Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Apr 5.
Published in final edited form as: Transfusion. 2020 Aug 8;60(9):2097–2107. doi: 10.1111/trf.15960

Naturally Occurring Point Mutation Cys460Trp Located in the I-EGF1 Domain of Integrin β3 Alters the Binding of Some anti-HPA-1a Antibodies

Sarah Theresa Holzwarth 1,*, Behnaz Bayat 1,*, Jieqing Zhu 2, Roongaroon Phuangtham 1,3, Lars Fischer 4, Doris Boeckelmann 5, Lida Roeder 1, Heike Berghöfer 1, Silke Schmidt 1, Gregor Bein 1, Sentot Santoso 1
PMCID: PMC8020566  NIHMSID: NIHMS1678835  PMID: 32770549

Abstract

Background:

Fetal and neonatal alloimmune thrombocytopenia (FNAIT) is caused by the destruction of platelets in the fetus or newborn by maternal platelet alloantibodies (Abs), mostly against human platelet antigen (HPA)-1a. Recent studies indicate that two anti-HPA-subtypes exist: type I reacts with epitopes residing on the plexin-semaphorin-integrin (PSI) and type II with PSI/integrin epidermal growth factor 1 (I-EGF1) domains of the β3 integrin. Here, we evaluated whether a Cys460Trp mutation in the I-EGF1 domain found in a Glanzmann thrombasthenia patient can alter the binding of anti-HPA-1a Abs.

Methods:

Stable HEK293 cell lines expressing wild-type and mutant αIIbβ3 and αvβ3 were generated to prove the reactivity of different antibodies against HPA-1a.

Results:

Flow cytometry analysis of wild-type (Cys460) and mutant (Trp460) expressed on HEK293 cells showed equal surface expression of αIIbβ3 and αvβ3. When tested with mutant αIIbβ3 cells, reduced binding was observed in type II but not in type I anti-HPA-1a Abs. These results could be confirmed with platelets carrying Cys460Trp mutation. Interestingly, reduced binding of type I Abs was detected with mutant αvβ3 cells. Both Ab types were found in maternal sera from FNAIT cases by an antigen-capture assay using HEK293 transfected cells.

Conclusions:

These observations confirm the existence of type I and type II anti-HPA-1a Abs. Furthermore, this study underlines different immunogenicity of HPA-1a antigen(s) residing on either αIIbβ3 or αvβ3. Further analysis of FNAIT cases from mothers having a fetus with and without intracranial bleedings using such an approach may highlight the functional relevance of different anti-HPA-1a subtypes.

Introduction

Fetal and neonatal alloimmune thrombocytopenia (FNAIT) is a serious bleeding condition that occurs as a result of transplacentally transported maternal alloantibodies (Abs) reacting with human platelet antigen (HPA) expressed on fetal platelets. In Caucasian populations, FNAIT is mostly caused by anti-HPA-1a Abs and has an incidence of 10/100,000 live births in the European Union (1). The clinical presentation of FNAIT varies from asymptomatic thrombocytopenia to severe clinical complications. The most serious one is intracranial hemorrhage (ICH) occurring in around 20% of severe FNAIT cases (2–4) and leading to fetal death or persistent neurological sequelae in neonates (5). In the last decade, the relationship between anti-HPA-1a quantity and FNAIT severity has been studied extensively (6–8). However, this issue remains a subject of debate (9). Recent data indicated that anti-HPA-1a Abs are heterogeneous with respect to their binding sites, indicating that a different subtype of anti-HPA-1a Abs could also contribute to the clinical severity of FNAIT (10).

HPA-1 alloantigen system is formed by a single amino acid, Leu33Pro, substitution located in the flexible plexin-semaphorin-integrin (PSI) domain of the integrin β3 chain (11). On the cell surface, however, the β3 chain can form heterodimers with either of the αIIb or αv subunits, which function as fibrinogen or vitronectin receptors, respectively. The integrin αIIbβ3 complex is exclusively expressed on platelets, whereas αvβ3 is expressed on a variety of cell types, including endothelial cells and trophoblasts (12, 13).

Recently, we found that immunized mothers can develop three different anti-HPA-1a Abs subtypes, reactive with either HPA-1a epitopes expressed solely on the common β3 chain independently of the α subunits (neither αIIb nor αv) or anti-HPA-1a Ab subtypes that recognize compound epitopes formed by the β3 chain together with the αIIb or αv chain. In mothers who had a foetus/child with FNAIT-associated ICH, we found anti-HPA-1a Abs that reacted specifically with αvβ3 compound epitopes (10)

Recent high-resolution mapping of HPA-1a antigen using transgenic mice harboring humanized residues within the PSI and integrin epidermal growth factor-1 (I-EGF1) domains showed that some anti-HPA-1a Abs recognized antigenic determinants solely express on the polymorphic PSI domain (termed as Type I) and some anti-HPA-1a Abs require additionally specific amino acid residues within the I-EGF1 domain (termed as Type II (14). Besides the polymorphic amino acid 33, Type I anti-HPA-1a Abs (such as monoclonal Ab [mAb] SZ21) require three additional residues located in positions 30, 32, and 39 of the PSI domain (APL33D residues). In contrast, Type II anti-HPA-1a Abs, such as mAbs B2G1 and 26.4, need additionally the I-EGF1 domain, especially the residues H446 and Q470, respectively (14). Furthermore, recent crystallographic study of monomeric β3 showed a different conformation of PSI and I-EGF1 domains in the extended (activated) and bent (resting) states. Interestingly, Type II anti-HPA-1a Abs could stabilize the bent conformation of integrin β3 and thus exert functional inhibition of integrins αIIbβ3 and αvβ3, which may contribute to the severity of bleeding and pathogenesis of ICH (15).

