Origins of the human B-cell lineage

By comparing fetal and adult B-lymphopoiesis, O’Byrne et al,¹  in this issue of Blood, identified a pre-pro-B–cell subset that marks the earliest stages of B-cell lineage commitment in utero.

The origins and developmental progression of the B-cell lineage have been studied at substantial resolution for many years in mice.^2,3  Many of the phenotypic and functional distinctions that demarcate early stages of B-cell development in mice are also applicable to differentiating B-cell subsets in adult human bone marrow.⁴  The pre-pro-B–cell subset, marking earliest stages of B-cell differentiation in mice, was discovered and functionally characterized 28 years ago⁵ ; however, developmental origins of human B-cell development remained elusive.

Although previous studies in humans focused on adult bone marrow,⁴  O’Byrne et al compared fetal liver and bone marrow to adult bone marrow samples. The authors identified pre-pro-B cells (CD19⁺CD10⁻CD34⁺) as the first committed B-cell precursor subset that lacks T-cell and myeloid potential, adding an important new observation to the emerging picture of human B-lymphopoiesis. Although abundant in fetal tissues, pre-pro-B cells are exceedingly rare in adult bone marrow (see figure) and differ from their fetal counterparts functionally and transcriptionally. Leveraging state-of-the-art technology, including single-cell RNA-seq, ATAC-seq for transcriptome profiling and to survey chromatin accessibility, O’Byrne et al reconstructed developmental trajectories that place the pre-pro-B–cell subsets downstream of early lymphoid progenitors (ELPs; CD19⁻CD10⁻CD34⁺CD127⁺) and upstream of pro-B cells (CD19⁺CD10⁺CD34⁺). Unlike ELPs, pre-pro-B cells did not retain non–B-lymphoid potential, suggesting that fetal pre-pro-B cells likely represent the earliest committed B-cell precursor. Despite the lack of T-cell and myeloid potential, pre-pro-B cells continued to express genes that are commonly expressed in hematopoietic stem or multipotent progenitor cells, myeloid, or T cells. Surface expression of CD19 represents the main phenotypic distinction between ELP and pre-pro-B cells. Given that CD19 expression requires activity of the PAX5 transcription factor,⁶  a central driver of B-cell lineage commitment,⁷  it is likely that initiation of B-cell lineage commitment mirrors the onset of PAX5 expression in pre-pro-B cells.

The existence of an even earlier B-lymphoid precursor population cannot be excluded, but it seems unlikely that these rare progenitors would be functionally and phenotypically distinct from ELPs and pre-pro-B cells in a meaningful way. Pre-pro-B cells and pro-B cells functionally overlap in that both subsets give rise to mature B-cell development and actively undergo V(D)J recombination of immunoglobulin V_H region genes. Although D-J_H rearrangements were detected in both pre-pro-B cells and pro-B cells, rearrangement of V_H-to-DJ_H joints was specific for pro-B cells.

Studying matched liver and bone marrow samples from the same fetuses, the authors discovered 2 waves of early B-lymphopoiesis. Pre-pro-B cells first emerged in the fetal liver ∼7 weeks postconception, which was followed by a proliferative burst of pre-pro-B cells, mainly in fetal bone marrow. At 20 weeks postconception, pre-pro- and pro-B cells occupied nearly half of the hematopoietic progenitor cell pool in fetal bone marrow. The extent of proliferative expansion during these earliest stages of B-cell development was previously unrecognized and raised the possibility that pre-pro-B cells may be vulnerable to malignant transformation.

Adding to the significance of the discovery of this subset as developmental origin of the B-cell lineage, fetal pre-pro-B cells may also represent the cell of origin of acute lymphoblastic leukemia (B-ALL) developing in utero or early in postnatal life.⁸  For instance, B-ALL subtypes defined by MLL- or ETV6-RUNX1 rearrangements originate from a premalignant clone during fetal development.⁹  Like pre-pro-B cells, MLL-rearranged B-ALL clones typically carry immunoglobulin loci with rearranged D-J_H but not V_H-to-DJ_H segments and lack CD10 expression.¹⁰  Both pre-pro-B cells and MLL-rearranged B-ALL share a “mixed lineage” phenotype, although pre-pro-B cells, unlike MLL-rearranged B-ALL, lack myeloid lineage potential. Mirroring abundance of pre-pro-B cells in fetal bone marrow, MLL-rearranged B-ALL arises in utero and represents the vast majority of cases of infant B-ALL.⁹  In addition, O’Byrne et al highlight striking similarities in gene expression between fetal pre-pro-B cells and MLL-rearranged infant B-ALL.¹  Based on a series of phenotypic and functional comparisons, the authors draw a fascinating connection between fetal pre-pro-B cells as the origin of human B-cell development and their potential role as cell of origin of MLL-rearranged infant B-ALL.