Final Report Summary - EPIMIRNA (MicroRNAs in the Pathogenesis, Treatment and Prevention of Epilepsy)
Executive Summary:
Epilepsy is a brain disease characterized by recurring, unprovoked seizures that affects about 6 million people in Europe. Temporal lobe epilepsy is the most common syndrome in adults and the most prone to drug-resistance. A number of priorities exist to improve the lives of patients with temporal lobe epilepsy, including the need for more effective anti-seizure drugs and disease-modifying therapies. Evidence has emerged that epilepsy and epileptogenesis (the process underlying its development) feature large-scale changes in gene activity. If the brain molecule(s) that drive or coordinate multiple genes could be identified, this would significantly improve our understanding of epileptogenesis and may lead to new thinking about the treatment of epilepsy. If the same brain molecules can be detected in a blood sample or other fluid, this could provide new ways to diagnose epilepsy or predict response to therapies.
The EpimiRNA project chose to focus on a newly-discovered molecule that coordinates gene networks - microRNA (miRNA). These are short stretches of RNA – closely related to DNA – which function in cells by attaching to the much longer messenger RNA (mRNA) transcripts that are the instructions for making proteins. The miRNAs fine-tune protein levels in all cells, including the brain, and a particularly important feature is that each miRNA regulates many different mRNAs; they are the “conductors of the protein orchestra”. The EpimiRNA project began in 2013. Originally 16 partners from eight different European countries plus the USA and Brazil, the consortium brought together for the first time experts in pre-clinical and clinical epilepsy research with experts on human genetics, chemistry, drug discovery and mathematical modelling. The project was complemented by research-active small and medium enterprises (SMEs) developing therapeutics, diagnostics and devices. The EpimiRNA project had five main objectives. We would measure the levels of every active miRNA in the epileptic brain. Then, develop mathematical models and technologies to explain how they controlled brain excitability. We would look in blood and brain fluids to see if any miRNAs were unique to epilepsy or could predict responses to therapies. We would look for variation in the sequence of miRNAs in the genetic code of 3500 patients and controls. Finally, we would test whether altering amounts of miRNA in the brain could reduce seizures or reverse epilepsy in models of epilepsy.
EpimiRNA successfully completed all its main objectives. We generated searchable databases containing over 5 billion data-points cataloguing all functional (active) miRNAs in three models of epilepsy and brain tissue from patients. Epilepsy caused changes to nearly half of all the brain-expressed miRNAs. We developed new models to predict targets of miRNAs and obtained direct evidence that the protein landscape of the epileptic brain is shaped by miRNAs. We identified miRNA “biomarkers” in blood and brain fluid and developed a prototype device to quickly detect miRNAs that can support the diagnosis of epilepsy. We found a new site deep within the brain that, when electrically stimulated, protects against epilepsy development. We found that temporal lobe epilepsy patients do not have rare variants or mutations in miRNAs (or the enzymes which produce them) in their genomes. Finally, we have identified 10 new molecules that produce anti-seizure effects in pre-clinical models of epilepsy. One of these stops epileptic seizures after a single injection, for which a patent was awarded. Two of our compounds are now in pre-clinical development as treatments for epilepsy. Our industry partners were able to bring compounds toward pre-clinical development, test new brain stimulation devices, refine and improve RNA technology, develop new data storage and management systems, and build a prototype device to simultaneously record brain activity and deliver therapies to patients.
EpimiRNA has generated important breakthroughs and advances in our understanding of miRNAs in the development of epilepsy and demonstrated this can support diagnosis and treatment (or even cure). We urge further research to understand miRNA control of brain excitability and translational and applied aspects, including diagnosis and treatment epilepsy.
Project Context and Objectives:
1.2.1 Background & Aims
Epilepsy is the most common serious neurological disorder for which there is no cure (European Written Declaration on Epilepsy, 2011). The disease is characterized by recurring, unprovoked seizures. There are 6 million people with active epilepsy in Europe and 300.000 further cases each year. The costs are estimated at nearly €14 billion a year in the EU. Temporal lobe epilepsy (TLE) is the most common form of epilepsy in adults. TLE can develop as a result of brain trauma, infection, status epilepticus (prolonged, continuous seizure), and is frequently associated with hippocampal pathology (hippocampal sclerosis; HS). People with epilepsy have increased mortality, experience high levels of unemployment and under-employment, and are exposed to stigma and prejudice. Epilepsy – particularly when poorly controlled by medication – not only damages health, but disrupts many other aspects of living, imposing very significant physical, psychological and social burdens on individuals, families and caregivers. In addition, diagnosis of epilepsy remains principally based on clinical history and examination. Electroencephalogram (EEG) recording, brain imaging and genetic testing provide important supports but mis-diagnosis rates remain unacceptably high.
A number of priorities exist to improve the lives of patients with epilepsy, including the need to discover anti-epileptogenic and disease-modifying therapies. The identification of measurable diagnostic indicators of epilepsy or response to therapies (e.g. biomarkers) and the epileptogenic process would also transform discovery of disease-modifying therapies and the diagnosis of epilepsy in future. However, no molecules to date could account for the complex gene expression landscape of the epileptic brain and be both biomarker and therapeutic targets. Do such molecules exist?
The EpimiRNA project focused on a completely new class of molecule that had recently been discovered inside cells called microRNA (miRNA). These are short stretches of RNA – the chemical cousin of DNA – which function by binding onto the much longer protein-coding messenger RNA (mRNA) transcripts that are the instructions for making proteins (DNA is the blueprint for the mRNA which is then read and a protein is produced). Through a series of important breakthroughs in cell biology, it was revealed that miRNAs fine-tune gene expression (e.g. proteins levels) in all cells. A particularly unique feature of miRNAs is that they are multi-targeting - a single miRNA can bind dozens of different mRNAs-, producing effects ranging from mild to strong suppression of protein levels. The human genome contains about 2000 different miRNAs. They are initially produced as longer precursor RNAs before undergoing a series of enzyme reactions that result in a duplex miRNA. One strand of this – the mature miRNA – is selected and loaded into a binding pocket by Argonaute (AGO), a protein that function as the vehicle of miRNA-directed control of gene expression.
The AGO protein next begins a search process, trafficking along mRNAs until it finds a site where the miRNA matches the complementary sequence in the mRNA. If this happens, the mRNA is either prevented from being read or is degraded by the cell. The brain contains more miRNA than any other organ and malfunction of the miRNA system resulted in brain mal-development. Experiments soon clarified why: miRNAs control about half of all genes and ensure protein levels are tightly regulated, including those that ensure the physical connections between neurones (synaptic communication).
In 2010, reports emerged that levels of certain miRNAs were altered in a part of the brain – the hippocampus – long-known to be a trigger point for seizures. The same year, studies in rats reported that levels of certain miRNAs changed in the blood after a seizure. The next two years saw the first evidence that blocking an individual miRNA using synthetic DNA-like oligonucleotides (“antagomirs”) could protect the brain from damage or seizures. Around the same time, an antagomir – miravirsen - successfully completed a clinical trial (for Hepatitis C), heralding a new era in medicine – therapeutic miRNA targeting.
The EpimiRNA project began in 2013 – originally 16 partners from eight European countries, the USA and Brazil. The consortium was new, with few of the partners ever having worked together before, and had a very innovative approach: A core team of experts in pre-clinical and clinical epilepsy research, then experts in human genetics, cell biology, RNA detection (sequencing), protein analysis and mathematical modelling. The EpimiRNA consortium was complemented by research-active small and medium enterprises (SMEs) developing RNA therapeutics (InteRNA), and diagnostics and devices for the treatment of epilepsy (Dixi, Cerbomed, Bicoll, BCPlatforms). Devices for non-pharmacological treatment of epilepsy – brain stimulation techniques including vagus nerve and deep brain stimulation (VNS, DBS) – had become of increasing clinical importance, yet lack of understanding of their mechanisms of action limited our knowledge of which patients benefit from their use. EpimiRNA aimed to track miRNA changes after seizures (including in humans), and in response to treatment, explain how miRNA influences epileptogenesis (e.g. epilepsy development) and ictogenesis, investigate genetic variants in TLE patients to learn if this caused their epilepsy, and ultimately use the discoveries to establish miRNA-based therapeutics and biomarkers to track, treat and prevent epilepsy.
1.2.2 Main objectives of the project
The overarching objective of the EpimiRNA project was to understand the contribution miRNA makes to the development and maintenance of the epileptic state, their potential to diagnose the disease, and opportunity to target these to treat or prevent epilepsy.
Objective 1 Identify conserved changes in functioning miRNAs in epileptogenesis and determine the mechanism(s) by which miRNA changes contribute to epileptogenesis
Gaps in our understanding of the epileptogenic process strongly limited our ability to develop more effective treatments. The multi-targeting properties of miRNAs represent a novel means to coordinate epileptogenesis but it was not certain that miRNA were really shaping gene expression in epilepsy. EpimiRNA aimed to perform a systematic, unbiased and quantitative analysis of functioning miRNA in three of the best animal models of epilepsy. EpimiRNA would determine the impact of miRNA changes on the “proteome” – the complete collection of all proteins in a cell or tissue - and measure how miRNA manipulations altered the physical structure and function of different brain cell types. Together, our approach would produce a major advance in our understanding of which miRNAs contribute to epileptogenesis and explain the underlying mechanisms.
Objective 2 Identify the miRNAs that are functional in the human TLE brain and evaluate how non-pharmacological interventions including brain stimulation modulate miRNA
Although human data on miRNAs in epilepsy existed before EpimiRNA, small group sizes, incomplete coverage of miRNAs and lack of evidence of function limited insights; a systematic approach was needed. Additionally, biomarkers of the epileptogenic process would be essential for the development of disease-preventative therapies in the future. Circulating miRNAs offer novel, mechanistic biomarkers of epilepsy but it was unknown if levels in blood or other biofluids were altered in patients or responsive to therapy. EpimiRNA would undertake a study to identify and validate miRNA biomarkers of epilepsy and identify the targets of the miRNAs in brain tissue. EpimiRNA would also investigate whether miRNA levels change in biofluid samples from patients receiving novel therapies - brain stimulation - for treatment of seizures, and try and find new brain stimulation sites. This would not only identify potential biomarkers of pharmacoresistance (or seizures) but would also provide an opportunity to improve prediction of efficacy of brain stimulation (patient stratification).
Objective 3 Characterize genetic variation in miRNAs in human TLE
Human epilepsy genetics has been a powerful tool for discovering genes responsible for inherited and new forms of epilepsy and can sometimes predict adverse reactions to anti-epileptic drugs. But for the majority of patients, variation and errors in traditional genes such as ion channels do not explain their epilepsy. Sequence variation in miRNAs, or in the various genes involved in miRNA biogenesis or function could be causally important in human epilepsy. EpimiRNA would undertake the first targeted sequencing focused on miRNAs and on the genes involved in their biogenesis, specifically in TLE with HS, a major syndrome prone to pharmacoresistance. Focusing on a single, well-defined population would maximize the likelihood that we identify significant risk variants and the large-scale collaborative nature of EpimiRNA could ensure high numbers of samples could be analysed.
Objective 4 Use systems biology to integrate miRNA bioinformatic and functional data to explain how miRNA expression changes trigger epileptogenesis
Predicting effects of miRNAs on gene expression is a major challenge - single miRNAs often produce only small changes in protein production but when several miRNAs target the same mRNA additive effects can produce strong suppression. There are often sharp, bi-directional changes in miRNA levels within specific cell types adding to the complexity of understanding miRNA effects on epileptogenesis. This had not been factored in any epilepsy studies to date. A key task of EpimiRNA would be to establish biologically meaningful connections and link these to disease processes by generating mathematical models that take account of the behaviour of entire gene pathways. Together, this would allow us to better predict the impact of miRNAs on brain function, seizures and identify novel targets.
Objective 5 Identify novel miRNA-modulating molecules as future therapeutics for epilepsy
MiRNAs represent an entirely new target for the treatment and prevention of epilepsy. Oligonucleotides targeting miRNAs have been tested in humans and these so-called “antagomirs” are highly potent and long-lasting - a single dose inhibits a miRNA for more than one month. Just how many high-value miRNA targets exist for epilepsy was, however, unknown. Reports were few and no study had validated an effect in another model. EpimiRNA would develop and test miRNA inhibitors using antagomir and other approaches, targeting those miRNAs causally important for epileptogenesis or attempting to reverse already-established epilepsy. A further innovation would be to measure miRNAs in biofluids in the animals during the epilepsy monitoring studies and record how gene expression changes in response to miRNA-based interventions to understand mechanisms of action.
1.2.3 Work Strategy
The discovery of miRNA un-locked a previously unknown layer of gene expression control with important implications for epilepsy. Changes in levels of miRNAs or mutations might be a key contributor to the altered gene expression landscape in the epileptic brain. Targeting miRNA could be a way to protect the brain from seizures.
Detecting circulating miRNA in blood or other biofluids could help with diagnosis. The five Objectives were developed into tasks that fell into a series of 11 work packages (WPs). Each WP had a leader, an expert in the main scientific and technical demands of the experiments, along with several other teams bringing expertise to ensure the research was achieved:
WP1: Characterise miRNAs in brain and biofluids from three different animal models to identify the miRNA changes that occur during epileptogenesis.
This work package was led by the Coordinator’s team at RCSI and UNIMAR and UNIVR generated the samples from the different preclinical models. The AU team analysed miRNA in brain tissue whereas RCSI analysed blood miRNA. SDU performed protein analysis.
WP2: Characterise the miRNA profiles in brain tissue from epilepsy patients and miRNA levels in biofluids before and after seizures in patients. Also, develop a new combined intracortical EEG-microelectrode.
Led by Prof Hamer’s group at UKER, the device development was performed with DIXI and collaborations across all clinical sites provided the samples with miRNA analysis and bioinformatics performed by RCSI teams.
WP3: Define miRNA changes following experimental brain stimulation as well as determine whether brain stimulation in patients or volunteers produces a miRNA signature that can be used as a biomarker.
Led by clinical co-coordinator, Prof Rosenow, samples were analysed by AU and a clinical trial performed using Cerbomed’s brain stimulation device.
