Revisiting the biology of infant t(4;11)/MLL-AF4+ B-cell acute lymphoblastic leukemia

Abstract

Infant B-cell acute lymphoblastic leukemia (B-ALL) accounts for 10% of childhood ALL. The genetic hallmark of most infant B-ALL is chromosomal rearrangements of the mixed-lineage leukemia (MLL) gene. Despite improvement in the clinical management and survival (∼85-90%) of childhood B-ALL, the outcome of infants with MLL-rearranged (MLL-r) B-ALL remains dismal, with overall survival <35%. Among MLL-r infant B-ALL, t(4;11)+ patients harboring the fusion MLL-AF4 (MA4) display a particularly poor prognosis and a pro-B/mixed phenotype. Studies in monozygotic twins and archived blood spots have provided compelling evidence of a single cell of prenatal origin as the target for MA4 fusion, explaining the brief leukemia latency. Despite its aggressiveness and short latency, current progress on its etiology, pathogenesis, and cellular origin is limited as evidenced by the lack of mouse/human models recapitulating the disease phenotype/latency. We propose this is because infant cancer is from an etiologic and pathogenesis standpoint distinct from adult cancer and should be seen as a developmental disease. This is supported by whole-genome sequencing studies suggesting that opposite to the view of cancer as a "multiple-and-sequential-hit" model, t(4;11) alone might be sufficient to spawn leukemia. The stable genome of these patients suggests that, in infant developmental cancer, one "big-hit" might be sufficient for overt disease and supports a key contribution of epigenetics and a prenatal cell of origin during a critical developmental window of stem cell vulnerability in the leukemia pathogenesis. Here, we revisit the biology of t(4;11)+ infant B-ALL with an emphasis on its origin, genetics, and disease models.

Figure 1

Proposed model for the oncogenic conversion of MLL fusions. (A) The physiologic situation of MLL functions. Taspase1-cleaved MLL is assembled into the holo-complex and binds to target promoter regions. This occurs via the N-terminally bound MEN1/LEDGF protein complex that allows binding to many transcription factors. The PHD domain is able to read histone core particles, whereas the SET domain allows it to write epigenetic signatures (H3K4_me2/3). Associated CREBP and MOF are able to acetylate nucleosomes. CYP33 allows switching into the repressor mode by enabling the docking of a Polycomb group complex composed of BMI1, HPC2, CtBP, and several HDACs. This enables the removal of acetyl groups from nucleosomes or transcription factors to shut down gene transcription. (B) In the case of a chromosomal translocation, the intrinsic regulatory mechanism of MLL becomes destroyed. The disrupted MLL portions are fused to protein sequences deriving from a large amount of different partner genes (n > 80). The N-terminal portion of MLL retains the ability to bind MEN1 and LEDGF, and thus, to bind to target promoter regions. Depending on the fusion sequence (AF4, AF5, LAF4, AF9, ENL, AF10), MLL-X fusions may recruit the endogenous AF4 complex that contains P-TEFb and the histone methyltransferases DOT1L, NSD1, and CARM1. This activates gene transcription and results in enhanced epigenetic signatures (H3K79_me2/3). The C-terminal portion retains CREBBP and MOF binding capacity, as well as the SET domain. In some cases (AF4, AF5, LAF4), the N-terminal fused protein sequences allow binding to P-TEFb and directly to the largest subunit of RNA polymerase II to enhance the process of transcriptional elongation. In addition, the fused protein sequences still bind NSD1 and DOT1L.

Figure 2

Two-hit cancer model in infant t(4;11)⁺ B-ALL. MA4 fusion is the first and driver oncogenic event. The very short latency of the disease indicates that secondary cooperating hits, if required, are expected to arise prenatally or very early after birth.