Background
Arterial thrombosis developing on the atherosclerotic lesions causes heart attack and stroke that are currently the most common cause of death. Platelets play central roles in hemostasis and in arterial thrombosis by forming thrombi [
1]. Considerable efforts in developing antiplatelet or antithrombotic drugs have been directed towards integrin αIIbβ3 since it is exclusively expressed in the platelet/megakaryocyte lineage and serves as a final common pathway of platelet aggregation in response to various agonists.
Arterial thrombus formation is initiated by platelet adhering to the injured vessel wall at high shear rates via the interaction of glycoprotein Ib (GPIb) with von Willebrand factor (vWF) [
2]. The adherent platelets need to be further activated primarily mediated by integrin αIIbβ3 so as to make firm association (spreading) and to recruit more platelets (aggregation). The αIIbβ3 activation is tightly regulated by inside-out signals initiated by various receptor/ligand interactions such as glycoprotein VI (GPVI)/collagen, P2Y
1 or P2Y
12/adenosine diphosphate (ADP) and protease-activated receptor (PAR)/thrombin, etc. The transduction of inside-out signals transforms the integrin αIIbβ3 receptor from low to high affinity that facilitates ligand (fibrinogen) binding and thus platelet aggregation [
3]. Ligand binding to integrin αIIbβ3 triggers outside-in signaling that promotes platelet cytoskeletal reorganization leading to the firm association of platelets with the vessel wall or with each other. Integrin αIIbβ3 bidirectional signaling is primarily dependent on the complex and dynamic associations between the cytoplasmic proteins and the short cytoplasmic tail of integrin β3. For instance, talin and kindlin bind to the β3 NxxY motifs and play central roles in inside-out signaling [
4], and the association of c-Src with the β3 tail, the RGT residues in particular [
5,
6], is obligatory for outside-in signaling [
5]. The conventional antagonists prevent ligand binding to αIIbβ3 receptor and thus actually inhibit the platelet functions regulated by both inside-out and outside-in signals.
Genetically manipulated mouse models with an impaired integrin αIIbβ3 outside-in signaling [
3,
7‐
14] showed a significantly compromised potential of arterial thrombosis with a normal or only slightly prolonged bleeding time, indicating that selective inhibition of outside-in signaling may represent a rational antithrombotic strategy. Compared with the genetic manipulation, which seems currently impossible to be applied as a therapy, synthetic molecules with “drug-like” properties have the capability of being developed into drugs. We have previously established that disruption of the Src/β3 interaction by the myristoylated RGT peptide selectively inhibited outside-in signaling [
15]. A recent study also showed that the myr-FEEERA peptide disrupted Gα13/β3 interaction, ultimately hampering c-Src activation and thereby inhibiting outside-in signaling [
16]. These data, in light of the established mechanisms in which c-Src is pivotal for outside-in signaling by interacting with β3 [
5,
17,
18], suggest that the Src/β3 association may be a potential target for the development of antithrombotics that do not cause excessive bleeding. The synthesized molecules, such as the RGT peptide, may thus serve as a paradigm for the conception of novel antithrombotic therapies even though the peptide itself may unlikely be directly applied as a drug.
The myristoylation modification renders the peptide membrane permeable but concomitantly integrates the peptide into the cell membrane that may enhance its ability to compete with β3 for c-Src [
19]. It thus became essential to define whether the RGT peptide relies on the membrane anchorage to exert its effects. This important feature is a very basic prerequisite for designing further applications of the peptide or its analogues. Activated c-Src is implicated in a wide range of cellular functions, and there is evidence that binding of some peptides to different domains of c-Src has direct impacts on its kinase activity most likely because of the conformational regulations. For example, Nef binding to the Src homology 3 (SH3) domain or pYEEI to Src homology 2 (SH2) [
20] induced an increased activity of Src-family kinases, as well as RGT peptide binding to SH3 domain of c-Src primed the kinase [
21]. Before moving towards the in vivo and clinical studies using the structure-based small molecular mimetics, the consequences of the RGT peptide, or its analogues, binding to c-Src in the context of its kinase activity need to be carefully defined. A vast body of experimental data, mostly in vivo, indicated that outside-in signaling was essential to stable platelet adhesion and aggregation [
8,
9,
16], and impaired stabilization of the developing thrombi under flow conditions might most likely be ascribable to the disruption of outside-in signaling [
22,
23]. The RGT peptide has been reported to selectively inhibit outside-in signaling under static condition [
15], whether and if so, how it influences thrombus formation under flow condition is yet an important but unanswered question.
By using reduction-sensitive peptides, the present study aimed to address whether the RGT peptide depends on membrane anchorage to influence Src-regulated signaling and whether it alters the kinase activity of c-Src. In addition, the effect of selective blockade of outside-in signaling in human platelets by the RGT peptide on thrombus formation and growth under flow was tested. These efforts allow evaluating the potential of the intracellular delivery of the RGT analogues for antithrombotic therapies where an effective inhibition on thrombosis is accomplished together with an adequate hemostasis.