WP4: Identify the gene targets of the identified miRNAs and assess the impact of manipulating these miRNAs on animal behaviour and brain excitability.
Prof Pasterkamp’s team with InteRNA, UCL’s Prof Schorge, UNIMAR and RCSI studied how various miRNAs control brain structure and function at a cell and subcellular level.
WP5: Determine the genetic variation in miRNAs, their biogenesis enzymes and the hippocampal target genes in epilepsy patients.
Prof Goldstein’s and Prof Lopes-Cendes’ teams performed the genetic study of miRNA variation in epilepsy patients, with DNA provided by multiple partners and collaborators.
WP6: Develop an integrated data storage and retrieval platform for the EpimiRNA datasets and support data mining, hypothesis testing and biomarker identification.
Prof Kjems and team at AU performed RNA-sequencing and developed several technologies including iCLIP. Data storage requirements were provided by partner BCPlatforms.
WP7: Develop computational models that can predict effects of miRNA changes on network excitability which can be fed back for testing and biomarker prediction.
Prof Prehn’s team at RCSI supported the project’s bioinformatics needs, developing new tools to search for targets of miRNAs.
WP8: Perform in vivo manipulations of miRNAs and assess the impact on epileptogenesis in animal models.
Prof Schorge’s team with RCSI tested numerous miRNA inhibitors and other compounds and explored how these new drug types affect seizure susceptibility and epilepsy in models.
WP9: Develop in vitro screening assays to identify natural compounds that display miRNA inhibitory properties that could be developed as anti-seizure drugs.
Dr Lamottke and Bicoll led the investigation of plant-based compounds for epilepsy and developed an anti-seizure small molecule with the teams at UCL and RCSI.
WP10, WP11: Ensure the dissemination of EpimiRNA results, inter-site training and exploitation of discoveries as well as project management.
Coordinated by RCSI with support from ARTTIC, this focused on scientific and communication achievements and the smooth running of this large and complex project.
Project Results:
1.3.1 Overview
The main results of the EpimiRNA project fall into six categories according to the main objectives. These are summarised as:
(1) Identification of novel miRNAs in epilepsy. The EpimiRNA project has generated genome-wide datasets identifying all functioning miRNA in experimental and human epilepsy. This includes every miRNA in the rodent hippocampus and how its level changes during epilepsy development in three different models and every miRNA in the human epileptic brain and its target(s).
(2) Deep mechanistic understanding of how miRNAs change neuronal networks in the brain. Our studies have identified the targets of numerous new miRNAs and mechanisms by which this regulates brain excitability. By locating the miRNA within causal pathways and identifying their target(s) our results explain how miRNA changes influence the epileptogenic process and lead to lasting changes in brain excitability. The data have proven effective for predicting novel targets for anti-epileptogenesis strategies.
(3) Identification of miRNA biomarkers of epileptogenesis and seizures. We have identified miRNAs in blood and cerebrospinal fluid that distinguish TLE patients from controls and other neurological disorders. These findings could be used to develop tests to support diagnosis, help predict disease risk and course/prognosis, direct treatment and avoid adverse side-effects. miRNA biomarkers of non-pharmacological treatments (e.g. brain stimulation) were found and may improve use parameters to help patients better control their seizures with fewer side effects.
(4) Novel therapeutics by targeting miRNAs. We have tested more than 10 different miRNA inhibitors (antagomirs) during our project, of which more than half produced anticonvulsant effects in animal models. Two of the antagomirs prevented spontaneous recurrent seizures. This has led to a patent and preclinical development of a miRNA-based disease-modifying therapy for epilepsy, with the hope to reach clinical development in future.
(5) Investigation of genetic variation in miRNAs. By searching 3500 patient and control genomes, we have discovered that mutations or variations in miRNAs, their biogenesis genes and their main targets in the hippocampus are unlikely to cause human TLE. This resolves a major question in our pursuit of the genetic causes of epilepsy. It could provide potential healthcare savings in terms of genetic testing and refines genetic counselling so patients have a better understanding of the cause and prognosis of their condition.
(6) Technological advances and resources. We have made a number of important technological advances and have delivered training of these, including techniques to characterise miRNAs with small RNAseq, as well as generating a systems biology framework for miRNA. EpimiRNA has delivered advances in oligo/DNA-based and traditional small chemical scaffolds. The biomarker findings have driven development of new technology for rapid, direct, detection of epilepsy-associated miRNA. EpimiRNA has generated research databases and a new data integration platform for RNA-seq, genetics and proteomics.
1.3.2 Objective (1) Identification of novel miRNAs in epilepsy
One of the most ambitious, multi-partner collaborative projects within EpimiRNA was to catalogue every miRNA that was functional in the epileptic brain. We would do this for three common animal models of epilepsy and the human brain. This ambitious goal was fully achieved. We produced a dataset on the relative amounts of every miRNA in the brain of a rat or mouse and whether and when this changes in epilepsy. We have identified the target(s) of every miRNA in the human epileptic brain.
Measuring every functional miRNA in the brain: AGO-Seq
The approach the EpimiRNA project took to measure every functioning miRNA required us to pull down the protein responsible for miRNA function. The protein is called AGO (argonaute). This is done by mixing an antibody raised to detect the AGO protein with the brain sample. This can then be separated and the miRNA extracted and identified. For the identification step, we were able to use RNA sequencing, a powerful technique that identifies all miRNA sequences and can tell you the relative amounts of each. The team at AU developed the technique and, once perfected, began processing tissue for AGO-Seq. Meanwhile, the teams at RCSI, UNIMAR and UNIVR were generating brain tissue samples from each of their animal models. We wanted to know what every miRNA was doing at every important step during the development of epilepsy and we therefore included time points that corresponded to when epilepsy was developing, the day of the first spontaneous seizure and when the animals had active epilepsy. We monitored every animal, so we knew precisely when their epilepsy emerged and how many seizures they were having. The brain tissue samples were finished during the second year of the project and were shipped to AU for analysis. By the end of the third year we had our complete miRNA functional catalogue.
Comprehensive profiles of functional miRNAs in three animal models of epilepsy
Altogether, the AU team generated 240 datasets – over 3 billion data-points - on the relative amount of functional miRNAs in three rodent models, at multiple different time points and three different regions of the hippocampus (the brain region involved in human TLE). One of the first checks we made was to look for miRNAs known to be abundant in the hippocampus. One of these, let7c is especially enriched. When we checked levels we found it to be among the highest of any miRNA (~15 on our scale that ran from 0 to 17). miR-9 is another abundant miRNA in the brain. Levels of miR-9 were also around 15 on our scale. Then we checked for a less abundant miRNA, miR-134-5p (one we had previously linked to epilepsy). Levels of miR-134-5p averaged around 7 on the scale. Finally, we looked up miRNAs that were only expressed outside the brain. One of these is miR-122, a liver-enriched miRNA. We couldn’t detect any miR-122 in any of the brain samples at any time in any model. We also looked at levels of miRNAs between the models. We had two mouse models but the particular strain of mouse was different. We would still expect, however, that the amount of any miRNA should be quite similar. This is exactly what we found. Taking miR-9 for example, in the RCSI mouse the average level was ~15.2 on the scale. In the mice from UNIVR it was 14.9. Another, miR-124-3p, was found at 11.7 in both mice.
We now started to look for miRNA changes during epilepsy development. We first checked for some of the miRNAs that had already been linked to epilepsy. We started with miR-146a and miR-132, which had been reported by multiple teams to be upregulated in epilepsy. We found both to be upregulated in all three models, but the timing was quite different. miR-132 levels went up almost immediately whereas miR-146a was much later.
We also had data on the amounts of different miRNAs within three different parts of the hippocampus. This is useful because the hippocampus of humans with TLE shows patterns of neuronal loss and glial scaring that are not uniform. Within the EpimiRNA data set we found sets of miRNAs that were very evenly expressed across the three regions as well as examples of miRNAs that were at higher levels within one or other part.
In addition to providing a rich dataset on all functional miRNAs in the normal brain and subfield-specific enrichment, the project generated findings on miRNA levels in epilepsy. Data from analysis of the mouse model from RCSI showed there were 476 miRNAs detected. This is close to the approximately 500 known miRNAs in the mouse genome. This confirms the brain as the most miRNA-enriched organ. From among the detected miRNAs in the RCSI model, 225 were altered at one or more time points during epilepsy. Thus, epilepsy induces changes to nearly half of all the brain-expressed miRNAs.
The results from the other two models were similar. From the UNIVR mouse model we detected 481 distinct miRNAs and of these, 291 showed a change in the epilepsy model. In the rat model we detected 419 of which 147 were altered in epilepsy. We were also able to study what happened to different miRNA families - sets of miRNAs produced from the same precursor DNA sequence. We found strong evidence that miRNA families are co-regulated in the epilepsy models.
Finally, we focused on miRNAs that were common to all three models at two main time points. The latent period (the period of time after the insult has occurred but before a first epileptic seizure occurs) and the chronic epilepsy period. We found eight miRNAs were common to all three models in the chronic epilepsy phase, all of which were identical in humans. Some were already known to us, namely miR-132-3p and miR-146a-5p. Then we had six miRNAs that were potentially novel. These were selected for functional testing.
Mapping the brain targets of miRNAs in human epilepsy
In parallel with the above efforts, the UNIMAR and AU teams started a project to map all the miRNAs in the human epileptic brain and their targets. We wanted to know which miRNAs were present in the human epileptic brain and what they targeted. This could not be done in the same way as the previous study and needed a new technique, called iCLIP. The principle is to bind together the miRNA to whatever it is targeting, extract the complex and then identify both. The difficulty was that iCLIP uses a number of complex chemical processing steps and many of the early efforts failed. Eventually, during the third year of the project the AU team fixed the technical problems and it worked. The team then took a set of human brain samples from TLE patients and mapped every miRNA to every targets.
The result is the first ever complete catalogue of miRNAs in human epilepsy with their targets. This unique resource allows EpimiRNA partners (and eventually any researcher) to look up any miRNA of interest and find what it binds to or look up a target of interest and determine which miRNAs are binding to it. The results are also semi-quantitative so more abundant miRNAs tend to have higher numbers of targets linked to them. The dataset can also be used to generate new approaches to drug therapy. For example, you could look up miRNAs that are targeting a particular under-produced protein in epilepsy and then deliver a miRNA inhibitor to the tissue to increase protein levels.
In addition, we used small RNAseq to identify the differentially expressed miRNAs in human resected brain tissue (the iCLIP technique tells you what the miRNA is attached to but is not a suitable method to determine how much of a particular miRNA is present. For this we compared hippocampus from samples with sclerosis versus those without. The results were the identification of numerous differentially expressed miRNAs including known as well as novel miRNAs.
The human miRNA-seq data was subjected to differential expression analysis, showing 85 miRNAs significantly differentially expressed between cases of temporal lobe epilepsy cases with and without hippocampal sclerosis (TLE-HS vs. TLE-nonHS). This included lower levels of miR-22-3p in the sclerotic samples, consistent with a protective role (which we experimentally verified in Jimenez-Mateos et al. (2015)).
Main conclusions and future objectives: EpimiRNA identified and measured all known miRNAs in the rat, mouse and human brain in epilepsy. We identified the targets of many of these miRNAs. The results are the first ever spatiotemporal atlas of functional miRNA changes in epilepsy. This advances knowledge of miRNAs, emphasizes the profound changes to miRNA in epilepsy and can be used to select miRNA to target for seizure control.
1.3.3 Objective (2) Deep mechanistic understanding of how miRNA changes de-stabilise neuronal networks in brain
Introduction
At the time the EpimiRNA project started, a handful of miRNAs had been found to influence seizures in rodents. This was based on observations of the effects of blocking or boosting the amount of the miRNA in the brain and recording a change in severity of seizures. Nobody knew at the time the mechanism by which a given miRNA could exert such an influence. This is important because understanding mechanisms of action can lead us to further drug targets. Even if the miRNA itself is not druggable, perhaps we can design drugs that work on the targets (proteins) instead. However, studying miRNA mechanisms is challenging because each miRNA has many possible targets. A miRNA action in a seizure model may be due to strong effects on one target or weak effects on many. EpimiRNA set itself the challenge of determining the mechanisms of as many epilepsy-associated miRNAs as they could. Overall, the project solved the molecular mechanism of seizure effects for three miRNAs, partially solved the mechanism of several more and for some others we now have strong leads but experimental proof will require further efforts.
miR-22 regulates epileptic seizures by suppressing the P2X7 receptor
Partners in EpimiRNA were the first to solve the molecular mechanism by which an individual miRNA regulated epileptic seizures. The RCSI team had discovered miR-22-3p in part of the brain circuit involved in seizures which normally showed no signs of pathology. We reasoned that miR-22-3p might protect this part of the brain and we were able to prove this. Blocking miR-22-3p using an antagomir resulted in mice developing a more severe form of epilepsy than normal (in the RCSI model, where the agent kainic acid (KA) is injected into the amygdala region of the brain). We also noticed that the brains of mice given the miR-22-3p inhibitor had an increased inflammatory response. If miR-22-3p worked to block brain inflammation, could we find the target(s)? Unfortunately, miR-22-3p has one of the highest numbers of predict targets of any miRNA and is made in every cell type in the brain. It would be impossible to guess which target of miR-22-3p was important - we needed to actually identify what miR-22-3p was binding to in this part of the brain. For this, we turned to our AGO pull-down technique and found one target of miR-22-3p attached to AGO with a known role in inflammation: P2X7 receptor. This protein is found on the outside of certain cells and is normally inactive. But, if enough ATP binds to it (which can happen if nearby cells are damaged), the receptor signals to release powerful inflammatory molecules.
