(uPA) in prostate cancer - Nature

Oncogene (2013) 32, 2282–2291 & 2013 Macmillan Publishers Limited All rights reserved 0950-9232/13 www.nature.com/onc

ORIGINAL ARTICLE

CFTR suppresses tumor progression through miR-193b targeting urokinase plasminogen activator (uPA) in prostate cancer C Xie1,9, XH Jiang1,2,9, JT Zhang1, TT Sun1, JD Dong1, AJ Sanders3, RY Diao1,4, Y Wang1, KL Fok1, LL Tsang1, MK Yu1, XH Zhang1, YW Chung1, L Ye3, MY Zhao1, JH Guo1, ZJ Xiao1, HY Lan5, CF Ng6, KM Lau7, ZM Cai4, WG Jiang3 and HC Chan1,2,8 Cystic fibrosis (CF) transmembrane conductance regulator (CFTR) is expressed in the epithelial cells of a wide range of organs/ tissues from which most cancers are derived. Although accumulating reports have indicated the association of cancer incidence with genetic variations in CFTR gene, the exact role of CFTR in cancer development and the possible underlying mechanism have not been elucidated. Here, we report that CFTR expression is significantly decreased in both prostate cancer cell lines and human prostate cancer tissue samples. Overexpression of CFTR in prostate cancer cell lines suppresses tumor progression (cell growth, adhesion and migration), whereas knockdown of CFTR leads to enhanced malignancies both in vitro and in vivo. In addition, we demonstrate that CFTR knockdown-enhanced cell proliferation, cell invasion and migration are significantly reversed by antibodies against either urokinase plasminogen activator (uPA) or uPA receptor (uPAR), which are known to be involved in various malignant traits of cancer development. More interestingly, overexpression of CFTR suppresses uPA by upregulating the recently described tumor suppressor microRNA-193b (miR-193b), and overexpression of pre-miR-193b significantly reverses CFTR knockdownenhanced malignant phenotype and abrogates elevated uPA activity in prostate cancer cell line. Finally, we show that CFTR gene transfer results in significant tumor repression in prostate cancer xenografts in vivo. Taken together, the present study has demonstrated a previously undefined tumor-suppressing role of CFTR and its involvement in regulation of miR-193b in prostate cancer development. Oncogene (2013) 32, 2282–2291; doi:10.1038/onc.2012.251; published online 16 July 2012 Keywords: CFTR; miR-193b; uPA

INTRODUCTION Cystic fibrosis (CF) transmembrane conductance regulator (CFTR) is a Cl and HCO3 conducting anion channel expressed in the epithelial cells of a wide variety of tissues, mutations of which cause CF, a common life-threatening autosomal recessive disease among Caucasians.1,2 Although progressive lung disease characterized by chronic inflammation and bacterial infection is the prime cause of morbidity in CF, the clinical consequences of absent or reduced CFTR ion channel function result in complex, multi-organ characteristics of the CF phenotype, which are sitespecific but vary considerably in severity and age of onset.3 Over the past three decades, the most striking result of advances in the care of patients with CF has been the dramatic improvement in survival. For instance, whereas the median survival of CF patients in the United States was only 16 years in 1970, it has risen to over 32 years in 2005.4 As a consequence, the natural history of the disease is changing and new clinical complications are emerging. There has been accumulating evidence indicating the association of cancer incidence with the genetic variations in the CFTR

gene. Large cohort studies in North American and European patients with CF found that there was a marked increase in the risk of malignancies affecting the gastrointestinal tract, pancreas and hepatobiliary system.5–7 However, CF gene mutations have also been reported to be associated with a lower risk in several cancers, such as melanoma,8 breast cancer,9 colon cancer,10 prostate cancer11 and lung cancer.12 The contradictory findings might be due to the differences in sample size, study design and various CFTR mutations included in the individual studies. On the other hand, the CFTR gene has been reported to be frequently hypermethylated in both cancer cell lines and primary tumor samples.13–16 These results indicate that DNA methylationmediated transcription silencing of CFTR may promote carcinogenesis. Regardless of all these indications, the exact role of CFTR in cancer development and the possible underlying mechanisms have not been elucidated. We undertook the present study to investigate the role of CFTR and the underlying mechanisms as a putative tumor suppressor by gene manipulation in prostate cancer cell lines and tumor

1 Epithelial Cell Biology Research Center, School of Biomedical Sciences, The Chinese University of Hong Kong, Shatin, Hong Kong; 2Key Laboratory for Regenerative Medicine (Jinan University-the Chinese University of Hong Kong), Ministry of Education of The People’s Republic of China, Guangzhou, The People’s Republic of China; 3Department of Surgery, Metastasis and Angiogenesis Research Group, Cardiff University School of Medicine, Cardiff, UK; 4Shenzhen Key Laboratory of Male Reproduction Medicine and Genetics, Peking University Shenzhen Hospital, Shenzhen, China; 5Li Ka Shing Institute of Health Sciences, Department of Medicine and Therapeutics, Faculty of Medicine, The Chinese University of Hong Kong, Shatin, Hong Kong; 6Department of Surgery, Division of Urology, The Chinese University of Hong Kong, Shatin, Hong Kong; 7Department of Anatomical and Cellular Pathology, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, Hong Kong and 8Sichuan University-The Chinese University of Hong Kong Joint Laboratory for Reproductive Medicine, West China Second University Hospital, Chengdu, China. Correspondence: Professor WG Jiang, Department of Surgery, Metastasis and Angiogenesis Research Group, Cardiff University School of Medicine, Cardiff, UK or Professor HC Chan, Epithelial Cell Biology Research Center, School of Biomedical Sciences, The Chinese University of Hong Kong, Shatin, Hong Kong. E-mail: [email protected] or [email protected] 9 These authors contributed equally to this work. Received 6 November 2011; revised 29 March 2012; accepted 13 April 2012; published online 16 July 2012

CFTR in the pathogenesis of prostate cancer C Xie et al

2283 xenograft mice. Our data show that CFTR suppresses prostate cancer progression both in vitro and in vivo. In addition, we reveal that the tumor-suppressing effect of CFTR is, at least in part, associated with its ability to alter the recently described tumor suppressor microRNA-193b (miR-193b) targeting urokinase-type plasminogen activator (uPA), which is known to be involved in various malignant traits of cancer development.

