Page 1 of 41

Accepted Preprint first posted on 23 September 2014 as Manuscript ERC-14-0152

1

CUX1 - a modulator of tumour aggressiveness in pancreatic neuroendocrine neoplasms

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Short title: CUX1 in pancreatic neuroendocrine neoplasms

20 21

Key words: CUX1, pancreatic neuroendocrine neoplasms, angiogenesis, RIP1Tag2, insulinomas

22

Word count: 4.490

23 24 25 26 27 28 29 30 31 32 33 34 35 36

Sebastian Krug1, Benjamin Kühnemuth1, Heidi Griesmann1, Albrecht Neesse1, Leonie Mühlberg1, Michael Boch1, Juliane Kortenhaus1, Volker Fendrich2, Dominik Wiese2, Bence Sipos3, Juliane Friemel4, Thomas M. Gress1 and Patrick Michl1

1

Department of Gastroenterology, Endocrinology and Metabolism, PhilippsUniversity Marburg, Germany 2 Department of Surgery, Philipps-University Marburg, Germany 3 Department of Pathology, Eberhard-Karls-University Tübingen, Germany 4 Department of Pathology, University Hospital Zurich, Switzerland

Corresponding author: Prof. Dr. Patrick Michl Department of Gastroenterology, Endocrinology and Metabolism Philipps-University Marburg Baldingerstraße 35043 Marburg Germany Phone: +49 6421 5866460 Fax: +49 6421 5868922 Email: [email protected]

37 38 39

1

Copyright © 2014 by the Society for Endocrinology.

Page 2 of 41

40

ABSTRACT

41

Pancreatic neuroendocrine neoplasms (PNENs) constitute a rare tumour entity, and

42

prognosis and treatment options depend on tumour hallmarks such as angiogenesis,

43

proliferation rate and resistance to apoptosis. Molecular pathways that determine the

44

malignant phenotype are still insufficiently understood and have limited the use of

45

effective combination therapies in the past. Here, we aim to characterize the impact

46

of the oncogenic transcription factor CUX1 on proliferation, resistance to apoptosis

47

and angiogenesis in murine and human PNENs. Expression and function of CUX1

48

was analyzed using knockdown and overexpression strategies in Ins-1 and Bon-1

49

cells, xenograft models and a genetically engineered mouse model of insulinoma

50

(RIP1Tag2). Regulation of angiogenesis was assessed using RNA profiling and

51

functional tube formation assays in HMEC-1 cells. Finally, CUX1 expression was

52

assessed in a tissue microarray of 59 human insulinomas and correlated with

53

clinicopathological

54

progression in a time- and stage-dependent manner in the RIP1Tag2 model, and

55

associated with pro-invasive and metastatic features in human insulinomas.

56

Endogenous and recombinant CUX1 expression increased tumour cell proliferation,

57

tumour growth, resistance to apoptosis, and angiogenesis in vitro and in vivo.

58

Mechanistically, the pro-angiogenic role of CUX1 was mediated via upregulation of

59

effectors such as HIF-1α and MMP-9. CUX1 mediates an invasive pro-angiogenic

60

phenotype and is associated with a malignant behaviour in human insulinomas.

data.

CUX1 expression was

upregulated

during tumour

61 62 63

2

Page 3 of 41

64

INTRODUCTION

65

Gastroenteropancreatic

66

heterogeneous group of solid cancers, based on embryogenic, morphological and

67

biochemical findings (Modlin et al. 2008; Rindi and Wiedenmann 2012). The

68

incidence and prevalence of neuroendocrine neoplasms (NEN) is constantly

69

increasing with an estimated number of 5.76/100,000 and 35/100,000, respectively

70

(Lawrence et al. 2011; Yao et al. 2008). The subgroup of pancreatic neuroendocrine

71

neoplasms (PNENs) is one of the top three entities in NEN with a reported

72

prevalence between 6-25% (Lawrence et al. 2011; Pape et al. 2008). The prognosis,

73

time to progression and response to treatment of PNENs largely depend on tumour

74

hallmarks such as proliferation rate and resistance to drug-induced apoptosis.

75

Metastasized PNENs show a poor 5-year survival rate of 37.6% with a median

76

overall survival of 18.9 months (Falconi et al. 2012; Larghi et al. 2012; Panzuto et al.

77

2011; Panzuto et al. 2012).

78

The most frequent genetic alterations in PNENs are inactivating mutations of the

79

MEN1 and DAXX/ATRX gene, dysregulation of the PI3K/AKT/mTOR signalling

80

pathway, and overactivation of growth factors and their receptors, such as VEGF

81

PDGF, IGF and c-Kit (Grothey and Galanis 2009; Jiao, et al. 2011; Missiaglia et al.

82

2010). Increased tumour angiogenesis is a hallmark histological feature of PNENs

83

and is generally correlated with aggressive tumour biology and disease progression

84

(Grothey and Galanis 2009). Preclinical data in an established genetically engineered

85

mouse model of insulinomas (RIP1Tag2) revealed promising activity of the multi-

86

tyrosine-kinase inhibitor sunitinib, which has been approved by the FDA for the

87

treatment of progressive, unresectable, locally advanced or metastatic PNENs

neuroendocrine

neoplasms

(GEP-NENs)

are

a

3

Page 4 of 41

88

(Olson et al. 2011; Raymond et al. 2011). However, although sunitinib significantly

89

prolonged the progression-free survival (PFS) in metastasized PNENs, it failed to

90

improve the overall survival (OS), and surrogate markers that predict the response to

91

treatment are currently not available (Raymond et al. 2011). Therefore, further

92

molecular characterization and functional validation of novel targets is urgently

93

needed and holds promise to classify PNENs into several prognostic and therapeutic

94

subgroups that can be employed for combined and personalized treatment strategies

95

in the future (Capurso et al. 2009; Capurso et al. 2012).

96

The transcription factor Cut homeobox 1 (CUX1), also called CCAAT-displacement

97

protein (CDP) or Cut-like1 (CUTL1), is located on chromosomal band 7q22.1 and is

98

expressed in multiple isoforms. Hence, CUX1 is involved in repressive and activating

99

transcriptional processes depending on the promoter regulation. Recent reports

100

suggest an important role for CUX1 in tumourigenesis and tumour progression

101

(Hulea and Nepveu 2012). We could show that CUX1 directly mediates tumour cell

102

proliferation, migration and invasiveness and also acts via the target genes GRIA3

103

and WNT5A in the pancreatic ductal adenocarcinoma (PDAC) (Aleksic et al. 2007;

104

Griesmann et al. 2013; Ripka et al. 2007; Ripka et al. 2010a; Ripka et al. 2010b).

105

Moreover, CUX1 expression is strongly associated with a less differentiated

106

phenotype and decreased survival in patients with breast cancer (Michl et al. 2005).

107

So far, the expression and function of CUX1 in PNENs have not been investigated.

108

Here, we aimed to elucidate the effects of CUX1 on tumour growth and progression

109

in neuroendocrine tumour cell lines as well as in a xenografts and the RIP1Tag2

110

model in vitro and in vivo. Additionally, we investigated potential mechanisms by

111

which CUX1 mediates angiogenesis and assessed the CUX1 expression in a series

112

of human insulinomas. 4

Page 5 of 41

113

MATERIALS AND METHODS

114

Material and Cell Lines

115

All media contained 10% fetal bovine serum and 40 µg/ml gentamicin. The cell-lines

116

were cultured in the indicated media: human Bon-1 carcinoid cells were a kindly gift

117

of R. Göke, University of Marburg, Germany (Parekh, et al. 1994, passage 10-30):

118

DMEM/HAM's F12 medium; rat Ins-1 insulinoma cells, donated by C. B. Wollheim,

119

University of Geneva, Switzerland (Asfari et al. 1992, Lankat-Buttgereit et al. 2004,

120

passage 80): RPMI1640 medium supplemented with additional 1 mM sodium

121

pyruvate, 10 mM HEPES and 0.05 mM 2-ß-mercaptoethanol; HMEC-1 human

122

immortalized microvascular endothelial cells (courtesy of Dr. R. Köhler, Department

123

of Internal Medicine - Nephrology, Philpps-University, Marburg, Germany (Ades, et

124

al. 1992)): DMEM High Glucose with L-Glutamine and 1 mM sodium pyruvate. All

125

cells were cultured in a humidified atmosphere containing 5% CO2 at 37°C. Cell

126

number was analyzed by counting the cells in a Neubauer chamber. Thereafter, in

127

our laboratory, the cells were used at passage 8-11. Conditioned media were

128

obtained from stably CUX1 expressing Bon-1 cells cultured for 3 days with

129

approximately 70% confluency. The amphotropic packaging cell line LinX was

130

maintained in Dulbecco’s modified Eagle’s medium, 10%fetal bovine serum, 250

131

µg/ml gentamicin, and 100 µg/ml hygromycin B (Roth, Karlsruhe, Germany). 5-

132

Fluorouracil (5-FU) was purchased from Sigma (Sigma-Aldrich, Taufkirchen,

133

Germany), tumour necrosis factor–related apoptosis-inducing ligand (TRAIL) was

134

obtained from Lilly (Bad Homburg, Germany). Cells synchronization was not

135

performed before functional assays.

136

5

Page 6 of 41

137

Plasmids, Small Interfering RNA and Retroviral Infection

138

CUX1 was transiently suppressed by using two different oligonucleotides (CUX1_1

139

and CUX1_2) as described previously (Michl et al. 2005). All assays were confirmed

140

using two different siRNA sequences to minimize the risk of possible off-target

141

effects. siRNA oligonucleotides were purchased from Ambion (Austin, Texas, USA).

142

Both silencing sequences resulted in a knockdown efficiency >70% (Ripka et al.

143

2010a). Stable Bon-1- and Ins-1 cells were generated using retroviral systems. For

144

retroviral expression CUX1 was amplified and cloned into pENTR-vector using

145

pENTR/D-TOPO Cloning Kit (Invitrogen) and recombined into a Gateway competent

146

pQCXIP (gift by T. Stiewe). To produce retroviruses, we transfected LinX packaging

147

cells with 5 µg of retroviral vectors; 48 and 72 hours after transfection, the retrovirus-

148

containing supernatant was harvested, filtered, and supplemented with 8 µg/ml

149

polybrene (Sigma). Target cells were transduced by spin infection at 1500 rpm, 37°C

150

for 1 hour and selected with puromycin and G418.

151

RNA Isolation, cDNA Synthesis, and Real-Time PCR

152

Murine RIP1Tag2 RNA-Isolation was done as previously described (Fendrich, et al.

153

2011). RNA was extracted using the RNeasy Mini Kit (Qiagen), and first-strand cDNA

154

was synthesised using random hexamer primers and Superscript II reverse

155

transcriptase (Invitrogen). Quantitative RT-PCR analysis was performed with an

156

Applied Biosystems 7500 Fast Real time PCR using the SYBR Green PCR Master

157

Mix kit (Applied Biosystems, Wellesley, Massachusetts, USA) according to the

158

manufacturer’s instructions. Sequence specific primer pairs were designed using the

159

Primer Express software (Applied Biosystems). RPLP0 was used as internal

160

standard (XS-13 forward 5’-GTC GGA GGA GTC GGA CGA G-3’; reverse 5’-GCC 6

Page 7 of 41

161

TTT ATT TCC TTG TTT TGC AAA-3’). Primer sequences for CUX1 are forward 5’-

162

GAC GTG TTG CGC ACT TAA CG-3’ and reverse 5’-ACG ACA TAG ATT GGG CTT

163

AAT GCT-3’. Human Angiogenesis pathway-focused gene expression profiling

164

comprising 84 genes was performed using the RT2 Profiler PCR Array System

165

(Qiagen, Hilden, Germany) according to the manufacturer’s instructions. A list of the

166

genes

167

(http://sabiosciences.com/Manual/1062756).

168

Immunoblotting

169

Cells were lysed in RIPA buffer supplemented with complete protease inhibitor

170

cocktail (Roche, Mannheim, Germany). Nuclear and cytoplasmic fractions were

171

performed using the ProteoJET Cytoplasmic and Nuclear Protein Extraction Kit

172

(Fermentas, St. Leon-Rot, Germany). Immunoblots were performed as described

173

previously (Ripka et al. 2010a).

174

Viability, Proliferation and Apoptosis Assays

175

Bromodeoxyuridine (BrdU) incorporation was measured by using the colorimetric

176

BrdU Cell Proliferation ELISA (Roche Diagnostics, Indianapolis, IN) according to the

177

manufacturer’s instructions. Histone-associated DNA fragments were quantified

178

using Cell Death Detection ELISA PLUS (Roche) according to the manufacturer’s

179

instructions. Cell viability was determined by using the MTT Cell Proliferation Kit

180

(Roche Diagnostics, Mannheim, Germany) according to the manufacturer’s

181

instructions.

182

Cell migration experiments

183

Cell migration experiments were performed as described previously (Michl et al.

184

2005). Briefly, 40,000 cells in 500 µl per inset of a 24-well plate were used and

analysed

in

this

profiler

is

available

online

7

Page 8 of 41

185

incubated for 4 h. Inserts were then rinsed with PBS, dried at room temperature and

186

fixed with methanol. Migration was measured by counting 10 fields of view under a

187

microscope covering about 70–80% of the insert area of the fixed and stained cells

188

(0.2% crystal violet) migrating towards the conditioned media and normalized to the

189

number of migrated cells towards mock conditioned media. Number of migrated cells

190

were also validated via TimeLapseAnalyzer program (Huth et al. 2010).

191

Tube formation assays

192

BD Matrigel™Basement Membrane Matrix Growth Factor Reduced (BD Bioscience)

193

was thawed and plated according to the supplier's protocol. 20,000 cells of the

194

human microvascular endothelial cell line HMEC-1 suspended in the appropriate

195

conditioned media of CUX1 over-expressing Bon-1 cells were plated onto the

196

Matrigel. After incubation for 24 h at 37 °C the wells were photographed at a 4-fold

197

magnification with a Canon 450 EOS camera connected to the microscope (Olympus

198

IMT2). Length and junctions of the generated tubes were analyzed with the

199

TimeLapseAnalyzer program (Huth et al. 2010). Each experiment was performed at

200

least in triplicates with independently obtained conditioned media.

201

Immunohistochemistry, construction of tissue arrays and evaluation

202

In brief, paraffin sections were stained after antigen retrieval (microwave in antigen-

203

unmasking solution, Vector Laboratories, Burlingame, California, USA) with rabbit

204

polyclonal anti-CUX1 (1:200), as described previously (Michl et al. 2005). In addition,

205

CD31 (1:20; Dianova; Hamburg, Germany) and Ki-67 (1:200; Clone SP6; Thermo

206

Fischer Scientific; Fremont; USA) antibodies were used. Antibody binding was

207

visualised using a biotinylated secondary antibody, avidin-conjugated peroxidase

208

(ABC method; Vector Laboratories), 3,39-diaminobenzidine tetrachloride (DAB) as a 8

Page 9 of 41

209

substrate, and H&E as counterstain. The insulinoma tissue samples (FFPE) were

210

obtained from the neuroendocrine tumour archives of the Departments of Pathology

211

in Zürich (Switzerland), Düsseldorf (Germany) and Kiel (Germany) from 1975 – 2006

212

according to the guidelines of the local ethics committee. 1.5-mm tissue cores were

213

taken from representative areas and inserted into four paraffin blocks. The 4 MTAs

214

contained a total of 286 cores and additional human tonsil orientation/control cores.

215

The construction was performed using MTA1 tissue arrayer equipment (Beecher

216

Instruments, Sun Prairie, WI, USA). MTAs were routinely processed for paraffin

217

sectioning.

218

immunohistochemical stains were performed according to routine methods. The

219

analysis was performed by a pathologist with expertise in endocrine and pancreatic

220

tumours (B.S.) in a blinded fashion regarding tumour parameters. The intensity of the

221

reactions was scored as mild, moderate or strong (score 1, 2 or 3, respectively). The

222

proportion of the positive cells in ducts and tumour areas was estimated in percent

223

and divided into scores (80% -4). The final score

224

was determined as a product of the intensity of the staining and the proportion of

225

positive cells (minimum 0, maximum 12) (Remmele.et al. 1986). The scores for cores

226

derived from the same tissue were averaged.

227

Animals and Xenografts

228

Athymic female nude mice (nu/nu mice) were purchased from Charles River

229

(Sulzfeld, Germany). Mice were housed in a climate-controlled SPF facility. All animal

230

experiments were approved by the local government authorities and were performed

231

according to the guidelines of the animal welfare committee. Nude mice were

232

injected subcutaneously with 1 × 106 CUX1 over-expressing or wild type Bon-1 cells

Paraffin

sections

were

deparaffinized

and

rehydrated

and

9

Page 10 of 41

233

(106 cells of each clone) dispersed in 0.1 ml of normal saline at each flank. Six mice

234

per group were injected with cells stably transfected with either CUX1 or empty

235

vector. Tumours were measured during the course of observation (twice a week) and

236

after the animals had been sacrificed.

237

Statistics

238

All data are presented as mean ± standard error of the mean. Two-tailed paired

239

Student's t test was used for statistical evaluation of the data. The one-way and non-

240

parametric ANOVA test was used to calculate the p value for more than two groups.

241

A p value < 0.05 was considered significant.

