Inhibition of TGF-β and CXCR4 Signaling Promotes Cell Cycle Arrest and Apoptosis In Leukemic Cells Co-Cultured with Stromal Cells

Abstract 1814

Transforming growth factor β1 (TGF-β1) is an essential regulator of cell proliferation, survival, and apoptosis, depending on the cellular context. In the bone marrow (BM) microenvironment, high levels of TGF-β1 are known to affect cell-to-cell interactions between cancer and stromal cells and to stimulate cancer cells survival. We have previously reported pro-survival effects of TGF-β1 in myelo-monocytic leukemia cells (Xu et al., Br J Haematol. 2008). In turn, the chemokine CXCL12 and its receptor CXCR4 play crucial roles in cell migration and stroma/leukemia cell interactions. CXCR4 contains TGF-β-inducible element Smad-binding sites in the proximal enhancer region, and TGF-β has been reported to up-regulate CXCR4 expression in several cancer cells.

In this study, we investigated the anti-proliferative effects of TGF-β neutralizing antibody in the context of BM microenvironment. Blocking TGF-β signaling with anti-TGF-β neutralizing antibody 1D11 (Genzyme) but not with isotype control antibody 13C4 decreased expression of the TGF-β downstream signaling targets plasminogen activator inhibitor-1 (PAI-1, mRNA) and Smad2 phosphorylation, in a dose-dependent manner (1-10 μg/mL) in three myelo-monocytic leukemia cell lines MV4;11, U937 and THP-1 cells. Treatment with rhTGF-β1 (2ng/mL, 72hrs) reduced cytarabine (0.5-2 μM) induced apoptosis (decrease in specific apoptosis measured by annexin flow cytometry: MV4;11, 9.2 +/− 2.3 %; U937, 14.5 +/− 0.7%; THP-1, 9.7 +/− 1.0 %, at 72 hrs). Further, rhTGF-β1 inhibited serum starvation-induced apoptosis and promoted G₀/G₁ cell cycle arrest (subG_1, MV4;11: 17.8% vs 2.4%; U937, 45.9% vs 24.2%; G₀/G_1, MV4;11: 30.4% vs 50.6%, U937, 12.9% vs 35.6%, at 72 hrs), which was effectively reversed by anti-TGF-β antibody 1D11. Co-culture of leukemic cells with bone marrow (BM)-derived mesenchymal stromal cells (MSC) in the presence of rhTGF-β1 further increased the proportion of cells in G₀/G₁ phase (MV4;11: 75.5%, U937: 39.7%), which was partially reversed by 1D11. Under MSC co-culture condition, 1D11 treatment induced prominent G₂/M accumulation (proportion of G₂/M cells, 1D11 vs 13C4: MV4;11, 6.8% vs 27.3%, U937, 10.5% vs 15.0%) with minimal apoptosis induction. These results indicate that TGF-β inhibition prevents accumulation of AML cells in a quiescent Go state, but does not induce apoptosis in cells lodging in the BM microenvironment.

We next evaluated effects of 1D11 antibody on CXCL12/CXCR4-mediated migration. While treatment with 1D11 only partially inhibited cell migration to CXCL12 in the presence of TGF-β1, CXCR4 antagonist AMD3100 completely diminished cell migration to CXCL12 with moderate increase of apoptosis in MSC co-cultures. Based on these findings, we are currently testing the hypothesis that the simultaneous blockade of TGF-β and CXCL12/CXCR4 signaling will inhibit AML cell migration and promote chemotherapy-induced cell killing of AML cells lodging in the BM microenvironment. The in vivo studies to test these hypotheses are ongoing and will be presented.