Activated Fibroblasts Protect Acute Myeloid Leukemia Cells from Chemotherapy

Introduction: It has become clear that cancer is not only just a disease of the genes, but the tumor microenvironment (TME) plays an important role in cancer progression (Borriello L & DeClerck YA, 2014). In leukemia, the bone marrow plays a special role in environment-mediated drug resistance as it is not only a sanctuary protecting tumor cells from cytotoxic drugs, but also a source of many stromal cells that colonize primary tumors and contribute to the pre-metastatic niche (Borriello L & DeClerck YA, 2014; Meads MB, et al, 2008). As we all know, there exists two distinct BM niches in this significant microenvironment: ‘osteoblastic (endosteal)’ and ‘vascular’ niches (Tabe Y & Konopleva M, 2014; Konopleva M, et al, 2009). Besides these two, scholars transferred to focus on the stromal in bone marrow，which may have importance effects on leukemic cells (Wu S, et al, 2005; Campana D, et al, 2005). Shriram V. Nath et al. found that there was a negative correlation between reticulin fiber density (RFD) at diagnosis in childhood ALL and white blood cell count in peripheral blood. Similarly, there was a negative correlation between RFD and the percentage of blast cells in blood. They speculated fiber in the bone marrow leading to poor prognosis in acute lymphoblastic leukemia (ALL) via anchoring leukemia cells in bone marrow stromal (Nath SV, et al. 2011). But the role of fibers and its mainly producing cells fibroblasts has not been previously studied in acute myeloid leukemia (AML). Therefore, we retrospectively investigate the bone marrow biopsies of primary AML patients, to analyze the relationship between the RFD and prognosis in adult AML, and to identify and quantitative analyze protective effects of the fibroblasts on leukemia cells from chemotherapy in vitro.

Methods: Makers of activated fibroblasts were stained by immunohistochemistry on bone marrow biopsies; The RFD were evaluated based on the grid point method of computer; Obtained activated fibroblasts by induced BM-MSC of primary AML with human recombinant TGFβ1.(4) Analyzed protective effects of activated fibroblasts by co-cultured with leukemia cells such as THP-1/K562, futher explored the mechanism via SB431542, a specific inhibitor of TGFβ signaling pathway.

Result:

RFD in primary AML was significant higher than control. Patients with higher RFD indicated poorer prognosis; However, collagenI coated plates showed no effects on survival rate of the leukemic cells. Here, we demonstrated that the general makers of activated fibroblasts: FSP1, α-SMA and FAP in primary AML were significant higher than control, implied that there existed amount of activated fibroblasts in patients; Activated fibroblast protect both THP-1 and K562 from apoptosis with treatment of Ara-c, and arrested them in phase of G0/G1. Moreover, this protect effects of activated fibroblast will be cancelled with SB431542.

Conclusion:

Activated fibroblasts and fibrous tissue were both proliferated abnormally in primary AML, and the later was significantly associated with relapse-free survival and overall survival of patients, indicating that RFD could be used as an important factor of poor prognosis and should guide clinical intervention. In vitro induced activated fibroblasts could be able to protect leukemia cells from chemotherapy. The possible mechanisms might be that the soluble cytokines TGFβ1, but not collagen I, which were secreted by activated fibroblasts, contributed to leukemia protection. In addition to this, activated fibroblasts could also change cell cycle of tumor cells, make them more arrested in G0/G1 phase, thus decreased the their chemotherapy sensitivity.