cd155/cd96 promotes immunosuppression in lung ... › content › 10.1101 › 688812v1.full.pdf ·...

1

CD155/CD96 promotes immunosuppression in lung adenocarcinoma 1

(LUAD) 2

Weiling He2, 3†, Hui Zhang1, 2†, Shuhua Li1†,Yongmei Cui1 † , Ying Zhu4, 3

Junfeng Zhu1, Yiyan Lei5, Run Lin6, Di Xu7, Zheng Zhu8, Wenting Jiang1, Han 4

Wang1, Zunfu Ke1, 2* 5

1Department of Pathology, The First Affiliated Hospital, Sun Yat-sen 6

University, Guangzhou 510080, China 7

2Institute of Precision Medicine, The First Affiliated Hospital, Sun Yat-sen 8


3Department of Gastrointestinal Surgery, The First Affiliated Hospital, Sun 10

Yat-sen University, Guangzhou 510080, China 11

4Department of Radiology, The First Affiliated Hospital, Sun Yat-sen 12


5Department of Thoracic Surgery, The First Affiliated Hospital, Sun Yat-sen 14


6Department of Radiology, The First Affiliated Hospital, Sun Yat-sen 16


7Department of Thoracic Surgery, The Central Hospital of Wuhan, Wuhan 18

430014, China 19

8 Department of Pathology, Longgang Central Hospital of Shenzhen, Affiliated 20

Longgang Hospital of Zunyi Medical University, Shenzhen 518116, Chin 21

22

Running Title: CD155-dependent immune suppression in LUAD. 23

24

†These authors contributed equally to this work. 25

certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was notthis version posted July 1, 2019. ; https://doi.org/10.1101/688812doi: bioRxiv preprint

https://doi.org/10.1101/688812

2

* Corresponding author: 26

Zunfu Ke 27

1Department of Pathology, First Affiliated Hospital, Sun Yat-sen University, 28

No. 58, ZhongShan Second Road, Guangzhou 510080, China 29

2Institute of Precision Medicine, The First Affiliated Hospital, Sun Yat-sen 30


Tel: 86-20-87331780; 32

Fax: 86-20-87331780; 33

E-mail: [email protected] 34

35

36 37

38

39

40

41

42

43

44

45

46

47

48

49


https://doi.org/10.1101/688812

3

Abstract 50

Lung adenocarcinoma (LUAD) remains one of the leading causes of death in 51

patients with cancer. The association of CD155 with CD96 transmits an 52

inhibitory signal and suppresses antitumor immune response. This study 53

investigates the effect of CD155/CD96 on immune suppression in LUAD. We 54

demonstrate that LUAD patients with high CD155 expression suffer from 55

immune suppression and experience a poor prognosis, which coincides with 56

an inhibited AKT-mTOR signaling pathway in CD8 T cells and subsequently 57

up-regulated CD96 expression. Moreover, the inhibition effect can be 58

reversed by CD96 blocking antibody. High CD155 expression inhibited the 59

release of IFNγ from CD8 cells. Moreover, Blocking CD96 restored IFNγ 60

production in CD8 T cells and neutralized the inhibition of IFNγ production in 61

CD8 T cells mediated by CD155. Animal experiments showed that CD155-62

mediated LUAD growth might depend on its suppression antitumor immune 63

response in the tumor microenvironment in PDX mice. In conclusion, our 64

results suggest that LUAD cells suppress antitumor immune response in the 65

tumor microenvironment through CD155/CD96. CD155/CD96 could be a 66

potential therapeutic target for LUAD patients. 67

68

Key words: lung adenocarcinoma, CD155,CD96, immune suppression, 69

tumor microenvironment 70

71

Abbreviations: 72

LUAD: lung adenocarcinoma; IFNγ: interferon gamma; PDX: patient-derived 73

xenograft; NSCLC: non-small cell lung cancer; PRR: poliovirus receptor–74


https://doi.org/10.1101/688812

4

related; MDSCs: myeloid-derived suppressor cells; PRR: poliovirus receptor–75

related; STR: short tandem repeat; IRS: immunoreactive score; SI: staining 76

intensity; PP: percentage of positive cells; RT-PCR: reverse transcription-77

polymerase chain reaction; PBS: phosphate-buffered saline; PBMCs: 78

peripheral blood mononuclear cells; SDS–PAGE: sodium dodecyl sulfate-79

polyacrylamide gel electrophoresis; rCD155: recombinant human CD155; 80

LUAD cells: lung adenocarcinoma cells; TILs: tumor-infiltrating lymphocytes; 81

GzmB: granzyme B; IL-2 (Interleukin-2); TNF-α：tumor necrosis factor-alpha; 82

PI: propidium Iodide; PDX: patient-derived xenograft; TIGIT: T cell 83

immunoreceptor with Igand ITIM domains; WBC: white blood cells; MFI: mean 84

fluorescence intensity; HPF: high power field 85

86

Introduction 87

Lung cancer is the leading cause of death among cancer patients. 88

More than 1 million deaths are related to lung cancer annually (Ding et al, 89

2008), and approximately 1.2 million new lung cancer cases are diagnosed 90

each year (Jemal et al, 2011). Lung adenocarcinoma (LUAD) is the most 91

common type of non-small cell lung cancer (NSCLC). Surgery remains the 92

first-line treatment and the most successful option in located disease of LUAD 93

(Molina et al, 2008). Despite a decline in the cancer death rate over the past 94

two decades, LUAD has a 5-year survival rate of 10-15% in stage IV due to its 95

late-stage diagnosis and lack of effective therapeutic options (Imielinski et al, 96

2012). 97

Immune escape represents the failure of the immune system in 98

preventing carcinogenesis (Kim et al, 2007; Swann & Smyth, 2007). The 99


https://doi.org/10.1101/688812

5

tumor microenvironment is composed of regulatory T cells, tumor-associated 100

macrophages and myeloid-derived suppressor cells (MDSCs) (Noy & Pollard, 101

2014; Sato et al, 2005; Talmadge & Gabrilovich, 2013), which are the 102

signature of chronic inflammation and immune suppression in the tumor milieu 103

(Crespo et al, 2013). Recently, the development of cancer immunotherapy 104

that uses tumor-targeted monoclonal antibodies has achieved broad 105

therapeutic efficacy (Brahmer et al, 2012). The application of monoclonal 106

antibody against PD-1/PD-L1 was associated with longer progression-free 107

and overall survival and fewer treatment-related adverse events than was 108

platinum-based combination chemotherapy in patients with previously 109

untreated advanced NSCLC (Herbst et al, 2016; Reck et al, 2016). Given the 110

only 20-30% response rate of lung cancer when targeting PD-1/PD-L1, we 111

speculated that other molecules or mechanisms might be involved in limiting 112

the application of current immunotherapies for lung cancer patients. 113

CD155, a member of poliovirus receptor–related (PRR) family, was 114

initially identified as a receptor for poliovirus in humans (Mendelsohn et al, 115

