Integrin αIIb-Mediated PI3K/Akt Activation in Platelets

Integrin αIIbβ3 mediated bidirectional signaling plays a critical role in thrombosis and haemostasis. Signaling mediated by the β3 subunit has been extensively studied, but αIIb mediated signaling has not been characterized. Previously, we reported that platelet granule secretion and TxA2 production induced by αIIb mediated outside-in signaling is negatively regulated by the β3 cytoplasmic domain residues R724KEFAKFEEER734. In this study, we identified part of the signaling pathway utilized by αIIb mediated outside-in signaling. Platelets from humans and gene deficient mice, and genetically modified CHO cells as well as a variety of kinase inhibitors were used for this work. We found that aggregation of TxA2 production and granule secretion by β3Δ724 human platelets initiated by αIIb mediated outside-in signaling was inhibited by the Src family kinase inhibitor PP2 and the PI3K inhibitor wortmannin, respectively, but not by the MAPK inhibitor U0126. Also, PP2 and wortmannin, and the palmitoylated β3 peptide R724KEFAKFEEER734, each inhibited the phosphorylation of Akt residue Ser473 and prevented TxA2 production and storage granule secretion. Similarly, Akt phosphorylation in mouse platelets stimulated by the PAR4 agonist peptide AYPGKF was αIIbβ3-dependent, and blocked by PP2, wortmannin and the palmitoylated peptide p-RKEFAKFEEER. Akt was also phosphorylated in response to mAb D3 plus Fg treatment of CHO cells in suspension expressing αIIbβ3-Δ724 or αIIbβ3E724AERKFERKFE734, but not in cells expressing wild type αIIbβ3. In summary, SFK(s) and PI3K/Akt signaling is utilized by αIIb-mediated outside-in signaling to activate platelets even in the absence of all but 8 membrane proximal residues of the β3 cytoplasmic domain. Our results provide new insight into the signaling pathway used by αIIb-mediated outside-in signaling in platelets.


Introduction
Integrins are a and b heterodimeric receptors required for numerous essential biological processes [1]. The megakaryocyteand platelet-specific integrin aIIbb3 is essential for normal hemostasis [2]. Many integrins on platelets in suspension, including aIIbb3, are unable to bind their ligands or signal in their low-affinity state [1]. Transformation from the resting state to the active or high-affinity state typically results from integrin-mediated inside-out signaling initiated indirectly by activation of other receptors [1]. This transformation induced by inside-out signaling is controlled by the interaction between the membrane proximal, highly conserved regions of the cytoplasmic domains of the a and b subunits [3][4][5][6]. Disruption of this interaction by mutation results in the constitutive activation of the affected aIIbb3 heterodimers expressed in CHO and 293T cells [4,7,8]. Agonist-induced physiologic disruption of this interaction appears to be caused by the binding of talin [9], Kindlin [10] or other proteins [1,11] to the cytoplasmic domain of b3.
The active, oligomerized receptors can also initiate and propagate integrin-mediated outside-in signaling [1]. Binding of a specific ligand to active aIIbb3 (the gamma chain of Fg does not contain the RGD sequence) induces 'outside-in' signaling and results in platelet aggregation, granule secretion and thromboxane A2 (TxA2) production [12]. The contents of the granules, and TxA2 amplify the response to activation, and thereby recruit more platelets to the injury site to participate in thrombus formation [13]. The important roles of b3 in outside-in signaling have been well characterized [14][15][16]. For example, the b3 cytoplasmic domain can bind to a variety of molecules such as c-Src, FAK, and others that propagate outside-in signaling [14][15][16][17][18][19][20][21][22]. Several studies have reported that the calcium-and integrin-binding protein (CIB) can bind to the aIIb cytoplasmic domain and promote b3 mediated outside-in signaling [19][20][21]. More recently, an interesting observation shows that Ga13 directly binds to the b3 cytoplasmic domain and promotes b3 mediated outside-in signaling [14]. And SHARPIN, a a-subunit cytoplasmic domain associated protein, inhibits the recruitment of talin and Kindlin to the integrin, and consequently inhibits the critical switching of b1integrins from inactive to active conformations [22]. Compared with the deep understanding of b3 subunit-mediated outside-in signaling, little is known about the signaling components utilized by aIIb-mediated outside-in signaling.
In 2005, we reported that aIIb-mediated outside-in signaling is enhanced in platelets of a patient lacking the terminal 39 residues of the b3 cytoplasmic tail (b3D724) [23]. This signaling was detected as thromboxane A2 (TxA2) production and granule secretion, and required ligand cross-linking of aIIbb3 and platelet aggregation. This enhanced outside-in signaling was specifically inhibited by a palmitoylated version of a peptide corresponding to b3 cytoplasmic domain residues R724-R734 (sequence RKE-FAKFEEER). Those results demonstrate that aIIb-mediated outside-in signaling resulting in TxA2 production and granule secretion is negatively regulated by RKEFAKFEEER, a sequence of residues in a membrane distal region of the b3 cytoplasmic domain. In this study, we identify some of the signaling molecules that mediate aIIb outside-in signaling.

