http://informahealthcare.com/plt ISSN: 0953-7104 (print), 1369-1635 (electronic) Platelets, 2014; 25(8): 639–642 ! 2014 Informa UK Ltd. DOI: 10.3109/09537104.2013.852660
SHORT COMMUNICATION
Circulating levels of soluble EMMPRIN (CD147) correlate with levels of soluble glycoprotein VI in human plasma G. J. Pennings1,2, A. S. C. Yong1,2, C. Wong1,2, M. Al-Tamimi3,4, E. E. Gardiner3, R. K. Andrews3, & L. Kritharides1,2 Vascular Biology Group, The ANZAC Research Institute, Concord Repatriation General Hospital, Hospital Rd, Concord, NSW, Australia, 2The University of Sydney, Sydney, NSW, Australia, 3Australian Centre for Blood Diseases, Monash University, Melbourne, Australia, and 4Department of Immunology and Microbiology, Hashemite University, Zarqa, Jordan
Platelets Downloaded from informahealthcare.com by University of Colorado Libraries on 12/30/14 For personal use only.
1
Abstract
Keywords
Extracellular matrix metalloproteinase inducer (EMMPRIN; CD147), which binds to the plateletspecific collagen receptor glycoprotein (GP) VI, is expressed in a range of cell types including platelets and leukocytes, and has been implicated in neoplastic disease and atherosclerotic coronary disease. Both CD147 and GPVI can be shed from cell membranes and detected in plasma. However, while the relationship between soluble CD147 (sCD147), soluble GPVI (sGPVI) and standard markers of platelet activation has received little attention, such analysis may help reveal pathways mediating release of sCD147. We investigated the relationship between sCD147 and platelet markers including sGPVI, soluble and platelet-bound CD62P (P-selectin), active aIIbb3 (assessed by PAC-1 binding) and platelet CD147 in 25 patients with stable angina pectoris (SAP), 13 patients with no coronary artery disease (CAD) and 10 healthy donors. Plasma levels of sCD147 significantly correlated with sGPVI (r ¼ 0.46, p ¼ .004), but did not correlate with any other platelet markers examined. Linear regression analysis identified that sCD147 levels could be predicted by sGPVI levels ( ¼ .445, p ¼ 0.003) and age ( ¼ 0.304, p ¼ 0.038), but were independent of potential clinical confounders such as CAD, diabetes and medication usage. As sCD147 strongly correlates with platelet-specific sGPVI, a common platelet source and/or mechanism of release may contribute to sCD147 levels in vivo.
EMMPRIN, platelet, soluble CD147, soluble GPVI
Introduction CD147 is a member of the immunoglobulin superfamily expressed on the surface of platelets, haematopoietic cells, endothelial cells and tumour cells. Surface expression of CD147 is elevated on circulating monocytes in myocardial infarct patients [1] and on circulating platelets, monocytes, granulocytes and lymphocytes in patients with coronary artery disease (CAD) [2]. In tumour cells, CD147 may be exported in microvesicles [3]; however, the release of soluble CD147 (sCD147) in the vasculature also involves proteolytic cleavage by matrix metalloproteinases (MMPs) [4]. Whether either of these mechanisms of release occurs in platelets or leukocytes is as yet unknown. The platelet collagen receptor, GPVI, also a member of the immunoglobulin superfamily, is constitutively expressed exclusively on platelets, and is co-associated with the Fc receptor g-chain (FcRg) required for stable surface expression [5]. GPVI mediates thrombus formation at sites of vascular injury [6]. Binding of collagen or other GPVI ligands to the GPVI/FcRg complex results in platelet activation, aggregation and metalloproteinase-dependent shedding of GPVI from the platelet surface [7]. The resultant 55 kDa soluble fragment of GPVI (sGPVI,) is detectable in human plasma and can be assessed by ELISA [8].
History Received 23 August 2013 Revised 29 September 2013 Accepted 2 October 2013 Published online 18 November 2013
CD147 has been found in atherosclerotic plaques [9] and elevated levels have been observed in platelet and leukocytes from patients with CAD [1, 2]. Levels of sCD147 are increased in serum from patients with inflammatory diseases such as systemic sclerosis [10] but have not been examined in CAD in tandem with other platelet markers that are known to be upregulated such as sCD62P (soluble P-selectin), activated integrin aIIbb3 (GPIIbIIIa) or sGPVI [11, 12]. Examining the relationship between sCD147 levels and markers of platelet activation such as increased CD62P expression on platelets, elevated plasma sCD62P, activation of aIIbb3 (detected by PAC-1 binding) and elevated plasma sGPVI may provide us with important information regarding the cellular origin of sCD147 in inflammatory disease states. In 2009, Seizer and colleagues demonstrated for the first time that CD147 was a receptor for GPVI [13]. As the relationship between sCD147 and sGPVI (two receptor fragments derived from metalloproteolysis) is not known, the aims of this study were to investigate the relationship between sCD147 and sGPVI as well as standard platelet activation markers and to investigate potential clinical characteristics that may affect sCD147 levels.
Methods Reagents
Correspondence: Prof. Leonard Kritharides, Department of Cardiology, Concord Repatriation General Hospital, Hospital Road, Concord, NSW 2139, Australia. Tel: +61 2 9767 6296. Fax: +61 2 9767 6994. E-mail: [email protected]
Preconjugated fluorescent antibodies CD42b-PE, CD62P-PE, IgG1-PE, IgM-FITC and PAC-1-FITC were from BD Biosciences, Franklin Lakes, NJ. Dulbecco’s modified phosphate buffered saline (DPBS) and Hank’s buffered salt solution
640
G. J. Pennings et al.
Platelets, 2014; 25(8): 639–642
Platelets Downloaded from informahealthcare.com by University of Colorado Libraries on 12/30/14 For personal use only.
Table I. Clinical characteristics.
