Β8 Integrin Mediates Pancreatic Cancer Cell Radiochemoresistance 2 3 Sha Jin1 †, Wei-Chun Lee1 †, Daniela Aust2-4, Christian Pilarsky5, Nils Cordes1, 4,6-8 *

Author Manuscript Published OnlineFirst on July 23, 2019; DOI: 10.1158/1541-7786.MCR-18-1352 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Jin and Lee et al., page 1

1 β8 integrin mediates pancreatic cancer cell radiochemoresistance 2 3 Sha Jin1 †, Wei-Chun Lee1 †, Daniela Aust2-4, Christian Pilarsky5, Nils Cordes1, 4,6-8 *

5 1OncoRay – National Center for Radiation Research in Oncology, Faculty of

6 Medicine and University Hospital Carl Gustav Carus, Technische Universität Dresden,

7 Helmholtz-Zentrum Dresden - Rossendorf, 01307 Dresden, Germany

8 2Institute for Pathology, University Hospital Carl Gustav Carus, Technische

9 Universität Dresden, Dresden 01307, Germany

10 3NCT Biobank Dresden, University Hospital Carl Gustav Carus, Technische

11 Universität Dresden, Dresden 01307, Germany

12 4German Cancer Consortium (DKTK), Partner Site Dresden, Heidelberg 69120,

13 Germany

14 5Department of Surgery, Universitätsklinikum Erlangen, Friedrich-Alexander

15 Universität Erlangen, 91054 Erlangen, Germany

16 6Department of Radiotherapy and Radiation Oncology, Faculty of Medicine and

17 University Hospital Carl Gustav Carus, Technische Universität Dresden, 01307

18 Dresden, Germany

19 7Institute of Radiooncology - OncoRay, Helmholtz-Zentrum Dresden - Rossendorf,

20 Dresden, 01328 Dresden, Germany

21 8German Cancer Research Center (DKFZ) – Partner Site Dresden, 69192 Heidelberg,

22 Germany

24 Running Title: β8 integrin in pancreatic cancer cells 25 26 Conflict of interest: The authors declare no potential conflicts of interest. 27 28 Word count: 4926; Total number of Figures: 7; References: 41

Downloaded from mcr.aacrjournals.org on September 23, 2021. © 2019 American Association for Cancer Research. Author Manuscript Published OnlineFirst on July 23, 2019; DOI: 10.1158/1541-7786.MCR-18-1352 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Jin and Lee et al., page 2

29 30 Keyword: Pancreatic cancer; β8 integrin; radiochemoresistance; cell survival; 31 intracellular vesicle trafficking 32 33 Financial support: The RADIATE project has received funding from the European 34 Union’s Horizon 2020 research and innovation programme under the Marie 35 Sklodowska-Curie grant agreement No. 642623

36 37 Author Notes: 38 † S. Jin and W.-C. Lee contributed equally to this paper. 39 40 41 * Corresponding author: 42 Prof. Dr. Nils Cordes 43 OncoRay – National Center for Radiation Research in Oncology, Technische 44 Universität Dresden, Fetscherstrasse 74 / PF 41, 01307 Dresden, Germany 45 Phone: +49 (0)351–458–7401. 46 Fax: +49 (0)351–458–7311. 47 E-mail: [email protected] 48 49

Jin and Lee et al., page 3

50 Abstract

51 Pancreatic ductal adenocarcinoma (PDAC) stroma, composed of extracellular matrix

52 (ECM) proteins, promotes therapy resistance and poor survival rate. Integrin-

53 mediated cell/ECM interactions are well known to controls cancer cell survival,

54 proliferation and therapy resistance. Here, we identified β8 integrin in a high-

55 throughput knockdown screen in three-dimensional (3D), ECM-based cell cultures for

56 novel focal adhesion protein targets as critical determinant of PDAC cell

57 radiochemoresistance. Intriguingly, β8 integrin localizes with the golgi apparatus

58 perinuclearly in PDAC cells and resection specimen from PDAC patients. Upon

59 radiogenic genotoxic injury, β8 integrin shows a microtubule-dependent perinuclear-

60 to-cytoplasmic shift as well as strong changes in its proteomic interactome regarding

61 the cell functions transport, catalysis and binding. Parts of this interactome link β8

62 integrin to autophagy, which is diminished in the absence of β8 integrin. Collectively,

63 our data reveal β8 integrin to critically co-regulate PDAC cell radiochemoresistance,

64 intracellular vesicle trafficking and autophagy upon irradiation.

66 Implications

67 This study identified β8 integrin as essential determinant of PDAC cell

68 radiochemosensitivity and as novel potential cancer target.

Jin and Lee et al., page 4

70 Introduction

71 Pancreatic ductal adenocarcinoma (PDAC) is one of the five most lethal

72 malignancies in the world. While the 5-year overall survival rate is about 15-20% for

73 resectable patients (1,2) most patients presenting at late stage with a treatment-

74 refractory disease survive significantly less. Neoadjuvant chemotherapy (e.g.

75 gemcitabine, nab-paclitaxel, FOLFIRINOX (folinic acid, 5-fluorouracil, irinotecan, and

76 oxaliplatin) is administered to patients, whose tumors seem irresectable or borderline

77 resectable. Alternatives such as radiation therapy and biologicals were neither

78 systematically evaluated in large clinical trials nor showed great advantages over

79 standard of care yet (2,3).

80 One putative cancer target in PDAC are PDAC cell/extracellular matrix (ECM)

81 interactions as PDACs are stroma-rich tumors (4). This stroma consists of numerous

82 ECM components such collagens, laminins, fibronectin, and hyaluronic acid (5). Cells

83 interact with ECM components via cell adhesion receptors that coalesce with

84 receptor tyrosine kinases, adapter and signaling molecules to form focal adhesion

85 complexes. These membranous multiprotein complexes are essentially co-regulating

86 key cell functions like survival, cell death, proliferation, metastasis and therapeutic

87 resistance (6,7). In previous work, we and others have documented the

88 radiochemosensitizing potential of integrin and focal adhesion protein targeting, as

89 the major family facilitating cell/ECM interactions. Examples for preclinically identified

90 novel targets are β1 integrin (8), αvβ3 integrin (9), αvβ6 integrin (10), FHL2 (11),

91 APPL1 and 2 (12), Caveolin 1 (8,13), as well as small integrin binding ligand N-linked

92 glycoproteins (called SIBLINGs) or secreted protein acidic and rich in cysteine (called

93 SPARC) families (14). Particularly the anti-integrin approaches were exploited for

94 molecular imaging (15), while biologicals inhibiting focal adhesion proteins (FAPs)

95 have not found their way into the clinic. Owing to the critical function of focal

Jin and Lee et al., page 5

96 adhesions for prosurvival signaling in normal and cancer cells, we established a high-

97 throughput esiRNA-based screening for more physiological three-dimensional (3D)

98 cell cultures to more systematically characterize the role of FAPs for PDAC

99 radiochemoresistance. Intriguingly, we identified β8 integrin as top druggable target

100 eliciting radiosensitization of PDAC cells.

101 β8 integrin, a 769-amino acids containing type I transmembrane protein,

102 consists of a large extracellular domain, a VWFA domain, four cysteine-rich repeats

103 and a short cytosolic domain (16,17). Recent studies demonstrated that, unlike other

104 β integrins, the cytosolic tail of β8 integrin shares no apparent homology and does

105 not directly influence cell adhesion. This suggests β8 integrin signaling to be distinct

106 from other β integrins (16). A connection between β8 and αV integrin as well as the

107 TGFβ signaling cascade has been reported (18). Others exhibited β8 integrin

108 involvement in liver cancer resistance to gefitinib (19), interactions with EphB4

109 receptor (20), as well as dependency of differentiation and radiosensitivity on β8

110 integrin in glioma cells (21).

111 Based on the fact that PDAC is highly therapy-refractory, there appears an

112 imbalance in survival and death mechanisms per se and under treatment. Partly, the

113 therapy resistance found in PDAC arises from autophagy (22). Autophagy is a highly

114 conserved catabolic process involving the formation of double-membraned vesicles

115 known as autophagosomes that engulf cellular proteins and organelles for delivery to

116 the lysosome (23). The impact of autophagy seems tissues and cancer type

117 dependent. In general, autophagy has two opposing function. One is cytoprotective

118 eliciting therapeutic resistance; the other one is cytotoxic inducing autophagic cell

119 death. Recent studies have shown that autophagy as a prosurvival and resistance

120 mechanism against chemotherapy treatment in PDAC (24,25).

Jin and Lee et al., page 6

121 In the present study, we explored the function of β8 integrin in PDAC therapy

122 resistance and unraveled parts of its contributing molecular mechanism. In PDAC cell

123 lines and PDAC primary cell cultures grown in 3D lrECM as well as in clinical

124 samples from PDAC patients, we observed that β8 integrin inhibition sensitizes

125 PDAC cells to x-rays and gemcitabine and localizes perinuclearly with the golgi

126 apparatus. Further, our data demonstrate that ionizing radiation induces a

127 microtubule-dependent perinuclear-to-cytoplasmic shift of β8 integrin and changes in

128 the composition of the β8 integrin interactome to transport, catalysis and binding

129 upon irradiation. Parts of the protein interactome of β8 integrin facilitate a connection

130 to autophagy, which is diminished in the absence of β8 integrin.