Glanzmann thrombasthenia (GT) is a rare inherited bleeding disorder caused by platelet function defect. The hallmark of this disease is severely reduced platelet aggregation in response to platelet agonists due to the lack of platelet αIIbβ3 expression or function (16). Recently, we found a compound heterozygous Cys460>Trp mutation located in the I-EGF1 domain of integrin β3 in a child with severe GT (17). In this study, we aimed to evaluate the binding of different anti-HPA-1a Abs using HEK293 cells expressing wild-type and mutant αIIbβ3 and αvβ3 integrin.

Case Report:

A 3-month-old male neonate (gestational age 41 + 1 weeks) was referred with spontaneous petechial bleeding (284×109/L) to the skin and bilateral hematoma after intramuscular vaccination in both legs. Analysis of platelet function analyzer-100 showed prolonged bleeding time, and platelet aggregation was impaired in response to all stimuli including adenine diphosphate (2%), collagen (24%), arachidonic acid (13%), and epinephrine (13%). In contrast, platelet aggregation to ristocetine was almost normal (67%; lower limit of normal is 70%) leading to the diagnosis of GT. Furthermore, reduced αIIbβ3 expression was found on the surface of neonate platelets leading to the diagnosis of GT. Molecular analysis revealed a compound heterozygous missense mutation in the ITGB3 gene (see Results). At the age of 11 months, the patient presented with mucosal bleeding, which could be stopped by platelet and red blood cell transfusions and not by tranexamic acid and recombinant activated factor VII administration.

Material and Methods

MAbs against αIIbβ3 complex (clone Gi5) was produced and characterized in our laboratory (18). MAbs SZ22 and SZ21 specific for αIIb and β3 subunits, respectively, were purchased from Beckman Coulter (Marseille, France). Hybridoma producing mAb AP3 directed against the β3 subunit was purchased from American Type Culture Collection (ATCC HB-242) and cultured in our laboratory. MAb against αvβ3 complex (clone 23C6) (19) was purchased from Millipore (Temecula, CA, USA). MAb against glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was from Halcyon (South Logan, UT, USA). Human mAbs specific for HPA-1a 26.4 and B2G1 were kindly provided by Dr. Bjørn Skogen (Prophylix Pharma AS, Tromsø, Norway) and Dr. Celdric Ghevaert (NHS Blood and Transplant, Cambridge, UK), respectively.

Anti-HPA-1a sera were collected from mothers with FNAIT without intracranial bleeding. Serum from a healthy blood donor with blood group AB was used as control. Standards of anti-HPA-1a Abs were purchased from the National Institute for Biological Standards and Control (NIBSC 03/152, Potters Bar, Hertfordshire, UK).

Quantification of αIIbβ3 surface expression by flow cytometry

Platelet-rich plasma was prepared from acid citrate dextrose anticoagulated blood by centrifugation at 800 g for 20 min at room temperature. The surface expression αIIbβ3 on platelets was quantified with different mAbs using calibrated bead suspension (Biocytek, Marseille, France) according to the manufacturer’s instructions on a FACSCalibur flow cytometer (Becton and Dickinson, Heidelberg, Germany).

Immunoblotting analysis of platelets

Immunoblotting with platelet lysates was performed as previously described (20). Platelet lysate (200 μg/mL protein) was separated on 7.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) under nonreducing conditions and transferred onto polyvinylidene fluoride (PVDF) membranes. After blocking with Tris buffer containing 1.5% bovine serum albumin (BSA) for 60 min at room temperature, membranes were stained overnight with mAbs (200 μg/mL) in blocking solution buffer and visualized by the use of peroxidase-labeled donkey anti-mouse immunoglobulin G (IgG; Dianova, Hamburg, Germany; dilution 1:50,000) and chemiluminescence system (ECL Plus, Thermo Scientific, Rockford, IL, USA). Molecular weight standard (Thermo Fisher) was run in parallel. Chemiluminescence was analyzed on FlourChem2 Imaging System (Alpha Innotec, Kasendorf, Germany).

Nucleotide sequencing

Full-length sequencing of ITG2B and ITGB3 was carried out as described previously (21). Briefly, αIIb and β3 coding regions of genomic DNA were polymerase chain reaction (PCR) amplified with primers corresponding to intronic sequences surrounding all exons of the ITG2B and ITGB3 genes. PCR was carried out using a FastStart High Fidelity PCR System (Roche Diagnostic Corporation, Indianapolis, IN, USA). Before sequence analysis, all PCR products were purified with a QIAquick PCR purification kit (Qiagen Sciences, Valencia, CA, USA). Automated sequence analysis was performed in both directions on a genetic analyzer (ABI 3100; Applied Biosystems, Foster City, CA, USA).

Production of mutant β3 construct by site-directed mutagenesis

A full-length construct for the β3 mutant isoform (Trp460) was produced by site-directed mutagenesis of wild-type β3 in pcDNA3.1/V5-His-TOPO/Geneticin (Invitrogen, Carlsbad, CA, USA) mammalian vector using Quick Change II Kit (Agilent Technologies, Santa Clara, USA) as previously described (21). Site-directed mutagenesis primers 5′-ATCTGTGGAGCATCCGGAACCTGGGTACCAA-3′ and 5′TTGGTACCCAGGTTC CGG ATGCTCCACAGAT-3′ were used for PCR amplification (Eurofins MWG Operon, Ebersberg, Germany). After denaturation for 30 s at 95°C, amplification consisted of 12 cycles (denaturation at 95°C for 30 s, annealing at 55°C for 1 min, and extension at 65°C for 12 min). PCR products were digested with Dpn endonuclease for 1 h at 37°C and transfected into DH5α high-efficiency competent Escherichia coli (Life Technologies, Carlsbad, CA, USA). Plasmid DNA from positive clones was verified by nucleotide sequencing as described earlier.

