Donor source-dependent B cell receptor repertoire patterns visualized by the pGen-SHM plot reveal differences in adaptive immune reconstitution after hematopoietic stem cell transplantation

Background A reliable method to monitor adaptive immune reconstitution after allogeneic hematopoietic stem cell transplantation (HSCT) has not been established, leaving many aspects of donor-source-dependent differences in immune reconstitution unclear. Here, we developed a new method, "probability of generation (pGen)-somatic hypermutation (SHM) plot (pGen-SHM plot)", to objectively and visually monitor long-term immune reconstitution after HSCT. Methods We conducted longitudinal B cell receptor (BCR) repertoire analysis on peripheral blood mononuclear cells (PBMCs) from six cord blood transplantation (CBT) (21 samples) and six bone marrow transplantation (BMT) patients (14 samples) from two months to approximately two years post-HSCT. PBMCs from 12 healthy adults were used as controls. pGen value, representing the probability of sequence generation by the recombination process, was calculated using OLGA. SHM rates were determined by nucleotide-level identity to the reference V gene. Maturation of repertoire was visualized by plotting the two-dimensional probability density distribution of pGen and SHM rates using Kernel density estimation, which we termed the pGen-SHM plot. Results After CBT, the proportions of both high pGen (> 1e-15) and low SHM (< 4%) strongly correlated with post-HSCT days (r = -0.86 and r = -0.91, respectively) and gradually approached levels observed in healthy adults. Furthermore, the plot analysis revealed a continuous transition from the pGen high SHM low to pGen low SHM high population during the post-CBT period. This finding was validated using publicly available datasets of healthy individuals aged 0 to 15 years, which showed a similar age-related shift. In contrast, the proportions of high pGen (> 1e-15) and low SHM (< 4%) after BMT were already close to healthy adult levels by Day 67 and showed little change over two years. Remarkably, the plot analysis identified two distinct populations in the early post-BMT, pGen high SHM low and pGen low SHM high ; the latter disappeared at around six months while the former gradually shifted toward pGen low SHM high , resembling the pattern observed after CBT. We then performed a detailed analysis of the two populations identified in the pGen-SHM plot after BMT. The pGen low SHM high population showed a lower frequency of IgG1-switched sequences and higher frequency of IgG2-switched sequences compared with the pGen high SHM low population. These findings suggest that pGen low SHM high represents graft-derived mature B cells, whereas pGen high SHM low corresponds to newly generated and maturing B cells derived from hematopoietic stem cells. Conclusion The pGen-SHM plot objectively and visually captured distinct patterns of adaptive immune reconstitution after HSCT, depending on the donor source.