Phosphatidylinositol-3-Kinase p110d Uniquely Promotes Gain-Of-Function PTPN11-Induced GM-CSF Hypersensitivity In a Model Of Juvenile Myelomonocytic Leukemia

Juvenile myelomonocytic leukemia (JMML) is a fatal leukemia affecting children under the age of 4 years and is characterized by myelomonocytic cell overproduction and hypersensitivity to GM-CSF. The only curative therapy is allogeneic stem cell transplantation; however, half of children relapse after this aggressive therapy. Approximately 85% of JMML patients bear loss-of-function (LOF) mutations in NF1 or CBL or gain-of-function (GOF) mutations in KRAS, NRAS, or PTPN11. Typically, these mutations are non-overlapping, with the net effect being Ras hyperactivation. Children bearing somatic GOF mutations within PTPN11, which encodes the protein tyrosine phosphatase, Shp2, exhibit the poorest prognosis. GOF Shp2 (Shp2D61Y and Shp2E76K) induces hyperactivation of both the Ras-MEK- Erk and PI3K-Akt pathways. While the Ras-MEK-Erk pathway clearly contributes to the pathogenesis of JMML, we hypothesize that the PI3K-Akt pathway cooperates with the Ras-MEK-Erk pathway to promote JMML. Recently published work indicates that genetic disruption of the PI3K regulatory subunit, p85a, reduces GOF Shp2-induced hypersensitivity to GM-CSF. However, as PI3K regulatory subunits cannot be easily inhibited pharmacologically, we examined the contribution of class IA PI3K catalytic subunits in GOF Shp2-induced JMML.

Shp2^D61Y/+;Mx1Cre+ mice were crossed with mice bearing conditional knockout of p110a (Pik3ca^flox/flox) or bearing a kinase dead mutant of p110d (Pik3cd^D910A/D910A). Shp2^D61Y/+;Mx1Cre-, Shp2^D61Y/+;Mx1Cre+, Shp2^D61Y/+;Mx1Cre+; Pik3ca^flox/flox, and Shp2^D61Y/+;Mx1Cre+; Pik3cd^D910A/D910A mice were treated with polyI;polyC, and 8 weeks post-treatment, animals were euthanized followed by evaluation of spleen size, hypersensitivity of bone marrow low density mononuclear cells (LDMNCs) to GM-CSF, frequency of bone marrow phenotypically-defined common myeloid, granulocyte-monocyte, and megakaryocyte-erythroid progenitors (CMPs, GMPs, and MEPs), and GM-CSF-stimulated Erk and Akt activation. Genetic disruption of p110a failed to normalize GOF Shp2-induced splenomegaly, GM-CSF hypersensitivity in proliferation assays and methylcellulose-based progenitor assays, or hyperphosphorylation of Erk or Akt. In contrast, genetic ablation of p110d kinase activity significantly reduced spleen size, normalized progenitor hypersensitivity to GM-CSF, and reduced both Akt and Erk hyperactivation. Additionally, genetic inhibition of p110d normalized the skewed hematopoietic progenitor distribution reported in the Shp2^D61Y/+;Mx1Cre+ mice, while genetic disruption of p110a failed to do so. This unique function of p110d in the context of GOF Shp2-expressing mice is significant, as p110d expression is restricted to hematopoietic cells and p110d bears transforming properties independent of Ras.

While previously published work indicates that the PI3K p110a and p110d inhibitor, GDC-0941, inhibits proliferation of GOF Shp2-expressing cells, we tested if the potent p110d-specific inhibitor, GS-9820, is similarly effective. GOF Shp2-expressing bone marrow LDMNCs treated with GS-9820 demonstrated significantly reduced proliferation in a dose-dependent fashion, while GS-9820 failed to inhibit the proliferation of WT Shp2-expressing cells. GS-9820 treatment decreased Akt phosphorylation (S473 and T308) as well as reduced Erk phosphorylation, indicating that p110d inhibition also reduces signaling within the Ras-MEK-Erk pathway. While PI3K activates the canonical Akt-mTORC1 pathway, it also positively feeds back to the Ras-MEK-Erk pathway via activation of Rac-Pak-MEK; therefore, we evaluated if p110d inhibition adds to or is redundant with MEK inhibition. Treatment of GOF Shp2-expressing hematopoietic cells with the MEK inhibitor, PD0325901, effectively reduced proliferation, and addition of GS-9820 further significantly reduced proliferation, indicating that p110d works cooperatively with MEK to promote GOF Shp2-induced disease.

Collectively, our findings suggest that PI3K catalytic subunit p110d functions in a Ras-MEK-Erk pathway-independent manner to promote GOF Shp2-induced hypersensitivity to GM-CSF, and suggest that PI3K p110d inhibition in combination with MEK inhibition may be a novel, optimal approach for the treatment of JMML.