Generation of stable transfected HEK293F expressing mutant αIIbβ3 and αvβ3 integrin

HEK293F cells were grown in Dulbecco’s modified eagle medium (Capricorn Scientific, Ebsdorfergrund, Germany) supplemented with 10% fetal calf serum and 1% penicillin/streptomycin (PAN-Biotech, Aidenbach, Germany) and were transfected with β3 (wild-type or mutant) together with full-length wild-type αIIb or αv in pcDNA4/HisMax©-TOPO® Zeocin (Invitrogen) vector using jetOptimus Transfection reagent (Polyplus, Illkirch, France). After 4 weeks of cell culture in the presence of Geneticin (400 μg/mL) and Zeocin (200 μg/mL), cell lines were analyzed for αIIbβ3 and αvβ3 surface expression by flow cytometry. Stable transfectants expressing high αIIbβ3 and αvβ3 densities on the cell surface were sorted with Alexa Flour 488-labeled mAb Gi5 (anti-αIIbβ3 complex) using a BD FACSAria III with FACSDiva 6.1.3 software (Becton and Dickinson).

Flow cytometry analysis of platelets and HEK293F transfected cells

Platelets were isolated from ethylenediaminetetraacetic acid anticoagulated blood by differential centrifugation as previously described. Aliquots of 20 × 106 washed platelets were incubated with 0.4 μg mAbs in 50 μL phosphate-buffered saline (PBS) containing 0.2% BSA (PBS-BSA) for 30 min at 4°C. After washings with 300 μL washing buffer (6,200 g, 1 min), platelets were stained with fluorescein isothiocyanate (FITC)-labeled goat anti-human IgG (Dianova; dilution 1:50) for 30 min at 4°C. Labeled cells were washed, resuspended in 200 μL CellFIX (Becton and Dickinson), and analyzed by flow cytometry using FACSCanto II and FACSDiva (Becton and Dickinson) and Flowing 2.5.1 Software (Perttu Terho, Turku Bioimaging).

Adherent HEK293F transfected cells were detached using Accutase and were resuspended in Dulbecco PBS (Thermo Fisher Scientific, Langenselbold, Germany) containing 0.2% BSA (Serva, Heidelberg, Germany; PBS-BSA). Aliquots of 3 × 105 cells were incubated with 20 μL mAb (20 μg/mL) for 30 min at 4°C. After washing with PBS-BSA (6,200 g, 1min), cells were labeled with Alexa Flour 488 conjugated donkey anti-mIgG (Invitrogen; dilution 1:50) or FITC-labeled goat anti-human IgG (Dianova; dilution 1:50) for 30 min at 4°C. Labeled cells were analyzed by flow cytometry as described above.

Analysis of anti-HPA-1a abs using HEK293 transfected cells by an antigen-capture assay

Aliquots of 100 μL HEK293 transfected cells (3 × 105 cells) were incubated with 50 μL serum or mAb against HPA-1a (20 μg/mL) and 10 μL capture mAb AP3 (anti-β3; 20 μg/mL) for 30 min at 37°C. After solubilization with 100 μL lysis buffer (10 mM Tris buffer in isotonic saline, pH 7.4) containing 0.5% Triton-X100 (Sigma, Munich, Germany) for 30 min at 4°C, cell lysates were centrifuged by 16,200 g for 30 min. Aliquots of 70 μL cell lysates were analyzed by mAb immobilization of platelet antigens as previously described (22). The cutoff was calculated by the use of control sera derived from healthy blood donors. All experiments were run in duplicates, performed twice and the results were given as arithmetic means of optical density. Mock-transfected HEK293 cells were run as a control.

Statistics

Statistical comparisons were made using an unpaired, two-tailed Student’s t-test. A p-value of <0.05 was assumed to represent statistical significance.

Results

Molecular analysis of αIIb (ITG2B) and β3 (ITGB3) genes

Recently, we described the molecular genetic characterization of several patients with GT (17). One of these patients (GT2) was found to carry two compound heterozygous novel missense mutations in ITGB3 (NM_000212.2) gene: one mutation c.31T>C located in exon 1 and one mutation c.1458C>G in exon 10 that lead to amino acid substitution Trp11Arg and Cys486Trp (mature Cys460Trp), respectively (17). The first mutation Trp11Arg located in the signal peptide may interfere with the translocation of β3 protein to or through cell membranes leading to the absence of αIIbβ3 on the platelet surface (23). Meanwhile, a GT type 1 patient (<5% αIIbβ3 expression) carrying c.31T>C mutation in a homozygous state was reported (24). The second mutation, Cys460Trp in the mature protein, is located in the I-EGF1 domain, which is known to harbor distinct amino acids responsible for the formation of HPA-1a epitopes (14, 15). The Cys460Trp mutation may hamper some anti-HPA-1a Ab binding and therefore was selected for this study.

Figure 1A shows the nucleotide sequencing analysis of the patient’s ITGB3 gene. Heterozygous C>G transition at position 1458 located in exon 10 leading to missense substitution Cys460Trp (TGC>TGG) in the I-EGF1 domain of integrin β3 was found. In accordance with the pedigree, this mutation was also detected in the father (data not shown). Furthermore, analysis of nucleotide 176 in exon 3 of the GT patient displayed a homozygous C encoding for Leu33, which is known to be responsible for the formation of the HPA-1aa phenotype (Figure 1A). Analysis of the parent’s DNA revealed homozygous HPA-1aa father and heterozygous HPA-1ab mother indicating that both HPA-1a alleles inherited to the GT son are defective carrying either Cys460Trp or Trp11Arg mutation. This result is illustrated in Figure 1B. Flow cytometry analysis with mAb AP3 against β3 integrin subunit showed reduced αIIbβ3 surface expression on GT platelets when compared with normal platelets (Figure 1C). In contrast, paternal platelets expressed normal αIIbβ3. Platelet glycoprotein quantification using calibration beads by flow cytometry revealed 11,662 and 41,583 copies of αIIbβ3 on the platelet of GT patient and his father, respectively. In the control experiments, platelet from normal healthy blood donors expressed 41–65,000 copies of αIIbβ3.

Figure 1.

