Platelet Clearance Research Articles

Adenovirus Induced Thrombocytopenia: The Role of P-Selectin and von Willebrand Factor in Virus- Mediated Platelet Clearance.

Acute thrombocytopenia has been consistently reported following IV administration of adenoviral vectors (Ad) but the mechanism responsible for this phenomenon has not been elucidated. Thrombocytopenia appears 24 hours after IV administration of Ad and is vector dose dependent. In this study, we have assessed the potential roles of the adhesive proteins P-selectin and von Willebrand Factor (VWF) on the aggregation and clearance of platelets following virus administration. We have addressed the question of whether the thrombocytopenia is due to a direct effect of the virus on platelets or an indirect effect related to interaction of platelets with other proteins or cells modified by the virus. We assessed platelet count in a group of Balb/c and C57Bl/6 mice over 1 week period following Ad administration and performed a detailed examination of the events within the first 24 h after Ad injection, the period that precedes the appearance of thrombocytopenia. We examined the effect of Ad on expression of the platelet activation marker P-selectin and the formation of platelet leukocyte aggregates (PLA) by means of flowcytometry after incubation of adenovirus with mouse platelets in vitro, and following Ad administration in vivo. To assess the role of VWF in Ad-induced thrombocytopenia we measured plasma VWF levels one hour after injection of Ad. Further investigations involved comparison of platelet counts, platelet activation, and the formation of PLA in a group of VWF KO mice. All studies have been performed with a replication deficient E1/E3-deleted Ad 1x 1011 viral particles/mouse. Our in vitro studies have shown that Ad directly activates mouse platelets as shown by increased expression of P-selectin. The average index of platelet activation for platelets stimulated by Ad was 2519.4 compared to 128.2 for resting platelets (n=5, p<0.02). Flow cytometric analysis of CD41 (platelets) and CD45 (leucocytes) double stained positive events indicated that Ad stimulation induced PLA when compared to the unstimulated samples. Our in vivo studies have confirmed the development of significant thrombocytopenia in both Balb/c as well as C57Bl/6 WT mice (n=8, p=0.00001, n= 6, p=0.002) 24 hours following Ad administration. Significant P-selectin expression was documented in both strains (n=4,p=0.0003; n=3, p=0.0008 respectively) as well as significant PLA one hour following Ad (n=4, p=0.01; n=3, p=0.007). The VWF KO mice showed non-significant thrombocytopenia (n= 6, p=0.063) at 24 hours following Ad, significant P-selectin expression (n=3, p=0.0003), but no significant PLA formation at one hour (n=3 p=0.12) relative to pre-injection levels. Plasma VWF levels were significantly elevated in both Balb/c and C57Bl/6 WT mice one hour following administration of the virus (n= 3, p=0.02; n= 3, p= 0.001). The average plasma VWF levels were 48.1 U/mL at 1h compared to 5.7 U/mL pre injection in Balb /c mice and 85.9 U/mL compared to 6.1 U/mL in C57Bl/6 mice. These studies have shown that Ad can act as an inducer of mouse platelet activation and as a promoter for platelet-leukocyte association both in vitro and in vivo. We have demonstrated a role for Ad in stimulating VWF release from the endothelium, and have shown that VWF has a critical role in platelet activation and clearance following Ad administration. We conclude that P-selectin and VWF proteins are directly involved in interactions between endothelial cells, platelets and leukocytes, a complex interaction that can explain at least in part the mechanisms underlying Ad-mediated thrombocytopenia.

Immune Thrombocytopenia Mediated by Anti-GPIbα Antibodies May Occur Via an FcR-Independent Pathway: A Potential Explanation for Refractory Cases to IVIG Therapy.

Intravenous immunoglobulin G (IVIG) is used to treat immune thrombocytopenia (ITP). While benefit has been observed in many cases, for unknown reasons some patients are refractory to this therapy. ITP is characterized by platelet clearance mainly mediated by antibodies against platelet surface glycoproteins (GP) GPIIbIIIa and GPIbα. Approximately 70–80% patients have antibodies against GPIIbIIIa, 20–40% patients have antibodies against GPIbα complex, and some patients have antibodies to both. Phagocytotic macrophages activated via their Fc receptors (FcR) are thought to be responsible for the clearance of opsonized platelets. However, anti-GPIbα antibodies may be able to induce ITP in an FcR-independent manner. Since the prevailing view of the mechanism of action of IVIG is Fc receptor blockade, we hypothesized that IVIG may not be effective at preventing thrombocytopenia induced by anti-GPIbα antibodies. Thrombocytopenia was induced in BALB/c mice using monoclonal antibodies against mouse GPIbα (p0p3, p0p4, p0p5, p0p9, and p0p11). Pretreatment with different doses of IVIG (1, 2, and 4g/kg of mouse) failed to prevent ITP in all anti-GPIbα-treated mice, except for p0p4. These results were repeated with p0p3, p0p4, and p0p5 using 2g/kg IVIG in C57BL/6 mice, and using 2g/kg of a different IVIG preparation. It has been shown that F(ab)2' fragments of the p0p antibodies can induce thrombocytopenia to the same extent as intact antibodies. To confirm that these anti-GPIbα antibodies act in an FcR-independent manner, we demonstrated that p0p3, p0p4, and p0p5 are able to induce thrombocytopenia in mice lacking activating Fc receptors (FcRγ chain −/−). Interestingly, IVIG was still able to prevent thrombocytopenia induced by p0p4 in these FcRγ chain −/− mice. We demonstrated in vitro that there are no anti-idiotype antibodies against any of the anti-GPIbα antibodies used present in IVIG. Thus, the amelioration of thrombocytopenia in p0p4-treated mice is not due to a neutralization of the antibody by IVIG. In contrast to the anti-GPIbα antibodies, anti-GPIIbIIIa monoclonal antibodies (JON1, JON2, and JON3) that induce thrombocytopenia were sensitive to IVIG treatment, which is consistent with earlier studies using other monoclonal antibodies against GPIIbIIIa. Also, JON2 was unable to induce thrombocytopenia in FcRγ chain −/− mice, confirming that this anti-GPIIbIIIa antibody is FcR-dependent. Our results suggest that ITP induced by anti-GPIbα antibodies may be mediated by FcR-independent mechanisms, many of which may be refractory to IVIG therapy. Thus, IVIG may not be as effective in treating patients with predominantly anti-GPIbα antibodies.