1989). The CD155 transcript is ubiquitously expressed in humans and mice 116

(Koike et al, 1990; Mendelsohn et al, 1989). Recently, CD155 has been 117

identified as the ligand of the Ig-like receptors Tactile (CD96) and DNAM-1 118

(CD226) on T cells. The interaction of CD155 with CD96 transmits an 119

inhibitory signal. The interaction of CD155 with CD226 enhances the immune 120

response (Bottino et al, 2003; Shibuya et al, 1996; Yu et al, 2009). Expression 121

of CD155 has been shown to be elevated in many types of tumors (Brooks et 122

al, 2017; Iguchi-Manaka et al, 2016) and is involved with immune suppression 123

in melanoma (Inozume et al, 2016). Recently, CD96-/- mice displayed hyper-124


https://doi.org/10.1101/688812

6

inflammatory responses, resulting in resistance to carcinogenesis and 125

experimental lung metastases (Chan et al, 2014). Blocking CD96 can improve 126

tumor control in mice (Blake et al, 2016). However, how CD155/CD96 is 127

involved in immune response in the tumor microenvironment of LUAD 128

remains unknown. 129

In the present study, we found that CD155 expression was increased in 130

tumor tissue and was associated with immune suppression in the tumor 131

microenvironment of LUAD. Blocking CD155/CD96 interaction restores CD8 T 132

cell effector functions by reversing CD8 T cell exhaustion, suggesting a 133

possible therapeutic role of CD155/CD96 in fighting LUAD. 134

135

Results 136

CD155 expression was associated with immune suppression in the 137

tumor microenvironment of LUAD 138

To investigate CD155 expression characteristics in LUAD tissues, we 139

first performed western blotting to compare its expression between tumor and 140

para-tumor lung tissues. We found that CD155 expression was substantially 141

stronger in LUAD tissues than that in para-tumor normal lung tissues (Figure 142

1A). The increased CD155 expression in LUAD tissues was confirmed by IHC 143

(Figure 1B). Lung adenocarcinoma cells (LUADCs) were isolated from tumor 144

tissues in 6 CD155high and 6 CD155low LUAD patients using NanoVelcro as 145

described previously (Lin et al, 2014) (Figure S1). The primary LUADCs 146

isolated from tumor tissues were also CD155 positive as measured by 147

immunofluorescence (Figure 1C). Further, CD155 high expression in tumor 148

tissues predicted a poor prognosis in LUAD (Figure S2). CD155, as the 149


https://doi.org/10.1101/688812

7

molecule of a co-inhibitory pathway, suppresses the anti-melanoma immune 150

response (Inozume et al, 2016). We speculate that the low survival rate in 151

CD155high patients might be due to the immune suppression in the tumor 152

microenvironment mediated by CD155. 153

To further explore the involvement of CD155 in the antitumor immune 154

response in LUAD, we first studied the association of CD155 with the immune 155

response in the tumor microenvironment of LUAD. CD155high and CD155low 156

were identified according to its expression intensity as described in the 157

methods section (Figure 1D). We found that the transcripts of inhibitory 158

molecules, such as CD96, TIGIT, Pdcd1 and Lag-3, were substantially higher 159

in CD155high patients, while CD226 was lower in CD155high patients (Figure 160

1E and S3). In contrast, gene expression of T cell effector function-associated 161

molecules was essentially lower in 13 CD155high patients compared with that 162

in 11 CD155low patients (Figure 1E-F). Thus, high expression of CD155 in 163

LUAD might be associated with immune suppression in tumor 164

microenvironment. 165

166

LUADCs impaired the effector functions of CD8 T cells through direct 167

cell-cell contact 168

Tumor-infiltrating lymphocytes (TILs) predict a better prognosis in 169

patients with colorectal cancer (Pages et al, 2005). Here, we showed that a 170

high density of CD8 T cell infiltration was associated with better survival in 171

patients with LUAD (Figure S4), which was in accordance with a previous 172

report (Kawai et al, 2008). CD8 T cells are the effector cells in antitumor 173

immune response. However, TILs are functionally exhausted in the tumor 174


https://doi.org/10.1101/688812

8

microenvironment, which has not fully been understood yet. To elucidate how 175

cancer cells influence CD8 T cells in the tumor microenvironment, we first 176

performed a cell-cell contact co-culture of CD8 T cells (from the same 177

CD155high LUAD patient) and LUADCs (from 6 independent CD155high 178

patients) in vitro (Figure 2A). We found that GzmB (granzyme B) and Perforin 179

expression on CD8 T cells was inhibited, as measured by flow cytometry, 180

when co-cultured with LUADCs (Figure 2B-E). Further, production of IL-2 181

(Interleukin-2), TNFα (tumor necrosis factor-alpha) and IFNγ cytokines in CD8 182

T cells was also suppressed (Figure 2F-I). However, cytokine production was 183

not affected when we separately cultured the T cells (from the same 184

CD155high LUAD patient) and LUADCs (from 6 independent CD155high 185

patients) using a cell culture insert with a 0.4-μm pore size (Figure 2J-K). 186

Thus, cancer cells impaired CD8 T cell effector function through cell-cell 187

contact, which might contribute to immune suppression in the tumor 188

microenvironment of LUAD. 189

190

LUADCs suppressed CD8 T cell function through CD155 191

CD155 expression in the melanoma correlated with immune suppression 192

in the tumor microenvironment [21]. To confirm the hypothesis that CD155 193

might be involved in T cell inhibition mediated by LUADCs, we first 194

determined CD155 expression in LUADC1-6 by western blotting. Compared 195

with CD155-low level in BEAS-2B cells, LUADC1-6 cells appear CD155-high 196

(Figure 3A). CD155 expression in LUADCs was further confirmed by flow 197

cytometry (Figure 3B). In a T cell-cancer cell co-culture system (Figure 3C), 198

LUADCs decreased the expression of p-AKT and p-mTOR in CD8 T cells, as 199


https://doi.org/10.1101/688812

9

measured by flow cytometry. The phosphorylation of S6K and 4EBP1 in CD8 200

T cells was also decreased by cancer cells (Figure 3D). Downregulation of 201

CD155 in LUADCs by RNAi abolished the inhibition on CD8 T cells. The 202

suppression of AKT, mTOR, S6K and 4EBP1 in CD8 T cells was reversed by 203

knocking down CD155 in LUADCs (Figure 3D). These data demonstrated that 204

LUADCs suppress CD8 T cell effector function, which could lead to poor 205

antitumor immune response. Further, IFNγ production in CD8 T cells was 206

decreased when co-cultured with LUADCs. Knocking down CD155 in cancer 207

cells could reverse the inhibition (Figure 3E-F). Additionally, CD155 208

upregulation in LUADCs further suppressed IFNγ production in CD8 T cells 209

compared with that in LUADC-vector cells (Figure 3G-H). In summary, CD155 210

mediated LUADCs’ suppression on CD8 T cell function. 211

212

CD155-independent tumor growth in LUADCs 213

CD155 was stably downregulated in LUADCs (from 6 independent 214

CD155high patients) or overexpressed in LUADCs (from 6 independent 215

CD155low patients). Knockdown or overexpression was confirmed by western 216

blotting and flow cytometry (Figure S5). We found that neither knockdown nor 217

overexpression of CD155 affected cancer cell proliferation, as measured by 218

cell number (Figure S6A-B). We used PI (Propidium Iodide) to investigate the 219

cell cycle by flow cytometry and observed that CD155 knockdown did not 220

change the cell cycle distribution of cancer cells. Further, CD155 221

overexpression showed no effects on the cell cycle (Figure S6C-E). To study 222

whether tumor growth was affected by CD155 in vivo, we utilized non-invasive 223

imaging. LUADCs were stably transfected with luciferase and inoculated into 224


https://doi.org/10.1101/688812

10

NOG mice subcutaneously. The data showed no difference in tumor growth 225

between mice that received LUADCs-vector or LUADCs-CD155 cells (Figure 226

S6F-G). 227

228

CD155 promotes tumor growth in LUAD tumor-bearing mice by 229

impairing the antitumor immune response 230

The tumor microenvironment is infiltrated with tumor-associated 231

macrophages, myeloid-derived suppressor cells and regulatory T cells 232

(Hanahan & Weinberg, 2011). The stromal cells also play critical roles in 233

shaping the tumor microenvironment (Hanahan & Coussens, 2012). To better 234

duplicate the tumor microenvironment similar to LUAD patients, we used a 235

PDX mouse model as described in the method section to study the immune 236

response in the tumor microenvironment. Previous study shows that 237

CD155/96 is important for NK cells in antitumor immune response. In this 238

mouse model, the CD56+ NK cells were sorted out from the injected PBMC 239

(from 6 independent CD155low patients), excluding the antitumor effects by NK 240

cells. To investigate whether CD155/CD96 affects immune reaction and tumor 241

growth, PDX mice were treated with rCD155 or vehicle (Figure 4A). We first 242

confirmed the binding of rCD155 to CD96 and that decreased IFNγ production 243

in CD8 T cells (Figure S7). The infiltration of CD8 T cells in the tumor 244

microenvironment was decreased to some extent by rCD155 treatment as 245

measured by IHC (Figure 4B-C). To measure the effector function of the 246

tumor infiltrated CD8 T cells, CD8 T cells were isolated from the tumor tissue 247