Figure 1

Figure 1

Figure 1

Open in a new tab

(A): Nucleotide sequencing analysis of patient’s ITGB3. Partial nucleotides located in exon 10 (left panel) and exon 3 (right panel) are shown. The nucleotide substitution C>G at position 1458 in exon 10 and the corresponding amino acid change from C460 (TGC) to W460 (TGG) are indicated. The positions of three disulfide bonds C437-C457, C448-C460, and C462-C471 are presented. The amino acid L33 encoded by CT176G located in exon 3 corresponding to homozygous human platelet antigen-1aa is shown. (B) The pedigree of the index family. The human platelet antigen (HPA)-1a phenotypes and the mutations on the ITGB3 gene of the father, mother, and son (Glanzmann thrombasthenia patient) are indicated. The suspected HPA-1a alleles carrying Cys460Trp and Trp11Arg of the son inherited from the father and mother, respectively, are shown. (C) Flow cytometry analysis of patient’s platelets. Washed platelets from a healthy blood donor (ND), Glanzmann thrombasthenia (GT) patient, and his father (FA) were incubated with monoclonal antibody (mAb) AP3 against β3. After washings, platelets were stained with fluorescein isothiocyanate-labeled secondary Ab and analyzed by flow cytometry. Ctr: isotype control.

These results showed that platelets carrying αIIbβ3 carrying Cys460Trp mutation can express mutant αIIbβ3 on the cell surface. Therefore, our patient can be classified as GT type 2 (25).

Effect of the Cys460Trp mutation on the expression of integrin β3

Full-length cDNA constructs encoding for the mutant β3 form were generated to further prove the effect of Cys460Trp mutation on integrin β3 surface expression. This mutant β3 construct was transfected together with wild-type αIIb or wild-type αv into HEK293 cells, and the expression of αIIbβ3 and αvβ3 on the surface of stable transfected cells was analyzed by flow cytometry using mAbs that reacted against β3 (mAb AP3), αIIbβ3 (mAb Gi5), and αvβ3 (mAb 23C6) complex. Similar to wild-type, both mutant-transfected cells expressed a comparable amount of αIIbβ3 and αvβ3 heterodimers on the cell surface (Figure 2A), confirming that Cys460Trp mutation did not significantly impair the surface expression of αIIbβ3 and αvβ3 integrin. In accordance with the previous result, immunoblotting analysis (Figure 2B) showed that the mutant β3 migrated slightly slower in comparison with wild-type β3 isoform, indicating that the disruption of one disulfide bond due to Cys460Trp mutation alters the secondary structure of integrin αIIbβ3. Similar results were found with αvβ3 transfected cells (data not shown).

Figure 2:

Figure 2:

Figure 2:

Open in a new tab

(A) Flow cytometry analysis of HEK293 cells transfected with wild-type and mutant αIIbβ3. HEK293 cells were transfected with either wild-type β3 (C460) or mutant β3 (W460) together with wild-type αIIb constructs. After selecting and sorting, stable transfected cells were analyzed with monoclonal antibodies (mAbs) Gi5 against αIIbβ3 complex and AP3 against β3 subunit. Mouse IgG (mIgG) was run in parallel as a control. Bound Abs were stained with fluorescent-labeled secondary Ab and analyzed by flow cytometry. (B) Immunoblotting analysis of HEK293 cells transfected with wild-type and mutant β3. Stable transfected HEK293 cells expressing mutant αIIbβ3 (lanes 1) or wild-type (lanes 2) were lysed. Cell lysates were separated on 7.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis under nonreducing conditions and transferred onto a polyvinylidene fluoride membrane. The membrane was then incubated with monoclonal antibody (mAb) AP3 against β3 or mouse IgG (mIgG) as control. After washings, bands were then visualized using a peroxidase-labeled secondary Ab and a chemiluminescence system. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was stained as an internal control. Note the molecular weight difference between wild-type and mutant β3.

Effect of the Cys460Trp mutation on HPA-1a antigenic determinants

Figure 3 shows the three-dimensional structure analysis of the PSI domain harboring polymorphic residue 33 (Leu or Pro) and the I-EGF1 domain, which can also control the formation of HPA-1a or HPA-1b alloform. Theoretically, the Cys460Trp mutation may disturb the conformation of I-EGF1 itself and the adjacent PSI domain, leading to the destruction of HPA-1a epitope(s).

Figure 3:

Figure 3:

Open in a new tab

Possible steric consequence of the β3-C460W mutation on the HPA-1a epitope Top view of the HPA-1a epitope composed of plexin-semaphorin-integrin and integrin epidermal growth factor 1 domains. The critical residues involved in anti-HPA-1a binding are shown as blue sticks. Disulfide bonds are gray sticks. The crystal structure of wild-type β3 is shown on the left. The predicted steric consequence of the C460W mutation was made in-silico using PyMol and shown on the right. The C460W mutation may directly disrupt the HPA-1a epitope through its bulky side chain or indirectly disturb the integrin epidermal growth factor 1 conformation because of the disruption of the C448-C460 disulfide bond.

To prove the possible effect of this mutation on HPA-1a epitopes, three different mAbs specific for HPA-1a (SZ21, 26.4, and B2G1) were tested with platelets from the GT patient, his father, and a normal donor by flow cytometry (Figure 4). Interestingly, only mAb SZ21, but not mAbs 26.4, and B2G1, showed positive reactions to GT platelets. In contrast, paternal platelets reacted with all three mAbs. However, reduced binding was obtained with mAbs 26.4 and B2G1 but not with mAb SZ21 when compared with normal platelets. These results indicated that Csy460Trp mutation alters the binding of some mAbs against HPA-1a.