If miR-22-3p normally blocks the P2X7 receptor, this could explain the increase in inflammation in the miR-22-3p-inhibited mice. Promisingly, we found higher levels of the P2X7 receptor in the mice given the miR-22-3p inhibitor. But this did not prove that the P2X7 receptor was the actual culprit. To prove the mechanism, we turned to a genetically-altered mouse that lacked the P2X7 receptor. When we injected the miR-22-3p inhibitor into these mice, we no longer saw the increase in seizures. We also showed that when the miRNA inhibitor was injected along with a drug that blocks the P2X7 receptor, we again didn’t see any change in seizures. So, the mechanism by which blocking miR-22-3p increased epilepsy had to be because of an increase in the P2X7 receptor. A final and unexpected discovery was that inhibiting miR-22-3-p regulated numbers of neurons in the hippocampus and their physical characteristics. We published three papers on these findings:
Jimenez-Mateos et al. microRNA targeting of the P2X7 purinoceptor opposes a contralateral epileptogenic focus in the hippocampus. Sci Rep 5:17486 (2015)
Engel et al. A calcium-sensitive feed-forward loop regulating the expression of the ATP-gated purinergic P2X7 receptor via specificity protein 1 and microRNA-22. Biochim Biophys Acta Mol Cell Res. 1864: 255-266 (2017). Engel et al. MicroRNA-22 controls aberrant neurogenesis and changes in neuronal morphology after status epilepticus. Front Mol Neurosci (accepted)
miR-324-5p regulates seizure initiation by suppressing a potassium channel
Potassium channels are critical for neurons to be able to signal to one another. By allowing movement of the positively charged potassium ion across the cell’s outer membrane, they help generate an electrical charge. This same movement also re-sets the electrical charge after a signal has been sent. Not surprisingly, several forms of epilepsy have been found to be caused by mutations in the genes that code for potassium channels. EpimiRNA partners, together with a team from Emory University led by Dr C Gross, uncovered a key role for a miRNA in setting the levels of one of these channels, called Kv4.2 in the brain. Together, we showed the channel is controlled by miR-324-5p. When the miRNA is removed the amount of the channel on the surface of brain cells increases. We found that during a seizure, miR-324-5p blocks enough Kv4.2 from getting to the surface. As a result, seizures are worse in mice in which miR-324-5p is inhibited. Blocking miR-324-5p had no effect in mice lacking the gene that coded for Kv4.2 (Kcnd2), solving the mechanism. Researchers are now exploring methods for delivering miR-324-5p as a potential approach to seizure suppression. The study was published in the fourth year of the project:
Gross et al. MicroRNA-mediated down-regulation of the potassium channel Kv4.2 contributes to seizure onset. Cell Rep 17, 37 – 45 (2017).
Seizure-suppressive effects of antagomirs targeting miR-134-5p are mediated via Limk1
The most potent anti-seizure effects we have observed for a miRNA inhibitor are those that block miR-134-5p. As we worked on answering key remaining questions about our antagomir targeting miR-134-5p (Ant-134), it was important to determine the mechanism of the seizure-suppressive effects. We already had a good idea of the probable target. Lim kinase (Limk1) is a key enzyme that controls the shape of contact points between neurons, where signals pass from one cell to another. The mouse and rodent Lim kinase have a predicted binding site for miR-134-5p and earlier studies in cell models had showed that when Limk1 is blocked we lost the protective effects of Ant-134. But was this how Ant-134 worked in the intact mouse brain? This was difficult to prove because we did not have a genetic model of Limk1 deficiency and there were no drugs selective for the enzyme. EpimiRNA therefore took a new approach. We designed a stretch of DNA-like molecules (Gapmers) that would stick perfectly to the Limk1 message. The normal reaction of a cell to the formation of such a duplex is to degrade the targeted message. We injected our Limk1-targeting molecule into mice along with Ant-134. When we did this we lost the anti-seizure effects of Ant-134. This proved the anti-seizure effects of Ant-134 depended on protecting Limk1. These findings are currently being prepared for publication.
The human Limk1 molecule does not contain such a good binding site for miR-134-5p. This could mean that other targets of miR-134-5p will be important in human tissue. We explored this idea by analysing the protein changes in the hippocampus of mice given Ant-134. This revealed a select number of proteins change after Ant-134. These could also be important to the mechanism of Ant-134 or relevant to whether or not the therapy will translate to humans.
miR-129-5p is a novel target for seizure control via a calcium pump and RNA binding protein
So far, it was mostly single gene targets that underlay the mechanism by which many of our miRNA inhibitors produced seizure-suppressive effects. This has changed thinking in the field about how miRNAs work, at least in the brain. However, this turned out not to be the case for the final miRNA for which EpimiRNA dedicated major efforts to elucidate the molecular mechanism. We had found that miR-129-5p was one of the most consistently upregulated miRNAs in our animal models. It was a neuron-expressed miRNA and, like miR-134-5p, was important for shaping the physical structure of neurons. No other team had yet looked at this miRNA. We could show that blocking miR-129-5p in mice strongly reduced seizures in the mouse IAKA model - we had a new epilepsy miRNA. The UNIMAR team led efforts to identify the targets of miR-129-5p. They found several. The first was a calcium pump, encoded by the Atp2b4 gene. This functions to extrude calcium from cells, a critical job since too much calcium can interfere with metabolism or cause cell death. Since miR-129-5p is increased in epileptic brain tissue (and Atp2b4 is down) this could be compromising how neurons handle calcium. Another target was an RNA binding protein called Rbfox. This class of proteins is critical for the distribution of RNA around a cell. While there is growing evidence that disturbances to RNA binding proteins cause neurodegenerative diseases, no-one had looked in epilepsy. The EpimiRNA project helped elucidate the mechanism by which this miRNA controls excitability and found two exciting new targets for seizure control. The findings were published in the fourth year of the project:
Rajman et al. A microRNA-129-5p/Rbfox crosstalk coordinates homeostatic downscaling of excitatory synapses. EMBO Journal 36:1770-1787 (2017)
Other miRNAs and their potential targets-mechanisms
The EpimiRNA project has made significant progress in understanding the mechanisms of at least four other miRNAs. This includes miR-135a, which was identified as dysregulated in human epileptic brain by the UMCU team. Within the EpimiRNA project, we conducted experiments that showed antagomirs targeting this miRNA can suppress epileptic seizures. The UMCU team has identified two important targets. One is a transcription factor (TF), which controls the activity of multiple genes, allowing this miRNA to shape an extensive gene expression landscape. The UMCU team also showed the miRNA functioned in the control of axon guidance, which may be altered following epileptic injury. Findings have been published and others are in preparation:
van Battum w et al. An image-based miRNA screen identifies miRNA-135s as regulators of CNS axon growth and regeneration by targeting Krüppel-like factor 4. J Neurosci 38:613-630 (2018).
Protein analyses provide evidence of strong influence of miRNAs during epilepsy development
By mid-way through the EpimiRNA project, we knew what happened to levels of every miRNA in the different models but it was uncertain whether many of the novel miRNAs were actually important. We chose to look at protein changes within the brain. Ultimately, if a miRNA is functional it should lower levels of proteins. We undertook an ambitious plan to look at all protein changes over the time-course of epilepsy, selecting the rat model from UNIMAR for this special analysis because of the larger size of the brain. The SDU team led the work to profile the protein changes using mass spectrometry technology. A total of 24 samples were processed, 9149 peptides were identified and 580 protein regulations were determined. Each time point had a unique set of protein changes and the amounts of protein that changed differed during the course of epilepsy development. The earliest time point showed only few changes, increasing thereafter. Remarkably, one of the time points showing the most changes was the day after the first epileptic seizure occurred.
To compare the proteomics data with the miRNA profiling data we also did a gene set enrichment analysis (GSEA). This will indicate if there is a significant enrichment of a group of proteins being targets of a specific miRNA among highly changing proteins. Essentially, you are taking a set of altered proteins and asking what miRNAs could explain what you are seeing. You then look if those are miRNA regulated in the model. The result was remarkable, identifying multiple key miRNAs we knew were implicated in epilepsy plus a set of miRNAs that have not yet been interrogated (not disclosed). The data are strong support for an important influence of miRNAs on the proteins in the hippocampus.
Main conclusions and future objectives Collectively, the EpimiRNA project solved the molecular mechanism by which multiple miRNAs control seizures. We also defined all protein changes during epilepsy development and show a strong predicted influence of miRNAs in shaping the protein landscape in the epileptic brain. The discoveries have re-shaped our thinking about the targeting preferences of these molecules. They have led us to new categories of gene that may be involved in epilepsy such as RNA binding proteins. Finally, we may be able to design drugs to the identified miRNA targets as anti-seizure or disease-modifying therapies of the future.
1.3.4 Objective (3) Identification of miRNA biomarkers of epileptogenesis and seizures
Introduction
A major focus and achievement of the EpimiRNA project was to identify miRNAs in biofluids that may be biomarkers of the epileptogenic process or a means to diagnose epilepsy. The project approached this challenge by analyzing miRNAs in biofluids from the three animal models of epilepsy and from human volunteers and patients. We collected blood samples over multiple time points from the animal models and from patients before and after they had a seizure. We also collected cerebrospinal fluid (CSF) samples which, being in direct contact with brain tissue, may be a rich source of biomarkers.
Choosing a miRNA screening technology
Our main hypothesis was that brain-specific miRNAs would be found in blood samples in the epilepsy models. Nevertheless, we approached the identification of miRNAs with an open mind and opted to initially screen for all known miRNAs before closing in on one or two that had the best biomarker potential. The main technology in the field for screening miRNAs at the start of the EpimiRNA project used a method that began by extracting small RNAs from a sample and mixed with reagents that include special “templates” that detect and amplify the miRNA resulting in a semi-quantitative measure of the amount of the miRNA. The EpimiRNA project required a screening platform that would work for the type of ultra-low amounts of RNA and small volumes we would have from the rodents and CSF. We selected a platform that was designed precisely for ultra-low abundance samples; OpenArray (OA). The OA technology provided us genome-wide coverage of miRNAs on an etched glass microchip-like plate into which the sample is added by robotic pipette. The semi-automated workflow then delivers the chips to a specialized PCR machine (12K Flex) for reading. An added advantage of the OA technology was speed, enabling us to screen multiple samples within a few hours.
Optimising and trouble-shooting the miRNA screening technology and other methodology
One of the first decisions we had to make was whether to use serum or plasma. Serum had the advantage that it required minimal processing (a blood sample is left to clot). However, there were concerns that the clotting process would alter miRNA profiles and introduce variability. We therefore opted for plasma and developed a standard operating procedure (SOP) for partners which we later made available to other researchers.
During the first year of the project, while the teams at RCSI, UNIMAR and UNIVR were generating the brain and blood samples from the animal models, the team at RCSI who would be responsible for biofluid work began tests of the OA platform. We first wanted to be certain of the platform’s reproducibility. We took a sample of plasma and split it, then ran it on two different OA chips, with a gap of three months in between. The OA results from the two chips were almost identical, confirming high reproducibility of the platform.
We also investigated whether the implanted telemetry unit, which we would be using to record brain activity and count seizures in mice, affected miRNA profiles in blood. This required a minor surgical procedure on the mice to place the device under the skin. Our tests showed only 1 miRNA out of all those screened differed in plasma from implanted mice. Thus, the OA platform was reproducible and our monitoring method would not bias results.
EpimiRNA’s rodent blood sampling technique
Another technical challenge for the project was how to obtain repeated samples of blood from mice. While the tail vein of the rat is large enough for repeat blood sampling, the mouse tail is much smaller and repeated sampling is not recommended. We adapted a technique in which blood is taken from a vein that runs down the cheek of the mouse (submandibular vein). A tiny needle is inserted into the cheek of the anesthetized mouse which results in a few drops of blood being released which can then be collected. We found this to be highly reproducible and a training video was generated so that all teams performed the procedure the same way. We now had the ideal plasma technique for the mouse OA studies.
Final tests of the OA platform
Our final tests were to investigate how many miRNAs were detectable in normal plasma and check if these were similar to what had been reported before. We ran a set of mouse plasma samples and could detect up to ~200 miRNAs in a sample. This was at least as good as other platforms and within the detected miRNAs we found plenty that were known to be abundant in plasma such as miR-16, miR-19b and miR-223. Equally important, we found very low or undetectable levels of the known brain-enriched miRNAs. When we ran human plasma samples we obtained very similar results. That is, we could often detect ~200 miRNAs in a given sample and the most abundant miRNAs were those that others had reported previously. For the actual experiments we applied strict and consistent criteria for what miRNAs were called present in a sample. This reduced the numbers of miRNAs called present to ~100 per plasma sample.
Plasma miRNA profile in healthy human volunteers
With the collection of human samples from epilepsy patients growing during the first year, we took the opportunity to understand the profile of miRNAs in healthy control plasma. We looked at several key variables: (1) morning versus afternoon sampling, (2) day-to-day differences and (3) male versus female miRNA profiles. We found no differentially expressed miRNAs between the morning or afternoon samples, no significant changes in miRNA levels from the same subject over a period of a month and no sex differences.
We concluded that, overall, plasma miRNA levels were very stable and that our platform compared well to others while having the advantage of speed and low input requirements. The results of these studies were published:
Mooney et al. High Throughput qPCR expression profiling of circulating microRNAs reveals minimal sex and sample timing-related variation in plasma of healthy volunteers. PLoS ONE 10:e0145316 (2015).
Identification of miRNAs in plasma in experimental epileptogenesis
By the third year of the project, the three different teams leading the animal model work had generated enough complete sets of blood samples that we could begin to profile these using the OA platform. We ran a time-series from each model, including a baseline sample (pre-insult), a sample taken after the animals had been subjected to status epilepticus but had not yet developed spontaneous recurrent seizures, and then finally a sample taken during the phase where active epilepsy was established.
We looked first at the baseline samples taken from the rodents prior to experimental manipulation. We detected an average of 103 miRNAs in the mouse IAKA model, 93 in the mouse PILO model and 88 in the rat PPS model. Focusing on the top 20 most abundant miRNAs, we found that 13 were common to all three models. We also found strong overlap between the most abundant miRNAs in rodent plasma and the most abundant in human plasma. Specifically, 7 of the 12 most abundant miRNAs in human plasma were found in the rodents (miR-19b-3p, miR-24-3p, miR-92a-3p, miR-146a-5p, miR-20a-5p, miR-223-3p and miR-191-5p). This suggests discoveries in the rodent models will translate to humans.