Results
Src Family Kinase (SFK) and PI3K are Key Factors Involved in aIIb-mediated Platelet Activation In previous work, we showed that the treatment of human b3D724 platelets in presence of fibrinogen with LIBS-specific mAb D3 and PT25-2, respectively caused aggregation, TxA2 production and granule secretion [23]. Here, we identify some of the signaling molecules that appear to be involved in these aIIbmediated responses by using kinase inhibitors. Many studies have demonstrated that Src family kinases [14,17,18], PI3K/Akt signaling [24] and MAPK signaling [25] play important roles in platelet activation. But, this direct SFK signaling and subsequent downstream signaling are thought to be dependent of the cytoplasmic domain of b3 [17]. Therefore, the broad spectrum Src family kinase (SFKs) inhibitor PP2, the PI3K inhibitor wortmannin and the MAPK inhibitor U0126 were used in this study.
As we reported previously [23], in contrast to normal platelets, which neither underwent aggregation (they agglutinated, no shape change- Fig. 1A) nor produced TxA2 or secreted the contents of their storage granules, the D724 platelets aggregated, produced TxA2, and secreted ATP and PF4 in response to treatment by D3 plus Fg. All these responses were inhibited by the 7E3 (anti human platelet aIIbb3) (Fig. 1A, B), which confirmed our previous results that D724 platelet TxA2 production and secretion induced by D3 plus Fg are aggregation-dependent.
Metabolic inhibitors provided new insight into aIIb outside-in signaling. The Src family kinase inhibitor PP2 and the PI3K inhibitor wortmannin eliminated shape change, and biphasic aggregation and therefore (because those responses require aggregation [23]) TxA2 production and granule secretion by D724 platelets in response to D3 plus Fg. These results demonstrate that SFK and PI3K signaling are required for the activation of D724 platelets induced by D3 plus Fg. In contrast, the COX inhibitor indomethacin blocked TxA2 production, but had little effect on aggregation and only a moderate effect on granule secretion. Apyrase, an ADP scavenger, had little effect on aggregation, no effect on TxA2 production and a small effect on PF4 secretion. These results revealed that under these conditions, D3 plus Fg induced D724 platelet activation is not obviously dependent on signal amplification mediated by TxA2 and ADP. TxA2 production, granule secretion, and aggregation, were slightly diminished in D724 platelets treated with MAPK inhibitor U0126 demonstrating that Erk1/2 apparently does not play an essential part in aIIb (D3 plus Fg induced) elicited D724 platelet activation (Fig. 1A, B). In summary, despite the fact that D724 platelets lack the membrane distal cytoplasmic domain of b3, aIIb elicited platelet aggregation, TxA2 production and granule secretion are SFK and PI3K dependent.

