Quantitative analysis of platelet surface HLA by W6/32 anti-HLA monoclonal antibody

Class I molecules of human major histocompatibility complex (HLA) are the most important antigenic system in determining the survival of transfused platelets in alloimmunized patients. Platelets with reduced expression of a specific type of HLA antigen may escape specific anti-HLA antibody-mediated destruction. By using 125I-labeled Fab fragments of W6/32 anti-HLA monoclonal antibody and competitive protein binding assays, we measured the range of total HLA concentrations on platelets. In 12 individuals examined, the mean number of HLA-A, B, and C molecules per platelet was 81,587 +/- 20,016 (mean +/- SD); its range was between 54,782 to 116,185 molecules per platelet. After treatment with chloroquine, 79.9 +/- 7.0% (mean +/- SD, n = 6) of HLA antigens were removed from platelets as determined by binding of 125I-W6/32 Fab. A similar result was obtained when HLA antigens on chloroquine-treated platelets were evaluated with immunofluorescence flow cytometry. In contrast, chloroquine treatment did not remove integral membrane protein such as P1A1 antigens on platelets. The presence of HLA antigens in the chloroquine eluate of platelets could be demonstrated to contain HLA antigens similar in mol wts to intact class I molecules by an immunoblotting technique. These data suggest that 70% to 80% of platelet HLA antigens are adsorbed and that such HLA antigens are not proteolytic products of integral membrane class I molecules. The origin of the adsorbed platelet HLA-antigens remains to be determined.     

Variation of class I HLA antigen expression among platelet density cohorts: a possible index of platelet age?

Platelet alloantigens and other surface markers were studied in platelet cohorts of different mean density, using monoclonal and polyclonal probes. High density (HD) platelets expressed 12% more P1A1 molecules (46,942) than low density (LD) platelets (41,892). However, LD platelets carried 42% more HLA-A2 molecules (6,267 +/- 184) than HD platelets (4,406 +/- 232) (P less than .01) and 55% more class I HLA antigens (17,034 +/- 2,062 v 11,007 +/- 2,190) (P = .05). The platelet subpopulations did not differ in their content of glycoprotein (GP)IIb/IIIa complex or Baka antigen. The difference in expression of class I HLA antigens on HD and LD platelets is consistent with two possibilities: either class I HLA molecules are acquired from plasma or they are released into plasma as platelets age in circulation. Accordingly, class I HLA molecules may provide a useful marker of platelet age.     

Monoclonal antibody-mediated inhibition of the human HLA alloimmune response to platelet transfusion is antigen specific and independent of Fcgamma receptor-mediated immune suppression

Presensitization of donor platelets with allo-specific immunoglobulin (Ig)G results in a diminished immune response against subsequent transfusions of platelets. To understand better the mechanism of how alloantibody presensitization results in a decreased alloimmune response, we have used murine monoclonal antibodies directed to polymorphic and non-polymorphic regions of human leucocyte antigen (HLA) as well as platelet-specific molecules. Here, we demonstrated that presensitization with anti-human HLA class I antibodies, as well as beta2-microglobulin-specific antibody, protected against alloantibody production to five subsequent untreated platelet challenges. Use of complement fixing, non-fixing or F(ab')2 fragments of HLA-specific antibody also resulted in complete inhibition of alloantibody production. This protection was not seen when the platelets were presensitized with monoclonal antibodies to CD42a (GPIX), CD32 (low-affinity IgG/Fcgamma receptor) or murine IgG and was thus independent of B-cell FcgammaRII-mediated immune suppression. The inhibition observed was independent of HLA alloantigenic specificity as antibodies directed at the beta2-microglobulin portion of HLA class I were as effective as antibodies against any of the HLA-alpha regions (either polymorphic or non-polymorphic) of class I. This work demonstrates that monoclonal antibody-mediated suppression of the human HLA alloimmune response to platelet transfusion is antigen specific and is independent of FcgammaRII-mediated immune regulation, complement fixing or HLA alloantigenic specificity.     

Platelet activation by cross-linking HLA class I molecules and Fc receptor.

