Deubiquitinases modulate platelet proteome ubiquitination, aggregation, and thrombosis

Abstract

Objective

Platelets express a functional ubiquitin-proteasome system. Mass spectrometry shows platelets contain several deubiquitinases, but whether these are functional, modulate the proteome, or affect platelet reactivity are unknown.

Approach and Results

Platelet lysates contained ubiquitin-protein deubiquitinase activity hydrolyzing both Lys₄₈ and Lys₆₃ poly-ubiquitin conjugates that was suppressed by the chemically unrelated deubiquitinase inhibitors PYR41 and PR619. These inhibitors acutely and markedly increased mono- and poly-ubiquitination of the proteome of resting platelets. PYR41 (i.v., 15 min) significantly impaired occlusive thrombosis in FeCl₃-damaged carotid arteries, and deubiquitinase inhibition reduced platelet adhesion and retention during high shear flow of whole blood through microfluidic chambers coated with collagen. Total internal reflection microscopy showed adhesion and spreading in the absence of flow was strongly curtailed by these inhibitors with failure of stable process extension, and reduced retraction of formed clots. Deubiquitinase inhibition also sharply reduced homotypic platelet aggregation in response to the incomplete agonists ADP and collagen acting through GPVI, but also to the complete agonist thrombin. Suppressed aggregation was accompanied by curtailed PAC-1 binding to activated IIb/IIIa and inhibition of P-selectin translocation to the platelet surface. Deubiquitinase inhibition abolished the agonist-induced spike in intracellular calcium, suppressed Akt phosphorylation, and reduced agonist-stimulated PTEN phosphatase phosphorylation. Platelets express the proteasome-associated deubiquitinases USP14 and UCHL5, and selective inhibition of these enzymes by b-AP15 reproduced the inhibitory effect of the general deubiquitinase inhibitors on ex vivo platelet function.

Conclusions

Remodeling of the ubiquitinated platelet proteome by deubiquitinases promotes agonist-stimulated intracellular signal transduction and platelet responsiveness.

Introduction

Platelet activation underlies thrombotic cardiovascular disease, responsible for significant death and disability in the developed world,^{1, 2} and modulation of platelet activation is an established therapeutic target for these diseases. Platelets are anucleate and lack transcription, so their activation is transduced by the proteome delivered to them during thrombopoiesis, although this is augmented by limited de novo protein synthesis.³ Conversely, inhibitors show platelets also possess a limited ability to reduce their proteome through the ubiquitin-proteasome proteolytic system that participates in their production during thrombopoiesis and contributes to the functions of activated cells.^4–6 Analysis of the platelet proteome by quantitative mass spectrometry⁷ identifies the expected components of the ubiquitin ligase system, but also identifies deubiquitinases at high copy number. These enzymes might modify the pattern of ubiquitin chains conjugated to the platelet proteome, but this is unstudied.

Covalent modification of proteins with ubiquitin is dynamic and reversible with six families of evolutionarily conserved deubiquitinases hydrolyzing these mono- and polymeric ubiquitin protein adducts.⁸ Deubiquitinases are isopeptidases that play pivotal roles in ubiquitin-mediated signaling pathways and deubiquitinase inhibitors alter diverse cellular functions, as anticipated from the range of processes employing ubiquitin adduction. Accordingly, some deubiquitinase inhibitors have therapeutic potential.⁹ The general deubiquitinase inhibitor PR619¹⁰ promotes autophagy, protein aggregation, and the unfolded protein response in nucleated cells.^{11, 12} A small molecule inhibitor of E1 ubiquitin activating enzyme, PYR41¹³, suppresses arachidonate-stimulated adhesion and migration of tumor cells on a collagen surface¹⁴, angiotensin II-mediated dendritic cell activation¹⁵, and NF-κB activation in tumor cells,¹³ However, PYR41 also leads to accumulation of ubiquitinated proteins ex vivo and in vitro by inhibiting deubiquitinases.¹⁶ The novel small molecule inhibitor b-AP15 that is highly specific for the proteasome-associated deubiquitinases USP14 and UCHL5 displays potent anti-tumor activity and induces cytotoxicity in multiple myeloma cells resistant to the proteasome inhibitor bortezomib.^{17, 18}