When we looked at all miRNAs in the baseline plasma samples from the three animal models we found extensive overlap, with 66 of the same miRNAs detected in all three models. This means that the circulating pool of miRNAs is very similar between the two rodent species. When we looked at the miRNAs present in samples from epileptic animals, we found 74 miRNAs were common between the models. This means that many of the same plasma miRNAs are affected the same way regardless of species or specific model. But we also found a small number of miRNAs that were unique to each model.
Comparing the lists – those found in all models at baseline and those in samples from epileptic animals - we identified 16 miRNAs unique to the epilepsy samples. When we checked databases on tissue-specific miRNAs we found many of these were brain-restricted. This strongly supported our hypothesis that brain-enriched miRNAs are released into the circulation in epilepsy and thus have biomarker potential. When we applied statistical tests to the lists of miRNAs from each model we identified several miRNAs that were altered at different stages of epilepsy in the different models.
We next began validating the OA-identified miRNAs. Since the amount of plasma from the animal models was limited we had to pick just a few miRNAs to focus on. We selected two miRNAs that were different in epilepsy versus baseline samples and three from among those that changed within a given model at a specific time point. We now switched from the OA platform to test each miRNA using individual miRNA assay kits. Two of the five did not show the expected change in at least two of the models whereas expression of three others were all validated as potential epilepsy biomarkers.
To extend this validation, we were interested in whether plasma levels of the miRNAs were sensitive to any common anti-epileptic drug (AED) therapies. This is important to know, since a miRNA that is strongly affected by medication maybe unreliable for diagnosis. For these experiments, mice were dosed daily with one of two common AEDs and plasma levels of the miRNAs were measured afterwards. We found no significant effect of AEDs on the tested miRNAs (Figure 14) suggesting they are stable biomarkers.
We were also interested in whether treating mice with a disease-modifying therapy would “normalise” (restore) the altered miRNA levels in plasma. If the miRNAs are biomarkers of the disease their levels should be corrected by a disease-modifying therapy. For these experiments, mice were injected with the miRNA inhibitor that we had in pre-clinical development (antagomir-134; Ant-134). This reduces spontaneous seizures by ~90% in the RCSI mouse model. Plasma levels of the validated miRNAs changed in mice given Ant-134 indicating the miRNA biomarkers respond to an anti-epileptogenic therapy.
Our final tests of the rodent-identified miRNAs was to look for them in plasma from TLE patients in adults. We had collected a number of plasma samples from epilepsy patients and we used a sub-set of these. Our results showed plasma levels of three of our miRNAs were altered in patient plasma, either at baseline or after seizure. Levels of another also showed a trend to lower levels, in line with the animal data, in samples after seizure. We used a statistical technique called receiver operating characteristics (ROC). to assess the ability of the miRNAs to distinguish between controls and patients with TLE. As an example, we found that post-seizure sample levels of one of our miRNAs could identify TLE patients with 86% confidence. This is a high-quality result for a single molecule biomarker.
Clinical studies to identify miRNAs biomarkers of TLE
In parallel with the animal model studies, we began our search for the best plasma miRNAs to diagnose human TLE. The main effort focused on sampling patients within specialist epilepsy monitoring units (EMU). Patients admitted to EMUs typically have complex, drug-resistant epilepsy and may be later offered neurosurgery to remove epileptic brain tissue. After admission, patients are continuously monitored by video and EEG to record epileptic events. After ethical approval was received, we began recruiting patients from the EMUs at epilepsy centres in two countries - Germany and Ireland. The standard protocol was tocollect a baseline sample after admission when the patient had been seizure-free for a few days. The research team would then collect one or more samples at different time points after a seizure occurred. We selected an early time point (1 h) and a later time point (24 h). As we had done with the rodent studies, plasma preparation protocols were harmonized between sites.
Our target was to recruit 60 patients for the profiling study. We reached this number quickly and by mid-way through the project had collected samples from over 150 epilepsy patients. We also started to collect samples from patients with other forms of epilepsy. This would allow us to determine whether a particular miRNA was unique to TLE patients or whether it was a broader biomarker of all epilepsies. We also collected samples from patients who had psychogenic non-epileptic seizures (PNES). These patients are particularly difficult to diagnose outside specialist centres. If the miRNAs we identified in the TLE patients were true biomarkers of epilepsy, their levels should not be altered in a patient with PNES since the seizure events are non-epileptic in origin (typically a manifestation of an underlying psychiatric condition). However, if the levels of the miRNA were similar between TLE and PNES patients then they might instead be reflecting medication or other factors (a PNES patient may be prescribed an AED on the basis that they are thought to have epilepsy).
During the third year of the project we profiled miRNAs in individual samples from epilepsy patients at baseline and after a seizure. In total, 96 plasma samples from the two centres were profiled by the OA platform. A set of miRNAs were identified that were differentially expressed in epilepsy samples compared to controls. Before moving to validate these, however, we decided to add an extra profiling study. The reason for this was that by this point the technology to directly sequence small RNAs from low-abundance sources such as plasma had become more widely available, costs were dropping and a dual approach would give EpimiRNA an advantage over what had come before. “RNA-seq” generates complete and quantitative analysis of all miRNAs and can identify novel miRNAs. We therefore ran RNA-seq, pooling the plasma samples from each centre to reduce costs and compared control to pre- and post-seizure. We found nearly 100 more miRNAs in the RNA-seq data compared to the OA platform. Results were also highly reproducible between Centres. Importantly, we found 11 additional miRNAs that were up- or down-regulated in epilepsy and these were then advanced with the OA-identified miRNAs for validation.
We performed an extensive set of validation experiments. First, we confirmed the profiling results for 15 miRNAs using the original samples, running individual miRNA assays. From these studies we became most interested in three miRNAs. The level of miR-27a-3p and miR-328-3p were consistently changed in samples collected after a seizure. Levels of miR-654-3p were different from controls in both baseline and after-seizure samples. We then tested these in a larger cohort comprising over 100 TLE patients and 100 controls. We also measured their level in plasma from patients with three other forms of seizure disorder and patients with PNES. This confirmed that levels of miR-27a-3p and miR-328-3p were different in TLE patient plasma but revealed they were also altered in patients with genetic generalised epilepsy (GGE). Importantly, the miRNAs were not significantly different to controls in samples from PNES patients. This means that measuring the level of these miRNAs could help diagnose epilepsy and help identify possible PNES cases. Statistical analyses suggested the miRNAs could correctly predict epilepsy with ~70% accuracy. While encouraging, we sought ways to increase the diagnostic accuracy of these biomarkers.
We tested for levels of the three miRNAs in plasma samples from one of our rodent models. Testing in rodent blood means we can exclude certain likely confounders that are impossible to avoid in human studies such as medication. We found that levels of all three miRNAs followed the same pattern in mice as we had seen in the patient samples. We also examined what happened to levels of the three miRNAs in mice given the disease-modifying Ant-134 therapy. Plasma levels of the three miRNAs were all corrected – normalised – in mice given the disease-modifying Ant-134 treatment.
Another technological advance that had appeared since the start of the project was digital PCR (dPCR). This allowed a researcher to count the individual copies of an RNA molecule in a sample. Thus, you could estimate how many copies of the three miRNAs would be present in a drop of blood. We put this technology to the test and ran dPCR analyses on our three miRNAs. The approach worked, and we learned that miR-27a-3p was the most abundant of the three in plasma, with ~6000 copies per sample of plasma in a healthy person. As expected, levels were significantly different in both TLE patients and patients with GGE, with readings of ~2000 copies in an equivalent volume in an epilepsy patient. Interestingly, the count in a PNES patient was ~5000, close to normal.
Making a prototype blood test for epilepsy
Project partners began thinking about how our findings on plasma miRNA could be developed into something resembling a clinical test. One of the partners (RCSI) started working with colleagues on a way to directly measure RNA without the need for template-directed amplification. If this were possible, it would dramatically simplify testing and reduce the sample-to-answer time. A prototype device was eventually developed that comprised a “capture” oligonucleotide that would trigger an electrochemical reaction proportional to the amount of miRNA. Because we had not yet finished screening miRNAs using the OA, we designed the first prototype to detect miR-134-5p. This was one of the miRNAs our project was most focused on because of its therapeutic potential. However, since miR-134-5p was also brain-enriched it could also be a potential biomarker. We tested some of the samples collected in the video-EEG unit using the new device. It gave almost identical results to those when measured by standard PCR but in a fraction of the time. These results were published in 2015: Spain et al. Direct, non-amplified detection of microRNA-134 in plasma from epilepsy patients. RSC Advances 5, 90071-90078 (2015).
We later adapted the device so it could detect the three novel miRNAs. We showed that we could measure all three miRNAs simultaneously with the device, termed “TORNADO” (Theranostic one-step RNA detector). This system is in commercial development.
Superior diagnostic accuracy by analysis of how miRNAs are transported in blood
Despite this progress, we felt we had to improve on the accuracy of our miRNA biomarkers. The breakthrough came from wanting to understand how miRNAs reach the circulation from the brain. There was evidence at the time that miRNAs were released from cells in tiny vesicles called exosomes. Other studies suggested that most circulating miRNA was bound to the AGO protein. Neither mechanism of transport had been explored in epilepsy. We therefore took a set of our plasma samples and split them and then either extracted the AGO protein or prepared exosomes using a special centrifuge. What we found was remarkable. First, there was generally more miRNA within exosomes than in the AGO-bound fraction. More importantly, when we ran the statistical analysis of diagnostic accuracy it noticeably improved our ability to tell patients from controls. This time the miRNAs gave results in the mid-80% range. Using a further statistical method called logistic regression and considering all three miRNAs combined gave a result of 93% confidence of diagnosis. This means that if we first separate this portion from the rest of the plasma we can get a more accurate diagnosis. We are now in discussion to develop a miRNA-based test with industry partners.
Raoof, Bauer, El Naggar et al. Dual-center, dual-platform microRNA profiling identifies potential plasma biomarkers of adult temporal lobe epilepsy. eBiomedicine (2018).
EpimiRNA performs the first analysis of miRNAs in CSF from epilepsy patients
In parallel with the plasma studies, partners in EpimiRNA collaborated to study miRNAs in CSF. As mentioned, we were particularly interested in CSF since it might be an even better source of epilepsy biomarker. Also, we were aware that no other group in the world had studied miRNAs in CSF samples from patients with epilepsy. However, our collection of CSF samples from the original partner hospitals was not sufficient. The clinical co-coordinator of EpimiRNA, Prof Felix Rosenow, reached out to a colleague who specialised in CSF analyses (Dr Peter Körtvelyessy, Magdeburg). We now had enough CSF samples from TLE patients plus samples from patients after status epilepticus. This is an interesting patient group because they have experienced prolonged seizures but may not necessarily have epilepsy. Another group of samples were collected from patients with other neurological diseases for the purposes of comparison. This included patients with Alzheimer’s disease (AD) and Multiple Sclerosis (MS). Last, we had a set of controls. In total we obtained over 100 samples of CSF from different individuals.
We began by using the OA platform to screen 45 samples from controls, TLE patients and SE patients. It was immediately obvious that miRNAs were much less abundant in CSF than in plasma. In fact, we detected only around half as many miRNAs in CSF as we had in plasma. We detected a number of significantly regulated miRNAs. This included five that were different in TLE patients compared to controls (one down and four up regulated) and 15 miRNAs in samples from patients with status epilepticus compared to controls (seven down and eight up regulated). Thus, despite detecting far fewer miRNAs in CSF the numbers of differentially expressed miRNAs remain significant. This is consistent with our hypothesis that CSF may be a rich source of potential biomarkers.
As with our plasma study, we selected a set of three miRNAs to focus on during validation. These were miR-19b-3p, a potential biomarker of TLE and miR-21-5p and miR-451, which were different in patients with status epilepticus. Levels of these were not different to controls in the patients who had other neurologic diseases meaning they were specific for epilepsy and seizure-related disorders. Again, we used dPCR to get an idea of the copy number of these miRNAs in CSF. By far the most abundant was miR-451, with an average of ~200 copies per sample in status epilepticus compared to under 20 for TLE patients and other neurological diseases. Diagnostic accuracy of over 80% was achieved for the miRNAs.
Our final experiments using the CSF were to look at the miRNAs within the exosome and AGO-bound fractions. What we found was that for each neurological disease the proportion of the miRNA in one form or another was quite different.
As an example, miR-19b-3p was most abundant in the AGO-fraction in TLE patients but was more highly expressed in the exosome fraction in MS patients. These findings support the development of a diagnostic test for epilepsy based on CSF miRNAs. Indeed, we tested TORNADO and confirmed that the device could determine the concentration of a miRNA in CSF. The device estimated the concentration of miR-134-5p in CSF at between 100 nM and 10 mM in patients with TLE or status epilepticus. This level is higher than in plasma, suggesting CSF is a unique source of miRNA biomarkers for diagnosis. The findings have been published:
Raoof et al. Cerebrospinal fluid microRNAs are potential biomarkers of temporal lobe epilepsy and status epilepticus. Sci Rep 7, 3328 (2017)
McArdle et al. “TORNADO” – Theranostic OneStep RNA Detector; microfluidic disc for the direct detection of microRNA-134 in plasma and cerebrospinal fluid. Sci Rep 7, 1750 (2017)
Biomarker profiles of human brain stimulation
Our final biomarker project concerned looking for evidence of miRNA changes in biofluids in response to brain stimulation. Our partner Cerbomed had a transcutaneous VNS stimulator (tVNS) in clinical testing for drug-resistant epilepsy, led by Prof Felix Rosenow. Does tVNS change miRNA profiles? We collected blood from volunteers before and after tVNS to determine if the stimulation altered biofluid levels of miRNAs. We focused on a sub-set of miRNA that had been shown in the earlier projects to be biomarkers of seizures or epilepsy. We also included an analysis of cytokines because both VNS therapy and epilepsy are associated with changes to inflammation. We ran samples from subjects with tVNS stimulation as well as sham-stimulated controls. The main finding was that none of the miRNAs were altered within the plasma samples of the healthy volunteers given active tVNS exposure. However, one miRNA (not disclosed) was found to be significantly upregulated in drug-refractory epilepsy patients given active tVNS. This suggests that plasma miRNA levels can be used as a biomarker of tVNS therapy in drug-resistant epilepsy. Of further interest, two of the cytokines changed in tVNS patients. The clinical use and biomarker potential of tVNS for patients with epilepsy remains uncertain, however. The device may be most useful in predicting who may benefit from implanted VNS.
