Background: Most severe cases of neonatal alloimmune thrombocytopenia (NAIT) in the Caucasian population are due to anti-HPA-1a. The precise influences of antibody amount or affinity on severity of NAIT remain to be elucidated. In an earlier study, we assessed the interaction between anti-HPA-1a and HPA-1a antigen by surface plasmon resonance technology (SPRT) using IgG fractions recovered from antibody (ab) positive HPA-1b maternal sera using protein-A chromatography. However, high background in some of the IgG fractions prohibited the determination of antibody affinity (Bessos et al, Trans Apher Sci, In Press). The aim of this study was to investigate the affinity of anti-HPA-1a purified directly from plasma using immobilised HPA-1a antigen.

Methods: Native glycoprotein (GP) IIb/IIIa was extracted and purified from HPA-1a1a platelet donors, while recombinant PSI domain of HPA-1a GPIIIa (rec-HPA-1a) was obtained by conventional cloning techniques. CNBr activated Sepharose 4B was coupled either with native GPIIb/IIIa or with rec-HPA-1a and used for the immunopurification of anti-HPA-1a directly from 4 female plasma samples positive for anti-HPA-1a ab. The plasma samples were pre-adsorbed with non-coupled Sepharose 4B prior to mixing with immobilised HPA-1a antigen. The acid eluted abs were dialysed, concentrated and assessed by ELISA (using HPA-1a GPIIb/IIIa coated plates), silver staining (using reduced gels) and SPRT-Biacore X, using CM5 sensor chips bound with HPA-1a1a or HPA-1b1b antigen at 600–800 response units (RU). GPIIIa AP3 monoclonal ab (Mab) & CamTran 007 recombinant anti-HPA-1a (r-anti-1a) were used as controls in SPRT. Affinity and dissociation constants (Ka and Kd respectively) of abs tested in serial doubling dilutions were determined using the computer software BIAevaluation 3.2 RC1 (Biacore) - Langmuir 1:1 mathematical model.

Results: Control AP3 Mab bound to both HPA-1a & -1b chips in SPRT yielding respective Ka values of 2x105 & 1.7x104, and respective Kd values of 2.6x10−3 & 6.7x10−5; while control r-anti-1a bound specifically to HPA-1a chips yielding Ka and Kd of 2.9x105 & 8.3x10−4. Antibodies purified from plasma using immobilised rec-HPA-1a demonstrated higher binding in the ELISA and purer samples by silver staining (IgG heavy and light chains) compared to abs purified using immobilised native GP. This was reflected in SPRT where anti-HPA-1a purified using immobilised native GP exhibited high background that prohibited Ka and Kd determinations, whereas abs purified using immobilised rec-HPA- 1a bound specifically to HPA-1a chips, enabling Ka and Kd determinations which ranged from 6.9x103 to 7.9x104 for Ka, and from 1.3x10−4 to 3.4x10−4 for Kd. Anti-HPA-1a purified by immobilised rec-HPA-1a also interacted specifically with CM 5 sensor chips bound with rec-HPA-1a but not rec-HPA-1b, although with lower affinity constants compared to chips bound with native HPA-1a GP, yielding Ka of 4.2x102 to 9.1x103, & Kd of 3.4x10−3 to 8.6x10−5.

Conclusion: This is the first report of the direct purification of anti-HPA-1a from plasma by immobilised HPA-1a antigen. Our study shows that immunopurification of plasma using immobilised rec-HPA-1a yields pure anti-HPA-1a antibodies that bind specifically in SPRT to chips coupled with both native or recombinant HPA-1a, allowing determination of antibody affinity and dissociation constants. This novel development would enable much needed analysis of anti-HPA-1a binding mechanisms in NAIT involving patients with different disease outcomes.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author