Inhibition of the proteasome quells the ultimate step of ubiquitin-mediated protein degradation, but layers of regulated processes lie upstream of this proteolytic machine. We determined whether ubiquitination of the platelet proteome was dynamic and whether modification of ubiquitin-protein adducts contributes to platelet function. We find platelets contain active deubiquitinases that regulate ex vivo platelet aggregation, adhesion, and activation, and that deubiquitinase inhibition reduced occlusive thrombosis in vivo. Specific inhibition of proteasome-associated deubiquitinases recapitulated modulation of platelet function by general deubiquitinase inhibitors, highlighting the active participation of the proteasome in platelet activation.⁵ Our data suggest that the modulation of platelet deubiquitinase activity is a potential new approach to anti-platelet therapy.

Materials and Methods

Materials and Methods are available in the online-only Data Supplement

Results

Platelets express functional deubiquitinases

The platelet proteome is decorated with poly-ubiquitin chains that direct select proteins to the proteasome for degradation after activation,⁵ suggesting the pattern of ubiquitin conjugation should be dynamic. We treated platelet lysates with HA-tagged ubiquitin vinyl sulfone,¹⁹ which covalently adducts deubiquitinases during hydrolysis of this isopeptidase substrate, to determine whether platelets contain enzymes that actively modify conjugated poly-ubiquitin chains. We found that platelets contain up to 10 differently sized proteins labeled by this mechanism-based deubiquitinase inhibitor (Fig. 1A). The general deubiquitinase inhibitor PYR41 and a structurally unrelated deubiquitinase inhibitor PR619 decreased catalytic labeling of each of these active platelet deubiquitinases. We directly assayed poly-ubiquitin deconjugating enzymatic activity in platelet lysates to find platelets expressed ubiquitin conjugate isopeptidase enzymatic activity that was inhibited by PYR41 and PR619 (Fig. 1B).

Figure 1. Platelets express functional deubiquitinases.

(A) Active deubiquitinases. Western blot of anti-HA antibody after active site–directed labeling of deubiquitinases with the mechanism based inhibitor HA-tagged Ub-vinylsulfone. Lysates from washed platelets were treated with or without PYR41 or PR619 prior to labeling (n=3). (B) Reduction of deubiquitinase activity by inhibitors. The isopeptidase activity of platelet deubiquitinases was assayed using a luminogenic substrate in the lysates of untreated and PYR41 or PR619 treated platelets (n=8; *p≤ 0.05). PYR41 or PR619 pretreatment limits hydrolysis of both (C) Lysine48- linked and (D) Lysine63- linked heptameric chains ( —> ) as detected by western blotting with anti-ubiquitin antibody (P4D1, n=2). (E) PYR41 or PR619 pretreatment (30 min) increased ubiquitination of the platelet proteome. Immunoblot of platelet lysates, treated or not with PYR41 or PR619, probed with anti-ubiquitin antibody (P4D1, n=3). The left lane shows commercial ubiquitin contains both monomeric (~8 kDa) and dimeric species.

Polyubiquitin chains most commonly polymerize through K₄₈, and in general these proteins are then recognized by the proteasomal system. Alternatively, polymerization through K₆₃ generally alters protein distribution. Deubiquitinases are selective for these isomeric ubiquitin chain linkages,⁸ and platelet lysates catalyzed rapid disassembly of heptameric K₄₈-linked (Fig. 1 C) and K₆₃-linked (Fig. 1D) poly-ubiquitin chains. The deubiquitinase inhibitors, PYR41 and PR619 protected some of the short K₄₈ polymers from proteolysis, but not the K₆₃ polymer. Treating intact platelets for 30 min with either PYR41 or the PR619 increased the content of polyubiquitin conjugated proteins relative to pattern of conjugation found in quiescent platelets (Fig. 1E). The increase in decorated proteins was apparent for most of the protein bands, but was prominent in the more slowly migrating proteins that no longer appeared as distinct bands. Thus, inhibition of deubiquitinase activity shows quiescent platelets must rapidly and continuously cycle ubiquitin through polymeric ubiquitin chains adducting select proteins.

