1

TITLE: Functional consequences of platelet binding to T lymphocytes in 1

inflammation 2

3

Carlos Zamora*, Elisabet Cantó*, Juan C. Nieto*, M. Angels Ortiz*, Cesar 4

Diaz-Torné†, Cesar Diaz-Lopez†, Josep M. Llobet†, Candido Juarez‡, Sílvia 5

Vidal* 6

7

* Dep. Immunology of Institut Recerca Hospital S. Pau, IIB-Sant Pau, 8

Barcelona, Spain 9

† Unit of Rheumatology of Hospital S. Pau, Barcelona, Spain 10

‡ Dep. Immunology of Hospital S. Pau, Barcelona, Spain 11

12

Summary sentence: Platelet binding is a new regulatory mechanism of T 13

cell function. Higher frequency of platelet binding to lymphocytes is 14

associated with low severe inflammatory profile 15

16

Running title: Platelet binding to T lymphocytes 17

18

Corresponding author: Sílvia Vidal 19

Avda. Antoni M.Claret, 167; Barcelona-08025, Spain; 20

Telf. +(34) 935537544 21

Fax +(34) 935537598 22

svidal@santpau.cat 23

24

Key words: Human, T cell activation, Rheumatoid arthritis 25

26

Total character count: 18000 27

Total number of figures: 5 28

Total number of references: 43 29

Total number of words in abstract: 191 30

Total number of words in summary sentence: 25 31

32

33

34

2

Abbreviations 35

Anti-CCP Anti-cyclic citrullinated peptide antibodies 36

CFSE 5(6)-carboxyfluorescein diacetate succinimidylester 37

CRP C-reactive protein 38

DAS28-CRP Disease activity score-C reactive protein 39

ESR Erythrocyte sedimentation rate 40

iGMFI Integrated geometric fluorescence mean 41

Plt Platelet 42

RA Rheumatoid arthritis 43

RF Rheumatoid factor44

3

Abstract 45

Expression of the scavenger receptor CD36 on lymphocytes is intriguing. We 46

observed that a minor subpopulation of lymphocytes expressed CD36 on cell 47

surface. We investigated the source of CD36 and also the proliferation and 48

cytokine production of these CD36+ CD4+ lymphocytes. Flow cytometry 49

analysis and immunofluorescence microscopy showed that CD36+ platelets 50

were responsible for CD36 detection on lymphocytes. CD36 was then used as a 51

tool to characterize lymphocytes with bound platelets. Activation-induced 52

proliferation was lower in CD4+ lymphocytes with bound platelets than 53

lymphocytes without bound platelets. IL-17 and IFNγ production was also 54

reduced in lymphocytes with bound platelets. We then studied the presence of 55

CD36+ CD4+ lymphocytes in rheumatoid arthritis (RA) patients. We observed 56

that percentage of CD4+ lymphocytes with bound platelets was higher on RA 57

patients than in healthy donors. RA patients with higher titers of anti-cyclic 58

citrullinated peptide antibodies (anti-CCP), rheumatoid factor levels (RF) and 59

cardiovascular risk index presented a lower percentage of CD4+ lymphocytes 60

with bound platelets. These patients also had higher IL-17 and IFNγ production. 61

These results suggest that platelet binding modifies lymphocyte function. This 62

binding could be a regulatory mechanism in RA that confers a less severe 63

phenotype. 64

65

66

67

68

69

70

71

4

Introduction 72

CD36 is a transmembrane glycoprotein member of the class B scavenger 73

receptor family. This receptor has been found on macrophages/monocytes, 74

microglia, microvascular endothelium, cardiac and skeletal muscle, adipocytes 75

and platelets [1]. In these cells, CD36 recognizes a wide variety of ligands that 76

support its scavenger role. We have recently found a minor subpopulation of 77

lymphocytes that express CD36 (unpublished results). The presence of such a 78

scavenger receptor on a lymphocyte is intriguing. Vallejo et al. were not able to 79

detect CD36 in T lymphocytes by western blot [2]. However, recent reports 80

showed that B and γδ T cells can express CD36 [3-4]. Other authors described 81

that after mitogenic stimulation, CD36 was upregulated on peripheral blood 82

lymphocytes [5]. 83

The subpopulation of positive CD36 lymphocytes could have a characteristic 84

function. In the recent literature, the description and analysis of subpopulations 85

provide useful markers to predict disease development or activity. Th17 86

lymphocytes, for instance, are involved in the pathogenesis of rheumatoid 87

arthritis (RA), partially through IL-17 production [6-7]. IL-17 is highly expressed 88

in RA synovium [8] where it contributes to angiogenesis [9]. IL-17 has also 89

become a predictor of microvascular function and arterial compliance in RA 90

[10]. CD161 has been described as a marker of a T lymphocyte subpopulation 91

producing IL-17 [11]. The analysis of CD4+ CD161+ by flow cytometry on whole 92

blood has been associated to RA disease activity [12]. Other T lymphocyte 93

subpopulations, like Th1, produce IFNγ and TNFα. These two inflammatory 94

mediators are also associated with RA, supporting the microenvironment in RA 95

joint [13]. Another type of relevant cells in inflammatory diseases is Treg that 96

produced IL-10 and inhibits T cell expansion in vivo [14] and regulates Th1 and 97

Th2 response [15]. Tregs can be identified by flow cytometry based on their 98

CD4+ CD25 (high) CD127 (low/-) phenotype. The percentage of cells with Treg 99

phenotype has been associated with RA disease activity [16]. 100

To clarify the controversy about CD36 expression on T lymphocytes and the 101

function of CD36+ cell subset: Firstly, we analyzed whether CD36 expression 102

was extrinsic or intrinsic to lymphocytes. Secondly, we studied the functional 103

capability of this subpopulation selecting T lymphocytes with this marker. 104

Finally, we analyzed their role in autoimmune pathologies such as RA. 105

5

Materials and methods 106

107

Samples and isolation of peripheral blood mononuclear cells 108

Whole blood from healthy donors (n=16) and RA patients (n=20) was collected 109

in BD Vacutainer tubes (BD Pharmingen, Franklin Lakes, NJ), containing 110

heparin as anticoagulant. Diagnosis was based on the American College of 111

Rheumatology diagnostic criteria for rheumatoid arthritis [17]. Disease activity 112

was assessed using the Disease activity score - C reactive protein (DAS28-113

CRP) [18]. The mean DAS28-CRP was 2.81 (range 1.13-5.08) and the mean 114

disease duration was 12.6 years (range, 2-40). Sixteen RA patients were 115

receiving methotrexate (7.5-22.5 mg), 6 of them were receiving prednisone too 116

(1.25-5 mg), and 4 RA patients were receiving leflunomide (20 mg). We 117

obtained clinical parameters of RA patients, C reactive protein (CRP) and 118

erythrocyte sedimentation rate (ESR). The cardiovascular risk index was 119

calculated according to age at the time of the study, gender, smoking history, 120

systolic arterial blood pressure, and total serum cholesterol level [19]. Written 121

consent was obtained and ethical approval for the study was granted by the 122

Institutional Ethics Committee. 123

Human peripheral blood mononuclear cells (PBMC) were isolated from 124

peripheral blood by gradient centrifugation on Ficoll-Hypaque (Lymphoprep, 125

AXIS-SHIELD PoCAs, Oslo, Norway). After centrifugation, cells were collected, 126

washed twice in PBS and resuspended in complete medium (RPMI 127

supplemented with 10% fetal calf serum, Biological Industries, Kibbutz Beit 128

Haemek, Israel, 2 mM glutamine, 100 U/ml penicillin and 100 μg/ml 129

streptomycin ,Biowhittaker, Verviers, Belgium). 130

131

Peripheral blood mononuclear cells and whole blood staining 132

PBMC (3 x 105 cells) were washed and resuspended in 100 μl of staining buffer 133

