Inhibition of RHOA activity preserves the survival and hemostasis function of long-term cold-stored platelets Free

Stem cell transplantation recipients and oncology patients require continued use of prophylaxis with platelet transfusion to reduce bleeding.¹ Cold-stored platelets are cleared from host circulation after a few hours.² Therefore, conventional cold-stored platelets are deemed unusable for bleeding prophylaxis in amegakaryocytic thrombocytopenia, for which donor platelets are to be stored at room temperature (RT) for no longer than 7 days to prevent bacterial contamination.³ Unfortunately, the short shelf-life of RT-stored platelets limits the availability of platelets for transfusion⁴ and does not abrogate the risk of sepsis.⁵ Chemical-induced pathogen reduction is approved but does not eliminate the risk of bacterial contamination–associated sepsis either⁶ and hampers platelet product potency, resulting in increased platelet transfusion needs to achieve equivalent platelet count increments.^7,8 Cold storage would significantly reduce the proliferation of micro-organisms,⁹ allowing a longer period of storage¹⁰ and ameliorating the current seasonal shortages for platelet component inventories at the local level.^11,12 

Refrigeration induces the so-called “cold storage platelet lesion.” The cold storage platelet lesion is believed to be caused by activation of multiple G-protein–coupled temperature-gated transient receptor potential (TRP) channels that act as nonselective cation channels gated by cold stimuli,^13,14 cold-induced arachidonic acid–dependent 14-3-3ζ-GPIbα association,^15,16 formation of lipid raft microdomains^17,18 and activation of p38 MAPK.¹⁹ The variety of cold-sensing mechanisms in mammalian platelets^14,20 also complicates any direct targeting strategy. Cold storage results in multiple irreversible cytoskeletal changes on the platelet membrane, including cytoskeletal rearrangements and shape changes, including filopodia or lamellipodia, accompanied by an increase in the fraction of polymerized actin (F-actin).^21-24 Ultimately, platelet clearance depends on platelet cytoskeleton-connected, microscale cholesterol-rich lipid raft domains,^17,18 proapoptotic activation signals^16,25 and GPIbα glycosylation lesions,^26-29 recognizable by host macrophage αMβ2 integrin,^27,30 host hepatic lectin-binding receptors,^28,30 and von Willebrand factor (VWF) binding–dependent GPIbα signaling.^31-33 

We and others have shown that RAS homolog family, member A (RHOA) and RAS-related C3 botulinum toxin substrate 1 (RAC1) are essential for various aspects of signaling from platelet receptors to the intracellular pathways that drive platelet function,^34-36 and for inside-out signaling to modulate platelet receptors.^34-41 Furthermore, our group, by using rational and structure–activity relationship approaches, has previously identified RHOA^42,43 and RAC1^34,37,44,45 activity inhibitors with relatively low affinity for the small GTPase substrates that can reversibly inhibit platelet activity. In this study, we have now identified that the activation of RHOA is necessary for the signals responsible for the cold storage–induced platelet lesion, and that the genetic loss of RHOA makes platelets insensitive to refrigeration-induced clearance. Inhibition of RHOA ex vivo during cold storage preserves platelet survival and clotting activity of transfused platelets at similar levels as RT-stored platelets.

Details on animal and human platelet components can be found in the supplemental Methods, available on the Blood website.

Platelet transfusion, murine platelet survival, and tail bleeding time after aspirin challenge are described in the supplemental Methods.

Western blot analyses, immunofluorescence microscopy, and flow cytometry analyses are described in the supplemental Methods. Experimental details on cell division cycle 42 (CDC42), RAC, and RHOA activation analysis methods can be found in the supplemental Methods.

Analysis of S-G04 plasma levels was performed by mass spectrometry/high-performance liquid chromatography. The details can be found in the supplemental Methods and supplemental Table 1.

The statistical analyses are described in the supplemental Methods.