RESULTS CFTR expression is downregulated in prostate cancer samples and cell lines Recently, we have observed that the expression of CFTR is markedly reduced in aged rat prostate (Supplementary Figure 1). Moreover, inhibition of CFTR function by a specific CFTR inhibitor, CFTR-inh172 (inh172), results in increased cell proliferation as determined by PCNA staining (Supplementary Figure 2). These observations prompted us to speculate that repression of CFTR function could confer a high risk of prostate cancer development. We therefore set out to examine CFTR expression in cancer tissue microarray, analyzing 62 cases of human prostate cancer sample and 10 cases of normal tissue. The immunohistochemical staining of CFTR was scored in a semiquantitative manner by examination of the cytoplasmic and membrane staining intensity. Our results showed that in normal tissue, 8/10 (80%) of the samples exhibited a CFTR positive immunohistochemical staining scored at þ þ (2) or þ þ þ (3). However, in prostate cancer samples, most of the cases (51/62, 82.2%) have weak þ or negative±CFTR expression (Figures 1a and b, Supplementary Table 1). We next examined the expression of CFTR in three prostate cancer cell lines (PC-3, DU145, LNcap) and one immortalized prostate epithelial cell line (PNT1A) by RT–PCR. As shown in Figure 1c, the CFTR mRNA level in cancer cell lines was significantly lower compared with normal prostate epithelial cells. These data are supportive of previous report showing CFTR is hypermethylated at its promoter region in the prostate cancer cell lines.15 Taken together, these results indicate that CFTR expression is downregulated in both prostate cancer tissues and cell lines. CFTR is negatively associated with tumorigenic activities in cancer cell lines To explore the role of CFTR in cancer development, we subsequently assessed the effects of altered CFTR expression on tumorigenic activities in prostate cancer cell lines (PC-3, DU-145 and LNcap). First, to examine whether CFTR per se has tumorsuppressive activity in cancer cell lines, we characterized prostate cancer cell lines (LNcap, PC-3) that were transfected with an expression vector harboring the full-length CFTR cDNA or empty vector and observed a substantial increase in CFTR protein expression (Figure 1d). Consequently, transfection with CFTR significantly decreased the growth rate of both cell lines compared with control transfectants (Figures 1e and f). Moreover, the ectopic overexpression of CFTR in both cell lines rendered the cells more susceptible to serum-starvation-induced apoptosis (Figures 1g and k), whereas suppressed cell adhesion, invasion and motility (Figures 1h–j and l–n). To knockdown CFTR expression in the cancer cell lines, we transfected cells with pEF6/VS-His vector carrying a CFTR-specific hammerhead ribozyme 2 to generate stably transfected cells (referred to CFTRrib2 in this study) and observed a substantial reduction in CFTR protein levels in PC-3 and DU-145 cells (Figure 2a). Our results showed that knockdown of CFTR promoted cell growth in PC-3 and DU-145 cells (Figures 2b and c). In addition, inhibition of CFTR significantly enhanced cell adhesion and motility in both PC-3 and DU-145 cells (Figures 2d–g). Taken together, CFTR inhibits multiple tumorigenic properties of cancer cells in vitro. & 2013 Macmillan Publishers Limited

CFTR suppresses tumor development in vivo To evaluate the tumor-suppressing role of CFTR in vivo, PC-3 cells with manipulated CFTR expression were grown as subcutaneous tumors in nude mice. Tumors in mice implanted with PC-3 cells stably overexpressing CFTR (n ¼ 7) exhibited markedly reduced tumor weight at the experimental endpoint (4 weeks) compared with tumors in mice implanted with PC-3 cells transfected with control vector (n ¼ 6) (Figure 3a). On the contrary, tumors in mice implanted with cells carrying hammerhead ribozyme targeting CFTR (n ¼ 4) showed a significantly increased tumor burden compared with tumors in mice (n ¼ 4) implanted with control cells that were transfected with the vector control (Figure 3c). Of note, we observed that CFTR manipulation in these cancer cell lines significantly affected the tumor size without changing the incidence of tumor formation, suggesting that CFTR suppresses tumor progression rather than tumor initiation in vivo. To understand the effect of CFTR on tumor growth, the subcutaneous tumor sections of either control vector or CFTR-transfected PC-3 cells were stained for expression of the PCNA, which is a nuclear protein associated with proliferation and Bcl-2, which is an antiapoptotic gene at the experimental endpoint. As shown in Figure 3b, immunohistochemical analysis of xenograft tumors revealed that CFTR-overexpressing tumors, displayed low frequency of mitotic figures and antiapoptosis capacity. Thus, in line with the in vitro observations, the in vivo results support the notion that CFTR suppresses cancer development. Involvement of uPA in mediating the tumor-suppressing effects of CFTR What might be the mechanism(s) mediating the tumor-suppressing effects of CFTR? Overwhelming evidence have demonstrated that the cell surface-associated uPA-mediated signaling is causatively involved in tumor progression and metastasis of many types of cancers, including prostate cancer, by exerting multifaceted functions via either direct or indirect interactions with integrins, endocytosis receptors and growth factors.17,18 Thus, we hypothesized that the multiple tumor-suppressing effects of CFTR might be mediated by uPA. Indeed, we observed that CFTR knockdown markedly increased uPA protein expression and uPA secretion in PC-3 cells (Figures 4a and b), indicating that uPA activity is CFTR dependent in these cancer cells. If the tumorsuppressing effects of CFTR are largely due to alteration in uPA expression/activities, the effects observed with CFTR knockdown should be reversed by altering uPA activity. To test this, we treated both control and CFTR knockdown prostate cancer cells (PC-3 and DU-145) with uPA or uPA receptor (uPAR)-neutralizing antibodies and evaluated their changes in tumorigenic phenotypes. Our results showed that neutralization with either uPA or uPAR antibodies in CFTR knockdown PC-3 (Figures 4c–e) or DU-145 (Supplementary Figures 3a–c) dramatically reversed the CFTR knockdown-enhanced cell proliferation, migration and invasion in these cells, indicating that upregulation of uPA expression/activity is the major mechanism leading to the observed increased malignancies induced by CFTR knockdown in these cancer cells. More importantly, the regulatory role of CFTR on uPA expression was confirmed in xenograft tumors. Immunohistochemical staining from CFTR-overexpressing xenograft tumors revealed remarkably decreased expression of uPA and uPAR compared with the tumors from vector control cells (Figure 3b). Taken together, these results suggest that CFTR inhibits cancer development through downregulation of uPA acvitity. The tumor-suppressing effect of CFTR involves miR-193b targeting uPA MicroRNAs (miRNAs) have been shown to regulate a wide variety of genes at a post-transcriptional level in response to changes in cellular microenvironment, such as ion fluxes mediated by ion Oncogene (2013) 2282 – 2291


2284

CFTR

CFTR

25 GAPDH

0

Normal

Cancer

e

d

f

eGFP3 CFTR

PC-3 eGFP3

CFTR

CFTR -tubulin

Viable cells (% increase)

500 LNcap

5

50

LNcap LNcapCFTR

400 300

** **

200 100

400 300

40

100

20 10

PC-3eGFPC3 PC-3CFTR

m 120 80 ** 40 0

PC-3eGFPC3 PC-3CFTR

0

j 20

10

** 0

LNcapeGFPC3 LNcapCFTR

l ***

3 days 5 days

5 days

40 20

** LNcapeGFPC3 LNcapCFTR

n 250

50

0

3 days

60

0


75

25

1 day

80 Migrated cells

**

Invaded cells

10

0


30

0

20

Invaded cells

Apoptosis %

Adherent cells

2

Adherent cells

Apoptosis (%)

4

0

k

***

6

1 day

i

30

**

200

**

PC-3eGFPC3 PC-3CFTR

Migratedcells

h

*

0 0

8

PC-3eGFPC3 PC-3CFTR

500

eGFPC3

0

g

p PM T1 a

75

c ca

3 2 1

*

100

Ln

b

PC -3 D U -1 4

Cancer

Percentage (%)