and measured IFNγ production by flow cytometry. rCD155 treatment 248

decreased IFNγ production in CD8 T cells isolated from tumor tissue (Figure 249


https://doi.org/10.1101/688812

11

4D-E). The IFNγ transcript was also decreased by rCD155 treatment as 250

measured by RT-PCR (Figure 4F). To further understand the functional status 251

of CD8 T cells in the tumor microenvironment, we measured TNF-α, GzmB 252

and Perforin expression in the tumor microenvironment by RT-PCR. We found 253

that rCD155 treatment decreased TNF-α, GzmB and Perforin expression in 254

the tumor microenvironment (Figure 4G). Importantly, rCD155 treatment 255

promoted tumor growth in the PDX mice (Figure 4H). 256

To further investigate the function of CD155 on immune response in 257

the tumor microenvironment of LUAD, CD155 was stably overexpressed in 258

LUADC cells (from 6 independent CD155low patients). Cells were inoculated 259

into NOG mice subcutaneously. The mice were subsequently reconstituted 260

with or without human PBMCs. We found that tumors in NOG mice that 261

lacked human PBMC reconstitution grew substantially faster than those in 262

mice carrying a human immune system (Figure S8A-B). Further, in mice 263

reconstituted with human PBMCs, tumor growth was substantially faster in 264

mice that received LUADC-CD155 cells than in those that received LUADC-265

vector cells (Figure S8C-D). Moreover, mice that received LUADC-CD155 266

cells showed a poorer prognosis than those that received LUADC-vector cells 267

(Figure S8E). 268

Therefore, these results demonstrated that CD155-mediated LUAD 269

growth might depend on its suppression antitumor immune response in the 270

tumor microenvironment. 271

272

Dysregulation of CD96/CD226 identified reduced CD8 T cell effector 273

functions in LUAD. 274


https://doi.org/10.1101/688812

12

CD96, as a co-inhibitory molecule, competes with the co-stimulatory 275

molecule of CD226 for binding to CD155. The lost balance between these two 276

molecules might lead to a hypo- or hyper-immune response. We found that 277

CD8 T cells from LUAD patients expressed higher levels of CD96 compared 278

with healthy controls (HC). CD96 expression was even higher in TILs than in 279

T cells from the circulation (Figure 5A-B). CD226 expression was substantially 280

lower in T cells from LUAD patients than that in HC. CD226 expression in 281

TILs was further decreased compared with that from the circulation (Figure 282

5C-D). The phosphorylation of AKT and mTOR was lower in CD96+ CD8 T 283

cells from LUAD patients than that in CD96- CD8 T cells. The phosphorylation 284

of mTOR downstream molecules also decreased in CD96+ cells compared 285

with that in CD96- cells, as measured by flow cytometry and western blotting 286

(Figure 5E). To compare the effector functions of CD96+ and CD96- CD8 T 287

cells, intracellular cytokine production was measured by flow cytometry. IFNγ 288

and TNFα production in CD96+ and CD96- T cells was confirmed by flow 289

cytometry (Figure 5F-G). CD96 expression in CD8 T cells from LUAD patients 290

was associated with decreased T cell effector functions. The increased CD96 291

expression in CD8 T cells from LUAD patients might be closely associated 292

with the poor immune response in tumor microenvironment. 293

294

LUAD cells suppress effector functions of CD8 T cell through 295

CD155/CD96 296

The tumor microenvironment is infiltrated with exhausted T cells 297

(Crespo et al, 2013; Jiang et al, 2015). CD96, a recently identified co-298

inhibitory molecule that competes with CD226 for the ligand CD155, was 299


https://doi.org/10.1101/688812

13

increased on CD8 T cells in the LUAD tumor microenvironment (Chan et al, 300

2014). To determine how CD96 expression on CD8 T cells is regulated in the 301

tumor microenvironment by cancer cells, we performed cell-cell contact T cell-302

cancer cell co-culture as described above and found that CD96 expression on 303

CD8 T cells was increased when co-cultured with CD155high cancer cells 304

(from 6 independent CD155high patients) (Figure 6A), whereas the co-305

stimulatory receptor CD226 was inhibited by CD155high cancer cells (Figure 306

6B). As shown in Figure 3D, AKT-mTOR pathway in CD8 T cells was inhibited 307

by culturing with CD155high cancer cells. We found that blocking CD96 can 308

rescue the inhibition. The phosphorylation of AKT-mTOR was increased by 309

blocking CD96 as measured by western blot (Figure 6C). Also, the 310

phosphorylation of 6SK and 4EBP1, which are downstream molecules of 311

mTOR, was also increased in CD8 T cells when CD96 was blocked, as 312

measured by flow cytometry (Figure 6D). Blocking CD96 in the T cell-cancer 313

cell co-culture system increased IFNγ production in CD8 T cells (Figure 6E). 314

IFNγ production in CD8 T cells was decreased when co-cultured with 315

CD155high LUAD cells, which was further decreased when CD155 was 316

overexpressed in CD155high LUAD cells (Figure 6F). However, CD96 317

blockade could neutralize the inhibition mediated by CD155 expression 318

(Figure 6F). Thus, LUAD cells suppress the T cell response through 319

CD155/CD96 signaling in the tumor microenvironment. 320

321

Discussion 322

In the tumor microenvironment, the upregulation of negative signals on 323

T-cell responses suppresses the antitumor immune response and promotes 324


https://doi.org/10.1101/688812

14

tumor progression (Zou & Chen, 2008). In this study, we showed that CD155 325

expression was increased in LUAD tumor tissue, and LUAD cells suppressed 326

the immune response in the tumor microenvironment through CD155/CD96. 327

However, CD155 itself showed no effects on cancer cell proliferation in vitro 328

or tumor growth in vivo. 329

CD8 T cells are the effector cells in the antitumor immune response in 330

most tumor models (Rosenberg et al, 2004). However, T cells that have 331

infiltrated into the tumor microenvironment are exhausted and functionally 332

impaired (Ahmadzadeh et al, 2009; Jiang et al, 2015). In the current study, we 333

found that LUAD cells impaired the effector functions of CD8 T cells through 334

cell-cell contact, which explained the immune suppression in the tumor 335

microenvironment. In addition to tumor-associated macrophages, MDSC and 336

regulatory T cells, cancer cells can impair CD8 T cell functions directly and 337

escape immune attack. 338

CD155 functions as a cell adhesion molecule that enhances glioma cell 339

migration (Sloan et al, 2005), and its expression is increased at the RNA and 340

protein levels in tumor tissues (Chandramohan et al, 2017; Masson et al, 341

2001). The association of CD155 to TIGIT (T cell immunoreceptor with Ig and 342

ITIM domains) transmits a negative signal to T cells and suppresses immune 343

response (He et al, 2017). In the current study, we found that CD155 344

expression was significantly increased in LUAD tissue. The primary LUAD 345

cells isolated from patients with LUAD expressed high levels of CD155, which 346

correlated with immune suppression in the tumor microenvironment. The 347

CD155-associated hypo-immune response in the tumor microenvironment of 348

LUAD could lead to immune invasion and promote tumor growth. In addition, 349


https://doi.org/10.1101/688812

15

CD155high patients showed poorer survival than CD155low patients, which was 350

in accordance with a previous report (Atsumi et al, 2013). Interestingly, we 351

found that CD155 expression did not affect cancer cell proliferation in vitro or 352