Similarly to previous reports,^46-48 we found that cold storage of pediatric-size, acid-citrate-dextrose anticoagulated, 100% plasma-containing human platelet products stored in gas-permeable, polyolefin bags under agitation⁴⁹ induces platelet activation as assessed by the expression of activated GPIIb/IIIa, membrane P-selectin, and membrane phosphatidylserine residue exposure (PS⁺) on days 7 and 14 of storage (supplemental Figure 1A-D). Cold-stored platelets lose membrane GPIbα and GPVI (supplemental Figure 1E-F), with concomitant increase in shedding of overall and PS⁺ microparticles^50,51 (supplemental Figure 1G-H) and increased hepatocyte-mediated platelet phagocytosis (supplemental Figure 1I) compared with RT-stored platelets. This data set is consistent with the reported loss of hemostatic activity of long-term stored platelets in aspirin-treated individuals^50,52 and their reduced survival in vivo.² 

The response of platelets to cold storage is complex, and TRP receptors seem to play a role in early temperature-induced activation of platelets.¹⁴ To determine whether Rho family members, known effectors of TRP receptors,^53,54 are activated during cold storage, we first analyzed the level of activation of the Rho family members RHOA, RAC, and CDC42. We found that murine RHOA and RAC1, but not CDC42, are activated by cold storage (Figure 1A). Murine platelet RAC is first activated in as early as 15 minutes of cold storage, followed by RHOA after 30 minutes. Rationally designed, cell-permeable pharmacological inhibitors of RHOA and RAC that prevent their activation in platelets and are deprived of platelet toxicity have been developed by our group.^34,42,55 The RHOA inhibitor R-G04⁴² and the RAC inhibitor NSC23766 (NSC)³⁷ are specific and effective in platelets at concentrations of 75 and 50 μM, respectively. Similarly, the Cdc42 activity-specific inhibitor (CASIN) is specific and effective in platelets at a concentration of 10 μM⁵⁵ (supplemental Figure 2A). We analyzed whether the combination of these 3 inhibitors (R-G04, NSC, and CASIN) may result in a beneficial effect by preventing platelet clearance of murine cold-stored platelets. Our data indicated that the inhibition of these 3 Rho subfamily GTPases resulted in platelet survival of unconditioned recipients similar to that of RT-stored platelets (Figure 1B-D).

Whole blood–derived, 5 donor–pooled platelet components were generated and stored at different temperatures, in agitation.⁵⁶ These components were manufactured following blood banking standardized operating protocols containing the platelet additive solution PAS-3M (67%) and plasma (33%),⁵⁶ adapted to a pool-split study design. Like short-term stored murine platelets, 7-day cold-stored human platelets maintained high activity of RHOA and RAC1, but not CDC42 (Figure 1E), which could be inhibited by R-G04 and NSC, respectively. The combination of R-G04 and NSC did not further inhibit RHOA or RAC1 compared with the use of single specific inhibitors (Figure 1F).

To determine the effect of these inhibitors on human platelets, we used our previously validated model of immunodeficient mouse transfusion.⁵⁶ Therapeutic, platelet-rich plasma–derived, pooled human platelets were stored refrigerated (1-6°C) for 7 days in presence and/or absence of specific inhibitors (supplemental Figure 2B) and transfused in irradiated, thrombocytopenic NSG (nonobese diabetic, severe combined deficient, γ-common chain deficient) mice (supplemental Figure 2C). Although human cold-stored platelets in the presence of a vehicle or the CDC42 inhibitor CASIN were rapidly cleared (area under the curve reduced ∼57% over RT-stored platelets), platelets stored in the presence of the RHOA inhibitor R-G04 or the combination of R-G04 with the RAC inhibitor NSC retained survival in vivo at levels identical to control RT-stored human platelets. Cold storage in the presence of the RAC inhibitor NSC only partially prevented cold-induced platelet clearance (Figure 1G-H). To confirm these data in a model of platelet clearance dependent on activated macrophages, we performed in vitro analysis of macrophage-induced phagocytosis of stored human platelets with or without inhibitors in which platelets were washed at the end of the day 7 of storage and incubated with activated, differentiated Tamm-Horsfall protein 1 (THP-1) macrophages. Like the in vivo transfusion model, cold-stored platelet phagocytosis was prevented by storage in the RHOA inhibitor R-G04 but not the CDC42 inhibitor CASIN. Storage in the presence of the RAC inhibitor NSC only partly prevented cold-stored platelet phagocytosis (Figure 1I).