Normal

Viable Cells (% increase)

a

200 150 100 50 0

** PC-3eGFPC3 PC-3CFTR

Figure 1. Overexpression of CFTR suppresses tumorigenic phenotype in prostate cancer cell lines. (a) Immunohistochemical staining of cell nuclei (blue) and CFTR protein (brown) in representative images from prostate tissue microarray. It can be seen that CFTR expression is decreased in prostate cancer sample compared with normal tissue. (b) Distribution of CFTR immunoreactivity scores in normal and tumors from the tissue microarray. *Po0.05 (w2-test). Staining intensity of þ þ þ , þ þ , þ B± was scored at 3, 2, 1, respectively. (c) CFTR expression is downregulated in prostate cancer cell lines compared with normal prostate epithelial cell line as determined by RT–PCR. (d) Western blot analysis of CFTR in control vector (eGFP-C3) or CFTR-transfected LNcap and PC-3 cells. Image shown is the representative of three independent experiments. (e, f ) Cell growth analysis of LNcap and PC-3 cells transfected with CFTR or peGFP-C3 as detected by MTS assay. Cell growth is presented as percentage increase: Absorbancedn absorbanced0/absorbanced0 and data are from three independent experiments, **Po0.01; *Po0.05. (g, k) Flow cytometric analysis of AnnexinV/PI-positive cells in peGFP-C3 or CFTR-transfected LNcap (g) and PC-3 (k) cells. The control or CFTR-overexpressing cells were serum starved for 24 h and apoptotic cells were determined by AnnexinV/PI staining. Data are mean±s.e.m. of three independent experiments, ***Po0.001. (h, l) Adhesion assay for PC-3 and LNcap cells transfected with peGFP-C3 or CFTR. Mean total number of PC-3 or LNcap cells with control or overexpressed CFTR levels adhered to the matrigel-coated plates are shown. Data are from three independent experiments, **Po0.01. (i, m) Invasion assay for PC-3 and LNcap cells transfected with peGFP-C3 or CFTR. Mean total number of cells with control or overexpressed CFTR levels invaded 24 h later are shown. Data are from three independent experiments, **Po0.01. (j, n) Motility for PC-3 and LNcap cells transfected with peGFP-C3 or CFTR. Mean total number of PC-3 and LNcap cells with control or overexpressed CFTR levels migrated from the cytodex-2 beads and adhered to the base of the well overnight are shown. Data are from three independent experiments, **Po0.01.

channels like CFTR, and are known to be involved in cancer development.19,20 Specifically, miR-193b has been recently identified as an epigenetically regulated putative tumor suppressor in various types of cancer21–23 and shown to target uPA.24 Thus, we hypothesized that the multiple tumorsuppressing effects of CFTR might be mediated by uPA through Oncogene (2013) 2282 – 2291

miR-193b. Indeed, we observed a direct association of CFTR with miR-193b expression in prostate cancer cell lines. As shown, knockdown of CFTR led to suppression of miR-193b expression, whereas the exogenous expression of CFTR produced the opposite effect in both PC-3 and LNcap cell lines (Figures 5a and b, Supplementary Figure 4). More interestingly, when we & 2013 Macmillan Publishers Limited


2285 800

DU-145 Viable cells (% increase)

CFTR β-tubulin

600 400

**

200 0 0

600 ***

150 100 50 0

*

PC-3CFTRrib2

300

**

**

400 300 200 100

200

100

PC-3pEF/His PC-3CFTRrib2

200

100

0

0

0 PC-3pEF/His

5 days

300 Adherent cells

200

3 days

500 Migrated cells

Adherent cells

250

**

PC-3pEF/His PC-3CFTRrib2

Migrated cells

PC-3

DU145pEF/His DU145CFTRrib2

DU145pEF/His DU145CFTRrib2

Figure 2. CFTR knockdown promotes tumorigenic activities in prostate cancer cell lines. (a) Western blot analysis of CFTR in control vector (pEF-His) and CFTRrib2-transfected PC-3 and DU-145 cells. Image shown is the representative of three independent experiments. (b, c) Cell growth analysis of PC-3 and DU-145 cells transfected with pEF-His or CFTRrib2 as detected by MTS assay, cell growth is presented as percentage increase: Absorbancedn absorbanced0/absorbanced0 and data are from three independent experiments, **Po0.01. (d, f ) Adhesion assay for PC-3 and DU-145 cells transfected with pEF-His or CFTRrib2. Mean total number of PC-3 and DU-145 cells with control or reduced CFTR levels adhered to the matrigel-coated plates. Data are mean±s.e.m. from three independent experiments, **Po0.01; ***Po0.001. (e, g) Motility for PC-3 and DU-145 cells transfected with pEF-His or CFTRrib2. Mean total number of PC-3 and DU-145 cells with control or reduced CFTR levels migrated from the cytodex-2 beads and adhered to the base of the well. Data are mean±s.e.m. from three independent experiments, *Po0.05, **Po0.01.

ectopically transfected pre-miR-193b in PC-3 CFTRrib2 cells, we observed significant or complete reversal of the CFTR knockdownenhanced proliferation, migration or invasion (Figures 5c–e), respectively, in PC-3 cells. Importantly, the forced overexpression of miR-193b also completely abrogated the CFTR knockdownelevated uPA activity in PC-3 cells (Figure 5f), as expected of uPA being a target of miR-193b. In consistency with the in vitro data, the regulatory role of CFTR on miR-193b was confirmed in xenograft tumors. The results showed that xenograft tumors from CFTR knockdown PC-3 cells expressed decreased level of miR-193b, whereas CFTR-overexpressing tumors exhibited increased expression of miR-193b (Figures 5g–i). Taken together, these results suggest that CFTR inhibits cancer development primarily through an epigenetic mechanism involving miR-193b that targets uPA. Therapeutic potential of CFTR in cancer Next, to evaluate the therapeutic potential of CFTR gene restoration in vivo, we delivered CFTR DNA directly to the xenograft tumors of athymic nude mice using the ultrasoundmicrobubble technique.25 As shown in Figures 6a and b, treated tumors underwent considerable volume regression, as compared with the control injected with empty vector. The antitumor effect of CFTR gene treatment was particularly potent, as the histological analysis of the residual masses indicated the presence of diffuse necrosis areas almost free of surviving cells (Figure 6c, H&E). In addition, the expression of both PCNA and Bcl-2 in treated xenograft tumors was dramatically reduced compared with those injected with control vector (Figure 6c). Besides, in consistence with the previous data, our western blot analysis demonstrated that the expression of both uPA and uPAR in CFTR-treated animals was significantly downregulated compared with that in control mice (Figure 6d). Thus, in vivo delivery and restoration of CFTR gene in tumor appears to have therapeutic potential. & 2013 Macmillan Publishers Limited

