Hyperactive Ptpn11 Mutations Alters Leukemic Stem Cell Frequency through Mcl-1 Overexpression

Shp2 is a non-receptor protein-tyrosine phosphatase encoded by PTPN11 and implicated in the Ras, JAK-STAT and PI3K pathways. Activating mutations in Shp2 are found in patients with developmental disorders such as Noonan and LEOPARD syndrome, as well as, hematologic malignancies. Although rare in most other solid tumors, Shp2 mutations are common in juvenile myelomonocytic leukemia (JMML) accounting for ~35% of cases. To understand its role as a cooperating mutation in AML we sequenced PTPN11 in human samples. Here we report that Shp2 mutations are present in human AML at a rate of 6.6% (6/91) in the ECOG E1900 dataset. To investigate the biological function of Shp2 mutations we asked how this functions in a cooperative model of leukemogenesis with the MLL-AF9 fusion protein. Despite showing increased genetic stability compared to other leukemias, MLL leukemias commonly contain type I mutations that can functionally cooperate resulting in more aggressive leukemias. These mutations often occur in genes encoding components of the Ras pathway including mutually exclusive mutations of NRAS, KRAS, PTPN11 and NF1 and account for ~37% of MLL rearranged leukemias. However, the mechanisms of cooperation with MLL fusion proteins are poorly understood. We found that the Shp2E76K activating mutation commonly found in humans significantly accelerates MLL-AF9 mediated leukemogenesis. The E76K mutation results in structural changes that confer increased phosphatase activity to the Shp2 protein and increased Ras signaling. We attribute the MLL-AF9/Shp2E76K cooperation to a more rapid leukemic initiation as evidenced in colony formation assays using mouse bone marrow HSPCs. Cells transduced with MLL-AF9/Shp2E76K expanded faster than MLL-AF9 cells at early stages following transduction, indicating more efficient transformation of myeloid progenitors than MLL-AF9 alone. Cytokine independent growth is achieved in MLL-AF9 cells following expression of Shp2E76K through the constitutive activation of the IL3 signaling pathway and ERK phosphorylation. Importantly, addition of Shp2E76K significantly accelerated MLL-AF9 mediated acute myeloid leukemia in mice, indicating activated Shp2 cooperates with MLL-AF9 in vivo. In addition, leukemic stem cell frequency was increased by greater than 4 fold due to Shp2E76K expression. As Shp2 is reported to regulate an anti-apoptotic gene program, we investigated these in the context of MLL-AF9 leukemic cells with and without Shp2E76K. While Bcl2, Bcl_XL and Mcl-1, were upregulated in Shp2E76K cells, Mcl-1 showed the highest upregulation in response to Shp2E76K. Further, expression of Mcl-1 with MLL-AF9 in colony assays phenocopies expression of Shp2E76K suggesting that, mechanistically, Shp2 mutations may cooperate through activation of an anti-apoptotic gene program, primarily through Mcl-1. Finally, we asked how leukemic cells bearing Shp2E76K would respond to small molecule inhibition of Mcl-1. MLL-AF9 leukemic cells expressing Shp2E76K were desensitized to small molecule mediated Mcl-1 inhibition consistent with increased Mcl-1 protein. These data were confirmed in human cells where U937 cells, which contain an activating Shp2 mutation, exhibited resistance to Mcl-1 inhibition compared to ML2 or K562 cell which both bear wild type Shp2. Together, these data suggest patients with hyperactive Shp2 signaling may respond poorly to drugs targeting Mcl-1 due to an overabundance of the protein.