To ensure that the R-G04 effect on preventing clearance of cold-stored platelets is attributed to RHOA inhibition, we generated RHOA-deficient murine platelets and analyzed whether the effect of R-G04 is on target. Platelet RHOA deficiency was induced by administration of 2 (10 mg/kg every other day) intraperitoneal doses of polyinosinic:polycytidylic acid (polyI:C) to Mx1-Cre;RhoA^flox/flox mice (and controls).⁵⁷ As previously reported,⁵⁸ by day 3 after the last dose of polyI:C, the platelet count of Mx1-Cre;RhoA^flox/flox mice was reduced by ∼50% (supplemental Figure 3A), and platelet RHOA expression was reduced by ∼90% (Figure 2A). Murine RHOA–deficient platelets did not show significant differences in binding of relevant lectins binding sialic acid, N-acetylglucosamine, or galactose-N-acetylglucosamine⁵⁹ (supplemental Figure 3B-C), or in the expression of membrane β3 integrin (CD61) (supplemental Figure 3D-E), even when cold-stored (supplemental Figure 3D-E). Platelets were labeled with carboxyfluorescein succinimidyl ester.³⁰ This labeling was found not to interfere with the binding of lectins relevant to cold storage–associated glycation structures (supplemental Figure 3B-E). Platelets were adjusted to a dose of 3 × 10⁸ (in 200 μL of phosphate-buffered saline) and transfused in irradiated, thrombocytopenic, congenic C57BL/6 mice (supplemental Figure 4A). Survival kinetics analysis demonstrated that murine RhoA^Δ/Δ platelets were completely insensitive to cold storage, having an in vivo platelet life span similar to that of RT-stored platelets (Figure 2B-C). The cold insensitivity of RHOA-deficient platelets was confirmed in our in vitro macrophage-mediated phagocytosis assay, and R-G04, NSC, or their combination did not further reduce the level of cold-stored platelet phagocytosis (Figure 2D). Collectively, these results demonstrate that RHOA is a genuine target for intervention and its expression is at the root of cold storage–induced damage, and that R-G04 activity is RHOA dependent.

We have previously demonstrated that a simple platelet wash process of R-G04-treated platelets results in loss of its in vitro antiaggregant effect, demonstrating the reversibility of the effect of R-G04 on platelet function.⁴² Parenteral administration of R-G04 to mice for up to 14 days at daily doses of up to 30 mg/kg does not result in any apparent toxicity or bleeding diathesis in rodents or dogs (data not shown),^60,61 suggesting that R-G04 may have a limited, short-term effect when infused in vivo. We hypothesized that in vivo dilution of R-G04–containing platelets should result in a rapid off-rate, leading to the release of R-G04 from its intracellular target and reversibility of its effect in vivo. Therefore, we analyzed whether murine R-G04 cold-stored platelets were hemostatically effective to correct the prolonged bleeding time of aspirin-treated mice. The transfusion of 3 × 10⁸ R-G04 cold-stored congenic platelets (in a volume of 0.2 mL, corresponding to ∼1/10 of the mouse volume) at 24 hours after administration of aspirin was able to restore bleeding times to the same levels as those of RT-stored platelets (Figure 3A). These data indicate that RHOA activity inhibitor R-G04, mimicking the RhoA^Δ/Δ effect, can maintain the survival of functional platelets that are able to substantially correct the bleeding time of aspirin-treated mice. Furthermore, these data expand upon our previous reports on low, rapidly reversible binding of R-G04 to its target.^42,43 

