Transfusion-related acute lung injury (TRALI), the leading cause of mortality associated with blood transfusion, usually results from passive transfer of antibodies present in donated blood to a patient. TRALI can be triggered by antibodies specific for Class I or Class II HLA antigens, human neutrophil antigens (HNA) and possibly other targets. For reasons not well understood, antibodies specific for the leukocyte antigen HNA-3a cause particularly severe, often fatal TRALI. It would be highly desirable therefore to be able to screen blood donors routinely for HNA-3a antibodies.

HNA-3a/b antigens are carried on choline transporter-like protein 2 (CTL2), an apparent 10 membrane-spanning protein with 5 extracellular loops and N and C intracellular termini. The HNA-3a/b polymorphism is created by an R/Q substitution at position 154 in the first of these extracellular loops (Loop 1). In solid phase assays, about one-half of HNA-3a antibodies implicated in TRALI recognize Loop 1 peptides containing R154 (Type 1 antibodies). The remaining antibodies (Type 2) are non-reactive with peptides despite reacting well against full length CTL2. We studied reactions of Type 1 and Type 2 HNA-3a antibodies against soluble recombinant CTL2 fragments, human CTL2, mouse CTL2, and human/mouse CTL2 chimeras expressed in HEK293 cells to characterize the basis for HNA-3a antibody heterogeneity. The following observations were made:

  • 1) Only Type 1 antibodies react with detergent-solublized CTL2 in solid phase assays.

  • 2) A soluble recombinant fragment derived from the first extracellular (EC) loop (R154) of human CTL2 reacts only with Type 1 antibodies.

  • 3) Mouse CTL2 is 91% identical to human CTL2 and contains the R154 residue critical for HNA-3a expression. Type 1 antibodies recognize mouse CTL2, but Type 2 antibodies do not.

  • 4) Chimeric CTL2 containing human sequence in EC Loop 1 and mouse sequence in Loops 2-5 (H1M) reacts only with Type 1 antibodies. The reciprocal construct with mouse sequence in EC loop 1 and human sequence in Loops 2-5 (M1H) reacts with both Type 1 and Type 2 antibodies.

  • 5) Chimeric CTL2 containing human sequence in EC Loops 1-2 and mouse sequence in Loops 3-5 (H2M) reacts only with Type 1 antibodies.

  • 6) Chimeric CTL2 containing human sequence in EC Loops 1-3 and mouse sequence in Loops 4-5 (H3M) reacts with both Type 1 and Type 2 antibodies.

These findings show that Loop 1 peptides containing R154 are sufficient for Type 1 antibodies to recognize CTL2 and the Type 1 epitope survives detergent solubilization of the protein. However, Type 2 antibodies require human sequence in EC Loops 1-3 for binding and the epitope they recognize does not survive detergent treatment. Moreover both Type 1 and Type 2 recognize the M1H chimera with the entire EC loop 1 sequence derived from mouse. The simplest explanation for these observations is that Type 2 HNA-3a antibodies need to contact human amino acid residues in EC Loop 3 in addition to Loop 1 for tight binding to CTL2. An alternative possibility is both Type 1 and 2 antibodies recognize only Loop1, but EC Loops 2 and 3 are required to hold Loop 1 in a configuration suitable for Type 2 antibody binding. In either case, it appears that at least the first 3 EC loops of CTL2 (R154) need to be in a configuration that closely mimics their natural state in the cell membrane in order to be recognized by Type 2 HNA-3a antibodies. Considerable ingenuity will be required to engineer a target capable of detecting both Type 1 and Type 2 HNA-3a antibodies in a format suitable for large-scale blood donor screening.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution