Purpose: Known genomic donor-recipient mismatches (mm) that influence allograft outcomes code for polymorphic major and minor HLA proteins recognized by adaptive immune cells. SIRPA alleles code for signal regulatory protein-α (SIRPα) variants that modulate innate cell function by binding with different affinities to CD47. Whether SIRPα mm detected by innate immune cells also affect allograft outcomes is unclear.
Methods: We tested how SIRPA gene mm affects renal allograft pathology and survival in mice by transplanting NOD or NOD.B6sirpα kidneys into WT.B6 or B6.CD47-/- mice. We next analyzed ~5000 publicly available human genomes for SIRPA SNPs that translate to amino acid polymorphisms in the CD47-binding domain of SIRPα. Finally, we analyzed the effect of SIRPA donor-recipient haplotype mm on allograft histology and outcomes in 375 renal transplant recipients with serial biopsies (protocol, for-cause), DSA testing and 7-yr follow-up.
Results: 
Murine model: When compared to B6 mice that received NOD renal allografts, those that received NOD.B6.SIRPα grafts developed significantly reduced chronic alloimmune injury manifesting as nodular cellular infiltrates and interstitial fibrosis. In addition, significantly less T cell, monocyte, monocyte-derived DC (mono-DC), and macrophage infiltrates were present in protected grafts, while no difference in alloantibodies was observed.  Similarly, B6.CD47-/- recipients were also protected from chronic alloimmune injury. Thus, elimination of SIRPα mm or CD47-binding abrogates chronic rejection. 
5000 Genomes project: We found 13 haplotypes that account for 99% variation in the human population. Absence or presence of a 3-nucleotide deletion that affects CD47 binding divided the top haplotypes into two groups: v1 and v2, respectively, referred to as A and B. A flow cytometry-based binding assay in healthy volunteers confirmed a greater affinity of SIRPα  BB haplotype to CD47 compared to others. 
Clinical transplantation: We tested the hypothesis that renal transplantation between a recipient with low-affinity SIRPα-A haplotype and donor with high-affinity B haplotype (SIRPα BàA mm) increases the risk of chronic alloimmune injury and premature allograft loss. Analysis of the 375 donor-recipient pairs showed that 24% patients had SIRP BàA mm, while 38% were fully matched. SIRP BàA mm was associated with significantly increased Banff ≥1A AR and premature IFTA by the 1st post-transplant year independent of potential confounders (HLA mm, DSA, DGF and KDPI). SIRP BàA mm was also associated with significantly worse overall and death-censored graft survival independent of the confounders. Finally, mediation analysis showed that poor graft survival in SIRP BàA mm group was mediated by premature IFTA.
Conclusion: Donor-recipient SIRPα mismatches that modulate innate alloimmune responses are associated with increased AR, premature IFTA and increased graft loss and are therefore important determinants of transplant success.