(PBS supplemented with 0.5 % BSA, Calbiochem Merk KGaA, Darmstadt, 134

Germany). PBMC were incubated with anti CD4-PECy7, CCR5-PE ,CCR6-PE, 135

CCR7-FITC, HLA-DR-PE, (BioLegend, San Diego, CA), CD8-PEDy647, 136

CD14-PEDy647, CD36-FITC (Clone TR9), CD41a-FITC, CD45RA-PEDy647 137

(Immunotools, Friesoythe, Germany), CD3-PECy7, CD25-PE, CD36-PE (clone 138

CDB38), CD45RA-FITC, CD45RO-FITC, CCR4-PE, CXCR3-PE (BD 139

6

Pharmingen, Frankin Lakes, NJ) monoclonal antibodies and the corresponding 140

isotype controls for 20 minutes in the dark. Samples were then washed with 141

staining buffer and resuspended in PBS. 142

Whole blood cells (100 μl) were incubated for 20 minutes at room temperature 143

in the dark with anti CD4-PECy7, CD8-PEDy647 and CD36-FITC monoclonal 144

antibodies. Red blood cells were lysed and white cell fixed using TQ-prep 145

System (Coulter Corp., Miami, Fla) to be analyzed by flow cytometry. 146

147

Immunofluorescence microscopy analysis 148

To assess CD36 expression on monocyte and lymphocyte surface, PBMC (2 x 149

105 cells) were stained with anti CD3-FITC (Immunotools) or anti CD14-FITC 150

(Immunotools) with anti CD36-PE (BD Pharmingen), CD41a-PE (Immunotools) 151

or corresponding isotype control. A drop of cells in PBS was then left to dry on a 152

glass slide at room temperature. Expression of these markers was analyzed 153

using NIKON ECLIPSE 80i microscopy. 154

155

Proliferation assay of lymphocyte 156

To determine CD4+ lymphocyte proliferation, PBMC were labeled with 5(6)-157

carboxyfluorescein diacetate succinimidylester (CFSE) (Sigma, S. Louis, 158

Missouri). Adjusted 107 cells/ml were stained with 7 μM of CFSE and incubated 159

for 15 minutes at 37ºC. After incubation, cells were washed with pre-warmed 160

medium and incubated for 30 minutes at 37ºC to ensure complete modification 161

of the probe. Cells were then washed with pre-warmed medium, resuspended 162

with complete medium at 5 x 106 cells/ml and incubated with anti-CD3, CD28 163

and CD2 T cell activation/expansion kit for 72 hours and 7 days according to the 164

manufacturer’s instructions (Milteny Biotec, CA, USA). After culturing, PBMC (3 165

x 105 cells) were collected, and stained with anti CD4-PECy7 and anti CD36-PE 166

monoclonal antibody (BD Pharmingen). Cells were then washed with staining 167

buffer and resuspended in PBS. CD4+ CD36- and CD4+ CD36+ history cycles 168

of proliferation were analyzed by flow cytometry. 169

170

Determination of intracellular cytokine production 171

PBMC were cultured with anti-CD3, CD28 and CD2 T cell activation/expansion 172

kit for 48 hours according to the manufacturer’s instructions. Cytokine 173

7

production on healthy controls was assessed on CD3+ CD36- and CD3+ 174

CD36+ cells by intracellular staining of IL-17-PE (eBioscience, San Diego CA, 175

USA), IFNγ-PE (BD Pharmingen) IL-10-PE (eBioscience) and TNFα-PE (BD 176

Pharmingen) using a Cytofix/Cytoperm and Perm/Wash kit from BD 177

biosciences. Five hours before analysis, PMBC were cultured with PMA (50 178

ng/ml)/ Ionomycin (1 μg/ml) in the presence of BD GolgiStop TM Protein 179

Transport Inhibitor. Cells were collected and stained. Intracellular cytokine 180

production was analyzed on gated subpopulations by flow cytometry. 181

182

Determination of cytokine plasma levels 183

Plasma samples from healthy donors and RA patients were recollected and 184

kept at -80ºC until use. IL-17 (Peprotech, London, UK), TNFα, IFNγ, and IL-10 185

(BD Pharmingen) levels were tested using a specific ELISA kit according to the 186

manufacturer’s instructions. All cytokines were quantified with standard curves 187

provided by the corresponding ELISA kit. The limits of detection were: 10 pg/ml 188

for IL-17 and IL-10, 20 pg/ml for TNFα, and 32 pg/ml for IFNγ. 189

190

Flow cytometry analysis 191

Surface expression of CD36 and CD14 was analyzed on monocytes and 192

lymphocytes (gated according to morphology by forward and side scatter 193

parameters). Other markers were analyzed on CD3+, CD4+ and CD8+ positive 194

cells. Naïve and memory T cells were selected according to CD3+ CD45RA+ 195

and CD3+ CD45RA- respectively. These analyses were performed on Beckman 196

Coulter F500 flow cytometer. We calculated the percentages of positive cells (% 197

cells) and geometric mean fluorescence intensity (GMFI) of each individual 198

marker using EXPO™ 32 MiltiCOMP Software (Beckman coulter). The 199

integrated geometric mean fluorescence intensity (iGMFI) determines the 200

amount of cytokines produced in each subpopulation and was calculated by 201

multiplying the percentage of positive cytokine-producing cells by the GMFI [20]. 202

Stimulation index in CFSE culture was measured calculating the ratio of 203

proliferating vs non-proliferating cells [21]. 204

205

206

8

207

Quantification of anti-CCP and Rheumatoid factor levels 208

Anti-cyclic citrullinated peptide antibodies (anti-CCP) were determined in 209

plasma using ELIA Unicap (Phadia Laboratory Systems, Uppasala, Sweden) 210

and Rheumatoid factor (RF) was determined using Nephelometry system 211

(Beckman Coulter). Levels of anti-CCP antibodies and RF were expressed as 212

ELIAunits/ml and IU/ml respectively. Anti-CCP antibodies were positive with 213

levels > 20 ELIAunits/ml and RF were positive with levels >26 IU/ml. 214

215

Statistics 216

Statistical analyses were performed using the probes t-test, Mann-Whitney and 217

Fisher exact test on Graph Pad Prism 5 software. Non-parametric correlation 218

Spearman test was applied to the analysis. Data were presented as mean and 219

± SEM deviation. P values <0.05 were considered significant. 220

221

222

9

Results 223

224

CD36 expression on lymphocytes 225

Most CD14+ monocytes from Ficoll-isolated peripheral blood expressed CD36 226

on their cell surface (90.8 ± 1.83%). Unexpectedly, 6.23 ± 1.23% of 227

lymphocytes (CD14 negative) were also CD36+. To rule out the possibility of 228

unspecific staining of the monoclonal antibody (clone CB38), the analysis was 229

repeated with another clone (TR9: 89.5 ± 1.74% on gated CD14+ monocytes 230

and 6.59 ± 1.55% on lymphocytes) (Fig. 1A). Monocytes and platelets express 231

CD36 on their surface. Since lymphocytes can interact with circulating platelets 232