Deubiquitination of the platelet proteome promotes thrombosis

We investigated whether deubiquitinase inhibition altered platelet function during arterial thrombosis created by injuring the wall of murine carotid arteries in situ with FeCl₃. This damage results in rapid platelet accretion with formation of a platelet-rich occlusive barrier at the site of injury.^{20, 21} Typically, complete cessation of flow through the artery occurred 12 min after the brief exposure to ectopic FeCl₃ in animals treated with the DMSO vehicle (Fig. 2A). However, disruption of ubiquitin metabolism by intravenous injection of PYR41 15 min prior to vessel injury significantly lengthened the time to occlusion to 26 min, consistent with the delay induced by inhibition of the platelet proteasome.⁵

Figure 2. Deubiquitinase inhibitors suppress platelet activation and thrombosis.

(A) The deubiquitinase inhibitor PYR41 prolongs the time to vascular occlusion. Mice were injected with PYR41 or DMSO and thrombosis was induced by application of FeCl₃ 15 min later to a surgically exposed murine carotid artery as described in “Methods”. Time to complete cessation of blood flow in the murine carotid artery was determined using intravital microscopy (n=5 experimental, 3 control; **p≤ 0.01). (B) PYR41 or PR619 pretreatment blocked platelet adhesion to collagen at high shear. Calcein-AM labeled blood, treated or not with PYR41 or PR619, was perfused over immobilized type 1 collagen fibrils (150 μg/ml) at 67.5 dyne/cm² for 3 min. Images are representative fields taken from three independent experiments that yielded similar results (n=3). (C) Area of platelet attachment after PYR41 or PR619 treatment. Platelet area in panel B was quantified by ImagePro plus software and results are plotted as area of platelet adhesion in square microns (n=3; ***p≤ 0.001).

We modeled platelet accretion ex vivo by flowing whole human blood through a collagen-coated microfluidic channel that generates high shear. Fluorescently labeled platelets in whole blood were immobilized along the length of the chamber, as shown in a typical video frame captured at the distal end of the chamber after 3 min of flow (Fig. 2B, Supplementary video I). Preincubating blood with either PYR41 or PR619 reduced the number of platelets adhering in the collagen-coated chamber by 80% (Fig. 2C), and this difference was significant (p< 0.05). These outcomes show the adhesive phenotype of activated platelets depends on rapid deubiquitination of their proteome.

Deubiquitinase inhibition impairs the function of isolated platelets

We purified and washed human platelets, and then assayed a range of functions to determine whether deubiquitinase inhibitors directly act on platelets to alter their function, and not just nucleated cells of the vasculature or blood. Aggregometry showed that aggregation induced by the incomplete agonists ADP (Fig. 3A) or collagen (Fig. 3B) was strongly suppressed by PYR41 and abolished after pre-treatment with PR619. These two inhibitors had an identical effect on the more robust aggregation induced by the complete agonist thrombin (Fig. 3C). These effects were stimulus-dependent because neither compound by itself promoted aggregate formation (Fig. 3D). The reduction of stimulated adhesion by each deubiquitinase inhibitor was statistically significant (Figs. 3E–G). Both PYR41 and PR619 suppressed platelet adhesion, and the formation and extension of pseudopods in response to thrombin stimulation (Supplementary video 2). The inhibitory effect of PYR41 and PR619 on deubiquitinase activity in platelet lysates was concentration dependent, and correlated to their level of thrombin-induced aggregation (Supplementary Fig. IA and IB). The inhibitory effect of PYR41 and PR619 was optimal at 0.2 U of thrombin, and was lost at higher concentrations (data not shown).

Figure 3. Deubiqutinase inhibition reduces platelet responses to agonist stimulation.

(A) PYR41 or PR619 pretreatment attenuated (A) ADP- (B) collagen- or (C) thrombin-induced aggregation. (D) Neither PYR41 nor PR619 stimulated platelet aggregation in the absence of agonists. PYR41 and PR619 imposed a reduction in ADP- (E) collagen- (F) and thrombin- (G) induced aggregation that was statistically significant (n=3; ***p≤ 0.001). (H) PYR41 (top) or PR619 (bottom) reduced thrombin stimulated adhesion and spreading. Interaction of control or thrombin- (0.2 U) stimulated platelets with a glass substrate was imaged by total internal reflection microscopy after 5 min (n=3). Videos of this interaction are presented as supplementary data. (I) Platelet area after PYR41 or PR619 treatment. Platelet surface area in panel (H) was quantified by Image-Pro plus software (n=3; ***p≤0.001).