Gender Male Female Risk Factors for CAD Hypertension Family history Hypercholesterolemia Diabetes Smoking (former and current) Medication Aspirin Clopidogrel Warfarin Beta-blocker ACE inhibitor/angiotensin II Anti-lipid Nitrates Age - Mean SD
n
%
30 18
62 38
22 9 30 11 17
46 19 63 23 35
30 63 21 44 2 4 15 31 20 42 25 52 13 27 60 13 yr
(10x, HBSS) were from Invitrogen, Carlsbad, CA. Formaldehyde (16% solution) was from ProSciTech, Kirwan, Queensland. Nonpyrogenic, endotoxin-free water and 0.9% sodium chloride solution were from Baxter, Deerfield, IL. Patient selection and blood collection A total of 48 subjects (mean age 60 2 years) comprising 25 unselected patients with confirmed stable angina pectoris (SAP) and 23 controls were recruited from Concord Repatriation General Hospital. Control subjects (n ¼ 23, 52 18 years) were recruited from healthy volunteers (n ¼ 10) or from patients without significant CAD (530% stenosis) upon conventional coronary angiography or computed tomography (n ¼ 13) [2]. Patients (n ¼ 25, 66 9 years) were classified as SAP if they presented with clinically stable angina and significant CAD (450% stenosis) upon conventional coronary angiography (Table I). Blood (15 mL) was collected by venipuncture by gentle aspiration without a tourniquet using a 20G intravenous cannula as previously described [2, 14, 15]. Briefly, the first 5 mL was discarded. The remaining 10 mL blood was distributed between EDTA and CTAD (buffered sodium citrate theophylline adenosine dipyridamole) blood collection tubes (BD Biosciences, Franklin Lakes, NJ). Tubes and samples were kept at 4 C throughout sample transport and processing. The EDTA and CTAD tubes were centrifuged at 1000 g for 15 min at 4 C; the plasma was then removed and stored at 80 C for assessment of soluble markers by ELISA. This study was approved by the institutional human research ethics committee. Written informed consent was obtained from all subjects.
correction at 600 nm. The lower limit of detection, intra-assay CV and inter-assay CV were as follows: sCD147 (7.6 pg/mL, 4.5% and 8.5%), sGPVI (1 ng/mL, 4.4%, 11.6%) [8] and sCD62P (0.5 ng/mL, 5.2%, 8.9%). Flow cytometry Flow cytometry was performed as previously described under resting conditions [15–18]. Briefly, DPBS and antibodies (isotype control – IgG or IgM; or platelet markers – CD42b, CD62P or PAC-1) were aliquoted into the appropriate tubes with 5 mL of CTAD anticoagulated blood (final volume, 50 mL). GPVI expression was determined by Zenon labelling as per the manufacturer’s instructions (Invitrogen, Carlsbad, CA) of anti-GPVI mAb (clone 1A12) and processed in the same manner as the other antibodies [19]. After 20 min at 4 C, samples were fixed with ice-cold 0.16% formaldehyde solution in saline. Acquisition was performed on a BD FACSCalibur (BD Biosciences). The platelet population was collected based on light scatter characteristics and confirmed by expression of CD42b. The results were analyzed with WinMDI 2.9 flow cytometry analysis software (The Scripps Institute, La Jolla, CA). The isotype control was used to establish positive events based on a gate/marker of 0.5% background expression. Statistics The D’Agostino and Pearson omnibus normality test and histograms were used to verify normality of the data. Correlations between sCD147 and sGPVI or platelet markers were performed using Pearson’s correlation. Independent samples t-test and linear regression (SPSS Statistics v20.0; SPSS Inc., Chicago, IL) were used to establish whether clinical characteristics/markers and/or medication may have an effect on sCD147 expression and the order in which to include them in a linear regression model to determine the predictors of sCD147 expression.
Results Correlations of sCD147 with markers of platelet activation and clinical characteristics There was a clear correlation between levels of sCD147 and sGPVI (r ¼ 0.456, p ¼ 0.004) (Figure 1). No significant correlations were observed between sCD147 and markers of platelet activation, including sCD62P, platelet-bound CD62P, PAC-1 binding and platelet-bound CD147 (Table II). No correlations were observed between sCD147 and white cell count (WCC), platelet count or conventional markers of systemic inflammation [neutrophil-lymphocyte ratio (NLR) or C-reactive protein (CRP)] (Table II). sGPVI levels correlated with sCD147 (as above), sCD62P (r ¼ 0.448, p ¼ 0.002, n ¼ 44), membrane-bound GPVI (r ¼ 0.530, p ¼ 0.020, n ¼ 19) and white cell count (r ¼ 0.422, p ¼ 0.004, n ¼ 44). No correlation was observed between sGPVI and other markers of platelet activation.
Measurement of sCD147 and other enzyme linked immunosorbent assays
Univariate analysis and linear regression
sCD147 was assessed using a commercially available ELISA antibody pair (R&D Systems, Minneapolis, MN) and optimized for use with human EDTA anticoagulated plasma. Levels of sCD147 were assessed as per the manufacturer’s instructions, with diluted plasma (1/100) used in place of cell lysates. sGPVI was assessed in EDTA anticoagulated plasma as previously described [8]. sCD62P was assessed using a commercially available ELISA kit according to the manufacturer’s instructions (R&D Systems, Minneapolis, MN) using CTAD anticoagulated plasma. sCD147 and sCD62P plates were read on a Wallac-Victor 3 plate reader (PerkinElmer, Walthan, MA) at 450 nm with a
Clinical characteristics were assessed for their potential effect on sCD147 and on sGPVI expression to look for confounders of the correlation (Tables I and III). Variables with a p value less than or equal to 0.1 were assessed for inclusion in the model. CAD was associated with both sCD147 (very weakly) and sGPVI (Table III). Age and clopidogrel weakly associated with sCD147, whereas diabetes and anti-lipid medication associated with sGPVI (Table III). Only sGPVI ( ¼ 0.445, p ¼ 0.003) and age ( ¼ 0.304, p ¼ 0.038) significantly affected sCD147 expression based on assessment by linear regression. The overall model fit was
sCD147 correlates with sGPVI
DOI: 10.3109/09537104.2013.852660
641
Table III. Univariate analysis for soluble CD147 and soluble GPVI. Soluble CD147 (ng/mL)
Platelets Downloaded from informahealthcare.com by University of Colorado Libraries on 12/30/14 For personal use only.