[22] we speculated that the CD36 expression on lymphocytes was derived from 233

bound CD36+ platelets. Our results showed that those CD36+ lymphocytes 234

were also CD41a+ (a specific marker of platelets) (Fig. 1B). We observed that 235

both CD4+ CD36+ and CD8+ CD36+ lymphocytes were CD41a+. 236

Immunofluorescence microscopy finally confirmed that CD36 expression on 237

lymphocytes was due to bound platelets. In contrast to the whole surface CD36 238

staining of monocytes, the CD36 staining of the CD36+ lymphocyte minority 239

was in a small protruding area of the surface. The image was compatible with a 240

platelet bound to a lymphocyte. A comparable image was observed staining 241

with CD41a+. These observations confirmed that the source of this marker is 242

the presence of platelets bound to lymphocytes (Fig. 1C). We observed that 243

stoichiometry between platelet and T cells were 1 to 1 in the majority of cells 244

(73%). 245

We next analyzed CD36 expression on CD4+ and CD8+ cells from whole blood. 246

We observed a comparable CD36 expression on CD4+ and CD8+ cells from 247

whole blood and Ficoll isolated mononuclear cells (data not shown). This result 248

confirms that the binding between lymphocytes and platelets was neither a 249

separation artifact nor dependent on the proportion of platelets per lymphocyte. 250

CD36 expression on lymphocytes was then a convenient marker to study 251

lymphocytes with bound platelets. 252

253

254

255

10

Phenotyping of CD3 naive and CD3 memory lymphocytes with bound 256

platelets 257

We firstly investigated whether the presence of bound platelets on naïve and 258

memory lymphocytes is restricted to a characteristic phenotype. The 259

percentage of naïve CD3+ lymphocytes with bound platelets was not different 260

from memory CD3+ lymphocytes without platelets (naïve CD3+: 61.1 ± 3.33 of 261

Plt-ng vs 55.4 ± 6.05 % of Plt-pos; memory CD3+: 38.9 ± 5.29 % of Plt-ng vs 262

44.6 ± 6.37 of Plt-pos). A significant higher percentage of CXCR3+ and CCR4+ 263

cells was observed in the naïve CD3+ cells with platelets than in naïve CD3+ 264

cells without platelets (CCR4: 2.33 ± 0.53 of Plt-ng vs 4.71 ± 0.85 of Plt-pos, 265

p=0.02; CXCR3: 8.81 ± 1.93 of Plt-ng vs 13.98 ± 3.47 of Plt-pos, p=0.04). In 266

contrast, a significant higher percentage of CD25+, CCR5+, CCR4+ and HLA-267

DR+ cells were observed in memory CD3+ cells with bound platelets than in 268

memory CD3+ without platelets (CD25: 3.43 ± 0.65 of Plt-ng vs 7.47 ± 1.61 of 269

Plt-pos, p=0.01; HLA-DR: 1.98 ± 0.23 of Plt-ng vs 4.83 ± 0.55 of Plt-pos, 270

p=0.01) and chemokine receptor CCR5 (32.4 ± 3.05 of Plt-ng vs 44.11 ± 4.69 of 271

Plt-pos, p=0.005) and CCR4 (15.37 ± 2.24 of Plt-ng vs 20.86 ± 3.04 of Plt-pos, 272

p=0.04) (Table 1). 273

We next evaluated whether the presence of bound platelets was associated to 274

Th1 (CCR5 and CXCR3) or Th2 (CCR4) markers. Significant higher 275

percentages of CCR5+ were observed in the CD4+ cells with bound platelets 276

than in CD4+ cells without platelets, suggesting that platelet are predominantly 277

bound to cells expressing certain Th1 markers (CCR5: 11.99 ± 1.01 of Plt-ng vs 278

21.59 ± 1.81 of Plt-pos, p=0.0004; CXCR3: 20.44 ± 2.09 of Plt-ng vs 25.76 ± 279

2.74 of Plt-pos, p=0.14). 280

We finally observed no differences in the percentage of naïve, effector and 281

central memory in CD4+ T cells with or without platelets (data not shown). 282

283

Reduced TCR-induced proliferation of CD4+Plt-pos lymphocytes 284

We compared the TCR-induced proliferation of CD4+ Plt-pos and CD4+ Plt-ng 285

cells. PBMC from healthy donors were labelled with CFSE and stimulated with 286

anti-CD3, CD28 and CD2 T cell activation/expansion for 72 hours and 7 days 287

(Fig. 2). After 72 hours and 7 days of culture, we observed that fewer CD4+ Plt-288

11

pos than CD4+ Plt-ng lymphocytes reached a higher number of proliferation 289

cycles (72h, cycle 3: 8.18 ± 3.24% of CD4+ Plt-pos vs 17.67 ± 1.89% of CD4+ 290

Plt-ng, p<0.05; 7d, cycle 4: 7.87 ± 2.9% of CD4+ Plt-pos vs 28 ± 0.33 of CD4+ 291

Plt-ng, p<0.05) (Fig. 2B and 2C). Stimulation index (SI) was also calculated in 292

CD4+ Plt-ng and CD4+ Plt-pos. CD4+ with platelets had a lower SI than CD4+ 293

without platelets (72 hours: 1.49 ± 0.17 for CD4+ Plt-ng vs 0.55 ± 0.17 for CD4+ 294

Plt-pos, p<0.01; 7d: 4.22 ± 0.37 for CD4+ Plt-ng vs 1.74 ± 0.05 for CD4+ Plt-295

pos, p<0.01) (Supplemental figure 1). The binding of platelets to lymphocytes 296

can therefore reduce the proliferation capacity of the CD4+ cells. 297

298

Binding of platelets modified the cytokine production of CD3+ 299

lymphocytes 300

PBMC from healthy donors were stimulated with anti-CD3, CD28 and CD2 T 301

cell activation/expansion kit for 48h. Five hours before cytokine production 302

analysis, PMA/Ionomycin and BD GolgiStop TM Protein Transport Inhibitor was 303

added to the lymphocyte culture. We then compared the intracellular cytokine 304

production of CD3+ Plt-pos and CD3+ Plt-ng lymphocytes. 305

Although there were no differences in the percentage of cells producing IL-17, 306

IFNγ, IL-10 and TNFα, binding platelets modified the amount of cytokine levels 307

in lymphocytes (iGMFI). CD3+ Plt-ng cells expressed higher levels of IL-17 and 308