DISCUSSION Although there has been recent interest in the risk of various cancers in CF patients and carriers of CFTR mutations, studies on the role of CFTR in the pathogenesis of cancer are very limited. It has been reported that CFTR inhibits MUC4 expression, which then suppresses the growth and metastasis of pancreatic tumor cells.26,27 Another study has proposed that DF508 CFTR mutation may protect against breast cancer, based on evidence that elevated extracellular ATP (adenosine triphosphate) is known to inhibit breast cancer cell line growth and that CFTR pumps ATP out of epithelial cells.9 However, the finding of this study was not supported by the subsequent clinical report showing that patients with CF do not seem to have a lower risk of malignancy in the breast.28 The contradictory finding could be explained by the possibility that the aberrant function of DF508 protein could have effects on tumorigenic behavior in some unknown way. In the present study, we have demonstrated a previously undefined tumor-suppressing role of CFTR in prostate cancer. Our gain and loss of functional studies clearly demonstrate that normal function of CFTR suppresses cancer progression, whereas inhibition of CFTR expression promotes cancer development both in vitro and in vivo. Of particular interest, besides its established effects on cell proliferation and apoptosis, we have observed that CFTR is critical for the regulation of cell adhesion, invasion and migration in cancer. Thus, loss of CFTR function inevitably results in accelerated tumor progression. It has been well established that uPA/uPAR axis has a central role in cell proliferation, angiogenesis, extracellular matrix degradation, invasion and metastasis during cancer development.29 Moreover, overexpression of uPA and uPAR has been shown to be associated with adverse prognosis in various malignancies.17 In prostate cancer, altered expression of uPA and PAI-1 is associated with adverse pathological features, such as higher Gleason sum and lymph nodal and skeletal metastasis.30,31 In our study, we have demonstrated that the expression and Oncogene (2013) 2282 – 2291


2286 CFTR

CFTR

eGFPC3

PCNA

Tumor weight (g)

0.50

0.25 *

0.00 PC-3 CFTR Bcl-2

PC-3 eGFPC3 0.20

0.10 uPA

Tumor weight (g)

* 0.15

0.05

0.00

uPAR

PC-3 pEF/His PC-3 CFTRrib2

Figure 3. CFTR expression levels affect tumorigenicity in vivo. (a) In vivo xenograft assay for PC-3 cells transfected with empty vector or CFTR. Tumor size were measured 4 weeks after injection of 2 106 tumorigenic PC-3 cells transfected with peGFP-C3 or CFTR vectors with matrigel (1:1). Data shown are mean from six mice in the control vector group and seven mice in the CFTR-overexpressing group, P ¼ 0.027. (b) Immunohistochemical staining for CFTR, PCNA, Bcl-2, uPA and uPAR of xenograft tumors from either control vector transfected or CFTRtransfected PC-3 cells. Scale bar, 100 mm. (c) In vivo xenograft assay for PC-3 cells transfected with empty vector or CFTRrib2. Tumor size was measured 4 weeks after injection of 2 106 tumorigenic PC-3 cells transfected with pEF-His or CFTRrib2 vectors with matrigel (1:1). pEF/His (n ¼ 4) vs CFTR rib2 (n ¼ 4); P ¼ 0.049.

activity of uPA is inversely associated with CFTR expression in prostate cancer cell lines (Figures 4a and b). Moreover, CFTR overexpressing dramatically suppressed uPA/uPAR expression in tumor xenografts (Figures 3b and 6d). In addition, we have shown that CFTR knockdown-enhanced cell proliferation, cell invasion and migration are virtually reversed by either uPA or uPAR antibodies (Figures 4c–e and Supplementary Figure 3), indicating that the tumor-suppressive role of CFTR is attributed to the inhibition of uPA-mediated multiple malignant traits in cancer cells. Interestingly, an increasing number of proteins have been found to interact directly or indirectly with CFTR to form macromolecular complexes, indicating CFTR might have more complex and extensive role than we thought.12 Indeed, besides uPA, we have found that overexpression of CFTR in PC-3 cell line downregulates a variety of genes involved in tumor invasion and migration, such as TWIST1, FGFR2, ITGA1, THBS1 (data not shown). However, more research is required to identify the mechanism underlying the association between CFTR and these genes. An interesting finding from the present study is that the regulation of uPA by CFTR involves an epigenetic mechanism, namely miR-193b, which has previously been shown to target Oncogene (2013) 2282 – 2291

uPA.24 Although oversexpression or knockdown of CFTR resulted in miR-193b upregulation and downregulation, respectively, in both cancer cell lines and tumor xenografts (Figures 5a, b and g–i and Supplementary Figure 4), forced overexpression of miR-193b significantly reversed CFTR knockdown-enhanced proliferation, migration or invasion (Figures 5c–e), and abrogated the CFTR knockdown-elevated uPA activity in PC-3 cells (Figure 5f). More interestingly, in an attempt to evaluate the correlation of CFTR and miR-193b, we determined both CFTR and miR-193b expression in six cases of prostate cancer samples, in which 80% of the cancer cells were ensured. As shown in Supplementary Figure 6, our quantitative PCR results revealed a significant correlation of CFTR with miR-193b. Collectively, these data suggest that CFTR suppresses tumor progression through miR-193b targeting uPA. Of note, uPA is also known to be activated by NFkB, a well-known transcription factor involved in inflammation32,33 and cancer.34 More interestingly, defective CFTR has been shown to result in endogenous activation of NF-kB,35–37 thus, it is possible that apart from miR-193b, CFTR may regulate uPA through NFkB. Indeed, in this study, we found that CFTR knockdown lead to enhanced activation of NFkB in prostate cancer cell lines & 2013 Macmillan Publishers Limited


2287

Rib uPA uPAR

Viable Cells (% increase)

uPA activity/unit

PC-3 pEF/His

600

***

40

30

20

10

β-Tubulin 0

400

200

0 0

2

4

6

Days #

75

#

# *

*

400

*

# #

CFTRrib2 PC-3

#

500

*

Invaded Cells

Migrated Cells

pEF/His

His His+anti uPA His+anti uPAR Rib Rib+anti uPA Rib+anti uPAR

300 200

* 50

25

100 0

uP AR

ib

+a nt i

ib

uP A

R

+a nt i ib R

is

R

uP AR

uP A

nt i

+a nt i H

H

is

nt +a ib

+a

is H

R iu

PA

uP

A

R ib

R PA

ti +a n ib R

R

is H

H

is

+a

nt

+a

nt

iu

iu

PA

H is

0

Figure 4. Involvement of uPA in the tumor-suppressing effects of CFTR. (a) The protein expression of uPA is upregulated in CFTR knockdown PC-3 cells compared with vector control cells. Image shown is the representative of three independent experiments. (b) UPA activity in CFTR knockdown PC-3 cells compared with vector control cells. The experiment was performed three independent times in triplicate, ***Po0.001. (c–e) PC-3 pEF-His or CFTRrib2-transfected cells were incubated with either 10 mg/ml uPA (Santa Cruz, SC-6830) or uPAR (Santa Cruz, SC-10815) neutralizing antibody, 10 mg/ml IgG was used as antibody control (c) MTS assay shows that the enhanced cell growth in PC-3 CFTRrib2 is significantly reversed by either uPA or uPAR antibody. CFTR rib2 compared with pEF-His, *Po0.05; CFTR rib2 treated with uPA or uPAR antibodies vs pEF-His þ IgG, #P40.05. (d) Modified Boyden chamber cell migration assay shows that knockdown of CFTR significantly promotes cell migration in 72 h. However, the treatment with either uPA or uPAR antibody nearly reverses the enhanced migration in PC-3 CFTRrib2 cells, CFTR rib2 treated with uPA or uPAR antibodies vs CFTRrib2 treated with IgG, *Po0.05. CFTR rib2 treated with uPA or uPAR antibodies vs pEF-His þ IgG, #P40.05. (e) Invasion assay shows that the treatment with either uPA or uPAR antibody significantly reverses the enhanced invasion in PC-3 CFTRrib2 cells, CFTR rib2 treated with uPA or uPAR antibodies vs CFTRrib2 treated with IgG, *Po0.05. CFTR rib2 treated with uPA or uPAR antibodies vs pEF-His þ IgG, #P40.05.