tumor growth in vivo. Further, tumor growth was greater in NOG mice without 353

a human immune system than in mice with a reconstituted human immune 354

system. CD155 overexpression in cancer cells resulted in greater tumor 355

growth and poorer survival in tumor-bearing NOG mice reconstituted with 356

human immune system. CD155 promoted tumor growth in mice reconstituted 357

with the human system through limiting the immune response in the tumor 358

microenvironment. These data further suggest that CD155 mediates LUAD 359

progression by manipulating the immune system, which is independent of 360

cancer cell proliferation. 361

To better understand how CD155 regulates the immune response 362

mediated by CD8 T cell in the tumor microenvironment, we established a PDX 363

model using patient-derived tumor tissues and reconstituted the mice with 364

PBMC from the same donors. To exclude the anti-tumor effect mediated by 365

CD56+ NK cells, CD56+ NK cells in PBMC were sorted out before injection. 366

We found that rCD155 treatment decreased CD8 T cell in the tumor 367

microenvironment and IFNγ production in the infiltrated CD8 T cells was 368

decreased by rCD155. In addition, rCD155 decreased the transcripts of IFNγ, 369

TNF-α, GzmB and Perforin in the tumor tissue. The suppressed immune 370

response contributed to tumor growth in the PDX mice. Thus, enhancing the 371

antitumor immune response targeting CD155 in the tumor microenvironment 372

could be a good therapeutic strategic for LUAD. 373


https://doi.org/10.1101/688812

16

The balance between the inhibitory signal CD96/CD155 and the co-374

stimulatory signal CD226/CD155 is important for immune homeostasis (Chan 375

et al, 2014; Gao et al, 2017). CD96 blockade inhibits experimental metastases 376

(Blake et al, 2016) and prevents tumor from relapsing in a transgenic 377

pancreatic ductal adenocarcinoma mouse model (Brooks et al, 2017). In the 378

current study, we found that CD96 expression was increased in CD8 T cells 379

from LUAD patients. CD226, the co-stimulatory molecule, was decreased in 380

CD8 T cells from LUAD. The increased expression of CD96 and decreased 381

expression of CD226 contributed to a hypo-immune response, which impaired 382

the antitumor immune response in LUAD. The activity of AKT-mTOR signaling 383

was decreased in CD96+ CD8 T cells, implying the low activity of CD8 T cell 384

function. We confirmed this by measuring cytokine production in CD8 T cells. 385

CD96+ CD8 T cells expressed substantially lower levels of IFNγ and TNFα 386

than CD96- CD8 T cells did. LUAD cells might suppress the immune response 387

by inducing the imbalance of CD96/CD226 expression on CD8 T cells in the 388

tumor microenvironment. 389

The tumor microenvironment has played a critical role in shaping the 390

immune response (Whiteside, 2008). We showed that LUAD cells induced the 391

expression of CD96 and suppression of CD226 on CD8 T cells when they 392

were co-cultured together. Blocking CD96 restored CD8 T cell function as 393

inhibited by LUAD cells. LUAD cells impaired CD8 T cell function, and CD155 394

overexpression in cancer cells further decreased impaired CD8 T cell function. 395

Blocking CD96 neutralized CD155-mediated inhibition of CD8 T cells. LUAD 396

cells induced the imbalance between CD96 and CD226 expression on CD8 T 397


https://doi.org/10.1101/688812

17

cells. These data demonstrated a mechanism in which LUAD cells suppress 398

immune response in the tumor microenvironment through CD155/CD96. 399

In conclusion, our findings provide insights into the mechanism that 400

CD155 facilitates tumor growth by impairing the antitumor immune response 401

in the tumor microenvironment through CD96. Targeting CD155/CD96 to 402

unleash CD8 T cells in the tumor microenvironment could be a novel 403

therapeutic alternative for LUAD patients. 404

405

Materials and methods 406

Patients 407

Peripheral blood and primary tumor tissue samples were collected from 408

clinically and pathologically verified lung cancer patients at the First Affiliated 409

Hospital, Sun Yat-sen University. Age- and sex-matched blood was collected 410

from healthy donors. Fresh tumor tissues were collected from patients with 411

advanced LUAD. The study was approved by the Institutional Review Board 412

of First Affiliated Hospital, Sun Yat-sen University. All animal procedures were 413

approved by the ethics committee of the First Affiliated Hospital, Sun Yat-sen 414

University and performed in accordance with the guidelines provided by the 415

National Institute of Health Guide for Care and Use of Animals. Consent forms 416

were obtained from each patient. Demographic characteristics of the included 417

patients were described in Supplementary Table 1. 418

419

Cell lines and cell culture 420

The normal human bronchial epithelial cell line BEAS-2B was obtained 421

from ATCC (Maryland, USA). Cell lines were authenticated by cell viability 422


https://doi.org/10.1101/688812

18

analysis, short tandem repeat (STR) profiling, and isoenzyme analysis. Cell 423

lines were screened for mycoplasma contamination as described previously 424

(Li et al, 2014). Cells were grown in RPMI 1640 medium supplemented with 425

10% fetal bovine serum and kept in a humidified atmosphere at 37°C with 5% 426

CO2. 427

428

Western blotting 429

Cells were collected, and the proteins were extracted, separated by SDS-430

polyacrylamide gels and then electro-transferred onto polyvinylidene difluoride 431

membranes. The membranes were washed with TBST, blocked with 10% 432

nonfat milk in TBST and incubated with anti-CD155 (Abcam, Hong Kong) or 433

anti-β-actin (Cell Signaling Technology, USA) primary antibodies at 4°C 434

overnight. The membranes were then washed and incubated with horseradish 435

peroxidase conjugated anti-rabbit IgG (Abcam, Hong Kong) at room 436

temperature for 60 min. Signals were detected by enhanced 437

chemiluminescence (ECL). 438

439

Immunohistochemistry 440

Paraffin-embedded tissues were cut into 4-μm sections. The slides 441

were deparaffinized, and antigen retrieval was performed. The slides were 442

incubated with anti-CD8 (1:500) and anti-CD155 (1:100) (Abcam, Hong Kong) 443

primary antibodies at 4°C overnight. The sections were then incubated with 444

an HRP-conjugated secondary antibody for 1 h at room temperature. 445

Peroxidase was visualized with 3,3’ diaminobenzidine, and the slides were 446

counterstained with hematoxylin. For immunofluorescence staining, frozen 447


https://doi.org/10.1101/688812

19

sections were fixed with acetone for 15 min. The slides were incubated with 448

anti-CD45 (1:200), anti-CK (1:200), and anti-CD155 (1:100) primary 449

antibodies (Abcam, Hong Kong) at 4°C overnight and then visualized with 450

Alexa Fluor® 488 anti-rabbit (1:200), Alexa Fluor® 546 anti-mouse (1:200) 451

(Thermo Fisher, USA), anti-rat-TRITC (1:200) (Abcam, Cambridge, UK) 452

secondary antibodies. Images were acquired using a fluorescence 453

microscope (Toshiba, Japan). The number of infiltrated CD8 T cells was 454

counted from 5 different high-power areas. Protein expression levels were 455

evaluated semiquantitatively based on staining intensity and distribution using 456

the immunoreactive score (IRS) as described previously (Nagata et al, 2004) 457

as follows: IRS =SI (staining intensity) × PP (percentage of positive cells). The 458

