At the protein level, PANC-1 and Capan-2 cells showed high expression of β6, while CFPAC-1 cells showed low β6 expression (Fig. 3a). To evaluate the effect of β6 in pancreatic carcinoma cells, PANC-1 and Capan-2 cells were treated with antisense β6 transfectants, while a β6 plasmid was used in CFPAC-1 cells. As shown in Fig. 3b, antisense β6 transfectants downregulated β6 expression in PANC-1 and Capan-2 cells, while β6 was effectively upregulated in β6 transfected CFPAC-1 cells. The results of the CCK8 assay showed that β6 downregulation in PANC-1 and Capan-2 cells was associated with decreased cell viability (Fig. 3c, d),

Fig. 3 β6 promoted proliferation and invasion of pancreatic carcinoma„ cells. a Integrin αvβ6 expression in PANC-1, Capan-2 and CFPAC-1 pancreatic carcinoma cells. Integrin αvβ6 expression was high in PANC-1 and Capan-2 cells and low in CFPAC-1 cells at the mRNA and protein levels. β6 expression was high in PANC-1 and Capan-2 cells and low in CFPAC-1 cells at the protein level. The expression of αv integrin was high in all three cells. b Western blotting showed that β6 expression was significantly suppressed after antisense β6 transfection of PANC-1 and Capan-2 cells, and markedly upregulated in β6-transfected CFPAC-1 cells. The transfection did not significantly affect the αv integrin expression. c–e CCK8 assay revealed that silencing of β6 significantly decreased cell viability in PANC-1 and Capan-2 cells, and overexpression of β6 markedly increased the proliferation of CFPAC-1 cells. f–h PANC-1 and Capan-2 cells transfected with antisense β6 and CFPAC-1 cells transfected with β6 plasmid were used for Transwell invasion assays. Five fields of cells in the lower side were counted (×200 magnification). Data represent the mean ± SD of three independent experiments. *P<0.05; **P<0.01; ***P<0.001

Tumor Biol.

Tumor Biol.

whereas β6 overexpression significantly promoted the proliferation of CFPAC-1 cells (Fig. 3e). To evaluate the effect of β6 on cell-invasive capability, we performed Transwell invasion assays. Preliminary experiments showed that the cell-invasive ability was independent of the number of cells plated, suggesting that cell proliferation does not significantly affect the outcome of invasion assays (data not shown). The number of PANC-1 and Capan-2 cells invading to the lower side of the membrane decreased dramatically in PANC-1 and Capan-2 cells transfected with siRNA against β6 compared with that in wild-type cells (Fig. 3f, g). Conversely, β6 overexpression markedly increased the invasive ability of CFPAC-1 cells (Fig. 3h). These data suggested that integrin β6 could effectively promote the malignant behaviors of pancreatic carcinoma cells, such as proliferation and the invasive capacity. β6 integrin increased MMP-9 expression at both mRNA and protein levels In our previous study, we showed that β6 integrin induces MMP-2, MMP-3, and MMP-9 in human colorectal carcinoma [18]. To identify MMPs induced by β6 integrin in pancreatic carcinoma, the mRNA expression of MMP-2, MMP-3, and MMP-9 was assessed in pancreatic carcinoma cells by quantitative real-time PCR. The results showed that silencing of β6 in PANC-1 and Capan-2 cells significantly downregulated MMP-9 mRNA expression (Fig. 4a, b), whereas β6 overexpression in CFPAC-1 cells upregulated MMP-9 mRNA expression (Fig. 4c). As MMPs secreted in the supernatant contribute to the high invasiveness of cancer cells, we examined the amount of MMP-9 in the supernatant of pancreatic cells by ELISA. The results showed that in PANC-1 and Capan-2 cells, the MMP-9 protein was significantly downregulated after suppression of β6 expression (Fig. 4d, e). In CFPAC-1 cells, MMP-9 expression was higher in β6-transfected cells than in mock-transfected cells (Fig. 4f). β6-induced MMP-9 upregulation was modulated by the ERK-ETS1 signaling pathway In our previous study, we reported a direct link between β6 integrin and ERK2 and showed that direct binding of these two molecules is essential for β6-induced ETS1 activation. In addition, the promoters of different MMP genes contained the ETS binding sequence (EBS). To explore the mechanism underlying the β6 integrin-mediated induction of MMP-9 expression, activated ERK and ETS1 levels were assessed in pancreatic cells after silencing or overexpression of β6 integrin. Both ERK activation and ETS1 activation were decreased in PANC-1 and Capan-2 cells

