Evaluation of the Role of PlGF and the Therapeutic Potential of Anti-PlGF in BCR/ABL+ Leukemia.

Introduction of imatinib and second-generation BCR/ABL inhibitors has revolutionized treatment of patients with Philadelphia chromosome positive (Ph+) leukemia, but leukemia cells persist even in successfully treated patients, and some patients develop resistance and ultimately relapse. The reasons for these drawbacks are not entirely resolved, but besides BCR/ABL the host stroma potentially plays an important (independent) role. Placental Growth Factor (PlGF), a homologue of VEGF, was already proven to be abundantly secreted by stromal cells in solid tumors. Therefore, we investigated the role of PlGF, and the therapeutic potential of αPlGF, a monoclonal antibody against PlGF, which we recently reported to have a broad anti-tumoral potential in pre-clinical models of solid tumors (Fischer et al., Cell, 2007), in BCR/ABL+ lymphoid and myeloid leukemia. First, we studied expression of PlGF by 5 different human and murine BCR/ABL+ leukemia cell lines (Bv-173, BaF3, 32D, K562, KCL22) in vitro and found that neither of the cell lines secreted PlGF protein, but expressed its target receptor VEGFR-1. In contrast, primary murine adherent bone marrow stromal cells (BMDSC) expressed abundant amounts of PlGF protein (up to 105 pg/ml/105 cells), indicating a potential stroma-related function of PlGF. Second, we analyzed whether PlGF could induce proliferation and thereby revealed dose-dependent induction of proliferation by recombinant PlGF in all analyzed leukemia cell lines. This pro-proliferative effect of PlGF was nearly completely abrogated by both αPlGF and an extracellular anti-VEGFR-1 antibody, thus indicating that it is mediated primarily by VEGFR-1. Third, we studied potential paracrine interactions between BMDSCs and leukemia cells by performing co-culture experiments. Remarkably, coculture of BMDSC with leukemia cells significantly induced proliferation of both cell types. We hypothesized, that this induction of proliferation might be mediated by PlGF and indeed found its nearly complete abrogation upon addition of αPlGF to the co-cultures. Furthermore, BMDSCs significantly upregulated PlGF secretion (2.1 fold; N=3; P=0.005) when cultured in presence of leukemia cells. Thus, we conclude, that stromal derived PlGF promotes proliferation of leukemia cells in a paracrine fashion and at the same time acts as autocrine pro-proliferative signal for stromal cells. To test this hypothesis in vivo, we established 3 different murine models of BCR/ABL+ myeloid and lymphoid leukemia. Subsequently, we analyzed PlGF protein as present in blood and bone marrow of diseased mice in comparison to healthy mice, and detected no PlGF protein in the peripheral blood of healthy mice and low amounts of PlGF protein in their bone marrow. In contrast, leukemic mice showed PlGF protein (76.5 ± 18.4 pg / ml plasma; N=7) in their circulation at levels comparable to mice bearing solid tumors, and, interestingly more than 8.9 fold (N=7; P<0.0001) elevated PlGF levels in their bone marrow, compared to healthy mice, again indicating that PlGF represents a stroma derived, novel pathogenetic factor in Ph+ leukemia. In order to investigate the therapeutic potential of PlGF inhibition in murine BCR/ABL+ leukemia, we treated mice bearing leukemia induced by syngeneic BCR/ABL+ BaF3 cells with αPlGF, and found a significant prolongation of median survival by 18 days (N=9; P=0.015), compared to control antibodies. Encouraged by these positive results, we established models of imatinib-sensitive and imatinib-resistant (T315I mutant) CML by transducing primary bone marrow cells and subsequent transplantation into lethally irradiated recipient mice. Interestingly, also in these aggressive models, we found a significant prolongation of survival of diseased mice induced by blockade of PlGF compared to controls (median survival prolongation in wt BCR/ABL induced leukemia 5 days; N=11; P=0.002; in T315I mutant 4 days; N=12; P=0.039). Bone marrow histology and phenotypic analysis by FACS revealed decreased infiltration of spleen and bone marrow with leukemia cells (reduction in the bone marrow by 38% and in the spleen by 24%). In conclusion, these data indicate that PlGF represents a stromal derived factor promoting the progression of Ph+ leukemia, independent of BCR/ABL mutational status, and might represent a novel target produced by the leukemic stroma, potentially useful adjunct to BCR/ABL kinase inhibitors.