The consequences of cold platelet storage include impaired α2β3 integrin signaling, vesicle/microparticle shedding, dilation of the open canalicular system, and activation of actomyosin-based cytoskeletal rearrangements.^23,24,62-64 We have previously demonstrated that R-G04 can prevent integrin-dependent outside-in signaling.⁴² We analyzed whether RHOA inhibition prevents cold storage–dependent outside-in signaling and major cytological consequences of cold storage–induced damage. We first analyzed the spreading of platelets on immobilized fibrinogen to determine the integrity of the αIIbβ3 integrin signaling–mediated cytoskeletal response of 7-day cold-stored, washed human platelets. We found that 7-day cold-stored human platelets spread over immobilized fibrinogen contained microaggregates and platelet fragments (<1-μm diameter). Among nonaggregated, nonmicroparticle, polymerized actin–expressing platelets (Figure 3B, vehicle-treated, cold platelets, outlined area), the platelet spread area was significantly diminished (∼70%), with underdeveloped stress fibers and loss of cortical actin polymerization (Figure 3B-C). Conversely, R-G04 cold-stored platelets preserved their spreading area and morphology at levels similar to those of their RT-stored counterparts with or without R-G04 (Figure 3B-C). Transmission electron microscopy analyses demonstrated that the microparticle formation and vesicle shedding of cold-stored platelets were significantly prevented by storage in the presence of R-G04 (Figure 3D-E). Similarly, we observed that the dilation of the open canalicular system was mostly prevented in R-G04 cold-stored platelets (Figure 3F), and the activation of myosin light chain through Ser19 phosphorylation was completely prevented by RHOA pharmacological targeting, whereas targeting of RAC only resulted in partial inhibition (Figure 3G). These data support the hypothesis that R-G04 treatment suffices for preventing major cytological and structural effects of cold storage–induced damage and outside-in signaling affecting actomyosin activation.

We hypothesized that pharmacological targeting of RHOA and/or RAC1 may prevent inside-out signaling in platelets and thus avert undesirable platelet structural conformation changes that result in host clearance. GPIbα is the major subunit of the receptor for VWF, which localizes platelets at sites of vessel damage in collaboration with collagen receptors GPVI and integrin α2β1. Cold storage exposes underlying galactose and β-N-acetyl-d-glucosamine residues and disrupts anchorage to the membrane-cytoskeleton, thereby inducing VWF binding–dependent GPIbα clusters. GPIbα-containing clusters make platelets susceptible to recognition by host lectin-binding receptors^28,30 and induce platelet apoptosis, as 14-3-3ζ dissociates from Bad and associates with clustered GPIbα, leading to the proapoptotic cascade of Bad activation, cytochrome c release, caspase 9 activation, and phosphatidylserine exposure that results in increased macrophage-dependent phagocytosis.³³ We first wanted to determine whether cold storage in R-G04 prevents GPIbα clustering on the membrane. Consistent with the reported effect of refrigerated storage on platelet membrane GPIbα,^28,30 we found that cold storage induced the formation of membrane areas of higher GPIbα expression (Figure 4A, arrows). This abnormal distribution of GPIbα was prevented by R-G04 and, at a lesser degree, by NSC23766. On the basis of this observation and our current understanding of the distribution of glycoproteins on the platelet membrane surface in microdomains of lipid rafts, we hypothesized that the process of lipid raft formation in specialized membrane microdomains was at the root of the effect of RHOA targeting. To determine whether RHOA activation affects the composition of lipid rafts, we analyzed the expression of flotillin-1 present in planar lipid rafts and caveolin-1 contained in caveolar lipid rafts (reviewed by Doherty and McMahon⁶⁵ and Hansen and Nichols⁶⁶). From platelets digested with the nonionic detergent Triton X-100 and separated by sucrose gradient centrifugation, we found that the membrane of refrigerated murine (Figure 4B; supplemental Figure 4B) and human platelets stored for 7 days (Figure 4C-D) was enriched in lipid rafts containing both flotilin-1 and caveolin-1. R-G04 phenocopied the effect of the cholesterol-depleting agent methyl-cyclodextrin to prevent the cold storage–induced formation of lipid rafts (supplemental Figure 4C) and phagocytosis of cold-stored platelets (supplemental Figure 4D). Comparatively, R-G04 alone, but not NSC, completely prevented the effect of cold storage on the generation of planar and caveolar lipid rafts in murine (Figure 4B) and human platelets stored for 3 or 7 days (Figure 4C-D). Conversely, fasudil, a potent, specific inhibitor of ROCK, one of multiple downstream effectors of RHOA activity (reviewed by Feng et al⁶⁷), did not phenocopy the effect of R-G04 and did not prevent the phagocytosis of cold-stored platelets (supplemental Figure 4E). Finally, R-G04 was able to prevent the recruitment of 14-3-3ζ to GPIbα to lipid raft microdomains induced by cold storage (supplemental Figure 5A-B). Taken together, these data indicate that RHOA inhibition in cold-stored platelets prevents abnormal distribution of membrane GPIbα and lipid raft formation and recruitment of 14-3-3ζ to GPIbα to lipid raft microdomains, which lie at the root of the processes of host recognition, apoptosis, and platelet clearance.

