Resistance of cancer cells to chemotherapy is a major issue in acute myeloid leukemia (AML) and is usually due to a clonal selection of resistant leukemic cells. Recently, a horizontal transfer of chemoresistance among tumor cells has been reported via extracellular vesicles (EVs). EVs are vesicles ranging from 0.03 to 1µm generated by almost all cell types. They carry cellular components such as nucleic acids, membrane and cytosolic proteins. By interacting with cells and by transferring their content, these vesicles could modify target cells' phenotype. Because of their presence in biological fluids, a lot of studies have tried to use EVs as biomarkers.

The aim of this research is to investigate the role of EVs in chemoresistance in AML. For this purpose, the sensitive strain of the promyelocytic leukemia HL60 cell line was compared to its multiresistant strain: HL60/AR. The latter overexpresses multidrug resistance-associated protein 1 (MRP1), an efflux pump conferring resistance to anthracyclines. The transfer of this resistance between the two strains is obtained by treating HL60 with EVs generated by HL60/AR.

The production of EVs by HL60/AR was confirmed by transmission electron microscopy. EVs from around 100 to 800 nm were highlighted. EVs from HL60/AR were then isolated by ultracentrifugation and their interaction with HL60 was investigated by labeling EVs with a fluorescent lipid membrane linker. The transmission of this fluorescence to HL60 was monitored by flow cytometry and demonstrated a binding or an incorporation of EVs.

In order to check if this interaction has an impact on HL60 resistance to daunorubicin, HL60 were treated with EVs, once for 20 hours or four consecutive times for 48 hours each. The resistance to daunorubicin of EVs-treated HL60 was then assessed by a MTT cytotoxicity assay and revealed that EVs increase survival of cells to daunorubicin. The resistant profile between the cells treated once or four times with EVs were somewhat different and highlighted a potential implication of nucleic acids cargo in this resistance transfer. Indeed, nucleic acids transferred by EVs would need more than 20 hours to impact target cells.

Afterwards, we investigated the changes in phenotype that could be responsible for this increased resistance of EVs-treated cells. First, the expression of MRP1 was evaluated by a flow cytometry drug retention assay based on the fluorescence exhibited by daunorubicin and consisting in analysis by flow cytometry of the daunorubicin retention of cells with or without the presence of a specific MRP1 inhibitor. This experiment emphasized an expression of MRP1 after EVs treatment. Oxidative stress also plays an important role in daunorubicin cytotoxicity. Thus reactive oxygen species (ROS) generation assays were performed on EVs-treated cells. These cells showed a decreased production of daunorubicin-induced ROS.

We finally analyzed the microRNAs content of isolated EVs. This was performed by TaqMan® Human MicroRNA Array on EVs from HL60/AR but also on EVs generated by HL60. This experiment aimed at highlighting the microRNAs cargo potentially involved in this chemoresistance transfer but also at identifying microRNAs differentially expressed between the two strains to serve as biomarker of chemoresistance in vivo. Out of the 380 microRNAs tested, 32 presented a differential expression between HL60 and HL60/AR. Among them, 5 microRNAs presented a more than 8-fold change between the two strains: miR-19b, miR-20a, miR-488, miR-519a and miR-523. miR-519a was the only microRNA more expressed in EVs produced by HL60, compared to EVs from HL60/AR.

In conclusion, this study demonstrates the ability of resistant leukemic cells to transfer, at least partially, their chemoresistance via EVs. This could be due to a direct transfer of MRP1 transported by EVs but also to a nucleic acid transfer. In this context, some microRNAs have been highlighted for their high differential expression in EVs between sensitive and chemoresistant cells. These results bring out a potential clinical use of these circulating microRNAs as chemoresistance biomarkers in AML.

Disclosures

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.

Sign in via your Institution