Akt Phosphorylation in D724 Platelets Treated with D3 Plus Fg is Dependent on a SFK(s) and PI3K
Since wortmannin eliminated aIIb outside-in signaling-dependent aggregation, TxA2 production and granule secretion, activation of the PI3K/Akt pathway in D724 platelets treated with D3 plus Fg was investigated. In contrast to the absence of Akt Ser473 phosphorylation in normal platelets, Akt Ser473 was phosphorylated in D724 platelets in the presence of Fg in response to D3 ( Fig. 2A). Phosphorylation of Akt Ser473 was diminished in the presence of indomethacin plus apyrase indicating that Akt Ser473 phosphorylation may be partially dependent on signal amplification. The Src family kinase inhibitor PP2 and the PI3K inhibitor wortmannin each totally blocked the Akt Ser473 phosphorylation, as well as shape change, aggregation, TxA2 produciton and secretion (Fig. 1). These results demonstrate that D3 plus Fg induced Akt phosphorylation in D724 platelets is SFK(s) and PI3K dependent; and, that D3+Fg induced platelet activation apparently is Akt-dependent-though further work is required to establish that point.
Our previous work showed that the palmitoylated peptide RKEFAKFEEER (p-RKEFAKFEEER corresponds to b3 cytoplasmic tail residues 724-734) can block aIIb-mediated platelet aggregation, TxA2 production and granule secretion in response to stimulation by D3 plus Fg without preventing activation of aIIbb3 [23]. Here, our data reveal that p-RKEFAKFEEER, in contrast to the palmitoylated scrambled control peptide (p-EAERKFERKFE), totally blocks the phosphorylation of Akt Ser473 in D724 platelets in response to D3 plus Fg (Fig. 2C), thereby demonstrating that aIIb-mediated activation of PI3K/Akt apparently is also negatively regulated by b3 cytoplasmic domain residues R 724 KEFAKFEEER 734 .
PAR4 Agonist Peptide (AYPGKF) Induced Akt Phosphorylation in Platelets is aIIbb3 Dependent and Regulated by the b3 R 724 KEFAKFEEER 734 Cytoplasmic Domain We have shown that aIIbb3 outside-in signaling induced by cthrombin in normal human platelets is negatively regulated by p-RKEFAKFEEER, which implies that aIIb-mediated outside-in signaling is also involved in normal platelet activation [23]. cthrombin activates platelets through PAR family receptors on the platelet surface. Here, we characterize aIIb-mediated outside-in signaling in normal platelet activation mediated by PAR family receptors by using mouse platelets. We used the specific PAR4 agonist peptide AYPGKF instead of c-thrombin because PAR4 is the receptor that mediates signaling in mouse platelets [24,26]. We used Tp deficient mouse platelets to block the amplification of signaling induced by TxA2, and used b3 deficient mouse platelets to investigate the function of aIIbb3 in PAR4 mediated platelet activation. As shown in Figure 3A and 3B, wild type mouse platelets aggregated in response to AYPGKF, and Akt was phosphorylated. In contrast, Tp 2/2 platelets neither aggregated as extensively, nor phosphorylated Akt as extensively as occurred in wild type platelets, in response to stimulation by AYPGKF. Also, apyrase totally inhibited Tp 2/2 platelet aggregation and Akt phosphorylation induced by AYPGKF. These results implied that PAR4 mediated platelet activation and Akt phosphorylation are amplification dependent. However, exogenous Fg eliminated the inhibitory effects of apyrase. Therefore, a-granule secretion is ADP-dependent, and Akt phosphorylation caused by AYPGKF apparently is aggregation dependent. This conclusion was confirmed by the ability of the mAb 1B5 (prevents Fg binding to mouse aIIbb3) to prevent platelet aggregation and Akt phosphorylation in platelets stimulated by the PAR4 agonist peptide in the presence of Fg (Fig. 3A, B). Further confirmation of this conclusion was provided by the inability of b3 deficient platelets (which cannot aggregate) stimulated with AYPGKF to phosphorylate Akt (Fig. 3C, D). These results demonstrate that PAR4 peptideinduced Akt phosphorylation in platelets is aIIbb3 and aggregation dependent.