Desialylated GPIbα²⁸ and irreversibly clustered β-N-acetylglucosamine (β-GlcNAc)–terminated glycans on GPIbα²⁷ are well recognized glycation defects induced by cold storage of platelets. Glycosyltransferases and glycosidases are responsible for the assembly, processing, and turnover of glycans, and work together to modify glycan structures (reviewed by Rini et al⁶⁸). A significant fraction of GPIbα is located intracellularly and can translocate to the membrane upon platelet activation.⁶⁹ We hypothesized that the hypogalactosylation and hyposialylation of GPIbα observed in cold-stored platelets may be associated with changes in the platelet spatial distribution of key glycosylation enzyme activities. Although the role of neuraminidase and sialidase activities in cold-stored platelets has been reported,^29,70-75 the role of glycosyltransferases required for modifications of glycans susceptible to cleavage has not been studied in refrigerated platelets. Analysis of the cellular distribution of the human uridine diphosphate (UDP)-N-acetyl-α-D-galactosamine:polypeptide N-acetylgalactosaminyltransferases (GalNAc) T1 and T3,⁷⁶ 2 enzyme activities crucial for generating sugar moieties susceptible to neuraminidase-mediated cleavage, demonstrated that cold storage of human platelets resulted in the membrane clustering of GalNAc-T1/3 (Figure 5A-B). Clustering of GalNAc-T1/3 is preventable by the presence of R-G04, or partly by NSC, during cold storage (Figure 5A-B). A washing step did not result in any significant effect over the effects identified in unwashed platelets. Analyses of the activities of UDP-galactosyltransferase and sialyltransferase, responsible for the generation of moieties susceptible to sialidase activity, in membrane lipid raft extracts of cold-stored platelets further confirmed that these key glycosyltransferases translocated to platelet membrane microdomains (Figure 5C-D; supplemental Figure 6A-B). R-G04 and partly NSC prevented the mislocalization of these glycosyltransferase activities (Figure 5C-D; supplemental Figure 6A-B), further highlighting the role of RHOA/RAC in cold storage platelet lesions.