IFNγ than CD3+ Plt-pos cells (iGMFI IL-17: 67.94 ± 25.89 in CD3+ Plt-ng vs 309

19.5 ± 10.73 in CD3+ Plt-pos, p<0.05; iGMFI IFNγ: 340 ± 97.53 in CD3+ Plt-ng 310

vs 172 ± 57.66 in CD3+ Plt-pos, p<0.05). No differences were observed in IL-10 311

or TNFα levels between the two lymphocyte populations (iGMFI IL-10: 1.36 ± 312

0.49 on CD3+ Plt-ng vs 8.54 ± 3.62 on CD3+ Plt-pos; iGMFI TNFα: 245 ± 137.1 313

on CD3+ Plt-ng vs 153 ± 35.34 on CD3+ Plt-pos) (Fig 3). These results suggest 314

that the binding of platelets decreased the production of Th1 and Th17 315

cytokines. 316

317

CD4+ Plt-pos and cytokine plasma levels from healthy donors and RA 318

patients 319

Activation molecule P-selectin has a significantly higher expression on platelets 320

from active RA patients than from remission RA patients or healthy donors [23]. 321

12

P-selectin can bind to its ligand PSGL1 expressed on leukocytes [24-25]. We 322

observed that platelets bound to lymphocytes in RA patients present higher 323

levels of P-selectin than healthy donors (37.11 ± 6.18 for RA patients vs.13.35 ± 324

1.76 for healthy donors p<0.05) (data not shown). We then compared the 325

percentage of T lymphocytes with bound platelets in the peripheral blood of RA 326

patients and healthy donors. The percentage of CD8+ Plt-pos was similar in 327

healthy donors and RA patients (data not shown). However, the percentage of 328

CD4+ Plt-pos in RA patients was significantly higher than in healthy donors 329

(7.81 ± 0.84% and 4.46 ± 0.51% respectively, p<0.01). According to the CD4+ 330

Plt-pos percentage, we distinguished two groups of RA patients. One group 331

(group-H) presented higher CD4+Plt-pos percentage (10.61 ± 0.68%) than 332

healthy donors. The other group (group-L) presented CD4+ Plt-pos levels (3.79 333

± 0.44%) comparable to healthy donors (4.46 ± 0.51%) (Fig. 4A and 4B). 334

Clinical and laboratory parameters were compared between these two groups 335

of RA patients (Supplemental table 1). Cardiovascular risk index were 336

significantly different between these groups. Plasma IL-17 and IFNγ levels in 337

group-L of RA patients presented higher plasma levels of IL-17 (189.7 ± 70.83 338

pg/ml) and IFNγ (147.3 ± 36.7 pg/ml) than healthy donors and group-H RA 339

patients (IL-17: 38.28 ± 18.24 pg/ml for healthy donors and 51.13 ± 11.26 pg/ml 340

for group-H RA patients, p<0.05; IFNγ: 43.25 ± 20.33 pg/ml for healthy donors 341

and 67.76 ± 15.55 pg/ml for group-H RA patients, p<0.05). These results 342

suggest that increased levels of CD4+ lymphocytes with bound platelets have a 343

regulatory effect in the excessive Th1 and Th17 cytokine levels of RA patients. 344

We did not observe a significant correlation between the proportion of CD4+ Plt-345

pos with serum levels of cytokines in healthy donors. However, a negative 346

correlation was observed between CD4+ Plt-pos and IFNγ in RA patients (r=-347

0.47, p<0.05). 348

349

Percentage of CD4+Plt-pos lymphocytes in RA patients segregated 350

according to clinical parameters 351

We analyzed the association between CD4+ Plt-pos and antibody levels and 352

clinical parameters in RA patients (anti-CCP, RF, CRP, ESR, DAS 28 and score 353

cardiovascular risk index). Negative or moderate-low anti-CCP titer (<340) RA 354

13

patients showed a higher percentage of CD4+ Plt-pos than high anti-CCP titer 355

RA patients (10.01 ± 0.93% for titers <20, 9.98 ± 1.72% for 20-340 titers and 356

5.64 ± 0.64% for titers >340, p<0.05) (Fig 5A). Using a rank based correlation, 357

anti-CCP antibodies inversely correlated with CD4+ Plt-pos (R= -0.443, 358

p=0.05). RA patients with lower RF levels (<26) showed a higher percentage of 359

CD4+ Plt-pos than RA patients with RF levels >160 RF (10.08 ± 1.42% and 360

5.49 ± 1.11% respectively, p<0.05)(Fig 5B). RA patients with lower CRP levels 361

(≤10 mg/L) showed a higher percentage of CD4+ Plt-pos than RA patients with 362

CRP>10 mg/L (9.74 ± 1.5% and 4.19 ± 0.54% respectively, p<0.05) (Fig 5C). 363

No differences in CD4+ Plt-pos percentages were observed when patients were 364

segregated according to ESR levels and DAS28 (Fig 5D and 5E). RA patients 365

with a lower cardiovascular risk index (≤3) presented a higher percentage of 366

CD4+ Plt-pos lymphocytes than patients with higher cardiovascular risk index 367

(>5) (9.01 ± 1.16% and 3.84 ± 0.58% respectively, p<0.05). CD4+ Plt-pos in RA 368

patients segregated according to medication were not significantly different 369

(data not shown) 370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

387

14

388

389

Discussion 390

Our results demonstrated that the CD36 expression on lymphocytes was 391

derived from bound platelets. Platelet binding reduced the capability of 392

lymphocytes to proliferate and to produce Th1/Th17 cytokines. Less IL-17, IFNγ 393

and cardiovascular risk was observed in a group of RA patients with a higher 394

percentage of CD4+ cells with bound platelets. 395

Vallejo et al. demonstrated that T lymphocytes do not express CD36 [2]. 396

However, other reports described a subpopulation of B, T and NK lymphocytes 397

expressing CD36 [3-4]. Some authors have already suggested the presence of 398

platelet-leukocyte complexes [26-27]. Accordingly, we observed that these 399

platelets were the source of the unexpected CD36 expression on this 400

subpopulation of lymphocytes. We were able to finally confirm this result by 401

microscopy. Presence of CD36 on lymphocytes was not due to trogocytosis of 402

membrane chunks of antigen presenting cells because platelet-bound 403

lymphocytes also expressed CD41a, a specific marker of platelets, absent on 404

antigen presenting cells. Most of the CD36+ lymphocytes had one bound 405

platelet (73%). Undergoing analysis of bound-platelet characteristics will provide 406

some explanation to this stoichiometry. 407

Staining lymphocytes with anti-CD36 antibodies proved to be a good tool to 408

analyze the functional consequences of platelet binding. Significant phenotype 409

differences were observed between CD3+ Plt-pos and CD3+ Plt-ng in naïve 410

and memorty T cells. CD3+ Plt-pos memory cells present a higher expression of 411

DR+, CD25+, CCR5+ and CCR4+, while CD3+ Plt-pos naïve present a higher 412

expression of CXCR3+ and CCR4+. We are currently analyzing a predominant 413

subpopulation of CCR5+ CD4+ T lymphocytes among those with bound 414

platelets. 415

Multiple ligands could be involved in this platelet-lymphocyte binding. P-selectin 416

(CD62P) on activated platelets can bind to its ligand PSGL-1 (CD162), 417

expressed on most leukocytes [25]. GPIb-CD11b [28] and GPIIb/IIIa-418

CD11/CD18 [29] could also be involved in platelet-leukocyte binding. It has 419

been shown that platelet-lymphocyte binding, initiated by platelet activation, can 420

be abolished by P-selectin blockade and reduced by inhibition of GPIIb/IIIa, 421

15

CD11b or CD40L [22]. We have found that <15% of platelets bound to 422

lymphocytes are P-selectin positive and this seem contradictory. However, we 423

can not rule out that bound platelet were activated when bound to lymphocyte 424

and then CD62P were released for the surface of platelets due to shedding of 425

this molecule [30]. 426

Only 3% of circulating lymphocytes carried bound platelets. This low percentage 427

could be due to the binding of few free activated platelets on circulation [27]. 428