(Supplementary Figure 5a). In addition, we have shown that NFkB inhibitor reverses the upregulation of uPA in CFTR knockdown PC3 cells (Supplementary Figure 5b). Taken together, it appears that CFTR may suppress uPA-mediated tumor activities through both genetic and epigenetic mechanisms. Another interesting observation in this study is the agedependent downregulation of CFTR in rat prostate (Supplementary Figure 1), which is consistent with the age-dependent prevalence of prostate hyperplasia and cancer.38,39 As castration results in suppressed CFTR expression and testosterone treatment upregulates CFTR expression both in post-castration rat prostate tissue and in cultured primary prostate epithelial cells (Supplementary Figures 7b and c), it is likely that in aging prostate, reduced levels of testosterone may be responsible for the loss of CFTR, either through direct inhibition of CFTR promoter or induction of CFTR gene methylation, which confers cellular susceptibility to malignant transformation. The regulation of CFTR expression by testosterone is of important clinical relevance. Clinically, most patients receiving androgen deprivation therapy tend to develop a more aggressive type of cancer eventually, which has been explained by the heterogenenous responsiveness to androgen in different subpopulations of cancer cells.40 Our findings that androgen upregulates CFTR and that repressed CFTR expression promotes cancer development provides an alternative mechanism underlying the observed dilemma in prostate cancer treatment, which may profoundly change the clinical practice regarding treatment strategy towards prostate cancer. It is particularly noteworthy that in our experimental models, overexpression of CFTR suppressed cancer development in both androgen dependent (LNcap) and androgen-independent (PC-3) cancer cell lines and delivery of CFTR to prostate cancer xenografts & 2013 Macmillan Publishers Limited

was able to induce a marked tumor regression. Thus, CFTR may have considerable therapeutic potential, regardless of androgen dependence. Although the mechanisms regulating CFTR expression are poorly understood, the CFTR gene is reported to be frequently hypermethylated in various primary cancer samples and cell lines. Interestingly, a recent study investigating 139 cases of primary non-small-cell lung cancer clearly revealed that CFTR gene methylation was significantly high in cancer patients and associated with poorer survival in young patients.16 Although there is no report regarding CFTR methylation status in prostate cancer patients to this end, CFTR promoter has been reported to be hypermethylated in prostate cancer cell line PC-3, DU-145 and LNcap,15 which is associated with its low expression level in these cell lines. The findings that CFTR is downregulated in prostate cancer cell line and patients’ samples, along with CFTR overexpression suppresses prostate cancer progression, warrant future investigations on CFTR methylation or mutation in prostate cancer patients, which are likely to provide attractive and novel targets for prognosis and therapeutic intervention of prostate cancer.

MATERIALS AND METHODS Cell lines and tumor samples We obtained PC-3, LNcap, DU-145 and PNT1A cell lines from American Type Culture Collection (Manassas, VA, USA) and cultivated them in the recommended media. We obtained prostate cancer samples from radical prostatectomies at the Department of Anatomical and Cellular Pathology, Prince of Wales Hospital, the Chinese University of Hong Kong. All samples were collected with the informed consent of the patients and the study Oncogene (2013) 2282 – 2291


2288 *** Viable Cells (% increase)

0 F/ H is (

C

C FT

pE

PC -3

PC -3

co n R rib ) R (c rib o (p n) re -1 93 b)

FP

FT C

PC -3

R

pe G

rib

H F/ pE

60

0

2

is

0.0

*

C FT

***

1

**

120

R

0.5

180

FT

1.0

PC -3

2

miR193b Relative Expression

miR193b Relative Expression

1.5

4 3 2 1

b)

-1

(c rib

(p

re

rib R FT

93

on

)

n) co

FP

1c G tl FP G 2 FP G 3 F C P4 FT C R1 FT C R2 FT C R3 FT R 4

0

0

C

pE

miR193b Relative Expression in Xenografts

1_ c PE tl F PE -2 F PE -3 FR 4 ib R -1 ib R 2 ib R 3 ib -4

0.0

5

10

4

control vector

*

CFTR manipulation

3 2 1 0

*

CFTR knockdown

CFTR overexpression

G

0.5

PE

F/ H is C FT (co n C R FT rib ) R (c rib on (p ) re -1 93 b)

F/ H is (c o C R FT rib n) R ( c rib on (p ) re -1 93 b) FT

pE

C

1.0

F-

miR193b Relative Expression in Xenografts

1.5

20

R

0

0

***

FT

25

**

30

C

50

40

uPA activity (unit)

60

**

F/ H is (

120

**

75

pE

100

**

miR193b Relative Expression in xenografts

**

Invaded cells

Migrated Cells

180

Figure 5. MiR-193b mediates the effect of CFTR on uPA. (a) The expression of miR-193b is significantly repressed in CFTR knockdown PC-3 cells compared with vector control cells by Q-PCR analysis, ***Po0.001. (b) The expression of miR-193b is significantly increased in CFTRoverexpressing PC-3 cells compared with the vector control cells by Q-PCR analysis, ***Po0.001. (c–f ) PC-3 CFTRrib2 cells were transfected with either control Pre-miRNA or Pre-miR-193b, and compared with PC-3 pEF-His cells transfected with control Pre-miRNA. (c) MTS assay shows that the enhanced cell growth in PC-3 CFTRrib2 is significantly reversed by overexpression of pre-miR-193b, **Po0.01; *Po0.05. (d) Cytodex-2 bead motility assay shows that the overexpression of pre-miR-193b nearly reverses the enhanced migration in PC-3 CFTRrib2 cells, **Po0.01. (e) Invasion assay shows that the overexpression of pre-miR-193b completely reverses the enhanced invasion in PC-3 CFTRrib2 cells in 24 h, **Po0.01. (f) uPA activity assay shows that the overexpression of pre-miR-193b completely reverses the enhanced uPA activity in PC-3 CFTRrib2 cells, ***Po0.001; **Po0.01. (g) MiR-193b expression in four control and four CFTR knockdown PC-3 xenografts as measured by Q-PCR. (h) MiR-193b expression in four control and four CFTR overexpression PC-3 xenografts as measured by Q-PCR. (i) Quantitative data for (g) and (h) showing miR-193b is increased in CFTR-overexpressing xenografts, whereas decreased in CFTR knockdown xenografts, *Po0.05.

was approved by the Ethics Committee of Prince of Wales Hospital. Microtissue array was purchased from US Biomax, Inc (MC5003, PR808 Rockville, MD, USA).

In vitro growth assay The growth rate of cells was measured using the CellTiter 96 MTS Aqueous One Solution Cell Proliferation Assay (Promega, Madison, WI, USA) at the indicated time points. In brief, 3000 cells in 200 ml of normal medium were seeded into each well of a 96-well plate. The results were expressed as a percentage increase and calculated by comparing the absorbance obtained for each incubation period using the following equation: Percentage increase ¼ absorbancedn absorbanced0/absorbanced0. Three independent experiments in triplicate were performed.

