The formation of blood and immune cells is sustained throughout life by hematopoietic stem cells (HSCs). Yet, a detailed, quantitative understanding of the output and activity of individual human HSCs is only emerging. Here, somatic mitochondrial DNA (mtDNA) mutations were used as natural barcodes to trace clonal output and dynamics for up to three years. Mitochondrial single-cell ATAC sequencing-based (mtscATAC-seq) was applied to matched bone marrow and peripheral blood samples from three patients undergoing allogeneic stem cell transplantation (alloHSCT) to evaluate hematopoietic regeneration. Clonal reconstitution dynamics were compared to longitudinal profiles from blood samples of three healthy individuals, collectively capturing chromatin accessibility and mutational profiles of over 770,000 cells. Stable cell type compositions were observed for healthy individuals over time, highlighting the robustness of homeostatic hematopoiesis. Following alloHSCT, a sequential pattern of immune reconstitution was observed, with innate immune cells, such as monocytes, repopulating the blood earlier than adaptive immune cells. B cells, for example, were detected only months to years post-transplantation. By approximately two years after transplantation, immune cell compositions reached those observed in healthy individuals. Tracking mtDNA variants over time revealed a higher degree of fluctuation in mutational profiles during early post-transplantation time points, likely reflecting a less stable and more dynamic output during initial engraftment. At later time points, clonal outputs appeared more stable and resembled those of healthy individuals. Despite immune cell expansions in both healthy and transplanted individuals, no strong lineage biases were detected, nor was evidence found for selection-driven dominance or absence of single mtDNA variants. Subtle differences in the mitochondrial mutational signature were detected in alloHSCT patients, most likely related to treatment-associated mutagenesis. In addition to its role as a clonal tracking marker, mtDNA showed high sensitivity in chimerism analysis, but also revealed limitations for phylogenetic inference when used in isolation. Together, these findings provide one of the first longitudinal, single-cell resolved assessments of human hematopoietic clonal dynamics under physiological homeostasis and post-transplantation regeneration. By leveraging mtDNA, new insights into human stem cell behavior were generated, while also highlighting important considerations for future lineage tracing and diagnostic applications.
