HLA-B leader sequences influence graft-versus-host-disease

From the Petersdorf laboratory, Clinical Research Division

Bone marrow or haemopoietic cell transplantation (HCT) is an effective treatment for multiple blood cancers. However, transplantation comes at a risk of acute or chronic graft-versus-host-disease (GVHD), where donor cells recognize and attack host cells and tissues, sometimes leading to life threatening complications. One important step in limiting GVHD risk is matching donor and recipient human leukocyte antigens (HLA) that present antigen to initiate immune responses and define tissue combability. When a patient does not have an HLA matched family member who could serve as the donor, it is not always feasible to find a perfect match due to large variation in HLA types. Each person inherits one HLA-A, HLA-B, HLA-C, HLA-DRB1, HLA-DQB1 and HLA-DPB1 gene (or allele) from each of their parents, leading to up to twelve alleles that must be matched. HLA-B is particularly difficult to match as it is the most polymorphic gene locus in the human genome; therefore, it has the potential for the most variation between donor and patient. However, when donors are mismatched at HLA-B with the patient, the risk of severe acute GVHD is much higher than when the donor is matched. Dr. Petersdorf (Clinical Research Division) recognized that HLA-B had another feature that made it distinct from other HLAs: it contains a divergent leader sequence that differs among individuals. During HCT, this leader sequence could potentially be involved in immune system recognition of HLA-E, a gene that is recognized by both NK and T cells, important cells in GVHD. The authors hypothesized that HCT with mismatched leader sequences would influence the rate of GVHD. The results of their study were recently published in The Lancet Haematology.   

The Human Leukocyte Antigen (HLA) complex is located on human chromosome 6. HLA-B leader sequences have diverged and vary between individuals.
The Human Leukocyte Antigen (HLA) complex is located on human chromosome 6. HLA-B leader sequences have diverged and vary between individuals. From Wiki Commons

The authors retrospectively analyzed a cohort of over 33,000 patients who received an HCT between 1988 and 2016 over multiple countries. As expected, the rate of GVHD increased as the number of HLA mismatches increased. A single HLA-B mismatch between patient and donor occurred in 1,457 HCT and was associated with higher levels of acute GVHD than any other single mismatch. In order to investigate the contribution of HLA-B leader sequence mismatch, the authors analyzed these single HLA-B mismatched patient/donor pairs. As each patient and donor have two HLA-B alleles, one HLA-B allele would be shared between donor and patient, and one would be different. The shared allele would also mean a shared leader sequence. However, the mismatched allele could either have a shared leader sequence or a mismatched leader sequence. The leader sequence divergence corresponds to a leader peptide with either a methionine (M) or a threonine (T). The authors found that GVHD occurrence is significantly lower when the mismatched patient and donor HLA-B alleles had the same leader sequence (either both M or both T), compared to a mismatched leader sequence. Interestingly, the authors also found that regardless of the mismatched HLA-B allele leader sequence, if the shared allele had an M leader sequence, patients had significantly increased occurrence of grade 3-4 GVHD than patient/donor pairs with shared T leaders. Finally, the authors studied the risks associated with different HLA-B-mismatched donors based on the leader sequences of the patient. They found that when a patient is TT, they tolerate an HLA-B mismatch better when the donor is also TT compared to MT. Likewise, MT patients have better outcome when the donor is mismatched against the M-leader HLA-B in the patient and matched for the HLA-B with the T-leader. 

Collectively, these data suggest that when possible, matching HLA-B alleles with T leaders rather than M leaders should be prioritized, as this would ultimately benefit the patient. Additionally, matching the leader sequences of mismatched alleles should be done whenever possible. However, one exception does apply: the authors found that MM patients benefit from an MT donor rather than an MM donor, though the samples numbers are small. This is because 91.7% of the HCT patients analyzed in the study had either TT or MT genotypes, suggesting few people have the risky MM genotype. Dr. Petersdorf summarized the importance of their results: “The success of transplantation is influenced by genetic variation that resides outside of the regions of the HLA-B gene that are currently assessed in clinical practice. The HLA-B leader provides novel information on GVHD responses, and may enhance the selection of unrelated donors for future patients.” Interestingly, while leader sequences correlated with GVHD occurrence, there were no discernable correlations between rate of relapse and leader sequences suggesting an alternate mechanism of relapse compared to GVHD. This study highlights a clinically feasible path to better match donors to patients in need of lifesaving HCT. 

This work was supported by funding from the National Institutes of Health, the US Office of Naval Research, Swiss FNRS, the Philanthropy Settlement Foundation, the Frederick National Laboratory for Cancer Research, and the Intramural Research Program of the National Institutes of Health, Frederick National Lab, Center for Cancer Research.

Fred Hutch/UW Cancer Consortium members Effie Petersdorf and Ted Gooley contributed to this work.

Petersdorf EW, Carrington M, O'hUigin C, Bengtsson M, De Santis D, Dubois V, Gooley T, Horowitz M, Hsu K, Madrigal JA, Maiers MJ, Malkki M, McKallor C, Morishima Y, Oudshoorn M, Spellman SR, Villard J, Stevenson P; International Histocompatibility Working Group in Hematopoietic Cell Transplantation. 2020. Role of HLA-B exon 1 in graft-versus-host disease after unrelated haemopoietic cell transplantation: a retrospective cohort study. The Lancet Haematology. 7(1):e50-e60.