SI was determined as follows: 0, negative; 1, weak; 2, moderate; and 3, 459

strong. The PP was defined as follows: 0, 80% positive cells. Ten visual fields from different areas of 461

each tumor were used for the IRS evaluation. Negative control slides were 462

included for each staining. An IRS score that reached 3.0 was recognized as 463

high expression; other scores were considered low expression in this study. 464

465

Reverse transcription-polymerase chain reaction (RT-PCR) 466

RNA was extracted according to the manufacturer’s instructions 467

(Qiagen, USA). Taq DNA polymerase (Fermentas, USA) was used for cDNA 468

synthesis. Real-time PCR was performed using SYBR Green I (Roche, USA). 469

Amplification was performed as follows: preheating at 95°C for 10 min; 470

denaturing at 95°C for 15 s; and annealing and extension at 65°C for 45 s for 471


https://doi.org/10.1101/688812

20

a total of 35 cycles. β-actin was used as control. The primers used are listed 472

in Supplementary Table 2. 473

474

Flow cytometry 475

Cells isolated from tumor tissues, PBMCs isolated from the matched 476

patients with LUAD or healthy controls were stained with FITC-anti-CD4, PE-477

anti-CD8, APC-conjugated anti-CD96, PE-CY7-anti-Granzme B, APC-CY7-478

anti-Perforin, APC-anti-CD226 antibodies (Biolegend, USA). For intracellular 479

cytokine staining, T cells (1×106 cells/ml) were stimulated with 500 ng/ml 480

PMA and 1 μg/ml ionomycin at 37°C with 5% CO2, 1 μg/ml Brefeldin for 4 h. 481

Cells were collected and stained with BV650-conjugated anti-IFNγ antibody 482

(Biolegend, USA). The data were acquired using a cytometer machine (BD 483

Fortessa, USA). 484

485

Cell isolation 486

Fresh tumor tissue samples were obtained from patients with LUAD who 487

underwent surgical resection of tumor or from tumor-bearing mice. Samples 488

were minced and digested with type I collagenase (2 mg/ml) and DNase (40 489

U/ml) in RPMI 1640. Cells were filtered through a cell strainer and washed 490

with phosphate-buffered saline (PBS) twice. Peripheral blood mononuclear 491

cells (PBMCs) from the matched patients with LUAD were isolated by density 492

gradient centrifugation. Cells were first enriched for CD8 T cells using 493

EasySep™ human total or naïve CD8 T Cell enrichment kits (STEMCELL 494

Technologies, Vancouver, Canada). CD56- and CD56+ PBMC fractions were 495

sorted by flow cytometry. Cells were checked for purity (>97%) by flow 496


https://doi.org/10.1101/688812

21

cytometry. CD96+ and CD96- CD8 T cells were sorted using BD influx, and 497

the purities were checked (>95%). 498

499

Plasmids and retroviral infection 500

CD155 constructs were generated by sub-cloning PCR-amplified full-501

length human CD155 cDNA into pcDNA3.1. Stable cell line (5×106 cells) 502

expressing CD155 was selected via treatment with 0.5 μg/ml puromycin for 10 503

days beginning 48 h after infection. Following selection, cancer cell lysates 504

prepared from the pooled cell populations in sampling buffer were fractionated 505

by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS–PAGE) 506

to detect protein levels via western blotting. To deplete CD155, shRNA 507

sequences were cloned into pGV248 to generate pGV248/CD155-shRNA 508

(containing a green fluorescent protein reporter gene) targeting CD155. A 509

negative vector (pGV248/control-shRNA) was similarly constructed with an 510

unrelated shRNA sequence. DNA sequencing was used to verify all inserted 511

sequences. Transduction efficiencies were confirmed by western blot. 512

Transduction efficiency is measured by flow cytometry and indicated as 513

percentage of successfully transduced GFP-positive cells. 514

515

Co-culture 516

To study CD8 T cell functions affected by cancer cells, CD8 T cells (1517

×106 cells/ml) were co-cultured with cancer cells (5×106 cells) in which 518

CD155 was either knocked down or overexpressed. Cells were stimulated 519

with anti-CD3/CD28 beads and co-cultured with cancer cells in 48-well plates 520

at a ratio of 5:1. Anti-CD3/CD28 beads were still present during T-cell/LUAD 521


https://doi.org/10.1101/688812

22

co-culture. Human CD96 blocking antibody (5 μg/ml) or isotype control 522

(Biolegend, USA) was included in some experiments. 523

524

Xenograft mouse model 525

NOD.Cg-PrkdcscidIl2rgtm1Sug/JicCrl (NOG) mice (Weitonglihua 526

Experimental Animal Co., Ltd, Beijing, China) are severe combined 527

immunodeficient mice. Immune reaction of human immune cells against 528

human tumors and the underlying mechanisms can be tested using this 529

humanized mouse model. To better understand the immune response in the 530

tumor microenvironment in LUAD patients, we used a patient-derived 531

xenograft (PDX) mouse model. Freshly collected human tumor tissues (from 6 532

independent CD155low patients) were cut into 0.5-cm3 pieces and 533

subcutaneously engrafted into the NOG mice. One week later, CD56- PBMC 534

(1×107 cells/ml)) from the same donors were sorted by flow cytometry and 535

injected into the mice intraperitoneally. To study whether CD155 regulates 536

tumor growth in vivo, PDX mice were treated with recombinant human CD155 537

(rCD155, 5 mg/kg, R&D, USA) 3 times a week. 538

To further investigate the effect of CD155 on CD8 T cell immune 539

response in the tumor microenvironment, NOG mice received PBMC (1×107 540

cells/ml)) from the same patients intraperitoneally for immune reconstitution. 541

NOG mice were then subcutaneously inoculated with 5×106 LUADC-Vector or 542

LUADC-CD155 cells (from 6 independent CD155low patients). 543

The mice were monitored every 2 days for signs of morbidity and 544

mortality. Tumor size was measured by a caliper, and tumor volume was 545

calculated using the formula volume as follow: (length×width2)×π/6. In vivo 546


https://doi.org/10.1101/688812

23

bioluminescence imaging was performed using the IVIS100 system. The 547

Living Image acquisition and analysis software (Caliper Life Sciences) were 548

used together as described previously (Olsen et al, 2017). 549

550

Statistics 551

The data were expressed as the mean±SEM. Statistical analyses were 552

performed using SPSS 16.0 (Chicago, IL, USA). The differences between 553

groups were assessed by an unpaired, two-tailed Student’s t test. To adjust 554

for multiple testing, in addition to individual p-values, we used Hochberg's 555

step-down method to control for a family-wise-error rate at the 0.05 level. 556

Survival curves were plotted using the Kaplan-Meier method and compared 557

with the log-rank test. Bivariate correlation analysis was demonstrated as 558

Spearman's rank correlation coefficient. Where appropriate, one-way ANOVA 559

was used and pair-wise comparison using Tukey's method to adjust for 560

multiple testing was applied. Two-tailed p

24

improving prognosis is attracting more and more attention in the basic 572

research and clinic application. 573

574

Results 575

In this study, we found that CD155 expression was significantly 576

increased in tumor tissue and associated with decreased immune response, 577

leading to poor survival in lung adenocarcinoma patients. Lung 578

adenocarcinoma cells suppressed CD8 T cell function through CD155. 579

Recombinant human CD155 protein inhibited immune response in the tumor 580

microenvironment and promoted tumor growth in a patient-derived xenograft 581

mouse model. In addition, we detected increased CD96 (co-inhibitory receptor) 582

