Herein we observed increased SEMA A gene

Herein, we observed increased SEMA3A gene expression in BMSCs of all patients compared to the control group. In the literature, we found two studies of SEMA3A in AML: The first reported that epigenetic changes increases SEMA3A expression and contributes to EZH2 silencing in patients with secondary AML (Yang et al., 2015). The second described lower expression of SEMA3A in nonadherent apoptosis inducer from bone marrow stroma of AML patients compared to the control group (Wang et al., 2005). Thus, not much is known of the role of SEMA3A in the bone marrow microenvironment and more studies are needed to affirm SEMA3A involvement in the pathogenesis of MDS and AML. In order to investigate VEGFA and SEMA3A effects in leukemic cells, we proceeded to co-culture of HS5 stroma cells overexpressing SEMA3A and KG1 leukemic cells or CD34+ cells overexpressing VEGFA. The VEGFA overexpression increased proliferation and clonogenicity of KG1 cells and increased CD34+ cell proliferation; however, had no influence in apoptosis. On the other hand, co-culture of KG1 or CD34+ cells with HS5 overexpressing SEMA3A decreased KG1 proliferation and clonogenicity and decreased CD34+ cell proliferation. These data are consistent with the anti-angiogenic effects of SEMA3A described in solid tumors (Tamagnone, 2012; Vacca et al., 2006). When we combined KG1 cells overexpressing VEGFA and HS5 cells overexpressing SEMA3A, the effect of SEMA3A was dominant over VEGFA, causing decreased proliferation and colony formation (Tamagnone, 2012; Vacca et al., 2006). In co-culture of KG1 cells overexpressing VEGFA and HS27 cells overexpressing SEMA3A, the effect of SEMA3A was also dominant over VEGFA, causing decreased KG1 proliferation. In co-culture assays using normal CD34+ cells, the SEMA3A effect was also dominant and caused decreased CD34+ cell proliferation. In contrast to KG1 cells, VEGFA overexpression did not increase the colony formation of CD34+ cells isolated from umbilical cord blood. However, increased clonogenicity of bone marrow mononuclear cells from high-risk MDS patients, after treatment with recombinant VEGF, has been previously described (Bellamy et al., 2001). These discrepant results may be related to different signaling pathways activated in normal and leukemic cells. List and colleagues reported increased colony formation of KG1 cells dependent on the PI3K/Akt pathway and independent of ERK (List et al., 2004). Diversely, Koistinen et al. described increased clonogenicity of the leukemic line OCI/AML-2-dependent on nitric oxide (Koistinen et al., 2001). Leukemic cells have frequent mutations that may contribute to dysregulation of signaling pathways (van Gisbergen et al., 2015), and are not present in CD34+ normal cells, therefore, we hypothesized that leukemic cells are more committed and increased VEGFA may be enough to stimulate clonal growth. On the other hand, normal CD34+ cells may be more effective in controlling mechanisms of clonogenicity induced by VEGFA overexpression. Thus, our results suggest that the model proposed for solid tumors, wherein SEMA3A inhibits the effects caused by VEGFA (Chekhonin et al., 2013; Bender & Mac Gabhann, 2015) also applies to AML. The overexpression of VEGFA confers advantages to leukemic cells, by increasing their proliferation capacity, and SEMA3A inhibits these effects, preventing the uncontrolled proliferation of leukemic cells. The mechanism mediating this interaction has not yet been elucidated, however there is speculation that this could be through the competition between VEGFA and SEMA3A for NRP1 receptor binding (Pellet-Many et al., 2008). To investigate this hypothesis, we co-cultivated KG1 and HS5 cells treated or not with VEGFA and SEMA3A recombinant proteins and obtained protein extracts. We then immunoprecipitated anti-NRP1 receptor and performed immunoblotting with anti-VEGFR2 and anti-Plexin A4 receptor antibodies. Our results indicated that in the presence of the two recombinant proteins, the NRP1 receptor preferentially forms complexes with Plexin A4 rather than VEGFR2. The formation of NRP1 SEMA3Aplexin A4-complex is essential for SEMA3A signaling (Goshima et al., 2012). These data suggest that the dominant effect of SEMA3A may be related to the competition for NRP1 binding.