242

243

RESULTS

244

CUX1 is expressed in human insulinomas and is upregulated during tumour

245

progression in the RIP1Tag2 model

246

Immunohistochemistry revealed that CUX1 is strongly expressed in a series of

247

human insulinomas and shows a particular intense immunoreactivity at the invasive

248

front and in highly proliferative areas (Fig. 1A). To assess CUX1 expression during

249

tumour progression, we employed an established genetically engineered mouse

250

model of insulinoma (RIP1Tag2 model). The RIP1Tag2 model is a tumour

251

progression model, which has been generated to develop invasive neuroendocrine

252

pancreatic islet neoplasms via hyperplastic and angiogenic precursor stages that

253

closely recapitulate the human disease development (Hanahan 1985). In this murine

254

model, CUX1 mRNA and protein was increasingly expressed during tumour

255

progression with the highest levels in invasive insulinomas in comparison to islet

256

hyperplasia and native islets (Fig. 1B and C: statistical analysis of relative CUX1 10

Page 11 of 41

257

mRNA expression: control vs. CUX1; mean 1.0 vs. 1.7; P = 0.011). These results

258

show that CUX1 is highly expressed in murine and human PNENs and suggest that

259

CUX1 may play a role in tumour progression.

260

CUX1 modulates

261

neuroendocrine tumour cell lines

262

In order to study the functional role of CUX1 in neuroendocrine tumour cells, CUX1

263

was genetically ablated in Ins-1 and Bon-1 cells by transient transfection with two

264

different siRNAs. Knockdown of CUX1 resulted in a significant reduction of cell

265

viability as well as proliferation rate in both cell lines (Fig. 2A-C, Suppl Fig 1A-C:

266

statistical analysis for siCUX1 vs siC in Ins-1 for MTT: mean: 65 vs. 100; P = 0.013

267

and BrdU: mean: 81 vs 100; P = 0.023 and Bon-1 for MTT: mean: 74 vs. 100; P =

268

0.027 and BrdU: 68 vs. 100; P = 0.012). Next, we overexpressed CUX1 by stable

269

transfection of Ins-1 and Bon-1 cells using two different clones with adequate empty

270

vector control for each cell line (Fig. 2D and Suppl. Fig. 1D). MTT in vitro assays (Fig.

271

2E and Suppl. Fig. 1E: statistical analysis for Mock vs. CUX1 overexpression after

272

72h in Ins-1: mean: 461 vs. 657; P = 0.002 and Bon-1: mean 609 vs. 741; P = 0.023)

273

and BrdU experiments (Fig. 2F and Suppl. Fig. 1F: Mock vs. CUX1 in Ins-1: mean:

274

100 vs. 123; P = 0.0003 and Bon-1: mean: 100 vs. 138; SD 19; P = 0.009)

275

unequivocally showed robust and statistically significant increase in proliferation in

276

both cell lines compared to the relevant mock control group.

277

Next, we investigated the effect of CUX1 on basal and TRAIL-induced apoptosis in

278

Ins-1 and Bon-1 cells using the same experimental setting. Knockdown of CUX1

279

induced basal apoptosis rates in both cell lines, as determined by cell death detection

280

assay (Fig 3A and Suppl. Fig 2A: statistical analysis siCUX1 vs. siC in Ins-1: mean:

proliferation, basal

and TRAIL-induced apoptosis in

11

Page 12 of 41

281

166 vs 100; P = 0.001 and Bon-1: mean: 133 vs. 100; P = 0.021). Interestingly, Ins-1

282

cells were more sensitive than Bon-1 cells to suppression of CUX1, resulting in a

283

significant increase in apoptosis without any apoptotic stimulus. In addition,

284

knockdown of CUX1 significantly enhanced TRAIL-induced apoptosis, whereas

285

treatment with TRAIL alone had only a minor effect on apoptosis induction, as

286

measured by effector caspase-3/7 activity (Fig. 3B and Suppl. Fig. 2B: statistical

287

analysis for Ins-1: mean: 100 vs. 200 and 124 vs. 227; siC/siCUX1 P = 0.019;

288

siC+TRAIL/siCUX1+TRAIL P = 0.0043 and Bon-1: mean: 100 vs. 298 vs. 110 vs.

289

413; siC/siC+TRAIL P = 0.001; siCUX1/siCUX1+TRAIL P = 0.0031). TRAIL-induced

290

apoptosis was more pronounced in Bon-1 cells, probably due to the fact that these

291

cells were highly proliferative and less differentiated. To confirm the anti-apoptotic

292

effects of CUX1 we additionally performed Western blots for cleaved caspase-3:

293

Knockdown of CUX1 enhanced cleaved caspase-3 levels (Suppl. Fig. 3A).

294

To exclude the possibility that only siRNA-mediated suppression of endogenous

295

CUX1

296

overexpression of CUX1 is able to protect from TRAIL- and 5-FU-induced apoptosis.

297

Upon CUX1 overexpression, basal apoptosis rates were strongly reduced, as

298

determined by cell death detection assays (Fig. 3C and Suppl. Fig. 2C: statistical

299

analysis for Ins-1: Mock/CUX1, Mock/CUX1+5-FU and Mock/CUX1+TRAIL: mean:

300

100 vs. 78; 147 vs. 101; 156 vs 62; P = 0.04; P = 0.011; P = 0.0023 and Bon-1:

301

Mock/CUX1, Mock/CUX1+5-FU and Mock/CUX1+TRAIL: mean: 100 vs. 43; 30 vs.

302

32; 167 vs 129; P = 0.003; P = 0.8; P = 0.024). Furthermore, CUX1 expression

303

protected Ins-1 and Bon-1 cells from TRAIL-induced apoptosis (Fig. 3C, Suppl. Fig.

304

2C and Suppl. Fig. 3B). Interestingly, while CUX1 affected cell viability and protected

305

from TRAIL-induced apoptosis in both cell lines, it affected 5-FU-induced apoptosis

affects

apoptosis

due

to

off-target

effects,

we

assessed

whether

12

Page 13 of 41

306

only in Ins-1 cells (Fig. 3C and Suppl Fig. 2C). Taken together, our experiments

307

indicate that CUX1 mediates pro-proliferative and anti-apoptotic effects in

308

neuroendocrine tumour cell lines.

309

CUX1 increases tube formation in HMEC-1 cells via tumour cell-derived

310

secreted pro-angiogenic mediators

311

One of the most important features of neuroendocrine neoplasms is the high

312

vascularization which is established and maintained by a complex molecular and

313

cellular crosstalk between neoplastic cells and endothelial cells. To assess potential

314

angiogenic effects of CUX1, we employed an established tube formation assay.

315

Conditioned medium from CUX1 overexpressing neuroendocrine tumour cells was

316

used to stimulate immortalized human microvascular endothelial cells (HMEC-1) in

317

order to assess the ability of endothelial cells to generate vessels. After 24h

318

stimulation with Ins-1 and Bon-1 cell supernatants, HMEC-1 cells significantly

319

induced tube formation in a CUX1 dependent manner (Fig. 4A-C and Suppl. Fig. 4A

320

and B: statistical analysis of Ins-1 Mock CM and CUX1 CM: Length: mean: 100 vs.

321

116; P = 0.044 and junctions: mean: 100 vs. 128; P = 0.022 and Bon-1: Length:

322

mean: 100 vs. 115; P = 0.012 and junctions: mean: 100 vs. 132; P = 0.006). Analysis

323

of the assays with the TimeLapseAnalyzer program from at least 5 independent

324

experiments demonstrated a remarkable increase in tube length and junctions (Fig.

325

4A-C and in Suppl. Fig. 4A and B). In contrast to tube length and formation of

326

junctions, HMEC-1 migration was not altered significantly (Suppl. Fig. 5: statistical

327

analysis of Ins-1-CM: mean: 100 vs. 118; P = 0.08 and Bon-1-CM: mean 100 vs. 120;

328

P = 0.06)

13

Page 14 of 41

329

To elucidate the transcriptional alterations underlying this CUX1-dependent pro-

330

angiogenic phenotype, we used a human angiogenesis pathway-focused gene

331

expression profiler comprising 84 genes to examine potential CUX1-regulated targets

332

on mRNA level in CUX1 overexpressing Bon-1 cells (Suppl. Table 1). Among others,

333

the

334

metalloproteases MMP-9 were consistently induced by CUX1. Both genes could be

335

confirmed as modulated by CUX1 on mRNA and protein level (Figure 5A-B:

336

statistical analysis for mRNA expression relative to GAPDH in Bon-1 Mock and

337

CUX1 overexpressing cells: HIF-1α: mean: 0.09 vs. 0.26; P = 0.1; MMP-9: 1.0 vs.

338

1.9; P = 0.05). These data suggest that CUX1 expression stimulates angiogenesis in

339

endothelial cells in vitro, with a number of differentially regulated pro-angiogenic

340

effector molecules pointing towards an important mechanistic cross-talk between the

341

secretome of neuroendocrine tumour cells and endothelial cells during angiogenesis.

342

CUX1 expression correlates with tumour mediating hallmarks in murine and

343

human insulinomas

344

To validate our in vitro findings on CUX1 as regulator of proliferation and

345

angiogenesis, xenograft experiments were performed in nude mice using CUX1

346

transfected Bon-1 cells. Expression of CUX1 was confirmed by Western blot analysis

347

in n=4 mice each at endpoint (Figure 6B). CUX1 tumours showed a marked trend

348

towards increased tumour volumes 7 weeks after implantation which, however, did

349

not reach statistical significance due to the limited number of animals (Fig. 6A:

350

statistical analysis: mean: 134 vs. 226; P = 0.26).

