Developmental Stage Specific B-Progenitor Expansion in Normal Fetal Bone Marrow Is Absent in Down Syndrome: Implications for Infant ALL

Although distinct fetal and adult B-cell lineage development has been demonstrated in murine studies, human fetal B-lymphopoiesis remains poorly understood. Previous work from our lab identified a population of PreProB progenitors (CD34+CD19+CD10-) in fetal liver (FL), co-existing with adult-type CD34+CD19+CD10+ ProB progenitors (Roy, A., Cowan G et al. PNAS: 109, 17579-17584; 2012). This supports other work demonstrating a PreProB progenitor population in normal cord blood (Sanz, E., et al. PNAS: 107, 5925-5930; 2010) and suggests that fetal pathways of B-cell development are also likely to be markedly different from adult bone marrow (AdBM) in humans. FL PreProB progenitors give rise solely to B-cells in vitro and display a gene expression pattern of successive activation of B-cell transcription factors as determined by Fluidigm RQ-PCR. In addition, increasing evidence indicates that infant ALL and many cases of childhood ALL arise in fetal life, suggesting that ontogeny-related changes in B-cell development may provide the cellular and microenvironmental context for in utero leukemia initiation.

We therefore investigated B-cell development in normal human fetal BM from the onset of BM hematopoiesis in late first trimester. The composition and function of the early lymphoid and committed B-progenitor compartments of fetal BM(12-22 weeks; n=20) were compared with FL (n=25) at the same gestation, paediatric (Paed) BM (n=6) and AdBM (n=7), by multiparameter flow cytometry, differentiation in stromal co-culture assays and clonogenic assays. All stages of B-cell development were demonstrable in human fetal BM. However, there was a significantly higher frequency of B-progenitors in fetal BM (45.8±2.7% of CD34+ cells) compared to FL (10.3±0.97%; p<0.0001), PaedBM (28.2±4.2%; p=0.001) and AdBM (25.8±2.8%; p<0.0001). As in FL, both CD34+CD19+CD10- (PreProB) and CD34+CD19+10+ (ProB) progenitors were identified in fetal BM. PreProB progenitors were significantly higher in fetal BM (21.9±2.3% of CD34+ cells) compared to FL (3.8±0.4%), PaedBM (4.2±0.9%) and AdBM (3.4±0.9%) (p<0.0001 for all). Fetal BM PreProB progenitors gave rise solely to B-cells when co-cultured on MS5 stromal cells with FLT3L, SCF and IL7 in contrast to multilineage output of fetal BM HSC and LMPP cultured under identical conditions. Furthermore, fetal BM Lin-CD10-CD34+ cells cultured in vitro acquire CD19 expression before CD10 expression (n=6) consistent with a unique, fetal-specific B-cell differentiation pathway. The progressive decline in the proportion of PreProB progenitors from 47.8% to 36.8%, 14.8% and 13% of total B-progenitors in fetal BM, FL, PaedBM and AdBM respectively points towards a developmental stage-specific emergence of these progenitors. Finally, the fetal BM B-progenitor: B-cell ratio falls rapidly from ~4:1 at 12 weeks gestation to 1:1 at 18 weeks gestation as mature B-cell production in fetal BM gradually increases. This is consistent with a developmentally-regulated drive to B-progenitor proliferation, at the expense of differentiation, early in the second trimester which might represent a target population vulnerable to leukemic transformation in fetal life. Since children with Down syndrome (DS) do not develop infant ALL, we examined DS fetal BM as they may lack susceptible target cells for leukemic transformation. Consistent with this, PreProB progenitors in DS fetal BM (n=7) were >6-fold lower than normal fetal BM (3.3±0.8% vs. 21.9±2.3% of CD34+ cells; p<0.0001). Furthermore, CD19+ cells with an infant ALL-associated immunophenotype (CD19+CD10-CD20-) are detectable in normal fetal BM mononuclear cells, but are rarely found in DS fetal BM.

Conclusion: There is a marked expansion of PreProB progenitors in normal second trimester human fetal BM which is virtually absent in DS fetal BM, in normal PaedBM and in normal AdBM. We suggest that developmentally-regulated, functional and molecular characteristics of these fetal-specific B-progenitors may provide the 'oncogenic' cellular context necessary to co-operate with genetic events, such as MLL rearrangements, to induce ALL in infants without DS.