Outside-in signaling induces cytoskeletal rearrangement, with extension of filopodia and lamellapodia, when platelets interact with a glass substrate. Thrombin stimulation enhances these responses, which was apparent in individual video frames captured by total internal reflection microscopy of fluorescently labeled platelets (Fig. 3H, Supplementary video II). Both PYR41 and PR619 abolish the adhesion and spreading of unstimulated platelets, while significantly (Fig. 3I) reducing these responses after thrombin stimulation.

Agonist activation of glycoprotein αIIbβ3 is reduced by deubiquitinase inhibition

Agonist stimulation alters the conformation of the platelet β3 integrin αIIbβ3 glycoprotein complex that enhances platelet-platelet and platelet-endothelial cell interactions.²² The antibody PAC-1 recognizes a neoepitope in activated αIIbβ3, and thrombin stimulation greatly increased PAC-1 binding to the surface of washed human plateles compared to quiescent control cells (Figs. 4A). Pre-treating platelets with PYR41 for 15 min suppressed PAC-1 binding to such an extent that αIIbβ3 activation was not statistically distinct from unstimulated platelets (Fig. 4B). PR619 also effectively reduced thrombin-stimulated expression of PAC-1 (Figs. 4E, F). Inhibition of deubiquitinase activity by either PYR41 (Figs. 4C, D) or PR619 (Figs. 4G, H) also fully suppressed thrombin stimulation of alpha granule secretion detected by the translocation of P-selectin from these granules to the platelet surface. Thrombin at the concentration used above activates platelets primarily through the PAR1 receptor, but activation of platelets through a second G protein-coupled receptor, the P2Y₁₂ receptor for ADP, was also abolished by deubiquitinase inhibition (Supplementary Fig. II). ADP stimulation increased binding of PAC-1 to platelets (Supplementary Fig. II A), although not to the extent induced by thrombin, and PYR41 effectively reduced the extent of αIIbβ3 activation after ADP activation to nearly quiescent levels (Supplementary Fig. IIB). PYR41 pretreatment also suppressed P-selectin exocytosis after ADP stimulation (Supplementary Fig. I C) and this reduction was statistically significant (Supplementary Fig. II D). PR619 was equally effective in suppressing platelet integrin activation (Supplementary Fig. II E, F).

Figure 4. Inhibitors of platelet deubiquitinases suppress protease activated receptor (PAR) mediated αIIbβ3 activation and degranulation.

(A) PYR41 pretreatment suppressed stimulated αIIb/β3 activation. Flow cytometry using PAC-1 antibody, that binds active αIIbβ3, on platelets treated or not with PYR41 stimulated or not with thrombin (0.2 U; n=3). (B) Mean fluorescence intensity (MFI) of PAC-1 binding to platelets in panel A (n=3; ^*p≤ 0.05). (C) PYR41 pretreatment reduced P-selectin surface expression upon activation. Flow cytometry using PE-conjugated-anti-P-selectin antibody to platelets treated or not with PYR41 with or without thrombin (0.2 U; n=3) stimulation. (D) Mean fluorescence intensity of anti-P-selectin binding to platelets in panel C (n=3; ^*p≤ 0.05). (E) PR619 pretreatment suppressed thrombin-stimulated αIIbβ3 activation. Flow cytometry using PAC-1 antibody binding to platelets treated or not with PR619 with or without 0.2 U thrombin (n=3). (F) Mean fluorescence intensity of PAC-1 binding on platelets in panel E (n=3; ^**p≤ 0.01). (G) PR619 pretreatment reduced P-selectin surface expression after thrombin activation. Flow cytometry using PE-conjugated-anti-P-selectin antibody binding to platelets treated or not with PR619 with or without thrombin (0.2 U; n=3). (H) Mean fluorescence intensity of anti-P-selectin binding to platelets in panel G (n=3; ^**p≤ 0.01).