131

132

133 Materials and Methods

134 Antibodies and reagents

135 The antibodies against β1 integrin (WB 1:1000, Abcam, ab179471), β8 integrin

136 (IF 1:200, WB 1:1000; Abcam, ab80673, Rb, recognizes aa 614-663 (this antibody

137 was general used in this study); Abnova, H00003696-M01, Mo, recognizes aa 392 –

138 503 (this antibody was only used where specifically indicated), GM130 (IF 1:100, WB

139 1:1000, BD, 610822), MEK1/2 (WB 1:1000, Cell signaling, 4694s), γH2AX (WB

140 1:1000, Cell signaling, 9718s), αV integrin (IF 1:250, Novus, NB100-2618), APPL2

141 (IF 1:100, Sigma, SAB1400605), Caveolin 1 (IF 1:100, BD, 610407), LC3B (IF 1:500,

142 Sigma, SAB4200361), β actin (WB 1:10000, Sigma, A5441), HRP-conjugated

143 secondary antibody (GE, Rb NXA931, Mo NXA931), Alexa Fluor 488– or Alexa Fluor

144 594–conjugated secondary antibodies (1:500, Life Technologies) were purchased

145 and used as indicated.

146

Jin and Lee et al., page 7

147 Cell culture

148 Human pancreatic ductal adenocarcinoma (PDAC) cell lines BxPC3,

149 MiaPaCa2, Panc-1, Patu8902 were purchased from ATCC, Capan-1, patient-derived

150 primary cell lines (PacaDD119, PacaDD137, PacaDD159) and COLO357 cells were

151 a kind gift from Chr. Pilarsky (TU Dresden). Origin and stability of the cells were

152 routinely monitored by short tandem repeat analysis (microsatellites). Established cell

153 lines were cultured in Dulbecco’s modified Eagle’s medium (DMEM, Gibco,

154 61965026) with 10% fetal calf serum (FCS, PAN, A15-101) and 1% non-essential

155 amino acids (Gibco, 11140050). Patient-derived primary cell cultures (PacaDD119,

156 PacaDD137, PacaDD159) and COLO357 cells were cultured in DMEM with 33% K-

157 SFM (Gibco, 17005034) and 13% FCS. All cells were maintained at 37°C with 8.5%

158 CO2 at pH 7.4. In all experiments, asynchronously growing cells were used. All cells

159 were tested negative for mycoplasma by using a mycoplasma detection kit from

160 Minerva biolabs (Venor®GeM OneStep).

161

162 Construct of β8 integrin

163 Human β8 integrin generated by PCR-based amplification from cDNA ORF

164 Clone, purchased from Sino Biological (HG10367-M) with specific primers

165 (pAcGFPC1-ITGB8-XhoI-F and pACGFPC1-ITGB8-KpnI-R). Constructs were flanked

166 with Xho I and Kpn I restriction sites and inserted into the Xho I and Kpn I sites of

167 pAcGFP-C1 (Clontech).

168

169 Plasmid DNA transfection

170 Cells were plated onto uncoated 35 mm dishes with a 0.17 mm glass bottom

171 (MatTek) and allowed to reach 60 – 70% confluency. pAcGFPC1-ITGB8 (β8 / AcGFP)

172 and pAcGFPC1 empty vector (control, data not shown) were introduced into the cells

Jin and Lee et al., page 8

173 using Lipofectamine LTX (Invitrogen, 15338100) according to the manufacture´s

174 protocol. Briefly, cells were incubated in 250 µl OptiMEM with a DNA-mix of 250 µl

175 (containing plasmid DNA of 1–2 µg / µl and 10 µl Lipofectamine LTX with 15 µl with

176 PLUS Reagent in 240 µl OptiMEM). Transfection media was removed after 4 h and

177 cells were further incubated in fresh medium.

178

179 esiRNA transfection

180 250,000 cells were seeded per well of 6-well plate, esiRNAs or nonspecific

181 RLUC esiRNA of 1 µg / ml final concentrations (Eupheria Biotech) were transfected

182 by using oligofectamine (Invitrogen, 12252011) according to the manufacturer’s

183 protocol. Knockdown efficiency of β8 integrins was measured by Western blotting.

184

185 Radiation exposure

186 Irradiation was delivered at room temperature using 2, 4 or 6 Gy single doses

187 of 200-kV X-rays (Yxlon Y.TU 320; Yxlon; dose rate∼1.3 Gy/min at 20 mA) filtered

188 with 0.5 mm Cu as published (11). The absorbed dose was measured using a Duplex

189 dosimeter (PTW).

190

191 3D high-throughput esiRNA-based screening (3DHT-esiRNAS) against focal

192 adhesion proteins and data analysis

193 3DHT-esiRNAS was performed in MiaPaCa2 cells in 96-well plate. The

194 esiRNAs of 10 ng / µl final concentration were transfected using oligofectamine. The

195 cells were subcultured at 24 h post-transfection time. Cell suspensions were mixed

196 with laminin-rich extracellular matrix (IrECM, final conentration at 0.5 mg/ml) and

197 seeded into a new 96-well plate pre-coated with 1 % agarose. After 1 d incubation,

198 3D IrECM cultured cells were irradiated with 6 Gy x-rays. Grown tumoroids

Jin and Lee et al., page 9

199 containing 50 cells or more were counted at 8 d post-irradiation times. The

200 percentage of tumoroid formation (N) is given by:

201

202 where n is the number of tumoroids in each well treated with target esiRNAs, n0 is

203 average of control samples without irradiation or with 6 Gy. Average of six data sets

204 were collected from three independent screens.

205

206 3D tumoroid formation

207 In addition to the 3D high-throughput screen, as described above, 3D tumoroid

208 formation was determined under identical growth conditions after irradiation (2 – 6 Gy

209 x-rays) or Gemcitabine treatment (for 24 h; PBS as control) as published previously

210 (26,27). For Gemcitabine, media were removed, cells repeatedly washed and fresh

211 media added to allow tumoroid growth over 8 d. Each point on the survival curves

212 represents the mean surviving fraction from at least three independent experiments.

213

214 Western blot analysis

215 Cells were lysed with modified RIPA buffer consisting of 50 mM Tris-HCl (pH

216 7.4), 1% Nonidet-P40, 0.25% sodium deoxycholate, 150 mM NaCl, 1 mM EDTA, 1

217 mM NaVO4, 2 mM NaF (all Sigma-Aldrich), complete protease inhibitor cocktail

218 (Roche) as described previously (28). Total protein amount was measured by BCA

219 assay (Thermo Fisher Scientific). After SDS–PAGE and transfer of proteins onto

220 nitrocellulose membranes (GE Healthcare), probing of specific proteins was

221 accomplished using indicated primary antibodies and horseradish peroxidase-

222 conjugated donkey anti-rabbit and sheep anti-mouse antibodies. Enhanced

Jin and Lee et al., page 10

223 chemiluminescent reagent (GE Healthcare) was used for detection of proteins on x-

224 ray films (GE Healthcare) and protein expression levels were measured by

225 densitometry.

226

227 Protein fractionation assay

228 The Subcellular Protein Fractionation Kit for Cultured Cells (Calbiochem,

229 ProteoExtract® Subcellular Proteome Extraction Kit, 539790) was used according to

230 the manufacturer’s protocol. In brief, cells lysis and fractionation were conducted after

231 24 h plating. Equal loading was ensured by total protein concentration measurement

232 using the BCA assay (Pierce). Fractionation efficacy was confirmed with detection of

233 β1 integrin for the membrane fraction, MEK1/2 for the cytosol fraction and γH2AX for

234 the nuclear fraction.

235

236 Immunofluorescence staining

237 In brief, cells were grown on coverslip or 35 mm dishes with a 0.17 mm glass

238 bottom (MatTek), fixed with 3.7 % PFA at room temperature for 15 min,

239 permeabilized with 0.25 % Triton X–100 for 10 min and washed with PBS as

240 published (27). Then, cells were incubated with blocking buffer (1% BSA in PBS) at

241 room temperature for 1 h followed by incubation with primary and secondary antibody

242 in blocking buffer in the dark at room temperature for 1 h. The cells were washed with

243 PBS three times between each step. Then the cover slides were mounted using

244 ProLong™ Diamond Antifade Mountant with DAPI. Samples were stored in dark until

245 microscopy.

246

247 Microscopy and image analysis

Jin and Lee et al., page 11

248 To investigate localization of β8 integrin and expression levels of LC3,

249 samples were imaged by a Zeiss Axioscope1 epifluorescence microscope using a

250 40x/0.75 or 100x/1.25 oil objective. Images for colocalization analysis were acquired

251 by a Zeiss LSM 510 meta microscope usin g 63x/1.2 water immersion objective and

252 were analyzed by Pearson correlation analysis using Coloc 2 plugin of Fiji. Images

253 for translocation and expression of β8 integrin were recorded by spinning disc

254 confocal microscopy (Olympus IX83 microscope with Yokogawa CSU–X1 Confocal

255 Spinning Disc unit, and iXon Ultra 897 EMCCD Camera) with 60x/1.2 NA water

256 immersion objective. Z-stacks with a step of 1 µm were acquired per image.

257 Maximum intensity projection of z–stacks was processed using Fiji. All the detected

258 β8 integrin signals from the whole cell were analyzed. The cell size and nucleus size

259 were measured on bright field images. Coordinates of each β8 integrin (Xn, Yn) were

260 determined by the position of maximum intensity. Distance of each β8 integrin to

261 nucleus center (D) is given by,

262

263 Where x0, y0 are the local position of nucleus center. Relative β8 integrin to nucleus

264 center distance (rel. D),

265

266

267

268 Sulforhodamine B (SRB) Assay

269 BxPC3 cells grown in a 96-well plate were treated with colchicine, paclitaxel,

270 chloroquine and gemcitabine with serial dilution. After three days, cells were fixed

271 with trichloroacetic acid. After washing with tap water, 50 μl of 0.04% (wt/vol) SRB

Jin and Lee et al., page 12

272 solution was added to each well and incubated for 1 h. Then, plates were rinsed four

273 times with 1% (vol/vol) acetic acid to remove unbound dye. Tris base solution (pH

274 10.5; 100 μl of a 10 mM solution) was added to solubilize the protein-bound dye.

