Functional characterization of PI3K C2 domain mutations detected in breast cancer circulating tumor cells and metastatic cells.

BACKGROUND: In breast cancer, over one third of all patients harbor a somatic mutation in the PIK3CA gene, encoding the p110α catalytic subunit of the phosphatidylinositol 3-kinase (PI3K) in their tumor cells. Circulating tumor cells (CTCs) are cells shed from the primary tumor into the blood stream. Recently, the long-term stable breast cancer CTC-ITB-01 cell line with tumorigenic and metastatic capacity was established from liquid biopsy derived cells. The oncogenic hotspot PIK3CA mutation H1047R (kinase domain) was detected in the primary tumor, CTCs and metastasis of the same patient. Other PIK3CA mutations located within the C2 domain (E418K and E453K) were detected in the CTCs and the vaginal metastasis but not in the primary tumor. The goal of our study was to functionally characterize the impact of the rare E418K and E453K mutations within the C2 domain that were not detected in the primary tumor. METHODS: PIK3CA mutations E418K, E453K, H1047R were generated by site-directed mutagenesis and stably overexpressed in breast cancer cells by lentiviral transduction. Subsequent signaling pathway activation was examined by western blot analysis. The impact of PIK3CA mutations on biological processes was studied by live cell imaging using the Incucyte Zoom system. Structural modeling was conducted in Pymol. The membrane localization of the mutants was evaluated by separating the cytosolic and membrane fraction using ultracentrifugation. Drug susceptibility of CTC-ITB-01 cells was analyzed by live cell imaging. RESULTS: Western blot analysis of human MDA-MB-231, MCF-7 and T47D breast cancer cells stably overexpressing either the PIK3CA wildtype (WT) or one of the E418K, E453K or H1047R mutants revealed a significant increase in AKT phosphorylation in both C2 mutants (E418K and E453K) and the kinase domain mutant H1047R. Functional analysis showed a significantly increased proliferation of MDA-MB-231 cells overexpressing the E453K and H1047R mutants. Migration was increased in all cells overexpressing WT and each of the mutants. Interestingly, invasion and chemotaxis were only enhanced in the MDA-MB-231 cells overexpressing the C2 domain mutants, i.e. E418K and E453K. In addition, membrane localization of the two C2 domain mutants was increased. Structural modeling of the E453K mutation suggests a disruption of the interaction between the negative regulatory domain of the p85α subunit and the p110α catalytic subunit as a potential mechanism leading to the observed activation of PI3K/AKT/mTOR signaling. Dual targeting of AKT/mTOR pathway by MK2206 and RAD001 leads to very strong synergistic effects (IC50 MK2206: 148 nM, IC50 RAD001: 15 nM) with respect to proliferation in the CTC-ITB-01 line through apoptosis induction. CONCLUSIONS: Our results demonstrate that PIK3CA C2 domain mutations activate PI3K downstream AKT signaling and can increase proliferation, migration and invasion after stable lentiviral transduction. Although both investigated mutations - E418K and E453K - are located within the C2 domain, a different molecular mechanism can be proposed. The PIK3CA mutated CTC-ITB-01 shows a high susceptibility against dual inhibition of AKT/mTOR. Further studies are required to fully elucidate the oncogenic potential of rare PIK3CA mutations.

Leukemia-associated mutations in SHIP1 inhibit its enzymatic activity, interaction with the GM-CSF receptor and Grb2, and its ability to inactivate PI3K/AKT signaling.

The inositol 5-phosphatase SHIP1 is a negative regulator of the PI3K/AKT pathway, which is constitutively activated in 50-70% of acute myeloid leukemias (AML). Ten different missense mutations in SHIP1 have been described in 3% of AML patients suggesting a functional role of SHIP1 in AML. Here, we report the identification of two new SHIP1 mutations T162P and R225W that were detected in 2 and 1 out of 96 AML patients, respectively. The functional analysis of all 12 AML-associated SHIP1 mutations, one ALL-associated SHIP1 mutation (Q1076X) and a missense SNP (H1168Y) revealed that two mutations i.e. Y643H and P1039S abrogated the ability of SHIP1 to reduce constitutive PI3K/AKT signaling in Jurkat cells. The loss of function of SHIP1 mutant Y643H which is localized in the inositol phosphatase domain was due to a reduction of the specific activity by 84%. Because all other SHIP1 mutants had a normal enzymatic activity, we assumed that these SHIP1 mutants may be functionally impaired due to a loss of interaction with plasma membrane receptors or adapter proteins. In agreement with this model, we found that the SHIP1 mutant F28L located in the FLVR motif of the SH2 domain was incapable of binding tyrosine-phosphorylated proteins including the GM-CSF receptor and that the SHIP1 mutant Q1076X lost its ability to bind to the C-terminal SH3 domain of the adapter protein Grb2. In addition, SHIP1 mutant P1039S which does not reduce PI3K/AKT signaling anymore is located in a PXXP SH3 domain consensus binding motif suggesting that mutation of the conserved proline residue interferes with binding of SHIP1 to a so far unidentified SH3 domain containing protein. In summary, our data indicate that SHIP1 mutations detected in human leukemia patients impair the negative regulatory function of SHIP1 on PI3K/AKT signaling in leukemia cells either directly by reduced enzymatic activity or indirectly by disturbed protein interaction with tyrosine-phosphorylated membrane receptors or adapter proteins. These results further support a functional role of SHIP1 as tumor suppressor protein in the pathogenesis of AML.

The inositol 5-phosphatase SHIP1 is a nucleo-cytoplasmic shuttling protein and enzymatically active in cell nuclei.

The inositol 5-phosphatase SHIP1 is a negative regulator of signaling processes in hematopoietic cells. SHIP1 mediates its regulatory function after relocalization from the cytoplasm to the plasma membrane where it converts its substrate PI(3,4,5)P(3) to PI(3,4)P(2) thereby terminating PI3-kinase mediated signaling. In addition, SHIP1 converts Ins(1,3,4,5)P(4) to Ins(1,3,4)P(3) thereby regulating inositol phosphate metabolism. Here we report, that SHIP1 can be detected in nuclear puncta of Jurkat cells by confocal microscopy after expression of SHIP1 from a tetracycline inducible vector. SHIP1-containing nuclear puncta partially co-localize with FLASH, a multifunctional nuclear protein that has been linked to apoptotic signaling and transcriptional control. Nuclear localization was confirmed for endogenously expressed SHIP1 in the myeloid leukemia cell line TF1. In addition, enzymatically active SHIP1 was found in nuclear fractions of Jurkat cells with a similar specific activity as cytoplasmic SHIP1. Further analysis revealed that SHIP1 is a nucleocytoplasmic shuttling protein which is actively imported into and exported out of the nucleus. Nuclear import is mediated by two canonical nuclear localization signals (NLS) i.e. K(327)KSK and K(547)KLR. Mutational inactivation of each NLS motif inhibited nuclear import and reduced the proliferation of cells indicating a functional role of nuclear SHIP1 for cell growth. Our data indicate that SHIP1 is partly localized in the nucleus and suggest that SHIP1 plays a role for nuclear phosphoinositide and/or nuclear inositol phosphate signaling.