We published a review on the state-of-the-art of miRNAs as biomarkers in collaboration with the EpiTarget consortium: Enright et al. Discovery and validation of blood microRNAs as molecular biomarkers of epilepsy – ways to close current knowledge gaps. Epilepsia Open (in press)
Main conclusions and future directions: EpimiRNA has generated important insights into the potential use of miRNAs as diagnostic biomarkers. There is good evidence that both plasma and CSF contain unique levels of several miRNAs that are not found in other neurological conditions. The levels of these can be absolutely measured using either dPCR or TORNADO. We cross-validated our animal findings in human samples and human findings with animal model results. We showed for the first time that a miRNA-based disease-modifying therapy altered plasma levels of miRNA biomarkers. We identified a brain-enriched miRNA that is altered in plasma of TLE patients given tVNS therapy. These discoveries could prompt the development of a multi-miRNA biomarker panel for diagnosis of epileptogenesis or epilepsy or to support brain stimulation device use.
1.3.5 Objective (4) Novel therapeutics by targeting miRNAs
Introduction
A major objective achieved by the EpimiRNA project was to discover new molecules that could suppress seizures or have disease-modifying effects and move these towards preclinical development. Foremost, we focused on inhibitors of miRNAs, termed antagomirs. Over the course of the project, we tested more than 10 different miRNA inhibitors in animal models. We split our efforts three ways. We tested inhibitors against a set of miRNAs which EpimiRNA partners had already become interested in due to links to epilepsy. Second, we screened inhibitors targeting the entire set of miRNAs we had discovered to be common-to-all-models. Finally, we made efforts to answer important questions around clinical translation of a miRNA inhibitor. Did our inhibitors work in species besides mice? How important is the structure of the antagomir? How would a patient even take such as drug?
Over the course of the project we identified 8 miRNA inhibitors that suppressed seizures triggered by chemicals or brain stimulation. Two more prevented spontaneous recurrent seizures (ie. epileptic seizures). We answered critical questions on the path to preclinical development. Partners at RCSI secured a US patent for one of the miRNA inhibitors. The EpimiRNA project also discovered an anti-seizure compound from Bicoll and by the end of the project had this accepted for analysis by the Epilepsy Therapy Screening Program run by the National Institutes of Health in the USA.
Effects of miRNA inhibitors targeting five epilepsy-associated miRNAs
Ant-22 We had become interested in miR-22-3p as one of the most upregulated miRNAs in protected brain regions in the mouse model at RCSI. Blocking miR-22-3p using antagomirs caused spontaneous seizure rates to increase in mice, indicating the miRNA had anti-seizure properties. We realised that if we could increase the amount of the miRNA in the brain it could reduce seizures. We showed that injecting miR-22-3p itself into the mouse brain could reduce epileptic seizures. However, the dose-range at which this worked was extremely narrow. We think that while inhibiting a miRNA is largely safe while introducing more of a miRNA could impact on cell functions in difficult to predict ways and trigger an inflammatory response. From this point onward, EpimiRNA focused on inhibiting up-regulated miRNAs. These data were published (Jimenez-Mateos et al. Sci Rep 2015)
Ant-129 miR-129-5p was among a set discovered by the UNIMAR team to control the strength of signalling at synapses and was also up-regulated in our animal models. The RCSI team tested antagomirs against the neuron-enriched miR-129-5p in their mouse model. Mice given the antagomir were profoundly protected against seizures induced by
the chemical kainic acid. Recordings of brain activity showed much-reduced seizure severity and there were fewer signs of brain injury from the seizures in those mice. These data
were published (Rajman et al. EMBO Journal 2017).
Ant-324 miR-324-5p was discovered as targeting the Kv4.2 potassium channel. The RCSI team tested antagomirs against miR-324-5p (Ant-324), again using their kainic acid model.
Seizure onset was delayed in mice given the antagomir and, again, there were few signs of brain injury from the seizures in those mice. These data were published (Gross et al. Cell
Rep 2016).
Ant-135a miR-135a-5p was originally found by the UMCU team to be increased in the hippocampus of patients with TLE. We checked brain sections from our mouse model and found it was strongly increased in neurons. We developed a technique to deliver an antagomir into the brain of an already-epileptic mouse. We implanted a device that recorded seizures and once we had a stable baseline we injected the inhibitor. Within a day we started to see a reduction in epileptic seizures while the control group continued to have seizures.
We waited for two weeks and then studied the results. The Ant-135a inhibitor significantly reduced epileptic seizures by over 60% and seizures were shorter in duration.
Ant-134 At the time of starting the project, we already knew that injecting an inhibitor of miR-134-5p (Ant-134) before or immediately after kainic acid in mice had a dramatic effect on seizures. Within EpimiRNA, we embarked on a comprehensive study to address important remaining questions that might otherwise be a barrier to preclinical development. First, the RCSI team showed that Ant-134 could reduce seizures in the PTZ test, a common test in mice for identifying anticonvulsive drugs. The UNIMAR team showed that injecting Ant-134 after triggering epilepsy using electrical stimulation of the brain dramatically reduced epilepsy in rats. The UCL team showed that Ant-134 reduced epileptic-like activity in brain slices. These results were published in the fourth year of the project: Reschke et al. Potent anti-seizure effects of locked nucleic acid antagomirs targeting miR-134 in multiple mouse and rat models of epilepsy. Mol Ther Nucleic Acids 6:45-56 (2017)
Another question was how important was the exact chemical composition of the Ant-134? Would small differences in the chemistry affect potency? We tested this by ordering Ant-134 from a different company and testing alongside our current batch. Both produced nearly identical seizure suppression. We then turned to safety. Since the target, miR-134-5p controlled the shape of contact points between neurons it was possible that we could be interfering with normal brain function. The UCL team led studies to record neuronal behaviour after injection of Ant-134. All physiological properties were found to be normal, while the molecule nevertheless blocked epileptiform events. Findings were published: Morris et al. Spared CA1 pyramidal neuron function and hippocampal performance following antisense knockdown of microRNA-134. Epilepsia 59:1518-1526 (2018)
Overcoming the delivery challenge
We next thought about delivery. Brain tissue is separated from the contents of the blood by a specialised barrier (blood-brain-barrier). This blocks large molecules from passing through. Ant-134 was too big to pass if injected into the blood. Our first idea was to inject into the spinal fluid using a route similar to how CSF is obtained. We found that Ant-134 can reach deep into the brain when given like this in mice and blocked the target miRNA. Since only a single injection of Ant-134 is needed to stop seizures, this could be acceptable to a patient with drug-resistant TLE. But we had a second idea. What if the blood-brain barrier is open in epilepsy? Then a systemic injection might be enough to get it into the brain. We studied the opening of the blood-brain barrier in the model at RCSI, finding it opened profoundly (Figure 23). With this knowledge, we triggered status epilepticus in mice and gave them a single injection of Ant-134 into the belly of the mouse. We tracked the mice for the next three months to be sure any effects were permanent. Epileptic seizure rates were dramatically reduced by Ant-134 indicating a disease-modifying effect. At the final time point of assessment (3 months), Ant-134-injected mice had 99% less seizures than the control group.
Ant-134 can prevent seizures in already-epileptic mice
The final study with Ant-134 was to test whether it could reduce seizures when injected into already-epileptic mice. This was one of the most clinically realistic scenarios, where a patient has recently presented with epilepsy and you could give them a disease-modifying therapy. We triggered epilepsy in mice and waited two weeks. We then injected Ant-134 or a control molecule and tracked the seizures in the mice. Within two days seizure rates had fallen by over 80% in the mice and stayed this way for two weeks. We continued recordings a month later. During these recordings, now many weeks after the single injection of Ant-134, mice had 90% fewer seizures than the controls. There was no sign of seizure rates recovering. Taken together, the work on Ant-134 suggests this is one of the most potent disease-modifying therapies identified to date. Having secured a patent for this invention we subsequently began collaborations with industry partners toward the preclinical development of this new therapy for epilepsy.
Anti-seizure and neuroprotective effects of novel miRNAs discovered by EpimiRNA
By mid-way through the project we had a list of 8 miRNAs that were upregulated in all three epileptic mice, six of which had not yet been targeted in an animal model. We therefore embarked on a systematic effort to test antagomirs against all six miRNAs. For this we needed a high-throughput model and turned again to the mouse model at RCSI. Experiments were run in batches (a so-called “block design”) with six animals a day each given a different antagomir. Two other mice received either saline or a scrambled version of the antagomir. This was repeated until we had ten sets of experiments. Our results showed three of the antagomirs produced powerful anti-seizure effects in the model. Pilot tests of a “combination” of the three antagomirs was not superior to one alone. Analysis of brain injury revealed five of the six antagomirs protected the hippocampus from damage.As we had done for Ant-134, we were also keen to test the safety of the antagomirs on normal neuronal function. The UCL team again led studies to record neuronal behaviour after injection of the three novel antagomirs. All physiological properties were found to be normal.
Non-miRNA-In addition to the miRNA focus, the EpimiRNA project included an SME, Bicoll that had expertise in drug discovery. During the project they rationally designed a small molecule compound for seizure control. During the fourth and fifth year of the project, teams at UCL and RCSI were able to test the molecule, beginning with in vitro tests in brain slices and moving later to an in vivo mouse model. The compound demonstrated strong anti-epileptic effects in both. This included suppressing epileptic activity in human brain tissue and reducing seizures in the PTZ seizure model in mice. The compound was accepted into pre-clinical screening by the NIH’s epilepsy therapy screening program.
Main conclusions and future objectives
Collectively, the EpimiRNA project identified 10 new molecules with anti-seizure effects, advanced one into preclinical development, obtained a patent, moved another anti-seizure
compound into the USA’s epilepsy therapy screening programme and developed new industry partnerships.
1.3.6 Objective (5) Investigation of genetic variation in miRNAs.
While TLE is largely considered an acquired form of epilepsy there is evidence that there is a genetic contribution. That is, that variation or mutation in certain genes may explain a portion of the TLE population. EpimiRNA undertook the first ever investigation to determine whether genetic variation in miRNAs underlies human TLE. The main focus was on the coding sequences of the miRNAs (over 1500) but we designed the study to also look at the over 40 human miRNA biogenesis pathway genes and we looked for variation in the 3’ UTR regions of all known genes expressed in the hippocampus (~90,000 sequences). Variant or mutationof a miRNA binding site in a target transcript is potentially as important as a mutation in the miRNAs themselves, because the introduced sequence change may reduce (or increase) silencing and alter gene expression, leading to a functional change in a cell.
We aimed to sequence 850 TLE patients with matched number of controls and we were also able to include other epilepsy syndromes. However, this exceeded the total number of DNA samples available within the project teams so we collaborated with colleagues at other European and non-European sites and were able to reach (and exceed) our targets. We
eventually sequenced over 3000 patients and controls from the different partner sites. The sequencing workload was spread between the CU team in the USA and the UNICAMP team in Brazil. Sequencing began in the first year of the project and was finished by the fourth year at both sites. The teams then began the process of carefully checking the quality of the sequencing and the analysis of the data. The teams leading the genetics analysis used different approaches to the analysis. However, we did not identify any single miRNA genes that passed the significance thresholds set in our main analysis. Nor did we identify miRNA biogenesis genes or the targets. This would indicate that, within the parts of the genome sequenced and with the sample size we had used, there are no sites that when burdened with genetic variation confer strong enough epilepsy risk. The results were similar when we combined the results of the European-decent subjects with those of Brazilian decent - no single gene passed significance.
1.3.7 Objective (6) Technological advances and resources.
The EpimiRNA project has delivered a number of important technological advances and overseen training of these. The project has also generated a new and unique biorepository of samples from across the various partners that can support research long into the future.
Technical advance - iCLIP analysis of human epileptic brain. A key objective of the project was to identify with certainty the targets of all known miRNAs in the human brain. The technique for this – iCLIP – depends on being able to selectively pull-down the AGO protein from a sample while maintaining the association between the miRNA and the target. If the procedure is not performed correctly the miRNA or target may be missed. The AU team perfected an iCLIP protocol for human post-surgery tissue. This protocol, once published, will be of value to any researchers in the neuroscience community. The data from the study will be available to the research community
Data base resource iCLIP database of all miRNA-target interactions in the human epileptic brain
The iCLIP study has generated a rich dataset that can we searched either by miRNA or by target mRNA. This database will be made freely available upon publication of a companion research article.
Technical advance Systems modelling frameworks for miRNA research
We established an unbiased, systems-level pipeline to identify and rank miRNA-mRNA target interactions (MTIs) from a curated in-house database. Using a common programming language (R), the pipeline compares miRNA sequences across species, collates miRNA-target interactions from multiple databases, identifies transcription factors among targets, and facilitates prioritisation/filtering of MTIs via a bespoke scoring system (‘MTI score’).
The filtered miRNA-target information can be used to generate network interaction graphs and perform enrichment/pathway and similar analyses and has been used in multiple EpimiRNA publications. We also developed an integrative model which enabled us to deduce the dynamics between miRNA and mRNAs based on their expression profile using Bayesian methods. The network was generated using 190 miRNAs and 198 mRNA containing 2100 edges between miRNAs-mRNAs and mRNAs-mRNAs (Different colour identifies different modules. Text size of each node depicts its influence in connectivity of the network).