275 Measurement of absorbance was accomplished at 510 nm in a microplate reader

276 (Tecan).

277

278 Histology

279 Patient material was examined according to the ethic approval (EK 378092017;

280 dated 02.01.2018) provided by the ethic committee of the Technische Universität

281 Dresden. The tissue used was provided by the tissue bank for tumor and normal

282 tissues of the UCC/NCT Dresden and used in accordance with the regulations of the

283 tissue bank. For immunofluorescence analysis, paraffin sections were deparaffinized

284 in xylene and rehydrated. Antigen retrieval was performed in 10 mM citric acid, pH

285 6.0, at 98°C for 15 min and sections were stained with hematoxylin and eosin or

286 antibodies against β8 integrin. For tyramide based immunofluorescence detection,

287 the TSA-kit (Thermo Fisher Scientific) was used according to the manufacturer’s

288 instructions. Sections were mounted using ProLong Gold Antifade Mountant with

289 DAPI (Thermo Fisher Scientific) for nuclear counterstaining. Images were acquired

290 using an Axioscope1 plus fluorescence microscope (Zeiss).

291

292 Immunoprecipitation

293 Immunoprecipitation were performed as previously published (26). In brief, 1

294 mg of whole cell lysates (Cell Lysis Buffer, Cell Signaling Technology) were

295 incubated with specific antibodies at 4°C for 1 h. Beads were washed with Cell Lysis

296 Buffer (Cell Signaling Technology) and added to lysates. After overnight incubation at

Jin and Lee et al., page 13

297 4°C, immunoprecipitates were washed, mixed with sample buffer, and loaded on

298 SDS gels.

299

300 Sequential immunoprecipitation/mass spectrometry-based proteomics (IP-MS)

301 and data analysis

302 To identify β8 integrin interacting proteins, mass spectrometric analysis was

303 carried out at the Max Planck Institute of Molecular Cell Biology and Genetics Mass

304 Spectrometry (MPI-CBG MS) Facility (Dresden, Germany) and performed as

305 published (29). In brief, immunoprecipitates were separated by gel electrophoresis,

306 in-gel digested with trypsin and peptides recovered from the gel matrix analyzed on a

307 LTQ Orbitrap XL mass spectrometer (Thermo Fischer Scientific) coupled on-line to

308 Ultima3000 LC system (Dionex) via a TriVersa robotic ion source (Advion

309 BioScience). PANTHER (Protein ANalysis THrough Evolutionary Relationships)

310 classification system (http://www.pantherdb.org/) was used to category interactomes

311 of β8 integrin according to their functions. Connections between the identified β8

312 integrin interactome and autophagy were analyzed by Autophagy Regulatory

313 Network (http://arn.elte.hu/).

314

315 Statistics

316 All results represent mean ± standard deviation (SD) of at least three

317 independent experiments. Unpaired, two-sided Student’s t-test was performed by

318 Microsoft Excel. A p-value is less than 0.05 was considered significant.

319

320

321 Results

Jin and Lee et al., page 14

322 High-throughput RNAi screen in 3D PDAC cells identifies potential focal

323 adhesion protein targets involved in radioresistance

324 The role of focal adhesion proteins (FAPs) in PDAC therapy resistance has

325 not been comprehensively unraveled. By means of a 3D high throughput RNA

326 interference (3D-HTP-RNAi) screen (Fig. 1A, and Supplementary Tab. 1), we

327 sought to identify novel potential FAP candidates whose depletion potently

328 diminishes PDAC tumoroid forming ability under basal (Fig. 1B) or irradiation

329 conditions (Fig. 1C). We found basal tumoroid forming ability to be significantly

330 (p<0.05) compromised upon depletion of, for example, Synemin, JUB, ITGB5, CRK,

331 GAB1 and Caveolin 1 relative to controls (Fig. 1B) . Upon 6 Gy x-rays, we

332 surprisingly found none of the FAP depletions to result in a significantly enhanced

333 radiosensitivity of MiaPaCa2 cells (Fig. 1D). We proceeded with the top ten

334 candidates for further evaluation based on the following criteria (Supplementary Fig.

335 1): (i) novelty in the context of radiosensitivity, (ii) current knowledge about the

336 candidates and, (iii) higher RNA expression levels in PDAC compared to normal

337 tissue according to Oncomine data base. Our analyses favored β8 integrin as a novel

338 and potent FAP candidate whose depletion enhanced, although not significantly,

339 PDAC cell radiosensitivity in our screen.

340

341 β8 integrin is overexpressed and located in the perinuclear region in PDAC

342 Next, we performed an Oncomine data (https://www.oncomine.org) analysis

343 and found a significant, 3.1-fold upregulation of β8 integrin mRNA expression level in

344 PDAC as compared to normal pancreas (Fig. 2A). Furthermore, an analysis using

345 OncoLnc. (http://www.oncolnc.org) for analysis of TCGA data base of the PDAC

346 patient cohort suggested high β8 integrin expression levels to result in shorter

347 survival of PDAC patient as compared to lower expression levels (30) (Fig. 2B). Then,

Jin and Lee et al., page 15

348 we determined the protein expression level of β8 integrin in six established PDAC

349 cell lines (BxPC3, Capan1, Colo357, Mia PaCa2, Panc1, Patu8902) and three

350 human PDAC patient primary cell cultures (PacaDD119, PacaDD137 and

351 PacaDD159) grown in IrECM 3D (Fig. 2C and D). The expression of β8 integrin

352 varied in a cell line-dependent manner (Fig. 2C and D). To characterize the

353 subcellular location of β8 integrin, immunofluorescence staining was performed in a

354 panel of human pancreatic cancer cell lines indicating β8 integrin to be located in the

355 perinuclear area in PDAC cells (Fig. 3A, B and D). In contrast, a clear β8 integrin

356 accumulation at the plasma membrane was observable in glioblastoma cell lines (Fig.

357 3C and E). Next, we examined cytosol, membrane and nucleus protein fractions

358 revealing β8 integrin localized in cytosol and membranes in contrast to our β1

359 integrin control found only in the membrane fraction (Supplementary Fig. S2A and

360 B). Intriguingly, confirmatory data for a perinuclear localization in situ are presented

361 by β8 integrin staining in resection specimen of human PDAC (Fig. 3F and G). To

362 further address this abnormal localization, we screened the International Cancer

363 Genome Consortium (ICGC; icgc.org) database for mutations in the ITGB8 gene

364 (https://icgc.org/ZyH). In 52 out of 472 donors, 83 mutations are found of which 67

365 are in introns and 9 in exons. The COSMIC database provided no ITGB8 gene

366 mutations in any of the cell lines used in the presented study

367 (https://cancer.sanger.ac.uk/cosmic/gene/samples?all_data=&coords=AA%3AAA&dr

368 =&end=770&gd=&id=5149&ln=ITGB8&seqlen=770&sn=pancreas&src=gene&start=1

369 #positive). These findings demonstrate a β8 integrin expression in PDAC cells

370 without functionally critical gene mutations and, as compared to other integrin

371 subunits, an unusual subcellular localization that warrants further investigation.

372

Jin and Lee et al., page 16

373 β8 integrin is a critical regulator of cellular sensitivity to ionizing radiation and

374 cytotoxic drugs

375 To confirm that β8 integrin plays an essential role in PDAC

376 radiochemoresistance, we effectively silenced β8 integrin in a panel of PDAC cell

377 lines (Fig. 4A and B), detected unaltered tumoroid growth (Fig. 4C and E) but

378 observed a highly significant decrease in tumoroid growth upon irradiation relative to

379 controls (Fig. 4D and E and Supplementary Fig. 3A-D). In addition, β8 integrin-

380 depleted PDAC cell lines (BxPC3, Patu8902) exhibited enhanced chemosensitivity to

381 gemcitabine (Fig. 4F and supplementary Fig. S3E). Collectively, our data suggest a

382 critical function in PDAC cell survival after x-ray irradiation and gemcitabine treatment.

383

384 Changes in the composition of the β8 integrin interactome to transport,

385 catalysis and binding upon irradiation

386 To define the interacting proteins likely to be causative for the perinuclear

387 localization of β8 integrin, sequential immunoprecipitation-mass spectrometry was

388 employed. Firstly, we found 133 proteins bound to β8 integrin in unirradiated cells

389 (Fig. 5A and Supplementary Tab. 2). Intriguingly, the interactions changed

390 dramatically at 2 h after 6-Gy x-rays revealing 637 interacting proteins (Fig. 5A), of

391 which only 32 proteins were present under both untreated and irradiation conditions.

392 This tremendous change in the interactome stimulated us to perform a GO analysis

393 based on the protein molecular functions according to PANTHER classification

394 system (see categorization of proteins in Supplementary Tab. 2). We calculated the

395 increase rate between the protein number in unirradiated versus 6-Gy irradiated cells.