transfected with siRNA against β6. Meanwhile, overexpression of β6 integrin increased ERK and ETS1 activation in CFPAC-1 cells (Fig. 5a). Unlike its effect on colon cancer cells [26], β6 had no significant effect on the PI3K/ Akt pathway in pancreatic cancer cells (data not shown). To confirm the involvement of the ERK-ETS1 signaling pathway in β6 integrin-induced MMP-9 upregulation, the signaling pathway was blocked at several levels and MMP-9 expression was detected. The specific ERK inhibitor PD98059 and the receptor recycling and vesicular transport inhibitor primaquine were used to demonstrate that ERK activity is essential for β6-induced ETS1 activation. The results showed that both PD98059 and primaquine effectively suppressed the expression of p-ETS1, with no obvious effect on t-ETS1. In CFPAC-1 cells, PD98059 and primaquine effectively reverted β6 integrin-induced ETS1 activation (Fig. 5b). To evaluate the effects of the ERK-ETS1 signaling pathway on MMP-9 expression, we first treated PANC-1 and Capan-2 cells with PD98059, primaquine or ETS1 siRNA, which resulted in the downregulation of MMP-9 expression at the mRNA (Fig. 5c, d) and protein (Fig. 5g, h) levels. Then, β6 antisense stable PANC-1 and Capan-2 cells were transfected with a pcDNA-ETS1 plasmid containing the ETS1 gene and the pcDNA 3.1 plasmid as a negative control. The results showed that ETS1 overexpression reversed the suppression of MMP-9 expression induced by β6 silencing, and this was validated by realtime PCR (Fig. 5e, f) and ELISA (Fig. 5i, j). In β6 p l a s m i d - t r a n s f e c t e d C F PA C - 1 c e l l s , P D 9 8 0 5 9 , primaquine, or ETS1 siRNA downregulated MMP-9 expression at both mRNA (Fig. 5k) and protein (Fig. 5l) levels. Taken together, these data indicated that the β6-induced MMP-9 upregulation is modulated by the ERK-ETS1 signaling pathway in pancreatic cells.

β6-induced activation of the ERK-ETS1 signaling pathway was required for the proliferation and invasion of pancreatic carcinoma cells To determine whether ERK-ETS1 activation is essential for β6-induced invasion of pancreatic carcinoma cells, PANC-1 cells were treated with the β6 neutralizing antibody 10D5, PD98059, or ETS1 siRNA, and the Transwell assay was performed to detect changes in invasive ability. The results showed that these treatments significantly decreased the invasion of PANC-1 cells (Fig. 6a, b). In CFPAC-1 cells overexpressing β6, inhibition of β6 by 10D5, inhibition of ERK activation by PD98059, or inhibition of ETS1 activity by ETS1 siRNA significantly decreased the invasive capability of cells (Fig. 6c, d).

Tumor Biol. Fig. 4 β6 integrin increased MMP-9 expression at both mRNA and protein levels. a–b MMP-9 mRNA was significantly decreased in PANC-1 cells (a) and Capan-2 cells (b) after transfection with antisense β6, while the expression of MMP-2 and MMP-3 showed no significant changes. c MMP-9 mRNA was significantly increased in β6 overexpressing CFPAC-1 cells. d–e The protein level of MMP-9 was detected by ELISA in PANC-1 cells (d) and Capan-2 cells (e) transfected with antisense β6. f The protein level of MMP-9 was detected by ELISA in CFPAC-1 cells transfected with β6 plasmid. Data represent the mean±SD from three independent experiments. *P<0.05; **P<0.01

The results of the CCK8 assay showed that the ERK-ETS1 pathway is essential for β6-induced proliferation of pancreatic cells, as treatment with 10D5, PD98059, or ETS1 siRNA effectively suppressed the viability of PANC-1 cells

(Fig. 6e) or β6 transfected CFPAC-1 cells (Fig. 6f). Taken together, these data indicated that the β6-induced activation of the ERK-ETS1 signaling pathway is required for the proliferation and invasion of pancreatic carcinoma cells.