351

In line with our in vitro data, CUX1 expression in vivo caused a significant increase in

352

tumour cell proliferation (Mock vs. CUX1: mean: 14.6 vs. 37.5; P = 0.0045), as

hypoxia-induced

transcription

factor

HIF-1α

as

well

as

the

matrix

14

Page 15 of 41

353

assessed by ki-67 immunohistochemistry, but failed to increase mean vessel density

354

(Mock vs. CUX1: mean: 26.4 vs. 38.6; P = 0.074), as determined by CD31

355

immunohistochemistry (Figure 6C and E). Despite the fact that CUX1 expressing

356

Bon-1 cells resulted in larger, and more proliferative xenograft tumours, systematic

357

histological analysis of H&E sections showed a significant reduction of necrotic areas

358

in CUX1-expressing xenografts (Figure 6D: statistical analysis Mock vs. CUX1:

359

mean: 49.7 vs. 24.8; P = 0.027).

360

As CUX1 is involved in the tumour progression in vitro and in vivo, we sought to

361

elucidate the clinical relevance of CUX1 in tumour progression and metastasis by

362

studying CUX1 expression in human insulinoma samples. To this end, we used

363

MTAs containing 59 different human insulinomas tissues for immunohistochemical

364

analysis. In the entire cohort, 39 female and 20 male patients were included (Suppl.

365

Tab 2). 52 primary tumours, 3 lymph nodes and 4 liver metastases were

366

immunohistochemically investigated. Overall, 45 tumours exhibited a benign

367

behaviour in comparison to 14 malignant tumours. CUX1 showed a distinct nuclear

368

expression in islets, as well as benign and malignant insulinomas. Out of 59

369

insulinomas 58 were CUX1 positive. Importantly, malignant tumours revealed

370

significantly increased immunoreactivity for CUX1 compared to those with benign

371

behaviour (immunoreactivity score IRS 8.8 vs 5.8; P = 0.002) (Fig. 7 A-C).

372

Metastases scored higher for CUX1 positivity than primary tumours (P = 0.018),

373

independently of their localization (lymph node or liver). In addition, a significant

374

correlation between CUX1 expression and G1/G2 grading was observed (P = 0.038)

375

pointing out a connection between CUX1 and proliferation, since tumour grading is

376

based on the Ki-67 proliferation index. Among the 14 patients with malignant

377

insulinoma five disease-related deaths occurred, none in the benign group (Suppl. 15

Page 16 of 41

378

Tab 2). The overall survival of 36.0 months in patients with malignant insulinomas

379

illustrated the poor prognosis of this tumour entity.

380

In summary, our in vivo data show that CUX1 mediates tumour progression and

381

angiogenesis in murine neuroendocrine tumours and is associated with malignant

382

behaviour in human insulinomas.

383

384

DISCUSSION

385

PNENs represent the second-most common malignant disease of the pancreas with

386

the majority of patients presenting at a metastasized and unresectable stage. The

387

mTOR inhibitor everolimus, and the multi-target anti-angiogenic agent sunitinib have

388

recently been introduced for the treatment of progressive, unresectable, locally

389

advanced or metastatic PNENs but still fail to prolong overall survival (Raymond et

390

al. 2011; Yao et al. 2011). Therefore, novel therapeutic approaches and predictive

391

biomarkers are needed to classify patients into subgroups to better predict treatment

392

success and guide by molecular rationales. Furthermore, molecular and biochemical

393

characterization of novel targets in PNENs may aid to devise new treatment

394

approaches and treatment combinations that can be deployed upon primary and

395

secondary chemoresistance.

396

CUX1 has been identified as a transcriptional activator as well as a transcriptional

397

repressor and its activity has been associated with tumour progression and

398

shortened survival in several tumour entities (Michl et al. 2005; Ripka et al. 2010a).

399

Based on several studies on PDAC, CUX1 co-opted Src and other downstream

400

signaling molecules such as WNT5A to induce an invasive phenotype in pancreatic

401

cancer (Aleksic et al. 2007; Griesmann et al. 2013; Ripka et al. 2007; Ripka et al. 16

Page 17 of 41

402

2010b). Moreover, CUX1 was described as transcriptional target of TGFB mediating

403

cell motility via influencing migration and invasion in diverse in vitro and in vivo

404

systems (Michl et al. 2005). Previous results indicated that CUX1 is also a major

405

mediator of PI3K/Akt-induced tumour cell survival (Ripka et al. 2010a).

406

Although PDAC and PNENs are distinctly different in terms of molecular evolution,

407

clinical behaviour, prognosis and treatment options, there seems to be a

408

considerable molecular overlap in terms of dysregulated pathways in PDAC and

409

PNENs. For instance, the PI3K/Akt/mTOR pathway, the TGFB pathway and Src

410

family kinase activity have been implicated in the pathogenesis of PNENs (Capdevila

411

and Tabernero 2011; Di Florio, et al. 2011; Di Florio, et al. 2007; Ghayouri et al.

412

2010; Missiaglia et al. 2010). Therefore, we aimed to investigate the expression and

413

molecular function of the oncogenic transcription factor CUX1 in neuroendocrine

414

tumours.

415

We show that CUX1 is a mediator of tumour progression in two neuroendocrine

416

tumour cell lines and xenograft tumours by promoting proliferation, tumour growth

417

and resistance to apoptosis. Importantly, 5-FU-containing regimens represent a

418

current standard of care treatment in combination with streptozocin (Falconi et al.

419

2012; Moertel et al. 1992), and in fact CUX1 protected 5-FU induced apoptosis in the

420

insulinoma cell-line thus indicates the potential therapeutic relevance of our findings.

421

Interestingly, in xenografts experiments CUX1 overexpression reduced necrotic

422

areas which might be attributable to CUX1-induced improved perfusion and

423

angiogenesis in this experimental setting.

424

Furthermore, we identify CUX1 as novel modulator of tumour angiogenesis in vitro

425

and in vivo by paracrine stimulation of endothelial cells via a CUX1 dependent, pro-

426

angiogenic secretome in neuroendocrine tumour cells. Surprisingly, the classic 17

Page 18 of 41

427

mediators of angiogenesis, VEGF, FGF, and PDGF revealed no difference upon

428

CUX1 expression. However, we found that MMP-9 and HIF-1α were both

429

upregulated in CUX1 overexpressing cells, and both proteins have been implicated in

430

tumour progression and angiogenesis (Couvelard et al. 2008). For instance, MMP-9

431

has been identified as critical modulator of the angiogenic switch in the RIP1Tag2

432

model (Bergers et al. 2000). Interestingly, MMP-9 does not only promote

433

angiogenesis, it also interacts with VEGF facilitating its expression and release

434

(Bergers et al. 2000).

435

Finally, to evaluate the clinical significance of CUX1 in human PNENs, we performed

436

CUX1 immunohistochemistry on a large cohort of human insulinomas. Our data

437

indicate that CUX1 is strongly expressed in the majority of insulinoma tissues.

438

Moreover, malignant invasive insulinomas expressed more CUX1 compared to those

439

with benign behaviour and metastatic tumours were slightly more positive for CUX1

440

than primary tumours. In analogy to our findings CUX1 was reported to be

441

overexpressed in breast and pancreatic PDAC, and inversely correlated with the

442

patient outcome in breast cancer (Michl et al. 2005; Ripka et al. 2010a). Data on the

443

clinical follow-up of our patients confirmed that surgical cure is feasible in most of the

444

benign insulinoma cases. However, malignant insulinomas presented with an

445

aggressive behaviour associated a shortened overall survival of approximately 35

446

months. Based on our data with insulinomas, we are currently planning to investigate

447

and correlate the CUX1 expression with clinicopathological features and outcome in

448

patients with human non-functional pancreatic neoplasms.

449

In summary, our data show for the first time an involvement of the transcription factor

450

CUX1 in neuroendocrine neoplasms. CUX1 promotes carcinogenesis via regulation

451

of growth, apoptosis and angiogenesis by multiple mechanisms. Mechanistically, we 18

Page 19 of 41

452

find that CUX1 promotes tumour angiogenesis via paracrine stimulation of

453

endothelial cells. Tumour angiogenesis represents a hallmark feature of PNENs and

454

anti-angiogenic drugs are currently the most promising candidates of PNENs.

455

Therefore, CUX1-dependent pathways may constitute attractive targets for future

456

therapeutic strategies in PNENs.

457

458

DECLARATION OF INTEREST

459

The authors declare that there is no conflict of interest that could be perceived as

460

prejudicing the impartiality of this scientific work.

461

462

FUNDING

463

This work was supported in part by grants of the Deutsche Forschungsgemeinschaft

464

(DFG) (to PM), the Behring-Roentgen Foundation (to PM), the Deutsche Krebshilfe

465

(to PM) and the European Commisson FP7 grant (Collaborative Project ‘EPC-TM-

466

Net, to PM and TG). This publication reflects only the authors' views. The European

467

Community is not liable for any use that may be made of the information herein.