396 The number of interacting proteins increased in different categories as follows:

397 transporter activity by 95%, antioxidant activity by 100%, catalytic activity appr. by

398 84%, signal transducer activity appr. by 83%, structural molecule activity appr. by

Jin and Lee et al., page 17

399 84%, receptor activity appr. by 64%, binding appr. by 68%, and translation regulator

400 activity appr. by 86%, respectively (Fig. 5B). Considering both factors, the absolute

401 number of proteins and their relative increase before and after irradiation, we

402 reckoned that the function of β8 integrin undergoes a critical shift towards

403 transport/binding and catalysis.

404 Moreover, we checked a small selection of putative proteins related to

405 transport (golgi apparatus reported by GM130; APPL2, Caveolin 1) and stress

406 (mitochondria) and published interaction partners (αV integrins) (18). In contrast to

407 mitochondria, the previously reported interacting integrin αv, APPL2 and Caveolin 1,

408 only GM130 co-localized to β8 integrin as indicative from the calculated Pearson

409 correlation (0.55 ± 0.04) (Fig. 5C and D and Supplementary Fig. 4).

410 As another possible explanation for the abnormal β8 integrin localization,

411 genetic mutations were evaluated in ITGAV. The ICGC database revealed 54 out of

412 472 donors with mutations (https://icgc.org/Zyr). In total, 78 gene mutations can be

413 found, of which 60 are located in introns and 3 are categorized missense. Similar to

414 ITGB8, no mutations were recognized in any of the used PDAC cell lines using

415 COSMIC database

416 (https://cancer.sanger.ac.uk/cosmic/gene/samples?all_data=&coords=AA%3AAA&dr

417 =&end=1049&gd=&id=5530&ln=ITGAV&seqlen=1049&sn=pancreas&src=gene&start

418 =1#positive). OncoLnc. (http://www.oncolnc.org) based analysis of the PDAC patient

419 cohort indicated high αV integrin expression to significantly correlate with shorter

420 survival of PDAC patient as compared to lower expression levels

421 (http://www.oncolnc.org/kaplan/?lower=10&upper=90&cancer=PAAD&gene_id=3685

422 &raw=ITGAV&species=mRNA) (Supplementary Fig. S5).

423

424 β8 integrin translocated from perinuclear area to cytosol upon irradiation

Jin and Lee et al., page 18

425 To contextually connect the β8 integrin interactome data with β8 integrin

426 behavior upon irradiation, we next conducted immunofluorescence staining followed

427 by analysis of positive particle intensity and the distance from the cell center to the

428 particles. We correlated this by multiplying with a correction factor, which was

429 determined by measuring the ratio of cell size to nucleus size. Under normal

430 conditions, β8 integrin was located in the perinuclear area. However, upon irradiation,

431 the distance of β8 integrin particles increased from center to nucleus indicating a

432 movement of β8 integrin-positive particles from perinuclear area to cell membrane.

433 This dynamic was accompanied by a significant, appr. 4-fold elevation of β8 integrin

434 expression at already 2 h as well as 24 h after genotoxic stress induced by 6-Gy x-

435 ray irradiation (Fig. 6A and B). Further, both average and maximum distance as

436 measure of perinuclear-to-cytoplasmic translocation were significant at 24 h but not 2

437 h post irradiation (Fig. 6A, C and D).

438 The association of the β8 integrin interactome proteins led us to speculate that

439 the microtubule system is involved in the radiogenic β8 integrin perinuclear-to-

440 cytoplasmic shift. Therefore, we pretreated cells with the microtubule assembly

441 inhibitor colchicine and found a loss of the significance change in average and

442 maximum distance of β8 integrin (Fig. 6E-G). This finding suggests microtubule-

443 dependence of the radiogenic perinuclear-to-cytoplasmic translocation of β8 integrin.

444

445 β8 integrin interactome connects to autophagy

446 Based on our observations that PDAC cells are radiochemosensitized by β8

447 integrin targeting and β8 integrin protein interaction partners are associated not only

448 with transporter activity and binding but also with catalytic activity, we hypothesized

449 β8 integrin to function, for example, in autophagy. Hence, candidate proteins involved

450 in autophagy were identified in the list of interacting proteins (Fig. 7A). While electron

Jin and Lee et al., page 19

451 transfer flavoprotein alpha subunit (ETFA), Thymopoietin (TMPO) and Glycoprotein

452 P43 (RBMX) dispersed from β8 integrin at 2 h after 6-Gy x-rays (Fig. 7A), DnaJ

453 homolog subfamily C member 7 (DNAJC7), aldehyde dehydrogenase 9 family

454 members A1 (ALDH9A1) already present under non-irradiated conditions showed an

455 enhanced binding to β8 integrin at 2 h after irradiation. Intriguingly, we also found

456 newly connecting proteins such as MAPK8, PRKAR1A, SAFB2, PRKAG1, DNAJB1,

457 SMC1A, MYH4, RHEB, FKBP1A, RPA2, HADHA, HACD3 and DIP2B categorized as

458 proteins regulating autophagy according to the PANTHER classification system.

459 Collectively, our interactome findings suggest β8 integrin to contribute to the

460 regulation of autophagy after exposure to x-rays in PDAC cells.

461

462 Depletion of β8 integrin reduces autophagy induction

463 Next, we investigated whether β8 integrin per se or together with the

464 microtubule system affects LC3B expression as key readout for autophagy. In

465 addition to the microtubule assembly inhibitor colchicine, we pretreated cells with the

466 microtubule stabilizing agent paclitaxel and the autophagy inhibitor chloroquine (31)

467 (Supplementary Fig. S6A-C). Unirradiated esiRNA control cells showed non-

468 significant elevation in LC3B intensity under colchicine and chloroquine (Fig. 7B, 7C

469 and Supplementary Fig. S7), while paclitaxel led to a slight reduction of LC3B

470 intensity (Fig. 7B). In contrast, β8 integrin depletion elicited a significant decline in

471 LC3B fluorescence intensity under exposure to all three agents, including PBS,

472 relative to esiRNA controls (Fig. 7B). Upon irradiation, the intensity of LC3B was not

473 significantly increased in esiRNA control cells independent of PBS or a specific agent

474 relative to unirradiated esiRNA controls (Fig. 7B). Similar to non-irradiated cells, β8

475 integrin-depleted and irradiated cells showed a significant reduction in LC3B intensity

476 (Fig. 7B).

Jin and Lee et al., page 20

477 To determine whether the conditions tested for LC3B fluorescence intensity

478 impacts on PDAC cell survival, cells were depleted of β8 integrin without and in

479 presence of colchicine, paclitaxel or the autophagy inhibitor chloroquine. Intriguingly,

480 3D lrECM PDAC tumoroid formation remained unaffected upon treatment (Fig. 7D),

481 while irradiated and treated PDAC cell cultures under removal of β8 integrin showed

482 significant lower tumoroid formation. This effect was not significantly different

483 between the two microtubule-inhibiting agents, but significantly lower for chloroquine

484 (Fig. 7E). The salient findings are that β8 integrin regulates LC3B expression in a

485 microtubule-independent manner and that the interplay between β8 integrin and

486 autophagy measured by LC3B intensity becomes functionally effective for cell

487 survival only upon stress as induced here by x-ray irradiation.

488

489

490 Discussion

491 Integrin-mediated cell/ECM interactions fundamentally co-regulate cancer cell

492 resistance to therapy. Owing to the massive stroma in PDAC, we conducted a high-

493 throughput knockdown screen in three-dimensional (3D), ECM-based cell cultures to

494 identify novel focal adhesion protein targets. Here, we show that β8 integrin (i) is

495 overexpressed in PDAC relative to normal pancreas, (ii) is critically involved in

496 radiochemoresistance in PDAC cells, (iii) co-localizes with the golgi apparatus

497 perinuclearly in PDAC cells and in resection specimen from PDAC patients, (iv)

498 shows a microtubule-dependent perinuclear-to-cytoplasmic shift and strong changes

499 in its proteomic interactome regarding the cell functions transport, catalysis and

500 binding upon radiogenic genotoxic injury, and (v) connects with parts of its

501 interactome to autophagy.

Jin and Lee et al., page 21

502 β8 integrin does not show an apparent homology to other β integrin subunits in

503 its cytosolic tail questioning its role in adhesion to ECM and downstream signaling

504 (16,17). Nonetheless, β8 integrin is overexpressed in various cancer types originating

505 from brain (21), lung (32), prostate (20), and gastrointestinal tract (19,33) as well as

506 pancreas as shown here. Intriguingly, while other β integrins are located in the cell

507 membrane, β8 integrin is found in the perinuclear region in 3D IrECM grown PDAC

508 cultures as well as in resection specimen from PDAC patients. Our characterization

509 of this perinuclear localization revealed co-localization with the golgi apparatus

510 marker GM130 but neither with the reported integrin αV to form one of the primary

511 receptors for latent-TGFβ1 and latent-TGFβ3 (18) nor mitochondria, APPL2 or

512 Caveolin 1. Genetic mutations in ITGB8 as well as its reported binding partner ITGAV

513 were not detected in regions inevitably entailing dysfunctionality in the protein using

514 ICGC and COSMIC databases. Hence, the underlying mechanism possibly involving

515 abnormal function of golgi and endoplasmic reticulum proteins requires further

516 exploration.

517 In addition, β8 integrin is documented to impact on cells from different tumor

518 types. In HepG2 liver cancer cells, β8 integrin contributes to resistance to gefitinib via

519 multidrug transporters and apoptosis regulators (19). It interacts with EphB4 receptor

520 in prostate cancer cells (20) and controls differentiation and radiosensitivity of glioma

521 cells (21). Moreover, β8 integrin has central roles in promoting glioma initiation in

522 vitro and in vivo and its inhibition reduces glioma cell self-renewal ability, stemness

523 and migration ability (21,34). These observations are in line with our data

524 demonstrating that β8 integrin targeting sensitizes PDAC cells grown under 3D

525 lrECM-based conditions to irradiation and chemotherapy.