Accordingly, in vitro activation of platelets increases the lymphocyte-platelet 429

binding (12% for Th lymphocytes and 15% by Tc lymphocytes) [22]. Activation 430

of T lymphocytes also increases the binding of platelets [22] probably because 431

activation upregulates PSGL-1 expression. This upregulation could explain the 432

superior capacity of platelet binding by activated lymphocytes [24]. Accordingly, 433

we could speculate that a higher percentage of CCR5+ CD4+ lymphocytes 434

have bound platelets, because they express more PSGL-1. Previous reports 435

with juvenile idiopathic arthritis patients showed that CCR5+ CD4+ lymphocytes 436

from synovial fluid have higher PSGL-1 expression than CCR4+ and CXCR3+ 437

CD4+ lymphocytes [31]. 438

After activation, lymphocytes with bound platelets have a reduced proliferation 439

and IFNγ and IL-17 production. Previous reports showed that co-culture with 440

activated platelets reduces the activation-induced proliferation of lymphocytes 441

[32-33] and IL-2 and IL-6 release [33]. However, other authors have shown that 442

co-culturing activated platelets with lymphocytes increases the release of IFNγ, 443

TNFα and IL-10 [32]. These reports suggest that soluble molecules released by 444

platelets could be activating or inhibiting factors [34]. One of these molecules, 445

PF4, was able to modulate Treg and non-Treg cells but in opposite manners. 446

PF4 inhibits non-Treg cell proliferation and also reduced Treg ability to inhibit 447

non-Treg proliferation [35]. In our system, we cannot exclude the presence of 448

inhibitory soluble factors released by activated platelets. However, the inhibitory 449

effects were more evident in lymphocytes with bound platelets. Furthermore, 450

our experiments were performed with isolated PBMC after eliminating most free 451

platelets through the Ficoll gradient. 452

Our results suggest that platelet binding to lymphocytes could be relevant in the 453

pathology of inflammatory diseases. We distinguished two groups of RA 454

16

patients according to the percentage of CD4+ Plt-pos. One group with CD4+Plt-455

pos was comparable to healthy donors and the other group had a higher 456

CD4+Plt-pos percentage. This was not surprising because platelets in active 457

RA patients show higher expression of P-selectin than those in remission or 458

healthy controls [23], observing in our experiment that platelets bound to 459

lymphocyte in RA patients present higher percentage of activated platelet than 460

healthy donors. The plasma levels of IL-17 and IFNγ were significantly higher in 461

patients with normal CD4+ Plt percentage. This result was not observed in 462

patients with higher CD4+ Plt-pos percentage, concordant with the lower 463

production of IL-17 and IFNγ by lymphocytes with bound platelets. It has been 464

described that RA pathology is caused partially by an increased Th1/Th17 465

response, with high levels of inflammatory cytokines on serum [36-37]. We 466

suggest that binding of platelets could then be a regulatory mechanism to 467

correct the altered production of inflammatory cytokines in RA patients. 468

RA patients with the highest levels of CRP, rheumatoid factor and anti-CCP 469

antibodies are those with the lowest CD4+Plt-pos levels. Binding of platelets to 470

CD4+ lymphocytes might therefore be associated with a less severe phenotype 471

of RA disease. These results are intriguing because platelets have been shown 472

to participate in the pathogenesis of RA and atherogenesis, a complication of 473

RA [38]. Free activated platelets shed vesicles called microparticles; during 474

inflammatory arthritis, these microparticles are abundant in synovial fluid and 475

can exacerbate inflammation via IL-1. This cytokine can promote release of 476

inflammatory cytokines such as IL6 and IL8 [39]. Moreover, platelets also 477

produce prostaglandins. PGI2 is increased in synovial fluid from the joints of RA 478

patients and contributes to arthritis in animal models [40]. 479

Patients with the highest cardiovascular risk also had the lowest levels of CD4+ 480

Plt-pos. However, it has been described that platelets are associated with 481

cardiovascular risk. CD62P levels on platelets increase in cardiovascular 482

ischemia associated with atherosclerotic complications [41-42]. This apparent 483

contradiction could be explained by a different function between free and bound 484

activated platelets. We should also consider other factors apparently unrelated 485

to platelets, like anti-CCP antibodies and IL-17, which contribute to the 486

development of endothelial dysfunction and cardiovascular disease [10, 43]. In 487

these cases, our patients with the highest levels of CD4+ Plt-pos could have 488

17

had a lower cardiovascular risk because they had a lower production of IL-17 489

and anti-CCP antibodies. 490

Further experiments are needed to decipher the mechanisms of platelet binding 491

and determine how and when these platelets can regulate lymphocyte 492

functions. Similarly, we need to study the influence of platelets on other 493

subpopulations. Ongoing experiments will describe whether this mechanism is 494

RA specific or whether it could be extended to other inflammatory pathologies. 495

The platelet-lymphocyte interaction could open new therapeutic approaches to 496

regulate the chronic inflammation. 497

498

499

500

501

502

503

504

505

506

507

508

509

510

511

512

513

514

515

516

517

518

519

520

521

522

18

523

524

Authorship 525

CZ Flow cytometry, ELISAs, cell culture and data analysis. Manuscript 526

preparation 527

EC Flow cytometry 528

JCN Flow cytometry 529

MAO ELISAs 530

CDT Collection of samples and clinical parameters 531

CDL Collection of samples and clinical parameters 532

JL Collection of samples and clinical parameters 533

CJ Analysis of results 534

SV Analysis of results, conception and design. Manuscript preparation 535

536

Acknowledgements 537

This research was supported by the Ministerio Educacion y Ciencia 538

(BFU06/15063) and Fondo de Investigaciones Sanitarias (PS09/132). S.V. was 539

supported by “Fondo Investigaciones Sanitarias” and participant in the Program 540

for Stabilization of Investigators of the “Direcció d'Estrategia i Coordinació del 541