468

469

ACKNOWLEDGEMENTS

470

We are grateful to Eleni Archontidou-Aprin for excellent technical assistance.

471 472

FIGURE LEGENDS

473

Figure 1. CUX1 expression of human und murine insulinoma tissue. (A)

474

Representative immunohistochemical stainings of CUX1 in human invasive 19

Page 20 of 41

475

insulinomas demonstrating a prominent expression of the target (upper row 10-fold

476

and bottom row 40-fold magnifications; n = 10). Arrows marked the invasive front and

477

highly proliferative areas (B). Increasing CUX1 expression in correlation to the islet

478

tumour stage ranging from normal islets, hyperplastic islets to invasive insulinomas,

479

depicted into different magnifications (upper row 10-fold and bottom row 20-fold; n =

480

10). (C) CUX1 mRNA levels as measured via RT-PCR of five 10-week old RIP1Tag2

481

mice (islet hyperplasia stage) compared to five 13-week old mice (invasive

482

insulinoma stage). Data were normalized to RPLP0 and are representative of at least

483

five independent experiments and presented as relative expression levels ± SD

484

(mean: 1.0 vs. 1.74; n = 5, SD = 0.2; P = 0.011). *P < 0.05; IHC,

485

immunohistochemistry; W, week.

486

Figure 2. CUX1 navigates the proliferation in insulinoma cells. (A) Ins-1 cells

487

were transiently transfected with CUX1 siRNA (siCUX1) or control siRNA (siC).

488

Protein isolation and Western Blot were performed demonstrating the CUX1

489

knockdown. (B) Relative cell viability was measured via MTT after 24h (n = 4; mean:

490

65 vs. 100; SD 7.5; P = 0.013). (C) Proliferation after suppression of CUX1 in Ins-1

491

cells after 24h was assessed by BrdU assay (n = 4; mean: 81 vs 100; SD 14.9; P =

492

0.023). (D) Confirmation of stable overexpression of CUX1 in Ins-1 cells by Western

493

Blot analysis. (E) The relative cell viability of CUX1 overexpressing (CUX1) and

494

empty vector control-transfected cells (Mock) was detected after 24h (n = 6; mean:

495

100 vs. 123; SD 4) and 72h (n = 6; mean: 461 vs. 657; SD 13; P = 0.002) via MTT

496

assay. (F) Similarly, the relative proliferation of CUX1-overexpressing cells in

497

comparison to mock-transfected cells as quantified by BrdU assay (n = 5; mean: 100

498

vs. 123; SD 16; P = 0.0003). Averages of the two stable clones each were evaluated 20

Page 21 of 41

499

and presented. Data are representative for at least three independent experiments

500

and are shown as mean ± SD. *P < 0.05 as compared with control cells.

501

Figure 3. CUX1 overcomes basal and drug-induced apoptosis in insulinoma

502

cells. (A) Knockdown of CUX1 in Ins-1 cells via siRNA (siCUX1) in comparison to

503

empty vector (siC) (n = 3; mean: 166 vs 100; SD 10; P = 0.001). Apoptosis was

504

determined by using a specific cell death detection ELISA quantifying the histone-

505

bound DNA fragmentation. (B) Apoptosis after knockdown of CUX1 by siRNA and

506

apoptosis induced by TRAIL (100ng) after 24h, as assessed by a caspase-3/7

507

activation assay (n = 4; mean: 100 vs. 200 and 124 vs. 227; SD 26, 35, 24; siC/

508

siCUX1 P = 0.019; siC+TRAIL/siCUX1+TRAIL P = 0.0043). (C) Apoptosis in Ins-1

509

cells stably overexpressing CUX1 and mock-transfected control cells +/- TRAIL and

510

+/- 5-FU (0.2mM) for 24h as assessed by quantification of histone-bound DNA

511

fragmentation (n = 3; mean: 100 vs. 78; 147 vs. 101; 156 vs 62; SD 1 vs. 3.7; 4.6 vs.

512

0.6; 5.7 vs. 2.3; Mock/CUX1 P = 0.04; Mock/CUX1+5-FU P = 0.011;

513

Mock/CUX1+TRAIL P = 0.0023). Data are representative for at least three

514

independent experiments and are shown as mean ± SD. *P < 0.05 as compared with

515

control cells.

516 517

Figure 4. CUX1 augments angiogenesis in tube formation assays. (A)

518

Representative pictures of HMEC-1 cells incubated with mock or CUX1 conditioned

519

medium of Bon-1 cells. Analysis of tube formation assay by determining length (red

520

lines) and junctions (yellow dots) of tubes evaluated with TimeLapseAnalyser

521

program (B-C). Calculation of length (mean: 100 vs. 116; SD 14.6; P = 0.044) and

522

junctions (mean: 100 vs. 128; SD 18.9; P = 0.022) of pictures were done from at least

523

9 independent experiments. Photographs were taken with 4-fold magnification. Tube 21

Page 22 of 41

524

formation was performed with conditioned medium from CUX1 stably transfected Ins-

525

1 cells compared to mock-transfected cells. *P < 0.05 as compared with control cells.

526

Figure

527

neuroendocrine cells. (A) qRT-PCR analysis for HIF-1α and MMP-9 mRNA in

528

CUX1 overexpressing and mock transfected Bon-1 cells. HIF-1α: n = 3; mean: 0.09

529

vs. 0.26; SD 0.009 vs. 0.11; P = 0.1; MMP-9: n = 3; mean: 1.0 vs. 1.9; SD 0.23 vs.

530

1.2 P = 0.05. (B) Confirmation of the CUX1-dependent regulation of HIF-1α and

531

MMP-9 on protein level by Western Blot analysis. Actin served as a control for equal

532

protein loading.

533

Figure 6. CUX1 influences tumour growth via acceleration of proliferation and

534

neovascularization in vivo. (A) Nude mice were injected subcutaneously with 1 ×

535

106 CUX1 over-expressing or mock-transfected Bon-1 cells (106 cells of each clone).

536

Six mice per group were injected. Tumours were measured ex vivo after the animals

537

had been sacrificed (mean: 134 vs. 226; SD 42 vs. 65; P = 0.26). (B) Representative

538

expression of CUX1 in protein lysates of CUX1 overexpressing or mock transfected

539

Bon-1 xenografts, for which material for protein lysates was available. (C-E)

540

Immunohistochemistry of Ki-67 and CD31 as well as H&E evaluation in tumour

541

tissues +/- CUX1 overexpression. (C) The proliferation index of the xenograft

542

tumours was estimated using the antibody against ki-67 (n = 8; mean: 14.6 vs. 37.5;

543

SD 3.1 vs. 4.8; P = 0.0045). (D) Necrosis areas were quantified and evaluated

544

comparing CUX1- and mock-transfected xenografts (n = 8; mean: 49.7 vs. 24.8; SD

545

12.2 vs. 3.1; P = 0.027). (E) Proportion of endothelial cells per visual field in sections

546

of CUX- or mock-tumours as measured using the CD31 antibody which detects

5.

CUX1

modulates

expression

of

angiogenic

markers

in

22

Page 23 of 41

547

endothelial cells (n = 8; mean: 26.4 vs. 38.6; SD = 2.4 vs. 4.6; P = 0.074). Data are

548

shown as mean ± SD. *P < 0.05 as compared with mock plasmid.

549

Figure 7. CUX1 expression is increased in malignant insulinomas. (A) CUX1

550

expression was evaluated using the Remmele and Stegner immunoreactive score as

551

described in methods. (B) Comparison of representative CUX1 IHC in benign and

552

malignant insulinomas in 10-fold magnification. (C) CUX1 score assessed in human

553

benign and malignant insulinomas (mean: 5.8 vs. 8.8; SEM: 0.47 vs. 0.98; P =

554

0.002). Data are shown as mean ± SD (A) and as mean ± SEM (C). *P < 0.05 as

555

compared to benign insulinomas.

556

557

SUPPLEMENTARY FIGURE LEGENDS

558

Suppl. Figure 1. CUX1 navigates the proliferation in Bon-1 cells. (A-B) Bon-1

559

cells were transiently transfected with CUX1 siRNA (siCUX1) or control siRNA (siC).

560

After evaluation of the CUX1 knockdown by Western Blot relative cell viability was

561

measured via MTT after 24h (n = 4; mean: 74 vs. 100; SD 13.9; P = 0.027). (C)

562

Proliferation after suppression of CUX1 in Bon-1 cells after 24h was assessed by

563

BrdU assay (n = 4; mean: 68 vs. 100; SD 23; P = 0.012). (D) CUX1 overexpression

564

was detected in Western Blot analysis. This blot was exposed at a lower exposure

565

time compared to suppl. Figure 1A. (E-F) The relative cell viability of CUX1

566

overexpression (CUX1) and empty vector control (mock) as well as the relative

567

proliferation was assayed with MTT (n = 6; mean: 100 vs. 124 and 609 vs. 741; SD

568

19.6 and 78.2; 24h P = 0.13; 72h P = 0.023) and BrdU assay (n = 4; mean: 100 vs.