526 Little is known about the underlying mechanisms how β8 integrin controls

527 cancer cell survival. We therefore characterized the β8 integrin proteomic

Jin and Lee et al., page 22

528 interactome and found pronounced changes in its composition upon radiogenic

529 genotoxic injury. The deduced categorization of interacting proteins from this analysis

530 revealed that β8 integrin engages in transporter activity, antioxidant activity, catalytic

531 activity, signal transducer activity, structural molecule activity, receptor activity,

532 binding, and translation regulator activity in genotoxically injured cells. This wide

533 spectrum of functions stimulated us to focus on transport and catalysis.

534 Upon genotoxic stress, we found β8 integrin to undergo a perinuclear-to-

535 cytoplasmic translocation that seems cytoprotective, as a siRNA-mediated β8 integrin

536 depletion resulted in elevated cell death and radiosensitization. Mechanistically, the

537 cytoprotective process depended on functional microtubules. Disturbing microtubule

538 assembly by the microtubule inhibitor colchicine strongly reduced the perinuclear-to-

539 cytoplasmic translocation and induced radiosensitization. Which proteins from our β8

540 integrin interactome profile are responsible for these actions require further

541 investigation.

542 The interactome also included proteins involved in autophagy. Autophagy is a

543 regulated catabolic pathway to degrade cellular organelles and macromolecules. In

544 malignant tumors, autophagy can either function as a prosurvival response to stress

545 such as starvation, hypoxia, and chemo- and radiotherapy that is thought to mediate

546 resistance to anticancer therapies or as an antisurvival response (23,35). Intriguingly,

547 β8 integrin seems to be a critical autophagy regulator and autophagy is a

548 cytoprotective mechanisms in PDAC. These findings are in line with published work

549 describing high levels of basal autophagy and macropinocytosis arising from intense

550 metabolic rewiring (36). While the various proteins found in the β8 integrin

551 interactome also require further attention to unravel the underlying mechanism, some

552 are already well known determinants of the autophagy process such as RHEB

553 (37,38), DNAJB1 (39), FKBP1A (40), and HADHA (41). Importantly, we observed

Jin and Lee et al., page 23

554 increasing interactions of these autophagy associated proteins with β8 integrin upon

555 radiogenic genotoxic stress. Accordingly, the β8 integrin expression in PDAC cells

556 seems connected with autophagy detected by LC3 intensity. LC3 intensity was

557 clearly dependent on β8 integrin expression but neither significantly altered by the

558 two microtubule-inhibiting agents colchicine and paclitaxel nor the late phase

559 autophagy inhibitor chloroquine. Concerning tumoroid formation, it was interesting to

560 us to exhibit unchanged tumoroid forming capacity in β8 integrin-expressing cells that

561 were pretreated with these three agents. In contrast, radiosensitization occurred in

562 dependence of β8 integrin depletion without strong impact of microtubule or

563 autophagy inhibitors.

564 Taken together, β8 integrin is overexpressed in PDAC and plays an essential

565 role in PDAC cell radiochemosensitivity. Based on its translocation and regulation of

566 autophagy upon radiogenic genotoxic injury, β8 integrin seems to be critically

567 involved in prosurvival mechanisms. The β8 integrin driven molecular circuitry and

568 prosurvival signaling networks as well as its druggability warrant further investigations

569 to identify the potential of β8 integrin as potential cancer target.

570

571

572 Acknowledgements

573 The authors thank Inga Lange for excellent technical assistance and all the

574 members of the Core Facility Cellular Imaging (CFCI) at Faculty of Medicine Carl

575 Gustav Carus, Technische Universität Dresden for technical assistance. The tissue

576 used was provided by the “Tumor- und Normalgewebebank des UCC Dresden” and

577 used in accordance with the regulations of the tissue bank (and the vote of the ethics

578 committee from the Technische Universität Dresden). This project has received

Jin and Lee et al., page 24

579 funding from the European Union’s Horizon 2020 research and innovation

580 programme under the Marie Skłodowska-Curie grant agreement No 642623.

581

Jin and Lee et al., page 25

582 References

583 1. Narayanan V, Weekes CD. Molecular therapeutics in pancreas cancer. World J

584 Gastrointest Oncol. Baishideng Publishing Group Inc; 2016;8:366–79.

585 2. Birnbaum DJ, Bertucci F, Finetti P, Birnbaum D, Mamessier E. Molecular

586 classification as prognostic factor and guide for treatment decision of

587 pancreatic cancer. Biochim Biophys Acta - Rev Cancer [Internet]. 2018 [cited

588 2018 Nov 13];1869:248–55. Available from:

589 http://www.ncbi.nlm.nih.gov/pubmed/29499330

590 3. Conway JR, Herrmann D, Evans TJ, Morton JP, Timpson P. Combating

591 pancreatic cancer with PI3K pathway inhibitors in the era of personalised

592 medicine. Gut [Internet]. 2018 [cited 2018 Nov 13];gutjnl-2018-316822.

593 Available from: http://www.ncbi.nlm.nih.gov/pubmed/30396902

594 4. Weniger M, Honselmann KC, Liss AS. The Extracellular Matrix and Pancreatic

595 Cancer: A Complex Relationship. Cancers (Basel) [Internet]. 2018 [cited 2018

596 Nov 13];10:316. Available from: http://www.ncbi.nlm.nih.gov/pubmed/30200666

597 5. Feig C, Gopinathan A, Neesse A, Chan DS, Cook N, Tuveson DA. The

598 Pancreas Cancer Microenvironment. 2012;18:4266–77.

599 6. Dickreuter E, Cordes N. The cancer cell adhesion resistome: mechanisms,

600 targeting and translational approaches. Biol Chem [Internet]. 2017 [cited 2017

601 Dec 18];398:721–35. Available from:

602 http://www.ncbi.nlm.nih.gov/pubmed/28002024

603 7. Winograd-Katz SE, Fässler R, Geiger B, Legate KR. The integrin adhesome:

604 from genes and proteins to human disease. Nat Rev Mol Cell Biol [Internet].

605 2014 [cited 2014 Jul 10];15:273–88. Available from:

606 http://www.ncbi.nlm.nih.gov/pubmed/24651544

607 8. Cordes N, Frick S, Brunner TB, Pilarsky C, Grützmann R, Sipos B, et al.

Jin and Lee et al., page 26

608 Human pancreatic tumor cells are sensitized to ionizing radiation by

609 knockdown of caveolin-1. Oncogene. 2007;26:6851–62.

610 9. Seguin L, Kato S, Franovic A, Camargo MF, Lesperance J, Elliott KC, et al. An

611 integrin β3–KRAS–RalB complex drives tumour stemness and resistance to

612 EGFR inhibition. Nat Cell Biol [Internet]. 2014 [cited 2018 Nov 13];16:457–68.

613 Available from: http://www.ncbi.nlm.nih.gov/pubmed/24747441

614 10. Hezel AF, Deshpande V, Zimmerman SM, Contino G, Alagesan B, O’Dell MR,

615 et al. TGF- and v 6 Integrin Act in a Common Pathway to Suppress

616 Pancreatic Cancer Progression. Cancer Res [Internet]. 2012 [cited 2018 Nov

617 13];72:4840–5. Available from: http://www.ncbi.nlm.nih.gov/pubmed/22787119

618 11. Zienert E, Eke I, Aust D, Cordes N. LIM-only protein FHL2 critically determines

619 survival and radioresistance of pancreatic cancer cells. Cancer Lett. 2015;364.

620 12. Hennig J, McShane MP, Cordes N, Eke I. APPL proteins modulate DNA repair

621 and radiation survival of pancreatic carcinoma cells by regulating ATM. Cell

622 Death Dis. Macmillan Publishers Limited; 2014;5:e1199.

623 13. Hehlgans S, Eke I, Storch K, Haase M, Baretton GB, Cordes N. Caveolin-1

624 mediated radioresistance of 3D grown pancreatic cancer cells. Radiother Oncol.

625 2009;92:362–70.

626 14. Kaleağasıoğlu F, Berger MR. SIBLINGs and SPARC families: their emerging

627 roles in pancreatic cancer. World J Gastroenterol [Internet]. 2014 [cited 2018

628 Nov 13];20:14747–59. Available from: http://www.wjgnet.com/1007-

629 9327/full/v20/i40/14747.htm

630 15. Parsons CM, Sutcliffe JL, Bold RJ. Preoperative evaluation of pancreatic

631 adenocarcinoma. J Hepatobiliary Pancreat Surg [Internet]. 2008 [cited 2018

632 Nov 13];15:429–35. Available from: http://link.springer.com/10.1007/s00534-

633 007-1240-7

Jin and Lee et al., page 27

634 16. Nishimura SL, Sheppard D, Pytela R. Integrin alpha v beta 8. Interaction with

635 vitronectin and functional divergence of the beta 8 cytoplasmic domain. J Biol

636 Chem [Internet]. 1994 [cited 2017 Apr 19];269:28708–15. Available from:

637 http://www.ncbi.nlm.nih.gov/pubmed/7525578

638 17. Moyle M, Napier M a, McLean JW. Cloning and expression of a divergent

639 integrin subunit beta 8. J Biol Chem. 1991;266:19650–8.