Cell adhesion assay The 96-well plates were pre-coated with matrigel (5 mg/well, BD Matrigel, Franklin Lakes, NJ, USA) and air dried. This membrane was then rehydrated in 100 ml of serum-free medium for 40 min before cell seeding. After rehydration, 40 000 cells were seeded in 200 ml of normal medium and incubated for 40 min. Following incubation, non-adherent or loosely attached cells were washed off. Adherent cells were then fixed in 4% formaldehyde (v/v) for 5 min before being stained in 0.5% crystal violet solution (w/v) in distilled water. The remaining adhered cells were then visualized under the microscope and random fields counted. At least four random fields per well were counted and three independent experiments in quadruplicate were performed.

Cell motility assay Flow cytometric analysis Apoptosis was measured by FACS using AnnexinV/PI double staining (BD Biosciences, San Jose, CA, USA). The cells were seeded in culture flask and grown for 24 h. In all, 1 105 cells were harvested and washed with PBS and then resuspended in 1 binding buffer and stained with 5 ml FITC Annexin V and 5 ml PI according to manufacturing protocol. Flow cytometry was then performed by using SD FACS Calibur System (BD Biosciences). All experiments were performed at least in triplicates. Oncogene (2013) 2282 – 2291

Cell motility was assessed using a cytodex-2 bead motility assay. Cells (1 106) for each cell type were incubated with growth medium containing 100 ml of cytodex-2 beads (Pharmacia, Piscataway, NJ, USA) for 4 h to allow the cells to adhere to the beads. The beads were washed twice in 5 ml of normal growth medium to remove non-adherent or dead cells. After the second wash the beads were resuspended in 1 ml of growth medium. Two hundred microlitres of this solution was then added to a 24well plate containing a further 800 ml of normal medium and incubated overnight. Following incubation, any cells that had migrated from the & 2013 Macmillan Publishers Limited


2289

Control CFTR-treated

Tumor weight (g)

0.75

0.50

***

0.25

0.00 Control CFTR-treated H&E

PCNA

Bcl-2

CFTR 1

2

1

2 CFTR uPA

CFTR-treated

Control

Control CFTR-treated

uPAR GAPDH

Figure 6. Therapeutic potential of CFTR in cancer. (a, b) Effect on tumor growth of CFTR reconstitution. (a) Tumors generated 2 weeks after injection of 2 106 PC-3 cells were treated with bubble conjugated DNA containing peGFP-C3 (upper row in inset photo) or CFTR vectors (lower row). (b) Data are mean from 13 mice in the control vector group and 12 mice in the CFTR reconstituting group, ***P ¼ 0.0004. (c) H&E and immunohistochemical staining of PCNA, Bcl-2 and CFTR from control or CFTR-treated PC-3 xenograft tumors. Rare living tumor cell islands in a necrotic tumor treated with CFTR vector are indicated by dashed line, scale bar, 100 mm. (d) UPA and uPAR expression in two control and two CFTR-treated PC-3 xenografts as measured by western blot.

cytodex-2 beads and adhered to the base of the well were fixed in 4% formaldehyde (v/v) for 5 min, stained with 0.5% crystal violet (w/v) and counted, following removal of cytodex-2 beads through several extensive washes with PBS. At least four random fields were counted per well and triplicate wells were set up per sample. The entire experimental procedure was repeated three independent times.

Cell migration assay We tested cell migration in modified Boyden chambers containing porous (8 mm) polycarbonate. A total of 20 000 cells were added to the transwell inserts over the top of the artificial basement membrane. The plate was then incubated for 72 h at 37 1C, 5% CO2 and 95% humidity. After 72 h, the inserts were removed from the plate and the inside of the insert was cleaned thoroughly with tissue paper. Any cells that had migrated through the membrane and passed to the underside of the insert were fixed in 4% formaldehyde (v/v) in PBS for 5 min before being stained in 0.5% crystal violet solution (w/v) in distilled water. These cells could then be visualized under the microscope and random fields counted. At least three random fields per insert were counted and triplicate inserts were set up for each test sample. The experimental procedure was repeated for three times.

Cell invasion assay We tested invasion in modified Boyden chambers containing porous (8 mm), polycarbonate membranes (Corning Incorporated, Big Flats, NY, USA) coated with 500 mg/ml matrigel. A total of 20 000 cells were added to the transwell inserts over the top of the artificial basement membrane. The plate was then incubated for 24 h at 37 1C, 5% CO2 and 95% humidity. After 24 h, the inserts were removed from the plate and the inside of the insert was cleaned thoroughly with tissue paper. Any cells that had invaded through the membrane and passed to the underside of the insert were fixed in 4% formaldehyde (v/v) in PBS for 5 min before being stained in 0.5% crystal violet solution (w/v) in distilled water. These cells could then be visualized under the microscope and random fields counted. At least four random fields per insert were counted and triplicate inserts were set up for each test sample. The experimental procedure was repeated for three times.

In vivo assay Tumorigenicity was investigated by tumor xenograft experiments. The athymic female nude mice of 4–6 weeks old were provided from the Laboratory Animal Service Center (LASEC), the Chinese University of Hong & 2013 Macmillan Publishers Limited

Kong and maintained in filter-topped units. All experimental procedures were under ethical approval from LASEC. Approximately 100 ml suspension of CFTR knockdown or overexpressed cells and the same amount of vector control cells (about 2 106 in 2.0 mg/ml matrigel) was injected subcutaneously. Mice injected with saline were used as sham control. Tumor formation in nude mice was monitored over about 4 week period and the tumor weight measured. Mice with tumor size larger than 1 cm in any dimension were terminated. For ultrasound-mediated gene delivery of CFTR, the plasmid peGFP-C3 and peGFP-CFTR for injection were prepared using the EndoFree plasmid kit (Qiagen Inc., Valencia, CA, USA) according to the manufacturer’s instructions. In all, 2 106 PC-3 cells were subcutaneously injected into the animals. The tumors were allowed to grow to o0.5 cm in any dimension for following CFTR gene transfer mediated by ultrasound-microbubble technique.25 Briefly, after mixing peGFP-C3 or peGFP-CFTR with sulphur hexafluoride microbubbles (Bracco Imaging B.V., Geneva, Switzerland) at a ratio of 1:1 (v/v), the mixed solution containing 200 mg of plasmid in 200 ml was injected into tail vein. Immediately after injection, the ultrasound transducer was directly applied to the tumor site and a continuous-wave output of 1 MHz was applied for a total 5 min. The same procedure was repeated once a week for about 2 weeks. At the end of the experiment, the animals were killed and the tumor tissues were collected for further analysis. For castration experiments, the adult 12-week-old male SD rats weighed between 250 and 350 g were used. In brief, the male rats were castrated under anesthesia and kept under stable condition for 1 week. From the seventh day, 10 mg/ml testosterone diluted by sesame oil was subcutaneously injected into the back of rats for 10 days and the rats of control group were injected with sesame oil only. The rats were killed by CO2 and the ventral lobes of prostate obtained from each rat were weighted.