23

Page 24 of 41

569

138; SD 19; P = 0.009). Averages of the two stable clones were evaluated and

570

presented. *P < 0.05 as compared with control cells.

571

Suppl. Figure 2. CUX1 overcomes basal and drug-induced apoptosis in Bon-1

572

cells. (A) CUX1 depletion in Bon-1 cells via transient siRNA (siCUX1) compared with

573

siControl (siC). Apoptosis was determined by histone-bound DNA fragmentation (n =

574

3; mean: 133 vs. 100; SD 17; P = 0.021). (B) Apoptosis after knockdown of CUX1 by

575

siRNA and apoptosis induced by 100ng TRAIL after 24h, as assessed by a caspase-

576

3/7 activation assay (n = 4; mean: 100 vs. 298 vs. 110 vs. 413; SD 34, 17, 33; siC/

577

siC+TRAIL P = 0.001; siCUX1/siCUX1+TRAIL P = 0.0031). Bon-1 cells stably

578

overexpressing CUX1 and mock-transfected control cells were incubated with 100ng

579

TRAIL and 0,2mM 5-FU for 24h and apoptosis by quantification of histone-bound

580

DNA fragmentation was assessed (C) (n = 3; mean: 100 vs. 43; 30 vs. 32; 167 vs

581

129; SD 1 vs. 8.7; 2.3 vs. 15.3; 28 vs. 4.8; Mock/CUX1 P = 0.003; Mock/CUX1+5-FU

582

P = 0.8; Mock/CUX1+TRAIL P = 0.024). Data are representative for at least three

583

independent experiments and are shown as mean ± SD. *P < 0.05 as compared with

584

control cells.

585

Suppl. Figure 3. CUX1 protects from apoptosis. Knockdown of CUX1 facilitates

586

apoptosis measured in native and TRAIL-treated conditions as detected by Western

587

blot for cleaved caspase-3 (A). CUX1 overexpression rescues from basal and TRAIL-

588

induced apoptosis in two different clones of Mock1/2 and CUX1-1/2 clones in Bon-1

589

cells (B). Actin served as a control for equal protein loading.

590

Suppl. Figure 4. CUX1 augments angiogenesis in tube formation assays. Tube

591

formation subdivided in length (A) (n = 20; mean: 100 vs. 115; SD 9.7; P = 0.012)

592

and junctions (B) (n = 20; mean: 100 vs. 132; SD 14.7; P = 0.006) of tubes were 24

Page 25 of 41

593

evaluated with TimeLapseAnalyser program. Tube formation was performed with

594

conditioned medium from CUX1 stably transfected Bon-1 cells compared to mock

595

transfected cells. Data are representative for at least three independent experiments

596

and are shown as mean ± SD. *P < 0.05 as compared with control cells.

597

Suppl. Figure 5. CUX1 has no significant impact on HMEC-1 migration. Boyden

598

chamber assays were performed with Ins-1 (n = 3; mean: 100 vs. 118; SD 4; P =

599

0.08) and Bon-1 (n = 4; mean 100 vs. 120; SD 7; P = 0.06) conditioned medium (CM)

600

of CUX1 overexpressing and control (mock) clones. Data are representative for at

601

least three independent experiments and are shown as mean ± SD.

602

Suppl. Figure 6-8. Adapted western blots according to the ERC guidelines. Full-

603

size western blot images of knockdown as well as of overexpression experiments

604

were described (Suppl. Fig. 6). Representative western blots of xenograft protein

605

lysates revealing CUX1 and Actin expression (Suppl. Fig. 7). Full-size western blot

606

images of CUX1 target genes on protein level with Actin as internal control (Suppl.

607

Fig. 8).

608

Suppl. Table 1. Raw data of the pathway-focused gene expression profiling

609

comprising 84 genes using the RT2 Profiler PCR Array System in CUX1-

610

overexpressing (CUX) or Mock-transfected (Mock) Bon-1 cells. Data are shown as

611

mean +/-SD, sorted for the highest ratio CUX/Mock.

612

Suppl. Table 2. Clinicopathological features of 59 patients with insulinoma were

613

subdivided into gender, age, tumour size, grade, localization, biological behaviour,

614

CUX1 score, recurrence-free survival, overall survival as well as disease-related

615

death. 25

Page 26 of 41

616

REFERENCES

617

Ades EW, Candal FJ, Swerlick RA, George VG, Summers S, Bosse DC & Lawley TJ 1992

618

HMEC-1: establishment of an immortalized human microvascular endothelial cell line. J

619

Invest Dermatol 99 683-690.

620

Aleksic T, Bechtel M, Krndija D, von Wichert G, Knobel B, Giehl K, Gress TM & Michl P 2007

621

CUTL1 promotes tumor cell migration by decreasing proteasome-mediated Src

622

degradation. Oncogene 26 5939-5949.

623

Bergers G, Brekken R, McMahon G, Vu TH, Itoh T, Tamaki K, Tanzawa K, Thorpe P, Itohara

624

S, Werb Z, et al. 2000 Matrix metalloproteinase-9 triggers the angiogenic switch during

625

carcinogenesis. Nat Cell Biol 2 737-744.

626 627

Capdevila J & Tabernero J 2011 A shining light in the darkness for the treatment of pancreatic neuroendocrine tumors. Cancer Discov 1 213-221.

628

Capurso G, Fazio N, Festa S, Panzuto F, De Braud F & Delle Fave G 2009 Molecular target

629

therapy for gastroenteropancreatic endocrine tumours: biological rationale and clinical

630

perspectives. Crit Rev Oncol Hematol 72 110-124.

631

Capurso G, Fendrich V, Rinzivillo M, Panzuto F, Bartsch DK & Delle Fave G 2012 Novel

632

molecular targets for the treatment of gastroenteropancreatic endocrine tumors:

633

answers and unsolved problems. Int J Mol Sci 14 30-45.

634

Couvelard A, Deschamps L, Rebours V, Sauvanet A, Gatter K, Pezzella F, Ruszniewski P &

635

Bedossa P 2008 Overexpression of the oxygen sensors PHD-1, PHD-2, PHD-3, and

636

FIH Is associated with tumor aggressiveness in pancreatic endocrine tumors. Clin

637

Cancer Res 14 6634-6639.

638

Di Florio A, Adesso L, Pedrotti S, Capurso G, Pilozzi E, Corbo V, Scarpa A, Geremia R,

639

Delle Fave G & Sette C 2011 Src kinase activity coordinates cell adhesion and

640

spreading with activation of mammalian target of rapamycin in pancreatic endocrine

641

tumour cells. Endocr Relat Cancer 18 541-554.

26

Page 27 of 41

642

Di Florio A, Capurso G, Milione M, Panzuto F, Geremia R, Delle Fave G & Sette C 2007 Src

643

family kinase activity regulates adhesion, spreading and migration of pancreatic

644

endocrine tumour cells. Endocr Relat Cancer 14 111-124.

645

Falconi M, Bartsch DK, Eriksson B, Klöppel G, Lopes JM, O'Connor JM, Salazar R, Taal BG,

646

Vullierme MP, O'Toole D, et al. 2012 ENETS Consensus Guidelines for the

647

management of patients with digestive neuroendocrine neoplasms of the digestive

648

system: well-differentiated pancreatic non-functioning tumors. Neuroendocrinology 95

649

120-134.

650

Fendrich V, Wiese D, Waldmann J, Lauth M, Heverhagen AE, Rehm J & Bartsch DK 2011

651

Hedgehog inhibition with the orally bioavailable Smo antagonist LDE225 represses

652

tumor growth and prolongs survival in a transgenic mouse model of islet cell

653

neoplasms. Ann Surg 254 818-823; discussion 823.

654

Ghayouri M, Boulware D, Nasir A, Strosberg J, Kvols L & Coppola D 2010 Activation of the

655

serine/theronine protein kinase Akt in enteropancreatic neuroendocrine tumors.

656

Anticancer Res 30 5063-5067.

657

Griesmann H, Ripka S, Pralle M, Ellenrieder V, Baumgart S, Buchholz M, Pilarsky C, Aust D,

658

Gress TM & Michl P 2013 WNT5A-NFAT signaling mediates resistance to apoptosis in

659

pancreatic cancer. Neoplasia 15 11-22.

660 661

Grothey A & Galanis E 2009 Targeting angiogenesis: progress with anti-VEGF treatment with large molecules. Nat Rev Clin Oncol 6 507-518.

662

Hanahan D 1985 Heritable formation of pancreatic beta-cell tumours in transgenic mice

663

expressing recombinant insulin/simian virus 40 oncogenes. Nature 315 115-122.

664

Hulea L & Nepveu A 2012 CUX1 transcription factors: from biochemical activities and cell-

665

based assays to mouse models and human diseases. Gene 497 18-26.