Further insight into AYPGKF-induced signaling was provided by the use of metabolic inhibitors and peptides. AYPGKF induced aggregation of wild type murine platelets was inhibited by the SFK inhibitor PP2 and the PI3K inhibitor wortmannin (Fig. 4A). Phosphorylation of Akt was completely inhibited by PP2, and diminished 95% by wortmannin (Fig. 4B). The palmitoylated b3 peptide p-RKEFAKFEEER, but not the scrambled p-control peptide (p-EAERKFERKFE) inhibited AYPGKF-induced platelet aggregation (Fig. 4C) and Akt phosphorylation (Fig. 4D). The interpretation of the ability of p-RKEFAKFEEER to inhibit AYPGKF-induced aggregation and Akt phosphorylation is that both of these events are dependent on aIIb-mediated outside-in signaling. Therefore, PAR4 agonist peptide induced activation of normal mouse platelets appears to be caused by aIIb mediated outside-in signaling through SFK and PI3K/Akt signaling pathway.
The Role of the b3 Cytoplasmic Domain and the R 724 KEFAKFEEER 734 Sequence in CHO Cell Spreading on Immobilized Fg The palmitoylated peptide p-RKEFAKFEEER blocked secretion and Akt phosphorylation in human D724 platelets treated with D3 plus Fg, and in wild type murine platelets stimulated by the PAR4 agonist peptide. These results confirm our previous observation that the b3 R 724 KEFAKFEEER 734 sequence negatively regulates aIIb mediated outside-in signaling. To evaluate this conclusion and investigate the pathway used for aIIb mediated outside-in signaling and regulation using an alternative approach [3,4,27], three CHO cell lines stably expressing either human aIIbb3, aIIbb3-scramble [a form of aIIbb3 containing a scrambled version of the R 724 KEFAKFEEER 734 sequence, (E 724 AER-KFERKFE 734 )], or aIIbb3-D724 were constructed. These stable CHO cell lines were evaluated for their ability to spread on immobilized Fg. As with CHO cells transfected with control (no insert) vectors, CHO cells expressing aIIbb3-D724 did not spread on immobilized Fg, while CHO cells expressing normal aIIbb3 (designated b3) or aIIbb3-scramble spread fully on immobilized Fg (Fig. 5A). These results demonstrate that the b3 cytoplasmic domain plays a critical role in aIIbb3-mediated CHO cell spreading, and that the b3 sequence R 724 KEFAKFEEER 734 is not required for spreading because CHO cells expressing aIIbb3scramble (E 724 AERKFERKFE 734 ) rather than the normal b3 sequence R 724 KEFAKFEEER 734 , spread normally. Spreading by CHO cells expressing normal aIIbb3 supports the conclusion that aIIb outside-in signaling is not required for the spreading of CHO cells expressing normal aIIbb3 on immobilized Fg. Apparently as with platelets [28,29], truncated b3 expressed in CHO cells is unable to mediate the interaction of aIIbb3-D724 with the cytoskeleton, so those CHO cells are unable to spread on immobilized Fg (Fig. 5, [28]).
aIIb mediated PI3K/Akt signaling in CHO cells. Akt phosphorylation was also investigated in all these cell lines in suspension, treated with D3 plus Fg. There was no Akt phosphorylation in the absence of stimulation (resting condition) (Fig. 5B). After stimulation with D3 plus Fg, there was a very low level of Akt phosphorylation in CHO cells expressing normal aIIbb3. In contrast, Akt phosphorylation was significantly enhanced in CHO cells expressing either aIIbb3-scramble or aIIbb3-D724. These latter results provide in vivo confirmation, in the absence of p-RKEFAKFEEER, that aIIb-mediated activation of the PI3K/Akt pathway is regulated by b3 cytoplasmic domain sequence R 724 KEFAKFEEER 734 . In view of the results obtained with the modified CHO cells in suspension, it cannot reasonably be argued that the inhibitory results obtained by treating AYPGKF stimulated wild type mouse platelets with p-RKEFAK-FEEER were the result of interference of the palmitoylated peptide with b3-mediated inside-out or outside-in signaling. This follows from the fact that replacing the normal b3 sequence R 724 KEFAKFEEER 734 with a scrambled version of that sequence (E 724 AERKFERKFE 734 ) eliminated the ability of the cytoplasmic domain of b3 to prevent Akt phosphorylation.