Gene overexpression and knockdown For the CFTR knockdown system, hammerhead ribozymes were designed based on the secondary structure of CFTR using the Zuker RNA mFold program,41 targeting at a specific GUC or AUC site. The ribozymes were generated by touch-down PCR with the primers as follows: Forward 50 -CTGCAGAGAA GGCATAAGCCTATGCCTACTGATGAGTCCGTGAGGA-30 , Reverse 50 -ACTAGTACCC GGATAACAAGGAGGAACGCTCTATCGCGATTTATTTCGTCCTCACGGACT-30 . The ribozymes were then cloned into the pEF6/V5-His vector (Invitrogen, Carlsbad, CA, USA). For knock down experiments, 1 mg DNA was transfected into cells by Easyjet Plus electroporator (EquiBio, Kent, UK) and selected by 5 mg/ml Blasticidin S for 2–3 weeks. The stable transfectants were maintained in medium containing 0.5 mg/ml Blasticidin S. The peGFP-C3 Oncogene (2013) 2282 – 2291


2290 plasmid expressing wild-type CFTR was kindly provided by Professor Tzyh-Chang Hwang (University of Missouri-Columbia). For overexpression experiments, the cells were transfected with 2 mg DNA and 7 ml Lipofectamine 2000 reagent (Invitrogen), the transfected cells were selected in full medium containing G418 (Calbiochem, Schwalbach, Germany) at 800 mg/ml for LNCap cells and 500 mg/ml for PC-3 cells. After 2–3 weeks of selection, G418-resistant cells growing in single colony were isolated and collected for western blotting. Stable transfectants were then maintained in 500 mg/ml G418 for LNCap cells and 300 mg/ml for PC-3 cells for subsequent studies. The pre-miR-193b and control-miRNA were purchased from Applied Biosystem (Carlsbad, CA, USA, PM12383). MiRNA transfections were performed using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s protocol.

Immunohistochemistry and immunofluoresence staining We performed immunohistochemistry experiments on 5 mm thick formalinfixed, paraffin-embedded tissue slices. Bcl-2, PCNA, uPA and uPAR were from Santa Cruz Biotech (Santa Cruz, CA, USA), and CFTR was from Almone Labs (Jerusalem, Israel). The universal biotinylated secondary antibody in UltraVision One HRP Polymer detection kit (Thermo Fisher Scientific, Fremont, CA, USA) was applied to the sections, and incubated at room temperature for 1 h in dark room according to the manufacturer’s instructions. The tissue microarray slides were deparaffinized in xylene, rehydrated through graded alcohol, immersed in 3% hydrogen peroxide for 10 min to block endogenous peroxidase activity and antigen retrieved by pressure cooking for 3 min in citrate buffer (pH ¼ 6). For blocking nonspecific binding, the slides were preincubated with 10% normal goat serum at room temperature for 20 min. Subsequently, the slides were incubated with anti-CFTR antibody (Almone labs, 1:200) and/or uPA antibody (Santa Cruz, 1:200) overnight at 4 1C in a moist chamber. Cytoplasmic and/or membrane immunoreactivity for CFTR was scored using a semiquantitative method by evaluating the intensity of positive tumor cells. For CFTR, staining intensity of þ þ þ , þ þ , þ B± was scored at 3, 2, 1, respectively.

Protein extraction, western blotting, RNA extraction and quantitative real-time PCR Total or nuclear and cytoplasmic extracts were prepared as standard method. Protein concentration was determined using Bradford reagent. Antibodies used in this study were: anti-CFTR (Alexis Biochemicals, 1:800), anti-uPA and anti-uPAR (Santa Cruz, 1:200), anti-p65, anti-p50, anti-b tubulin, anti-GAPDH and anti-histone (Santa Cruz, 1:1000). CFTR and miR193b TaqMan primer and probes were obtained from Applied Biosystems. Total RNA was isolated and reversely transcribed using standard protocols. For quantitative PCR, assays were performed in triplicate on an Applied Biosystems 7500Fast Real-Time PCR System.

uPA activity ELISA assay uPA activity assay kit from Millipore (Billerica, MA, USA) was used according to the protocol provided by the manufacture.

Statistical analyses Data are presented as the mean±s.e.m. Student’s unpaired t-test or w2-test were used for two groups of statistical analysis and Pearson’s correlation test was used for correlation study. Differences between groups were analyzed using SPSS (New York, NY, USA). A P-value o0.05 was considered statistically significant.

CONFLICT OF INTEREST The authors declare no conflict of interest.

ACKNOWLEDGEMENTS The work was supported by the Focused Investment Scheme of the Chinese University of Hong Kong and National 973 project (2012CB944900), GRFCUHK466111, and the Fundamental Research Funds for the Central Universities (JiNan University). We are grateful to Professor Tzyh-Chang Hwang at Department of Biological Engineering, University of Missouri-Columbia, USA for providing CFTRpeGFP-C3 and control plasmids. The authors thank Mrs Huang Xiao ru and Mr Qin Wei from Li Ka Shing Institute of Health Sciences, The Chinese University of Hong Kong, for their helpful technical assistance.