640 18. Cambier S, Gline S, Mu D, Collins R, Araya J, Dolganov G, et al. Integrin αvβ8-

641 Mediated Activation of Transforming Growth Factor-β by Perivascular

642 Astrocytes. Am J Pathol. American Society for Investigative Pathology;

643 2005;166:1883–94.

644 19. Wang W-W, Wang Y-B, Wang D-Q, Lin Z, Sun R-J. Integrin beta-8 (ITGB8)

645 silencing reverses gefitinib resistance of human hepatic cancer HepG2/G cell

646 line. Int J Clin Exp Med. 2015.

647 20. Mertens-Walker I, Fernandini BC, Maharaj MS, Rockstroh A, Nelson CC,

648 Herington AC, et al. The tumour-promoting receptor tyrosine kinase, EphB4,

649 regulates expression of Integrin-β8 in prostate cancer cells. BMC Cancer.

650 BioMed Central; 2015;15:164.

651 21. Malric L, Monferran S, Delmas C, Arnauduc F, Dahan P, Boyrie S, et al.

652 Inhibiting Integrin β8 to Differentiate and Radiosensitize Glioblastoma-initiating

653 Cells. Mol Cancer Res. American Association for Cancer Research;

654 2018;molcanres.0386.2018.

655 22. Boone BA, Zeh HJ, Bahary N. Autophagy Inhibition in Pancreatic

656 Adenocarcinoma. Clin Colorectal Cancer [Internet]. 2017 [cited 2018 Jan 15];

657 Available from: http://www.ncbi.nlm.nih.gov/pubmed/29223362

658 23. Levy JMM, Towers CG, Thorburn A. Targeting autophagy in cancer. Nat Rev

659 Cancer [Internet]. 2017 [cited 2018 Jan 15];17:528–42. Available from:

Jin and Lee et al., page 28

660 http://www.ncbi.nlm.nih.gov/pubmed/28751651

661 24. Kang R, Tang D, Schapiro NE, Livesey KM, Farkas A, Loughran P, et al. The

662 receptor for advanced glycation end products (RAGE) sustains autophagy and

663 limits apoptosis, promoting pancreatic tumor cell survival. Cell Death Differ.

664 Nature Publishing Group; 2010;17:666–76.

665 25. Yang S, Wang X, Contino G, Liesa M, Sahin E, Ying H, et al. Pancreatic

666 cancers require autophagy for tumor growth. Genes Dev. Cold Spring Harbor

667 Laboratory Press; 2011;25:717–29.

668 26. Eke I, Deuse Y, Hehlgans S, Gurtner K, Krause M, Baumann M, et al.

669 β₁Integrin/FAK/cortactin signaling is essential for human head and neck cancer

670 resistance to radiotherapy. J Clin Invest [Internet]. 2012;122:1529–40.

671 Available from:

672 http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=3314473%7B&%7Dt

673 ool=pmcentrez%7B&%7Drendertype=abstract

674 27. Dickreuter E, Eke I, Krause M, Borgmann K, Van Vugt MA, Cordes N.

675 Targeting of β1 integrins impairs DNA repair for radiosensitization of head and

676 neck cancer cells. Oncogene. 2016;35.

677 28. Dickreuter E, Eke I, Krause M, Borgmann K, van Vugt MA, Cordes N. Targeting

678 of β1 integrins impairs DNA repair for radiosensitization of head and neck

679 cancer cells. Oncogene [Internet]. 2015 [cited 2015 Jun 16]; Available from:

680 http://www.ncbi.nlm.nih.gov/pubmed/26073085

681 29. Eke I, Zscheppang K, Dickreuter E, Hickmann L, Mazzeo E, Unger K, et al.

682 Simultaneous β1 integrin-EGFR targeting and radiosensitization of human

683 head and neck cancer. J Natl Cancer Inst. 2015;107.

684 30. Anaya J. OncoLnc: linking TCGA survival data to mRNAs, miRNAs, and

685 lncRNAs. PeerJ Comput Sci [Internet]. PeerJ Inc.; 2016 [cited 2018 Dec

Jin and Lee et al., page 29

686 20];2:e67. Available from: https://peerj.com/articles/cs-67

687 31. Degtyarev M, De Mazière A, Orr C, Lin J, Lee BB, Tien JY, et al. Akt inhibition

688 promotes autophagy and sensitizes PTEN-null tumors to lysosomotropic

689 agents. J Cell Biol [Internet]. 2008 [cited 2018 Nov 16];183:101–16. Available

690 from: http://www.ncbi.nlm.nih.gov/pubmed/18838554

691 32. Garber ME, Troyanskaya OG, Schluens K, Petersen S, Thaesler Z, Pacyna-

692 Gengelbach M, et al. Diversity of gene expression in adenocarcinoma of the

693 lung. Proc Natl Acad Sci U S A. National Academy of Sciences;

694 2001;98:13784–9.

695 33. Deng N, Goh LK, Wang H, Das K, Tao J, Tan IB, et al. A comprehensive

696 survey of genomic alterations in gastric cancer reveals systematic patterns of

697 molecular exclusivity and co-occurrence among distinct therapeutic targets.

698 Gut. BMJ Publishing Group; 2012;61:673–84.

699 34. Guerrero PA, Tchaicha JH, Chen Z, Morales JE, McCarty N, Wang Q, et al.

700 Glioblastoma stem cells exploit the αvβ8 integrin-TGFβ1 signaling axis to drive

701 tumor initiation and progression. Oncogene. Nature Publishing Group;

702 2017;36:6568–80.

703 35. Singh SS, Vats S, Chia AY-Q, Tan TZ, Deng S, Ong MS, et al. Dual role of

704 autophagy in hallmarks of cancer. Oncogene. Nature Publishing Group;

705 2018;37:1142–58.

706 36. Biancur DE, Kimmelman AC. The plasticity of pancreatic cancer metabolism in

707 tumor progression and therapeutic resistance. Biochim Biophys Acta - Rev

708 Cancer [Internet]. 2018 [cited 2018 Nov 15];1870:67–75. Available from:

709 http://www.ncbi.nlm.nih.gov/pubmed/29702208

710 37. Campos T, Ziehe J, Palma M, Escobar D, Tapia JC, Pincheira R, et al. Rheb

711 promotes cancer cell survival through p27Kip1-dependent activation of

Jin and Lee et al., page 30

712 autophagy. Mol Carcinog. Wiley-Blackwell; 2016;55:220–9.

713 38. Melser S, Chatelain EH, Lavie J, Mahfouf W, Jose C, Obre E, et al. Rheb

714 regulates mitophagy induced by mitochondrial energetic status. Cell Metab.

715 Elsevier; 2013;17:719–30.

716 39. Li G, Ma D, Chen Y. Cellular functions of programmed cell death 5. Biochim

717 Biophys Acta [Internet]. 2016 [cited 2018 Nov 15];1863:572–80. Available from:

718 https://linkinghub.elsevier.com/retrieve/pii/S0167488916000045

719 40. Ge D, Han L, Huang S, Peng N, Wang P, Jiang Z, et al. Identification of a novel

720 MTOR activator and discovery of a competing endogenous RNA regulating

721 autophagy in vascular endothelial cells. Autophagy [Internet]. 2014 [cited 2018

722 Nov 16];10:957–71. Available from:

723 http://www.tandfonline.com/doi/abs/10.4161/auto.28363

724 41. Zhu Y, Runwal G, Obrocki P, Rubinsztein DC. Autophagy in childhood

725 neurological disorders. Dev Med Child Neurol [Internet]. 2018 [cited 2018 Nov

726 15]; Available from: http://www.ncbi.nlm.nih.gov/pubmed/30417343

727

728

Jin and Lee et al., page 31

729 Legends to figure 1-7:

730

731 Figure 1. High-throughput esiRNA knockdown screen of 117 focal adhesion

732 proteins in 3D PDAC cell culture. (A) Workflow of esiRNA screen. (B-D) 3D

733 tumoroid formation in unirradiated (B) or 6 Gy x-rays irradiated cells (C) upon

734 knockdown (KD) of focal adhesion proteins. (D) Enhancement ratio of radiosensitivity

735 upon knockdown (tumoroid number at 6 Gy) / (tumoroid number at 0 Gy). Results for

736 β8 integrin depletion are shown in red. Black columns indicate non-specific esiRNA

737 control. All results show mean ± SD (n=3; two-sided t-test; *P<0.05; n.s., not

738 significant).

739

740 Figure 2. mRNA and protein expression of β8 integrin. (A) mRNA expression

741 level of β8 integrin in healthy and in PDAC pancreatic tissue using Oncomine

742 database (https://www.oncomine.org). Boxplot illustrate maximum, 75th percentile,

743 median, 25th percentile and the minimum; study by Badea et al.. (B) Correlation of β8

744 integrin gene expression and PDAC patient survival rate analyzed using OncoLnc.

745 (http://www.oncolnc.org). (C) Western blot analysis for β8 integrin expression from

746 whole cell lysates of tumoroids derived from 3D lrECM grown established PDAC cell

747 cultures and patient-derived primary cell cultures. (D) Densitometry analysis of “B”

748 normalized to β actin. All results show mean ± SD (n=3).