Our group has recently identified S-G04, a second-generation RHOA inhibitor and enantiomer of R-G04 (supplemental Figure 7A), as a reversible inhibitor with increased potency for inhibiting RHOA activation and aggregation and thromboxane production by multiple agonists.⁴² Like R-G04,⁴³ the predicted structural contacts of S-G04 include the binding pocket around the Trp58 residue of RHOA.⁴² However, S-G04 is more potent in the prevention of macrophage-dependent phagocytosis of cold-stored murine platelets (supplemental Figure 7B).

S-G04, similar to R-G04, is not toxic to mice when injected at concentrations of up to 10 mg/kg per day for up to 14 days (data not shown). Pharmacokinetic analysis of serum levels of S-G04 after administration of up to 10 nmol indicates that this small molecule is cleared from circulation within a few minutes after intravenous administration (supplemental Figure 7C).

Long-term cold storage of platelets in the presence of S-G04 at a concentration as low as 10 μM phenocopies the effect of 75 μM R-G04 and suffices to prevent platelet activation as measured by membrane exposure of P-selectin, phosphatidylserine residue exposure, and activated GPIIb/IIIa (supplemental Figure 7D-F). It also suffices to prevent the loss of integrity of cold-stored platelets (supplemental Figure 8A-B), the formation of microparticles exposing phosphatidylserine-enriched, negatively charged phospholipids (supplemental Figure 9A-C), maintains platelet pH (supplemental Figure 10A), and prevents cold storage–induced mitochondrial proton-leak and loss of mitochondrial bioenergetic health (biohealth)^56,77 (supplemental Figure 10B-C). The effect of S-G04 in preventing platelet loss, GPIIb/IIIa activation, P-selectin exposure, level of membrane expression of GPVI and GPIbα, mitochondrial dysfunction, or reactive oxygen species production was independent of whether cold-stored platelets were stored under agitation (supplemental Figures 11, 12A-C, and 13A-C), confirming that the agitation factor did not modify the surface receptor expression of long-term cold-stored platelets⁴⁷ and only marginally reduced their level of mitochondrial dysfunction and activation (supplemental Figure 13A-C).

We found that S-G04 prevents the phagocytosis of 7-day cold-stored human platelets by differentiated THP-1 macrophages (Figure 5E; supplemental Figure 14A-C) or by hepatocytes HepG2 (supplemental Figure 14D), similar to its effect on murine RHOA–deficient platelets (supplemental Figure 14E-G). Similar to the effect in preservation of the hemostatic function of cold-stored murine platelets, long-term cold storage of human platelets in the presence of 10 μM S-G04 compares well with 75 μM R-G04 in its ability to preserve transfused platelet hemostatic activity in vivo after 7 or 10 days of storage (Figure 5F-G) in a murine model of bleeding time correction by platelet xenotransfusion.⁵⁶ Finally, we also found that S-G04 (10 μM) or R-G04 (75 μM) treated cold-stored platelets (supplemental Figure 15A-C) were able to prevent cold storage–induced platelet clearance after xenotransfusion (Figure 6A) of platelets stored for 7 (supplemental Figure 15D-F), 10 (Figure 6B-C), and 14 days (Figure 6D-E), correlating with the bleeding time correction activity observed in our hemostatic function model (Figure 5F-G). Altogether, this data set indicates that the addition of either R-G04 or the more potent S-G04 to the cold-stored human platelets can maintain the survival and circulation of functional, effective human platelets with preserved hemostatic activity well within the second week of storage.