Oncogene (2013) 2282 – 2291

REFERENCES 1 Collins FS. Cystic fibrosis: molecular biology and therapeutic implications. Science 1992; 256: 774–779. 2 Riordan JR, Rommens JM, Kerem B, Alon N, Rozmahel R, Grzelczak Z et al. Identification of the cystic fibrosis gene: cloning and characterization of complementary DNA. Science 1989; 245: 1066–1073. 3 Cuthbert AW. New horizons in the treatment of cystic fibrosis. Br J Pharmacol 2011; 163: 173–183. 4 Wiehe M, Arndt K. Cystic fibrosis: a systems review. AANA J 2010; 78: 246–251. 5 McWilliams RR, Rabe KG, Olswold C, De AM, Petersen GM. Risk of malignancy in first-degree relatives of patients with pancreatic carcinoma. Cancer 2005; 104: 388–394. 6 Neglia JP, FitzSimmons SC, Maisonneuve P, Schoni MH, Schoni-Affolter F, Corey M et al. The risk of cancer among patients with cystic fibrosis. Cystic Fibrosis and Cancer Study Group. N Engl J Med 1995; 332: 494–499. 7 McWilliams RR, Petersen GM, Rabe KG, Holtegaard LM, Lynch PJ, Bishop MD et al. Cystic fibrosis transmembrane conductance regulator (CFTR) gene mutations and risk for pancreatic adenocarcinoma. Cancer 2010; 116: 203–209. 8 Warren N, Holmes JA, al-Jader L, West RR, Lewis DC, Padua RA. Frequency of carriers of cystic fibrosis gene among patients with myeloid malignancy and melanoma. BMJ 1991; 302: 760–761. 9 Abraham EH, Vos P, Kahn J, Grubman SA, Jefferson DM, Ding I et al. Cystic fibrosis hetero- and homozygosity is associated with inhibition of breast cancer growth. Nat Med 1996; 2: 593–596. 10 Padua RA, Warren N, Grimshaw D, Smith M, Lewis C, Whittaker J et al. The cystic fibrosis delta F508 gene mutation and cancer. Hum Mutat 1997; 10: 45–48. 11 Qiao D, Yi L, Hua L, Xu Z, Ding Y, Shi D et al. Cystic fibrosis transmembrane conductance regulator (CFTR) gene 5T allele may protect against prostate cancer: a case-control study in Chinese Han population. J Cyst Fibros 2008; 7: 210–214. 12 Li Y, Sun Z, Wu Y, Babovic-Vuksanovic D, Li Y, Cunningham JM et al. Cystic fibrosis transmembrane conductance regulator gene mutation and lung cancer risk. Lung Cancer 2010; 70: 14–21. 13 Ding S, Gong BD, Yu J, Gu J, Zhang HY, Shang ZB et al. Methylation profile of the promoter CpG islands of 14 ‘drug-resistance’ genes in hepatocellular carcinoma. World J Gastroenterol 2004; 10: 3433–3440. 14 Moribe T, Iizuka N, Miura T, Kimura N, Tamatsukuri S, Ishitsuka H et al. Methylation of multiple genes as molecular markers for diagnosis of a small, well-differentiated hepatocellular carcinoma. Int J Cancer 2009; 125: 388–397. 15 Mishra DK, Chen Z, Wu Y, Sarkissyan M, Koeffler HP, Vadgama JV. Global methylation pattern of genes in androgen-sensitive and androgen-independent prostate cancer cells. Mol Cancer Ther 2010; 9: 33–45. 16 Son JW, Kim YJ, Cho HM, Lee SY, Lee SM, Kang JK et al. Promoter hypermethylation of the CFTR gene and clinical/pathological features associated with non-small cell lung cancer. Respirology 2011; 16: 1203–1209. 17 McMahon B, Kwaan HC. The plasminogen activator system and cancer. Pathophysiol Haemost Thromb 2008; 36: 184–194. 18 Malinowsky K, Bollner C, Hipp S, Berg D, Schmitt M, Becker KF. UPA and PAI-1 analysis from fixed tissues—new perspectives for a known set of predictive markers. Curr Med Chem 2010; 17: 4370–4377. 19 Gregory PA, Bert AG, Paterson EL, Barry SC, Tsykin A, Farshid G et al. The miR-200 family and miR-205 regulate epithelial to mesenchymal transition by targeting ZEB1 and SIP1. Nat Cell Biol 2008; 10: 593–601. 20 Korpal M, Lee ES, Hu G, Kang Y. The miR-200 family inhibits epithelialmesenchymal transition and cancer cell migration by direct targeting of E-cadherin transcriptional repressors ZEB1 and ZEB2. J Biol Chem 2008; 283: 14910–14914. 21 Rauhala HE, Jalava SE, Isotalo J, Bracken H, Lehmusvaara S, Tammela TL et al. miR193b is an epigenetically regulated putative tumor suppressor in prostate cancer. Int J Cancer 2010; 127: 1363–1372. 22 Xu C, Liu S, Fu H, Li S, Tie Y, Zhu J et al. MicroRNA-193b regulates proliferation, migration and invasion in human hepatocellular carcinoma cells. Eur J Cancer 2010; 46: 2828–2836. 23 Chen J, Zhang X, Lentz C, bi-Daoud M, Pare GC, Yang X et al. miR-193b Regulates Mcl-1 in Melanoma. Am J Pathol 2011; 179: 2162–2168. 24 Li XF, Yan PJ, Shao ZM. Downregulation of miR-193b contributes to enhance urokinase-type plasminogen activator (uPA) expression and tumor progression and invasion in human breast cancer. Oncogene 2009; 28: 3937–3948. 25 Hou CC, Wang W, Huang XR, Fu P, Chen TH, Sheikh-Hamad D et al. Ultrasoundmicrobubble-mediated gene transfer of inducible Smad7 blocks transforming growth factor-beta signaling and fibrosis in rat remnant kidney. Am J Pathol 2005; 166: 761–771. 26 Singh AP, Chauhan SC, Andrianifahanana M, Moniaux N, Meza JL, Copin MC et al. MUC4 expression is regulated by cystic fibrosis transmembrane conductance regulator in pancreatic adenocarcinoma cells via transcriptional and post-translational mechanisms. Oncogene 2007; 26: 30–41.

& 2013 Macmillan Publishers Limited


2291 27 Singh AP, Moniaux N, Chauhan SC, Meza JL, Batra SK. Inhibition of MUC4 expression suppresses pancreatic tumor cell growth and metastasis. Cancer Res 2004; 64: 622–630. 28 Southey MC, Batten L, Andersen CR, McCredie MR, Giles GG, Dite G et al. CFTR deltaF508 carrier status, risk of breast cancer before the age of 40 and histological grading in a population-based case-control study. Int J Cancer 1998; 79: 487–489. 29 Smith HW, Marshall CJ. Regulation of cell signalling by uPAR. Nat Rev Mol Cell Biol 2010; 11: 23–36. 30 Cozzi PJ, Wang J, Delprado W, Madigan MC, Fairy S, Russell PJ et al. Evaluation of urokinase plasminogen activator and its receptor in different grades of human prostate cancer. Hum Pathol 2006; 37: 1442–1451. 31 Kumano M, Miyake H, Muramaki M, Furukawa J, Takenaka A, Fujisawa M. Expression of urokinase-type plasminogen activator system in prostate cancer: correlation with clinicopathological outcomes in patients undergoing radical prostatectomy. Urol Oncol 2009; 27: 180–186. 32 Sen R, Smale ST. Selectivity of the NF-{kappa}B response. Cold Spring Harb Perspect Biol 2010; 2: a000257. 33 Lawrence T. The nuclear factor NF-kappaB pathway in inflammation. Cold Spring Harb Perspect Biol 2009; 1: a001651.

34 Wu Y, Zhou BP. TNF-alpha/NF-kappaB/Snail pathway in cancer cell migration and invasion. Br J Cancer 2010; 102: 639–644. 35 Boncoeur E, Roque T, Bonvin E, Saint-Criq V, Bonora M, Clement A et al. Cystic fibrosis transmembrane conductance regulator controls lung proteasomal degradation and nuclear factor-kappaB activity in conditions of oxidative stress. Am J Pathol 2008; 172: 1184–1194. 36 DiMango E, Ratner AJ, Bryan R, Tabibi S, Prince A. Activation of NF-kappaB by adherent Pseudomonas aeruginosa in normal and cystic fibrosis respiratory epithelial cells. J Clin Invest 1998; 101: 2598–2605. 37 Vij N, Mazur S, Zeitlin PL. CFTR is a negative regulator of NFkappaB mediated innate immune response. PLoS One 2009; 4: e4664. 38 Hsing AW, Chokkalingam AP. Prostate cancer epidemiology. Front Biosci 2006; 11: 1388–1413. 39 Jemal A, Siegel R, Ward E, Murray T, Xu J, Thun MJ. Cancer statistics, 2007. Cancer J Clin 2007; 57: 43–66. 40 Niu Y, Altuwaijri S, Lai KP, Wu CT, Ricke WA, Messing EM et al. Androgen receptor is a tumor suppressor and proliferator in prostate cancer. Proc Natl Acad Sci USA 2008; 105: 12182–12187. 41 Zuker M. Mfold web server for nucleic acid folding and hybridization prediction. Nucleic Acids Res 2003; 31: 3406–3415.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

& 2013 Macmillan Publishers Limited

Oncogene (2013) 2282 – 2291