Departament Salut de la Generalitat de Catalunya.” 542

We thank Carolyn Newey for helpful editorial assistance and Ignasi Gich for his 543

statistical assistance. 544

This work has been performed within the Ph.D Immunology program of 545

Universitat Autònoma de Barcelona (UAB). 546

547

Conflicts of Interest disclosure 548

The authors declare no financial or commercial conflicts of interest associated 549

with this study. 550

551

552

553

554

555

556

19

557

558

559

References 560

1. Febbraio, M., Hajjar, D. P., Silverstein, R. L. (2001) CD36: a class B scavenger 561

receptor involved in angiogenesis, atherosclerosis, inflammation, and lipid 562

metabolism. J Clin Invest 108, 785-91. 563

2. Vallejo, A. N., Mugge, L. O., Klimiuk, P. A., Weyand, C. M., Goronzy, J. J. 564

(2000) Central role of thrombospondin-1 in the activation and clonal expansion 565

of inflammatory T cells. J Immunol 164, 2947-54. 566

3. McDonnell, M., Liang, Y., Noronha, A., Coukos, J., Kasper, D. L., Farraye, F. 567

A., Ganley-Leal, L. M. (2011) Systemic Toll-like receptor ligands modify B-cell 568

responses in human inflammatory bowel disease. Inflamm Bowel Dis 17, 298-569

307. 570

4. Lubick, K. and Jutila, M. A. (2006) LTA recognition by bovine gammadelta T 571

cells involves CD36. J Leukoc Biol 79, 1268-70. 572

5. Tassone, P., Tuccillo, F., Bonelli, P., Turco, M. C., Cecco, L., Cerra, M., Bond, 573

H. M., Barbieri, V., Venuta, S. (1998) CD36 is rapidly and transiently 574

upregulated on phytohemagglutinin (PHA)-stimulated peripheral blood 575

lymphocytes. Analysis by a new monoclonal antibody (UN7). Tissue Antigens 576

51, 671-5. 577

6. Harrington, L. E., Hatton, R. D., Mangan, P. R., Turner, H., Murphy, T. L., 578

Murphy, K. M., Weaver, C. T. (2005) Interleukin 17-producing CD4+ effector T 579

cells develop via a lineage distinct from the T helper type 1 and 2 lineages. Nat 580

Immunol 6, 1123-32. 581

7. Bettelli, E., Carrier, Y., Gao, W., Korn, T., Strom, T. B., Oukka, M., Weiner, H. 582

L., Kuchroo, V. K. (2006) Reciprocal developmental pathways for the 583

generation of pathogenic effector TH17 and regulatory T cells. Nature 441, 235-584

8. 585

8. Kotake, S., Udagawa, N., Takahashi, N., Matsuzaki, K., Itoh, K., Ishiyama, S., 586

Saito, S., Inoue, K., Kamatani, N., Gillespie, M. T., Martin, T. J., Suda, T. (1999) 587

IL-17 in synovial fluids from patients with rheumatoid arthritis is a potent 588

stimulator of osteoclastogenesis. J Clin Invest 103, 1345-52. 589

9. Pickens, S. R., Volin, M. V., Mandelin, A. M., 2nd, Kolls, J. K., Pope, R. M., 590

Shahrara, S. (2010) IL-17 contributes to angiogenesis in rheumatoid arthritis. J 591

Immunol 184, 3233-41. 592

20

10. Marder, W., Khalatbari, S., Myles, J. D., Hench, R., Yalavarthi, S., Lustig, S., 593

Brook, R., Kaplan, M. J. (2011) Interleukin 17 as a novel predictor of vascular 594

function in rheumatoid arthritis. Ann Rheum Dis 70, 1550-5. 595

11. Maggi, L., Santarlasci, V., Capone, M., Peired, A., Frosali, F., Crome, S. Q., 596

Querci, V., Fambrini, M., Liotta, F., Levings, M. K., Maggi, E., Cosmi, L., 597

Romagnani, S., Annunziato, F. (2010) CD161 is a marker of all human IL-17-598

producing T-cell subsets and is induced by RORC. Eur J Immunol 40, 2174-81. 599

12. Chalan, P., Abdulahad, W., Huitema, M. G., Kroesen, B. J., Brouwer, E., Boots, 600

A. M. H. (2011) Rheumatoid arthritis patients shown a decline in peripheral 601

blood CD4+CD161+T lymphocytes ANNALS OF THE RHEUMATIC DISEASES 602

70, A6-A6 603

13. Ruschpler, P. and Stiehl, P. (2002) Shift in Th1 (IL-2 and IFN-gamma) and Th2 604

(IL-10 and IL-4) cytokine mRNA balance within two new histological main-types 605

of rheumatoid-arthritis (RA). Cell Mol Biol (Noisy-le-grand) 48, 285-93. 606

14. Sundstedt, A., O'Neill, E. J., Nicolson, K. S., Wraith, D. C. (2003) Role for IL-10 607

in suppression mediated by peptide-induced regulatory T cells in vivo. J 608

Immunol 170, 1240-8. 609

15. Xu, D., Liu, H., Komai-Koma, M., Campbell, C., McSharry, C., Alexander, J., 610

Liew, F. Y. (2003) CD4+CD25+ regulatory T cells suppress differentiation and 611

functions of Th1 and Th2 cells, Leishmania major infection, and colitis in mice. J 612

Immunol 170, 394-9. 613

16. Kawashiri, S. Y., Kawakami, A., Okada, A., Koga, T., Tamai, M., Yamasaki, S., 614

Nakamura, H., Origuchi, T., Ida, H., Eguchi, K. (2011) 615

CD4+CD25(high)CD127(low/-) Treg cell frequency from peripheral blood 616

correlates with disease activity in patients with rheumatoid arthritis. J 617

Rheumatol 38, 2517-21. 618

17. Arnett, F. C., Edworthy, S. M., Bloch, D. A., McShane, D. J., Fries, J. F., 619

Cooper, N. S., Healey, L. A., Kaplan, S. R., Liang, M. H., Luthra, H. S., et al. 620

(1988) The American Rheumatism Association 1987 revised criteria for the 621

classification of rheumatoid arthritis. Arthritis Rheum 31, 315-24. 622

18. Aletaha, D. and Smolen, J. S. (2006) The definition and measurement of 623

disease modification in inflammatory rheumatic diseases. Rheum Dis Clin North 624

Am 32, 9-44, vii. 625

19. Conroy, R. M., Pyorala, K., Fitzgerald, A. P., Sans, S., Menotti, A., De Backer, 626

G., De Bacquer, D., Ducimetiere, P., Jousilahti, P., Keil, U., Njolstad, I., 627

Oganov, R. G., Thomsen, T., Tunstall-Pedoe, H., Tverdal, A., Wedel, H., 628

Whincup, P., Wilhelmsen, L., Graham, I. M. (2003) Estimation of ten-year risk of 629

21

fatal cardiovascular disease in Europe: the SCORE project. Eur Heart J 24, 630

987-1003. 631

20. Darrah, P. A., Patel, D. T., De Luca, P. M., Lindsay, R. W., Davey, D. F., Flynn, 632

B. J., Hoff, S. T., Andersen, P., Reed, S. G., Morris, S. L., Roederer, M., Seder, 633

R. A. (2007) Multifunctional TH1 cells define a correlate of vaccine-mediated 634

protection against Leishmania major. Nat Med 13, 843-50. 635

21. Milovanova, T. N. (2007) Comparative analysis between CFSE flow cytometric 636

and tritiated thymidine incorporation tests for beryllium sensitivity. Cytometry B 637

Clin Cytom 72, 265-75. 638

22. Li, N., Ji, Q., Hjemdahl, P. (2006) Platelet-lymphocyte conjugation differs 639

between lymphocyte subpopulations. J Thromb Haemost 4, 874-81. 640

23. Wang, F., Wang, N. S., Yan, C. G., Li, J. H., Tang, L. Q. (2007) The 641

significance of platelet activation in rheumatoid arthritis. Clin Rheumatol 26, 642