Development of a New Experimental Animal Model of Human Megakaryopoiesis Using Immunodeficient NOD/Shi-scid, IL-2Rγnull (NOG) Mice.

[Background] We developed novel small molecules with thrombopoietin (TPO) mimetic activities in vitro. These small molecules induced colony formation of megakaryocytes from human bone marrow cells but not murine and cynomolgus monkey cells, indicating strict species specificity. However, the lack of experimental animal models for human megakaryopoiesis hampered evaluation of whether these molecules can increase the number of human platelets in vivo. To solve this problem, we developed a new experimental animal model of human megakaryopoiesis using immunodeficient NOD/Shi-scid, IL-2Rγnull (NOG) mice.[Aims] We developed a new experimental animal model of human megakaryopoiesis using NOG mice to evaluate the in vivo efficacies of the novel human-specific TPO mimetics. We also compared the potentials of cord blood (CB), bone marrow (BM), and peripheral blood (PB)-derived CD34 positive cells at maintaining human megakaryopoiesis in NOG mice.[Methods] After 2.4-Gy X-ray irradiation, 1x105 CD34-positive hematopoietic cells from human CB, BM and PB were intravenously injected into NOG mice. The engraftment of human megakaryocytes and platelets in NOG mice was analyzed by flow cytometry and immunohistochemistry with human-specific antibodies for up to 6 months.[Results] Four weeks after transplantation, we found that around 1% of platelets in these mice were positive for human CD41 antigen (fibrinogen receptor) (n=8). We also confirmed 30–40% of megakaryocytes in the murine bone marrow were positive for human-specific megakaryocyte markers by immunohistochemistry. The percentage of human platelets in the murine blood was maintained at around 2% for 6 months after transplantation with human CD34-positive CB cells. In contrast, the engraftment of human platelets was transient and almost disappeared after 3 months, when CD34-positive cells from human BM and PB were injected into NOG mice. Interestingly, we found around 20 and 40% of murine bone marrow cells were positive for human CD45 three months after transplantation with PB- and BM-derived CD34-positive cells, respectively (n=7). These results indicated that CB cells have higher potencies to maintain human megakaryopoiesis than BM and PB cells in NOG mice. As the percentage of human platelets was much lower than that of human megakaryocytes in NOG mice transplanted with human CB cells, we speculated that the clearance of human platelets, which are larger than murine platelets, might be high or that the murine bone marrow microenvironment lacks factors related to platelet release from human megakaryocytes. We also analyzed the function of human platelets in NOG mice. ADP induced the activation of alpha IIb beta IIIa (fibrinogen receptor) and the expression of CD62P on human platelets in NOG mice in vitro, suggesting that human platelets produced from human CB-derived CD34 positive cells in NOG mice, are functional.[Conclusions] We succeeded to develop a new experimental animal model of human megakaryopoiesis using NOG mice. CB was the most suitable donor for maintaining human matured megakaryocytes and functional platelets in NOG mice. Currently, we are using this model to evaluate in vivo efficacies of novel human-specific TPO mimetics.

Differential Changes in Platelet VWF Receptor Following Refrigeration for Short or Long Periods.

One pathway that mediates the rapid removal of transfused platelets exposed to cold for short periods (≤ 2 h) is the clustering of the von-Willebrand factor (VWF) receptor complex. Clustered VWF receptors are recognized by the αMβ2 integrin on hepatic macrophages, which, through its lectin domain, binds exposed βN-acetylglucosamine (βGlcNAc) residues on the clustered GPIbα subunits (Cell 112: 87, 2003). Coverage of the exposed βGlcNAc residues by galactosylation prevents clearance of platelets from the circulation. We have now investigated detailed changes of the VWF complex on platelets during short (≤ 2 h) and prolonged refrigeration (≥ 48 h). The VWF receptor structure in chilled and rewarmed platelets was probed by flow cytometry using a panel of mAbs directed towards the GPIbα N-terminal (N–T) and GPIX (FITC-AN51 (aa36–59), PE-Cy5-HIP1 (aa59–81), FITC-SZ2 (aa268–282), and PE-GPIX mAb). Binding of the anti-GPIbα mAbs HIP1 and SZ2 as well as the anti-GPIX was diminished 40–50 % following 0.5 h of cooling when compared with room temperature (RT) platelets. The anti -GPIbα N-T mAb AN51 bound similarly to cooled and RT platelets. In contrast, the binding of all anti-VWF complex mAbs was decreased by 50–60 % after refrigeration of platelets in plasma for 48 h when compared to fresh RT platelets. Immunoblots demonstrated that loss of binding did not result from GPIbα shedding. VWF and fibrinogen binding were also increased by ~ 5 % or ~ 20 %, respectively, and P-selectin exposure increased by ~ 30 % in 48 h refrigerated platelets compared to the RT platelets stored for 48 h as determined by flow cytometry. These results suggest that the conformation of the VWF receptor is altered by long term refrigeration which leads to VWF binding and partial platelet activation. Loss of antibody binding is not simply due to GPIbα clustering as incubation of platelets with 5 μM latrunculin A for 1 h, which induces profound GPIbα clustering, increased binding by ~ 30 % compared to untreated platelets, but did not alter the diminished mAb binding to platelets induced by chilling. Refrigerated (≤ 14 days) platelets still adhere well to collagen-coated surfaces under flow (1000 seconds−1 for 2 min) demonstrating that VWF receptor function remains intact. We conclude that refrigeration of platelets in plasma for ≥ 48 h induces both conformational changes and clustering of the VWF complex, to prime it for activation by inducing VWF binding.