Gender M F CAD þ Age r
Figure 1. Soluble CD147 increase with increasing soluble GPVI. Levels of soluble CD147 correlate strongly with soluble GPVI levels (r ¼ 0.46, p ¼ 0.004, n ¼ 38) analyzed with Pearson’s correlation as assessed in a subset of the cohort (includes SAP patients, patients without significant CAD and healthy controls). Table II. Correlations for soluble CD147 and soluble GPVI. Soluble CD147
Platelet Markers Soluble GPVI Soluble CD147 Soluble CD62P Platelet CD62P PAC-1 binding Platelet CD147 Platelet GPVI Haematological Parameters White cell count Platelet count Neutrophil-lymphocyte ratio C-reactive protein
Soluble GPVI
r
p Value
n
r
p Value
n
0.456 – 0.231 0.143 0.300 0.164 0.202
0.004** – 0.162 0.393 0.067 0.326 0.576
38 – – – 0.456 0.004** 38 0.448 0.002** 38 0.159 0.282 38 0.021 0.885 38 0.036 0.810 10 0.530 0.020*
– 38 44 48 48 48 19
0.252 0.127 0.106 0.126
0.138 0.454 0.551 0.499
36 37 34 31
44 45 39 36
0.422 0.066 0.087 0.016
0.004** 0.666 0.597 0.928
*p50.05, **p50.01.
r2 ¼ 0.300. The following equation was determined for predicting sCD147 levels: sCD147 ¼ 2:45 þ ð0:16 sGPVIÞ þ ð0:27 AgeÞ: To ensure there were no clinical confounders, we also repeated the linear regression including four parameters that had some association with either sCD147 or sGPVI [clopidogrel usage ( ¼ 0.285, p ¼ 0.126), CAD ( ¼ 0.235, p ¼ 0.338), diabetes ( ¼ 0.589, p ¼ 0.560) or anti-lipid therapy ( ¼ 0.013, p ¼ 0.941)]. This confirmed that sCD147 levels were independently predicted by sGPVI ( ¼ 0.522, p ¼ 0.003), and more weakly by age, the significance of which was decreased due to over-fitting of the model ( ¼ 0.326, p ¼ 0.057). The overall model fit was r2 ¼ 0.362.
Discussion In this study, we demonstrate for the first time that sCD147 and sGPVI are strongly correlated. Statistical analysis shows that sCD147 expression is predicted by sGPVI expression and age, but not by other clinical parameters examined. The association observed between sCD147 and sGPVI could indicate that the source of the sCD147 is indeed the platelet since sGPVI is a platelet-specific marker. However, only a weak association, that was not significant, was observed between platelet GPIIbIIIa activation (assessed by PAC-1 binding) and
Risk Factors for CAD Hypertension þ Family history þ Hypercholesterolemia þ Diabetes þ Smoking þ
Soluble GPVI (ng/mL)
Mean (SD)
p Value
Mean (SD)
p Value
27.59 (11.1) 26.57 (7.5)
0.764
51.74 (27.1) 44.5 (23.5)
0.352
24.45 (5.1) 29.22 (11.9)
0.104
39.63 (22.4) 57.67 (26.1)
0.014*
0.321
0.049*
0.193
0.189
27.35 (11.1) 27.09 (8.9)
0.939
48.24 (25.3) 49.95 (27.0)
0.822
28.11 (10.5) 23.26 (5.1)
0.245
50.31 (25.8) 43.44 (26.5)
0.478
28.66 (12.4) 26.46 (8.4)
0.522
46.23 (26.9) 50.70 (25.4)
0.567
26.74 (10.4) 28.74 (8.3)
0.601
44.02 (24.1) 65.87 (25.2)
0.012*
26.80 (10.5) 27.84 (9.1)
0.756
45.39 (23.2) 55.65 (29.6)
0.191
0.789
42.22 (26.3) 53.11 (25.1)
0.159
0.079
45.48 (25.5) 53.59 (26.1
0.285
0.275
48.34 (27.8) 50.52 (21.6)
0.79
0.521
47.30 (25.3) 51.44 (27.0)
0.59
0.374
42.09 (25.3) 55.41 (25.1)
0.074
0.162
48.73 (26.7) 49.83 (24.2)
0.897
Medication Aspirin 27.78 (12.5) þ 26.88 (8.2) Clopidogrel 24.55 (6.8) þ 30.17 (11.9) Beta-blocker 25.63 (6.8) þ 30.25 (13.9) ACE inhibitor/Ang II 26.22 (6.8) þ 28.3 (12.5) Anti-lipid 25.74 (6.1) þ 28.5 (12.3) Nitrates 25.78 (7.6) þ 30.74 (13.8)
Ang II ¼ Angiotensin II inhibitors. *p50.05, bold text indicates values included in linear regression models.
sCD147. Also, no association was observed between the other platelet surface markers examined and sCD147 levels. Because this was also the case for sGPVI and the platelet markers – CD62P and PAC-1 binding – it is possible that the levels of platelet surface markers are disproportionately affected by the use of antiplatelet agents as has been previously shown [20]. Alternatively, there may be a common mediator of the release of sGPVI and sCD147. For example, activation of the metalloproteinases responsible for release of sGPVI from platelets may also result in the release of sCD147. Such a mechanism would be consistent with sCD147 being at least in part plateletderived. Because CD147 is expressed by other cells also involved in CAD progression, such as leukocytes and smooth muscle cells, further work is needed to determine the proportion of sCD147 derived from these cells. Does GPVI either indirectly or directly mediate the release of CD147? Although it has been determined that CD147 is a counter-receptor for GPVI, the implications of that binding have
Platelets Downloaded from informahealthcare.com by University of Colorado Libraries on 12/30/14 For personal use only.
642
G. J. Pennings et al.
not been examined [13]. Preliminary work has demonstrated that membrane-bound CD147 is not affected by GPVI ligands, collagen or convulxin (data not shown), which induce platelet activation and shedding of GPVI. However, it has not been determined whether binding of membrane-bound or soluble GPVI to CD147 leads to the release of sCD147. Further studies are required to understand the mechanisms involved. Soluble GPVI correlated strongly with WCC. However, this correlation was not specific to a single subset of leukocytes because sGPVI correlated with monocyte, neutrophil and lymphocyte counts (data not shown). A possible explanation is that cytokines which promote leukocytosis stimulate the release of GPVI from platelets; however, this remains to be directly investigated. We have determined that the association between sCD147 and sGPVI is not confounded by CAD or other comorbidities. Age, however, seems to be consistent in affecting the levels of CD147 whether platelet-bound [2] or soluble (this study). Because age does not affect leukocyte CD147 expression, this may further implicate the platelet as the source of sCD147. Finally, inflammatory markers such as IL-6 and CRP have been associated with age [21], although whether the elevated levels of these markers are caused by age-related disease or participate in the development of disease is unclear. sCD147 levels may be useful as a marker for ageing since the expression does not appear to be affected by the inflammatory state, as no correlations were observed between sCD147 levels and NLR, WCC or CRP (Table II). Further studies would be needed to determine the efficacy of sCD147 as a marker for ageing because little is known about sCD147 in disease. In addition, the relatively small size of our cohort is a limitation and our observations will require confirmation in larger populations In conclusion, the strong association observed between sCD147 and sGPVI is independent of the haematological and clinical parameters examined. sCD147 does not correlate with traditional markers of platelet activation such as CD62P expression or PAC-1 binding. As sCD147 correlates with plateletspecific sGPVI, a common platelet source and/or mechanism of release appear likely to contribute to sCD147 levels in vivo.
