Chromosome 21 Gain Is Dispensable for Transient Myeloproliferative Disorder (TMD) Development

Transient myeloproliferative disorder (TMD) is a hematopoietic disease, characterized by a clonal proliferation of immature megakaryoblasts in the neonatal period occurring in approximately 10% of newborns with Down syndrome (DS). Rarely, TMD occurs in non-DS newborns but then it is associated with somatic trisomy 21 (tri21). Tri21 together with in-utero gained mutations in the GATA1 gene encoding a myeloid transcription factor are thus considered essential in TMD.

Recently, we have identified a TMD with a typical manifestation and course in a newborn without DS/somatic tri21, which admits that tri21 is dispensable for TMD development. To elucidate the alternative TMD pathogenesis, we performed comprehensive genomic/transcriptomic profiling of this TMD case. We utilized high-density SNP array and whole exome and transcriptome sequencing (WES/RNAseq) to detect copy number changes, mutations and fusion genes. We did not find any aberrations on chromosome 21 and any fusion genes. Two focal intronic losses, likely representing benign germline variants, were found on chromosome X. In addition to 6 missense mutations affecting genes without established roles in hematopoietic disorders, we found in-frame deletions in the GATA1 and JAK1 genes. Both mutations are novel. The GATA1 D65_C228del mutation is predicted to result in an internally truncated protein - GATA1aber. Unlike GATA1s (resulting from GATA1 mutations in DS-TMD) which lacks the transactivation domain (TAD) but retains both Zinc fingers (ZF), GATA1aber lacks part of TAD and the N-terminal ZF. Nevertheless, we hypothesize that GATA1aber substitutes the pathogenetic role of GATA1s. The JAK1 gene encodes a non-receptor tyrosine-kinase engaged in the JAK/STAT signaling pathway. The identified mutation results in the loss of phenylalanine 636 (F636del), which is located in the pseudokinase domain and belongs to a conserved amino acid triad (F636-F575-V658) that is believed to mediate a structural switch controlling the JAK1 catalytic activity (Toms, Nat Struct Mol Biol, 2013). JAK1 mutations are implicated in various hematological malignancies including acute megakaryocytic leukemia, and we hypothesize that JAK1 F636del co-operates with GATA1aber on TMD pathogenesis via deregulation of cytokine/growth factor signaling.

We cloned the coding sequences of GATA1aber and JAK1 F636del and transfected them into a model cell line in which we confirmed the expression of both in-silico predicted proteins. Their subcellular trafficking was analogous to that of their wild type counterparts; GATA1aber was found in the nucleus and JAK1 F636del in both the nucleus and cytoplasm. Next, we assessed the kinase activity of JAK1 F636del. To distinguish auto- from trans-phosphorylation, we utilized the JAK1 F636del construct harboring an inactivating mutation of an ATP-binding site (K908G). The JAK1 F636del (but not JAK1 F636del + K908G) was autophosphorylated on Y1034/Y1035 and induced STATs phosphorylation both under steady-state conditions and following non-specific stimulation with PMA. However, at all studied time points all phosphorylation levels were lower compared to wild-type JAK1. Moreover, unlike constitutively active JAK1 V658I, JAK1 F636del did not confer IL3-independent growth to the murine B-cell progenitor cell line BAF3. Interestingly, the transforming potential of double-mutated JAK1 (JAK1 V658I + F636del) was enforced compared to JAK1 V658I. These data show that F636del does not lead to constitutive activation, but in the same time it is not functionally neutral. As the impact of F636del on JAK1 function may vary depending on upstream signaling, we are currently assessing JAK1 F636 kinase activity/transforming potential in BAF3 cells stably expressing the IL6 receptor, which (unlike the IL3 receptor) directly activates JAK1 upon ligand binding. In the future, we plan to study the impact of JAK1 F636del on GATA1s induced deregulation of erythroid/megakaryocytic lineage development and to demonstrate "GATA1s-like" function of GATA1aber.

To conclude, we identified two novel mutations affecting GATA1 and JAK1 as likely drivers in an alternative tri21-independent TMD pathogenesis. As the pathogenetic role of tri21 has been poorly understood so far, we believe that by clarifying an alternative mechanism of TMD development, we could improve our understanding of this intriguing disease in general.

Support: GAUK 86218