Two distinct pathways recognize modified platelet receptor GPIbα to remove refrigerated platelets in host livers. First, the hepatic asialoglycoprotein Ashwell-Morell receptor (AMR) recognizes desialylated GPIbα.²⁸ Second, α_Mβ₂ integrins on hepatic resident macrophages selectively recognize β-GlcNAc–terminated glycans on GPIbα.²⁷ To determine whether the effect of R-G04 and S-G04 on the prevention of platelet clearance of cold-stored platelets was dependent on the recognition of cold-stored platelets by cognate host liver cellular receptors, we performed experiments of platelet survival in which we induced the functional disruption of both hepatocyte- and macrophage-mediated clearance. We first performed xenotransfusion of 10-day cold-stored human platelets into NSG mice conditioned with sublethal irradiation, macrophage depletion, and/or the administration of the AMR inhibitor asialofetuin, and analyzed the platelet survival kinetics in vivo after transfusion. We found that like the depletion of macrophages (Figure 6D-E), the in vivo inhibition of the AMR alone (Figure 7A-C) did not prevent the clearance of platelets, whereas the cold storage of platelets in presence of R-G04 or S-G04 did (Figure 7A-C). A washing step to remove S-G04 did not modify the effect of S-G04 in preventing platelet clearance (Figure 7A-C). Only the combined depletion of macrophages and in vivo inhibition of AMR modified the survival of long-term cold-stored human platelets stored in the absence of R-G04 or S-G04 (Figure 7D-F), which survived at levels not significantly different from those of platelets stored in R-G04 or S-G04 with or without a washing step (Figure 7D-E compared with Figure 7B-C). These data supported that the effect of R-G04/S-G04 on cold-stored platelets can be significantly phenocopied by in vivo targeting of both macrophage and hepatocyte mechanisms of platelet clearance. Furthermore, the data indicate that platelet RHOA inhibitors target biochemical signals, resulting in the combined clearance of platelets by host hepatocytes and macrophages.

Our study identifies the key role of RHOA activation in short-term and long-term cold storage lesions of platelets. Cold storage induces activation of RAC1 and RHOA. However, although RAC1 activation is only partly responsible for cold storage–induced platelet clearance, RHOA activation alone explains the major hallmarks of cold storage lesions. RHOA-deficient platelets are cold-insensitive and resist cold-induced clearance in vitro and in vivo. The cell-permeant, reversible RHOA inhibitor R-G04, or its more potent enantiomer S-G04,⁴² prevent the cold storage–induced platelet damage and allow the transfusion of murine or human platelets able to survive at levels similar to those of RT-stored platelets and to shorten the bleeding time of aspirin-treated, thrombocytopenic mice after tail amputation. Our data demonstrate that platelet RHOA inhibition prevents the shedding of phosphatidylserine-exposed microparticles, lipid raft formation, and recruitment of 14-3-3ζ to GPIbα, and upon in vitro washing or in vivo transfusion-associated dilution, it can maintain the function of platelets. S-G04 prevents cold storage–induced damage regardless of whether storage involves agitation.^47,56 R-G04 and S-G04 would bind to their target RHOA in vitro and prevent cold-induced signaling resulting in irreversible biochemical, metabolic, and cytoskeletal changes in platelets. Once platelets are infused, S-G04 seems to be released from its target and rapidly cleared from serum. The rapid clearance of S-G04 is associated with functional recovery in aspirin-induced thrombocytopathy of irradiated, thrombocytopenic mice.

Platelet clearance depends on the downstream acquisition of defects in GPIbα glycosylation, recognizable by host receptors for phagocytosis and clearance.^26,28 However, UDP-galactosylation cannot prevent long-term cold-storage damage of human platelets. Cold-induced platelet apoptosis can also result from VWF binding–dependent GPIbα signaling. Although the depletion of arachidonic acid during refrigeration prevents the proapoptotic 14-3-3ζ–GPIbα association, this approach only partly improved the posttransfusion survival of refrigerated platelets.^15,16 Peptide interference of GPIbα interaction with VWF improves posttransfusion survival, but its use is impractical in long-term storage, and only partially prevents storage damage.^31,78,79 Only the inhibition of RHOA seems to be able to completely prevent the cold storage–induced acquisition of conformational changes in platelet glycoproteins and lipid rafts, preventing macrophage-mediated phagocytosis and in vivo clearance. R-G04 and S-G04 completely prevented the effect of cold storage on platelet survival at levels similar to the combined pharmacological depletion of macrophages and AMR inhibition in vivo, suggesting that preventing cold storage–induced lesions may represent a successful ex vivo strategy to prevent cold-induced platelet clearance. Interestingly, inhibition of ROCK, a major downstream effector of RHOA, does not phenocopy the effect of RHOA inhibition. This suggests that, as expected given the multiple downstream effects of cold-induced damage, multiple effectors downstream of RHOA relevant in platelet signaling, including but not limited to ROCK, may contribute to RHOA-mediated cold platelet lesions.⁸⁰ 