The effect on platelet activation of monoclonal antibodies directed against common determinants of the HLA class I heavy chain molecule was studied. Cross-linking W6/32, an anti-HLA class I of IgG2a subclass, led to platelet activation. Two other antibodies of the same subclass did not have this effect on platelets. The lack of activity of the F(ab')2 fragments suggests that the activation signal is mediated by the platelet Fc receptor (Fc gamma RII). Indeed, except for a higher sensitivity of W6/32 to aspirin and apyrase, activations by cross-linking IV-3 (an anti-Fc gamma RII) and W6/32 are similar at the level of InsP3 formation, calcium mobilization, pH modifications, and activation of protein kinase C and myosin kinase. When HLA class I molecules and Fc gamma RII are cross-linked together, platelet activation occurs. This is not observed when a control IgG2a is substituted for W6/32 or when CD9 and Fc receptor are cross-linked together. This suggests that HLA class I molecules and Fc gamma RII synergize to activate platelets.     

New Approach to Eliminate HLA Class I Antigens from Platelet Surface without Cell Damage: Acid Treatment at pH 3.0

A new method was studied for eliminating HLA class I antigens from the surface of platelets without damaging the cells. Platelets were exposed to an acid solution (pH 3.0) to eliminate the antigenicity of HLA class I antigens. The reduction in antigenicities of HLA class I common antigen and individual HLA class I antigens by acid treatment was marked. Patients' sera which contained multispecific HLA antibodies reacted with PBS-treated platelets, but not with acid-treated platelets. No changes were observed in the antigenicities of glycoprotein Ib or glycoprotein IIb/IIIa. The viability of acid-treated platelets was 83%. Ultrastructural investigations revealed no significant difference between the PBS-treated platelets and acid-treated platelets. The platelet function studies showed that the aggregation of acid-treated platelets induced by various agonists was only slightly reduced compared with PBS-treated platelets. We propose that acid-treated platelets are promising for clinical use in patients refractory to platelet transfusions and may be superior to chloroquine-treated platelets for analysis of the specificity of antiplatelet antibodies.     

Identification and characterization of a functional receptor for interferon-gamma on a megakaryocytic cell line

We have previously shown that human interferon-gamma (Hu-IFN-gamma) induces platelets to become efficient effector cells, capable of killing young larvae of the parasite Schistosoma mansoni. Recently, binding sites for IFN-gamma on platelets have been characterized. We show here the presence of high-affinity receptors for IFN-gamma on the surface of the human megakaryocytic Dami cell line. Scatchard analysis indicated the presence of about 11,000 binding sites per cell, with a kd of 3 +/- 0.5 x 10(-10) mol/L; the apparent molecular weight of the receptor was 90 Kd. Receptor-bound 125I Hu-recombinant IFN-gamma was rapidly internalized and degraded when the temperature was increased from 4 degrees C to 37 degrees C. The half-life of this receptor was about 7 hours, and pretreatment of cells with IFN-gamma or phorbol myristate acetate had very little effect on the surface receptor number and no detectable effect on IFN-gamma receptor messenger RNA (mRNA) expression. The receptor was functional, because 24 hours of treatment with IFN-gamma led to the increase of HLA class I mRNA expression and to the initiation of HLA class II mRNA expression. These effects were selective because platelet glycoprotein Ib, IIb, or IIIa mRNA expression and cell proliferation were unaffected.     

Simple method for differentiating between HLA and platelet-specific antibodies by flow cytometry.

Since platelets express both platelet-specific and class I HLA antigens, serum antiplatelet reactivity assessed by most platelet antibody techniques could be due to antibodies with either or both specificities. Flow cytometric analysis of sera for detection of antiplatelet antibody commonly employs a purified platelet preparation as target cells. A method is described for investigating sera for platelet antibodies by flow cytometry using a mixture of platelets and lymphocytes. The mixture of lymphocytes and platelets as target cells has the advantage of confirming the presence of the HLA antibodies in reactive sera. The concomitant use of platelets and lymphocytes treated with citric acid, pH3, or with chloroquine (to remove or alter surface HLA antigens without affecting platelet specific antigens) may further assist in identifying antiplatelet antibodies in alloimmunized patients. These techniques may also be useful in platelet crossmatching procedures.     

A Comparative Analysis of Methods Suitable for Detection of Platelet Alloantibodies

Abstract. Three HLA-specific antisera were comparatively analyzed by platelet micro-complement fixation and ¹⁴C-serotonin uptake and/or ¹⁴C-serotonin release studies with platelets. In general, there was a good correlation of HLA antibody specificity exhibited in all 3 test systems. Quantitative differences in the response of platelets to HLA antisera were observed and accounted for by differences in antigen density and different types of antibodies involved. Platelet autoantibodies could not be detected by means of ¹⁴C-serotonin release from platelets in plasma.     