666

Huth J, Buchholz M, Kraus JM, Schmucker M, von Wichert G, Krndija D, Seufferlein T, Gress

667

TM & Kestler HA 2010 Significantly improved precision of cell migration analysis in

27

Page 28 of 41

668

time-lapse video microscopy through use of a fully automated tracking system. BMC

669

Cell Biol 11 24.

670

Jiao Y, Shi C, Edil BH, de Wilde RF, Klimstra DS, Maitra A, Schulick RD, Tang LH, Wolfgang

671

CL, Choti MA, et al. 2011 DAXX/ATRX, MEN1, and mTOR pathway genes are

672

frequently altered in pancreatic neuroendocrine tumors. Science 331 1199-1203.

673

Larghi A, Capurso G, Carnuccio A, Ricci R, Alfieri S, Galasso D, Lugli F, Bianchi A, Panzuto

674

F, De Marinis L, et al. 2012 Ki-67 grading of nonfunctioning pancreatic neuroendocrine

675

tumors on histologic samples obtained by EUS-guided fine-needle tissue acquisition: a

676

prospective study. Gastrointest Endosc 76 570-577.

677

Lawrence B, Gustafsson BI, Chan A, Svejda B, Kidd M & Modlin IM 2011 The epidemiology

678

of gastroenteropancreatic neuroendocrine tumors. Endocrinol Metab Clin North Am 40

679

1-18, vii.

680

Michl P, Ramjaun AR, Pardo OE, Warne PH, Wagner M, Poulsom R, D'Arrigo C, Ryder K,

681

Menke A, Gress T, et al. 2005 CUTL1 is a target of TGF(beta) signaling that enhances

682

cancer cell motility and invasiveness. Cancer Cell 7 521-532.

683

Missiaglia E, Dalai I, Barbi S, Beghelli S, Falconi M, della Peruta M, Piemonti L, Capurso G,

684

Di Florio A, delle Fave G, et al. 2010 Pancreatic endocrine tumors: expression profiling

685

evidences a role for AKT-mTOR pathway. J Clin Oncol 28 245-255.

686

Modlin IM, Oberg K, Chung DC, Jensen RT, de Herder WW, Thakker RV, Caplin M, Delle

687

Fave G, Kaltsas GA, Krenning EP, et al. 2008 Gastroenteropancreatic neuroendocrine

688

tumours. Lancet Oncol 9 61-72.

689

Moertel CG, Lefkopoulo M, Lipsitz S, Hahn RG & Klaassen D 1992 Streptozocin-doxorubicin,

690

streptozocin-fluorouracil or chlorozotocin in the treatment of advanced islet-cell

691

carcinoma. N Engl J Med 326 519-523.

692

Olson P, Chu GC, Perry SR, Nolan-Stevaux O & Hanahan D 2011 Imaging guided trials of

693

the angiogenesis inhibitor sunitinib in mouse models predict efficacy in pancreatic

694

neuroendocrine but not ductal carcinoma. Proc Natl Acad Sci U S A 108 E1275-1284. 28

Page 29 of 41

695

Panzuto F, Boninsegna L, Fazio N, Campana D, Pia Brizzi M, Capurso G, Scarpa A, De

696

Braud F, Dogliotti L, Tomassetti P, et al. 2011 Metastatic and locally advanced

697

pancreatic endocrine carcinomas: analysis of factors associated with disease

698

progression. J Clin Oncol 29 2372-2377.

699

Panzuto F, Campana D, Fazio N, Brizzi MP, Boninsegna L, Nori F, Di Meglio G, Capurso G,

700

Scarpa A, Dogliotti L, et al. 2012 Risk factors for disease progression in advanced

701

jejunoileal neuroendocrine tumors. Neuroendocrinology 96 32-40.

702

Pape UF, Berndt U, Müller-Nordhorn J, Böhmig M, Roll S, Koch M, Willich SN &

703

Wiedenmann

B

2008

Prognostic

factors

of

long-term

outcome

in

704

gastroenteropancreatic neuroendocrine tumours. Endocr Relat Cancer 15 1083-1097.

705

Parekh D, Ishizuka J, Townsend CM, Haber B, Beauchamp RD, Karp G, Kim SW,

706

Rajaraman S, Greeley G & Thompson JC 1994 Characterization of a human pancreatic

707

carcinoid in vitro: morphology, amine and peptide storage, and secretion. Pancreas 9

708

83-90.

709

Raymond E, Dahan L, Raoul JL, Bang YJ, Borbath I, Lombard-Bohas C, Valle J, Metrakos P,

710

Smith D, Vinik A, et al. 2011 Sunitinib malate for the treatment of pancreatic

711

neuroendocrine tumors. N Engl J Med 364 501-513.

712

Remmele W, Hildebrand U, Hienz HA, Klein PJ, Vierbuchen M, Behnken LJ, Heicke B &

713

Scheidt E 1986 Comparative histological, histochemical, immunohistochemical and

714

biochemical studies on oestrogen receptors, lectin receptors, and Barr bodies in human

715

breast cancer. Virchows Arch A Pathol Anat Histopathol 409 127-147.

716 717

Rindi G & Wiedenmann B 2012 Neuroendocrine neoplasms of the gut and pancreas: new insights. Nat Rev Endocrinol 8 54-64.

718

Ripka S, König A, Buchholz M, Wagner M, Sipos B, Klöppel G, Downward J, Gress T & Michl

719

P 2007 WNT5A--target of CUTL1 and potent modulator of tumor cell migration and

720

invasion in pancreatic cancer. Carcinogenesis 28 1178-1187.

29

Page 30 of 41

721

Ripka S, Neesse A, Riedel J, Bug E, Aigner A, Poulsom R, Fulda S, Neoptolemos J,

722

Greenhalf W, Barth P, et al. 2010a CUX1: target of Akt signalling and mediator of

723

resistance to apoptosis in pancreatic cancer. Gut 59 1101-1110.

724

Ripka S, Riedel J, Neesse A, Griesmann H, Buchholz M, Ellenrieder V, Moeller F, Barth P,

725

Gress TM & Michl P 2010b Glutamate receptor GRIA3--target of CUX1 and mediator of

726

tumor progression in pancreatic cancer. Neoplasia 12 659-667.

727

Yao JC, Hassan M, Phan A, Dagohoy C, Leary C, Mares JE, Abdalla EK, Fleming JB,

728

Vauthey JN, Rashid A, et al. 2008 One hundred years after "carcinoid": epidemiology

729

of and prognostic factors for neuroendocrine tumors in 35,825 cases in the United

730

States. J Clin Oncol 26 3063-3072.

731

Yao JC, Shah MH, Ito T, Bohas CL, Wolin EM, Van Cutsem E, Hobday TJ, Okusaka T,

732

Capdevila J, de Vries EG, et al. 2011 Everolimus for advanced pancreatic

733

neuroendocrine tumors. N Engl J Med 364 514-523.

734

30

Page 311 of 41 Fig.

CUX1 IHC

B

CUX1 IHC from the RIP1Tag2 murine insulinoma model

islet

human invasive insulinomas

C

rel CUX1 mRNA expression

A

islet hyperplasia

2,5

*

2 1,5 1 0,5

0 10 W

13 W

invasive insulinoma

Ins-1

C

B 120

*

60 40

20

rel proliferation (%)

rel cell viability (%)

80

CUX1

Actin

Actin

E

siCUX1Kontrolle

F

700

*

*

600

80

60 40 20

500 400 300 200

100 80

60 40 20

0 2

120

*

100

1

160

140

0

0

Ins-1

CUX1

120 100

100

Page 32 of 41

D

rel proliferation (%)

A

rel cell viability (%)

Fig. 2

0 24h 24h 20.000

CUX1 Mock

72h 72h 20.000

1

2

Page 333 of 41 Fig.

*

C 180

*

160

250

rel apoptosis (%)

140

rel apoptosis (%)

rel apoptosis (%)

300 200 180 160 140 120 100 80 60 40 20 0

*

B

A

200 150 100 50

100

*

* *

80 60

40 20 0

0

siCUX1Kontrolle

120

siControl siControl + TRAIL

siCUX1

siCUX1 + TRAIL

Kontrolle

5-FU

TRAIL

Fig. 4

A

Page 34 of 41

B

C

Mock CM

*

140

140 120

100

rel junctions (%)

rel tube length (%)

120

CUX1 CM

*

160

80 60 40

20

100 80 60 40

20 0

0 Mock

CUX1

Mock

CUX1

Page 355 of 41 Fig.

A

B

CUX1

HIF-1α

MMP-9

Actin Mock

CUX1

Fig. 6

Page 36 of 41

A

350

tumor volume (mm3)

300

B

250 200

CUX1

150

Actin 100

CUX1

50

Mock

0 Mock

CUX1

C

D

E

*

*

Page 377 of 41 Fig.