Resource - EpimiRBase
EpimiRBase was released in 2015 and is a searchable database of all known epilepsy studies on miRNA. It allows any researcher to look up whether a particular miRNA has been identified, the species, model and whether any functional studies were undertaken. It has already been queried more than 1000 times and is available at www.epimirbase.eu
Resource - EpimiRNA biorepository
Over the course of the EpimiRNA project, blood samples were obtained from over 300 patients and healthy controls. These were processed to plasma and any remaining sample will be retained for future biomarker research on miRNAs. We also collected nearly 100 human CSF samples and over 100 human brain resections from drug-refractory epilepsy cases. From our animal model work we have also collected 100s of blood samples, also processed to plasma, and brain tissue samples. These will be used by partners within the project in the immediate future but will later be made available to other teams, where ethical approval and GDPR allow, to validate their own findings or to develop new hypotheses.
Potential Impact:
1.4.1 Impacts
The scientific, technological and non-scientific outputs from the EpimiRNA project have resulted in important impacts. These cover diverse areas including improved understanding of the causes of epilepsy, advances in technologies to detect and monitor miRNAs, resources (biorepository, databases) for the research community, new therapies in pre-clinical development and impacts on our various SME partners.
Key Impact 1: Improved understanding of the aetiology and mechanisms of epilepsy
A primary objective of the EpimiRNA project was to generate new knowledge about the epileptogenic process. Research by the consortium and our collaborators has delivered the first complete identification of functional, conserved miRNA changes that accompany epileptogenesis in animal models and in the brains of refractory TLE patients. This impact was achieved through EpimiRNA’s innovative project design, which brought together, for the first time, experts in RNA research and the required underpinning technologies (deep sequencing, quantitative proteomics), with leading experimental epilepsy researchers and multi-center clinical teams. Specialists in electrophysiology, cell imaging and computer modelling helped determine how the miRNA changes influence excitability, spanning neuron-to-network level assessment. The findings have had a transformative impact on our understanding of epileptogenesis and chronic epilepsy. We have uncovered the involvement of numerous miRNAs in brain excitability and this has generated new understanding of the epileptogenic process.
Among the highlights, we have generated a molecular catalogue that features abundance data on every miRNA expressed in the mouse and rat hippocampus and how these change as epilepsy develops. The data contain reassuring entries on miRNAs previously linked to epilepsy and miRNAs known to be enriched (or absent), but also many surprises including a set of novel miRNAs common to all animal models that had not been previously linked to epilepsy. The human iCLIP dataset are remarkable in their value, allowing anyone to explore the target(s) of every miRNA made in the human epileptic brain.
Our functional studies have uncovered numerous new miRNAs which can influence epilepsy or seizures and demonstrated the target(s) and mechanisms. This includes a miRNA (miR-22) that works to counter excessive neuroinflammation, a miRNA (miR-324) that restrains ion channel levels in neurons, a miRNA that regulates axon guidance (miR-135a), a miRNA that controls calcium levels and movement of other RNAs within neurons (miR-129). For the first time, we proved the mechanism by which miRNA inhibitors altered seizures. As a result, we now know that blocking miR-22 worsens seizures because it allows excessive build-up of the P2X7 receptor that promotes neuroinflammation. We know that the seizure-suppressive effects of inhibiting miR-134 result from protecting (restoring) levels of an enzyme that regulates the structure of synapses (Limk1). These findings not only offer miRNAs as novel targets but can lead to non-miRNA targets for seizure control. For example, EpimiRNA researchers have now shown that drugs that block the P2X7 receptor can have disease-modifying effects in drug-resistant epilepsy. The set of novel miRNAs that we identified as common-to-all models included several that were non-neuronal. Despite this, interference in their function altered seizures. Thus glial-based miRNAs as well as neuronally-expressed miRNAs are targets for seizure control. Again, the introduction and establishment of these concepts have had a transformative impact on the field, as evidenced by the commissioning of multiple articles on the topic in topic research journals (Lancet Neurology, Epilepsy Currents, Epilepsia Open, Nature Reviews Neurology (invited)).
At the time of the project start, miRNAs were a nascent area and minor curiosity within the epilepsy research field. Over the past five years our project’s research along with findings by other groups around the world, have projected miRNA research to the forefront of the epilepsy research agenda. This is evidenced by the numerous workshops scheduled for and by EpimiRNA partners to discuss this topic at European and international epilepsy congresses. Our consortium‘s approach represented a new scientific model for how to uncover the mechanisms of epileptogenesis. The project has been instrumental in shaping the direction of basic, pre-clinical and clinical epilepsy research. The findings have also served to support diagnostic and drug development.
All our miRNA-target data will eventually be lodged in an open-access repository that will allow researchers from outside the project to take advantage of the discoveries and drive new research in the future. Indeed, while we had time to study the functions of ~10 different miRNAs during the course of our project, it is clear that further data mining is likely to identify other miRNA mechanisms and targets.
Key Impact 2: Biomarkers for epilepsy diagnosis
It is widely acknowledged that the identification of biomarkers of epilepsy and the epileptogenic process would transform discovery of disease-modifying therapies and the diagnosis of epilepsy in the future. Through the EpimiRNA project we have shown that miRNAs fulfil several key requirements for a blood-based molecular biomarker being enriched in the brain, dysregulated in epileptic brain tissue and manipulation of miRNAs can have seizure-suppressive and disease-modifying effects in preclinical models. A second major impact of the EpimiRNA project is the identification and validation of miRNAs in blood and CSF samples in patients with epilepsy. The results of these studies could be used to develop a blood or CSF-based test for epilepsy. This could improve the time-to-diagnosis for new epilepsy patients and reduce mis-diagnosis rates. Importantly, EpimiRNA’s studies compared results in epilepsy patients to results in patients with conditions that can often be mistaken for epilepsy such as psychogenic non-epileptic attack. The miRNA biomarkers discovered by EpimiRNA were not altered in these patients. This significantly increased confidence in the specificity of the miRNAs for epilepsy and makes commercial development more likely in the future. Indeed, the commercial market opportunity may be significant. The global in vitro diagnostics market has sustained compound annual growth rate of over 5% in recent years and is predicted to exceed €60 billion. There are currently no molecular diagnostics on the market to predict epilepsy therefore our programme has strong potential to drive innovative new healthcare products that have commercial value.
The EpimiRNA project was the first to discover that one of the miRNAs we had shown to be a target for seizure control and disease-modification, miR-134, was also found at higher levels in plasma from epilepsy patients. We followed this work with the (unbiased) discovery of a triple-miRNA plasma “signature” unique to epilepsy patients in plasma and miRNAs unique to CSF. Because we could show that levels of the miRNAs change in response to an anti-epileptogenic therapy (in mice) these could also be useful for stratifying patients in any future clinical trial of a disease-modifying therapy. Although initial work suggests miRNAs may not be changed in blood samples from humans exposed to brain stimulation devices, we did find some molecules associated with inflammatory responses change suggesting there may be biomarkers of therapeutic brain stimulation.
EpimiRNA research also made important advances in how we can enhance the diagnostic accuracy of miRNAs as biomarkers. This could influence the technical requirements of any future miRNA-based clinical test for epilepsy. Our research showed that measuring the relative amounts of any given miRNA that are bound to the AGO protein or sealed inside microvesicles called exosomes offers superior diagnostic accuracy relative to measuring the overall amount of a miRNA. It may be preferable to pre-process a biofluid before proceeding to miRNA extraction and detection. Our finding shows that protein-bound miRNA levels change after seizures could also be useful in developing a test for seizures. Such a test might be useful beyond the diagnosis of epilepsy – it could help identify seizures in patients admitted to emergency departments who are unresponsive. Currently this requires lengthy and technically-demanding EEG monitoring or brain imaging. A blood test could rapidly and cheaply be used to determine whether a seizure occurred and could be administered by staff with minimal or no specific neurological training.
Research by EpimiRNA and colleagues also drove the development of a series of prototype point-of-care miRNA detection devices. The central innovation is the generation of a dramatically amplified signal upon nucleic acid hybridisation based on nucleic acid functionalized electrocatalytic nanoparticles. During the project we helped advance this technology from a relatively slow, early prototype that could only detect a single brain-enriched miRNA (miR-134) to a rapid, point-of-care test (“TORNADO”) that featured move of the sample through a multi-chamber microfluidic system to allow rapid (~1.5 hour) detection of three different miRNAs simultaneously. Members of EpimiRNA continue to work on the development of this technology and are in advanced discussion about a spin-off company to commercialise the device. EpimiRNA’s exploitation efforts have identified at least one of the major manufacturers of anti-seizure drugs that would support efforts to commercialize a miRNA-based diagnostic device. Molecular diagnostics is the fastest growing segment in the diagnostics industry. Our pilot studies suggest sensitivity rivals high performing PCR target amplification but far surpasses it in terms of speed (results in seconds), operator simplicity and overall cost.
Having demonstrated that we can monitor a disease-modifying therapy by a blood-based miRNA panel, the EpimiRNA project has accomplished the expected future industry strategy of co-developed drugs and diagnostics that are increasingly important to pharmaceutical development. Our biomarker studies will advance in-step with the microfluidics device development to maximise the potential that both become available at the same time. We foresee our miRNA biomarker findings could lead to investment and our Technology Transfer Offices are in discussion to identify potential commercial partners and to identify, capture and commercialise further intellectual property from the project.
Key Impact 3: Preventing the development of epilepsy after potentially epileptogenic brain insults
Preventing the epileptogenic consequences of brain injury is an urgent therapeutic priority. EpimiRNA addressed this need by focusing on the development of miRNAs as therapeutic targets. In contrast to small molecules that hit just one single target, miRNAs can exert a "pleiotropic effect" as individual miRNAs regulate many targets in a network/pathway. At the time that EpimiRNA began, only a single study had investigated the functional consequences of targeting miRNA on the development of epilepsy. By the end of the project, research by EpimiRNA and others outside the project had tested and reported effects of over 20 different miRNAs on seizures or epilepsy. EpimiRNA took a set of miRNAs that were already disclosed, as well as novel miRNAs from its discovery work, and systematically targeted these in animal models. We answered key questions that would be needed to know before any kind of preclinical development could proceed. Are they safe? Do they work in multiple models? Do they affect normal animal behaviour? How important is the chemical “backbone” of the oligonucleotide antagomir? By what route could they be delivered to patients? What is the mechanism by which the miRNA inhibitor works? Can they stop seizures in already-epileptic animals? The scale and multi-disciplinary expertise within the EpimiRNA project allowed us to address all of these questions. The results will be encouraging to biotechnology or pharmaceutical companies considering this area of therapeutics for epilepsy or other diseases of brain excitability.
Advancing previously-known targets into pre-clinical development The EpimiRNA project made important progress in moving a miR-134-based antagomir toward preclinical development. By pooling resources, we demonstrated that the Ant-134 antagomir suppressed seizures in three other animal models and worked in rats as well as mice. Importantly, we demonstrated that normal firing properties of neurons in the hippocampus are not changed by the treatment – the molecule suppresses pathological brain activity while sparing normal functions. This strongly increases the likelihood that the treatment would be safe. We also showed that you could deliver the antagomir into a region of damaged (ultimately epileptogenic) brain tissue by a single systemic injection. We solved the main mechanism by which the molecule suppressed seizures in mice. Finally, we were able to show that a single injection of the therapy into already-epileptic mice permanently suppresses spontaneous recurrent seizures in mice. This is a critical test because the largest category of patients likely to receive this treatment would be patients where epilepsy is pre-existing but for whom neurosurgical treatment is not possible or for whom brain stimulation was ineffective, Together, the EpimiRNA project achieved all its objectives and strongly increased the translational value of the therapy. Partners at RCSI were awarded a US patent for this technology in 2017. During the project, partners at RCSI undertook collaborative pre-clinical development projects with US and European-based industry partners interested in using the therapy to treat drug-resistant epilepsy. This remains an area of ongoing collaboration. EpimiRNA activities have ensured biotechnology and Pharma companies as well as academic teams working on AED discovery, are provided with a robust rationale and series of already comprehensively pre-clinically tested miRNA manipulations, which will form the basis for future therapeutic approaches to prevent the development of epilepsy after epileptogenic brain insults. We are optimistic that a miRNA-based therapy for epilepsy can emerge from this research.
Identifying new targets for preclinical development: EpimiRNA generated a portfolio of pre-clinical assessments of novel miRNA-targeting antagomirs with potential applications in seizure control, epileptogenesis and disease-modification in epilepsy. In total, 9 additional miRNA-based therapies were tested during the course of the project in one or more animal model. One of these, Ant-135a, also prevented spontaneous recurrent seizures in epileptic animals, although the potency and duration of the effect was not superior to Ant-134. The remaining 8 antagomirs have only been shown to reduce seizures when given as a pre-treatment. It will be necessary, in future research by EpimiRNA partners or others, to explore whether these can reduce spontaneous seizures in already-epileptic animals.
Key Impact 4: Translating molecular and cellular targets into drug discovery
The EpimiRNA project has helped develop a non-miRNA inhibitor for seizure control. This recently entered the National Institute’s of Health’s epilepsy therapy screening program. The achievement was a collaboration between chemists at SME Bicoll and two of the academic partners who showed that the Bicoll molecule, now designated EPM06, has acute seizure-suppressive effects in multiple models and reduces epileptiform activity in the human brain.
Another outcome from EpimiRNA was an in vitro assay that can be applied to drug development that is based on identifying chemicals (either individual or initially as a mix) with miRNA-modulatory effects. The assay, based on a reporter molecule in a cell line that responds to the presence of a miRNA inhibitor, was used to screen a plant-based library from SME Bicoll for effects on miRNA processing. In the future, the assay could be useful directly or following further development by our SMEs, partners or future interests (e.g. Pharma).
Key Impact 5: Additional scientific impacts: miRNA and systems biology
The EpimiRNA project has had an important impact on our understanding of the molecular mechanisms underlying miRNA expression and malfunction. We have discovered how the levels of several miRNAs are controlled and we have generated new thinking about the importance of individual targets versus large-scale effects and shown this is miRNA-specific. EpimiRNA has developed novel molecular tools that should facilitate miRNA research in the future. This includes the development of miRNA reporter mice in which the AGO protein is selectively expressed in different cell types in the brain allowing researchers to assign specific miRNA species to neurons or non-excitable cells such as glia. This should lead to major strengthening of the Europe-based laboratories and SMEs focused on identifying miRNA targets and therapeutic opportunities. EpimiRNA also brought outstanding researchers not previously working on epilepsy to the field of epilepsy research including the teams of Gerhard Schratt (UNIMAR), Jorgen Kjems (AU) and Jens Anderson (SDU).