Tumor Biol.

Tumor Biol.

ƒFig. 5

Integrin β6 induced the expression of MMP-9 via the ERK-ETS1 signaling pathway. a Western blotting showed that integrin β6 increased the expression of phosphorylated ERK (p-ERK) and phosphorylated ETS1 (p-ETS1), while total ERK and total ETS1 remained unchanged. b Western blotting showed that in PANC-1 and Capan-2 cells, the specific ERK inhibitor PD98059 and primaquine markedly downregulated pETS1, while in CFPAC-1 cells, PD98059 and primaquine effectively reverted the β6 induced ETS1 activation. c–d The mRNA expression of MMP-9 decreased markedly after treatment with PD98059, primaquine or ETS1 siRNA in PANC-1 cells (c) and Capan-2 cells (d). e–f After transfection with antisense β6, PANC-1 cells (e) and Capan-2 cells (f) were transfected with a pcDNA-ETS1 plasmid (with the pcDNA 3.1 plasmid as control), then MMP-9 mRNA expression was evaluated by real-time PCR. g–j The results of real-time PCR were confirmed by ELISA at the protein level. k–l In CFPAC-1 cells transfected with β6 plasmid, changes of MMP-9 expression were detected by real-time PCR and ELISA in cells pretreated with PD98059, primaquine or ETS1 siRNA. The results are shown as the mean±SD of three independent experiments. *P<0.05; **P<0.01; ***P<0.001

Suppression of established tumor growth in nude mice by β6 siRNA treatment To evaluate the effect of integrin β6 in vivo, we established a xenograft tumor model in nude mice bearing PANC-1 cells. Consistent with our in vitro findings, siRNAmediated silencing of β6 markedly suppressed the growth of PANC-1 xenograft tumors compared with the NC group (Fig. 7a). The overall mean tumor volume and tumor weight of the β6 siRNA-treated group were significantly smaller than those of the NC group (Fig. 7b–d). These results indicated that β6 silencing attenuated the growth of PDAC xenograft tumors.

Discussion PDAC is a high-grade malignancy with an extremely poor prognosis associated with the invasiveness and metastatic potential of tumor cells [1, 4, 27]. Therefore, improving our understanding of the mechanisms underlying the malignant behaviors of PDAC may facilitate the development of potential clinical strategies. In the present study, β6 upregulation was an independent unfavorable prognostic indicator in PDAC patients. Integrin β6 markedly promoted the proliferation and invasion of pancreatic carcinoma cells through the ERK-ETS1 signaling pathway. The epithelial-restricted integrin β6 is not constitutively expressed in the normal epithelium and benign epithelial tumor tissues, whereas it is upregulated in those that have undergone malignant transformation [19, 28]. β6 expression rates in carcinomas of the colon, breast, and stomach were previously reported as 37, 18, and 36.7 %, respectively, and increased β6 expression is associated with advanced tumor stage and lymphatic metastasis [15, 19,

29]. In the present study, the total expression rate of β6 in PDAC tissues was 53.8 %, and increased β6 expression was significantly related to larger tumor size, more advanced TNM stage, and lymph node metastasis. This is in agreement with previous observations that β6 expression levels are highest in PDAC among various human carcinoma types [30]. A characteristic staining pattern was also observed, with more intense staining at the invasive margins or tumor cells infiltrating the stroma. This typical staining pattern has been described in invasive colon carcinomas and oral squamous carcinomas and might reflect an interaction between the tumor cells and ECM components [19, 31]. In the present study, β6 upregulation was associated with shorter overall survival times, confirming the role of β6 as an independent unfavorable prognostic indicator for aggressive pancreatic carcinoma in humans. The potential implications of these findings are very meaningful for clinical practice. First, the expression of β6 could be used as a novel marker of poor prognosis in PDAC patients. Second, it may serve to identify patients who are at high risk of malignant evolution. This would help define those individuals that should be targeted for more aggressive intervention and treatment. To the best of our knowledge, this is the first report to show the prognostic value of β6 expression in this field. Our findings are consistent with recent reports in cervical squamous cell carcinomas and non-small cell lung cancer and might have potential implications for cancer-targeting therapy [31–33]. Enhanced β6 expression promotes cell proliferation, migration, invasion, and metastasis in many types of malignancies [10, 14, 16, 18, 21]. However, the role of integrin β6 in pancreatic carcinoma cells has remained underexplored. Based on our immunohistochemical findings, we extended the observations to investigate the effect of β6 on cellular malignant behaviors. First, we examined the expression β6 in three pancreatic adenocarcinoma cell lines at both mRNA and protein levels. High expression of β6 was observed in PANC-1 and Capan-2 cells, while CFPAC-1 cells showed low expression of β6. In the current study, we suppressed β6 function by antisense β6 transfection in PANC-1 and Capan-2 cells and overexpressed β6 in CFPAC-1 cells to investigate the effects of β6 in PDAC. Our results showed that β6 plays an important role in promoting tumor cell growth and invasiveness. Previous studies showed that β6 regulates the expression of proteinases such as MMP-2, MMP-3, and MMP-9, increasing matrix degradation and facilitating tumor cell invasion and metastasis [18, 23, 31]. Therefore, we explored the effect of β6 on MMP expression in PDAC. Our results showed that β6 could induce the expression of MMP-9, but not that of MMP2 or MMP-3 in PDAC cells. Our previous research found a direct link between ERK and the cytoplasmic domain of β6,

