Fetal defense against leukemia: a developmental shield Free

In this issue of Blood, Mendoza-Castrejon et al¹ present a paradigm-shifting study that addresses a long-standing paradox in pediatric leukemia biology: why are congenital and infant leukemias, particularly those driven by MLL (KMT2A) rearrangements (MLLr), so rare despite their early origin and potent oncogenic potential. Using sophisticated mouse models and multiomic profiling, the authors uncover a heritable, fetal-stage resistance to leukemic transformation that is enforced by the epigenetic regulator MLL3.

MLLr mutations are known to arise in utero and require only a few cooperating mutations to initiate transformation.² Yet, the incidence of congenital leukemia is remarkably low, just a few cases per million births. Mendoza-Castrejon et al hypothesize that fetal hematopoietic progenitors possess intrinsic mechanisms that suppress transformation, and their data compellingly support this idea.

Using a temporally controlled mouse model of MLL::ENL expression, the authors demonstrate that fetal induction of the fusion gene leads to a leukemia-resistant state that persists after birth. In contrast, postnatal induction results in rapid and fully penetrant acute myeloid leukemia (AML). This resistance is not due to lack of transgene expression but rather to a selective pressure that favors progenitors with low oncogene expression and enhanced myeloid differentiation. Mendoza-Castrejon et al show that fetal MLL::ENL expression leads to transcriptional reprogramming of hematopoietic progenitors, marked by reduced expression of stemness genes (Pbx1, Myct1, Nkx2-3, Hlf) and enhanced myeloid priming, that limits transformation potential. Moreover, cooperating mutations of the RAS pathway (Nras^G12D), a pathway commonly mutated in infant leukemias³ can override this fetal barrier.

A key finding is the identification of MLL3, a histone methyltransferase, as a critical enforcer of this resistance. MLL3 is known to promote differentiation and restrict self-renewal in hematopoietic stem and progenitor cells.⁴ Mendoza-Castrejon et al show that MLL3 does not enforce myeloid bias directly; instead, it modulates the transcriptional response to oncogenic stress, restricting the capacity of myeloid-primed cells to undergo malignant transformation. This epigenetic enforcement is heritable and persists across cell divisions, even after transplantation into adult recipients. Importantly, these heritable transcriptional changes without corresponding shifts in chromatin accessibility, indicate that the resistant state is maintained through targeted epigenetic regulation rather than broad chromatin remodeling. The study thus positions MLL3 as a developmentally regulated tumor suppressor that preserves lineage fidelity and suppresses leukemic transformation during fetal hematopoiesis.

Although Lin28b has previously been implicated as a fetal-specific suppressor of MLL::ENL–driven leukemogenesis, with studies showing that its expression can inhibit leukemic transformation in neonatal models,^5,6 loss-of-function studies revealed that its inactivation did not accelerate transformation following fetal MLL::ENL induction.⁷ The identification of MLL3 as a key epigenetic regulator adds a crucial layer to our understanding of developmental leukemogenesis.

This study also has implications for understanding genetic predisposition. Human germ line variants in MLL3 have been associated with infant leukemia,⁸ independent of MLL rearrangements. These findings suggest that inherited defects in MLL3 may lower the fetal barrier to transformation, providing a mechanistic explanation for rare cases of congenital leukemia.

Although Mendoza-Castrejon et al provide compelling evidence for a heritable fetal resistance to MLL::ENL–driven leukemogenesis, the implications of this work extend beyond the immediate findings, opening several avenues for future investigation. Notably, transcriptional changes occur without major shifts in chromatin accessibility, raising the possibility that MLL::ENL may bind different genomic targets depending on developmental stage. Investigating whether the fusion protein engages distinct cofactors or binding sites in fetal vs postnatal progenitors would be crucial. One important question is whether similar fetal resistance mechanisms can be observed in human hematopoietic progenitors, such as those derived from fetal liver or cord blood. Recent studies using the MLL::AF4 fusion in primary human fetal liver CD34⁺ cells show that infant acute lymphoblastic leukemia (ALL), but not childhood ALL, maintains fetal-specific gene expression programs, supporting the role of developmental context in human cells leukemogenesis.⁹ 

Another promising direction involves exploring whether MLL3 or its downstream pathways could be therapeutically leveraged to reinforce differentiation and suppress leukemic transformation. The role of inherited variants in MLL3 and other epigenetic regulators warrants further study, particularly in understanding how these genetic predispositions might modulate the fetal barrier to transformation and contribute to the rare occurrence of congenital leukemia.

This study has profound implications for our understanding of pediatric leukemia biology. It suggests that fetal hematopoiesis is not merely permissive to transformation, but that fetal hematopoietic cells also have mechanisms to actively resist transformation through lineage bias and epigenetic regulation. The identification of MLL3 as a tumor suppressor in this context aligns with human genetic data linking MLL3 variants to infant leukemia predisposition. Furthermore, the work highlights the importance of ontogeny in shaping leukemic potential. It underscores the need to consider developmental timing when modeling leukemia initiation and interpreting disease heterogeneity. Mendoza-Castrejon et al provide a compelling explanation for the rarity of congenital and infant leukemias, reframing fetal hematopoiesis as a context that not only shapes leukemia phenotype but actively resists transformation. Their identification of MLL3 as a key enforcer of this resistance adds a critical layer to our understanding of pediatric AML and offers new insights into the developmental origins of leukemia.

Conflict-of-interest disclosure: The author declares no competing financial interests.