All-Trans Retinoic Acid (ATRA) up-Regulates Cell Surface CD38 Expression Which Promotes Pro-Tumoral Mitochondrial Trafficking from Stromal Cells to Multiple Myeloma

Background

Multiple myeloma (MM) is malignancy highly reliant on its microenvironment. In this study, we investigated whether mitochondrial transfer occurred between bone marrow stromal cells (BMSC) and malignant plasma cells. We then used our observations as a platform to investigate the mechanisms controlling pro-tumoral mitochondrial transfer with a view to identifying druggable targets.

Methods

Primary MM cells were obtained from patients' bone marrow after informed consent and under approval from the United Kingdom Health Research Authority. Animal experiments were conducted under approvals from the UK Home Office and the University of East Anglia Animal Welfare and Ethics Review Board. Primary BMSC were also obtained from patient bone marrow, using adherence and characterised using flow cytometry. Mitochondrial transfer was assessed using two methods; a MitoTracker Green based staining of the BMSC (in-vitro), rLV.EF1.AcGFP-Mem9 labelling of the MM plasma membrane with MitoTracker CMXRos staining of the BMSC (in-vitro) and an in vivo MM NSG xenograft model. CD38 expression on MM cells was tested after ATRA treatment, using RT-qPCR and flow cytometry. Mitochondrial transfer levels were assessed when CD38 was over expressed using ATRA or inhibited using lentivirus targeted shRNA.

Results

We report that mitochondria are transferred from BMSC to MM cells. First, we cultured MM cells on MitoTracker Green labelled BMSC and found increased MitoTracker Green fluorescence in the MM cells. We then transduced MM with rLV.EF1.AcGFP-Mem9 lentivirus and stained BMSC with MitoTracker CMXRos and used wide field microscopy to show MM derived tunnelling nanotubles (TNT) formed between MM cells and BMSC, with red mitochondria located within the GFP-tagged TNT. Next, we engrafted the MM cell lines MM1S and U266 into NSG mouse, after isolation we detected the presence of mouse mitochondrial DNA in the purified MM population. Together, these data show that mitochondria are transferred from BMSC to MM cells.

We next analysed OXPHOS levels in MM cells grown on BMSC, using the seahorse extracellular flux assay. We found that the MM cells had increased levels of OXPHOS after culture with BMSC, which was also the case for MM cell lines analysed after isolation from NSG mice, showing the micro-environment of MM can alter the metabolism of the malignant cell. To examine whether the mitochondrial transfer process was controlled by CD38, we knocked down CD38 in MM cells using lentiviral targeted shRNA. We found reduced levels of mitochondrial transfer in CD38KD MM cells, with a consequent reduction of OXPHOS in the malignant cells.

Finally, as ATRA has previously been shown to increase CD38 expression in AML, we next quantified CD38 mRNA and surface glycoprotein level on malignant plasma cells with and without ATRA treatment. We found ATRA increased CD38 expression at the mRNA and protein levels and this resulted in an increase in mitochondrial transfer from BMSC to MM cells.

Conclusion

Here we show that CD38 mediated mitochondrial transfer in the MM micro-environment forms part of the malignant phenotype of multiple myeloma. This finding develops our understanding of the mechanisms which underpin the efficacy of CD38 directed therapy in MM.