Tumor Biol.

Tumor Biol.

ƒFig. 6

β6-induced activation of ERK-ETS1 signaling was required for the proliferation and invasion of pancreatic carcinoma cells. a 10D5, PD98059 or ETS1 siRNA significantly suppressed the invasive capability of PANC-1 cells (×200 magnification). b After transfection with β6 plasmid, the invasiveness of CFPAC-1 cells was markedly suppressed by 10D5, PD98059 or ETS1 siRNA (×200 magnification). c–d Statistical analysis of Transwell invasion assay results in (a) and (b). e 10D5, PD98059 or ETS1 siRNA significantly suppressed the proliferation of PANC-1 cells. f After transfection with β6 plasmid, the proliferation of CFPAC-1 cells was markedly suppressed by 10D5, PD98059 or ETS1 siRNA. The results are shown as the mean±SD from three independent experiments. *P<0.05; **P<0.01; ***P<0.001

and increased ERK activity was associated with direct binding. We also demonstrated that inhibition of ERK phosphorylation by PD98059 blocked the internalization of β6 and downstream signaling. Primaquine, a receptor recycling and vesicular transport inhibitor, also suppressed the effects of β6. The specific β6 neutralizing antibody 10D5 is an effective inhibitor of β6 function, as shown previously [34–36]. Furthermore, ETS1 activation, which is the downstream effect of β6-induced ERK activation, leads to the transcriptional activation of downstream targets including MMP-9 and ETS1 itself [18, 37]. Based on these data and the effects of PD98059, primaquine, 10D5, and ETS1 siRNA on inhibiting the ERK-ETS1 pathway, we designed a hypothetical model in which integrin αvβ6 internalizes into the cytoplasm and activates ERK, then activated ERK translocates into the nucleus and activates Ets-1, resulting in the proliferation, migration

Fig. 7 Treatment of β6 siRNA effectively suppressed the growth of subcutaneous xenograft tumors. a The growth curves of tumors in nude mice treated with negative control and β6 siRNA. b Representative image of isolated tumors. c The mean volume of tumors was significantly smaller in the β6 siRNA treatment group (n=6) than in the negative control group. d The mean tumor weight was significantly lower in the β6 siRNA treatment group (n=6) than in the negative control group

and invasion of PDAC cells, while the internalized integrin αvβ6 Bshuttles^ to the membrane (Fig. 8). The results of a xenograft tumor model support our conclusions in vitro, showing that the growth of PANC-1 xenograft tumors was markedly suppressed in response to β6 siRNA treatment. Using genetically engineered mouse models, Hezel et al. reported that in the presence of Smad4, αvβ6 interacts with TGF-β and functions as a tumor suppressor during PDAC progression, indicating that αvβ6 targeting therapy in PDAC patients may carry a risk [38]. However, it is also possible that anti-αvβ6 treatment could block the tumorpromoting downstream effects, suggesting αvβ6 as an effective target. The roles of αvβ6 depend on the genetic background and microenvironment of tumor cells; in addition, the animal model used for in vivo studies is important. Further research is needed to clarify these issues. Pancreatic carcinoma continues to be a severe disease with poor prognosis. At diagnosis, the resectability rate is low and these tumors are not responsive to chemotherapy or radiotherapy [5]. Molecular targeting strategies for cancer therapy are distinct from conventional chemotherapy and radiotherapy because of their potential for increased tumor specificity and because they are better tolerated [39]. The present study suggests that integrin β6 plays an important role in the progression of pancreatic carcinoma and contributes to reduced survival times. β6 upregulation was identified as an independent unfavorable prognostic indicator in PDAC. We also demonstrated that integrin αvβ6 Bshuttling^-induced activation of

