Acute kidney injury (AKI) is a silent killer characterized by a sudden loss of renal function resulting in disruptionof fluid, acid-base, and electrolyte homeostasis. It is a global burden associated with a 10-fold increase inmortality rate affecting between 7-17% of all hospitalized patients. A potent contributor to AKI is ischemiareperfusion injury, which e.g., is unavoidable during kidney transplantation. A proinflammatory monocyte infiltration is a crucial event of AKI, recruited by injured kidney epithelial cells, but the signalling mechanismsbetween kidney epithelial cells and bone marrow-derived monocytes are not fully understood. Extracellularvesicles (EVs) have emerged as new important cell-cell communicators, and they might create an importantmissing link in the development of AKI. EVs are small membrane-bound vesicles released from most cell typescapable of transferring cellular protein, lipids, and nucleic acids, which are believed to be important during bothphysiological and pathophysiological conditions. However, most of our current understanding is built on in vitroexperiments or in vivo experiments injecting supraphysiological concentrations of EVs, and cell-specific EVsare challenging to isolate and differentiate. To overcome this, my PhD has focused on characterizing a novelEV reporter mouse developed in our group, that enables cell-specific isolation and tracking of EVs. Concurrently, we have focused on mechanisms important for EV secretion regulation. EVs reflect the parental cells,from which they are secreted and are found in human plasma and urine, which enable them as novel easilyaccessible candidates for biomarkers able to detect and monitor disease progression as a liquid biopsy. Celltype-specific EV secretion is, however, highly inconsistent and to fully release the potential of EV biology weneed to obtain a better understanding of the mechanisms affecting cellular EV secretion.
Study I: In study I we developed and characterized a new EV reporter cell line and transgene mouse model,using a truncated version of the tetraspanin CD9 to display extravesicular enhanced green fluorescent protein(EGFP) tag and isolate cell-specific EVs. CD9truc-EGFP expression did not modify the size or release of EVsin vitro but enabled EV isolation from cell-conditioned medium using anti-GFP immunoprecipitation. To enablecell-specific EV isolation in vivo, we then created a transgenic EV reporter mouse, by inserting a double-floxedCD9truc-EGFP in the inverse orientation, which enables Cre recombinase-dependent EGFP expression. Wecrossed our mice with mice expressing Cre in the cardiomyocytes, kidney epithelium, and ubiquitously, andobtained cell-specific EGFP expression. By immunoprecipitation, we isolated only EVs in plasma from miceexpressing Cre recombinase ubiquitously and in the cardiomyocytes. In contrast to plasma, we were only ableto detect EVs in urine from mice expressing Cre recombinase ubiquitously and in the kidney epithelium. Thisindicates that EVs are fluid compartment restricted, and our new EV reporter mouse provides a novel tool toobtain unique insight into the physiological and pathophysiological functions of EVs in vivo.
Study II: In study II we used pharmacological interventions to mimic hypoxia by stabilizing hypoxia induciblefactor-1a (HIF-1a) and increasing mitochondrial ROS production by manipulating the electron transport chain(ETC) in our EV reporter cell line. We collected cell-conditioned medium from treated cells in vitro, and quantified EV concentrations with western blotting and nanoparticle tracking analyse. We found no significant effectof different HIF-1a stabilizers, but pharmacological intervention that increases mitochondrial ROS productionincreased EV secretion.Corroborating that ROS is involved, did rotenone, which decreases electron deposition through complex Iinhibition to the ETC, and TEMPO, an antioxidant mimicking the effects of superoxide dismutase, attenuatethe ROS-associated increase in EV secretion. We used antimycin A to mimic the constrained effect of cellular hypoxia on complex III of the ETC, which in turn increased ROS-dependent EV secretion. In addition, byblocking an early step in the mevalonate pathway with pitavastatin, we blocked the synthesis of coenzymeQ10, which shuffles electrons between complex I and III in the ECT. Pitavastatin had no effect alone butpotentiated the effect of increasing oxidative phosphorylation by DCA, resulting in an additional increase in EVrelease. Blocking the mevalonate pathway early interfere with the synthesis of other components relevant toEV secretion, so we treated our cells with 4-nitrobenzoic acid that specifically inhibits the condensation ofmevalonate and tyrosine metabolites, which also resulted in ROS dependent increase in EV secretion. Ourfindings suggest that mitochondrial ROS production contributes significantly to the regulation of cellular EVsecretion rates.
Conclusion: Study I demonstrate a novel transgenic mouse model that enables future in vivo isolation andtracking of cell-specific EVs. Furthermore, we found that EVs are body fluid-restricted, indicating that they areunable to cross the glomerular basal basement membrane of the kidney. In study II, we found that pharmacological stabilization of HIF-1a is insufficient to affect EV secretion rates, while pharmacological manipulationof the ETC in the mitochondria increases EV secretion in a ROS-dependent way. These findings contribute toa better interpretation of EV data and add new knowledge to understanding the highly diverse EV secretionrates between different cell types.