Profiles of Antibodies Specific to Platelet Glycoproteins in ITP Differ in Remission, Relapse and Stable Phases; and Anti-GpIIb/IIIa, -GpIb/IX Were Associated with Bleeding Manifestations.

BACKGROUND: In immune thrombocytopenic purpura (ITP), antiplatelet antibodies (APtAb) bind to antigens exposed on the platelet surface to mediate platelet clearance from circulation. GpIIb/IIIa, GpIb/IX and GpIV, common target antigens in ITP, are receptors for fibrinogen, von Willebrand factor and collagen, respectively, and are involved in platelet aggregation and adhesion, playing critical roles in hemostasis and thrombosis. It has been suggested that APtAb binding to these molecules interfere with their function, inducing platelet dysfunction and enhancing bleeding in ITP. In the present study, we investigated profiles of antibodies against these antigens in different stages of ITP, and in relation to bleeding manifestations.METHODS: We measured IgG and IgM APtAb against the three most prevalent glycoprotein target antigens, GpIIb/IIIa, GpIV and GpIb/IX, by PAICA [Macchi, et al; 1996, Thromb Haemost 76:1820]. We studied 130 patients with ITP. There were 87 women (66.9%) and 43 men (33.1%). ITP was classified into three stages, (i) acute/relapse, 28 pts (21.5%); (ii) chronic/stable (platelet counts <150,000 for at least 6 mo's), 69 pts (53.0%); (iii) remission, 33 pts (25.3%). The chronic/stable patients were further divided into bleeders (n=27) and non-bleeders (n=19). Those patients with visible multiple petechias or echymosis at more than three sites, or mucosal bleeding were defined as bleeders; non-bleeders had none of these signs.RESULTS: (1) IgG APtAb against GpIIb/IIIa, GpIV and GpIb/IX were all significantly more frequent in acute/relapse (77%, 70%, and 67%, respectively) than in chronic/stable ITP (52%,55%, and 44%), at p=0.005, p=0.003, p=0.04, respectively. (2) All IgG APtAb were significantly more frequent in acute/relapse and chronic/stable ITP compared to remission, p<0.0001. (3) IgM APtAb were detected in approximately 50% of patients in relapse and stable stages for all 3 antigens; there was no significant difference in frequency between acute/relapse and chronic/stable, but both were significantly higher than in remission. (4) Multiple antibodies (≥ 2 antigens positive) were observed in 50–70% of patients in both relapse and stable stages. (5) When we compared frequencies of APtAb between bleeders and non-bleeders, both IgG and IgM APtAb were significantly more frequent in bleeders, for all antigens except IgM GpIb/IX. Interestingly, 96.7% of bleeders had at least one APtAb of IgG or IgM class, while only 58% were positive in non-bleeders (p<0.0002). This difference persisted when reanalyzed to ensure equal mean platelet counts between bleeders and non-bleeders. It was also found that positivity for APtAb of IgG (but not IgM) class against the pair, GpIIb/IIIa and GpIb/IX (but not GpIV) was predictive of bleeding.CONCLUSIONS. Distinctive profiles of APtAb to GpIIb/IIIa, GpIb/IX, and GpIV were observed in different stages of ITP. IgG APtAb were frequently seen in acute/relapse and chronic/stable, both clases (IgG-IgM) were rarely seen in remission. Patients with signs of bleeding more frequently had IgG APtAb to the pair, GpIIb/IIIa and GpIb/IX, suggesting this pair of APtAb interferes with platelet hemostatic functions, enhancing the bleeding manifestations of ITP.

A Revisit of the Criteria To Define Remission in Patients with Idiopathic Thrombocytopenia Undergoing Splenectomy.