Discussion
Bidirectional aIIbb3 mediated signaling plays an essential role in hemostasis and thrombosis. Contrary to b3-mediated bidirectional signaling, aIIb mediated signaling has not been investigated in detail. Our previous work documented that either of two LIBS-specific mAbs (D3 and PT25-2) plus Fg (but none of the agents alone [23]) can induce human b3D724 platelets to undergo shape change, biphasic aggregation (not simply agglutination), secrete the contents of their storage granules and produce TxA2 via aIIb mediated outside-in signaling [23]. Furthermore, those responses were shown to be negatively regulated by the membrane distal b3 cytoplasmic domain residues R 724 KEFAK- FEEER 734 [23]. Here, we identify some signaling molecules activated by this aIIb-mediated outside-in signaling in platelets.
The data presented in Figures 1 and 2 demonstrate that the SFK inhibitor PP2, and PI3K inhibitor wortmannin block aIIb outside-in signaling-induced shape change, platelet aggregation, granule secretion and TXA2 production, thereby demonstrating that a Src family kinase(s) and PI3K/Akt signaling propagate aIIbmediated outside-in signaling, even in platelets lacking all but eight membrane proximal residues of the b3 cytoplasmic domain. The immunoblotting results in Figure 2A show that aIIb-elicited Akt phosphorylation was inhibited by PP2, and by wortmannin which confirmed the results of preliminary inhibitor screening, and also revealed that the Src family kinase function is upstream of PI3K/ Akt signaling as PP2 totally blocked Akt phosphorylation. We have reported that the palmitoylated peptide p-RKEFAKFEEER, the sequence of which corresponds to b3 cytoplasmic domain residues R 724 KEFAKFEEER 734 , negatively regulates aIIb-mediated outside-in signaling, including TxA2 production and secretion [23]. The data in Figure 2C demonstrate that the palmitoylated peptide p-RKEFAKFEEER, but not the palmitoylated control peptide (a scrambled version of RKEFAKFEEER with the sequence EAERKFERKFE) inhibited Akt phosphorylation in D724 platelets treated with D3 plus Fg. Therefore, aIIb elicited platelet aggregation, granule secretion and TxA2 production induced by D3 plus Fg treatment of D724 platelets is SFK-PI3K/Aktdependent.
The broad physiologic relevance of aIIb-elicited activation of the PI3K/Akt pathway in platelets is demonstrated by the data in Figures 3 and 4. PAR4 is one of a number of types of thrombin receptors in mouse platelets [30]. The peptide AYPGKF elicits PAR4-induced signaling that causes aIIbb3 activation [31]. The active aIIbb3 binds to Fg and initiates amplifying signaling that enhances platelet aggregation. Here, we found that Akt phosphorylation induced in platelets by a low level of AYPGKF is secretion and aggregation dependent. As with D724 platelets, p-RKEFAKFEEER, but not the control palmitoylated peptide, inhibited aggregation and Akt phosphorylation induced in wild type platelets by AYPGKF. So, the negative regulation of aIIbelicited activation of the PI3K/Akt pathway by the b3 cytoplasmic domain residues R 724 KEFAKFEEER 734 apparently is a normal aspect of the regulation of aIIbb3-mediated bidirectional signaling in platelets.
In order to confirm the function of b3 cytoplasmic domain residues R 724 KEFAKFEEER 734 on aIIb-mediated activation of the PI3K/Akt pathway by a means other than the use of p-RKEFAKFEEER, three CHO cell lines expressing integrin aIIbb3-WT, aIIbb3-scramble or aIIbb3-D724 were constructed. CHO cells expressing aIIbb3-D724 failed to spread on immobilized Fg demonstrating that the b3 cytoplasmic domain is required for CHO cell spreading on fibrinogen, as with platelets [28,31]. CHO cells expressing aIIbb3-WT or aIIbb3scramble, spread well. These results confirm that the b3 cytoplasmic domain plays a critical role in platelet spreading, but also provide the new insight that the b3 cytoplasmic domain sequence R 724 KEFAKFEEER 734 does not participate in CHO cell and probably platelet spreading in a sequence-specific manner. So, integrin aIIbb3 mediated cell spreading is mainly through b3-mediated outside-in signaling, not aIIb-mediated outside-in signaling. And as stated above, b3 cytoplasmic domain sequence R 724 KEFAKFEEER 734 is not required for cell spreading-though it isn't known whether or not deletion of those residues rather than their substitution with EAERK-FERKFE, a scrambled version of R 724 KEFAKFEEER 734 would enable spreading. In this regard, the ability of PMA to cause spreading of CHO cells expressing aIIbb3-D724 (Supplemental Figure S1) supports the conclusion that the b3 CD is required to attach to and/or activate the cytoskeleton of CHO cells, as with platelets [31]. In other words, PMA activation of PKC apparently activates the cytoskeleton thereby enabling the CHO cells expressing aIIbb3-D724 to spread. Activation of the cytoskeleton results in cell spreading even though the truncated b3 presumably cannot directly link to the cytoskeleton. The lack of activation of aIIbb3-D724 is not a barrier to CHO cell spreading on immobilized Fg as resting platelet aIIbb3 can support cell attachment to Fg [28,[32][33][34].
Regarding D3 plus Fg treatment of CHO cells in suspension, extensive Akt phosphorylation occurred in CHO cells expressing aIIbb3-D724, and substantially less occurred in CHO cells expressing aIIbb3-scramble in response to stimulation by D3 plus Fg (Fig. 5B), and there was no Akt phosphorylation in CHO cells expressing normal aIIbb3. These results demonstrate that b3 cytoplasmic domain residues R 724 KEFAKFEEER 734 negatively regulate aIIb-mediated activation of the PI3K/Akt pathway, and that other b3 cytoplasmic domain residues may also be involved in this regulation.
The potential physiological significance of the aIIb outside-in signaling is that the resulting platelet aggregation presumably could contribute to hemostasis and thrombosis in response to a low, physiologically relevant level of stimulation by an agonist produced/released in response to vascular injury. The aIIb signaling would initiate a positive feedback loop that amplifies the original weak activation induced by the injury, thereby facilitating hemostasis. Assuming the physiological relevance of botrocetin/ vwf in-vitro signaling studies [35], aIIb outside-in signaling elicited by vwf/GPIb interaction following arterial plaque rupture or endothelial injury probably would facilitate thrombus formation. This suggestion is supported by the observation that vwf/GPIb interaction initiates signaling that activates aIIbb3, and the subsequent platelet aggregation enhances TxA2 production, dense granule secretion and Akt phosphorylation [36]. In fact, aIIbb3 mediated aggregation resulted in about 4 times more TxA2 production than vwf/GPIb mediated agglutination did in the absence of aggregation [37]. Further work is required to test these conjectures.
A number of unanswered questions are raised by our observations. For example, the mechanism of the transfer of signaling information from the cytoplasmic tail of aIIb to a Src family kinase(s) and the PI3K/Akt pathway is unknown. Likewise, the identity of the SFK(s) activated by aIIb signaling is unknown. Also, the mechanism of negative regulation of aIIb-mediated outside-in signaling by key amino acids in b3 R 724 KEFAKFE-EER 734 sequence is unknown. Further work is required to resolve these issues.
In summary, we found that SFK(s) elicited PI3K/Akt signaling is utilized by aIIb-mediated outside-in signaling to activate platelets. Also, our data demonstrate that the negative regulation of aIIb-mediated outside-in signaling by the b3 cytoplasmic domain residues R 724 KEFAKFEEER 734 prevents activation of the PI3K/Akt pathway. Therefore, our results identified some of the components used by platelets to propagate aIIb-mediated outside-in signaling, an important aspect of platelet function that has not been characterized in detail.