749

750 Figure 3. Subcellular localization of β8 integrin in PDAC cells and tissue. (A)

751 Immunofluorescence staining of β8 integrin (green), actin stained with

752 phalloidin/Alexa 594 (red), and nucleus with DAPI (blue). (B) Endogenous β8 integrin

753 detected using two different anti-β8 integrin antibodies for nearly membrane domain

754 (aa 614-663, in green) and the headpiece (aa 392-503, red) in BxPC3 and MiaPaCa2

Jin and Lee et al., page 32

755 cells visualized by Alexa 405 and Alexa 647, respectively. In this study, the antibody

756 detecting the aa 614-663 sequence was general used. (C) Immunofluorescence

757 staining of β8 integrin (green) in glioblastoma cells. Black arrows indicate membrane

758 expression of β8 integrin. (D and E) BxPC3 (D) and U-251 MG (E) cells were

759 transfected with β8-AcGFP plasmid and then co-stained with anti-β8 antibody. White

760 line represents plasma membrane; blue line represents nucleus based on BF images.

761 (F and G) H & E (F) and immunofluorescence staining of integrin β8 (G; plus DAPI)

762 in human resection specimen from PDAC patients. White boxes indicate the same

763 area. Scale bars, (E and F) 100 µm and zoomed areas 1 & 2 in (F) 10 μm.

764

765 Figure 4. β8 integrin is critical for PDAC cell sensitivity to ionizing radiation

766 and chemotherapy. (A and B) esiRNA-mediated β8 integrin depletion analysis by

767 Western blot (A) and densitometry (B). (C and D) 3D tumoroid forming ability in

768 unirradiated (C) and in 6 Gy x-rays irradiated samples upon β8 integrin silencing. (E)

769 Representative phase contrast images of PDAC tumoroids (C and D). (F) 3D

770 tumoroid forming ability upon gemcitabine treatment in BxPC3 and Patu8902 cells.

771 All results show mean ± SD (n=3; two-sided t-test; *P<0.05, **P<0.01, ***P<0.005).

772

773 Figure 5. β8 integrin interactome and subcellular co-localization. (A) Sequential

774 immunoprecipitation-mass spectrometry-based analysis of β8 integrin interactome

775 was performed in unirradiated and 6-Gy x-rays irradiated Patu8902 cells. Time point

776 of analysis after irradiation was 2 h. (B) Comparative changes in the β8 integrin

777 interactome between control and irradiation upon categorization into different

778 molecular functions using the PANTHER classification system. (C)

779 Immunofluorescence co-staining of β8 integrin with GM130 (Golgi), mitochondria, αV

Jin and Lee et al., page 33

780 integrin, APPL2 and Caveolin 1. Scale bars, 10 µm. (D) Pearson correlation analysis

781 of “C”. Data show mean ± SD (n=3).

782

783 Figure 6. Translocation of β8 integrin from perinucleus area towards the

784 plasma membrane upon stress. (A) Representative maximum intensity projection

785 of z-stacks show β8 integrin subcellular localization in 6-Gy x-rays irradiated BxPC3

786 cells using immunofluorescence staining. Signals of β8 integrin were analyzed in the

787 whole cell. (B) Image based expression analysis of radiogenic induction of β8 integrin

788 expression. (C and D) Both analyses of “A” for average (C) and maximum (D)

789 distance between β8 integrin-positive particles and the center of the nucleus. (E)

790 Representative maximum intensity projection of z-stacks show β8 integrin subcellular

791 localization in 6-Gy x-rays irradiated cells pretreated with colchicine. (F and G) Both

792 analyses of “E” for average (C) and maximum (D) distance between β8 integrin-

793 positive particles and the center of the nucleus. Scale bars, 10 µm. β8 integrin, green;

794 nucleus (stained with DAPI), gray. Results show mean ± SD (n=3; data were

795 collected from 20-40 cells from three independent experiments; two-sided t-test;

796 ***P<0.005; n.s., not significant)

797

798 Figure 7. β8 integrin interaction with autophagy proteins upon irradiation and

799 reduced LC3 expression under β8 integrin silencing. (A) Autophagy associated

800 proteins identified using the (http://arn.elte.hu/) increasingly interact with β8 integrin

801 upon irradiation. (B and C) Representative images (B) and analyses (C) of LC3

802 expression at 24 h after 6-Gy x-rays irradiation in presence of colchicine (5 nM),

803 paclitaxel (1 nM) or chloroquine (25 µM) in β8 integrin-depleted and control cells. (D

804 and E) 3D tumoroid formation upon β8 integrin depletion in BxPC cells in presence of

805 colchicine (5 nM), paclitaxel (1 nM) or chloroquine (25 µM) plus/minus 6 Gy x-rays.

Jin and Lee et al., page 34

806 Scale bars; 10 µm. LC3B, red; nucleus, gray. Results show mean ± SD (n=3; two-

807 sided t-test; *P<0.05, ***P<0.005; n.s., not significant).

808

809

Fig. 1 A 0 Gy

KIF11 C20orf42 CORO1B FHL2 FLNA GRB2 PXN LASP1 HAX1 PFN1 LDB3 SSH3BP LPP SYNM Tumoroid SORBS1 NCK2 CASS4 NF2 CRKL PARVA ZFYVE21 esiRNA N = 100 x n / n0, ITGA5 SH2B1 ITGA8 SORBS2 THY1 SLC3A2 ZYX Perform FBLIM1 ACTN1 FERMT3 CTTN GNB2L1 PARVB ITGA4 ITGAM TGFB1I1 ITGB1 TRIP6 SVIL CALR GRB7 ARPC2 JUB SHC1 LIMS2 SORBS3 ITGAL ITGB6 ITGA1 KTN1 ITGA2 NRP2 ITGB3BP MSN BCAR1 NEDD9 CRK PALLD VCL ITGB5 SDCBP ITGA7 SMPX ITGAD SDC4 PPFIA1 counting & analyzing RLUC PLEKHC1 CORO2A GAB1 KEAP1 ITGA3 RDX transfection 3D cell culture N, normalized average tumoroid formation in percentage TES ITGAX TNS1 ITGB3 KCNH2 NEXN VASP ITGB1BP1 SYNM LIMS1 IRS1 ITGAE TENC1 CD151 ITGB8 ITGA11 NRP1 CD47 PKD1 ACTB NDEL1 CAV1 OSTF1 VIL2 ITGB4 CEACAM1 ITGA6 SH3KBP1 ITGA9 PVR ENAH CSRP1 ITGAV TLN1 ITGB2 TRPM7 NUDT16L1 ENG n, tumoroid number upon KD of target proteins ITGB7 ITGA10 LRP1 PLEKHC1 LPXN CFL1 6 Gy n0, average tumoroid number in control sample FAPs esiRNA library Incubation 24 h

Incubation 24 h Tumoroid formation

B Surviving upon KD of target proteins + 0 Gy (normalized on control)

n.s. * ** ***

100 formation / % 0

Avg. tumoroid Avg.

LPP NF2 JUB

TES ZYX VCL

PVR VIL2 PXN

CRK RDX

SVIL ENG

MSN IRS1

CFL1 TLN1 FHL2

LRP1 LDB3

THY1 FLNA TNS1 CD47 KTN1 PFN1

PKD1 LPXN HAX1 CAV1

SDC4 KIF11 NRP2 CRKL VASP ACTB GAB1 SHC1 CTTN NRP1 CALR NCK2

GRB7 GRB2 RLUC

ENAH NEXN

TRIP6 SMPX

ITGA7 ITGA1 ITGA2 ITGB3 ITGA4 ITGA3 LIMS2 ITGB5 ITGA5 ITGA9 ITGB1 ITGB2 ITGA6 ITGB8 ITGB7 ITGB6 ITGA8 ITGB4 SYNM SYNM LIMS1 ITGAL

ITGAE ITGAX ITGAV

ITGAD

LASP1 CD151

SH2B1 ITGAM PALLD

KEAP1 OSTF1 NDEL1

ACTN1 TENC1 CASS4

ARPC2 BCAR1 CSRP1

PARVA PARVB NEDD9 KCNH2 TRPM7

ITGA10 ITGA11 PPFIA1 SDCBP

FBLIM1

SLC3A2

GNB2L1

SSH3BP

FERMT3

SORBS2 TGFB1I1 SORBS3 SORBS1

C20orf42

CORO2A CORO1B ITGB3BP

ZFYVE21

SH3KBP1

PLEKHC1 PLEKHC1

ITGB1BP1

CEACAM1 NUDT16L1

C Surviving upon KD of target proteins + 6 Gy (normalized on control)

n.s. * ** ***

100 formation / % 0

Avg. tumoroid Avg.