IntroductionIn patient with immune thrombocytopenic purpura (ITP), the responses to splenectomy are traditionally categorized based on the platelet count. A platelet count of 150 x109/L is generally considered as complete remission. However, asplenic patients may have higher baseline platelet counts. To investigate whether the definition of the remission affects the interpretation of follow up data, we have studied patients with ITP undergoing splenectomy. The results were also compared with a meta-analysis of all the trials published in the past 30 years.Methods and MaterialsPhase I: A retrospective chart review was performed of all patients referred to a tertiary medical center before the splenectomy. The peak post-splenectomy count (PPSC) was defined as the platelet count within 100 days after splenectomy. The patients were followed for various durations. Long-term outcomes after splenectomy were categorized as remission or failure. A failure was defined as a platelet count of less than 30 x109/L after the first 100th post-operative day or the need for further long-term therapy to maintain the platelet count at a safe level. The clinical outcomes were independently assessed by two investigators. The relationship between PPSC and clinical outcome was assessed via logistic regression.Phase II: The published case series of patients with ITP were identified through MEDLINE® search. Study inclusion criteria were defined as cohort of adult patients with ITP or adult patients undergoing splenectomy. 2550 articles were found through MEDLINE® search. The results of the 51 publications involving 4198 patients were summarized.As the focus of this review is on the failure rates of splenectomy for patients with ITP, the published results were recalculated according to a set of pre-defined criteria with dichotomized variables (remission or failure). A failure was defined as a platelet count of less than 50 x109/L, 100 x109/L or 150 x109/L whatever the lowest platelet count used in the study or the need for further long term post-splenectomy therapy. Two reviewers extracted the information from the publications independently to ensure consistent reporting of the results.ResultsPhase I: There was a significant correlation between the PPSC and clinical outcome only when the PPSC was higher than 150 x109/L. Logistic regression indicated that for each increase in the PPSC of 10 x109/L there was a 5% risk reduction in the failure rate after splenectomy.Phase II: The long-term failure rates were extracted from published series. There has been no significant difference in the long-term failure rate of splenectomy when the initial remission criteria were defined as a post-splenectomized platelet count above 50 x109/L, 100 x109/L or 150 x109/L.DiscussionThe post-splenectomy platelet counts are determined by remaining platelet clearance mechanism(s); the perioperative medical treatments; and other concomitant medical conditions. The normal range of platelet count in asplenic population is poorly defined. As the consensus is not to treat asymptomatic ITP patients with a platelet count higher 30 x109/L, the parameter used to measure the long-term outcome of patients with ITP should derive from these clinical criteria. Simplified dichotomized variables (remission or failure) may provide a better description of the outcome of the treatments after splenectomy in patients with ITP.

Therapeutic expression of the platelet-specific integrin, αIIbβ3, in a murine model for Glanzmann thrombasthenia

AbstractIntegrins mediate the adhesion of cells to each other and to the extracellular matrix during development, immunity, metastasis, thrombosis, and wound healing. Molecular defects in either the α- or β-subunit can disrupt integrin synthesis, assembly, and/or binding to adhesive ligands. This is exemplified by the bleeding disorder, Glanzmann thrombasthenia (GT), where abnormalities of the platelet-specific integrin, αIIbβ3, prevent platelet aggregation following vascular injury. We previously used a retrovirus vector containing a cDNA cassette encoding human integrin β3 to restore integrin αIIbβ3 on the surface of megakaryocytes derived from peripheral blood stem cells of GT patients. In the present study, bone marrow from β3-deficient (β3–/–) mice was transduced with the ITGβ3-cassette to investigate whether the platelet progeny could establish hemostasis in vivo. A lentivirus transfer vector equipped with the human ITGA2B gene promoter confined transgene expression to the platelet lineage. Human β3 formed a stable complex with murine αIIb, effectively restoring platelet function. Mice expressing significant levels of αIIbβ3 on circulating platelets exhibited improved bleeding times. Intravenous immunoglobulin effectively diminished platelet clearance in animals that developed an antibody response to αIIbβ3. These results indicate the feasibility of targeting platelets with genetic therapies for better management of patients with inherited bleeding disorders.

Platelet homeostasis is regulated by platelet expression of CD47 under normal conditions and in passive immune thrombocytopenia

AbstractInteraction between target cell CD47 and the inhibitory macrophage receptor signal regulatory protein α (SIRPα) counteracts macrophage phagocytosis of CD47-expressing host cells. As platelets also express CD47, we asked whether inhibitory CD47/SIRPα signaling regulates normal platelet turnover and clearance of platelets in immune thrombocytopenic purpura (ITP). CD47-/- mice had a mild spontaneous thrombocytopenia, which was not due to a decreased platelet half-life as a result of increased expression of P-selectin, CD61, or phosphatidylserine. In contrast, CD47-/- platelets were rapidly cleared when transfused into CD47+/+ recipients, whereas CD47+/- platelets had a nearly normal half-life in CD47+/+ mice under nonautoimmune conditions. CD47-/- mice were more sensitive to ITP, as compared with CD47+/+ mice. In vitro, macrophage phagocytosis of immunoglobulin G (IgG)–opsonized CD47-/- platelets was significantly higher than that for equally opsonized CD47+/+ platelets. However, when SIRPα was blocked, phagocytosis of CD47+/+ platelets increased to the level of CD47-/- platelets. Phagocytosis of opsonized CD47+/- platelets was higher than that for CD47+/+ platelets, but lower than that for CD47-/- platelets, suggesting a gene-dose effect of CD47 in this system. In conclusion, we suggest that inhibitory CD47/SIRPα signaling is involved in regulating platelet phagocytosis in ITP, and that targeting SIRPα may be a new means of reducing platelet clearance in ITP.