768-71. 643

24. Moore, K. L., Stults, N. L., Diaz, S., Smith, D. F., Cummings, R. D., Varki, A., 644

McEver, R. P. (1992) Identification of a specific glycoprotein ligand for P-645

selectin (CD62) on myeloid cells. J Cell Biol 118, 445-56. 646

25. Furie, B. and Furie, B. C. (1995) The molecular basis of platelet and endothelial 647

cell interaction with neutrophils and monocytes: role of P-selectin and the P-648

selectin ligand, PSGL-1. Thromb Haemost 74, 224-7. 649

26. Furman, M. I., Benoit, S. E., Barnard, M. R., Valeri, C. R., Borbone, M. L., 650

Becker, R. C., Hechtman, H. B., Michelson, A. D. (1998) Increased platelet 651

reactivity and circulating monocyte-platelet aggregates in patients with stable 652

coronary artery disease. J Am Coll Cardiol 31, 352-8. 653

27. Joseph, J. E., Harrison, P., Mackie, I. J., Isenberg, D. A., Machin, S. J. (2001) 654

Increased circulating platelet-leucocyte complexes and platelet activation in 655

patients with antiphospholipid syndrome, systemic lupus erythematosus and 656

rheumatoid arthritis. Br J Haematol 115, 451-9. 657

28. Simon, D. I., Chen, Z., Xu, H., Li, C. Q., Dong, J., McIntire, L. V., Ballantyne, C. 658

M., Zhang, L., Furman, M. I., Berndt, M. C., Lopez, J. A. (2000) Platelet 659

glycoprotein ibalpha is a counterreceptor for the leukocyte integrin Mac-1 660

(CD11b/CD18). J Exp Med 192, 193-204. 661

29. Silverstein, R. L., Asch, A. S., Nachman, R. L. (1989) Glycoprotein IV mediates 662

thrombospondin-dependent platelet-monocyte and platelet-U937 cell adhesion. 663

J Clin Invest 84, 546-52. 664

30. Dole, V. S., Bergmeier, W., Patten, I. S., Hirahashi, J., Mayadas, T. N., Wagner, 665

D. D. (2007) PSGL-1 regulates platelet P-selectin-mediated endothelial 666

22

activation and shedding of P-selectin from activated platelets. Thromb Haemost 667

98, 806-12. 668

31. Issekutz, A. C., Quinn, P. J., Lang, B., Ramsey, S., Huber, A. M., Rowter, D., 669

Karkada, M., Issekutz, T. B. (2011) Coexpression of chemokine receptors 670

CCR5, CXCR3, and CCR4 and ligands for P- and E-selectin on T lymphocytes 671

of patients with juvenile idiopathic arthritis. Arthritis Rheum 63, 3467-76. 672

32. Gerdes, N., Zhu, L., Ersoy, M., Hermansson, A., Hjemdahl, P., Hu, H., 673

Hansson, G. K., Li, N. (2011) Platelets regulate CD4(+) T-cell differentiation via 674

multiple chemokines in humans. Thromb Haemost 106, 353-62. 675

33. Wang, Y. and Niu, J. (2008) Platelets inhibit in vitro response of lymphocytes to 676

mitogens. Immunol Lett 119, 57-61. 677

34. Fleischer, J., Grage-Griebenow, E., Kasper, B., Heine, H., Ernst, M., Brandt, E., 678

Flad, H. D., Petersen, F. (2002) Platelet factor 4 inhibits proliferation and 679

cytokine release of activated human T cells. J Immunol 169, 770-7. 680

35. Liu, C. Y., Battaglia, M., Lee, S. H., Sun, Q. H., Aster, R. H., Visentin, G. P. 681

(2005) Platelet factor 4 differentially modulates CD4+CD25+ (regulatory) versus 682

CD4+CD25- (nonregulatory) T cells. J Immunol 174, 2680-6. 683

36. Tukaj, S., Kotlarz, A., Jozwik, A., Smolenska, Z., Bryl, E., Witkowski, J. M., 684

Lipinska, B. (2010) Cytokines of the Th1 and Th2 type in sera of rheumatoid 685

arthritis patients; correlations with anti-Hsp40 immune response and diagnostic 686

markers. Acta Biochim Pol 57, 327-32. 687

37. Ziolkowska, M., Koc, A., Luszczykiewicz, G., Ksiezopolska-Pietrzak, K., 688

Klimczak, E., Chwalinska-Sadowska, H., Maslinski, W. (2000) High levels of IL-689

17 in rheumatoid arthritis patients: IL-15 triggers in vitro IL-17 production via 690

cyclosporin A-sensitive mechanism. J Immunol 164, 2832-8. 691

38. Boilard, E., Blanco, P., Nigrovic, P. A. (2012) Platelets: active players in the 692

pathogenesis of arthritis and SLE. Nat Rev Rheumatol 8, 534-42. 693

39. Boilard, E., Nigrovic, P. A., Larabee, K., Watts, G. F., Coblyn, J. S., Weinblatt, 694

M. E., Massarotti, E. M., Remold-O'Donnell, E., Farndale, R. W., Ware, J., Lee, 695

D. M. (2010) Platelets amplify inflammation in arthritis via collagen-dependent 696

microparticle production. Science 327, 580-3. 697

40. Honda, T., Segi-Nishida, E., Miyachi, Y., Narumiya, S. (2006) Prostacyclin-IP 698

signaling and prostaglandin E2-EP2/EP4 signaling both mediate joint 699

inflammation in mouse collagen-induced arthritis. J Exp Med 203, 325-35. 700

41. McCabe, D. J., Harrison, P., Sidhu, P. S., Brown, M. M., Machin, S. J. (2004) 701

Circulating reticulated platelets in the early and late phases after ischaemic 702

stroke and transient ischaemic attack. Br J Haematol 126, 861-9. 703

23

42. Yamazaki, M., Uchiyama, S., Iwata, M. (2001) Measurement of platelet 704

fibrinogen binding and p-selectin expression by flow cytometry in patients with 705

cerebral infarction. Thromb Res 104, 197-205. 706

43. Lopez-Longo, F. J., Oliver-Minarro, D., de la Torre, I., Gonzalez-Diaz de 707

Rabago, E., Sanchez-Ramon, S., Rodriguez-Mahou, M., Paravisini, A., 708

Monteagudo, I., Gonzalez, C. M., Garcia-Castro, M., Casas, M. D., Carreno, L. 709

(2009) Association between anti-cyclic citrullinated peptide antibodies and 710

ischemic heart disease in patients with rheumatoid arthritis. Arthritis Rheum 61, 711

419-24. 712

713

714

715

716

24

Figure Legends 717

718

Fig. 1. CD36 expression on monocytes and lymphocytes after Ficoll 719

isolation. PBMC from healthy donors (n=10) were Ficoll gradient isolated. Cells 720

were then stained with anti CD3, CD4, CD8, CD14, CD36 (clone CB38,and 721

clone TR9), and CD41a. (A) Expression of CD14-PEDy647, CD36-PE and 722

CD36-FITC were analyzed on monocytes and lymphocytes gated according to 723

FS and SS parameters and (B) Expression of CD41a-FITC and CD36-PE were 724

analyzed on CD4+ and CD8+ lymphocytes by flow cytometry gated on CD3+ 725

lymphocytes. (C) CD36 expression was analyzed on CD3+ and CD14+ cells 726

and CD41a on CD3+ cells using fluorescence microscopy. A representative 727

experiment is shown. Data are expressed as percentage ± SEM of analyzed 728

marker. 729

730

Fig. 2. Influence of binding platelet to CD4+ lymphocytes on activation-731

induced proliferation. PBMC from healthy donors were labelled with 7 μM of 732

CFSE. PBMC were then cultured with anti-CD3 plus CD2 and CD28 or medium 733

for 72h and 7d. After culturing, PBMC were collected and stained with anti CD4-734