Ethics Statement
After written, informed consent was obtained, and as part of a routine health exam, extra blood was collected for research purposes from the variant thrombasthenia patient. Approval for the provision of blood for research purposes was obtained from the Institutional Review Board of the Cincinnati Children's Hospital. After written informed consent, blood was also obtained from a healthy donor to serve as a source of control platelets. The animal research was approved by the Shanghai Jiao Tong University School of Medicine Animal Care and Use Committee (Approve No. SYXK2008-0050).

Mice
The TxA2 receptor deficient (Tp 2/2 ) mice were derived as described [40], wild type littermate siblings were used as controls. Dr. Steven Teitelbaum at Washington University supplied b3 deficient and littermate control mice.

Patient
The patient has a variant form of Glanzmann thrombasthenia designated here as b3D724 [28]. The patient is a young man with a life-long history of enhanced bruising, mucosal bleeding, and petechiae. He has a normal platelet count but a prolonged bleeding time. The patient contains 2 distinct mutant alleles for b3: an allele that contains a transition (CGA [R] to TGA [nonsense]) mutation that results in a truncated b3 lacking all but the 8 membrane proximal cytoplasmic domain residues. The other allele sustains a deletion of b3 T1181 resulting in a frame shift and a nonsense codon corresponding to amino acid 642. So, the abnormal platelets express only aIIbb3 complexes containing the truncated b3 subunit, which are otherwise structurally normal. These complexes are expressed at about 40% of the normal level. As a consequence of the truncation, the platelets do not aggregate to physiologic agonists because the truncated b3 subunit cannot mediate inside-out signaling to activate the altered aIIbb3 complex.