410. Adenovirus Activates Mouse Platelets and Induces Platelet Leucocyte Association

Adenovirus vectors have been used extensively in pre-clinical and clinical studies as agents for gene delivery. Acute, transient thrombocytopenia is a known adverse effect of adenovirus vector administration with a mechanism that currently remains unresolved. The fall in platelets appears 24 hours following IV administration of adenovector and is vector dose dependent. We have initiated experiments to investigate the mechanism of Ad induced thrombocytopenia. We examined the expression of the platelet activation marker P-selectin (CD-62p) on mouse platelets after incubation with adenovirus in vitro by means of flow cytometry. An indirect immunoflourescence approach was used with anti-mouse CD 62p monoclonal antibody and an FITC labeled rat antimouse-IgG. An appropriate negative isotype control and positive platelet activation control using thrombin were incorporated into these studies. The index of platelet activation (IPA), was used to quantify the activation effect of Ad on the platelets. This is the product of the mean channel fluorescence and the percentage of positive cells. We also examined Platelet leucocyte association (PLA) using the direct immunoflourescence and double colour staining of platelets and leucocytes using FITC labeled anti-mouse CD41 (platelets) and PE labeled anti-mouse CD 45 (common leucocyte antigen) respectively. Our study has shown that the mean IPA for platelets stimulated by adenovirus was 2519.4 compared to 128.2 for resting platelets (n=5, p<0.02). Flow cytometric analysis of CD41 (platelets) and CD45 (leucocytes) double stained positive events indicated that Ad stimulation induced PLA when compared to the unstimulated samples. PLA visualisation was also confirmed by means of direct fluorescence microscopy. These preliminary in vitro results indicate that the Adenovirus activates mouse platelets as it induces degranulation and expression of P-selectin and supports the recruitment of leucocytes presumably through an interaction with the leucocyte P-selectin glycoprotein ligand-1. This may have a role in mediating the accelerated platelet clearance by the reticuloendothelial system following adenovirus administration. In addition, the interaction with leucocytes may also stimulate pro-coagulant microparticle generation and thus lead to systemic hemostatic activation. To our knowledge, this is the first report that identifies Adenovirus as an inducer for mouse platelet activation and as a promoter for platelet leucocyte association. However, further in vitro and in vivo studies are still needed to better understand the mechanism of Ad induced thrombocytopenia; a complication that significantly limits the application of Ad mediated gene therapy.

Current methods of assessing platelet function: relevance to transfusion medicine

Current methods of assessing platelet function: relevance to transfusion medicine

Severe Thrombocytopenia, Type 2B von Willebrand Disease and Pregnancy

VON Willebrand’s Disease (vWD) is the most common inherited bleeding disorder in humans with an estimated incidence as high as 2–3% in the general population. Characterized by abnormal platelet interactions with the subendothelium or other platelets, the disease is caused by changes in the multimeric glycoprotein, von Willebrand’s Factor (vWF). Types 1 and 3 vWD are associated with relative or absolute quantitative defects in the protein, respectively. Type 1 vWD accounts for 70% of all cases and is likely to temporally improve in parturients as a result of an increase in vWF and factor VIII with pregnancy. Type 2 vWD has qualitative abnormalities and comprises 20–30% of all vWD diagnoses. A unique subtype, type 2B, accounts for less than 20% of all type 2 vWD and is characterized by an increased affinity of vWF for platelet glycoprotein Ib, resulting in spontaneous binding and clearance of both vWF and platelets. This feature allows type 2B, unlike most other vWD variants, to be exacerbated by pregnancy and to exhibit a poor or worsening response to desmopressin (DDAVP; Aventis, Bridgewater, NJ). We present the management of a parturient with type 2B vWD who had severe thrombocytopenia during the peripartum period.

Role of P-Selectin and GPIbα in the Fast and Delayed Clearance of Transfused Platelets.

Platelets become activated during preparation and storage of platelet concentrates (PCs) for transfusion. Flow cytometric assays of platelet activation can be employed for quantifying in vitro quality of PCs. It remains, however, unclear whether the level of in vitro platelet activation in stored PCs correlates with in vivo survival of the platelets after transfusion. Platelet surface glycoprotein (GP) Ibα and P-selectin (CD62) can be involved in regulation of posttransfusion PC clearance, mediating adhesive interactions of platelets with counter-receptors on leukocytes and endothelial cells. Recently, we described a rabbit model for analyzing posttransfusion kinetics of human PCs (Leytin et al, Transfusion 42:711, 2002, Transfusion 43:983, 2003). In the present work, we used this validated model for studying the implication of CD62 and GPIbα expression in posttransfusion PC clearance. Platelet activation in vitro was determined by flow cytometry using anti-CD62 and anti-GPIbα antibodies. PC clearance in vivo was evaluated in rabbits with inhibited reticuloendothelial system, as measured by 0.5 hr (R0.5), 24 hr (R24) and total (R∑) platelet recoveries, and survival time (ST). Correlations were analysed between in vitro assays of platelet activation and in vivo clearance of conventional (Day 2–5) and outdated (Day 7–8) PCs stored at 22°C, and refrigerated PCs. We found that the binding of anti-CD62 antibody was significantly increased in outdated and refrigerated PCs compared to conventional PCs, reflecting an increased exposure of CD62 on the platelet surface. In contrast, binding of anti-GPIbα antibody was significantly decreased during prolonged and refrigerated PC storage. The clearance of conventional (Day 2–5) PCs from the circulation can be described by a biphasic survival curve. The first (early) phase of platelet clearance is characterized by fast (≥ 14 x 109 platelets per hour) platelet removal, whereas the second (delayed) phase has a much slower rate of platelet clearance (approximately 0.4 x 109 platelets per hour). The biphasic survival curves were also obtained for outdated and refrigerated PCs, and were employed for determining fast (R0.5), delayed (ST and R24) and overall (fast + delayed; R∑) platelet clearance in vivo. We found that when stored PCs are cold-damaged, their in vivo viability decreased significantly, in comparison to conventional PCs, as reflected by the fast, delayed and overall platelet clearances. Viability of Day 7–8 PCs is also decreased, compared to Day 2–5 PCs, but only the fast and overall platelet clearance increased significantly. Negative correlation was observed between in vitro anti-CD62-binding to platelets and their fast, but not delayed, clearance. In contrast, anti-GPIbα-binding showed positive correlations with delayed, but not fast, platelet clearance. Overall clearance correlated better with anti-GPIbα- than with anti-CD62-binding. We also demonstrated that CD62 is shed from the platelet surface after transfusion, whereas GPIbα remains unchanged on the surface of circulating platelets. The data suggest that CD62 exposure during PC storage triggers fast CD62-mediated PC clearance. However, after CD62 shedding during platelet circulation, in vitro GPIbα alterations, such as cleavage, clustering or conformation changes, may determine long-term GPIbα-mediated PC clearance.

Tumor Necrosis Factor-α-Converting Enzyme (ADAM17) Mediates GPIbα Shedding from Platelets In Vitro and In Vivo.

Interaction of the platelet receptor GPIb-V-IX with von Willebrand factor exposed at a site of vascular injury is an essential step in the initiation of a hemostatic plug. Proteolytic cleavage (shedding) of the GPIbα subunit was first described more than 25 years ago, the protease mediating this event as well as its physiological function, however, have not been elucidated. We have recently reported that shedding of GPIbα and platelet clearance induced by platelet storage or mitochondrial injury involves a platelet-derived metalloproteinase(s). Here we show that GPIbα shedding in response to mitochondrial injury or physiological activation is inhibited in platelets obtained from chimeric mice, which express inactive tumor necrosis factor-α converting enzyme (TACEΔZn/ΔZn) in blood cells only. Shedding was also inhibited in mouse and human platelets in the presence of two potent TACE inhibitors, TAPI and TMI-1. Our data further suggest that TACE is important in the regulation of GPIbα expression in vivo, as we observed an ~ 90% reduction in soluble GPIbα (glycocalicin) in plasma of TACEΔZn/ΔZn chimeras as well as significantly increased levels of GPIbα on circulating platelets. In contrast, shedding of P-selectin from activated platelets was not affected by the mutation in TACE. Mitochondrial damage in TACEΔZn/ΔZn platelets did not result in reduced posttransfusion recovery or impaired hemostatic function of such platelets in mice, indicating that TACE not only mediates GPIbα shedding but that it may also be crucial for the development of cellular changes leading to platelet clearance. In conclusion, our data demonstrate that TACE is expressed in platelets and that it is the key enzyme mediating shedding of GPIbα. Inhibition of TACE during platelet storage could be a powerful means to improve the effectiveness of platelet transfusions.

The Macrophage αMβ2 Integrin αM-Lectin Domain Mediates the Phagocytosis of Chilled Platelets.

αMβ2 integrin receptors on myeloid cells mediate the adhesion or uptake of diverse ligands, and particular sequences on this integrin's alpha subunit are responsible for recognition specificity. Inhibition of neutrophil adhesion to platelet monolayers by αM-derived peptides has implicated an αM sequence (I-domain) that also binds iC3b, fibrinogen, ICAM-1 and other ligands. However, αMβ2 also recognizes βN-acetylglucosamine (βGlcNAc) residues on the platelet von Willebrand factor receptor (GPIbα) that are clustered after cooling, suggesting that the αM lectin domain is important for this recognition. The phagocytosis of chilled platelets can be reconstituted when CHO cells are transfected with αMβ2. Removal of the N-T portion of GPIbα, containing the exposed βGlcNAc residues, from the chilled platelet surface using the snake venom mocarhagin inhibits this phagocytosis, as does the addition of 1 mM β-methyl-N-acetyl-D-glucosamine. Replacement of the I domain, or the lectin domain of the αM-chain with the corresponding domain from the αx-chain (p150), reveals that the activity of the αMβ2 integrin toward chilled platelets resides within the lectin-domain, and does not require the I domain. Other evidences for this conclusion include: (1) Sf9 cells expressing αM lectin domain but not β2 bind 3-4 x more chilled platelets than control cells; βGlcNAc inhibits this binding. (2) Soluble recombinant αM lectin domain inhibits the phagocytosis of chilled platelets by αMβ2-expressing THP-1 cells while I-domain substrates (fibrinogen or LPM19c), recombinant I-domain protein, or a synthetic peptide that inhibits neutrophil adhesion to platelet monolayers, are without inhibitory effect. We conclude that chilled platelets are removed from blood by an interaction between βGlcNAc residues on clustered GPIbα and the lectin domain of αM-chain of the αMβ2 integrin. Since the phagocytosis and clearance of chilled platelets by hepatic macrophages has precluded refrigeration of platelets procured for transfusion, a more precise molecular definition of the recognition mechanism could inform strategies for its inhibition, thereby accommodating cold platelet storage.

An Externally Active Endogenous Platelet Galactosyl Trasferase (Gal-T).