Declaration of interest This work is supported by the National Health and Medical Research Council of Australia (NHMRC), NHMRC Program grant 482800 (LK) and NHMRC Postdoctoral Training Fellowship (AY). The authors report no declarations of interest.
Platelets, 2014; 25(8): 639–642
6.
7.
8. 9.
10.
11.
12.
13.
14.
15.
16.
17.
References 1. Schmidt R, Bultmann A, Fischel S, Gillitzer A, Cullen P, Walch A, Jost P, Ungerer M, Tolley ND, Lindemann S, et al. Extracellular matrix metalloproteinase inducer (CD147) is a novel receptor on platelets, activates platelets, and augments nuclear factor kappabdependent inflammation in monocytes. Circ Res 2008;102:302–309. 2. Pennings GJ, Yong ASC, Kritharides L. Expression of EMMPRIN (CD147) on circulating platelets in vivo. J Thromb Haemost 2010;8: 472–481. 3. Sidhu SS, Mengistab AT, Tauscher AN, LaVail J, Basbaum C. The microvesicle as a vehicle for EMMPRIN in tumor-stromal interactions. Oncogene 2004;23:956–963. 4. Tang Y, Kesavan P, Nakada MT, Yan L. Tumor-stroma interaction: Positive feedback regulation of extracellular matrix metalloproteinase inducer (EMMPRIN) expression and matrix metalloproteinasedependent generation of soluble EMMPRIN. Mol Cancer Res 2004; 2:73–80. 5. Tsuji M, Ezumi Y, Arai M, Takayama H. A novel association of Fc receptor gamma-chain with glycoprotein VI and their co-expression
18. 19.
20.
21.
as a collagen receptor in human platelets. J Biol Chem 1997;272: 23528–23531. Nieswandt B, Brakebusch C, Bergmeier W, Schulte V, Bouvard D, Mokhtari-Nejad R, Lindhout T, Heemskerk JW, Zirngibl H, Fassler R. Glycoprotein VI but not alpha2beta1 integrin is essential for platelet interaction with collagen. EMBO J 2001;20: 2120–2130. Gardiner EE, Karunakaran D, Shen Y, Arthur JF, Andrews RK, Berndt MC. Controlled shedding of platelet glycoprotein (GP)VI and GPIb-IX-V by ADAM family metalloproteinases. J Thromb Haemost 2007;5:1530–1537. Al-Tamimi M, Mu FT, Moroi M, Gardiner EE, Berndt MC, Andrews RK. Measuring soluble platelet glycoprotein VI in human plasma by ELISA. Platelets 2009;20:143–149. Major TC, Liang L, Lu X, Rosebury W, Bocan TM. Extracellular matrix metalloproteinase inducer (EMMPRIN) is induced upon monocyte differentiation and is expressed in human atheroma. Arterioscler Thromb Vasc Biol 2002;22:1200–1207. Yanaba K, Asano Y, Tada Y, Sugaya M, Kadono T, Hamaguchi Y, Sato S. Increased serum soluble CD147 levels in patients with systemic sclerosis: Association with scleroderma renal crisis. Clin Rheumatol 2012;31:835–839. Al-Tamimi M, Tan CW, Qiao J, Pennings GJ, Javadzadegan A, Yong AS, Arthur JF, Davis AK, Jing J, Mu FT, et al. Pathologic shear triggers shedding of vascular receptors: A novel mechanism for down-regulation of platelet glycoprotein VI in stenosed coronary vessels. Blood 2012;119:4311–4320. Bigalke B, Stellos K, Weig HJ, Geisler T, Seizer P, Kremmer E, Potz O, Joos T, May AE, Lindemann S, et al. Regulation of platelet glycoprotein VI (GPVI) surface expression and of soluble GPVI in patients with atrial fibrillation (AF) and acute coronary syndrome (ACS). Basic Res Cardiol 2009;104:352–357. Seizer P, Borst O, Langer HF, Bultmann A, Munch G, Herouy Y, Stellos K, Kramer B, Bigalke B, Buchele B, et al. EMMPRIN (CD147) is a novel receptor for platelet GPVI and mediates platelet rolling via GPVI-EMMPRIN interaction. Thromb Haemost 2009; 101:682–686. Yong AS, Pennings GJ, Chang M, Hamzah A, Chung T, Qi M, Brieger D, Behnia M, Krilis SA, Ng MK, et al. Intracoronary shearrelated up-regulation of platelet P-selectin and platelet-monocyte aggregation despite the use of aspirin and clopidogrel. Blood 2011; 117:11–20. Yong AS, Pennings GJ, Chung T, Brieger D, Lowe HC, Kritharides L. Intravascular blood sampling using the export catheter does not induce artifactual platelet activation. J Invasive Cardiol 2009;21: 159–161. Schmitz G, Rothe G, Ruf A, Barlage S, Tschope D, Clemetson KJ, Goodall AH, Michelson AD, Nurden AT, Shankey TV. European working group on clinical cell analysis: Consensus protocol for the flow cytometric characterisation of platelet function. Thromb Haemost 1998;79:885–896. Joseph JE, Harrison P, Mackie IJ, Isenberg DA, Machin SJ. Increased circulating platelet-leucocyte complexes and platelet activation in patients with antiphospholipid syndrome, systemic lupus erythematosus and rheumatoid arthritis. Br J Haematol 2001; 115:451–459. Chung T, Connor D, Joseph J, Emmett L, Mansberg R, Peters M, Ma D, Kritharides L. Platelet activation in acute pulmonary embolism. J Thromb Haemost 2007;5:918–924. Al-Tamimi M, Mu FT, Arthur JF, Shen Y, Moroi M, Berndt MC, Andrews RK, Gardiner EE. Anti-glycoprotein VI monoclonal antibodies directly aggregate platelets independently of Fcgamma RIIa and induce GPVI ectodomain shedding. Platelets 2009;20: 75–82. McKenzie ME, Malinin AI, Bell CR, Dzhanashvili A, Horowitz ED, Oshrine BR, Atar D, Serebruany VL. Aspirin inhibits surface glycoprotein IIb/IIIa, P-selectin, CD63, and CD107a receptor expression on human platelets. Blood Coagul Fibrinolysis 2003; 14:249–253. Ferrucci L, Corsi A, Lauretani F, Bandinelli S, Bartali B, Taub DD, Guralnik JM, Longo DL. The origins of age-related proinflammatory state. Blood 2005;105:2294–2299.
