A preleukemic phase of Transient Abnormal Myelopoiesis (TAM) occurs in approximately 28% of neonates with Trisomy 21 (Down Syndrome; DS), driven by acquired mutations in the transcription factor GATA1, resulting in an N-terminal truncated protein, GATA1 s. TAM can either be clinically overt or silent, and in the majority of cases the GATA1 s clone naturally disappears in the first few years of life. However, in around 10% of TAM patients the disease evolves into Myeloid Leukemia of Down Syndrome (ML-DS) with the acquisition of additional mutations.
We sought to functionally characterise mutations in JAK1, JAK2, JAK3, the thrombopoietin receptor (TPO-R, MPL), and CSF2RB (the common beta chain utilized by the IL-3, IL-5 and GM-CSF receptors) identified by whole exome and targeted resequencing in a recent study of 252 samples from 141 ML-DS and 111 TAM patients, including 12 with samples from both the TAM and ML-DS stages (Labuhn et al, under review). Mutations in genes involved in the JAK-STAT signalling pathway have been recurrently identified in myeloid malignancies, and in this patient cohort were found in both TAM and ML-DS samples, prompting the question of whether these variants contribute to the proliferative aspects of the disease.
We examined the 5 mutations identified in JAK1, 12 in JAK2, 16 in JAK3, 5 in MPL and 3 in CSF2RB from this cohort in a luciferase reporter assay for STAT5 transcriptional activity, which is frequently upregulated in proliferative diseases driven by constitutive JAK signalling. Additionally, we sought to dissect the molecular mechanism leading to the constitutive activity of the novel hotspot mutation identified in the transmembrane domain of CSF2RB, A455D, using both a luciferase reporter assay and a NanoBit protein-protein interaction assay to assess receptor chain requirements for dimerization.
Variants identified in the TAM stage did not drive constitutive signalling in our assays. In contrast, most of the mutations in our genes of interest from ML-DS samples resulted in constitutive STAT5 activity. Those that did not were, for the most part, found in patients harbouring additional known oncogenic mutations. The CSF2RB A455D mutant did not require the IL3-, IL5- or GM-CSF-R alpha chains for dimerization and constitutive activity.
These results demonstrate the importance of functionally characterising variants, and lead us to question whether mutations found in TAM have other functions contributing to the pathology of the disease, or are merely bystanders, and why they should occur in the same genes in ML-DS with an oncogenic result. Our examination of the novel CSF2RB A455D hotspot mutation's molecular mechanism of pathological signalling supports a model where this mutation drives CSF2RB dimerization in the absence of the alpha chain usually required for normal cytokine-stimulated signalling.