Proteasome-associated deubiquitinases regulate platelet activation

Platelets contain several cytoplasmic deubiquitinases, but also abundantly express the proteasome associated deubiquitinases USP14 and UCHL5 at 3100 and 1500 copies per platelet, respectively.⁷ Specific inhibitors of USP14 and UCHL5 have been developed since the levels of these two enzymes are upregulated in many cancers²³ and are considered to be targets for potential anti-cancer therapies. A small molecule inhibitor, b-AP15, was recently identified¹⁷ that specifically inhibited just these two proteasome bound deubiquitinases without affecting the catalytic activities of soluble deubiquitinases. This inhibitor is potent and available to intracellular enzyme with b-AP15 inhibiting hydrolysis of fluorogenic ubiquitin substrate by the 19S subunit of the proteasome with an IC50 of just 6.5 μM ²⁴. We treated washed platelets with b-AP15 to find this inhibitor induced accumulation of high molecular weight ubiquitin-conjugated proteins (Fig. 5A). We confirmed platelets contained both USP14 and UCHL5 targets of b-AP15 by recovering platelet proteasomes by high speed centrifugation, and then western blotting proteins of the isolated proteasome (Fig. 5B). Next we determined whether USP14 and UCHL5 regulated platelet function to find b-AP15 inhibited thrombin- (Fig. 5C), collagen- (Fig. 5D) and ADP- (Fig. 5E) induced aggregation. The inhibitory effect of b-AP15 on agonist-induced platelet aggregation was significant for each of these agonists (Fig. 5F), and the data show that platelet aggregation was not affected by b-AP15 treatment alone. The small molecule inhibitor b-AP15 also inhibited thrombin-induced αIIbβ3 activation detected by PAC-1 binding (Fig. 5G), as well as P-selectin exocytosis from platelet alpha-granules (Fig. 5H). Pretreatment with b-AP15 also blocked ADP-induced αIIbβ3 activation (Supplementary Fig. IIIA). USP14 and UCHL5 inhibition had a more profound effect on collagen-induced aggregation (Fig. 5C) than the general deubiquitinase inhibitors, so we tested the effect of inhibition of these enzymes on platelet adhesion to collagen under static conditions. b-AP15 pretreatment reduced the number of platelets adhering to a exposed collagen-coated surface (Supplementary Fig. IIIB), suggesting a vital role of proteasome-associated deubiquitinases in mediating platelet interactions with exposed collagen.

Figure 5. Pharmacologic inhibition of the proteasome-associated deubiquitinases USP14 and UCHL5 reduce platelet responsiveness to agonists.

(A) b-AP15 pretreatment increased ubiquitination of the platelet proteome. Immunoblot of platelet lysates, treated or not with b-AP15, probed with anti-ubiquitin antibody (P4D1, n=3). (B) Platelet proteasomes contained UCH14 and UCLH5. Western blot of platelet proteasomes recovered by high speed (178,000 × g, 2h) centrifugation. (C) Proteasome-associated deubiquitinases promote stimulated platelet aggregation. b-AP15 pretreatment inhibited thrombin-(n=3) (D) collagen- (n=3) and (E) ADP-induced aggregation (n=3). (F) b-AP15 mediated inhibition of stimulated platelet aggregation is statistically significant (n=3; ***p≤ 0.001, **p≤0.01, *p≤ 0.05). (G) b-AP15 pretreatment suppressed stimulated αIIbβ3 activation. Flow cytometry using PAC-1 antibody binding to platelets treated or not with b-AP15 and stimulated or not with 0.2 U thrombin (n=3). (H) b-AP15 pretreatment reduced P-selectin surface expression upon activation. Flow cytometry using PE-conjugated-anti-P-selectin antibody binding to platelets treated or not with b-AP15 with or without 0.2 U thrombin (n=3).

