Global immunopeptidome remodeling favors piga mutant clones in bone marrow failure

Abstract

Abstract The basis of PNH immune privilege remains elusive. Various models have been proposed, each consistent with some but not all clinical features. Proposed models suggest that absent GPI-anchored proteins (GPI-APs) on PNH cells are immune targets, and / or their loss impairs effector attack. It is also possible GPI deficiency itself triggers autoimmunity yet shields mutant clones. Observations and theoretical considerations point towards peptide presentation dysfunction possibly driven by degradation of unanchored pro-proteins, proteasome overload, altered antigen processing, and generation or displacement of target peptides. Whether such remodeling narrows or diversifies antigen presentation remains unclear. We hypothesized that PIGA loss constrains and reshapes the global peptidome, whereas PNH- cells retain a broader, IFN-γ–expanded repertoire, enhancing immune visibility and promoting selective PIGA clone outgrowth in marrow failure. HLA-I immunopeptidomes were isolated from isogenic TF-1 WT and PIGA-mutant cells (PIGA cells) (received from Brodsky's Lab, JHU) (2 basal and 3 IFN-γ 10 ng/mL × 48 h biological replicates, each with 3 technical replicates) by immunoprecipitation. Western blot normalization ensured equal HLA recovery on the affinity column, yielding matched MS loading for both lines. Peptides were classified as union (≥ 1 replicate), reproducible (all replicates/condition), or core (reproducible, lineage-unique). Across all conditions, WT and PIGA cells displayed markedly different peptidomes. We identified 38,724 unique peptides from 15,923 proteins: 36.4% were WT-exclusive, 19.3% PIGA-exclusive, and 44.3% shared—a 1.9-fold enrichment of WT-only antigens. Basally, WT presented 10,693 peptides from 8,165 proteins versus 9,094 from 6,348 in PIGA (−15% peptide diversity, −22% protein coverage). The core peptidome contained 991 peptides in WT (9.3%) versus 391 in PIGA (4.3%). While 82% of PIGA cores appeared in WT, only 61% of WT cores were found in PIGA, confirming mutant constraint. PIGA showed systematic reduction: 15% fewer total peptides, 61% fewer core peptides, and nearly 2-fold fewer lineage-exclusive antigens (7,460 vs 14,104). IFN-γ amplified this discordance. WT expanded to 26,272 peptides from 14,427 proteins (2.46× and 1.77× increases); PIGA reached 18,531 from 13,441 (2.04× and 2.12×), widening the gap from 15% to 29%. Using >2 replicate threshold, WT displayed 4,644 peptides versus 3,764 in PIGA; peptides detected in all three replicates: 1,186 in WT versus 1,000 in PIGA. WT gained 20,571 new peptides compared to 15,526 in PIGA (+32.5%) and preserved 53% (5,701) of basal peptides, while PIGA retained only 33% (3,005). Critically, of 22 basal core-exclusive peptides per genotype, 5 WT-exclusive resurfaced under IFN-γ but none from PIGA, suggesting inflammatory conditions irreversibly constrain mutant antigen presentation.To quantify genotype-biased presentation, we analyzed 8–15-mers, reproducible in both basal runs or in >2 IFN-γ replicates. MS1 intensities were averaged, and log₂ fold-changes were calculated (WT ÷ PIGA). Peptides with |log₂FC| ≥ 1 were considered differential. Of 6,704 IFNγ-reproducible peptides, 3,834 (57%) were differential—2,685 enriched in WT and 1,149 in PIGA. Top WT-biased epitopes included TEVDNYHFY (NF2L2, log₂FC +4.4); PIGA-biased peptides included SLPRLQSLRVP (ZMY15, −5.9). GPI-AP profiling revealed that WT cells basally presented 22 peptides from 10 proproteins, while PIGA cells showed 12 from 7. IFN-γ expanded WT to 97 from 38 and PIGA to 70 from 30 (4.4× vs 5.8×); Cumulatively, WT presented 111 GPI peptides from 40 proteins, compared with 81 from 34 in PIGA. Junction neoepitope analysis revealed 17 of >10,000 predicted ω-site peptides (0.17%) were detectable, ~ 94% only after IFN-γ. Ongoing analysis is comparing junctional and C-terminal peptide diversity between WT and PIGA-mutant. Notwithstanding the limitations of cell line analysis, our preliminary findings suggest that the peptidome of PIGA-mutant cells differs significantly from that of WT isogenic cells. Ongoing analysis with paired PNH+/- patient cells, to be presented at ASH, may reveal novel clues to the selective advantage of PIGA-mutant cells in immune attack.