Juvenile Myelomonocytic Leukemia (JMML) is a clonal hematopoietic disorder of childhood that is principally characterized by proliferation of cells of the granulocytic and monocytic lineages. Previous studies showed that Ras or Nf1 mutations contribute to ~65% cases of JMML. More recently, mutations in the protein tyrosine phosphatase Shp2 (PTPN11) have been found in ~35% of sporadic JMML, and at lower incidence in acute myeloid leukemia (AML), myelodysplastic syndrome (MDS) and childhood B-cell acute lymphoblastic leukemia (B-ALL). Interestingly, germline mutations of Shp2 cause the autosomal dominant genetic disorder Noonan Syndrome (NS) and NS patients may have an increased incidence of JMML. Nearly all disease-associated Shp2 mutations affect residues known to control catalytic activity. NS and JMML mutations can involve the same residues, but when they do, the latter are less conservative, suggesting that they may be more activated. We have compared the biochemical and biological effects of NS and JMML mutations. When produced as recombinant proteins in bacteria, nearly all NS mutations studied showed increased PTP activity. However, the JMML mutations E76K and D61Y have the highest activity, resembling the fully active mutant E76A that we defined earlier. Retroviral-mediated expression of E76K or D61Y, but not wild type (WT) Shp2 in murine bone marrow (BM) cells results in cytokine-independent myeloid colony outgrowth, as well as hypersensitivity to both IL-3 and GM-CSF. Notably, NS- associated mutants (e.g., D61G or N308D) also could transform BM cells to factor-independence, but yielded far fewer colonies. Transformation by the leukemia-associated E76K mutant required the phosphatase activity and intact FLVRE motifs in the Shp2 SH2 domains. Gab2, a major Shp2 SH2 domain binding protein, also was required for transformation. When retrovirally transduced BM cells were transplanted into lethally irradiated recipients, E76K and D61Y but not WT Shp2 evoked fatal myeloproliferative disease (MPD) in 30–40% of recipients characterized by splenomegaly, leukocytosis, neutrophilia and monocytosis with occasional anemia and thrombocytopenia. Another ~20% showed less severe MPD at early stage and ultimately succumbed to T-ALL at later times. Histological analysis of mice dying MPD revealed BM and spleen packed with myeloid cells and livers with perivascular myeloid infiltrates. Flow cytometry on BM and spleen confirmed the presence of MPD. Moreover, BM from E76K and D61Y transplanted recipients exhibited enhanced factor-independent colony formation. Furthermore, mast cells derived from E76K and D61Y transplanted mice exhibit increased proliferation and enhanced activation of Erk, Akt and Stat5 in response to IL-3. In contrast, mice with a knock-in mutation of NS (D61G) displayed only a mild MPD. These data support a model in which leukemia-associated mutants of Shp2 are strong hypermorphs that enhance signaling via βc cytokines, whereas NS mutants are less severe gain of function alleles. Our results provide the first evidence that Shp2 mutants have a causal role in leukemogenesis and therefore, Shp2 is the first bona fide PTP proto-oncogene.

Author notes

Corresponding author

Sign in via your Institution