NF2 JUB LPP

TES ZYX VCL

PVR PXN

VIL2

RDX CRK

SVIL ENG

IRS1 MSN

TLN1 CFL1 FHL2

LRP1 LDB3

TNS1 FLNA KTN1 PFN1 THY1

CD47

PKD1 LPXN CAV1 HAX1

SDC4 CTTN KIF11 NRP1 CALR CRKL NRP2 SHC1 GAB1 VASP ACTB NCK2

RLUC GRB7 GRB2

NEXN ENAH

SMPX TRIP6

ITGB7 ITGA7 ITGB1 ITGA8 ITGA6 ITGA3 ITGB3 ITGB8 ITGB4 SYNM ITGAL ITGA5 ITGA9 ITGA1 ITGA2 ITGB2 ITGB6 ITGA4 SYNM LIMS2 ITGB5 LIMS1

ITGAX ITGAE ITGAV

ITGAD

LASP1 CD151

ITGAM

PALLD SH2B1

KEAP1 NDEL1 OSTF1

TENC1 ACTN1 CASS4

CSRP1 BCAR1 ARPC2

TRPM7 KCNH2 PARVB PARVA NEDD9

ITGA10 ITGA11 SDCBP PPFIA1

FBLIM1

SLC3A2

GNB2L1

SSH3BP

FERMT3

SORBS2 SORBS1 TGFB1I1 SORBS3

C20orf42

CORO1B CORO2A ITGB3BP

ZFYVE21

SH3KBP1

PLEKHC1 PLEKHC1

ITGB1BP1

CEACAM1 NUDT16L1

D Enhancement ratio of radiosensitivity by KD (1 = no effect; >1 = radioprotection; <1 = radiosensitization) 2

1 0

6 Gy / untreated

NF2 JUB LPP

TES ZYX VCL

PXN PVR VIL2

CRK RDX

ENG SVIL

MSN

IRS1

CFL1 TLN1 FHL2

LDB3 LRP1

THY1 PFN1 KTN1 TNS1 FLNA

CD47

CAV1 LPXN PKD1 HAX1

SDC4 VASP SHC1 GAB1 KIF11 ACTB CALR CRKL CTTN NCK2 NRP2 NRP1

GRB2 GRB7 RLUC

NEXN ENAH

SMPX TRIP6

ITGAL SYNM ITGB2 ITGB7 ITGA3 ITGA4 LIMS2 ITGA8 ITGB6 ITGB5 ITGB4 ITGA1 ITGA9 SYNM ITGA7 ITGA2 ITGA5 ITGB1 ITGB3 ITGA6 LIMS1 ITGB8

ITGAV ITGAE ITGAX

ITGAD

LASP1 CD151

SH2B1 PALLD ITGAM

OSTF1 NDEL1 KEAP1

TENC1 CASS4 ACTN1

ARPC2 BCAR1 CSRP1

PARVB KCNH2 TRPM7 NEDD9 PARVA

SDCBP ITGA10 PPFIA1 ITGA11

FBLIM1

SLC3A2

GNB2L1

SSH3BP

FERMT3

SORBS1 SORBS2 TGFB1I1 SORBS3

C20orf42

ITGB3BP CORO1B CORO2A

ZFYVE21

SH3KBP1

PLEKHC1 PLEKHC1

ITGB1BP1

CEACAM1 NUDT16L1 Downloaded from mcr.aacrjournals.org on September 23, 2021. © 2019 American Association for Cancer Research. Author Manuscript Published OnlineFirst on July 23, 2019; DOI: 10.1158/1541-7786.MCR-18-1352 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Fig. 2

A β8 integrin B mRNA expression β8 integrin p = 4.8E-8 low high (Fold change: 3.1) 100

4 P=0.04; n=52 50 0 PDAC (n=39)

Overall survival / % 0 Log2 median- 0 1000 2000 3000

centered intensity -4 n.c. Days (n=39)

C D Patient–derived PDAC 1

0.5 of actin / AU of actin / rel. integrin β 8 COLO357 Capan–1 MiaPaca 2 PaCaDD137 PaCaDD159 PaCaDD119 BxPC3 Patu8902 Panc 1 β8 integrin 0

β actin Panc 1 BxPC3 Capan–1 Patu8902 COLO357 MiaPaca 2 PaCaDD119 PaCaDD137 PaCaDD159

Fig. 3

A PDAC cell lines C Glioblastoma cell lines β8 integrin DAPI Phalloidin Overlay β8 integrin BF Overlay BxPC3 U-251 MG DD-T4 Capan–1

D β8 / AcGFP anti-β8 Overlay BF COLO357 BxPC3

β8 / AcGFP1 vs. anti-β8 of whole cell area: Pearson's R = 0.6 MiaPaCa2 E β8 / AcGFP anti-β8 Overlay BF Panc1 U-251 MG β8 / AcGFP1 vs. anti-β8 of whole cell area: Pearson's R = 0.8

Patu8902 F PDAC patient biopsy – H & E PaCaDD119 PaCaDD137 PaCaDD159 G PDAC patient biopsy – IF

B anti-β8 Area1 Antibody 1 Antibody 2 BF Pearson's R (aa 614-663) (aa 392-503) (Ab.1 vs. Ab. 2):

0.9 BxPC3 Area2

0.64 MiaPaCa2

β8 integrin Nucleus

Fig. 4

A B β8 integrin β actin esi RNA control esi integrin β8:- + - + esi β8 integrin

BxPC3 1 Capan–1 COLO357 MiaPaca 2 0.5 Panc 1

Patu8902 Fold change PaCaDD119 0 PaCaDD137

BxPC3 Panc 1 Capan–1 Patu8902 COLO357MiaPaca 2 PaCaDD119PaCaDD137

C Unirradiated / D 6 Gy x-rays / plating efficiency surviving fraction esi RNA control esi RNA control esi β8 integrin esi β8 integrin 20 50 ** *** * ** *** *** * **

10 25 / % / % Tumoroid formation Tumoroid Tumoroid formation Tumoroid 0 0

BxPC3 Panc 1 BxPC3 Panc 1 Capan–1 Capan–1 Patu8902 Patu8902 COLO357MiaPaca 2 COLO357MiaPaca 2 PaCaDD119PaCaDD137 PaCaDD119PaCaDD137

E Unirradiated 6 Gy x-rays

esi RNA control esi β8 integrin esi RNA control esi β8 integrin

F 100 BxPC3 100 Patu8902

50 50 formation / % Rel. tumoroid

0 0 1 10 100 1000 10000 1 10 100 1000 10000 Gemcitabine / nM Gemcitabine / nM

esi β8 integrin esi RNA control

A β8 integrin interactome Fig. 5 6 Gy x-rays

Unirradiated

n Number of proteins: n = 101 =32 n = 605

B β8 integrin involved in molecular functions

Transporter activity 0 Gy Unirradiated & 6 Gy x-rays 6 Gy x-rays Antioxidant activity Catalytic activity Signal transducer activity Structural molecule activity Receptor activity Binding Translation regulator activity 0 50 100 150 200 Number of proteins C

GM130 Mitochondria Integrin αV APPL2 Caveolin 1 β8 integrin Nucleus β8 integrin Nucleus β8 integrin Nucleus β8 integrin Nucleus β8 integrin Nucleus

D 0.75 elation 0.5

son corr 0.25

Pear 0 GM130 Mitochondria Integrin αV APPL2 Caveolin 1

Fig. 6 β8 integrin β8 integrin + nucleus *** A B *** 8

4 n.c.

per cell 2 Rel. β 8 level 1

0 n.c. 2 h p. IR 24 h p. IR C *** 2 h p. IR 2

1 Norm. avg. distance 0 24 h p. IR n.c. 2 h p. IR 24 h p. IR D *** E β8 integrin β8 integrin + nucleus 2

1 n.c. Norm. max. distance 0 n.c. 2 h p. IR 24 h p. IR F n.s. ne 1 Colchici

. avg. distance 0.5 Norm

0 ne n.c. Colchicine Colchicine G 2 h p. IR 24 h p. IR 2 h p. IR

Colchici n.s. 1

ne 0.5 . max. distance Norm 24 h p. IR Colchici 0 n.c. Colchicine Colchicine 2 h p. IR 24 h p. IR

Downloaded from mcr.aacrjournals.org on September 23, 2021. © 2019 American Association for Cancer Research. E D PRKAR1A A ALDH9A1 PRKAG1 DNAJC7 FKBP1A DNAJB1 HADHA SMC1A MAPK8 Tumoroid formation HACD3 Tumoroid formation SAFB2 RBMX TMPO DIP2B MYH4 RHEB ETFA / % / % RPA2 10 50 25 20 0 0 Unirr 6 Gyx-rays 3 n.c. esi integr esi RNA esi integr esi RNA n.c. *** adiated Unirradiated 1.5 n.s. Col Intensity / AU control control Col in β8 in β8 n.s. Downloaded from Author manuscriptshavebeenpeerreviewedandacceptedforpublicationbutnotyetedited. n.s. 0 Author ManuscriptPublishedOnlineFirstonJuly23,2019;DOI:10.1158/1541-7786.MCR-18-1352 * Pac 6 Gyx-rays Pac 1.5 Chl Chl 3 mcr.aacrjournals.org B Norm. avg. C Flu. I. of LC3 Chloroquin (Chl) Paclitaxel (Pac) Colchicine (Col) n.c. 0 2 1 LC3 + DAPI Integrin β8 LC3 + DAPI Integrin β8 LC3 + DAPI Integrin β8 LC3 + DAPI Integrin β8 esi RNA control esi RNA n.c. Col control Unirradiated Unirr

Chl adiated esi integrin β8 esi integrin esi integr

Research. n.c. Col

Pac in β8 Chl esi RNA control esi RNA n.c. Col control 6 Gyx-rays Pac 6 Gyx-rays Chl esi integrin β8 esi integrin n.c. esi integr

Col Fig. 7

Pac in β8 Chl Author Manuscript Published OnlineFirst on July 23, 2019; DOI: 10.1158/1541-7786.MCR-18-1352 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

β8 integrin mediates pancreatic cancer cell radiochemoresistance

Sha Jin, Wei-Chun Lee, Daniela Aust, et al.

Mol Cancer Res Published OnlineFirst July 23, 2019.

Updated version Access the most recent version of this article at: doi:10.1158/1541-7786.MCR-18-1352

Supplementary Access the most recent supplemental material at: Material http://mcr.aacrjournals.org/content/suppl/2019/07/23/1541-7786.MCR-18-1352.DC1

Author Author manuscripts have been peer reviewed and accepted for publication but have not yet Manuscript been edited.

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mcr.aacrjournals.org/content/early/2019/07/23/1541-7786.MCR-18-1352. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.