br Experimental Procedures br Author Contributions br

Experimental Procedures
Author Contributions
Acknowledgments
Introduction We recently described an animal model in which primary genetic lesions in bone marrow (BM) stromal dihydrofolate reductase resulted in abnormal hematopoiesis in mice and the rare development of acute leukemia (Raaijmakers et al., 2010). While evaluating three leukemias that emerged in the setting of this aberrant stroma, we noted that two of them had a shared chromosomal abnormality. Reasoning that this represented a highly nonstochastic event, we assessed the open reading frames in the altered region and found that two of them putatively encoded transmembrane molecules. Since transmembrane molecules might represent a means by which an altered BM stroma could select for abnormal hematopoietic stem/progenitors, we focused on these molecules. One of these, C14ORF37, is the subject of this report. Using overexpression and knockdown constructs, we found that C14ORF37 encodes a protein with unique characteristics in modulating a specific aspect of the AKT/mTOR growth and differentiation pathway in hematopoietic cells. Activation of AKT in hematopoiesis as induced experimentally by deletion of Pten leads to a myeloproliferative syndrome and eventual loss of hematopoietic stem cells (Kharas et al., 2010; Yilmaz et al., 2006; Zhang et al., 2006). This is mTOR dependent, as loss of hematopoietic stem cells can be rescued by rapamycin. The mTOR complex implicated in this process appears to be mTORC1, since experimental deletion of Raptor (a canonical component of mTORC1) similarly resulted in hematopoietic failure as evidenced by a lack of reconstituting ability upon transplantation (Kalaitzidis et al., 2012; Magee et al., 2012). In addition, pharmacological inhibition of mTOR resulted in antileukemia effects in a mouse model of acute myeloid leukemia (AML) (Zeng et al., 2007, 2012). In some leukemia-initiating cells, however, the AKT pathway appears to play a different role. AKT activation at position S473 is associated with subsequent phosphorylation of FOXOs, which reduces the ability of the FOXOs to enter the nucleus and serve as transcriptional regulators. However, 40% of human AML shows a gene-expression signature consistent with Foxo transcriptional activity, and FOXO inhibition in human and mouse AML cells results in terminal differentiation (Sykes et al., 2011). Deletion of Foxo1, Foxo3, and Foxo4 in a mouse model of AML resulted in differentiation and loss of leukemia-initiating cells. Therefore, the AKT pathway may have complex roles in the proliferation and differentiation abnormalities associated with myeloid malignancies. Acute lymphocytic leukemia (ALL) is more clearly driven by activation of the AKT/mTOR pathway. Rapamycin was shown to reduce the leukemia burden in xenograft models of B cell ALL (B-ALL), and reducing AKT activity by anti-connective tissue growth factor (anti-CTGF) therapy resulted in decreased B-ALL (Lu et al., 2014). In T cell ALL (T-ALL), clonal dominance was shown in cells where AKT activation increased and was associated with resistance to antitumor glucocorticoid therapy (Blackburn et al., 2014). Another study showed that inhibition of AKT could overcome glucocorticoid resistance (Piovan et al., 2013). The T-ALL that emerged with either Pten deficiency or with NOTCH activation was dependent upon very specific elements of the AKT pathway, in particular, the mTORC2 complex. This complex includes RICTOR, among other proteins, and is upstream of AKT, phosphorylating AKT at S473. Deletion of Rictor is capable of inhibiting T-ALL associated with either Pten deletion (Kalaitzidis et al., 2012) or NOTCH activation (Lee et al., 2012). Therefore, mTORC2 is critical for some lymphoid leukemias. Here, we report that C14ORF37 encodes a protein that is capable of interacting with and inhibiting the function of mTORC2. This distinctive transmembrane molecule may be a modulator of growth-regulator signals conveyed from the BM microenvironment.