Inhibition of platelet deubiquitinases modulates signaling downstream of thrombin and collagen receptors

Akt (protein kinase B) is a serine/threonine kinase that is an established downstream effector of phosphatidylinositol 3-kinase (PI3K).²⁵ Platelets express two Akt isoforms, Akt1 and Akt2,^{26, 27} where targeted deletion of Akt1 diminishes responses to both thrombin and collagen,^{28, 29} while Akt2 deletion suppresses secretion and thrombosis.³⁰ We determined whether deubiquitinase activity modulated Akt activation. Stimulation with thrombin resulted in time dependent increase in Akt phosphorylation at both serine₄₇₃ and threonine₃₀₈ residues, which was reduced by pretreatment with either PYR41, PR619, or b-AP15 (Fig. 6A). Pre-treating platelets with any of the three deubiquitinase inhibitors resulted in equivalent reduction of collagen-induced phosphorylation of Akt serine₄₇₃ and threonine₃₀₈ (Fig. 6B), although the inhibition of threonine₃₀₈ phosphorylation by either agonist was more profound than the reduction at serine₄₇₃. In platelets, Akt activation is both PI3K dependent and independent, but sustained Akt activation is PI3K dependent.³¹

Figure 6. Deubiquitinase inhibitors impair platelet signaling following agonist stimulation.

(A) Deubiquitinase inhibitors reduced thrombin induced Akt activation. Platelets treated with or without deubiquitinase inhibitors were stimulated with 0.2 U thrombin for the stated times and immunoblotted using anti-pAKT ser₄₇₃, anti-pAKT thr₃₀₈, or anti-Akt antibodies (n=3). (B) Deubiquitinase inhibitors reduced collagen-induced Akt activation. Platelets treated with or without deubiquitinase inhibitors were stimulated with 2 μg collagen for the stated times and immunoblotting with anti-pAKT ser₄₇₃, anti-pAKT thr₃₀₈, or anti-Akt antibodies (n=3). (C) Deubiquitinase inhibitors blocked collagen-induced intracellular calcium release. Platelets labeled with Fura-2 AM were treated or not with deubiquitinase inhibitors before stimulation with 2 μg collagen with monitoring of the ratio of fluorescent emmision at 340 and 380 nmeters. (D) Deubiquitinase inhibition suppressed clot retraction. Ratio of the image surface area of images of thrombin-induced clots over time (n=3; ^**p< 0.01, *p<0.05). (E) Deubiquitinase inhibition reduces PTEN phosphorylation. Immunoblotting of platelets with anti-phospho PTEN-ser₃₈₀ antibody treated or not with deubiquitinase inhibitors before stimulation with 0.2 U thrombin or 2 μg collagen (n=3).

PI3K isoform β (PI3K β) is essential to promote PLCγ-2 activation downstream of GPVI-FcRγ receptor³² that leads to intracellular calcium release critical for sustained αIIbβ3 activation necessary for ex vivo clot retraction^{33, 34} and consolidation in vivo.^{32, 35} Collagen-induced intracellular calcium release was inhibited when platelets were preincubated with deubiquitinase inhibitors before stimulation by collagen (Fig. 6C). Deubiquitinase inhibitors also significantly reduced fibrin clot retraction (Fig. 6D), suggesting impaired outside in signaling from integrin αIIbβ3.

PTEN on chromosome 10 is a negative regulator of PI3K/Akt signaling³⁶ that reduces collagen-dependent platelet activation.³⁷ PTEN phosphatase activity is negatively regulated in nucleated cells by phosphorylation at serine₃₈₀,^{38, 39} although this has not been extended yet to platelets. We observed both thrombin and collagen increased PTEN phosphorylation at this residue (Fig. 7E), and that pretreatment with deubiquitinase inhibitors reduced agonist dependent increase in PTEN phosphorylation and inactivation.

Discussion

We found that protein ubiquitination in platelets is a dynamic process where acute inhibition of deubiquitinase activity caused a significant increase in ubiquitin modification of the proteome of quiescent cells within just a few minutes. This change occurred both with general deubiquitinase inhibitors, but also in response to b-AP15 inhibition of UCHL5 and USP14 of the platelet proteasome. Since these changes were apparent in the absence of agonist stimulation and intracellular signaling, we conclude the proteome of quiescent platelets undergoes a continuous, and rapid, cycle of ubiquitination and deubiquitination. Notably, deubiquitinase inhibition was without detectable effect on the total content of platelet proteins (Supplemental Fig. IV), nor did complete loss of platelet proteasome proteolytic activity alter the complement of platelet proteins (beyond release of filamin A to the soluble fraction).⁵