and decreased CD226 (co-stimulatory receptor) on CD8 T cells from lung 583

adenocarcinoma patients. In a T cell-cancer cell co-culture system, IFNγ 584

production in CD8 T cells was suppressed and blocking CD155/CD96 could 585

restore IFNγ production in CD8 T cells. 586

587 Impact 588

LUAD cells suppress antitumor immune response through 589

CD155/CD96-interaction. Moreover, the inhibition effect can be reversed by 590

CD96 blocking antibody, suggesting that CD155/CD96 can serve as a 591

potential treatment target for LUAD 592

593

ACKNOWLEDGEMENT 594

Author contributions 595

Conception and design: Zunfu Ke, Weiling He, Hui Zhang, Shuhua Li and 596

Yongmei Cui, 597


https://doi.org/10.1101/688812

25

Funding support: Zunfu Ke 598

Collection and assembly of data: Ying Zhu, Zheng Zhu, Junfeng Zhu, Yiyan 599

Lei, Run Lin, Di Xu, Wenting Jiang and Han Wang, 600

Data analysis and interpretation: Zunfu Ke , Weiling He and Hui Zhang 601

Manuscript writing: Zunfu Ke , Weiling He and Ying Zhu 602

Final approval of manuscript: All authors. 603

604

Authors’ disclosures of potential conflicts of interest: 605

All authors declare no potential conflicts of interest. 606

607

Funding 608

This work was supported by grants from YFC (2017YFC1308800), National 609

Natural Science Foundation of China to Zunfu Ke (30900650, 81372501, 610

81572260, 81773299, 81701834, 81502327, 81172232 and 31430030), and 611

Guangdong Natural Science Foundation (2011B031800025, 612

S2012010008378, S2012010008270, S2013010015327, 2013B021800126, 613

20090171120070, 9451008901002146, 2013B021800126, 2014A030313052, 614

2014J4100132, 2015A020214010, 2016A020215055, 201704020094, 615

2013B021800259, 2017B070705002, 16ykjc08 and 2015ykzd07). 616

617

618


https://doi.org/10.1101/688812

26

Reference 619

Ahmadzadeh M, Johnson LA, Heemskerk B, Wunderlich JR, Dudley ME, White DE, 620 Rosenberg SA (2009) Tumor antigen-specific CD8 T cells infiltrating the tumor 621 express high levels of PD-1 and are functionally impaired. Blood 114: 1537-1544 622 623 Atsumi S, Matsumine A, Toyoda H, Niimi R, Iino T, Sudo A (2013) Prognostic 624 significance of CD155 mRNA expression in soft tissue sarcomas. Oncology letters 625 5: 1771-1776 626 627 Blake SJ, Stannard K, Liu J, Allen S, Yong MC, Mittal D, Aguilera AR, Miles JJ, 628 Lutzky VP, de Andrade LF, Martinet L, Colonna M, Takeda K, Kuhnel F, Gurlevik E, 629 Bernhardt G, Teng MW, Smyth MJ (2016) Suppression of Metastases Using a New 630 Lymphocyte Checkpoint Target for Cancer Immunotherapy. Cancer discovery 6: 631 446-459 632 633 Bottino C, Castriconi R, Pende D, Rivera P, Nanni M, Carnemolla B, Cantoni C, 634 Grassi J, Marcenaro S, Reymond N, Vitale M, Moretta L, Lopez M, Moretta A (2003) 635 Identification of PVR (CD155) and Nectin-2 (CD112) as cell surface ligands for 636 the human DNAM-1 (CD226) activating molecule. The Journal of experimental 637 medicine 198: 557-567 638 639 Brahmer JR, Tykodi SS, Chow LQ, Hwu WJ, Topalian SL, Hwu P, Drake CG, 640 Camacho LH, Kauh J, Odunsi K, Pitot HC, Hamid O, Bhatia S, Martins R, Eaton K, 641 Chen S, Salay TM, Alaparthy S, Grosso JF, Korman AJ, Parker SM, Agrawal S, 642 Goldberg SM, Pardoll DM, Gupta A, Wigginton JM (2012) Safety and activity of 643 anti-PD-L1 antibody in patients with advanced cancer. The New England journal 644 of medicine 366: 2455-2465 645 646 Brooks J, Fleischmann-Mundt B, Woller N, Niemann J, Ribback S, Peters K, Demir 647 IE, Armbrecht N, Ceyhan GO, Manns MP, Wirth TC, Kubicka S, Bernhardt G, Smyth 648 MJ, Calvisi DF, Gurlevik E, Kuhnel F (2017) Perioperative, spatiotemporally 649 coordinated activation of T and NK cells prevents recurrence of pancreatic 650 cancer. Cancer research 651 652 Chan CJ, Martinet L, Gilfillan S, Souza-Fonseca-Guimaraes F, Chow MT, Town L, 653 Ritchie DS, Colonna M, Andrews DM, Smyth MJ (2014) The receptors CD96 and 654 CD226 oppose each other in the regulation of natural killer cell functions. Nature 655 immunology 15: 431-438 656 657 Chandramohan V, Bryant JD, Piao H, Keir ST, Lipp ES, Lefaivre M, Perkinson K, 658 Bigner DD, Gromeier M, McLendon RE (2017) Validation of an 659 Immunohistochemistry Assay for Detection of CD155, the Poliovirus Receptor, in 660 Malignant Gliomas. Archives of pathology & laboratory medicine 141: 1697-1704 661 662 Crespo J, Sun H, Welling TH, Tian Z, Zou W (2013) T cell anergy, exhaustion, 663 senescence, and stemness in the tumor microenvironment. Current opinion in 664 immunology 25: 214-221 665 666


https://doi.org/10.1101/688812

27

Ding L, Getz G, Wheeler DA, Mardis ER, McLellan MD, Cibulskis K, Sougnez C, 667 Greulich H, Muzny DM, Morgan MB, Fulton L, Fulton RS, Zhang Q, Wendl MC, 668 Lawrence MS, Larson DE, Chen K, Dooling DJ, Sabo A, Hawes AC, Shen H, 669 Jhangiani SN, Lewis LR, Hall O, Zhu Y, Mathew T, Ren Y, Yao J, Scherer SE, Clerc K, 670 Metcalf GA, Ng B, Milosavljevic A, Gonzalez-Garay ML, Osborne JR, Meyer R, Shi X, 671 Tang Y, Koboldt DC, Lin L, Abbott R, Miner TL, Pohl C, Fewell G, Haipek C, 672 Schmidt H, Dunford-Shore BH, Kraja A, Crosby SD, Sawyer CS, Vickery T, Sander S, 673 Robinson J, Winckler W, Baldwin J, Chirieac LR, Dutt A, Fennell T, Hanna M, 674 Johnson BE, Onofrio RC, Thomas RK, Tonon G, Weir BA, Zhao X, Ziaugra L, Zody 675 MC, Giordano T, Orringer MB, Roth JA, Spitz MR, Wistuba, II, Ozenberger B, Good 676 PJ, Chang AC, Beer DG, Watson MA, Ladanyi M, Broderick S, Yoshizawa A, Travis 677 WD, Pao W, Province MA, Weinstock GM, Varmus HE, Gabriel SB, Lander ES, 678 Gibbs RA, Meyerson M, Wilson RK (2008) Somatic mutations affect key pathways 679 in lung adenocarcinoma. Nature 455: 1069-1075 680 681 Gao J, Zheng Q, Xin N, Wang W, Zhao C (2017) CD155, an onco-immunologic 682 molecule in human tumors. Cancer science 108: 1934-1938 683 684 Hanahan D, Coussens LM (2012) Accessories to the crime: functions of cells 685 recruited to the tumor microenvironment. Cancer cell 21: 309-322 686 687 Hanahan D, Weinberg RA (2011) Hallmarks of cancer: the next generation. Cell 688 144: 646-674 689 690 He W, Zhang H, Han F, Chen X, Lin R, Wang W, Qiu H, Zhuang Z, Liao Q, Zhang W, 691 Cai Q, Cui Y, Jiang W, Wang H, Ke Z (2017) CD155T/TIGIT Signaling Regulates 692 CD8(+) T-cell Metabolism and Promotes Tumor Progression in Human Gastric 693 Cancer. Cancer research 77: 6375-6388 694 695 Herbst RS, Baas P, Kim DW, Felip E, Perez-Gracia JL, Han JY, Molina J, Kim JH, 696 Arvis CD, Ahn MJ, Majem M, Fidler MJ, de Castro G, Jr., Garrido M, Lubiniecki GM, 697 Shentu Y, Im E, Dolled-Filhart M, Garon EB (2016) Pembrolizumab versus 698 docetaxel for previously treated, PD-L1-positive, advanced non-small-cell lung 699 cancer (KEYNOTE-010): a randomised controlled trial. Lancet 387: 1540-1550 700 701 Iguchi-Manaka A, Okumura G, Kojima H, Cho Y, Hirochika R, Bando H, Sato T, 702 Yoshikawa H, Hara H, Shibuya A, Shibuya K (2016) Increased Soluble CD155 in 703 the Serum of Cancer Patients. PloS one 11: e0152982 704 705 Imielinski M, Berger AH, Hammerman PS, Hernandez B, Pugh TJ, Hodis E, Cho J, 706 Suh J, Capelletti M, Sivachenko A, Sougnez C, Auclair D, Lawrence MS, Stojanov P, 707 Cibulskis K, Choi K, de Waal L, Sharifnia T, Brooks A, Greulich H, Banerji S, 708 Zander T, Seidel D, Leenders F, Ansen S, Ludwig C, Engel-Riedel W, Stoelben E, 709 Wolf J, Goparju C, Thompson K, Winckler W, Kwiatkowski D, Johnson BE, Janne 710 PA, Miller VA, Pao W, Travis WD, Pass HI, Gabriel SB, Lander ES, Thomas RK, 711 Garraway LA, Getz G, Meyerson M (2012) Mapping the hallmarks of lung 712 adenocarcinoma with massively parallel sequencing. Cell 150: 1107-1120 713 714


