Abstract

Background: The mechanisms by which AML cells and other hematologic malignancies interact with the BM microenvironment to provide a chemoprotective effect remain largely unknown. Our hypothesis is that resistance and relapse of AML may be related to altered AML-BM niche interactions, and that interruption of these interactions might enhance chemosensitivity thus overcoming chemotherapy resistance. In the current studies we have generated an in vitro model of chemotherapy protection mediated by a bone marrow (BM) stromal cell line and attempted to determine soluble and cell associated factors that mediate this protection against genotoxic stresses such as chemotherapy.sought to answer if murine stromal cells provide chemoprotection to mCGPR/+ APL cells in vitro.

Methods: APL cells generated from mice in which a single copy of the human PML-RARa was knocked into the murine cathepsin G locus (mCGPR/+; Westervelt et al, Blood. 2003 Sep 1;102(5):1857-65) were cultured in 12-well plates (normal or 0.4 μm pore transwell) with or without murine M2-10B4 stromal cells. After 24 h, APL cells were left untreated or treated with AraC (40 mg/ml) or daunorubicin (DNR; 40 ng/ml) for 2 days. FACS was used to test APL cell viability, cell cycle status, and proliferation kinetics using annexin V/PI, acridine orange, and CFSE assays, respectively.

Results: We observed a significant survival benefit for mCGPR/+ APL cells co-cultured with M2-10B4 stromal cells (Table 1). This survival advantage was observed in both the absence (untreated) and presence of chemotherapeutic agents. Optimal protection against both chemotherapy agents (95% daunorubicin and 92% Ara-C) was seen when APL cells were in direct contact with stromal cells. Interestingly, the survival benefit afforded by the stromal cells was largely maintained following separation of the APL cells from the M20-10B4 cells by a 0.4 um transwell (Table 1). This observation suggests that the anti-apoptotic stimulus provided by the M2-10B4 cells was mediated, at least in part, by a soluble factor released from the stromal cells. To begin to identify candidate molecules that might mediate this anti-apoptotic effect, we harvested media from wells incubated 72 h in the presence or absence of M2-10B4 cells and measured the levels of 64 different molecules using a luminex bead assay. The five abundant molecules detected in this initial screen were VCAM-1 (22-fold), SCF (29-fold), VEGF (747-fold), MCP-1 (979-fold), and MCP-3 (907-fold). It should be noted that the APL binding to stroma was not associated with alterations of cell cycle (except increase in G1a) nor decreased proliferation kinetics (no change in CFSE striping profiles relative to untreated cells).

Conclusion: These results demonstrate that murine stromal cells confer a chemotherapy survival advantage for mCGPR/+ APL cells in vitro. Furthermore, antibodies and small molecules (such as mobilizing agents) that interrupt direct binding of leukemic cells to the BM microenvironment or block signaling pathways mediated by physical attachment or anti-apoptotic soluble stromal factors may be rationally used to sensitize leukemia cells to chemotherapy in vivo.

Table 1.

M2-10B4 stromal cells provide a survival advantage to mCG PR/+ APL cells

% Apoptotic cells(AnnexinV+/PI-)
UntreatedAra-C (40ng/mL)DNR (40ng/mL)
No stroma 27% (SD 1.3) 80.6% (SD 1.9) 80% (SD 2.2) 
Transwell 4.3% (SD 4.1) 12.2% (SD 0.2) 10.1% (SD 2.8) 
Stroma 3.3% (SD 0.9) 8.2% (SD 0.9) 5.3% (SD 0.3) 
% Apoptotic cells(AnnexinV+/PI-)
UntreatedAra-C (40ng/mL)DNR (40ng/mL)
No stroma 27% (SD 1.3) 80.6% (SD 1.9) 80% (SD 2.2) 
Transwell 4.3% (SD 4.1) 12.2% (SD 0.2) 10.1% (SD 2.8) 
Stroma 3.3% (SD 0.9) 8.2% (SD 0.9) 5.3% (SD 0.3) 

Author notes

Disclosure:Honoraria Information: Dr DiPersio: AnorMed, Genzyme, MGI Pharma.