We find protein deubiquitination promotes platelet reactivity. The general deubiquitinase inhibitors PYR41 or PR619 blocked inside-out activation of αIIbβ3 integrin, homotypic aggregation, and inhibited adhesion to collagen under flow at high shear. PYR41 additionally blocked formation of occlusive platelet-rich thrombi in damaged carotid arteries. Inhibition of deubiquitinase activity suppressed extension of membrane protrusions and cell spreading, and deubiquitinase inhibition blocked alpha granule release and hence blocked the agonist-stimulated increase of P-selectin on the platelet surface. Continual remodeling and turnover of ubiquitin modification of platelet proteins is thus a positive element in the transition from quiescent cells to activated platelets that is necessary for thrombosis. We conclude the widespread suppression of stimulated platelet function by deubiquitinase inhibition means that ubiquitin modification of the proteome must maintain platelets in an inactive state, and that either restructuring of the existing polyubiquitin decoration of the platelet proteome or that recycled ubiquitin is produced to allow its addition to new targets—western blotting shows intracellular free ubiquitin is limiting—releases tonic ubiquitin inhibition of platelet signaling and activation.

Deubiquitinase inhibitors affected signaling downstream of G protein coupled receptors as well as the GPVI receptor for collagen. Thrombin- and collagen-induced Akt phosphorylation was reduced by both the pan deubiquitinase inhibitors PYR41 and PR619, but also by the specific inhibitor of proteasome associated deubiquitinase activity b-AP15. Activation of PI3K and its effector, Akt, is essential in most aspects of collagen-induced platelet activation, and thrombin- and collagen-induced αIIbβ3 integrin outside-in signaling. The effect of PI3K/Akt axis on thrombin-induced aggregation is only seen at lower thrombin concentrations and is overcome at high agonist concentrations.^{33, 40} Akt activation in platelets is both PI3K-dependent and PI3K-independent,⁴¹ however PI3K isoform β (PI3K β) is mandatory for GPVI-FcRγ receptor-mediated Akt activation, platelet aggregation, intracellular calcium release, and degranulation.⁴² We find ubiquitin remodeling is an integral component of this cascade.

We attribute at least part of the reduction in phosphorylated Akt after pretreatment with deubiquitinase inhibitors to the upregulation of PTEN phosphatase activity. PTEN negatively regulates Akt activation, which negatively regulates collagen-induced platelet activation,³⁷ by converting phosphatidylinositol 3,4,5-trisphosphate (PIP₃) to phosphatidylinositol 4,5-bisphosphate (PIP₂). Phosphoinositide-dependent protein kinase 1 (PDK1) is activated by an increase in PIP₃ following PI3K activation, and activated PDK1 phosphorylates Akt at residue threonine₃₀₈.⁴³ Platelet specific knockdown of PDK1 and subsequent loss in Akt phosphorylation reduces thrombin-induced platelet aggregation, clot retraction and thrombosis.⁴⁰ This central phosphorylation cascade, we find, includes ubiquitin chain remodeling.

Significance.

Platelet proteins are modified by ubiquitination, which is enhanced in stimulated platelets. Platelets express functional deubiquitinase activity, whose inhibition rapidly increased ubiquitination of the platelet proteome and blocked agonist-stimulated adhesion, spreading, and stimulated aggregation. Intravascular injection of a deubiquitinase inhibitor slowed formation of occlusive platelet thrombi in damaged carotid arteries. General deubiquitinase inhibitors, or selective inhibition of proteasome-associated deubiquitinases, reduced signaling through the kinase cascade and blocked the increase in intracellular Ca⁺⁺ necessary for platelet activation. Deubiquitination is therefore necessary for platelet activation, and is a drugable target that would reduce thrombosis.

Abbreviations

Akt

v-akt murine thymoma viral oncogene homolog 1, protein kinase B

K₄₈

lysine 48

K₆₃

lysine 63

PE

phycoerythrin

PTEN

Phosphatase and tensin homologue-deleted

PI3K

phosphatidylinositol 3-kinase

Thr

thrombin

Ub

ubiquitin