Abstract
Accumulating evidence suggests that bone marrow stromal micro-environment act as sanctuary site for acute myeloid leukemia (AML) cells and affords protection from conventional chemotherapy agents. Recently, there has been considerable interest in extracellular vesicles (EVs) and their role in intercellular communication. We had previously reported that there is significant stromal cell mediated drug resistance in acute promyelocytic leukemia (APL) which was predominantly driven by up-regulation of the NF-kB pathway in the malignant cells (Ganesan S et al, Leukemia 2016). Though there are reports of similar stromal cell mediated protection against chemotherapy in non-M3 AML cells, there is limited data on the role EVs play, if any, in inducing such chemotherapy resistance. We undertook a preliminary study to evaluate the role of EVs in inducing resistance to chemotherapeutic agents in AML.
We demonstrated that direct stromal co-culture experiments with AML cell lines as well as primary AML cells showed a significant stromal mediated protective effect against both cytosine arabinoside (AraC) and daunorubicin (DNR) [U937: AraC p < 0.01, DNR p < 0.001 (n=5); primary AML cells: AraC p < 0.01, DNR p < 0.001 (n=26)]. Gene expression profiling (GEP) analysis of AML cell line U937 co-cultured with HS-5 revealed a significant enrichment of genes involved in PI3K-AKT signaling pathway (p < 0.0001) followed by NF-kB pathway genes. It has been previously reported that on direct co-culture with stromal cells there is activation of the PI3K-AKT pathway in AML cells leading to drug resistance (Tabe Y et al. Cancer Res 2007). It is also well recognized that PI3K-AKT pathway can drive NF-kB signaling, which is consistent with our earlier reported observation in APL. We hypothesize that this effect of PI3K-AKT activation could be mediated by EVs without direct malignant and stromal cell interaction as well. Towards this, we isolated stromal EVs as previously described (Suzanne M et al, Blood 2016). The size and concentration of EVs was characterized by NanoSight and flow cytometry (data not shown).
We demonstrated an activation of PI3K-AKT signaling by showing phosphorylation of AKT (P-AKTser473) in U937 cells when treated with stromal EVs (figure 1A). We further demonstrated that stromal EVs were able to mediate a protective effect against AraC and DNR similar to that seen on direct co-culture with HS-5 cells [AraC p < 0.01, DNR p < 0.01 (n=3)]. This was further confirmed by looking at the pathway genes using Q-PCR, where an increased expression of PI3K-AKT signaling was observed in AML cells upon treatment with stromal EVs. The activation PI3K signaling was also confirmed by showing down regulation of PTEN [p < 0.01 (n=5)] which is a negative regulator of PI3K-AKT signaling. Further, to demonstrate the role of stromal EVs induced PI3K-AKT signaling in drug resistance, we treated the leukemic cells with a small molecule inhibitor of PI3 kinase (LY294002-10uM) along with AraC and DNR in the presence of stromal EVs. We noted that inhibiting PI3K-AKT signaling along with AraC or DNR was able to overcome the resistance mediated by stromal EVs (figure 1B). The mechanism of inducing apoptosis by LY294002 along with DNR or AraC (in the presence of stroma or stromal EVs) probably involves a TP53 dependent apoptosis pathway as evidenced through upregulation of TP53 and its target apoptotic genes such as BAD, CASP9 and PUMA (AraC- figure 1C, DNR - data not shown).
Our study thus illustrates the complexity of cellular crosstalk between leukemic cells and its microenvironment in inducing drug resistance. This resistance can be mediated by non-contact setting via stromal EVs through up regulation of PI3K-AKT signaling in myeloid leukemic cells. Our data highlights the need for additional studies that are required to explore and characterize in detail the cargo of these EVs and the mechanisms by which they mediate drug resistance.
No relevant conflicts of interest to declare.