SHORT COMMUNICATION
Circulating levels of soluble EMMPRIN (CD147) correlate with levels of soluble glycoprotein VI in human plasma G. J. Pennings1,2, A. S. C. Yong1,2, C. Wong1,2, M. Al-Tamimi3,4, E. E. Gardiner3, R. K. Andrews3, & L. Kritharides1,2 Vascular Biology Group, The ANZAC Research Institute, Concord Repatriation General Hospital, Hospital Rd, Concord, NSW, Australia, 2The University of Sydney, Sydney, NSW, Australia, 3Australian Centre for Blood Diseases, Monash University, Melbourne, Australia, and 4Department of Immunology and Microbiology, Hashemite University, Zarqa, Jordan
Platelets Downloaded from informahealthcare.com by University of Colorado Libraries on 12/30/14 For personal use only.
1
Abstract
Keywords
Extracellular matrix metalloproteinase inducer (EMMPRIN; CD147), which binds to the plateletspecific collagen receptor glycoprotein (GP) VI, is expressed in a range of cell types including platelets and leukocytes, and has been implicated in neoplastic disease and atherosclerotic coronary disease. Both CD147 and GPVI can be shed from cell membranes and detected in plasma. However, while the relationship between soluble CD147 (sCD147), soluble GPVI (sGPVI) and standard markers of platelet activation has received little attention, such analysis may help reveal pathways mediating release of sCD147. We investigated the relationship between sCD147 and platelet markers including sGPVI, soluble and platelet-bound CD62P (P-selectin), active aIIbb3 (assessed by PAC-1 binding) and platelet CD147 in 25 patients with stable angina pectoris (SAP), 13 patients with no coronary artery disease (CAD) and 10 healthy donors. Plasma levels of sCD147 significantly correlated with sGPVI (r ¼ 0.46, p ¼ .004), but did not correlate with any other platelet markers examined. Linear regression analysis identified that sCD147 levels could be predicted by sGPVI levels ( ¼ .445, p ¼ 0.003) and age ( ¼ 0.304, p ¼ 0.038), but were independent of potential clinical confounders such as CAD, diabetes and medication usage. As sCD147 strongly correlates with platelet-specific sGPVI, a common platelet source and/or mechanism of release may contribute to sCD147 levels in vivo.
EMMPRIN, platelet, soluble CD147, soluble GPVI
Introduction CD147 is a member of the immunoglobulin superfamily expressed on the surface of platelets, haematopoietic cells, endothelial cells and tumour cells. Surface expression of CD147 is elevated on circulating monocytes in myocardial infarct patients [1] and on circulating platelets, monocytes, granulocytes and lymphocytes in patients with coronary artery disease (CAD) [2]. In tumour cells, CD147 may be exported in microvesicles [3]; however, the release of soluble CD147 (sCD147) in the vasculature also involves proteolytic cleavage by matrix metalloproteinases (MMPs) [4]. Whether either of these mechanisms of release occurs in platelets or leukocytes is as yet unknown. The platelet collagen receptor, GPVI, also a member of the immunoglobulin superfamily, is constitutively expressed exclusively on platelets, and is co-associated with the Fc receptor g-chain (FcRg) required for stable surface expression [5]. GPVI mediates thrombus formation at sites of vascular injury [6]. Binding of collagen or other GPVI ligands to the GPVI/FcRg complex results in platelet activation, aggregation and metalloproteinase-dependent shedding of GPVI from the platelet surface [7]. The resultant 55 kDa soluble fragment of GPVI (sGPVI,) is detectable in human plasma and can be assessed by ELISA [8].
History Received 23 August 2013 Revised 29 September 2013 Accepted 2 October 2013 Published online 18 November 2013
CD147 has been found in atherosclerotic plaques [9] and elevated levels have been observed in platelet and leukocytes from patients with CAD [1, 2]. Levels of sCD147 are increased in serum from patients with inflammatory diseases such as systemic sclerosis [10] but have not been examined in CAD in tandem with other platelet markers that are known to be upregulated such as sCD62P (soluble P-selectin), activated integrin aIIbb3 (GPIIbIIIa) or sGPVI [11, 12]. Examining the relationship between sCD147 levels and markers of platelet activation such as increased CD62P expression on platelets, elevated plasma sCD62P, activation of aIIbb3 (detected by PAC-1 binding) and elevated plasma sGPVI may provide us with important information regarding the cellular origin of sCD147 in inflammatory disease states. In 2009, Seizer and colleagues demonstrated for the first time that CD147 was a receptor for GPVI [13]. As the relationship between sCD147 and sGPVI (two receptor fragments derived from metalloproteolysis) is not known, the aims of this study were to investigate the relationship between sCD147 and sGPVI as well as standard platelet activation markers and to investigate potential clinical characteristics that may affect sCD147 levels.
Methods Reagents
Correspondence: Prof. Leonard Kritharides, Department of Cardiology, Concord Repatriation General Hospital, Hospital Road, Concord, NSW 2139, Australia. Tel: +61 2 9767 6296. Fax: +61 2 9767 6994. E-mail: [email protected]
Preconjugated fluorescent antibodies CD42b-PE, CD62P-PE, IgG1-PE, IgM-FITC and PAC-1-FITC were from BD Biosciences, Franklin Lakes, NJ. Dulbecco’s modified phosphate buffered saline (DPBS) and Hank’s buffered salt solution
640
G. J. Pennings et al.
Platelets, 2014; 25(8): 639–642
Platelets Downloaded from informahealthcare.com by University of Colorado Libraries on 12/30/14 For personal use only.
Table I. Clinical characteristics.