Human Washed Platelets Preparation and Aggregation
After informed consent was obtained, blood was collected from healthy donors and the variant thrombasthenia patient into empty syringes and then transferred to polypropylene centrifuge tubes containing 100 ml/mL (final concentration) Whites anticoagulant (2.94% sodium citrate, 136 mM glucose, pH 6.4), 0.1 mg/mL PGE1, and 1 U/mL apyrase. Platelet-rich plasma (PRP) was prepared by differential centrifugation. Washed platelets were prepared as described [23]. Aggregation was measured in the Lumi-Aggregometer (Chrono-Log, Havertown, PA) using washed platelets (300 ml) adjusted to approximately 10 6 platelets / ml with stirring at 1000 rpm.

Measurement of ATP Secretion
ATP secretion was evaluated using the CHRONO-LUME reagent (CHRONO-LOG Corp., Havertown, PA) as described [12]. After 5 min activation of the stirred platelets (1000 rpm), 10 ml reagent was added directly to each platelet suspension. The luminescence intensity was measured at luminescence setting of 60.001.

Measurement of TxA2 Production
TxB2, a stable metabolite of TxA2, was measured as described [12,42]. TxB2 was measured to indirectly estimate TxA2 production. After a 5 min aggregation period, platelets were removed by centrifugation in the presence of 5 mM EDTA. The supernatant fluid was collected and diluted 1:50 with the assay buffer supplied in the TxB2 enzyme immunoassay (EIA) kit (Assay Designs, Ann Arbor, MI). TxB2 was measured according the manufacture's protocol.

Measurement of PF4 Secretion
PF4 (platelet factor 4) is released from the a-granules by the activated platelets. The supernatant fractions of platelet suspensions were collected and assayed using Asserachrom PF4 quantitative enzyme-linked immunosorbent assay (ELISA) kit (Diagnostica Stago, Parsippany, NJ) as described [43].

Construction of Stable CHO Cell Lines, Spreading of the CHO Cell Lines on Immobilized Fg and Stimulation of Suspensions of the Cell Lines with D3 Plus Fg
Lenti-X cell (Clontech) was used as the lentivirus packaging cell. Lentiviruses were produced by co-transfecting Lenti-X cells with pLVX-IRES-ZsGreen/pLVX-IRES-puro, pspAx2 and pMD2G at 3:2:1 proportion. The lentiviruses were harvested after transfection, and depletion of the cell debris by centrifugation. CHO cells (Chinese Hamster Ovary cells) were infected with a lentivirus mixture of blank vectors, aIIb and b3, aIIb and b3-scramble, aIIb and b3-D724, respectively. Forty eight hours after infection, the infected CHO cells were cultured for two weeks in selective medium containing puromycin (7.5 mg/ml). The viable, GFP positive CHO cells were sorted by FACS. Then the sorted CHO cells were cultured in medium containing puromycin (5 mg/ml) for two weeks. This procedure resulted in the isolation of stable CHO cell lines expressing blank vectors, aIIbb3, aIIbb3-scramble, or aIIbb3-D724, respectively. Then, these cell lines were cultured in medium without serum for 12 hrs before further experiments.
The CHO cell lines were allowed to spread for 2 hrs at 37uC on chamber slides coated with 100 mg/ml Fg. Adherent cells were rinsed with PBS and fixed with 4% paraformaldehyde at room temperature. Then the cells were washed with labeling buffer (0.5% BSA, 0.5% Triton X-100 in PBS) and incubated with Rhodamine-phalloidin for 30 min at room temperature for F-actin staining. The slides were washed twice with PBS for the fluorescence microscopy survey. Cell spreading size was quantified using National Institutes of Health (NIH) ImageJ. Statistical significance between groups was determined using a Student t test.
For D3 plus Fg stimulation, the cells were detached from the plates, and resuspended in modified Tyrode solution containing 2 mM Ca 2+ , stimulated with D3 (5 mg/ml), Fg (100 mg/ml) at 37uC for 5 min with stirring at 1000 rpm. Lysates of CHO cells were prepared by adding an equal volume of 26 SDS loading buffer to each cell suspension.

Western Blot Analysis
Each platelet and CHO cell lysate was boiled at 100uC for 10 min, then resolved on a sodium dodecyl sulfate (SDS), 10% polyacrylamide gel and then transferred to PVDF membrane. Western blots were performed using anti-phospho-Akt (Ser473), or an anti-b-actin antibody at a 1:1000 dilution, followed by incubation with horseradish peroxidase-conjugated goat antirabbit antibody at a 1:5000 dilution. Blots were developed using Supersignal chemiluminescent substrate (Pierce, Rockford, IL).