Cold-dependent platelet activation results in the formation of micro- and macroaggregates during storage of clinical platelet units, resulting in up to 18% of wasted components.⁸¹ Slichter et al⁸² demonstrated that most bleeding time measurements in patients with thrombocytopenia transfused with conventional cold-stored platelets remained prolonged, whereas most bleeding times in patients receiving RT-stored platelets improved. Filip and Aster⁵² found that after 72 hours of storage, conventionally refrigerated platelets were less effective in elevating platelet counts and in shortening bleeding time than platelets stored at RT. More recently, Kogler et al⁵⁰ found that cold-stored platelets have a progressive decline in their survival upon autologous transfusion in humans within the first 10 days of storage, with even further decline in survival in the first 24 hours when they are cold-stored for up to 20 days. Consequently, platelets are allowed to be stored at 1 to 6°C with or without agitation for up to 14 to 21 days only in the treatment of active bleeding when conventional platelets are not available or their use is not practical. This effectively means that conventional cold-stored platelets will benefit only a very small number of patients and will not be available for bleeding prophylaxis, which represents up to 65% of the platelet products transfused in tertiary-care hospitals.^83,84 

Limitations of our study include the restriction of the analysis to pooled platelet-rich plasma–derived platelet products and the circumscription of the in vivo analysis to models of transfusion of murine congenic or human platelets in immunodeficient mice. Our data are to be confirmed in posttransfusion radiolabeled single donor–derived platelet recovery and survival analyses. Similarly, our analysis of bleeding time correction is limited to the reversal of prolonged bleeding time in aspirin-treated mice and has not been tested in models of trauma or surgical bleeding.

In summary, our data identify RHOA as the key molecular switch responsible for cold storage–induced damage leading to platelet clearance of refrigerated platelets. Inhibition of RHOA by the enantiomers R-G04 or S-G04 represents a promising approach to preventing cold storage–induced platelet lesions, their clearance upon transfusion, bacterial contamination, and extending their shelf-life well into the second week of storage.

The authors thank the Research Division and the Components Laboratory of Hoxworth Blood Center, University of Cincinnati, and Nikolai Timchenko from the Cincinnati Children’s Hospital Medical Center for providing HepG2 cells.

The authors thank the Translational Core Laboratory, mouse core and flow cytometry core of Cincinnati Children’s Hospital Medical Center, supported by the National Institute of Diabetes and Digestive and Kidney Diseases Center of Excellence in Hematology (grant U54 DK126108). This study was also funded by grants from the National Heart, Lung, and Blood Institute (P01HL158688 and R01HL147536).

Contribution: S.H., H.A., A.M.W., J.F.J., and X.Z. performed experiments; S.N. coordinated and generated platelet pools for testing; X.Z. and K.D.S. developed the mass spectrometry/high-performance liquid chromatography protocol for plasma determinations of inhibitor levels; S.H., Y.Z., and J.A.C. interpreted data and wrote the manuscript; and all authors reviewed the manuscript.

Conflict-of-interest disclosure: Y.Z. and J.A.C. own intellectual property described in this manuscript. The remaining authors declare no competing financial interests.

Correspondence: Jose A. Cancelas, Connell and O’Reilly Families Cell Manipulation Core Facility, Dana-Farber Cancer Institute, 450 Brookline Ave, Boston MA, 02215; email: jose_cancelas-perez@dfci.harvard.edu.