Figure 4:

Figure 4:

Open in a new tab

Analysis of monoclonal antibodies (mAbs) against human platelet antigen-1a with platelets by flow cytometry. Washed platelets from a healthy blood donor (ND), Glanzmann thrombasthenia patient, and his father (FA) were incubated with mAbs SZ21, 26.4, and B2G1 as indicated. After washings, platelets were stained with fluorescein isothiocyanate-labeled secondary Abs and analyzed by flow cytometry. Ctr: isotype control.

To confirm these findings, wild-type and mutant αIIbβ3 transfected HEK293 cells were generated and analyzed by flow cytometry. As shown in Figure 5 (top panel), mAb SZ21 reacted comparably strong with both wild-type and mutant αIIbβ3 cells, whereas the other two mAbs, 26.4 and B2G1, showed significantly decreased reactivity with mutant αIIbβ3 cells. Of note, the binding of mAb B2G1 to mutant αIIbβ3 cells was completely abolished. Accordingly, the Cys460Trp mutation affects predominantly the binding type II (such as 26.4 and B2G1) rather than type I anti-HPA-1a Abs (such as mAb SZ21).

Figure 5:

Figure 5:

Open in a new tab

Flow cytometry analysis of transfected HEK293 cells with anti-human platelet antigen-1a antibodies (Abs). Stable transfected HEK293 cells expressing wild-type αIIbβ3 or mutant αIIbβ3 (top panel) and wild-type αvβ3 or mutant αvβ3 (bottom panel) were incubated with monoclonal Abs SZ21, 26.4, and B2G1 as indicated. After washings, bound Abs were incubated with fluorescent-labeled secondary Abs and analyzed by flow cytometry.

Recent structural analysis of the PSI domain of αIIbβ3 and αvβ3 showed distinct conformational states, which could, in theory, be differentially recognized by the immune system (15). It is therefore conceivable that the binding of type I and type II anti-HPA-1a Abs onto αIIbβ3 and αvβ3 integrin may differ. To prove this hypothesis, anti-HPA-1a mAbs were tested with wild-type and mutant αvβ3 transfected HEK293 cells (Figure 5, bottom panel). In comparison with αIIbβ3, the binding of all anti-HPA-1a mAbs to mutant αvβ3 cells was significantly affected. Interestingly, mAb SZ21 that similarly bound to mutant αIIbβ3 as wild-type αIIbβ3 cells reacted only weakly with mutant αIIbβ3 cells. Thus, it seems as some type I Abs can react as type II Abs when β3 formed a complex with αv subunit.

The Cys460Trp mutation affected the binding of some anti-HPA-1a Abs

To further confirm the existence of these two subtypes of anti-HPA-1a Abs, sera from mothers with FNAIT were tested by an antigen-capture assay using wild-type and mutant cells as targets (Figure 7). Recently, we could demonstrate that anti-HPA-1a standard serum (NIBSC 03/152) reacted with the β3 subunit independently from the α subunits (26). Accordingly, NIBSC 03/152 (#1) represents as type I Abs and showed therefore similar reaction with wild-type and mutant αIIbβ3 transfected HEK293 cells. In contrast, 4/5 FNAIT sera showed significantly diminished reactivity with mutant cells indicating that most sera (#3, #4, #6, and #7) contained predominantly Type II anti-HPA-1a Abs. Only one serum (#5) consisted of Type I anti-HPA-1a Abs. Similar results were obtained with mutant αvβ3 transfected HEK293 cells. This observation indicates that most anti-HPA-1a sera contain Abs recognize epitopes formed by the PSI and I-EGF1 domains (type II).

Discussion

Although the HPA-1a alloantigen determinant is known to be controlled by Leu33Pro dimorphism located in the PSI domain of the integrin β3 (11), previous studies demonstrated that the I-EGF1 domain of αIIbβ3 also participates in the formation of HPA-1a epitopes. Some anti-HPA-1a Abs reacted with PSI domain alone (termed type I), but some required both PSI and I-EGF1 domains (termed type II) (27, 28, 14). However, the clinical relevance of these Ab subtypes is currently unclear (29).

In this study, we sought to prove the existence of type I and type II Ab subtypes by using HEK293-transfected cells expressing naturally occurring mutation Cys460Trp located in the I-EGF1 domain found in a severe GT patient (17).

Our first expression studies on platelets of the patient and his father using mAb AP3 reacted with monomorphic epitopes residing on residues 50 and 62 on the integrin β3 (30, 31) showed that the father, carrier of Cys460Trp mutation, still expressed normal amount of αIIbβ3 complex on the platelet surface. The compound heterozygous GT, carrier of Cys460Trp and Trp11Arg showed reduced αIIbβ3 expression, attributable to the fact that Trp11Arg mutation alter αIIbβ3 expression (23).

In accordance, our flow cytometry analysis showed that mutant-transfected HEK293 cells expressed a comparable amount of αIIbβ3 on the cell surface as wild-type cells. A similar result was obtained with αvβ3 transfected HEK293 cells indicating that the disruption of non-conserved C448-C460 disulfide bond did not alter αIIbβ3 and αvβ3 surface expression. Three disulfide bonds (C437-C457, C448-C460, and C462-C471) are known to participate in the three-dimensional structure of the I-EGF1 domain. The first disulfide bond C437-C457 is unique for integrin, whereas the two other disulfide bonds C448-C460 and C462-C471 are conserved in other EGF domains (32, 33). The crystal structures of either the full-length ectodomain (34) or the αIIbβ3 headpiece confirmed the disulfide bonds C448-C460 and C462-C471 (35).

The disruption of the unique C437-C457 disulfide bond resulted in reduced αIIbβ3 and αIIbβ3 surface expression but constitutively active receptors (36, 37). However, little is known about the structural and functional relevance of two non-conserved C448-C460 and C462-C471 disulfide bonds. Our study here indicated that the non-conserved C448-C460 does not seem to alter αIIbβ3 and αvβ3 surface expression. A previous study has documented the importance of Cys460 on the activation status of αIIbβ3 integrin (38). The question of whether this mutation can alter αIIbβ3 and αvβ3 integrin function leading to constitutively active receptors or not is intriguing and is subject to further study.

