Integrin αvβ3 Transduces Survival and Angiogenic Signals to T Cell Lymphomas and Is a Therapeutic Target

Malignant T cell proliferation, survival and drug resistant are dependent on a combination of external stimuli delivered by the microenvironment. Previous studies have shown that the transmembrane receptor integrin αVβ3 plays a critical role mediating the interaction of T cell lymphoma (TCL) cells with external signals. Integrin αVβ3 ligands include extracellular matrix-associated signaling proteins and soluble factors such as thyroid hormones (TH). We have also shown that THs stimulate the proliferation of TCLs through complimentary intracellular pathways involving the αVβ3 integrin. We therefore hypothesized that targeting integrin αVβ3 could represent a novel strategy to treat TCL patients. To determine survival pathways induced by TH on αVβ3 integrins, we evaluated the transcriptional changes (RNA-seq) induced by physiological concentrations of cell impermeable T3/T4 coupled to agarose (TH-AG) vs. control in the TCL cell line CUTLL1. We identified 123 up- and 5 down-regulated transcripts (p< 0.01), belonging to “angiogenesis” (e.g. VEFGB), “lymphocyte proliferation/differentiation” and “DNA replication/transcription” (e.g. DBP, IL4, EDF1, DOK2) pathways. Target-gene promoter analysis suggested that TH acting on αVβ3 integrin activated NFkB (that was later confirmed by EMSA-like assays). We then focused on the angiogenesis pathway since (i) VEGF expression and angiogenesis correlate with survival and prognosis in TCL patients, and (ii) we found a positive correlation between integrin αVβ3 and VEGFA and VEGFB expression in 169 cases of TCLs. We analyzed T3/T4-αVβ3-dependent increase of VEGFB and VEGFA in an extended panel of cell lines (n=5) representing the spectrum of immature and peripheral TCL. Similarly to CUTLL1, treatment with TH-AG increased VEGFB and VEGFA mRNA levels in Jurkat (TCL/L), HuT-78 (CTCL), OCI-Ly12 (PTCL-NOS) and Karpas299 (ALCL-ALK+) cells. Increase in VEGF production was completely abrogated by knocking-down either αV or β3components with specific si-RNAs (vs. siRNA control) in CUTLL1, HuT-78 and OCI-Ly12 cells. Moreover, exposing HMEC1 endothelial cells to conditioned medium from CUTLL1 cells treated with TH-AG vs control increased they proliferation and migration (to 481 ± 118 cells from 206 ± 82 cells, respectively, p= 0.01). Importantly, this higher migration was completely abrogated when conditioned medium was obtained from CUTLL1 cells knocked-down for either αV or β3 and treated with TH-AG. To determine whether targeting αVβ3integrin could be of therapeutic benefit for TCL, we developed TCL xenografts in SCID mice using CUTLL1 cells transfected with si-control, si-αV and si-β3, and monitored tumor growth and angiogenesis. We found that CUTLL1 transfected with si-αV and si-β3 developed significant smaller tumors than si-control. Also, tumors from integrin knocked-down cells, showed decreased tumor vascularization (by CD31) and VEGF expression. To determine the translational impact of this strategy, we assessed the effect of cilengitide, a selective αVβ3 integrin inhibitor in phase 3 for glioma, in pre-clinical models of PTCL-NOS and ALCL-ALK+. For PTCL-NOS we xenografted OCI-Ly12 cells in NOD/SCID mice (n=12) and for ALCL-ALK+ we developed a patient-derived tumorgraft (PDT) in NSG mice (n=8). Cilentide treatment for 10 days (vs. vehicle), at human equivalent dose, induced tumor remission in both models (p<0.001, T-test), without toxicity to normal tissues. Similarly to si-αV and si-β3 treated mice, the anti-lymphoma effect of cilengitide correlated with lower levels of angiogenesis and NFkB activation. In sum, we elucidated the mechanism by which integrin αVβ3 activation increases TCL proliferation by activating pro-survival pathways in malignant T cells while promoting angiogenesis. We also showed that genetic and pharmacological targeting of integrin αVβ3 induces anti-lymphoma effect in TCLs, including an ALCL-ALK+ PDT model obtained from an ALK inhibitor refractory patient.