Discussion
The antagonists target integrin αIIbβ3 receptor and inhibit thrombosis but compromise hemostasis as well. Targeting signaling pathways instead of receptors may reconcile this conflict. It has been established that model animals with impaired outside-in signaling feature a reduced potential of thrombosis without excessive bleeding, and the strategy of blocking outside-in signaling has thus a definite advantage in designing new antithrombotic therapies. The myr-RGT peptide is known to be capable of blocking outside-in signaling by targeting c-Src and is considered as a synthetic molecule that is closest to be developed into drugs by virtue of structure-based analogues. However, there are key issues to be addressed before proceeding with the studies. A very primary issue is whether the inhibition of outside-in signaling by myr-RGT peptide depends on its membrane anchorage since a precise knowledge of the peptide action is definitely necessary for conceiving a structural analogue, i.e. membrane associable or not. Previous studies indicated that the regulatory effect of the myristoylated peptides on the cellular biological functions depends on its anchorage to cell membrane in some reports [
33,
34] whereas it seems not the case in others [
25,
26]. Glutathione is abundant in platelet cytoplasm at millimolar levels [
35], which are sufficient to exhaust the capacity of the disulfide bonds of myr-AC ~ CRGT peptide at its working concentrations and confer an inability of the CRGT peptide to associate with the myristoylation moiety, and thereby with the cell membrane. In other words, the competition of the CRGT peptide with the β3 tail for c-Src is unlikely attributed to the proximity of these molecules achieved by their membrane integration. Results showed that myr-AC ~ CRGT peptide penetrated the cell membrane and yielded free CRGT peptide in platelet cytoplasm (Fig.
1). This peptide did not disrupt the integrity of the cell membrane (Additional file
5: Figure S5) but did exert an inhibitory effect on platelet spreading on immobilized fibrinogen, clot retraction, irreversible aggregation, and P-selectin expression, without affecting soluble fibrinogen binding to αIIbβ3 and reversible aggregation (Additional file
2: Figure S2 and Additional file
3: Figure S3). In contrast to active c-Src phosphorylating β3 Y
747 and Y
759 independent of its binding to the RGT sequence of the β3 tail in vitro (Fig.
2), c-Src in platelets is constitutively associated with β3 in an inactive form, and is activated upon signaling through αIIbβ3. Once c-Src is dissociated by myr-AC ~ CRGT from β3, it will not be activated through signaling and is thus unable to phosphorylate the β3 tail. Indeed, myr-AC ~ CRGT inhibited agonist-induced c-Src activation (Fig.
5a) and attenuated the phosphorylation of the Y
747 and Y
759 residues of β3 in thrombin-stimulated platelets (Fig.
5b). These data brought new information, in comparison with those from myr-RGT, that the cytoplasmic existence of free RGT peptide is sufficient to sequestrate c-Src to an extent capable of separating it from β3 (Figs.
3b and
4b). This study further provides evidence that myr-AC ~ CRGT treatment did not affect the level of c-Src activation in platelets deposited on the immobilized fibrinogen (Additional file
2: Figure S2D), but did cause a downregulation of the platelet RhoA activation in clots (Additional file
2: Figure S2G) as well as a substantial inhibition on collagen-induced platelet aggregation (Additional file
2: Figure S2H). These results suggest that the RGT peptide regulated outside-in signaling by competing with β3 for c-Src not necessarily owing to its membrane association but, rather, through sequestrating c-Src and further indicated that the membrane anchorage is thus not an issue of importance in conceiving “drug-like” molecules mimicking the active structure of RGT peptide.
In line with the previous reports, the present observations showed that active c-Src directly phosphorylates β3 Y
747 and Y
759 independent of its binding to the RGT sequence of the β3 tail (Fig.
2) indicating that activated c-Src may still be able to promote αIIbβ3 signaling even having been separated from the β3 tail and in view of the fact that the free RGT peptide sequestrates c-Src in cytoplasm and that some peptides binding to c-Src has direct impact on its kinase activity [
20,
21]. It became essential to clarify whether the RGT peptide binding is able to alter the enzymatic activity of c-Src and accordingly to change the biological processes correlating with this multi-functional kinase. Our data show that the phosphorylation levels of Y
416 and Y
527 as well as the association of Csk with c-Src were not affected in platelets in the presence of free cytoplasmic CRGT peptide (Fig.
4). In addition, in vitro assays for c-Src activity, in which the phosphoryl transfer reaction occurs exclusively on the Src substrate peptide since the CRGT peptide contains no tyrosine residue, also revealed that the CRGT peptide did not influence the potential of c-Src to catalyze its substrates (Fig.
4f). These results suggest that RGT peptide binding to c-Src was unable to alter the c-Src activity and the presence of the cytoplasmic CRGT peptide caused primarily the dissociation of the Src/β3 interaction without a pronounced direct effect on c-Src. This feature for the RGT peptide argues in favor of the safety of its potential application in antithrombotic therapies.