Interestingly, different results were observed when mAbs against HPA-1a were tested. Although mAb SZ21 (type I Abs) reacted equally with wild-type and mutant αIIbβ3 transfected HEK293 cells, type II Abs (mAbs 26.4 and B2G1) showed significantly reduced binding to the mutant-transfected HEK293 cells.

Recently, Zhi and coworkers elegantly demonstrated the use of transgenic mice approach in not only the polymorphic amino acid 33 in the PSI domain but also some residues in the I-EGF1 domain that are important for the formation of HPA-1a epitopes recognized by type II Abs (14). Although mAb 26.4 requires amino acid residue Q470 in the I-EGF1 domain for its binding, mAb B2G1 needs amino acid residues Q470 and H446. Accordingly, Cys460Trp mutation more significantly affects the binding of mAb B2G1 rather than mAb 26.4. Nonetheless, the results strongly indicated that conformational change(s) of the PSI/I-EGF1 interface can alter the binding of type II Abs. Of note, all anti-HPA-1a mAbs studied here reacted with the common monomeric β3 independently from the α subunits (26).

It is known that HPA-1a is expressed not only on αIIbβ3 but also on αvβ3 integrin, which is found on platelets and endothelial cells (39). Here, we showed, in contrast to αIIbβ3, that the Cys460Trp mutation on αvβ3 integrin reduced the reactivity of type II Abs as well as the binding of type I Abs (mAb SZ21). This observation suggested a determining effect of partner protein (αIIb or αv) on β3 confirmations. Theoretically, this fact can lead to the formation of different HPA-1a epitopes indicating different immunogenicity of αIIbβ3 and αvβ3 complexes, which may lead to different Ab profiles of HPA-1a immunized individuals. In accordance, we recently found a specific subtype of anti-HPA-1a Abs that reacted with compound epitopes formed by αv and β3 subunits. This Ab subtype can cause endothelial dysfunction and was frequently detected in severe FNAIT cases with ICH (10). More recently, we found that all mAbs against HPA-1a tested here (type I and type II) reacted with the β3 chain alone, independently from the α subunits. These mAbs were not capable to cause endothelial dysfunction as Abs against αvβ3 complex did (10, 26). In contrast, type I and type II Abs may have distinct effects on platelet functions (14).

Next, we tested the applicability of our transfected HEK293 cells to differentiate type I and type II anti-HPA-1a Abs in the serum of HPA-1a immunized mothers with FNAIT by an antigen-capture assay. Surprisingly, anti-HPA-1a WHO standard containing polyclonal Abs behaved as type I Ab, reacting with mutant αIIbβ3 and mutant αvβ3. By this approach, we found predominantly type II Abs in our FNAIT cohort. Only 1/5 FNAIT serum contained type I Abs. It is tempting to speculate that type II and type I Abs represent successive affinity maturation during the course of the immune response, where type I Abs represent greater affinity for the Leu33 HPA-1a epitope on β3 chain alone.

Taking together, these results underline the heterogeneous immune response against HPA-1a alloantigen. Anti-HPA-1a immunized mothers can develop Abs that react with different epitopes formed by the polymorphic amino acid Leu33 residing in the PSI domain alone (type I) or together with other topographically closer regions (such as I-EGF1 domain) of the β3 (type II) or with αIIb or αv integrin subunits (compound epitopes) leading to the production of different anti-HPA-1a Ab subtypes. These different subtypes may bind to platelets and/or endothelial cells by different affinities and may thereby trigger a different cellular response(s), leading to different clinical pictures. Since the composition and the concentration of different Ab subtypes may differ from one maternal serum to the other, this phenomenon may explain the inconsistent association between anti-HPA-1a Abs titer and clinical outcome of FNAIT (8, 40).

However, further studies are necessary to understand the impact of different anti-HPA-1a Ab subtypes on the development of severe bleeding. Characterizing the anti-HPA-1a profile in immunized mothers involved in FNAIT cases will help us to improve our laboratory diagnostic methods and to predict the severity of FNAIT and design new prophylactic strategies to prevent fatal bleeding complications.

Figure 6:

Figure 6:

Open in a new tab

Analysis of maternal anti-human platelet antigen (HPA)-1a antibodies (Abs) with wild-type and mutant αIIbβ3 and αvβ3 transfected HEK293 cells by an antigen-capture assay. Wild-type (WT) and mutant (MU) transfected αIIbβ3 and αvβ3 HEK293 cells were incubated with anti-HPA-1a standard serum (National Institute for Biological Standards and Control 03/152; columns 1), AB serum from a healthy blood donor as control (columns 2), or five different anti-HPA-1a serums from fetal and neonatal alloimmune thrombocytopenia cases without intracranial hemorrhage (columns 3–7), together with monoclonal Ab (mAb) AP3 as capture Abs. Non-transfected HEK293 cells (MOCK) were used as negative control. After lysis, rabbit anti-mouse immunoglobulin G (IgG) coated onto microtiter wells immobilized the trimolecular complex comprising mAb AP3, αIIbβ3, or αvβ3 antigen and anti-HPA-1a. Anti-HPA-1a Abs bound to WT or MU αIIbβ3 and αvβ3 were detected with enzyme-labeled goat anti-mouse IgG and analyzed by spectrophotometry (optical density at 490/620 nm). Data are given as mean from duplicates + 1SD.

What is known on this topic?

  • Anti-human platelet antigen (HPA)-1a antibodies are heterogeneous with respect to their epitopes.

  • Anti-HPA-1a antibody subtype reactive with compound epitopes formed on integrin αvβ3 can trigger endothelial dysfunction leading to intracranial hemorrhage.

  • Recently, two other anti-HPA-1a antibody subtypes have been identified: one that reacts with the plexin-semaphorin-integrin (PSI) domain (Type I) and one that reacts with epitopes formed by PSI and integrin epidermal growth factor 1 domains (Type II).