PECy7 and CD36-PE. (A) Representative experiment of history proliferation 735

cycles on CD4+ Plt-ng and CD4+ Plt-pos cells after 72h and 7d. (B) Bar graph 736

shown percentage of cells in each proliferation cycle after 72h and (C) 7 days of 737

culture with anti-CD3 plus CD2 and CD28 on CD4+ Plt-ng and CD4+ Plt-pos 738

cells (n=5). Mann-Whitney test was used for statistical analysis. * P <0.05 and 739

** P<0.01. 740

741

Fig. 3. Influence of binding platelet to CD3+ lymphocytes on cytokine 742

production. PBMC from healthy donors were cultured with anti-CD3 plus CD2 743

and CD28 or medium. After 48h, cells were cultured with PMA (50 744

ng/ml)/Ionomycin (1 μg/ml) in presence of BD GolgiStopTM Protein Transport 745

Inhibitor for five hours before analysis of cytokine production. (A) IL-17, (B) IL-746

10 (C) IFNγ and (D) TNFα iGMFI production cells was calculated gating on 747

CD3+ Plt-ng and CD3+ Plt-pos lymphocytes (n=5). Data are expressed as 748

iGMFI ± SEM. Mann-Whitney test was used for statistical analysis. * P<0.05. 749

25

Fig. 4. CD4+ Plt-pos and cytokine plasma levels in RA patients. PBMC 750

from healthy donors and RA patients were stained with anti CD4-PECy7 and 751

CD36-PE. (A) Graph shown percentage of Plt-pos ± SEM on CD4+ 752

lymphocytes from healthy donors (n=16) and RA patients (n=20). (B) 753

Representative experiment of CD36 expression on CD4+ lymphocytes from 754

healthy controls and groupH and groupL RA patients. (C) Bar graph showing IL-755

17, IFNγ and IL-10 plasma levels from healthy donors and RA patients (pg/ml). 756

T-test was used for statistical analysis. * P<0.05 ** P<0.01. 757

758

Fig. 5. CD4+ Plt-pos percentage according to levels of CCP, rheumatoid 759

factor, C-reactive protein, ESR, DAS28 and cardiovascular risk in RA 760

patients. We analyzed the percentage of CD4+ Plt-pos in those RA patients 761

segregated according to (A) anti-CCP, (B) RF, (C) CRP, (D) ESR, (E) DAS28 762

and (F) cardiovascular risk index. Data are expressed as mean of percentage ± 763

SEM. T-test was used for statistical analysis * P<0.05 764

CD3 naive* CD3 memory** Plt- Plt+ Plt- Plt+

Marker Mean ± SEM Mean ± SEM p value Mean ± SEM Mean ± SEM p value

CD25 0.86 ± 0.34 0.79 ± 0.24 0.86 3.43 ± 0.65 7.47 ± 1.61 0.01

CCR5 10.48 ± 1.76 17.16 ± 4.36 0.22 32.4 ± 3.05 44.11 ± 4.69 0.005

CCR6 4.06 ± 1.14 4.92 ± 1.01 0.28 35.42 ± 7.44 32.37 ± 6.53 0.09

CXCR3 8.81 ± 1.93 13.98 ± 3.47 0.04 35.32 ± 6.76 34.97 ± 6.96 0.86

CCR4 2.33 ± 0.53 4.71 ± 0.85 0.02 15.37 ± 2.24 20.86 ± 3.04 0.04

HLA-DR 1.11 ± 0.55 2.63 ± 1.06 0.08 1.98 ± 0.23 4.83 ± 0.55 0.01 *Percentage of each marker in the CD3+ naive cells with or without platelets (n=5) ** Percentage of each marker in the CD3+ memory with or without platelets (n=5) p value was calculated with Mann Whitney test

CD

14-P

EDy6

47C

D36

-PE

CD

36-F

ITC

Cell CountsG

ated

on

mon

ocyt

es

Gat

edon

ly

mph

ocyt

es

CD36 PE

CD

41a

FITC

Gat

edon

C

D4+

lym

phoc

ytes

CD

3 FI

TCC

D36

PE

Mer

ge

A.

B.

82.1

4 ±

3.16

%

CD

14 F

ITC

CD

36 P

EM

erge

C.

Gat

edon

C

D8+

lym

phoc

ytes

0.87 %

2.4

%

0.7

%

0.46 %

1.55

%

0.6

%

Isot

ype

co

ntro

l

0.09

±0.

02 %

90.8

±1.

83 %

6.23

±1.

23 %

89.7

5 ±

1.74

%

6.59

±1.

55 %

CD

3 FI

TCC

D41

a PE

Mer

ge

CD

4+ P

lt-ng

CD

4+ P

lt-po

s

72h

a-C

D3

+ C

D28 +

CD

2

Cell Counts

Med

ium

12

31

23

44

A.

B.

C.

12

354

61

23

546

CFS

E

7d a-C

D3

+ C

D28 +

CD

2

020406080100

*

*

Plt-

pos

Plt-

ng

*

4

3

2

1

% CD4+ CELLS

0204060Pl

t-ng

Plt-

pos

**

*

*

** 6

5

4

3

2

1

% CD4+ CELLS

Plt-ng Plt-pos0

20

40

60

80

100*

iGM

FI IL

-17

Plt-ng Plt-pos0

5

10

150.09

iGM

FI IL

-10

Plt-ng Plt-pos0

100

200

300

400

500*

iGM

FI IF

Nγ

A.

Plt-ng Plt-pos0

100

200

300

400

500

iGM

FI T

NFα

B.

C. D.

CD4+ Cell counts

CD

36-P

E

4.14

%

Hea

lthy

dono

rsR

A G

roup

LR

A G

roup

H

12.2

9 %

5.03

%

A.

B.

Gro

up H

Gro

up L

C.H

ealth

y do

nors

RA

pat

ient

s05101520

**CD

4 +

% Plt-pos

IFNγ

Health

y donors

RA Gro

up HRA G

roup L

050100

150

200

**

pg/mlIL

-17

Health

y donors

RA Gro

up HRA G

roup L

0

100

200

300

*0.

07

pg/ml

IL-1

0

Health

y donors

RA Gro

up HRA G

roup L

0510152025

pg/ml

A.

C.

D.

F.E.

B.

<20

20-3

40>3

4005101520

Anti-

CCP *

*

% CD4+ Plt-pos

DAS2

8

<2.

62.

6-4

>405101520

% CD4+ Plt-pos

C-re

activ

e pr

otein

<44-

10>1

005101520

*

% CD4+ Plt-pos

Rheu

mat

oid

fact

or

< 2

626

-160

>160

05101520*

% CD4+ Plt-pos

ESR

<66-

25>2

505101520

% CD4+ Plt-pos

Car

diov

asul

ar ri

sk in

dex

< 3

3-5

> 5

05101520*

% CD4+ Plt-pos