Anti-HNA-3a Alloantibodies Can Cause Differential Endothelial Damage through a Direct Neutrophil Activation Pathway

Abstract

INTRODUCTIONThe human neutrophil antigen (HNA) 3a/b is associated with a single amino acid substitution (R 154Q polymorphism) on the first extracellular loop of choline transporter-like protein 2 (CTL2). Antibodies against HNA-3a have been associated with severe and fatal transfusion-related acute lung injury (TRALI). Epitope mapping suggests the existence of two types of HNA-3a antibodies. Type I antibodies only require first extracellular loop of the CTL2 for expression of the antigen, type II antibodies require at least the first 3 extracellular loops of CTL2 in a native configuration (i.e. expressed on a cell membrane). This study aimed to evaluate the activity of type I and type II anti-HNA-3a antibodies in an in vitro TRALI model.STUDY DESIGN & METHODS: The granulocyte agglutination test (GAT) and granulocyte immunofluorescence test (GIFT) were used to confirm anti-HNA-3a activity in two donor sera (Q49 and Q50). Flow cytometry was used to confirm antibody binding to freshly isolated human or mouse neutrophils. A rabbit polyclonal antibody against an epitope in the third extracellular loop of CTL2's was coated onto the wells of an ELISA plate and neutrophil HNA-3a antigen was captured and sera tested. For the TRALI model, human lung microvascular endothelial cells (HLMVECs) were grown to confluence before being treated with lipopolysaccharide (LPS). Freshly isolated neutrophils from HNA-3aa homozygote donors were then added with 10% Q49 or Q50. Microscopic counting of Trypan blue staining was used to measure rate HLMVEC death. Significance was determined using a one-way ANOVA (p<0.05), followed by Dunnett's post-hoc test. RESULTS AND DISCUSSIONGAT and GIFT confirmed that both Q49 and Q50 contained an anti-HNA-3a antibody, and that neither contained antibodies against any other HNA or HLA molecules. In flow cytometry, both Q49 and Q50 bound to human neutrophils, but only Q49 bound to mouse neutrophils. In the ELISA, signal was detected only for Q49. These data provided evidence that Q49 was a type I antibody and Q50 was a type II antibody. Within the TRALI model, HLVMEC death was observed with both Q49 and Q50 only in the presence of LPS and neutrophils (Table 1). Previously reported HLMVEC permeability in the absence of neutrophils (Bayat et al. Arterioscler Thromb Vasc Biol. 2013) was not observed as in this study the outcome measured was HLMVEC death. In the present study, the effect for Q49 (type I) was significantly larger than for Q50 (type II; Table 1; p-value <0.0001). CONCLUSIONSIn this model, LPS and neutrophils were required for anti-HNA-3a mediated HLMVEC death thought to reflect events involved in the initiation of TRALI. The observation that type I antibodies elicited a stronger response provides a direction for further research. [Display omitted] DisclosuresNo relevant conflicts of interest to declare.