Gender Male Female Risk Factors for CAD Hypertension Family history Hypercholesterolemia Diabetes Smoking (former and current) Medication Aspirin Clopidogrel Warfarin Beta-blocker ACE inhibitor/angiotensin II Anti-lipid Nitrates Age - Mean SD
n
%
30 18
62 38
22 9 30 11 17
46 19 63 23 35
30 63 21 44 2 4 15 31 20 42 25 52 13 27 60 13 yr
(10x, HBSS) were from Invitrogen, Carlsbad, CA. Formaldehyde (16% solution) was from ProSciTech, Kirwan, Queensland. Nonpyrogenic, endotoxin-free water and 0.9% sodium chloride solution were from Baxter, Deerfield, IL. Patient selection and blood collection A total of 48 subjects (mean age 60 2 years) comprising 25 unselected patients with confirmed stable angina pectoris (SAP) and 23 controls were recruited from Concord Repatriation General Hospital. Control subjects (n ¼ 23, 52 18 years) were recruited from healthy volunteers (n ¼ 10) or from patients without significant CAD (530% stenosis) upon conventional coronary angiography or computed tomography (n ¼ 13) [2]. Patients (n ¼ 25, 66 9 years) were classified as SAP if they presented with clinically stable angina and significant CAD (450% stenosis) upon conventional coronary angiography (Table I). Blood (15 mL) was collected by venipuncture by gentle aspiration without a tourniquet using a 20G intravenous cannula as previously described [2, 14, 15]. Briefly, the first 5 mL was discarded. The remaining 10 mL blood was distributed between EDTA and CTAD (buffered sodium citrate theophylline adenosine dipyridamole) blood collection tubes (BD Biosciences, Franklin Lakes, NJ). Tubes and samples were kept at 4 C throughout sample transport and processing. The EDTA and CTAD tubes were centrifuged at 1000 g for 15 min at 4 C; the plasma was then removed and stored at 80 C for assessment of soluble markers by ELISA. This study was approved by the institutional human research ethics committee. Written informed consent was obtained from all subjects.
correction at 600 nm. The lower limit of detection, intra-assay CV and inter-assay CV were as follows: sCD147 (7.6 pg/mL, 4.5% and 8.5%), sGPVI (1 ng/mL, 4.4%, 11.6%) [8] and sCD62P (0.5 ng/mL, 5.2%, 8.9%). Flow cytometry Flow cytometry was performed as previously described under resting conditions [15–18]. Briefly, DPBS and antibodies (isotype control – IgG or IgM; or platelet markers – CD42b, CD62P or PAC-1) were aliquoted into the appropriate tubes with 5 mL of CTAD anticoagulated blood (final volume, 50 mL). GPVI expression was determined by Zenon labelling as per the manufacturer’s instructions (Invitrogen, Carlsbad, CA) of anti-GPVI mAb (clone 1A12) and processed in the same manner as the other antibodies [19]. After 20 min at 4 C, samples were fixed with ice-cold 0.16% formaldehyde solution in saline. Acquisition was performed on a BD FACSCalibur (BD Biosciences). The platelet population was collected based on light scatter characteristics and confirmed by expression of CD42b. The results were analyzed with WinMDI 2.9 flow cytometry analysis software (The Scripps Institute, La Jolla, CA). The isotype control was used to establish positive events based on a gate/marker of 0.5% background expression. Statistics The D’Agostino and Pearson omnibus normality test and histograms were used to verify normality of the data. Correlations between sCD147 and sGPVI or platelet markers were performed using Pearson’s correlation. Independent samples t-test and linear regression (SPSS Statistics v20.0; SPSS Inc., Chicago, IL) were used to establish whether clinical characteristics/markers and/or medication may have an effect on sCD147 expression and the order in which to include them in a linear regression model to determine the predictors of sCD147 expression.
Results Correlations of sCD147 with markers of platelet activation and clinical characteristics There was a clear correlation between levels of sCD147 and sGPVI (r ¼ 0.456, p ¼ 0.004) (Figure 1). No significant correlations were observed between sCD147 and markers of platelet activation, including sCD62P, platelet-bound CD62P, PAC-1 binding and platelet-bound CD147 (Table II). No correlations were observed between sCD147 and white cell count (WCC), platelet count or conventional markers of systemic inflammation [neutrophil-lymphocyte ratio (NLR) or C-reactive protein (CRP)] (Table II). sGPVI levels correlated with sCD147 (as above), sCD62P (r ¼ 0.448, p ¼ 0.002, n ¼ 44), membrane-bound GPVI (r ¼ 0.530, p ¼ 0.020, n ¼ 19) and white cell count (r ¼ 0.422, p ¼ 0.004, n ¼ 44). No correlation was observed between sGPVI and other markers of platelet activation.
Measurement of sCD147 and other enzyme linked immunosorbent assays
Univariate analysis and linear regression
sCD147 was assessed using a commercially available ELISA antibody pair (R&D Systems, Minneapolis, MN) and optimized for use with human EDTA anticoagulated plasma. Levels of sCD147 were assessed as per the manufacturer’s instructions, with diluted plasma (1/100) used in place of cell lysates. sGPVI was assessed in EDTA anticoagulated plasma as previously described [8]. sCD62P was assessed using a commercially available ELISA kit according to the manufacturer’s instructions (R&D Systems, Minneapolis, MN) using CTAD anticoagulated plasma. sCD147 and sCD62P plates were read on a Wallac-Victor 3 plate reader (PerkinElmer, Walthan, MA) at 450 nm with a
Clinical characteristics were assessed for their potential effect on sCD147 and on sGPVI expression to look for confounders of the correlation (Tables I and III). Variables with a p value less than or equal to 0.1 were assessed for inclusion in the model. CAD was associated with both sCD147 (very weakly) and sGPVI (Table III). Age and clopidogrel weakly associated with sCD147, whereas diabetes and anti-lipid medication associated with sGPVI (Table III). Only sGPVI ( ¼ 0.445, p ¼ 0.003) and age ( ¼ 0.304, p ¼ 0.038) significantly affected sCD147 expression based on assessment by linear regression. The overall model fit was
sCD147 correlates with sGPVI
DOI: 10.3109/09537104.2013.852660
641
Table III. Univariate analysis for soluble CD147 and soluble GPVI. Soluble CD147 (ng/mL)
Platelets Downloaded from informahealthcare.com by University of Colorado Libraries on 12/30/14 For personal use only.
