Exosomal lipids impact on tumoral cell behavior

The transfer of proteins or genetic material from exosomes and the subsequent effects on cell fate are well described. However lipids effects on cell behavior remain less documented. Few years ago Ristorcelli reported that the human pancreatic cancer SOJ-6 and BxPc-3 cells, 2 differentiated cancer cell lines are sensitive to exosomes whereas poorly differentiated MiaPaCa-2 and Panc-1 cells are (quite) insensitive. SOJ6 cells expressed exosomes, rich in lipidforming rafts, which further interacted with cell membrane to disturb the Notch-1 survival pathway. In particular the g-secretase, cleaving Notch-1 in the active Notch IntraCytoplasmic Domain (NICD), is sensitive to lipid microenvironment. These effects on Notch functioning down-regulate the phosphorylation of the pro-apoptotic PTEN and GSK-3b leading to their activation. GSK-3b inhibits mitochondrial PDH, up-regulates pro-apoptotic Bax protein, down-regulates anti-apoptotic Bcl2, Hes-1 (a nuclear target of the Notch pathway) and cyclin D1 expressions promoting the arrest of the cell cycle in the G0-G1 transition phase. To demonstrate the implication of exosomal lipids in promoting the death of SOJ-6 cells, Beloribi synthesized exosome-like nanoparticles (SELN) solely composed of lipids among which lipid-forming rafts predominate. The size of SELN, which appear stable with time, is compatible with a fusion/exchange with rafts in cell membranes. SELN co-localize with Notch-1 and with GangliosideM1 (GM1, a marker of rafts) at the level of plasma membrane and with Notch-1 and Rab5A at the level of early endosomes. Therefore data presented by Beloribi fit with a fusion/exchange of lipid SELN with rafts at the plasma membrane followed by endocytosis and/ or with a fusion/exchange with rafts of the endosomal membrane. Beloribi further showed that SELN mimicked the effects of exosomes and concluded that exosomal lipids alone could be responsible for the death of differentiated pancreatic tumor SOJ-6 cells. To further establish the involvement of rafts lipid the ratio Lo/Ld (i.e., lipids forming rafts versus non-raft lipids) was varied and higher the ratio Lo/Ld (i.e., the amount of raft lipids), greater the detrimental effects on SOJ-6 cells. Beloribi concluded that the raft lipids of SELN are responsible for SOJ-6 cells death. On another hand MiaPaCa-2 pancreatic cells, which are considered as a stem-like cancer (or initiating cancer) cell model as they are ALDH-positive and CD44positive (among others) were insensitive to exosomes and SELN. The difference in sensitivity to SELN could be due to different genetic backgrounds of SOJ-6 and MiaPaCa-2 cells. PTEN (a Notch partner) mutations have been involved in the resistance to Notch inhibition. However PTEN is rarely mutated in pancreatic adenocarcinomas and SOJ6 and MiaPaCa-2 cells expressed wildtype PTEN. Notch also represses stem/ progenitor cell expansion via p53 tumor suppressor protein. However p53 mutations have been observed in DNA specific binding domain in both exosomes-sensitive SOJ-6 and exosomesinsensitive MiaPaCa-2 cells and consequently cannot be the source of observed differences. Therefore the mechanisms of the MiaPaCa-2 cells resistance to exosomal lipids are somewhere else. In a recent paper Beloribi-Djefaflia aimed at deciphering the reason(s) of this resistance. In the presence of SELN, the expression of NICD decreases in MiaPaCa-2 cells but neither Hes-1, the nuclear target of NICD, nor the ratio Bax/Bcl2 were affected. Beloribi-Djefaflia concluded that Notch pathway was obviously affected in MiaPaCa-2 cells exposed to SELN without deleterious effects. Authors further showed that in MiaPaCa-2 cells SELN induced the activation of NF-kB, thus promoting the expression and the secretion of SDF-1a. Oncogenic Kras (a signature of pancreatic adenocarcinoma) acting as an upstream activator of Akt can stimulate the transcriptional activity of NF-kB which is constitutively activated in almost 70% of pancreatic cancers. Therefore authors showed that MiaPaCa-2 cells harbor the Kras pGly12Cys mutation and hypothesized that once the fusion had occurred nonrafts lipids of SELN may freely diffuse within the cell membrane compartment to impact on the constitutively active oncogenic Kras which exhibits stable interaction with the inner leaflet of plasma membrane at saturable non-raft sites. Such interaction is required for