A 2D-DIGE-based proteomic analysis reveals differences in the platelet releasate composition when comparing thrombin and collagen stimulations

OPEN SUBJECT AREAS: PROTEOMICS MOLECULAR BIOLOGY Received 25 September 2014 Accepted 12 January 2015 Published 3 February 2015 Correspondence and requests for materials should be addressed to A.G. (angel.garcia@ usc.es) A 2D-DIGE-based proteomic analysis reveals differences in the platelet releasate composition when comparing thrombin and collagen stimulations Paula Vélez1,2,3, Irene Izquierdo1,2,3, Isaac Rosa1,2,3 & Ángel Garcı́a1,2,3 1 Center for Research in Molecular Medicine and Chronic Diseases (CIMUS), Universidade de Santiago de Compostela, Santiago de Compostela, Spain, 2Instituto de Investigación Sanitaria (IDIS), Santiago de Compostela, Spain, 3Departament of Pharmacology, Faculty of Pharmacy, Universidade de Santiago de Compostela, Santiago de Compostela, Spain. Upon stimulation, platelets release a high number of proteins (the releasate). There are clear indications that these proteins are involved in the pathogenesis of several diseases, such as atherosclerosis. In the present study we compared the platelet releasate following platelet activation with two major endogenous agonists: thrombin and collagen. Proteome analysis was based on 2D-DIGE and LC-MS/MS. Firstly, we showed the primary role of thrombin and collagen receptors in platelet secretion by these agonists; moreover, we demonstrated that GPVI is the primary responsible for collagen-induced platelet activation/aggregation. Proteomic analysis allowed the detection of 122 protein spots differentially regulated between both conditions. After excluding fibrinogen spots, down-regulated in the releasate of thrombin-activated platelets, 84 differences remained. From those, we successfully identified 42, corresponding to 37 open-reading frames. Many of the differences identified correspond to post-translational modifications, primarily, proteolysis induced by thrombin. Among others, we show vitamin K-dependent protein S, an anticoagulant plasma protein, is up-regulated in thrombin samples. Our results could have pathological implications given that platelets might be playing a differential role in various diseases and biological processes through the secretion of different subsets of granule proteins and microvesicles following a predominant activation of certain receptors. P latelets are small anucleate cells that play a fundamental role in haemostasis. Undesired platelet activation and formation of arterial thrombi are implicated in many diseases, such as myocardial infarction and stroke1. More recently, platelets have been also shown to play a role in other diseases and biological processes, such as angiogenesis, cancer metastasis, or immune response2. Once activated, platelets release a high number of proteins and other biomolecules, which is known as the releasate. During the last decade, a few groups have applied various proteomic approaches to study in detail the platelet releasate3–6. Platelets were primarily stimulated with thrombin; in some cases microvesicles were removed prior to analysis3 whereas in others not5,6. Besides providing a repertoire of platelet secreted proteins, the study of the platelet releasate has led to the identification of proteins relevant to disease. For example, Coppinger and colleagues found some platelet-released proteins in human atherosclerotic plaques, which indicates they could be contributing to the pathogenesis of atherosclerosis3. Moreover, the impact of aspirin in the platelet releasate was also studied by the same group, leading to the conclusion that aspirin has a general moderating effect on the amount of protein released regardless of the agonist4. A recent report by Jonnalagadda and colleagues showed that platelet secretion is kinetically heterogeneous in an agonist-responsive manner7. In line with this, we tried to confirm the platelet secretome varies with the stimulus by comparing the platelet releasate following platelet activation with two major endogenous agonists: thrombin and collagen. Results The platelet releasate varies when comparing thrombin and collagen stimulations. Platelets were isolated following a standardized procedure that minimizes contamination with other blood cells or plasma proteins, as SCIENTIFIC REPORTS | 5 : 8198 | DOI: 10.1038/srep08198 1 www.nature.com/scientificreports well as activation during isolation8. Firstly, platelets were stimulated with the agonists at different concentrations to determine the minimum concentration needed to achieve maximum aggregation after 3 minutes. Aggregation of approximately 80% was achieved with the following concentrations: 0.75 U/mL of thrombin, and 30 mg/mL of collagen (Fig. 1A). Aggregation profiles were followed to make sure equal platelet aggregation levels were obtained with thrombin and collagen for each donor. Besides the proteomic analysis, we decided to study the contribution of each receptor to platelet activation/aggregation by the above agonists at the final concentrations that were used. Interestingly, a report by Wu and colleagues showed a few years ago that thrombin-induced platelet activation, at doses above 0.5 U/mL, cannot be effectively inhibited by just blocking either single thrombin receptor pathway but by blocking them all (PAR1, PAR-4, and GPIb)9. As a control, we tested the inhibition of the primary human thrombin receptor, PAR-1, and showed thrombininduced platelet aggregation is not inhibited by the PAR-1 specific antagonist SCH 79797 (2 mM) (Fig. 1B). On the other hand, platelet Figure 1 | Effect of PAR-1, GPVI and a2b1 inhibitors on thrombin- and collagen-induced platelet aggregation. (A) Representative platelet aggregation profiles following platelet activation with 0.75 U/mL Thrombin (shown in blue) or 30 mg/mL collagen (shown in black). (B) Effect of PAR-1 inhibition on thrombin-induced platelet aggregation, and of GPVI and a2b1 inhibition on collagen-induced platelet aggregation. Washed human platelets were pre-incubated with the inhibitors for 5 min, then 0.75 U/mL thrombin or 30 mg/mL collagen were added to trigger platelet aggregation. Results are presented as mean 6 SE (n54–6). *p,0.05 (Mann-Whitney test). Coll: collagen; Thr: thrombin; Fab-OM2: Fab fragment of the anti-GPVI monoclonal antibody OM2; BTT: BTT 3033; SCH: SCH 79797. SCIENTIFIC REPORTS | 5 : 8198 | DOI: 10.1038/srep08198 activation with 10 mM TRAP-6 (SFLLRN) - specific PAR-1 agonist was completely inhibited by 140 nM SCH 79797 (not shown). Regarding collagen platelet activation, we inhibited the GPVI receptor by using the Fab fragment of the anti-GPVI monoclonal antibody, OM2, which works as specific antagonist of the receptor10. As expected, OM2 Fab fragment at a final concentration of 1 mg/mL was able of completely inhibiting platelet aggregation induced by 5 mg/mL of the GPVI specific agonist collagen-related peptide (CRP) (not shown). Regarding collagen-induced platelet aggregation, a maximum OM2 (Fab fragment) concentration of 51 mg/mL was able to inhibit aggregation by (...truncated)