Gender M F CAD þ Age r
Figure 1. Soluble CD147 increase with increasing soluble GPVI. Levels of soluble CD147 correlate strongly with soluble GPVI levels (r ¼ 0.46, p ¼ 0.004, n ¼ 38) analyzed with Pearson’s correlation as assessed in a subset of the cohort (includes SAP patients, patients without significant CAD and healthy controls). Table II. Correlations for soluble CD147 and soluble GPVI. Soluble CD147
Platelet Markers Soluble GPVI Soluble CD147 Soluble CD62P Platelet CD62P PAC-1 binding Platelet CD147 Platelet GPVI Haematological Parameters White cell count Platelet count Neutrophil-lymphocyte ratio C-reactive protein
Soluble GPVI
r
p Value
n
r
p Value
n
0.456 – 0.231 0.143 0.300 0.164 0.202
0.004** – 0.162 0.393 0.067 0.326 0.576
38 – – – 0.456 0.004** 38 0.448 0.002** 38 0.159 0.282 38 0.021 0.885 38 0.036 0.810 10 0.530 0.020*
– 38 44 48 48 48 19
0.252 0.127 0.106 0.126
0.138 0.454 0.551 0.499
36 37 34 31
44 45 39 36
0.422 0.066 0.087 0.016
0.004** 0.666 0.597 0.928
*p50.05, **p50.01.
r2 ¼ 0.300. The following equation was determined for predicting sCD147 levels: sCD147 ¼ 2:45 þ ð0:16 sGPVIÞ þ ð0:27 AgeÞ: To ensure there were no clinical confounders, we also repeated the linear regression including four parameters that had some association with either sCD147 or sGPVI [clopidogrel usage ( ¼ 0.285, p ¼ 0.126), CAD ( ¼ 0.235, p ¼ 0.338), diabetes ( ¼ 0.589, p ¼ 0.560) or anti-lipid therapy ( ¼ 0.013, p ¼ 0.941)]. This confirmed that sCD147 levels were independently predicted by sGPVI ( ¼ 0.522, p ¼ 0.003), and more weakly by age, the significance of which was decreased due to over-fitting of the model ( ¼ 0.326, p ¼ 0.057). The overall model fit was r2 ¼ 0.362.
Discussion In this study, we demonstrate for the first time that sCD147 and sGPVI are strongly correlated. Statistical analysis shows that sCD147 expression is predicted by sGPVI expression and age, but not by other clinical parameters examined. The association observed between sCD147 and sGPVI could indicate that the source of the sCD147 is indeed the platelet since sGPVI is a platelet-specific marker. However, only a weak association, that was not significant, was observed between platelet GPIIbIIIa activation (assessed by PAC-1 binding) and
Risk Factors for CAD Hypertension þ Family history þ Hypercholesterolemia þ Diabetes þ Smoking þ
Soluble GPVI (ng/mL)
Mean (SD)
p Value
Mean (SD)
p Value
27.59 (11.1) 26.57 (7.5)
0.764
51.74 (27.1) 44.5 (23.5)
0.352
24.45 (5.1) 29.22 (11.9)
0.104
39.63 (22.4) 57.67 (26.1)
0.014*
0.321
0.049*
0.193
0.189
27.35 (11.1) 27.09 (8.9)
0.939
48.24 (25.3) 49.95 (27.0)
0.822
28.11 (10.5) 23.26 (5.1)
0.245
50.31 (25.8) 43.44 (26.5)
0.478
28.66 (12.4) 26.46 (8.4)
0.522
46.23 (26.9) 50.70 (25.4)
0.567
26.74 (10.4) 28.74 (8.3)
0.601
44.02 (24.1) 65.87 (25.2)
0.012*
26.80 (10.5) 27.84 (9.1)
0.756
45.39 (23.2) 55.65 (29.6)
0.191
0.789
42.22 (26.3) 53.11 (25.1)
0.159
0.079
45.48 (25.5) 53.59 (26.1
0.285
0.275
48.34 (27.8) 50.52 (21.6)
0.79
0.521
47.30 (25.3) 51.44 (27.0)
0.59
0.374
42.09 (25.3) 55.41 (25.1)
0.074
0.162
48.73 (26.7) 49.83 (24.2)
0.897
Medication Aspirin 27.78 (12.5) þ 26.88 (8.2) Clopidogrel 24.55 (6.8) þ 30.17 (11.9) Beta-blocker 25.63 (6.8) þ 30.25 (13.9) ACE inhibitor/Ang II 26.22 (6.8) þ 28.3 (12.5) Anti-lipid 25.74 (6.1) þ 28.5 (12.3) Nitrates 25.78 (7.6) þ 30.74 (13.8)
Ang II ¼ Angiotensin II inhibitors. *p50.05, bold text indicates values included in linear regression models.
sCD147. Also, no association was observed between the other platelet surface markers examined and sCD147 levels. Because this was also the case for sGPVI and the platelet markers – CD62P and PAC-1 binding – it is possible that the levels of platelet surface markers are disproportionately affected by the use of antiplatelet agents as has been previously shown [20]. Alternatively, there may be a common mediator of the release of sGPVI and sCD147. For example, activation of the metalloproteinases responsible for release of sGPVI from platelets may also result in the release of sCD147. Such a mechanism would be consistent with sCD147 being at least in part plateletderived. Because CD147 is expressed by other cells also involved in CAD progression, such as leukocytes and smooth muscle cells, further work is needed to determine the proportion of sCD147 derived from these cells. Does GPVI either indirectly or directly mediate the release of CD147? Although it has been determined that CD147 is a counter-receptor for GPVI, the implications of that binding have
Platelets Downloaded from informahealthcare.com by University of Colorado Libraries on 12/30/14 For personal use only.
642
G. J. Pennings et al.
not been examined [13]. Preliminary work has demonstrated that membrane-bound CD147 is not affected by GPVI ligands, collagen or convulxin (data not shown), which induce platelet activation and shedding of GPVI. However, it has not been determined whether binding of membrane-bound or soluble GPVI to CD147 leads to the release of sCD147. Further studies are required to understand the mechanisms involved. Soluble GPVI correlated strongly with WCC. However, this correlation was not specific to a single subset of leukocytes because sGPVI correlated with monocyte, neutrophil and lymphocyte counts (data not shown). A possible explanation is that cytokines which promote leukocytosis stimulate the release of GPVI from platelets; however, this remains to be directly investigated. We have determined that the association between sCD147 and sGPVI is not confounded by CAD or other comorbidities. Age, however, seems to be consistent in affecting the levels of CD147 whether platelet-bound [2] or soluble (this study). Because age does not affect leukocyte CD147 expression, this may further implicate the platelet as the source of sCD147. Finally, inflammatory markers such as IL-6 and CRP have been associated with age [21], although whether the elevated levels of these markers are caused by age-related disease or participate in the development of disease is unclear. sCD147 levels may be useful as a marker for ageing since the expression does not appear to be affected by the inflammatory state, as no correlations were observed between sCD147 levels and NLR, WCC or CRP (Table II). Further studies would be needed to determine the efficacy of sCD147 as a marker for ageing because little is known about sCD147 in disease. In addition, the relatively small size of our cohort is a limitation and our observations will require confirmation in larger populations In conclusion, the strong association observed between sCD147 and sGPVI is independent of the haematological and clinical parameters examined. sCD147 does not correlate with traditional markers of platelet activation such as CD62P expression or PAC-1 binding. As sCD147 correlates with plateletspecific sGPVI, a common platelet source and/or mechanism of release appear likely to contribute to sCD147 levels in vivo.