https://doi.org/10.1101/688812

28

Inozume T, Yaguchi T, Furuta J, Harada K, Kawakami Y, Shimada S (2016) 715 Melanoma Cells Control Antimelanoma CTL Responses via Interaction between 716 TIGIT and CD155 in the Effector Phase. The Journal of investigative dermatology 717 136: 255-263 718 719 Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D (2011) Global cancer 720 statistics. CA Cancer J Clin 61: 69-90 721 722 Jiang Y, Li Y, Zhu B (2015) T-cell exhaustion in the tumor microenvironment. Cell 723 death & disease 6: e1792 724 725 Kawai O, Ishii G, Kubota K, Murata Y, Naito Y, Mizuno T, Aokage K, Saijo N, 726 Nishiwaki Y, Gemma A, Kudoh S, Ochiai A (2008) Predominant infiltration of 727 macrophages and CD8(+) T Cells in cancer nests is a significant predictor of 728 survival in stage IV nonsmall cell lung cancer. Cancer 113: 1387-1395 729 730 Kim R, Emi M, Tanabe K (2007) Cancer immunoediting from immune 731 surveillance to immune escape. Immunology 121: 1-14 732 733 Koike S, Horie H, Ise I, Okitsu A, Yoshida M, Iizuka N, Takeuchi K, Takegami T, 734 Nomoto A (1990) The poliovirus receptor protein is produced both as 735 membrane-bound and secreted forms. The EMBO journal 9: 3217-3224 736 737 Li G, Wang Z, Ye J, Zhang X, Wu H, Peng J, Song W, Chen C, Cai S, He Y, Xu J (2014) 738 Uncontrolled inflammation induced by AEG-1 promotes gastric cancer and poor 739 prognosis. Cancer research 74: 5541-5552 740 741 Lin M, Chen JF, Lu YT, Zhang Y, Song J, Hou S, Ke Z, Tseng HR (2014) 742 Nanostructure embedded microchips for detection, isolation, and 743 characterization of circulating tumor cells. Accounts of chemical research 47: 744 2941-2950 745 746 Masson D, Jarry A, Baury B, Blanchardie P, Laboisse C, Lustenberger P, Denis MG 747 (2001) Overexpression of the CD155 gene in human colorectal carcinoma. Gut 748 49: 236-240 749 750 Mendelsohn CL, Wimmer E, Racaniello VR (1989) Cellular receptor for poliovirus: 751 molecular cloning, nucleotide sequence, and expression of a new member of the 752 immunoglobulin superfamily. Cell 56: 855-865 753 754 Molina JR, Yang P, Cassivi SD, Schild SE, Adjei AA (2008) Non-small cell lung 755 cancer: epidemiology, risk factors, treatment, and survivorship. Mayo Clinic 756 proceedings 83: 584-594 757 758 Nagata Y, Lan KH, Zhou X, Tan M, Esteva FJ, Sahin AA, Klos KS, Li P, Monia BP, 759 Nguyen NT, Hortobagyi GN, Hung MC, Yu D (2004) PTEN activation contributes 760 to tumor inhibition by trastuzumab, and loss of PTEN predicts trastuzumab 761 resistance in patients. Cancer cell 6: 117-127 762 763


https://doi.org/10.1101/688812

29

Noy R, Pollard JW (2014) Tumor-associated macrophages: from mechanisms to 764 therapy. Immunity 41: 49-61 765 766 Olsen SN, Wronski A, Castano Z, Dake B, Malone C, De Raedt T, Enos M, DeRose 767 YS, Zhou W, Guerra S, Loda M, Welm A, Partridge AH, McAllister SS, Kuperwasser 768 C, Cichowski K (2017) Loss of RasGAP Tumor Suppressors Underlies the 769 Aggressive Nature of Luminal B Breast Cancers. Cancer discovery 7: 202-217 770 771 Pages F, Berger A, Camus M, Sanchez-Cabo F, Costes A, Molidor R, Mlecnik B, 772 Kirilovsky A, Nilsson M, Damotte D, Meatchi T, Bruneval P, Cugnenc PH, 773 Trajanoski Z, Fridman WH, Galon J (2005) Effector memory T cells, early 774 metastasis, and survival in colorectal cancer. The New England journal of 775 medicine 353: 2654-2666 776 777 Reck M, Rodriguez-Abreu D, Robinson AG, Hui R, Csoszi T, Fulop A, Gottfried M, 778 Peled N, Tafreshi A, Cuffe S, O'Brien M, Rao S, Hotta K, Leiby MA, Lubiniecki GM, 779 Shentu Y, Rangwala R, Brahmer JR, Investigators K- (2016) Pembrolizumab 780 versus Chemotherapy for PD-L1-Positive Non-Small-Cell Lung Cancer. The New 781 England journal of medicine 375: 1823-1833 782 783 Rosenberg SA, Yang JC, Restifo NP (2004) Cancer immunotherapy: moving 784 beyond current vaccines. Nature medicine 10: 909-915 785 786 Sato E, Olson SH, Ahn J, Bundy B, Nishikawa H, Qian F, Jungbluth AA, Frosina D, 787 Gnjatic S, Ambrosone C, Kepner J, Odunsi T, Ritter G, Lele S, Chen YT, Ohtani H, 788 Old LJ, Odunsi K (2005) Intraepithelial CD8+ tumor-infiltrating lymphocytes and 789 a high CD8+/regulatory T cell ratio are associated with favorable prognosis in 790 ovarian cancer. Proceedings of the National Academy of Sciences of the United 791 States of America 102: 18538-18543 792 793 Shibuya A, Campbell D, Hannum C, Yssel H, Franz-Bacon K, McClanahan T, 794 Kitamura T, Nicholl J, Sutherland GR, Lanier LL, Phillips JH (1996) DNAM-1, a 795 novel adhesion molecule involved in the cytolytic function of T lymphocytes. 796 Immunity 4: 573-581 797 798 Sloan KE, Stewart JK, Treloar AF, Matthews RT, Jay DG (2005) CD155/PVR 799 enhances glioma cell dispersal by regulating adhesion signaling and focal 800 adhesion dynamics. Cancer research 65: 10930-10937 801 802 Swann JB, Smyth MJ (2007) Immune surveillance of tumors. The Journal of 803 clinical investigation 117: 1137-1146 804 805 Talmadge JE, Gabrilovich DI (2013) History of myeloid-derived suppressor cells. 806 Nature reviews Cancer 13: 739-752 807 808 Whiteside TL (2008) The tumor microenvironment and its role in promoting 809 tumor growth. Oncogene 27: 5904-5912 810 811


https://doi.org/10.1101/688812

30

Yu X, Harden K, Gonzalez LC, Francesco M, Chiang E, Irving B, Tom I, Ivelja S, 812 Refino CJ, Clark H, Eaton D, Grogan JL (2009) The surface protein TIGIT 813 suppresses T cell activation by promoting the generation of mature 814 immunoregulatory dendritic cells. Nature immunology 10: 48-57 815 816 Zou W, Chen L (2008) Inhibitory B7-family molecules in the tumour 817 microenvironment. Nature reviews Immunology 8: 467-477 818 819 820