EpimiRNA has already had an impact on miRNA bioinformatics, generating new data integration infrastructure pipelines and a platform to ensure data from RNA-seq and proteomics are optimally mined and pathways identified and understood. The project has taken advantage of developments in the field of systems biology by developing innovative computational and mathematical modelling tools that explore the impact of miRNA on biological function at a systematic level, and that integrate qualitative and quantitative miRNA profiling, mRNA/protein target profiling, as well as neurophysiological data. These models now allow researchers to explore and explain miRNA effects, predict therapeutic targets, and deliver novel prognostic markers. Modelling of miRNA effects not only at the single neuron level, but also at the neuronal circuit/tissue level is furthermore a significant step towards more ‘holistic’ modelling approaches, and therefore also a major project impact.
Key impact 6: Scientific excellence, training and reputation
The EpimiRNA project has already delivered science of the highest quality and impact during the funding period and several major publications are in late development. This includes publications in Lancet Neurology, discoveries about mechanisms of miRNAs in EMBO journal and Cell Reports, the miRNA database in Bioinformatics and a series of miRNA biomarker studies in Scientific Reports and eBiomedicine. This supports Europe’s scientific and global R&D ranking. High impact publications will also attract the best collaborators from academia and industry to Europe to ensure the partners attract researchers of the highest level. The project helped grow the research teams of the participating partners and to retain existing talent (e.g. emerging investigators such as Dr Tobias Engel and Dr Eva Jimenez-Mateos who since obtained permanent faculty positions at top institutes) as well as attract new foreign and domestic researchers. Multiple members of the work package leader laboratories trained and supervised PhD students to completion who benefitted from the research excellence of the project and opportunities to learn from multi-disciplinary teams across Europe. Our innovative programme contained strong training for researchers including in much sought-after areas including pre-clinical animal models, clinical studies/trials, drug delivery/therapeutics, RNA chemistry and nanotechnology. The multi-disciplinary, inter-sectorial development and applied aspects of our research programme, meant team members came out equipped to secure industry as well as academic or other posts.
Key outputs from the research programme have been the establishment of new genetic models, discoveries on molecular mechanisms, technologies for gene and drug delivery to the brain and other “frontier research”. This will form the basis of more competitive applications by the project teams to the various European (H2020) funding mechanisms where the excellence of the science and record of the applicant group is fundamental to success. The success of the EpimiRNA project already drove the establishment of two epilepsy-focused national research centres. In Ireland, the EpimiRNA Coordinator is the Director of the FutureNeuro research centre. Two other EpimiRNA investigators are principal investigators in the Centre (J Prehn, G Cavalleri). The centre focuses on the development of novel diagnostics and innovative therapeutics enabled by the emerging connected (eHealth) sector. The project secured ~8€Mio in support from the Irish government and a further ~3€Mio in industry-committed investment. The clinical co-coordinator Prof Felix Rosenow established CepTER, the Centre for Personalized Translational Epilepsy Research, which secured ~5€Mio in funding in Germany. These projects build upon discoveries and infrastructure that were established through EpimiRNA.
Key Impact 7: Clinical and patient impacts from EpimiRNA
The biomarker, therapeutic and genetic discoveries made by EpimiRNA may lead to direct benefits to patients from faster diagnosis and new treatments but it may be several years before these will be realised. However, a number of discoveries may enter clinical use more quickly. Brain stimulation (BS) devices are already clinically important non-pharmacological treatments for pharmacoresistant epilepsy. Several are already approved but are currently only moderately effective. EpimiRNA has made important contributions to our scientific understanding of BS for epilepsy. First, we have identified a brain region that, when stimulated at a certain frequency, delays the onset of epilepsy. These animal data could be relatively quickly tried in patients and the EpimiRNA clinical network comprising multiple sites in Germany and Ireland could be involved in a future clinical trial. This could lead to new clinical investigations of alternate brain stimulation sites that could produce more powerful clinical effects or even disease-modification. Our studies found select changes to miRNAs occur in the brain in association with this therapeutic stimulation. While we could not prove that an individual miRNA was responsible, the project has generated several leads for further investigation that could identify seizure-regulatory miRNAs.
During our studies we investigated the potential therapeutic effects of transcutaneous vagal nerve stimulation (tVNS). This was a promising technology that offered a non-invasive
form of brain stimulation that avoids the need for surgery that is necessary for a VNS device. Clinical testing during the project lifetime failed to show a therapeutic effect of tVNS on drug-resistant epilepsy. Nevertheless, the device might still have utility, for example, in helping to predict who will best respond to VNS therapy. We identified a brain-specific miRNA that changes levels in the blood of a drug-resistant epilepsy patient given the therapy.
Human epilepsy genetics has the power to explain disease risk, identify new drug targets and avoid adverse drug reactions. EpimiRNA systematically explored all potential variation in miRNA and the pathway of their biosynthesis and all known targets in the hippocampus. We did not find that variations in miRNA could explain epilepsy, thereby refining our understanding of the genetic contributions to this disease. This may direct future research toward other regions of the human genome or stimulate an ever-larger study size which would be better powered to identify an effect of miRNA variation on disease risk. Regardless, this improves our understanding of epilepsy and can lead to revised standards that potentially reduce time to diagnosis and treatment providing potential healthcare savings and refine genetic counselling so patients have a better understanding of the cause and prognosis of their condition.
EpimiRNA’s research could - in the longer-term - help reduce health spending by delivering breakthroughs and infrastructure for diagnostics, new treatments and monitoring solutions that improve clinical outcomes. A prompt genetic diagnosis impacts on healthcare costs by speeding up time-to-diagnosis and helping better manage and monitor treatment compliance. EpimiRNA’s identification of molecular factors underlying drug-resistant epilepsy and biomarkers thereof could result in better matching of patients to treatments and fewer demands on healthcare resources creating savings for both hospitals and insurers. A patient with refractory epilepsy is more likely than a controlled patient to have a comorbid condition like depression, twice as likely to be hospitalized and when they are in hospital to stay longer. If a disease-modifying treatment emerged as a result of EpimiRNAs’ research it would have a transformational impact on long-term healthcare costs.
Key Impact 8: Benefit to participating SMEs
EpimiRNA’s core scientific objectives were all achieved and this was supported by excellent participation from our SMEs. EpimiRNA afforded our SME partners several opportunities to develop and protect novel intellectual property in the areas of therapeutics, medical devices for epilepsy, miRNA research tools and data integration platforms.
For Dixi, the project provided an opportunity to work with leading epilepsy surgery centres and develop a prototype combined intracerebral microdialysis probe. Due to challenges in preclinical testing the product did not enter clinical testing during the EpimiRNA project. However, should the company be interested in pursuing the device beyond the end of the project this could help grow their portfolio which includes intracranial electrodes for use in patients. There have been no competitor products that combine recording and microdialysis functions. The device also has the potential to be further developed into a local drug delivery system, resulting in the desired “closed-loop” single diagnostic and therapeutic delivery system for patients. This could also be extended to other pathologies such as paroxysmal pain or brain tumors; very important future markets.
The global epilepsy neurostimulation market is estimated at $0.5 billion. Cerbomed’s tVNS device was trialed during the course of the project for drug-resistant epilepsy but was not found to be efficacious. We did, however, identify a miRNA biomarker that could be useful in the future to predict responders. The company is currently being acquired in a merger. While the future of the tVNS system is therefore uncertain, the project nevertheless delivered clinical data on safety and tolerability and EpimiRNA’s findings may encourage future research of similar devices.
Targeting small non-coding RNA such as miRNA is a growing area of biomedical research with clear therapeutic applications. During the EpimiRNA project, SME InteRNA performed RNA sequencing and miRNA research that has supported their interest in expanding beyond cancer therapeutics to the central nervous system. EpimiRNA provided InteRNA early integration with a new field of research and access to expertise from multiple RNA specialists. With EpimiRNA, InteRNA were able to develop their capabilities in RNA research and therapeutic applications which will broaden and increase their potential market.
Bicoll currently offers small molecule drug-like compounds from natural resources for the global pharmaceutical and biotech industries. Through EpimiRNA, Bicoll was able to expand their focus on medicinal chemistry and custom synthesis and screen compounds from within their catalogue that impact on epilepsy or target novel processes, including miRNAs. As a result of EpimiRNA, Bicoll now has a drug lead candidate accepted into the NIH’s Epilepsy Therapy Screening program. This provides access to a battery of preclinical test models and will evaluate whether their lead molecule has a realistic chance of being developed into a new medicine. If successful, Bicoll would be likely to partner with an existing company with expertise in the epilepsy field.
The global market for genetic data analysis and management software continues to grow, partly as a result of high throughput „omics“ data opening new markets for data management and analysis. EpimiRNA enabled BCPlatforms to move towards the area of managing data efficiently and securely including data from „omics“ hetereogenous biological data which increasingly comes from international multi-institutional groups (their primary market currently). The project resulted in BCPlatforms creating cost effective tools for „omics“ platforms that can merge RNA-seq, ChIP-seq, proteomics and genome sequencing, considerably widening their potential market.
Key impact 9: Formation of a “spin-out” company
Another important achievement of the EpimiRNA project was the development of a new spin-out company (SME) from partner AU. The company, named Omiics, is being led by Dr Morten Veno who was responsible for the RNA sequencing efforts and iCLIP technique that were developed within EpimiRNA. Omiics, will offer bespoke RNA discovery and informatics resources to academic and commercial entities.
Key impact 10: Broader European and societal benefits of EpimiRNA
Partners in EpimiRNA have extensively engaged with patient organisations and the lay public during the course of the project. The focus has been on disseminating findings and reducing the stigma associated with the disease. This includes appearances on national television programmes to explain the disease and the type of research being performed in the project as well as lecture series at host institutes for patients and families. Researchers in EpimiRNA have also performed visits to schools to de-mystify careers in science and introduce concepts about how the brain works and help them understand a disease such as epilepsy. Through the epiXchange conference, members of EpimiRNA engaged more closely with project officers at the EC to provide oversight of S&T developments and influence future funding directions. The EpimiRNA project has also helped deliver on commitments made at a national and European level to increase investment in epilepsy research, which was mandated by the European parliament’s acceptance of the written declaration on epilepsy.
The EpimiRNA project has maintained the leading position internationally into the contribution of miRNAs to epilepsy. At the time of project start, this was driven principally by European groups, including the first functional studies of miRNAs. EpimiRNA and miRNA-focused research within other projects (e.g. EpiTarget) have ensured that half of the published research on miRNAs in epilepsy has come from European labs. The project has enabled this momentum to continue and ensued European-based teams continue to make major breakthroughs. EpimiRNA has also catalyzed greater interactions between EU nationalities. The EpimiRNA consortium’s large-scale, multi-disciplinary focused approach has effectively consolidated the European competitive advantage and attracted investment from outside (for example, the USA’s Regulus therapeutics funding a pilot miRNA therapy project within one of the EpimiRNA teams). This dominance is important because breakthroughs often translate into patent filings and other outputs. Indeed, the Coordinator’s institute RCSI was awarded the US patent for a miRNA-based treatment for epilepsy in 2017.
In addition to the scientific and technical achievements, EpimiRNA was central to the development of the epiXchange initiative – a large collaborative network comprising the major FP7-funded epilepsy projects and a number of other major European epilepsy-associated research projects. This resulted in a high-profile event in Brussels in 2018 (see section 1.4.2 for more details) and has since led to establishment of a network that aims to foster cross-project collaboration on epilepsy and jointly pursue research funding in the future.
The EpimiRNA consortium is distinct from other large project approaches that focus just on biomarkers or groups of genes involved in a single process (e.g. inflammation, extracellular matrix, neuropeptides) that are ongoing both in Europe and in the USA. The EpimiRNA project has established a unique programme that brings distinctiveness to European epilepsy research. EpimiRNA nevertheless complements large European and International research efforts and partners in EpimiRNA have been active in other FP and Horizon-2020-funded programmes. 1.4.2 Measure to maximize impact: Main dissemination activities
EpimiRNA communicated about the project’s main objectives and disseminated its main results using a variety of tools. These communication tools were carefully designed to each target audience from the scientific community to patient organisations and policy-makers. EpimiRNA-related activities were promoted on Twitter using @Epilepsylab Twitter account from 2014 to 2016 and a dedicated project Twitter account (@EpimiRNA) from January 2016 onwards, with 295 and 151 followers respectively, as well as 117 and 143 tweets and retweets respectively.
Promoting miRNA research within the scientific community
At the start of the EpimiRNA project, research on miRNAs was at its infancy and very little was known about their role in epileptogenesis and chronic epilepsy. We contributed to uncover new miRNAs and demonstrate their mechanism of actions, making miRNA a credible approach to treat and diagnose epilepsy. Over the course of EpimiRNA, our research already resulted in more than 20 publications in peer-reviewed journals and many more publications are still to come. In addition to talks during scientific conferences and posters, EpimiRNA partners contributed to and led several workshops on epilepsy-miRNA related research at major European (European Congress on Epileptology (Stockholm 2014; Prague 2016, Vienna 2018)) and International conferences (International Epilepsy Congress (Barcelona 2017)) showing that miRNA in epilepsy is now a credible field of research and triggering interest for this research area.
Targeting patients/ lay audience to inform about state-of-the art of epilepsy research
Given the novelty of the miRNA approach to treat and diagnose epilepsy, the focus of the dissemination materials towards a non-scientific audience was always to explain first what miRNAs are and which role do they play at cellular level, before explaining the project objectives and main results. The main communication tools towards the lay audience have been the EpimiRNA website (updated constantly during the project life-time), the release and distribution of the project flyer (printed and electronic version), as well as the production of 3 short videos , available from the EpimiRNA website https://www.epimirna.eu/media/videos/
From 2015 onwards, several face-to-face events have been organised with patient organisations (e.g. Epilepsy Ireland, Ireland’s epilepsy charity) at the coordinator host institution to demystify epilepsy and explain how research on miRNAs could help tackle the many challenges that epilepsy patients are currently facing.
