MicroRNAs and extracellular vesicles in pre-clinical and clinical stroke studies

Stroke is a leading cause of death and disability worldwide and remains a largely unmet clinical need. Despite decades of pre-clinical stroke research there are only two licensed interventions: intravenous delivery of thrombolytic recombinant tissue plasminogen activator (rt-PA) within 4.5 hours of stroke or mechanical thrombectomy. However, the number of patients eligible to receive either treatment is limited. An alternative intervention is needed, one which directly address specific aspects of stroke pathophysiology. Through their ability to alter the expression of multiple genes involved in stroke pathophysiology microRNAs (miRNA or miR) offer a novel therapeutic intervention. miRNA expression is altered both in experimental stroke and in patients with stroke. It was initially hypothesised that modulation of specific dysregulated miRNAs would be therapeutically beneficial in pre-clinical experimental ischaemic stroke. Recently, active transport of miRNAs in extracellular vesicles (EV), such as exosomes, has been demonstrated pre-clinically between cells in atherosclerosis and cardiac hypertrophy disease settings. It was therefore hypothesised that miRNAs packaged in exosomes would differ between patients with stroke and patients without stroke, raising the potential for novel exosomal miRNAs to be used as biomarkers or therapeutic agents for modulation. In Chapter 3, investigations were carried out to test the hypotheses that modulation of either miR-494 or miR-21 would be therapeutically beneficial in pre-clinical in vitro models of stroke. While miR-494 expression was unchanged in brain tissue of spontaneously hypertensive stroke prone rats (SHRSP) harvested at either 24 or 72 hours following transient middle cerebral artery occlusion (tMCAO) its expression was successfully up-regulated in B50 neuronal and GPNT cerebral endothelial cell lines following delivery of miR-494 mimic in combination with siPORT, a lipid based transfection reagent. mRNA expression of putative miR-494 target genes (PTEN, MMP2 and MMP9) was investigated post-miR-494 modulation but there was no obvious change in their expression. Modulation of miR-494 expression did not appear to be therapeutically beneficial (or detrimental) when assessed by a cell survival assay (MTS). As the balance of evidence did not indicate that miR-494 would be a suitable target for modulation in experimental stroke subsequent similar experiments investigated the therapeutic potential of miR-21. Its expression was significantly increased in SHRSP brain tissue at 72 hours following ischaemic stroke. miR-21 expression was successfully increased in cerebral endothelial cells following delivery of miR-21 mimics (with siPORT). mRNA expression of putative target genes (PDCD4 and PTEN) was unchanged following miR-21 modulation and cell survival (assessed by MTS assay) was unaffected. Subsequent experiments looked at vessel reactivity of aortae taken from miR-21+/- and miR-21-/- mice in comparison to wild type (WT) mice. Treatment of vessels with L-NAME to block endogenous nitric oxide (NO) bioavailability resulted in unopposed contraction to U46619 in WT mice while there was no change in contraction in miR-21-/- mice aortae, consistent with reduced basal NO bioavailability, and a detrimental phenotype associated with the loss of miR-21 expression. As the data generated in this study were primarily neutral and gave no indication that either miR-494 or miR-21 would be therapeutically beneficial in the setting of ischaemic stroke, subsequent studies focussed on investigating exosomal miRNA in ischaemic stroke. In Chapter 4, exosomal miRNA expression was profiled in blood samples from stroke patients and subsequently in pre-clinical rodent stroke models. Patients with suspected stroke were recruited and a blood sample taken at 48 hours post-stroke. All participants gave full informed consent and the study was approved by the Scotland A Research Ethics Committee. Exosomes were isolated from 200 μL serum before RNA was extracted. A miRNA microarray was performed (OpenArray™ platform) on samples from 39 patients. Validation of results was performed by real-time quantitative polymerase chain reaction using samples from 173 patients to determine the expression levels of specific miRNAs. Microarray experiments identified 26 exosomal miRNAs that were significantly dysregulated between stroke and non-stroke patients or between specific TOAST subtypes and non-stroke controls. Of these, changes in 13 miRNAs were validated in the larger cohort and levels of 9 miRNAs (-27b, -93, -20b, -17, -199a, -30a, let-7e, -218 and -223) were found to be significantly increased in definitively diagnosed stroke patients as compared to non-stroke patients. Differences in exosomal miRNA expression were observed between TOAST subtypes with small vessel disease patients consistently having the highest levels of these miRNAs. miRNA expression did not correlate with baseline or day 7 NIHSS score, although there was a trend towards patients with better functional outcome post-stroke (as assessed by modified Rankin Score at 1 month) having a higher level of some exosomal miRNAs. Subsequently total and exosomal miR-17 family (miRNAs -17, -93 and -20b) expression was investigated in pre-clinical models of hypertension and stroke. Total circulating miR-17 expression was unchanged between the serum of normotensive WKY and hypertensive SHRSP rats, whilst exosomal miR-17 expression was significantly increased in SHRSP vs. WKY. miR-17 family expression was unchanged in peri-infarct brain tissue of SHRSP at both 24 and 72 hours post-tMCAO. Experiments profiling total and exosomal circulating miR-17 family expression in serum of SHRSP post tMCAO or permanent MCAO revealed that expression was variable and changes observed were not significant. Cellular expression of miR-17 family miRNAs was unchanged following hypoxic challenge in neuronal, glial and cerebral endothelial cell lines and exosomal miRNA expression was highly variable, with no changes detected as significant. This study both identified and validated (for the first time) changes in exosome packaged miRNA expression in patients with stroke across differing stroke subtypes. The pre-clinical experimental findings corroborate the human data and support a functional role for these findings. In Chapter 5 exosomal packaged miR-17 family miRNAs were delivered in pre-clinical models of ischaemic stroke (both in vivo and in vitro) to test the hypothesis that they would be therapeutically beneficial following in vitro hypoxic challenge or in vivo experimental stroke. Bioinformatics analysis highlighted a number of important target genes implicated in stroke pathophysiology for each miRNA including genes involved in the regulation of the cellular response to stress, apoptosis and angiogenesis. miRNAs were artificially loaded (by electroporation) into EVs harvested from SHRSP brain. While miRNA loaded EVs did not successfully modulate miRNA expression either in vivo or in vitro it is believed that this is a result of technical issues with the loading of the miRNAs into the EVs. This study should be repeated when miRNAs have been successfully loaded into EVs, as these experiments remain of interest. In summary, the findings presented in this thesis confirm that packaging of miRNAs into exosomes is significantly dysregulated in stroke patients and that as a result the circulating exosomal miRNA profile is altered. This will direct future studies looking into paracrine signalling in the setting of stroke and the modulation of specific miRNAs as a novel therapy in the setting of experimental stroke