(407) Exon 3 May Present Opportunity for Differential Antibody Binding to DQA And DQB P-Group Alleles

Body: We present the case of a liver transplant candidate appearing to produce antibodies against alleles included in self HLA-DQ7 DQA1*05 and DQB1*03 P-groups in a potential cautionary example of Luminex single antigen bead (LSA) anti-HLA antibody (HLAab) identification analysis.
The case patient was HLA typed by hybrid capture Next Generation Sequencing (h-NGS) and found to be DQ7(DQA1*05:05/DQB1*03:01). Patient serum was assayed neat with HLAab native antigen screen (FPRA) and LSA standard and extended panels. Supportive data were obtained by HLAab native antigen Luminex mixed bead (LmxMix) panel testing and AXE adsorption/elution studies with living donor cells from a h-NGS homozygous DQ7(DQA1*05:05/DQB1*03:19) donor and from a h-NGS heterozygous DQ7(DQA1*05:03/DQB1*03:01) + DQ7(DQA1*05:05/DQB1*03:01 donor) to attempt to isolate antibodies specific to non-self, P-group alleles harboring DQB1*03:19 or DQA1*05:03. Alignments were performed in IMGT.
The DQ7(DQA1*05:05/DQB1*03:01) patient FPRA was positive and seemed to correlate with a LSA HLAab DQ stack including high MFI reactivity against DQ7 beads representing self P-group heterodimers DQ7(DQA1*05:03/DQB1*03:01) and DQ7(DQA1*05:05/DQB1*03:19) (Figure 1a-b). DQA1*05:03 and DQB1*03:19 differ from self-alleles in exon 3 at positions as 160 and 185, respectively, and exhibit unique eplets 160S and 185I (Figure 2). The eplet registry notes that mismatch 160S is not antibody verified, while mismatch 185I is antibody verified, and both eplet mismatches are high exposition. Since no LSA self DQ7(DQA1*05:05/DQB1*03:01) bead was available, LmxMix with self DQ7 and mismatched P-group DQ7 antigen beads were assayed. Beads coated in only self DQ7(DQA1*05:05/DQB1*03:01) were non-reactive, while other P-group DQ7 beads were antibody bound (Figure 1c). Patient serum adsorption with DQ7(DQA1*05:03/DQB1*03:01) LDPB was weak and did not rule out antibody against P-group heterodimers, while DQ7(DQA1*05:05/DQB1*03:19) donor cells did not support antibody against the 185I eplet mismatch (Figure 1d-g).

Conclusion: Immunologic equivalence of P-group alleles defined by exon 2 for HLA Class II molecules should not be interpreted to mean that structural elements outside of this exon cannot be targeted by alloantibody. While likely rare, this case supports that patients can produce antibodies against heterodimers in P-groups with self-heterodimers.