Over 30 years ago, Jamison and Barber proposed that externally disposed glycosyl transferase (GT) activity mediates platelet adhesion and other functions. Subsequent work ruled out ecto-GT activity in nucleated cells and established the Golgi as the site of such enzymes, although no further studies examined platelets. We recently reported a platelet-associated galactosyl transferase (β4Gal-T) catalyzing galactose coupling in a β1,4 linkage to exposed N-acetylglucosamine (GlcNAc) residues on N-linked glycans of the GPIbα subunit of the von Willebrand factor receptor complex. This activity, also present in plasma, mediates galactose transfer to GPIbα following addition of UDP-galactose (UDP-Gal) alone. Since galactosylation of GPIbα prevents clearance of chilled platelets by liver macrophages after transfusion, sufficiently active endogenous endogenous β4Gal-T could facilitate platelet cold storage after the mere addition of substrate UDP-Gal (Science 301: 1531–1534, 2003). We now give evidence that human platelets contain a highly active (0.05 nmol/min) plasma membrane-associated ecto-β4Gal-T with properties of an isoform designated as type 1 (β4Gal-T1). A specific β4Gal-T1 antibody immunoprecipitates β4Gal-T1 activity from platelet lysates and the inhibition profile activity by α-lactalbumin and benzyl-β-D-GlcNAc are characteristic of β4Gal-T1. β4Gal-T1 has a transmembrane domain, and platelet β4Gal-T, although releasable by non-ionic detergent, remains cell- and membrane associated despite extensive washing. The platelet enzyme incorporates UDP-[C14]Gal into a single polypeptide with the apparent Mr of GPIbα, identifying this protein as its unique endogenous target. Nevertheless, washed platelets are as active as recombinant β4Gal-T1 in galactosylating exogenous substrates, benzyl-β-D-GlcNAc and ovalbumin. In conclusion, we have revived the concept of an ecto-glycosyl transferase on blood platelets and demonstrated that platelet β4Gal-T is probably sufficiently active to accommodate reliable enzymatic conversion of platelets by simple addition of UDP-Gal.

Distinctive Efficacy of IVIG in Ameliorating Thrombocytopenia Induced by Anti-Platelet GPIIbIIIa and GPIbα Antibodies.

Idiopathic autoimmune thrombocytopenia (ITP) is an autoimmune disease characterized by autoantibody-mediated platelet destruction. The major platelet antigens that autoantibodies are directed against in ITP are GPIIbIIIa and GPIbα. Intravenous immunoglobulin G (IVIG) is a treatment used in ITP; however, some patients are refractory to this treatment and the mechanisms of action of IVIG in this disease are not fully understood. We tested the hypothesis that IVIG may not be equally effective in preventing ITP caused by GPIIbIIIa versus GPIbα anti-platelet antibodies. Thrombocytopenia was induced in Balb/c mice using three monoclonal antibodies against either mouse platelet GPIIbIIIa (JON1, 2, −3) or against GPIbα (p0p 3, 4, −5). Mouse platelet counts were performed daily from day one (pre-treatment) until day five using a flow cytometric assay. The efficacy of IVIG in inhibiting monoclonal antibody-induced platelet destruction in mice was assessed by intraperitoneal injection of IVIG (2 g/kg) 24-hours before anti-platelet antibody injections (7 mg/mouse, i.v.) on day two. We found that pre-treatment with IVIG resulted in a significantly higher platelet count on day three in all groups of anti-GPIIbIIIa-injected animals (n=3/antibody; for JON1, P=0.002; JON2, P=0.015; JON3, P=0.002) as compared to albumin-treated controls. In contrast, IVIG failed to prevent platelet clearance and thrombocytopenia on day three in two anti-GPIbα antibody-treated groups (p0p3 and p0p5, n=3/antibody), but did inhibit platelet destruction in mice that were injected with p0p4 antibody (P=0.0003). Our results indicate that the efficacy of IVIG treatment in preventing ITP induced by these anti-platelet antibodies is affected by their antigen specificities. These data also suggest that patients with ITP mediated by GPIbα may be less responsive to IVIG treatment than those with ITP mediated by GPIIbIIIa antibodies.

Thrombophilia Associated with Anti-CD154 Monoclonal Antibody Treatment and its Prophylaxis in Nonhuman Primates

We agree with the findings of Koyama et al. (1) published in a recent issue of Transplantation, and would like to share published insight pertaining to FcγRIIA (CD32) mediated platelet aggregation in relation to CD40L/CD154 mABs. The authors describe the efficaciousness of ketorolac in reducing thromboembolism in cynomolgus monkey renal allograft recipients treated with anti-CD154 monoclonal antibody (mAb). Blockade of the CD40L/CD154 pathway has been hailed as a promising strategy in prolonging allograft survival in murine allograft models (2). This leads to the initiation of trials including anti-CD40L/CD154 mAb in the immunosuppressive regimen in human allograft recipients. These trials were, however, precociously terminated because of an unusually high incidence of thromboembolic complications in these allograft recipients. This clinical observation ignited the study presented in Transplantation. From unpublished data in their discussion section, Koyama et al. (1) speculate that the underlying mechanism of platelet activation/aggregation might be Fc-receptor dependent. Two pathologic steps are required for immune mediated thrombocytopenia and platelet activation/aggregation: 1) presence of Ab to a self or neo-Ag; and 2) triggering of Ab effector mechanisms that leads to accelerated platelet clearance and/or platelet activation/aggregation, representing the low affinity receptor FcγRIIA (CD32) on platelets. FcγRIIA (CD32) is a single-chain sialoglycoprotein, which interacts with the Fc-domain of IgG antibodies. When cross-linked, FcγRIIA (CD32) fully activates platelets for secretion and activation (3). The large majority of antibodies have no effect on platelets. By contrast, only a subset of antibodies are able to directly promote platelet activation, although it has been reported that binding of these antibodies, via their antigen-binding domain alone, suffices to induce platelet activation (4). It is of interest to acknowledge, in respect to the murine allograft model, that mice lack the genetic equivalent of human and simian FcγRIIA (CD32) and therefore have not demonstrated thrombosis after being confronted with CD40L/CD154 mABs. In our previous work (5), we were able to define a mechanism by which polyclonal antithymocyte-globulin (ATG)—a purified pasteurized gamma globulin suspension and a drug utilized as induction therapy in allograft recipients—is causing platelet depletion in allograft recipients. ATG (thymoglobulin) induces surface platelet activation markers CD62P and CD63, platelet aggregation, and secretion of platelet-bound sCD40L (CD154) in in vitro assays. AT10, a FcγRIIA blocking antibody, is able to abrogate these described effects. Our results find that ATG, containing activating Fc domains, was able to bind the FcγRIIA (CD32) receptor on platelets as well as autologous Fab-fragments (e.g. CD40L/CD154, MHC-class I, CD31) and thus was responsible for the activation and aggregation of platelets in these in vitro assays (5). We are tempted to believe that CD40L/CD154 related thrombophilia/thrombosis is mediated by platelet activation via CD40L/CD154, as well as the presence of functional human FcγRIIA (CD32) receptor. Georg Alexander Roth Andreas Zuckermann Walter Klepetko Ernst Wolner Hendrik Jan Ankersmit Bernhard Moser Ivo Volf Department of Surgery, General Hospital, Vienna Medical School, Vienna, Austria Department of Surgery, Columbia University, New York, New York Institute of Pathophysiology, Vienna Medical School, Vienna, Austria

Role of platelet surface glycoprotein Ibalpha and P-selectin in the clearance of transfused platelet concentrates.

Role of P-selectin (CD62) and glycoprotein (GP) Ibalpha in posttransfusion clearance of platelet concentrates (PCs) is unclear. Platelet (PLT) activation in vitro was determined by flow cytometry using anti-CD62 and anti-GPIbalpha. PC clearance in vivo was evaluated in an animal model using rabbits with an inhibited reticuloendothelial system, as measured by 0.5-hour (R(0.5)), 24-hour (R(24)), and total (R( summation operator )) PLT recoveries, and survival time (ST). Correlations were analyzed between in vitro assays of PLT activation and in vivo clearance of conventional (Days 2-5), outdated (Days 7-8), and refrigerated PCs. Binding of anti-CD62 to the PLT surface was significantly increased and of anti-GPIbalpha decreased in outdated and refrigerated PCs compared to conventional PCs. Negative correlation was observed between in vitro anti-CD62 binding and the fast (R(0.5)) PLT clearance, but not with delayed (R(24) and ST) clearance. In contrast, anti-GPIbalpha binding showed positive correlations with delayed but not with fast PLT clearance. Overall (R( summation operator )) clearance correlated better with anti-GPIbalpha than with anti-CD62 binding. CD62 density on the PLT surface was decreased after PC transfusion, whereas GPIbalpha density remained unchanged. These data suggest that CD62 exposure on the PLT surface during PC storage triggers fast CD62-mediated PC clearance, whereas in vitro GPIbalpha changes are involved in delayed GPIbalpha-mediated PC clearance.

Tetramolecular immune complexes are more efficient than IVIg to prevent antibody-dependent in vitro and in vivo phagocytosis of blood cells.

Intravenous immunoglobulins (IVIg) have immunomodulatory effects in vivo and are widely used in the treatment of autoimmune diseases, such as idiopathic thrombocytopenic purpura (ITP). The mechanisms by which IVIg can prevent platelet clearance in ITP patients are not fully understood but are known to require the participation of low affinity Fcgamma receptors (FcgammaRs), which interact poorly with monomeric immunoglobulin G (IgG). Given the importance of low affinity FcgammaRs in the treatment of ITP, we hypothesized that immune complexes (IC) produced in vitro could reproduce the effects of IVIg. Small-size tetramolecular IC were prepared using mouse monoclonal anti-human IgG and human Fc fragments. The effects of tetramolecular IC and IVIg on the in vitro and in vivo inhibition of phagocytosis of opsonized blood cells were compared. The results obtained showed that tetramolecular IC were at least six times more efficient than IVIg to prevent phagocytosis of opsonized red blood cells in vitro, and clearance of platelets in the thrombocytopenic mouse model.

Pathologic high shear stress induces apoptosis events in human platelets

Recently, it has been discovered that apoptosis of anucleate platelets can be induced by chemical agonists. Other studies demonstrated that mechanical forces (shear stresses) stimulate platelet activation and signaling in the absence of exogenous chemical stimuli. We analyzed whether shear stresses can trigger platelet apoptosis, a question that has not yet been studied. Using a cone-and-plate viscometer, we exposed human platelet-rich plasma to different shear stresses, ranging from physiologic arterial and arteriole levels (10–44 dyn/cm 2) to pathologic high levels (117–388 dyn/cm 2) occurring in stenotic vessels. We found that pathologic shear stresses induce not only platelet activation (P-selectin upregulation and GPIbα downregulation) but also trigger apoptosis events, including mitochondrial transmembrane potential depolarization, caspase 3 activation, phosphatidylserine exposure, and platelet shrinkage and fragmentation, whereas physiological shear stresses are not effective.This novel finding suggests that shear-induced platelet apoptosis can be mediated by mechanoreceptors, does not require nuclear participation, and may affect platelet clearance.