Tumor Biol. Fig. 8 Hypothetical model showing that integrin αvβ6 promotes malignant behaviors via the ERK-ETS1 pathway in PDAC cells. The internalization of integrin αvβ6 activated ERK by direct binding, then activated ERK translocated into the nucleus and promoted the phosphorylation of ETS1, which subsequently promoted the malignant behaviors of PDAC cells

the ERK-ETS1 signaling pathway was required for the proliferation and invasion of pancreatic carcinoma cells. Our data revealed that the Bshuttling^ of integrin αvβ6 contributes largely to the malignant behaviors of PDAC cells, and β6 perturbation might present a rational strategy for the treatment of patients with PDAC. Acknowledgments This study is supported by National Natural Science Foundation of China (no. 81272653) and by Natural Science Foundation of Shandong Province (no. ZR2015HM071). The authors thank Biogen Idec for the special antibodies and Professor Hui Wang for the assistance in statistical analysis. Compliance with ethical standards

6.

7. 8. 9.

10.

11.

Conflicts of interest None 12.

References 1. 2.

3.

4. 5.

Siegel R, Ma J, Zou Z, Jemal A. Cancer statistics, 2014. CA Cancer J Clin. 2014;64(1):9–29. doi:10.3322/caac.21208. Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, Rebelo M, et al. Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012. Int J Cancer. 2015;136(5):E359–86. doi:10.1002/ijc.29210. Rahib L, Smith BD, Aizenberg R, Rosenzweig AB, Fleshman JM, Matrisian LM. Projecting cancer incidence and deaths to 2030: the unexpected burden of thyroid, liver, and pancreas cancers in the United States. Cancer Res. 2014;74(11):2913–21. doi:10.1158/ 0008-5472.CAN-14-0155. Li D, Xie K, Wolff R, Abbruzzese JL. Pancreatic cancer. Lancet. 2004;363(9414):1049–57. doi:10.1016/S0140-6736(04)15841-8. Stathis A, Moore MJ. Advanced pancreatic carcinoma: current treatment and future challenges. Nat Rev Clin Oncol. 2010;7(3): 163–72. doi:10.1038/nrclinonc.2009.236.

13.

14.

15.

16.

17.