Declaration of interest This work is supported by the National Health and Medical Research Council of Australia (NHMRC), NHMRC Program grant 482800 (LK) and NHMRC Postdoctoral Training Fellowship (AY). The authors report no declarations of interest.
Platelets, 2014; 25(8): 639–642
6.
7.
8. 9.
10.
11.
12.
13.
14.
15.
16.
17.
References 1. Schmidt R, Bultmann A, Fischel S, Gillitzer A, Cullen P, Walch A, Jost P, Ungerer M, Tolley ND, Lindemann S, et al. Extracellular matrix metalloproteinase inducer (CD147) is a novel receptor on platelets, activates platelets, and augments nuclear factor kappabdependent inflammation in monocytes. Circ Res 2008;102:302–309. 2. Pennings GJ, Yong ASC, Kritharides L. Expression of EMMPRIN (CD147) on circulating platelets in vivo. J Thromb Haemost 2010;8: 472–481. 3. Sidhu SS, Mengistab AT, Tauscher AN, LaVail J, Basbaum C. The microvesicle as a vehicle for EMMPRIN in tumor-stromal interactions. Oncogene 2004;23:956–963. 4. Tang Y, Kesavan P, Nakada MT, Yan L. Tumor-stroma interaction: Positive feedback regulation of extracellular matrix metalloproteinase inducer (EMMPRIN) expression and matrix metalloproteinasedependent generation of soluble EMMPRIN. Mol Cancer Res 2004; 2:73–80. 5. Tsuji M, Ezumi Y, Arai M, Takayama H. A novel association of Fc receptor gamma-chain with glycoprotein VI and their co-expression
18. 19.
20.
21.
as a collagen receptor in human platelets. J Biol Chem 1997;272: 23528–23531. Nieswandt B, Brakebusch C, Bergmeier W, Schulte V, Bouvard D, Mokhtari-Nejad R, Lindhout T, Heemskerk JW, Zirngibl H, Fassler R. Glycoprotein VI but not alpha2beta1 integrin is essential for platelet interaction with collagen. EMBO J 2001;20: 2120–2130. Gardiner EE, Karunakaran D, Shen Y, Arthur JF, Andrews RK, Berndt MC. Controlled shedding of platelet glycoprotein (GP)VI and GPIb-IX-V by ADAM family metalloproteinases. J Thromb Haemost 2007;5:1530–1537. Al-Tamimi M, Mu FT, Moroi M, Gardiner EE, Berndt MC, Andrews RK. Measuring soluble platelet glycoprotein VI in human plasma by ELISA. Platelets 2009;20:143–149. Major TC, Liang L, Lu X, Rosebury W, Bocan TM. Extracellular matrix metalloproteinase inducer (EMMPRIN) is induced upon monocyte differentiation and is expressed in human atheroma. Arterioscler Thromb Vasc Biol 2002;22:1200–1207. Yanaba K, Asano Y, Tada Y, Sugaya M, Kadono T, Hamaguchi Y, Sato S. Increased serum soluble CD147 levels in patients with systemic sclerosis: Association with scleroderma renal crisis. Clin Rheumatol 2012;31:835–839. Al-Tamimi M, Tan CW, Qiao J, Pennings GJ, Javadzadegan A, Yong AS, Arthur JF, Davis AK, Jing J, Mu FT, et al. Pathologic shear triggers shedding of vascular receptors: A novel mechanism for down-regulation of platelet glycoprotein VI in stenosed coronary vessels. Blood 2012;119:4311–4320. Bigalke B, Stellos K, Weig HJ, Geisler T, Seizer P, Kremmer E, Potz O, Joos T, May AE, Lindemann S, et al. Regulation of platelet glycoprotein VI (GPVI) surface expression and of soluble GPVI in patients with atrial fibrillation (AF) and acute coronary syndrome (ACS). Basic Res Cardiol 2009;104:352–357. Seizer P, Borst O, Langer HF, Bultmann A, Munch G, Herouy Y, Stellos K, Kramer B, Bigalke B, Buchele B, et al. EMMPRIN (CD147) is a novel receptor for platelet GPVI and mediates platelet rolling via GPVI-EMMPRIN interaction. Thromb Haemost 2009; 101:682–686. Yong AS, Pennings GJ, Chang M, Hamzah A, Chung T, Qi M, Brieger D, Behnia M, Krilis SA, Ng MK, et al. Intracoronary shearrelated up-regulation of platelet P-selectin and platelet-monocyte aggregation despite the use of aspirin and clopidogrel. Blood 2011; 117:11–20. Yong AS, Pennings GJ, Chung T, Brieger D, Lowe HC, Kritharides L. Intravascular blood sampling using the export catheter does not induce artifactual platelet activation. J Invasive Cardiol 2009;21: 159–161. Schmitz G, Rothe G, Ruf A, Barlage S, Tschope D, Clemetson KJ, Goodall AH, Michelson AD, Nurden AT, Shankey TV. European working group on clinical cell analysis: Consensus protocol for the flow cytometric characterisation of platelet function. Thromb Haemost 1998;79:885–896. Joseph JE, Harrison P, Mackie IJ, Isenberg DA, Machin SJ. Increased circulating platelet-leucocyte complexes and platelet activation in patients with antiphospholipid syndrome, systemic lupus erythematosus and rheumatoid arthritis. Br J Haematol 2001; 115:451–459. Chung T, Connor D, Joseph J, Emmett L, Mansberg R, Peters M, Ma D, Kritharides L. Platelet activation in acute pulmonary embolism. J Thromb Haemost 2007;5:918–924. Al-Tamimi M, Mu FT, Arthur JF, Shen Y, Moroi M, Berndt MC, Andrews RK, Gardiner EE. Anti-glycoprotein VI monoclonal antibodies directly aggregate platelets independently of Fcgamma RIIa and induce GPVI ectodomain shedding. Platelets 2009;20: 75–82. McKenzie ME, Malinin AI, Bell CR, Dzhanashvili A, Horowitz ED, Oshrine BR, Atar D, Serebruany VL. Aspirin inhibits surface glycoprotein IIb/IIIa, P-selectin, CD63, and CD107a receptor expression on human platelets. Blood Coagul Fibrinolysis 2003; 14:249–253. Ferrucci L, Corsi A, Lauretani F, Bandinelli S, Bartali B, Taub DD, Guralnik JM, Longo DL. The origins of age-related proinflammatory state. Blood 2005;105:2294–2299.