Thrombosis occurs in the blood stream, and the shear forces play significant roles throughout the process. Selective inhibition of outside-in signaling was achieved by RGT peptide in human platelets based on experiments in static conditions [
15]. We thus became interested in testing the effect of myr-AC ~ CRGT on thrombus formation under flow conditions. We chose collagen to coat the micro-chambers because it is the major component of the subendothelial matrices and has thus been recommended by the Biorheology Subcommittee of the SSC of the ISTH [
36]. Platelet stable adhesion and irreversible aggregation regulated primarily through outside-in signaling [
30] contribute to the thrombus growth under flow. In a very recent publication [
37], Stalker and colleagues proposed a model for hemostatic plug formation in which the plugs are comprised of distinct regions defined by the degree of platelet activation, packing density, and stability. Platelets in the “core” are characterized by contact-dependent signaling, P-selectin exposure with a greater packing density while those in the unstable “shell” are loosely packed and less activated most likely dependent on the P2Y
12 receptors. Similar features may also exist in platelets that have undergone bidirectional or inside-out signaling through integrin αIIbβ3, respectively. Therefore, the RGT-treated platelets, in which only inside-out signals can be transduced, may still be able to form thrombi though less stable to resist the hydrodynamic drag forces. Indeed, at a shear rate of 125 s
−1, neither myr-AC ~ CRGT nor 7E3 affected the platelet adhesion, in line with the previous observations in which thrombasthenic platelets adhered normally to deendothelialized vessels at low shear rates [
38,
39]. The inability of myr-AC ~ CRGT to affect platelet adhesion and aggregation remained at the intermediate shear rates (500 s
−1) in contrast to an unequivocal inhibition induced by 7E3 (Fig.
6). At high shear rates (1500 and 5000 s
−1) myr-AC ~ CRGT partially inhibited thrombus formation, as evidenced by the smaller and thinner thrombi and less thrombus coverage as well (Fig.
6), whereas 7E3 could still vigorously inhibit platelet adhesion and aggregation leading to a drastic decrease of thrombus formation either in size or in coverage by showing only few platelets adhered on the surfaces. This was the first application of human platelets with specifically impaired outside-in signaling in such an observation indicating that the Src-regulated outside-in signaling played a pivotal role in inducing the firm linkage of platelets to the subendothelial matrices and to each other which renders the growing thrombi more resistant to the drag forces. Hemostasis usually occurs in small vessels such as arterioles, capillaries, and venules [
40] where blood loss will end as long as small numbers of platelets accumulate at the sites of injury [
37]. The flow experimental data allow us to infer that in venules (<500 s
−1), myr-AC ~ CRGT-treated platelets may adhere normally, while in arterioles and capillaries (1500 s
−1 or greater), the unaffected inside-out signaling may still enable these platelets to form thrombi to a size sufficient to seal the vessel damages in contrast to a much more profound inhibition of 7E3 [
31,
41]. This unique effect of myr-AC ~ CRGT on thrombus formation under flow may explain the consequences of the elimination of outside-in signaling in hemostatic potential manifested in genetically manipulated animals. Differently, pathological thrombosis takes place in stenotic medium-sized arteries with a luminal diameter exceeding 100 μm, usually in millimeters such as coronary or cerebral arteries. Coronary thrombosis model showed that shear rates reached as high as 84,000 s
−1 for a 65 % stenosis by diameter [
42]. Occlusion of these arteries requires bigger and more stable thrombi formed under these conditions. Significantly smaller and thinner thrombi formed by myr-AC ~ CRGT-treated platelets at high shear rates (1500 s
−1 and 5000 s
−1) (Fig.
6) indicate that the diminished platelet stable adhesion and irreversible aggregation impair the stabilization and growth of thrombi under higher flow and thus may prevent the complete occlusion of the stenotic medium-sized arteries in contrast to an adequate hemostatic situation that may be achieved in normal small vessels.
Integrins are αβ heterodimers and the β3 subunit forms αIIbβ3 and αvβ3 complexes; the former is expressed exclusively in megakaryocyte/platelet lineage and the latter is however expressed by multiple cell types [
43]. Integrin αvβ3 is also present in platelet, but only a few hundred copies in contrast to approximately 80,000 αIIbβ3 per platelet. Disruption of the Src/β3 interaction by the RGT peptide interferes with signal transduction through integrin αIIbβ3 in platelets and may also affect that through integrin αvβ3 in platelets and other cells such as melanoma cells or osteoclasts. The influence of the RGT-containing peptide on platelet function may be contributed by the global action of the peptide on the platelet β3 integrins while, as a commonly accepted concept, αIIbβ3 but not αvβ3 is crucial to platelet function owing to the overwhelming expression superiority. It would also be interesting to look at the effect of blocking the Src/β3 interaction in integrin αvβ3-expressing cells on tumorogenesis or bone resorption. That would be the goals for the future work.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
JH and XS performed the experiments and prepared the manuscript. WX analyzed data and contributed to paper writing. PL and ZL performed the experiments. XX designed the study, analyzed data, and wrote the paper. JH and XS are considered as co-first authors. All authors read and approved the final manuscript.