Li HY, Cui ZM, Chen J, Guo XZ, Li YY. Pancreatic cancer: diagnosis and treatments. Tumour Biol. 2015;36(3):1375–84. doi:10. 1007/s13277-015-3223-7. Bandyopadhyay A, Raghavan S. Defining the role of integrin alphavbeta6 in cancer. Curr Drug Targets. 2009;10(7):645–52. Felding-Habermann B. Integrin adhesion receptors in tumor metastasis. Clin Exp Metastasis. 2003;20(3):203–13. Thomas GJ, Nystrom ML, Marshall JF. Alphavbeta6 integrin in wound healing and cancer of the oral cavity. J Oral Pathol Med. 2006;35(1):1–10. doi:10.1111/j.1600-0714.2005.00374.x. Li HX, Zheng JH, Fan HX, Li HP, Gao ZX, Chen D. Expression of alphavbeta6 integrin and collagen fibre in oral squamous cell carcinoma: association with clinical outcomes and prognostic implications. J Oral Pathol Med. 2013;42(7):547–56. doi:10.1111/jop. 12044. Li W, Liu Z, Zhao C, Zhai L. Binding of MMP-9-degraded fibronectin to beta6 integrin promotes invasion via the FAK-Src-related Erk1/2 and PI3K/Akt/Smad-1/5/8 pathways in breast cancer. Oncol Rep. 2015. doi:10.3892/or.2015.4103. Liu S, Wang J, Niu W, Liu E, Wang J, Peng C, et al. The beta6integrin-ERK/MAP kinase pathway contributes to chemo resistance in colon cancer. Cancer Lett. 2013;328(2):325–34. doi:10. 1016/j.canlet.2012.10.004. Xue B, Wu W, Huang K, Xie T, Xu X, Zhang H, et al. Stromal cellderived factor-1 (SDF-1) enhances cells invasion by alphavbeta6 integrin-mediated signaling in ovarian cancer. Mol Cell Biochem. 2013;380(1–2):177–84. doi:10.1007/s11010-013-1671-1. Yang GY, Guo S, Dong CY, Wang XQ, Hu BY, Liu YF, et al. Integrin alphavbeta6 sustains and promotes tumor invasive growth in colon cancer progression. World J Gastroenterol. 2015;21(24): 7457–67. doi:10.3748/wjg.v21.i24.7457. Zhang ZY, Xu KS, Wang JS, Yang GY, Wang W, Wang JY, et al. Integrin alphanvbeta6 acts as a prognostic indicator in gastric carcinoma. Clin Oncol. 2008;20(1):61–6. doi:10.1016/j.clon.2007.09. 008. Wang B, Wang W, Niu W, Liu E, Liu X, Wang J, et al. SDF-1/ CXCR4 axis promotes directional migration of colorectal cancer c e l l s t h r o u g h u p r e g u l a t i o n o f i n t e g r i n al p h a v b e t a 6 . Carcinogenesis. 2014;35(2):282–91. doi:10.1093/carcin/bgt331. Gu X, Niu J, Dorahy DJ, Scott R, Agrez MV. Integrin alpha(v)beta6-associated ERK2 mediates MMP-9 secretion in

Tumor Biol. colon cancer cells. Br J Cancer. 2002;87(3):348–51. doi:10.1038/ sj.bjc.6600480. 18. Gao H, Peng C, Liang B, Shahbaz M, Liu S, Wang B, et al. β6 integrin induces the expression of metalloproteinase-3 and metalloproteinase-9 in colon cancer cells via ERK-ETS1 pathway. Cancer Lett. 2014;354(2):427–37. doi:10.1016/j.canlet.2014.08. 017. 19. Bates RC, Bellovin DI, Brown C, Maynard E, Wu B, Kawakatsu H, et al. Transcriptional activation of integrin beta6 during the epithelial-mesenchymal transition defines a novel prognostic indicator of aggressive colon carcinoma. J Clin Invest. 2005;115(2): 339–47. doi:10.1172/JCI23183. 20. Ahmed N, Niu J, Dorahy DJ, Gu X, Andrews S, Meldrum CJ, et al. Direct integrin alphavbeta6-ERK binding: implications for tumour growth. Oncogene. 2002;21(9):1370–80. doi:10.1038/sj.onc. 1205286. 21. Niu W, Liu X, Zhang Z, Xu K, Chen R, Liu E, et al. Effects of alphavbeta6 gene silencing by RNA interference in PANC-1 pancreatic carcinoma cells. Anticancer Res. 2010;30(1):135–42. 22. Agrez M, Chen A, Cone RI, Pytela R, Sheppard D. The alpha v beta 6 integrin promotes proliferation of colon carcinoma cells through a unique region of the beta 6 cytoplasmic domain. J Cell Biol. 1994;127(2):547–56. 23. Ahmed N, Pansino F, Clyde R, Murthi P, Quinn MA, Rice GE, et al. Overexpression of alpha(v)beta6 integrin in serous epithelial ovarian cancer regulates extracellular matrix degradation via the plasminogen activation cascade. Carcinogenesis. 2002;23(2):237–44. 24. Marechal R, Demetter P, Nagy N, Berton A, Decaestecker C, Polus M, et al. High expression of CXCR4 may predict poor survival in resected pancreatic adenocarcinoma. Br J Cancer. 2009;100(9): 1444–51. doi:10.1038/sj.bjc.6605020. 25. Zhang YJ, Dai Q, Sun DF, Xiong H, Tian XQ, Gao FH, et al. mTOR signaling pathway is a target for the treatment of colorectal cancer. Ann Surg Oncol. 2009;16(9):2617–28. doi:10.1245/ s10434-009-0555-9. 26. Defilles C, Lissitzky JC, Montero MP, Andre F, Prevot C, Delamarre E, et al. alphavbeta5/beta6 integrin suppression leads to a stimulation of alpha2beta1 dependent cell migration resistant to PI3K/Akt inhibition. Exp Cell Res. 2009;315(11):1840–9. doi: 10.1016/j.yexcr.2009.03.014. 27. Lee J, Hun Yun J, Lee J, Choi C, Hoon KJ. Blockade of dualspecificity phosphatase 28 decreases chemo-resistance and migration in human pancreatic cancer cells. Sci Rep. 2015;5:12296. doi: 10.1038/srep12296. 28. Breuss JM, Gillett N, Lu L, Sheppard D, Pytela R. Restricted distribution of integrin beta 6 mRNA in primate epithelial tissues. J Histochem Cytochem. 1993;41(10):1521–7.