Thrombocyte HLA molecules retain nonrenewable endogenous peptides of megakaryocyte lineage and do not stimulate direct allocytotoxicity in vitro

The origin and the function of HLA class I molecules present on the surface of human platelets are still unclear. In particular, it is controversial which fraction of these class I molecules represents integral membrane components derived from the megakaryocyte-platelet lineage versus soluble plasma HLA molecules acquired by adsorption. Results of the present study show that HLA-A2 ligands isolated from platelets possess the same peptide motif as described for HLA-A2-associated peptides obtained from nucleated cells. Sequencing of these platelet-derived peptides reveals that they originate mainly from ubiquitously expressed proteins also present in the megakaryocyte-platelet lineage. Moreover, one of these peptides derives from the GPIX protein, which is specifically expressed by platelets and their precursors. Platelet HLA molecules are unstable in vitro at 37 degrees C, but can be partially stabilized by addition of exogenous beta(2)-microglobulin and HLA class I binding peptide, suggesting that platelets cannot load HLA molecules with endogenous peptides. In in vitro experiments platelets were used to stimulate peripheral blood mononuclear cells. No allospecific cytotoxicity was observed after primary stimulation, or secondary restimulation, with allogenic resting or activated platelets, even in the presence of additional third-party helper activity. These data indicate that HLA class I molecules from platelets cannot directly induce allogenic CD8(+) cytotoxic T-cell response in vitro.     

A specific assay for anti-HLA antibodies: application to platelet donor selection

Alloimmunization to donor class I HLA antigens represents a major obstacle to successful platelet transfusion therapy. It is desirable to distinguish alloimmunization from nonimmunologic causes of poor platelet survival to assess the need for HLA-matched, single-donor platelets. We describe a new in vitro assay for anti-HLA antibodies and report its application to the problem of platelet crossmatching. In contrast to previously described crossmatch techniques, the immunobead assay is specific for anti-HLA antibodies. The assay was used to evaluate 51 single-donor platelet transfusions given to seven patients from 35 different donors. Recipient plasma was assayed for antibodies directed against HLA antigens present on donor platelets. A one-hour posttransfusion corrected count increment of greater than or equal to 7,500 was considered a successful outcome. Twenty-nine of 33 (87.9%) transfusion episodes associated with a negative immunobead assay had successful outcomes. The four unsuccessful transfusions were associated with potential nonimmunologic causes of poor platelet survival. Only two of 18 (11.1%) episodes associated with a positive assay had successful outcomes. Only one unsuccessful transfusion episode was associated with a negative immunobead assay and a positive radiolabeled antiglobulin test result, which suggested that isolated alloantibodies to antigens other than class I HLA antigens are not a common cause of platelet refractoriness. Platelets stored in suspension at 4 degrees C or frozen in liquid nitrogen were found suitable for crossmatch testing.     

Sequence of exons and the flanking regions of human bilirubin-UDP-glucuronosyltransferase gene complex and identification of a genetic mutation in a patient with Crigler-Najjar syndrome, type I.

Crigler-Najjar syndrome, type I is a heterogeneous disorder that may result from mutations of various regions of the bilirubin-UDP-glucuronosyltransferase gene complex that encodes two bilirubin-UDP-glucuronosyltransferase isoforms and a phenol-UDP-glucuronosyltransferase isoform in the human liver. The two bilirubin-UDP-glucuronosyltransferase messenger RNAs and the phenol-UDP-glucuronosyltransferase messenger RNA have identical 3' regions derived from four consecutive exons. The 5' region of each messenger RNA is unique and is derived from distinct single exons. By screening a human genomic library with probes corresponding to various regions of the messenger RNAs, we have isolated five cosmid clones containing overlapping segments of this large gene complex that spans at least 84 kb of the human genome. To facilitate the amplification of each exon by polymerase chain reaction and their adjacent splice junctions, we have delineated the intron-exon boundaries of the four common region exons and the two single exons that encode the unique regions of the two bilirubin-UDP-glucuronosyltransferase isoforms and have described sequences of the regions flanking each exon. All exons encoding the two bilirubin-UDP-glucuronosyltransferase isoforms and their splice junctions were amplified from the DNA of two control subjects and a Crigler-Najjar syndrome, type I patient. The DNA from the Crigler-Najjar syndrome, type I patient revealed a point mutation in exon 3 (a common region exon) resulting in a stop codon. RNA blot showed that the two bilirubin-UDP-glucuronosyltransferase messenger RNAs in the liver of the Crigler-Najjar syndrome, type I patient were of normal length but were reduced in concentration.(ABSTRACT TRUNCATED AT 250 WORDS)     