What this paper adds

  • Using naturally occurring point mutation Cys460Trp located in the integrin epidermal growth factor 1 domain found in a Glanzmann thrombasthenia patient, we could confirm the existence of Type I and Type II anti-HPA-1a antibodies using HEK293 transfected cells

  • Analysis of anti-human platelet antigen (HPA)-1a sera with these transfected HEK293 cells may highlight the functional relevance of different types of anti-HPA-1a antibodies.

Source of support

This work was supported by the German Society for Transfusion Medicine (DGTI to SS) and by the JLU TRAINEE Program of the Faculty of Medicine at Justus Liebig University Giessen (to STH). J.Z. was supported by Grant HL131836 from the Heart, Lung, and Blood Institute of the National Institutes of Health.

Footnotes

Conflict of Interest

No conflict of interest to declare.

References:

  • 1.Kamphuis MM, Paridaans NP, Porcelijn L, Lopriore E, Ospkes D. Incidence and consequences of neonatal alloimmune thrombocytopenia: a systematic review. Pediatrics 2014; 133:715–21. [DOI] [PubMed] [Google Scholar]
  • 2.Mueller-Eckhardt C, Grubert A, Weisheit M, et al. 348 cases of suspected neonatal alloimmune thrombocytopenia. Lancet 1989; 333: 363–6. [DOI] [PubMed] [Google Scholar]
  • 3.Kaplan C, Morel-Kopp M, Kroll H, et al. HPA-5b (Br(a)) neonatal alloimmune thrombocytopenia: clinical and immunological analysis of 39 cases. Br J Haematol 1991; 78: 425–9. [DOI] [PubMed] [Google Scholar]
  • 4.Bussel JB, Zacharoulis S, Kramer K, et al. Clinical and diagnostic comparison of neonatal alloimmune thrombocytopenia to non-immune cases of thrombocytopenia. Pediatr Blood Cancer 2005; 45: 176–83. [DOI] [PubMed] [Google Scholar]
  • 5.McQuilten ZK, Wood EM, Savoia H, et al. A review of pathophysiology and current treatment for neonatal alloimmune thrombocytopenia (NAIT) and introducing the Australian NAIT registry. Aust New Zeal J Obstet Gynaecol 2011; 51: 191–8. [DOI] [PubMed] [Google Scholar]
  • 6.Eksteen M, Tiller H, Averina M, et al. Characterization of a Human Platelet Antigen-1a–Specific Monoclonal Antibody Derived from a B Cell from a Woman Alloimmunized in Pregnancy. J Immunol 2015; 194: 5751–60. [DOI] [PubMed] [Google Scholar]
  • 7.Bessos H, Turner M, Urbaniak S. Is there a relationship between anti-HPA-1a concentration and severity of neonatal alloimmune thrombocytopenia? Immunohematology 2005; 21: 102–9. [PubMed] [Google Scholar]
  • 8.Killie M, Husebekk A, Kaplan C, et al. Maternal human platelet antigen‐1a antibody level correlates with the platelet count in the newborns: a retrospective study. Transfusion 2007; 47: 55–8. [DOI] [PubMed] [Google Scholar]
  • 9.Kjær M, Bertrand G, Bakchoul T, et al. Maternal HPA-1a antibody level and its role in predicting the severity of Fetal/Neonatal Alloimmune Thrombocytopenia: a systematic review. Vox Sang 2019; 114: 79–94. [DOI] [PubMed] [Google Scholar]
  • 10.Santoso S, Wihadmadyatami H, Newman P, et al. Anti-endothelial αvβ3 antibodies are a major cause of intracranial bleeding in fetal-neonatal alloimmune thrombocytopenia. 2016; 36: 1517–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Newman PJ, Derbes RS, Aster RH. The human platelet alloantigens, PI(A1) and PI(A2), are associated with a leucine33/proline33 amino acid polymorphism in membrane glycoprotein IIIa, and are distinguishable by DNA typing. J Clin Invest 1989; 83: 1778–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Hynes RO. Integrins: Versatility, modulation, and signaling in cell adhesion. Cell 1992; 69: 11–25. [DOI] [PubMed] [Google Scholar]
  • 13.Zhou Y, Fisher SJ, Janatpour M, et al. Human cytotrophoblasts adopt a vascular phenotype as they differentiate: A strategy for successful endovascular invasion? J Clin Invest 1997; 99: 2139–51. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Zhi H, Ahlen MT, Thinn AMM, et al. High-resolution mapping of the polyclonal immune response to the human platelet alloantigen HPA-1a (Pl A1). Blood Adv 2018; 2: 3001–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Thinn AMM, Wang Z, Zhou D, et al. Autonomous conformational regulation of β3 integrin and the conformation-dependent property of HPA-1a alloantibodies. Proc Natl Acad Sci 2018; 115: 9105–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.George JN, Caen JP, Nurden AT. Glanzmann’s Thrombasthenia: The Spectrum of Clinical Disease. Blood 1972; 75: 1383–95. [PubMed] [Google Scholar]
  • 17.Sandrock-Lang K, Johannes O, Wiegering V, et al. Characterisation of patients with Glanzmann thrombasthenia and identification of 17 novel mutations. Thromb Haemost 2015; 113: 782–91. [DOI] [PubMed] [Google Scholar]
  • 18.Santoso S, Kalb R, Walka M, Kiefel V, Mueller-Eckhardt C, Newman PJ. The human platelet alloantigens Br(a) and Br(b) are associated with a single amino acid polymorphism on glycoprotein Ia (integrin subunit alpha 2). J Clin Invest. 1993. ;92(5):2427–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Horton MA, Taylor ML, Arnett TR, Helfrich MH. Arg-Gly-Asp (RGD) peptides and the anti-vitronectin receptor antibody 23C6 inhibit dentine resorption and cell spreading by osteoclasts. Exp Cell Res. 1991; 195: 368–75. [DOI] [PubMed] [Google Scholar]
