Regulatory mechanisms of microRNA sorting into extracellular vesicles

Coronary artery disease is the most important cause of mortality worldwide. In the last decades, key mechanisms of atherosclerosis and plaque formation have been described.1, 2 It is commonly accepted that continuous damage of endothelial cells by oxidative stress causes endothelial apoptosis and dysfunction, which mark the beginning of atherosclerotic plaque formation.3, 4 If the aforementioned atherogenic factors outweigh atheroprotective factors, such as laminar fluid shear stress and intercellular anti-inflammatory signalling, plaque formation progresses and eventually causes vascular stenoses.5, 6 Based on these findings, multiple therapeutic strategies have been developed to improve vascular regenerative function in order to prevent atherosclerosis at an early stage.7-10 Nevertheless, atherosclerosis and its sequelae have remained the number one cause of global mortality in 2015.11 Recently, an increasing body of evidence has highlighted the pivotal role of extracellular vesicles (EVs) in the regulation of endothelial dysfunction, atherosclerotic plaque formation and thrombotic vessel occlusion.12 Extracellular vesicles are commonly classified into microvesicles, which have a size of 100-1000 nm and are released directly from the outer cell membrane and exosomes, which are below 100 nm in size and are released via the multivesicular endosome.13 The release of EVs allows cells to communicate with other adjacent or distant cells by transferring biologically active molecules into recipient cells.14 Interestingly, EVs emerge as major transport vehicles for microRNA (miRNA) in patient plasma, as the vesicular membrane protects the microRNA from degradation.15 Vice versa, the effects of EVs have been shown to be dependent on the miRNA content of the vesicles. miRNAs are strong regulators of gene expression and are thereby involved in multiple cellular signaling processes.16-18 Under physiological conditions, the transfer of vasculoprotective miRNAs, such as miRNA-126, via microvesicles promotes endothelial repair and atheroprotection.19, 20 The packing of microRNAs into extracellular vesicles and the release of the vesicles from their mother cells, however, depend on the state of the releasing cells. Under pathological conditions, such as hyperglycaemia, the incorporation of miRNAs in endothelial microvesicles is altered.21 In this context, it has been shown that microvesicular transfer of miRNAs, which are particularly protective for endothelial function, such as the miRNA-126 and the miRNA-26a, is decreased in diabetic patients.22 In line with this observation, the pro-regenerative effects of microvesicles are abrogated in hyperglycaemic conditions.23, 24 Therefore, it is of great importance to understand how miRNAs are packed into extracellular vesicles. In exosomes, it has been shown that RNA-binding proteins such as heterogeneous nuclear ribonucleoproteins Q and A2B1 (hnRNPQ and hnRNPA2B1) are involved in exosomal RNA export (Figure 1).25 Depending on post-translational SUMOylation, hnRNPA2B1 binds miRNAs which exhibit specific export motives in their sequences (exo-motives).26 Subsequently, the RNA-protein complex is packed into exosomes via protein sorting mechanisms, which depends on the RNA-binding ubiquitin E3 ligase and the adaptor-related protein complex 2.27 The exosomes are then released into the extracellular space and transfer the microRNAs to recipient cells. Recently, our group discovered that besides hnRNPQ and hnRNPA2B1, hnRNPU regulates miRNA packing into extracellular vesicles as well. Apart from hnRNPs, a relevant function in RNA binding and export has been proven for the Argonaute Protein 2 (Ago2) and the Y-box protein 1.28, 29 Interestingly in cancer cells, the RNA export via Ago2 can be inhibited via mutation of the KRAS-MEK signalling pathway, which underlines the relevance of EV RNA sorting mechanisms not only for cardiovascular research but for cancer research, as well.30 Besides RNA-binding proteins, ceramide signalling regulates vesicle formation and RNA sorting. Intracellular ceramide levels are primarily controlled by a group of enzymes named sphingomyelinases. Sphingomyelinases catalyse the hydrolysis of sphingomyelin to ceramide, which is the rate-limiting step of ceramide synthesis. Ceramide and sphingomyelin-1-phosphate have been detected in atherosclerotic plaques, where they contribute to plaque instability.31 Furthermore, sphingosine-1-phosphate acts as a vasoconstrictor and regulates endothelial barrier function as well as apoptosis in interaction with ceramide.32, 33 In the context of vesicle formation, ceramide facilitates the formation of lipid rafts through self-association, which serve as a basis for protein sorting complexes, such as the ESCRT.34 Additionally, ceramide triggers the budding of vesicles due to its cone-shaped structure (Figure 1). Inhibition of neutral sphingomyelinase 2 therefore causes a decline in exosomal miRNA export.35 Interestingly, a third regulatory mechanism of exosomal miRNA export has been identified: mRNA-miRNA interaction. miRNAs targeting abundantly expressed mRNAs are exported to a lower proportion than those miRNAs targeting mRNAs, which are sparsely expressed.36 This suggests that miRNA export is at least partially regulated passively via target mRNA expression (Figure 1). In summary, remarkable progress has been achieved in the characterization of molecular mechanisms, which regulate vesicle-based miRNA transfer. Nevertheless, all of the aforementioned sorting pathways only apply to a limited number of miRNA species and to certain cell types. A general understanding of miRNA packing into exosomes is still lacking. To the best of our knowledge, there is currently no direct analysis of miRNA packing into microvesicles available. The microvesicular miRNA transfer is crucial for the maintenance of vascular integrity, which is disturbed in pathological conditions, such as hyperglycaemia. Therefore, a deeper understanding of miRNA sorting and packing mechanisms into EVs is of high relevance to develop novel therapeutic strategies to combat vascular diseases. NW and FJ were supported by Deutsche Forschungsgemeinschaft (WE 4139/8-1). FJ and AZ were supported by the Medical Faculty of the Rheinische Friedrich-Wilhelms-University Bonn (BONFOR). FJ was supported by the “Familie Schambach” foundation and the German Society of Cardiology. The authors declare no conflict of interest.