HLA Antigens Shed from the Surface of Synthetic or Naturally Occurred Platelet-Derived Microparticles During Storage of Platelet Concentrate

The demand for standard platelet concentrates (PCs) has continued to increase in the recent years. Infusible platelet membranes (IPM) prepared from new or outdated human platelets have been developed as an alternative to standard PCs, with the additional advantage of long shelf life and increased viral safety. Reduction of HLA antigens on the IPM has been assigned as one of the probable advantages of this product. In re-examining this issue, we studied the existence of HLA class I on the surface of IPM microparticles. In comparison we also surveyed HLA expression on the surface of the naturally occurred platelet-derived microparticles (nPMPs) during 7 days storage. Intended for producing IPM, PCs obtained from Iranian blood transfusion organization were lysed; virally inactivated with wet heat in the presence of a heat stabilizer and then sonicated. IPMs were separated using centrifugation and liquid-stored in 4°C. The expression of HLA class I antigens was surveyed using flow cytometry technique. HLA molecules were present on the microparticles. Shedding of HLA antigens was demonstrated from the surface of the both liquid-stored IPM and nPMPs during storage. Storage of IPM in 4°C was accompanied with significant reduction of HLA molecules. It seemed that achievement of HLA-free IPM could be impossible unless chloroquine treated platelets were used to prepare these microvesicles.     

Severe Neonatal Alloimmune Thrombocytopenia due to Anti-HPA-3a

Background and Objectives: Neonatal alloimmune thrombocytopenia is usually attributable to HPA-la antibodies. We report a case of parental platelet antigen incompatibility associated with a severe neonatal thrombocytopenia secondary to alloimmunization to HPA-3a. Materials and Methods: Platelet antibodies were detected by the monoclonal antibody-specific immobilization of platelet antigens, genotyping of the platelets by PCR, and HLA typing by serologic procedures and PCR. Results: Genotyping of maternal and paternal platelets confirmed the incompatibility in the HPA-3a system. It is noteworthy that the mother is of the HLA type DRB3*0101, is ABO-incompatible with her husband, and also has HLA class I antibodies. Conclusion: Severe neonatal thrombocytopenia associated with HPA-3a alloimmunization is infrequent and all the factors mentioned above could have played a role in the severity of the disease.     

Platelet Antithrombins: Role of Thrombin Binding and the Release of Platelet Fibrinogen

The nature of platelet antithrombin was elucidated by comparison of thrombin binding and antithrombin activities of intact platelets and by purification of antithrombin from platelet lysates using glycerol osmotic lysis, ethanol precipitation and Sephadex gel filtration techniques. The major portion of the antithrombin and thrombin binding activity of intact platelets is lost after brief sonication. The antithrombin activity in destroyed platelets is found to be due to platelet fibrinogen. Treatment of platelets with PGE₁ (100 μg/ml) markedly inhibits (>80%) the release of platelet fibrinogen induced by thrombin. However, the PGE₁ treatment produced slight (<30%) but significant decrease of antithrombin activity of intact platelets, whereas the binding of thrombin to platelets was not affected by PGE₁ treatment. The amounts of thrombin bound to and inactivated by PGE₁-treated platelets at the same cell concentration are identical. The above results suggest that platelets contain at least two antithrombin activities. One, which accounts for the major portion of platelet antithrombin is mediated by thrombin binding to platelets. The other, which attributes to a lesser extent to platelet antithrombin activity, is due to the release of platelet fibrinogen. Also, antithrombin is readily demonstrated in a plasma medium indicating physiological significance of platelet antithrombin.     

Specific Interaction of HLA Antibodies (Eluates) with Washed Platelets

The mechanism of complement-independent action of HLA-A2 antibodies (eluates) on washed platelets was investigated. HLA-specific alteration was confirmed by serological (platelet micro-complement fixation), morphological (platelet spreading) and functional parameters (platelet aggregation, inhibition of collagen-induced platelet aggregation, [¹⁴C]serotonin release). In the presence of fibrinogen and calcium ions, HLA antibodies induced instantaneous platelet aggregation and release. Although no morphological (spreading) and functional changes (collagen-induced aggregation) were seen, these platelets did not aggregate or release when fibrinogen was subsequently added. When platelets—in the presence of fibrinogen—were incubated with antibody concentrations too low to induce platelet aggregation or release, specific reduction of platelet reactivity was observed by subsequent collagen aggregation. HLA-specific action of antibodies on washed platelets was inhibited by apyrase and acetyl-salicylic acid, indicating an active participation of platelets in HLA antibody-induced platelet alteration.     

