The molecular pathogenesis of TAM and ML-DS in children with Down syndrome has provided valuable insights into myeloid leukemogenesis. Current research suggests that the development of TAM and subsequent progression to ML-DS result from a three-step process: perturbation of fetal hematopoiesis by trisomy 21; an acquired GATA1 mutation; and the acquisition of additional oncogenic mutations.
It is increasingly clear that trisomy 21 results in genome-wide changes in gene expression that affect, directly or indirectly, most chromosomes. This results in perturbed hematopoiesis with increased numbers and clonality of HSCs, increased frequency of megakaryocyte and erythroid progenitors, and reduced numbers of granulocyte- and macrophage-committed progenitor cells.
Recent studies have shown that virtually all patients with TAM, and most patients with ML-DS, harbor mutations in exon 2 and, occasionally, in exon 3 of the hematopoietic transcription factor GATA1. GATA1 is a double–zinc finger DNA-binding transcription factor expressed primarily in hematopoietic cells. It is required for the development of red blood cells, megakaryocytes, mast cells, and eosinophils. A number of different mutations in GATA1 have been identified, including insertions, deletions, mis sense mutations, nonsense mutations, and splice site mutations. All of these mutations lead to a block in the expression of the full length 50-kd isoform of GATA1 but allow for the expression of a smaller, 40-kd isoform (GATA1s). This smaller isoform lacks the N-terminal transactivation domain, but retains both zinc fingers involved in DNA binding as well as interactions with its cofactor, friend of GATA1 (FOG1). Recent studies have shown that mutations that alter GATA1-FOG1 binding in the N-terminal zinc finger or result in the expression of the GATA1s isoform uncouple megakaryocyte growth and differentiation. Similar studies in cell lines derived from children with ML-DS have demonstrated that expression of GATA1 led to erythroid differentiation whereas expression of GATA1s did not alter the characteristics of the cell line. Taken together, current data suggests that the loss of GATA1 and expression of GATA1s directly contribute to leukemogenesis. Importantly, TAM and ML-DS reflect an interaction between GATA1s and trisomy 21. This likely reflects the combined effects of multiple genetic interactions, with studies implicating various genes on chromosome 21 such as ERG, ETS2, RUNX1, CHAF1B, and mir-125b2.
Although GATA1 mutations in the context of trisomy 21 may be sufficient to cause TAM, these mutations are not sufficient for the development of ML-DS, as evidenced by the latency period between resolution of TAM and the development of ML-DS, as well as the observation that not all children with TAM and GATA1 mutations will ultimately develop ML-DS. Therefore, a multistep pathogenesis model is proposed in patients with Down syndrome in whom ML-DS develops in clones with GATA1 mutations and additional cooperating mutations. This has been corroborated by genomic analyses of TAM and ML-DS cases. These studies have revealed common molecular functions in genes mutated upon progression to ML-DS. One such class increases cytokine signaling, with typical lesions affecting genes in receptors (MPL. CSF2RB, KIT), associated kinases (JAK2, JAK3), and downstream signal transduction molecules (NRAS, KRAS, NF1, SH2B3). Another class of mutations disrupts the cohesin complex (STAG2, RAD21, CTCF, SMC1A). This multifunctional complex regulates sister chromosome pairing during mitosis and also participates in epigenetic regulation of gene expression in a broad fashion. Epigenetic regulation is also implicated by a third large class of mutations (EZH2, SUZ12, KANSL1, DNMT3A), often reducing function of the PRC2 complex and/or DNA methylation. Importantly, some of these same genes can be mutated in TAM, but without the same deleterious biochemical or functional consequences; thus, mutations of unclear functional significance must be interpreted with caution in TAM.