821

822

823

824

825

826

827

828

829

830

831

832

833

834

835

836

837

838

839

840


https://doi.org/10.1101/688812

31

841 Figure and Figure legends 842

843

Figure 1. CD155 expression and immune suppression in the tumor 844

microenvironment of LUAD. (A), CD155 expression in tumor or para-tumor 845

lung tissues from patients (n=6) with LUAD was measured by western blotting. 846

All of 6 representative patients were considered CD155high. (B), CD155 847

expression in tumor or para-tumor lung tissues was detected by 848

immunohistochemistry (200×). (C), Primary LUADCs were isolated from 6 849


https://doi.org/10.1101/688812

32

CD155high patients and expanded as in Supplementary Figure 1. CD155 850

expression in primary LUADCs was detected by immunofluorescence. White 851

blood cells (WBC) were used as control. (D), CD155high or CD155low 852

expression was determined by immunohistochemistry (200×). (E), RNA was 853

extracted from tumor tissue. Gene expression was measured by RT-PCR and 854

shown as a heat map. (F), Gene expression of IL-2, IFNγ, TNFα and 855

granzyme B (GzmB) was summarized from 24 independent samples (13 for 856

CD155high and 11 for CD155low). Error bars show SEM. The data are shown 857

as the mean ± SEM, *p

33

864

Figure 2. LUAD cells impaired CD8 T cell effector functions. (A), Scheme 865

of CD8 T cells (from the matched CD155high LUAD patient) co-cultured with 866

LUADC1-6 from 6 independent CD155high patients in a cell-cell contact 867

manner. CD8 T cells were stimulated with beads for 3 days. Representative 868

results are shown as follows (B-I). (B), GzmB expression on CD8 T cells was 869

analyzed by flow cytometry. (C), Mean fluorescence intensity (MFI) of GzmB 870

in CD8 T cells was summarized. (D), Perforin expression on CD8 T cells was 871

analyzed by flow cytometry. (E), Percentage of Perforin-expressing CD8 T 872

cells was summarized. (F), IL-2 and TNFα production in CD8 T cells was 873


https://doi.org/10.1101/688812

34

measured by flow cytometry. (G), Percentages of IL-2- and TNFα-producing 874

CD8 T cells were summarized. (H), IFNγ production in CD8 T cells was 875

measured by flow cytometry. (I), Percentage of IFNγ+ CD8 T cells was 876

summarized. (J), Scheme of CD8 T cells (from the same CD155high LUAD 877

patient) co-cultured separately with LUADC1-6 from 6 independent CD155high 878

patients. A cell culture insert with 0.4-μm pore size was used to culture CD8 T 879

cells and tumor cells separately. IFNγ production in CD8 T cells was 880

measured by flow cytometry. Representative results are shown as follows (K). 881

(K), Percentage of IFNγ+ CD8 T cells was summarized. The data are shown 882

as the mean±SEM of 3 independent experiments, **p

35

886

Figure 3. LUAD cells suppress CD8+ T cell effector functions through 887

CD155. Representative results are shown as follows (A-H). (A), CD155 888

expression in LUADC1-6 cells (from 6 independent CD155high patients) was 889

detected by western blotting. (B), CD155 expression in LUADC was detected 890

by flow cytometry. (C), CD8 T cells (from the matched CD155high LUAD 891

patient) were stimulated with anti-CD3/CD28 beads and co-cultured with 892

LUADC. (D) CD8 T cells were co-cultured with LUAD cells that were treated 893

with Scramble or CD155 RNAi. CD8+ T cells cultured alone was used as 894


https://doi.org/10.1101/688812

36

control. Phosphorylation of AKT, mTOR, S6K and 4EBP1 in CD8+ T cells was 895

measured by flow cytometry and immunoblot, respectively. (E), CD8 T cells 896

(from the matched CD155high LUAD patient) were co-cultured with or without 897

LUADC cells that were treated with CD155-specific or scramble RNAi for 3 d. 898

CD8+ T cells cultured alone was used as control. IFNγ production in CD8 T 899

cells was measured by flow cytometry and summarized in F. (G), CD8 T cells 900

(from the matched CD155high LUAD patient) were co-cultured with LUADC in 901

which CD155 was stably upregulated for 3 d. IFNγ production in CD8 T cells 902

was measured by flow cytometry and summarized in H. The data are shown 903

as the mean±SEM of 3 independent experiments, **p

37

920

921

Figure 4. CD155 suppressed the immune response in the tumor 922

microenvironment and promoted tumor growth. (A), Scheme of the 923

patient-derived xenograft (PDX) experiment. Fresh tumor samples (from 6 924

independent CD155low patients) were collected and engrafted into NOG mice 925

subcutaneously. One week later, CD56- PBMC from the matched donors were 926

injected intraperitoneally for immune reconstitution. PDX mice were treated 927

with recombinant CD155 (rCD155, 5mg/kg) or vehicle. (B), The number of 928

CD8 T cells in the tumor microenvironment was measured by IHC (400×). (C) 929

The number of CD8 T cells per high power field (HPF) was summarized from 930

6 independent samples. (D), Tumor tissue was digested to generate a single-931

cell suspension. IFNγ production in CD8 T cells was measured by flow 932

cytometry. Representative flow plots were gated on CD8 T cells. (E), 933


https://doi.org/10.1101/688812

38

Percentages of CD8+IFNγ+ cells were summarized from 6 independent 934

samples. (F, G), RNA was extracted from tumor tissue. Transcripts of 935

IFNγ ,TNFα, GzmB and Perforin expression in the tissue was measured by 936

RT-PCR. The data were summarized from 6 independent samples. (H), Mice 937

were treated with rCD155 or vector, and tumor growth was monitored. The 938

data are shown as the mean±SEM, **p

39

949

Figure 5. Dysregulation of CD96/CD226 identified CD8 T cell exhaustion 950

in LUAD. (A-D), CD96 or CD226 expression on CD8 T cells from PBMCs of 951

24 matched patients with LUAD （13 CD155high and 11 CD155low） or healthy 952

controls (HC) (n=24) and from tumor-infiltrating lymphocytes (TILs) was 953

analyzed by flow cytometry. Percentages of CD8+CD96+ or CD8+CD226+ cells 954

in PBMCs (n=24) or TILs (n=24) were summarized and shown as bar graphs. 955

Error bars show SEM. (E), CD8+CD96+ or CD8+96- cells were sorted from 956

PBMCs in LUAD patient. Cells were stimulated with anti-CD3/CD28 beads. 957


https://doi.org/10.1101/688812

40

Phosphorylation of AKT, mTOR S6K and 4EBP1 was measured by cytometry 958

and immunoblot, respectively. (F, G), TNFα and IFNγ production in CD96+ or 959

CD96- CD8 T cells from LUAD patient was measured by flow cytometry. The 960

data are shown as the mean±SEM, **p

41

973

Figure 6. LUAD cells suppressed anti-tumor immune response through 974

the CD155/CD96 pathway. (A, B), CD8 T cells isolated from the matched 975

LUAD patient were stimulated with anti-CD3/CD28 beads and co-cultured with 976

LUADC1 from the same patien for 3 d. CD96 and CD226 expression on CD8 977

T cells was measured by flow cytometry. Percentage of CD8+CD96+ or 978

CD8+CD226+ T cells was summarized from 8 independent samples. (C), 979

Phosphorylation of AKT and mTOR in CD8 T cells were measured by western 980

blotting. The expression of p-S6K and p-4EBP1 was determined by flow 981

cytometry. (D, E), CD8+ T cells were co-cultured with autologous LUAD cells. 982


https://doi.org/10.1101/688812

42

Anti-CD96 antibody (5 μg/ml) was added in some experiments. CD8 T cells 983

cultured alone was used as control. IFNγ production in CD8 T cells was 984

measured by flow cytometry. Percentage of IFNγ+ CD8 T cells was 985

summarized from 6 independent samples (right panel). (F) CD155 or empty 986

vector was transfected into LUADC1. CD8 T cells were co-cultured with 987

LUADC1-CD155 or LUADC1-vector for 3 d. Anti-CD96 antibody (5 μg/ml) was 988

added in the experiments. IFNγ production in CD8 T cells was measured by 989

flow cytometry. The percentage of IFNγ+ CD8 T cells was summarized from 6 990

independent samples (right panel). The data are shown as the mean±SEM, 991

**p


https://doi.org/10.1101/688812

cd155/cd96 promotes immunosuppression in lung ... › content › 10.1101 › 688812v1.full.pdf ·...

Documents