29.

Arihiro K, Kaneko M, Fujii S, Inai K, Yokosaki Y. Significance of alpha 9 beta 1 and alpha v beta 6 integrin expression in breast carcinoma. Breast Cancer. 2000;7(1):19–26. 30. Sipos B, Hahn D, Carceller A, Piulats J, Hedderich J, Kalthoff H, et al. Immunohistochemical screening for beta6-integrin subunit expression in adenocarcinomas using a novel monoclonal antibody reveals strong up-regulation in pancreatic ductal adenocarcinomas in vivo and in vitro. Histopathology. 2004;45(3):226–36. doi:10. 1111/j.1365-2559.2004.01919.x. 31. Thomas GJ, Lewis MP, Hart IR, Marshall JF, Speight PM. AlphaVbeta6 integrin promotes invasion of squamous carcinoma cells through up-regulation of matrix metalloproteinase-9. Int J Cancer. 2001;92(5):641–50. 32. Elayadi AN, Samli KN, Prudkin L, Liu YH, Bian A, Xie XJ, et al. A peptide selected by biopanning identifies the integrin alphavbeta6 as a prognostic biomarker for nonsmall cell lung cancer. Cancer Res. 2007;67(12):5889–95. doi:10.1158/0008-5472.CAN-070245. 33. Liang B, Shahbaz M, Wang Y, Gao H, Fang R, Niu Z, et al. Integrinbeta6-targeted immunoliposomes mediate tumor-specific drug delivery and enhance therapeutic efficacy in colon carcinoma. Clin Cancer Res. 2015;21(5):1183–95. doi:10.1158/1078-0432. CCR-14-1194. 34. Ramos DM, But M, Regezi J, Schmidt BL, Atakilit A, Dang D, et al. Expression of integrin beta 6 enhances invasive behavior in oral squamous cell carcinoma. Matrix Biol. 2002;21(3):297–307. 35. Huang X, Wu J, Spong S, Sheppard D. The integrin alphavbeta6 is critical for keratinocyte migration on both its known ligand, fibronectin, and on vitronectin. J Cell Sci. 1998;111(Pt 15):2189–95. 36. Xue H, Atakilit A, Zhu W, Li X, Ramos DM, Pytela R. Role of the alpha(v)beta6 integrin in human oral squamous cell carcinoma growth in vivo and in vitro. Biochem Biophys Res Commun. 2001;288(3):610–8. doi:10.1006/bbrc.2001.5813. 37. Wang J, Wu J, Hong J, Chen R, Xu K, Niu W, et al. PKC promotes the migration of colon cancer cells by regulating the internalization and recycling of integrin alphavbeta6. Cancer Lett. 2011;311(1): 38–47. doi:10.1016/j.canlet.2011.06.025. 38. Hezel AF, Deshpande V, Zimmerman SM, Contino G, Alagesan B, O’Dell MR, et al. TGF-beta and alphavbeta6 integrin act in a common pathway to suppress pancreatic cancer progression. Cancer Res. 2012;72(18):4840–5. doi:10.1158/0008-5472.CAN-12-0634. 39. Leconet W, Larbouret C, Chardes T, Thomas G, Neiveyans M, Busson M, et al. Preclinical validation of AXL receptor as a target for antibody-based pancreatic cancer immunotherapy. Oncogene. 2014;33(47):5405–14. doi:10.1038/onc.2013. 487.