(603) Sharpening the AXE: Protocol optimization and troubleshooting

Aim: We sought to optimize the AXE (cell adsorption/elution) protocol through use of purified lymphocytes, determination of minimal cell counts required from different cell sources, and to troubleshoot post-AXE anti-HLA antibody identification result anomalies encountered.

Methods: AXE (Liwski et al., 2022) was modified with automated magnetic negative lymphocyte selection (RoboSep) and comparison of cell counts and sources (spleen=SP and deceased=DDPB or living donor peripheral blood=LDPB). Patient sera with known anti-HLA Class I and II antibody (HLAab) patterns were adsorbed with 2.5, 3.75, 5 and 10x10^6 lymphocytes. Minimally necessary cells for adsorption were determined. Pooled positive (PP) serum representing known, robust HLAab identification was utilized as positive control and negative human serum (NHS) as negative control. Serum eluates and washes were assayed for HLAab following AXE with Luminex single antigen beads (LSA). Donor cell incubation with only PBS was included in isolated cases, to determine if anything detected by post-AXE originated specifically from the donor cell alone, independent of patient serum contents.

Results: Minimal lymphocytes for AXE are noted in Table 1. PB and SP lymphocytes similarly adsorbed Class I HLAab, while SP optimally adsorbed Class II HLAab (Table 2a-d), likely related to B cell proportions and HLA Class II expression. Fewer target antigens on donor cells decreased AXE yield, while more antigen replication enhanced yield (i.e. SP > PB for Class II adsorption in the context of a donor with only a single HLA allele of interest). Interestingly, two HLA-A*02:01 cell (1 LDPB and 1 SP) adsorptions resulted in detection of an anti-HLA A2 107W eplet antibody pattern that was not present prior to AXE. PBS incubation (no serum) and elution off these cells revealed LSA binding from a product of the donor cells themselves, rather than patient serum (Figure 1a-d). Antibody elution buffer significantly damages cells, preventing its use to remove detectable donor cell LSA binding contaminants prior to patient serum incubation.

Conclusion: Source choice and number of isolated donor lymphocytes is important for AXE adsorption and elution efficiency. Spleen was superior for Class II HLAab adsorption. AXE elution can pull non-specific LSA binding contaminants off of some donor cells. It is important to analyze AXE eluate LSA data in the context of original serum LSA results.