Management of bleeding in a multi-transfused patient with positive HLA class I alloantibodies and thrombocytopenia associated with platelet dysfunction refractory to transfusion of cross-matched platelets.

Thrombocytopenia is a common condition in the critical care setting. Repetitive platelet transfusion might lead to formation of alloantibodies. HLA class I and human platelet antigen antibodies can lead to transfusion-refractory thrombocytopenia. Transfusion of cross-matched platelets often is effective in these patients. We report on the successful use of recombinant activated factor VII in an acute bleeding situation in a multi-transfused patient presenting with positive HLA class I alloantibody status and thrombocytopenia associated with platelet dysfunction refractory to even transfusion of cross-matched platelets. The 41-year-old female patient developed HLA class I antibodies during former episodes of massive transfusion. Her former medical history was empty concerning hemorrhagic events. During this specific bleeding episode the patient suffered from intractable profuse bleeding from the nasopharynx and oral cavity. Global coagulation tests were within the normal range. Platelet dysfunction was confirmed by PFA100. Initially the patient responded well to Desmopressin infusion, but after 36 h she became thrombocytopenic and refractory to even transfusion of cross-matched platelets. Recombinant activated factor VII was chosen as the last resort. Two identical boli of 160 microg/kg NovoSeven each were injected via a central line within an interval of 3 h. After the first injection bleeding was significantly reduced and vasopressor support discontinued. After the second bolus bleeding completely ceased and did not reoccur. We did not observe any side effects. The pluripotent hemostatic agent recombinant activated factor VII might be a new option in the treatment of hemorrhagic episodes in patients presenting with this rare disorder, especially when the patient is refractory to cross-matched platelets or matched platelets are not available.     

Human platelet activating antibodies.

Platelet antibodies identified in the plasma of three multiply transfused patients and a woman who had delivered a baby with neonatal alloimmune thrombocytopenia were investigated for their platelet activating properties. Three patients possessed multispecific HLA antibodies reactive with 90 to 100% of the cells on a lymphocytotoxic panel. These antibodies were also detected using the MAIPA assay and MAb w6/32, which recognizes an epitope common to all HLA class I molecules. In addition to HLA antibodies, three of the patients possessed platelet-specific antibodies that were identified by the MAIPA assay as anti-HPA-1a and anti-HPA-3a (one patient) and anti-HPA-1b (two patients). Each of the HLA antibodies when reacted with platelets expressing the corresponding HLA antigens, potently induced aggregation and release of ATP from dense granules. In contrast, the HPA-1b antibodies induced platelet agglutination, but failed to trigger ATP release. However, platelets coated with these latter antibodies were now refractory to subsequent stimulation by ADP. Similarly, when HLA antibodies were reacted with platelets to produce suboptimal activation, the platelets could now be stimulated only poorly or not at all by either epinephrine or thrombin. This was also true for anti-HPA-1b, which, although not inducing aggregation or ATP release by itself, was capable of almost completely blocking thrombin-induced platelet activation. The thrombin-inhibiting activity of these antibodies could partially be reversed by pretreating antibody-coated platelets with epinephrine immediately followed by stimulation with thrombin. These findings suggest that transfused platelets may either be activated or inhibited by reaction with various platelet antibodies. Therefore it is conceivable that the presence of platelet reactive antibodies in multiply transfused recipients may contribute to the increased thrombotic and hemorrhagic symptoms often observed among these patients.     

Acidified Chloroquine Treatment for the Removal of Class I HLA Antigens

With the aim of reducing the damage to platelets while effectively removing class I HLA antigens from their surfaces, we developed a new method using acidified chloroquine diphosphate. Platelets were treated with a 0.2 M solution of chloroquine diphosphate (pH 4.0). More than 90% of the platelets remained viable after treatment. While a marked reduction in reactions of acidified chloroquine-treated platelets with multispecific HLA antisera was noted in comparison with phosphate-buffered-saline-(PBS)-treated platelets, reactions with platelet-specific antibodies were preserved. This was demonstrated by immunofluorescence tests and solid-phase and monoclonal antibody immobilization of platelet antigen assays. Aggregation responses, though reduced in comparison with PBS-treated platelets, were still preserved after acidified chloroquine treatment. Ultrastructural analysis did not show any significant difference from PBS-treated platelets. We conclude that treatment of platelets with acidified chloroquine diphosphate is a simple and effective method for removing class I HLA antigens from their surfaces with minimal damage to their structure and function.