  • 20.Kroll H, Kiefel V, Santoso S, et al. Sra, a private platelet antigen on glycoprotein IIIa associated with neonatal alloimmune thrombocytopenia. Blood 1990; 76: 2296–302. [PubMed] [Google Scholar]
  • 21.Wihadmadyatami H, Röder L, Berghöfer H, Bein G, Heidinger K, Sachs UJ, Santoso S. Immunization against αIIbβ3 and αvβ3 in a type 1 variant of Glanzmann’s thrombasthenia caused by a missense mutation Gly540Asp on β3. Thromb Haemost. 2016;116:262–71. [DOI] [PubMed] [Google Scholar]
  • 22.Kiefel V, Santoso S, Weisheit M, Mueller-Eckhardt C. Monoclonal antibody-specific immobilization of platelet antigens (MAIPA): a new tool for the identification of platelet-reactive antibodies. Blood. 1987; 70: 1722–6. [PubMed] [Google Scholar]
  • 23.Kannan M, Ahmad F, Yadav B, et al. Molecular defects in ITGA2B and ITGB3 genes in patients with Glanzmann thrombasthenia. J Thromb Haemost 2009; 7: 1878–85. [DOI] [PubMed] [Google Scholar]
  • 24.Nurden AT, Pillois X, Fiore M, et al. Expanding the mutation spectrum affecting aIIbb3 integrin in Glanzmann Thrombasthenia: Screening of the ITGA2B and ITGB3 genes in a large international cohort. Human Mutations 2015; 36: 489–571. [DOI] [PubMed] [Google Scholar]
  • 25.Nurden AT, Fiore M, Nurden P, et al. Glanzmann thrombasthenia : a review of ITGA2B and ITGB3 defects with emphasis on variants, phenotypic variability, and mouse models. Blood 2011; 118: 5996–6005. [DOI] [PubMed] [Google Scholar]
  • 26.Bayat B, Traum A, Berghöfer H, et al. Current Anti-HPA-1a standard antibodies react with the β3 Integrin subunit but not with αIIbβ3 and αvβ3 complexes. Thromb Haemost 2019; 119: 1807–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Valentin N, Visentin GP, Newman PJ. Involvement of the Cysteine-Rich Domain of Glycoprotein IIIa in the Expression of the Human Platelet Alloantigen, PIA: Evidence for heterogeneity in the humoral response. Blood 1995; 85: 3028–33. [PubMed] [Google Scholar]
  • 28.Honda S, Tomiyama Y, Pelletier AJ, Annis D, Honda Y, Orchekowski R, Ruggeri Z, Kunicki TJ. Topography of ligand-induced binding sites, including a novel cation-sensitive epitope (AP5) at the amino terminus, of the human integrin b3 subunit. J Biol Chem. 1995; 270: 11947–54. [DOI] [PubMed] [Google Scholar]
  • 29.Stafford P, Ghevaert CJG, Campbell K, et al. Immunologic and structural analysis of eight novel domain-deletion β3 integrin peptides designed for detection of HPA-1 antibodies. J Thromb Haemost 2008; 6: 366–75. [DOI] [PubMed] [Google Scholar]
  • 30.Newman PJ, McEver RP, Doers MP, et al. Synergistic action of two murine monoclonal antibodies that inhibit ADP-induced platelet aggregation without blocking fibrinogen binding. Blood 1987; 69: 668–76. [PubMed] [Google Scholar]
  • 31.Peterson JA, Nyree CE, Newman PJ, Aster RH. A site involving the “hybrid” and PSI homology domains of GPIIIa (beta 3-integrin subunit) is a common target for antibodies associated with quinine-induced immune thrombocytopenia. Blood 2003; 101: 937–42. [DOI] [PubMed] [Google Scholar]
  • 32.Beglova N, Blacklow SC, Takagi J, et al. Cysteine-rich module structure reveals a fulcrum for integrin rearrangement upon activation. Nat Struct Biol 2002; 9: 282–7. [DOI] [PubMed] [Google Scholar]
  • 33.Takagi J, Beglova N, Yalamanchili P, et al. Definition of EGF-like, closely interacting modules that bear activation epitopes in integrin b subunits. Proc Natl Acad Sci USA 2001; 98: 11175–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Zhu J, Luo BH, Xiao T, Zhang C, Nishida N, Springer TA. Structure of a complete integrin ectodomain in a physiologic resting state and activation and deactivation by applied forces. Molecular Cell 2008; 32: 849–861. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Zhou D, Thinn AMM, Zhao Y, et al. Structure of an extended b3 integrin. Blood 2018; 132: 962–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Xiong JP, Mahalingham B, Alonso JL, et al. Crystal structure of the complete integrin αVβ3 ectodomain plus an α/β transmembrane fragment. J Cell Biol 2009; 186: 589–600 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Mor-Cohen R, Rosenberg N, Einav Y, et al. Unique disulfide bonds in epidermal growth factor (EGF) domains of β3 affect structure and function of αIIbβ3 and αvβ3 integrins in different manner. J Biol Chem 2012; 287: 8879–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Kamata T, Ambo H, Puzon-McLaughlin W, et al. Critical cysteine residues for regulation of integrin alphaIIbbeta3 are clustered in the epidermal growth factor domains of the beta3 subunit. Biochem J 2004; 378: 1079–82. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Leeksma O, Giltay J, Zandbergen-Spaargaren J, et al. The platelet alloantigen Zwa or PlA1 is expressed by cultured endothelial cells. Br J Haematol 1987; 66: 369–73. [DOI] [PubMed] [Google Scholar]
  • 40.Ghevaert C, Campbell K, Stafford P, et al. HPA-1a antibody potency and bioactivity do not predict severity of fetomaternal alloimmune thrombocytopenia. Transfusion 2007; 47: 1296–305. [DOI] [PubMed] [Google Scholar]

RESOURCES