A

B CUX1 score (0-12) WHO 2010

Mean

SD

Grade 1 (n=43)

5.93

3.42

Grade 2 (n=12)

8.07

2.76

5.84

3.1

8.42

3.91

10

1.73

malignant insulinoma

benign insulinoma

ANOVA

p=0.038

Site

Primary malignant (n=7) Node metastasis (n=3) Liver metastasis (n=4) Behaviour

9.25

3.77

Benign (n=45)

5.84

3.1

Malignant (n=14)

9.0

3.35

10

C

8

p=0.018

p=0.002

CUX1 score (0-12)

Primary benign (n=45)

* 6

4

2

0

benign (n=45)

malignant (n=14)

Page 38 of 41

Supplementary table 1:

Gen KDR LECT1 FGF2 S1PR1 PLG B2M TYMP TIMP1 ITGAV HIF1A HPRT1 THBS1 MDK TGFB1 EGF MMP9 RPL13A COL4A3 PDGFA TNFAIP2 MMP2 IL6 CXCL10 ID3 CXCL5 CXCL1 IL8 FGF1 HAND2 TNF HPSE JAG1 IFNA1 ANGPT2 NOTCH4 BAI1 PPC EPHB4 LEP LAMA5 NRP1 FLT1 PPC ANGPTL4 ANGPTL3 NRP2 EREG ANPEP ACTB ITGB3 CXCL3

Mean Mock 0,013837346 0,007721688 0,043175161 0,026771547 3,504173384 4,909381398 0,01800581 17,84490845 0,571943826 0,093686954 0,004714184 15,43611121 0,022932073 3,763276052 0,154680068 1,016875812 6,219372684 0,513442938 0,6452159 0,212295013 34,64207815 0,053412 0,264761595 91,81934352 0,033820142 11,09310346 2,800664411 0,094556463 0,546534335 2,033891684 0,097055608 0,033640826 0,164218747 0,735365598 2,694045551 0,132992158 30,6006187 0,319706749 0,458161754 11,26627601 4,049482545 0,215822157 29,49857275 1,41552582 0,039439692 0,172521014 0,003836563 0,041009064 0,43108846 0,827412705 0,289900816

Mean CUX1 Ratio CUX1/Mock 0,168438095 12,17271676 0,037797342 4,894958286 0,199741252 4,626300142 0,09681443 3,616318108 11,43628072 3,263617256 14,93239433 3,041604046 0,05403968 3,001235806 52,41819159 2,937431241 1,626008557 2,842951497 0,260299475 2,77839617 0,013067971 2,772053309 41,00395078 2,656365338 0,059082348 2,576406807 9,534658231 2,533605853 0,388957394 2,514592854 2,518483322 2,476687214 14,91771001 2,398587569 1,211313222 2,359197356 1,513066725 2,345054925 0,492667631 2,320674533 79,60963987 2,298061898 0,121121869 2,267690219 0,58234963 2,19952456 197,5754107 2,151784179 0,072574846 2,145906005 23,77335091 2,143074839 5,979530197 2,135039876 0,200099074 2,116186109 1,150723077 2,105490914 4,203531807 2,0667432 0,199982508 2,060494102 0,067929648 2,019262161 0,324747324 1,977528936 1,421776954 1,933428703 5,087251116 1,888331515 0,250608201 1,884383292 56,60119559 1,849674876 0,577784978 1,807234224 0,827110301 1,805280108 20,3245752 1,804018931 7,273477375 1,796149828 0,368382121 1,706878132 49,61065422 1,68179846 2,376190242 1,678662592 0,063886978 1,619865014 0,274278043 1,589823967 0,005979709 1,558610909 0,062107202 1,514475013 0,640728474 1,486303935 1,220849572 1,47550257 0,425706364 1,468455212

Page 39 of 41

CXCL9 FIGF TGFBR1 ENG SERPINF1 VEGFC PPC IGF1 PTGS1 CXCL6 PECAM1 TEK TIMP2 COL18A1 EFNA3 HGF TGFB2 CCL11 IL1B CDH5 CCL2 ID1 HGDC TGFA RTC GAPDH STAB1 SPHK1 RTC RTC PROK2 EFNB2 FGFR3 VEGFA THBS2 PLXDC1 EFNA1 TIMP3 IFNG PLAU PF4 IFNB1 ANGPT1 AKT1 PGF

0,01926469 2,326479071 0,248461962 0,383813717 195,8770203 3,108387946 29,07698893 0,973910288 0,552431522 0,695412669 0,120535803 0,019078198 60,52816292 1,881125362 0,091993947 0,047130951 4,884682411 6,535483236 0,614307364 0,241642153 0,249498416 0,199747762 0,450384039 1,849344647 0,9020174 1 1,413187915 5,87315419 0,899972943 1,097648425 0,527375163 4,579227355 0,498187794 0,23320982 0,090939798 0,211162131 0,332668985 0,218102024 0,036817819 1,245727824 0,032311543 0,072765842 0,327848037 0,022439502 38941362,12

0,026789887 3,093537348 0,316412332 0,486135509 243,3049216 3,855212173 35,88920977 1,184060093 0,668032187 0,826682992 0,142393193 0,02229003 70,18636988 2,170979681 0,104682002 0,05320726 5,500134018 7,313492056 0,679307377 0,264203674 0,272502166 0,215666759 0,4647737 1,892539123 0,912466702 1 1,349722977 5,541238284 0,827435548 0,93298015 0,428146722 3,624535407 0,366594835 0,171535392 0,064577481 0,145728713 0,22919747 0,149491596 0,025166241 0,838757086 0,018681299 0,040719567 0,172369266 0,011654533 1,127221941

1,390621219 1,329707792 1,273483992 1,266592327 1,242131013 1,240260945 1,234282197 1,215779428 1,209257908 1,18876608 1,181335247 1,168350883 1,159565506 1,154085594 1,13792272 1,128923965 1,125996238 1,119043809 1,105810245 1,09336749 1,092199984 1,079695498 1,031949756 1,023356639 1,011584369 1 0,955090942 0,94348592 0,919400472 0,849980858 0,811844683 0,791516805 0,735856718 0,735541032 0,710112431 0,690127118 0,688965549 0,685420489 0,683534277 0,673306857 0,578161778 0,559597279 0,525759639 0,519375755 2,89466E-08

Page 40 of 41

Patient gender age

size (cm)

grade localization (WHO)

biological behavior

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

1,5 2 1,7 1,9 1,2 1 1,5 1,5 1 2 1,2 1,7 1 1 1,2 2,5 2,5 1,1 1,5 0,8 1 3 1,2 3 1 1 1 1,3 1 3,5

1 1 1 1 2 1 1 1 1 1 1 2 1 1 1 1 1 2 1 1 1 1 1 2 1 1 1 1 2 2

benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign

f f f f f f f f f f m m f f f f m m f f f m m m f f m m f f

52 68 18 61 49 35 17 16 76 26 68 20 40 77 38 79 35 30 52 31 73 32 60 43 32 30 41 39 35 35

primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary

CUX1 score (012) 0 6 6 9 9 6 6 9 9 9 6 9 2 6 3 2 4 6 4 3 6 4 6 9 2 12 1 12 9 8

recurrence-free survival (month) 295 94 45 187 68 252 67 93 80 115 72 95 55 48 138 35 149 89 126 123 42 na 79 144 149 52 208 62 51 1

overall survival (month)

disease related death

295 94 45 187 68 252 67 93 80 115 72 95 55 48 138 35 149 89 126 123 42 na 79 144 149 52 208 62 51 156

no no no no no no no no no no no no no no no no no no no no no no no no no no no no no no

Page 41 of 41

31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 53 48

m f f m f f f m f f m m f f m m f f f f m f f

47 32 50 59 45 57 66 65 48 70 40 36 68 45 55 66 48 64 63 56 23 59 64

2,2 1,2 2,2 2 0,9 1,2 2,5 1,5 1,3 0,8 1,8 1,1 1,3 1,5 1,5 1,9 1,5 1 1,7 8 1 1,5 na

2 1 1 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 1 1 na

50

f

56

na

2

54

f

41

2,2

1

46 55 56 50

m m m f

66 36 82 56

na 4,5 2,5 na

na 2 2 2

primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary primary lymph node metastasis lymph node metastasis lymph node metastasis liver metastasis liver metastasis liver metastasis liver metastasis

benign benign benign benign benign benign benign benign benign benign benign benign benign benign benign malignant malignant malignant malignant malignant malignant malignant malignant

6 2 6 2 4 2 6 9 12 6 2 9 6 6 2 12 2 12 9 12 6 6 9

99 39 122 39 145 153 96 59 49 58 71 96 42 95 238 0 0 na na 0 0 20 na

99 39 122 39 145 153 96 59 49 58 71 96 42 95 238 4 1 na 84 27 27 45 na

no no no no no no no no no no no no no no no yes yes na na no yes no na

malignant 9

0

27

no

malignant 12

108

108

no

malignant malignant malignant malignant

0 0 0 0

4 54 24 27

yes no yes no

12 12 4 9

CUX1: a modulator of tumour aggressiveness in pancreatic neuroendocrine neoplasms.

Pancreatic neuroendocrine neoplasms (PNENs) constitute a rare tumour entity, and prognosis and treatment options depend on tumour-mediating hallmarks ...
3MB Sizes 0 Downloads 6 Views