International and European media coverage consisted in several articles being published in the EU’s Horizon magazine, as well as television, radio and press coverage, especially in Ireland where the host institution of the project coordinator is located.
Working hand in hand with epilepsy lobby organisations and other epilepsy-funded projects to promote an agenda for epilepsy research
EpimiRNA partners worked hand-in-hand with six other FP7-funded large collaborative projects (EpiTarget, Desire, EpiSTOP, Epixchange, Epicare and EpiPGX) and existing epilepsy lobby organisations (such as ILAE) to promote a future research agenda for epilepsy. These joint efforts were particularly helpful to target policy-makers at the European Commission and the European Parliament (especially the members European Advocates for Epilepsy working group). The main joint action consisted in the organisation of the epiXchange 2018 conference, a major international conference that was organised in Brussels (23rd May 2018) and that brought together representatives from the industry, patient organisations, research community and policy-makers. The programme of this conference aimed at showcasing the results of the FP7-funded projects and highlight existing research gaps. A paper is currently being drafted and will highlight the major achievements of the projects, as well as recommendations for future epilepsy research. This paper will be submitted shortly.
Some numbers from the epiXchange conference:
• 170 participants from 18 European and 5 non-European countries
• 250 watching the live-stream on Facebook
• Participation of officials from the EC, the European Brain Council, EMA, patient organisations (e.g. Epilepsy Foundation and Federazione Italiana Epilessie) and pharma (strong presence of UCB Pharma)
• 28 talks and 38 posters divided into 5 main themes
• 800 Tweets & retweets from October 2017 to August 2018 to promote the event
On the epixchange website. www.epixchange2018.eu all material from the conference can be found, including the full live-stream (and special cutouts to dedicated part of the conference) and presentations (as pdf).
Promote uptake of project results by industries outside the consortium
During the project life-time, we maintained continuous dialogue with potential industry partners to support the transition from basic research to the clinic. This has resulted into two active collaboration programmes with a pharmaceutical industry to drive the pre-clinical development of a new therapy epilepsy. At the time of writing this report, RCSI were contacted by a leading company that manufactures anti-epileptic drugs to explore potential development of a blood test for epilepsy diagnosis. Discussions are currently ongoing.
Training and harmonization of procedures to ensure high-quality research
Given the multidisciplinary research carried out in EpimiRNA, a number of training visits were organised between EpimiRNA partners during the project life-time. Among other things, partners were trained to perform high-quality hippocampal microdissections, RNAseq data.analysis in vitro methods for electrophysiology, ia KA model and antimir injections.
SOPs for key techniques used at multiple sites were prepared. These relate to techniques for freezing brain tissue, collecting blood and processing plasma and preparation of CSF.
List of Websites:
www.epimirna.eu; Coordinator: David Henshall, RCSI, dhenshall@rcsi.ie
Epilepsy is a brain disease characterized by recurring, unprovoked seizures that affects about 6 million people in Europe. Temporal lobe epilepsy is the most common syndrome in adults and the most prone to drug-resistance. A number of priorities exist to improve the lives of patients with temporal lobe epilepsy, including the need for more effective anti-seizure drugs and disease-modifying therapies. Evidence has emerged that epilepsy and epileptogenesis (the process underlying its development) feature large-scale changes in gene activity. If the brain molecule(s) that drive or coordinate multiple genes could be identified, this would significantly improve our understanding of epileptogenesis and may lead to new thinking about the treatment of epilepsy. If the same brain molecules can be detected in a blood sample or other fluid, this could provide new ways to diagnose epilepsy or predict response to therapies.
The EpimiRNA project chose to focus on a newly-discovered molecule that coordinates gene networks - microRNA (miRNA). These are short stretches of RNA – closely related to DNA – which function in cells by attaching to the much longer messenger RNA (mRNA) transcripts that are the instructions for making proteins. The miRNAs fine-tune protein levels in all cells, including the brain, and a particularly important feature is that each miRNA regulates many different mRNAs; they are the “conductors of the protein orchestra”. The EpimiRNA project began in 2013. Originally 16 partners from eight different European countries plus the USA and Brazil, the consortium brought together for the first time experts in pre-clinical and clinical epilepsy research with experts on human genetics, chemistry, drug discovery and mathematical modelling. The project was complemented by research-active small and medium enterprises (SMEs) developing therapeutics, diagnostics and devices. The EpimiRNA project had five main objectives. We would measure the levels of every active miRNA in the epileptic brain. Then, develop mathematical models and technologies to explain how they controlled brain excitability. We would look in blood and brain fluids to see if any miRNAs were unique to epilepsy or could predict responses to therapies. We would look for variation in the sequence of miRNAs in the genetic code of 3500 patients and controls. Finally, we would test whether altering amounts of miRNA in the brain could reduce seizures or reverse epilepsy in models of epilepsy.
EpimiRNA successfully completed all its main objectives. We generated searchable databases containing over 5 billion data-points cataloguing all functional (active) miRNAs in three models of epilepsy and brain tissue from patients. Epilepsy caused changes to nearly half of all the brain-expressed miRNAs. We developed new models to predict targets of miRNAs and obtained direct evidence that the protein landscape of the epileptic brain is shaped by miRNAs. We identified miRNA “biomarkers” in blood and brain fluid and developed a prototype device to quickly detect miRNAs that can support the diagnosis of epilepsy. We found a new site deep within the brain that, when electrically stimulated, protects against epilepsy development. We found that temporal lobe epilepsy patients do not have rare variants or mutations in miRNAs (or the enzymes which produce them) in their genomes. Finally, we have identified 10 new molecules that produce anti-seizure effects in pre-clinical models of epilepsy. One of these stops epileptic seizures after a single injection, for which a patent was awarded. Two of our compounds are now in pre-clinical development as treatments for epilepsy. Our industry partners were able to bring compounds toward pre-clinical development, test new brain stimulation devices, refine and improve RNA technology, develop new data storage and management systems, and build a prototype device to simultaneously record brain activity and deliver therapies to patients.
EpimiRNA has generated important breakthroughs and advances in our understanding of miRNAs in the development of epilepsy and demonstrated this can support diagnosis and treatment (or even cure). We urge further research to understand miRNA control of brain excitability and translational and applied aspects, including diagnosis and treatment epilepsy.
Project Context and Objectives:
1.2.1 Background & Aims
Epilepsy is the most common serious neurological disorder for which there is no cure (European Written Declaration on Epilepsy, 2011). The disease is characterized by recurring, unprovoked seizures. There are 6 million people with active epilepsy in Europe and 300.000 further cases each year. The costs are estimated at nearly €14 billion a year in the EU. Temporal lobe epilepsy (TLE) is the most common form of epilepsy in adults. TLE can develop as a result of brain trauma, infection, status epilepticus (prolonged, continuous seizure), and is frequently associated with hippocampal pathology (hippocampal sclerosis; HS). People with epilepsy have increased mortality, experience high levels of unemployment and under-employment, and are exposed to stigma and prejudice. Epilepsy – particularly when poorly controlled by medication – not only damages health, but disrupts many other aspects of living, imposing very significant physical, psychological and social burdens on individuals, families and caregivers. In addition, diagnosis of epilepsy remains principally based on clinical history and examination. Electroencephalogram (EEG) recording, brain imaging and genetic testing provide important supports but mis-diagnosis rates remain unacceptably high.
A number of priorities exist to improve the lives of patients with epilepsy, including the need to discover anti-epileptogenic and disease-modifying therapies. The identification of measurable diagnostic indicators of epilepsy or response to therapies (e.g. biomarkers) and the epileptogenic process would also transform discovery of disease-modifying therapies and the diagnosis of epilepsy in future. However, no molecules to date could account for the complex gene expression landscape of the epileptic brain and be both biomarker and therapeutic targets. Do such molecules exist?
The EpimiRNA project focused on a completely new class of molecule that had recently been discovered inside cells called microRNA (miRNA). These are short stretches of RNA – the chemical cousin of DNA – which function by binding onto the much longer protein-coding messenger RNA (mRNA) transcripts that are the instructions for making proteins (DNA is the blueprint for the mRNA which is then read and a protein is produced). Through a series of important breakthroughs in cell biology, it was revealed that miRNAs fine-tune gene expression (e.g. proteins levels) in all cells. A particularly unique feature of miRNAs is that they are multi-targeting - a single miRNA can bind dozens of different mRNAs-, producing effects ranging from mild to strong suppression of protein levels. The human genome contains about 2000 different miRNAs. They are initially produced as longer precursor RNAs before undergoing a series of enzyme reactions that result in a duplex miRNA. One strand of this – the mature miRNA – is selected and loaded into a binding pocket by Argonaute (AGO), a protein that function as the vehicle of miRNA-directed control of gene expression.
The AGO protein next begins a search process, trafficking along mRNAs until it finds a site where the miRNA matches the complementary sequence in the mRNA. If this happens, the mRNA is either prevented from being read or is degraded by the cell. The brain contains more miRNA than any other organ and malfunction of the miRNA system resulted in brain mal-development. Experiments soon clarified why: miRNAs control about half of all genes and ensure protein levels are tightly regulated, including those that ensure the physical connections between neurones (synaptic communication).
In 2010, reports emerged that levels of certain miRNAs were altered in a part of the brain – the hippocampus – long-known to be a trigger point for seizures. The same year, studies in rats reported that levels of certain miRNAs changed in the blood after a seizure. The next two years saw the first evidence that blocking an individual miRNA using synthetic DNA-like oligonucleotides (“antagomirs”) could protect the brain from damage or seizures. Around the same time, an antagomir – miravirsen - successfully completed a clinical trial (for Hepatitis C), heralding a new era in medicine – therapeutic miRNA targeting.
The EpimiRNA project began in 2013 – originally 16 partners from eight European countries, the USA and Brazil. The consortium was new, with few of the partners ever having worked together before, and had a very innovative approach: A core team of experts in pre-clinical and clinical epilepsy research, then experts in human genetics, cell biology, RNA detection (sequencing), protein analysis and mathematical modelling. The EpimiRNA consortium was complemented by research-active small and medium enterprises (SMEs) developing RNA therapeutics (InteRNA), and diagnostics and devices for the treatment of epilepsy (Dixi, Cerbomed, Bicoll, BCPlatforms). Devices for non-pharmacological treatment of epilepsy – brain stimulation techniques including vagus nerve and deep brain stimulation (VNS, DBS) – had become of increasing clinical importance, yet lack of understanding of their mechanisms of action limited our knowledge of which patients benefit from their use. EpimiRNA aimed to track miRNA changes after seizures (including in humans), and in response to treatment, explain how miRNA influences epileptogenesis (e.g. epilepsy development) and ictogenesis, investigate genetic variants in TLE patients to learn if this caused their epilepsy, and ultimately use the discoveries to establish miRNA-based therapeutics and biomarkers to track, treat and prevent epilepsy.
1.2.2 Main objectives of the project
The overarching objective of the EpimiRNA project was to understand the contribution miRNA makes to the development and maintenance of the epileptic state, their potential to diagnose the disease, and opportunity to target these to treat or prevent epilepsy.
Objective 1 Identify conserved changes in functioning miRNAs in epileptogenesis and determine the mechanism(s) by which miRNA changes contribute to epileptogenesis
Gaps in our understanding of the epileptogenic process strongly limited our ability to develop more effective treatments. The multi-targeting properties of miRNAs represent a novel means to coordinate epileptogenesis but it was not certain that miRNA were really shaping gene expression in epilepsy. EpimiRNA aimed to perform a systematic, unbiased and quantitative analysis of functioning miRNA in three of the best animal models of epilepsy. EpimiRNA would determine the impact of miRNA changes on the “proteome” – the complete collection of all proteins in a cell or tissue - and measure how miRNA manipulations altered the physical structure and function of different brain cell types. Together, our approach would produce a major advance in our understanding of which miRNAs contribute to epileptogenesis and explain the underlying mechanisms.
Objective 2 Identify the miRNAs that are functional in the human TLE brain and evaluate how non-pharmacological interventions including brain stimulation modulate miRNA
Although human data on miRNAs in epilepsy existed before EpimiRNA, small group sizes, incomplete coverage of miRNAs and lack of evidence of function limited insights; a systematic approach was needed. Additionally, biomarkers of the epileptogenic process would be essential for the development of disease-preventative therapies in the future. Circulating miRNAs offer novel, mechanistic biomarkers of epilepsy but it was unknown if levels in blood or other biofluids were altered in patients or responsive to therapy. EpimiRNA would undertake a study to identify and validate miRNA biomarkers of epilepsy and identify the targets of the miRNAs in brain tissue. EpimiRNA would also investigate whether miRNA levels change in biofluid samples from patients receiving novel therapies - brain stimulation - for treatment of seizures, and try and find new brain stimulation sites. This would not only identify potential biomarkers of pharmacoresistance (or seizures) but would also provide an opportunity to improve prediction of efficacy of brain stimulation (patient stratification).
Objective 3 Characterize genetic variation in miRNAs in human TLE
Human epilepsy genetics has been a powerful tool for discovering genes responsible for inherited and new forms of epilepsy and can sometimes predict adverse reactions to anti-epileptic drugs. But for the majority of patients, variation and errors in traditional genes such as ion channels do not explain their epilepsy. Sequence variation in miRNAs, or in the various genes involved in miRNA biogenesis or function could be causally important in human epilepsy. EpimiRNA would undertake the first targeted sequencing focused on miRNAs and on the genes involved in their biogenesis, specifically in TLE with HS, a major syndrome prone to pharmacoresistance. Focusing on a single, well-defined population would maximize the likelihood that we identify significant risk variants and the large-scale collaborative nature of EpimiRNA could ensure high numbers of samples could be analysed.
Objective 4 Use systems biology to integrate miRNA bioinformatic and functional data to explain how miRNA expression changes trigger epileptogenesis
Predicting effects of miRNAs on gene expression is a major challenge - single miRNAs often produce only small changes in protein production but when several miRNAs target the same mRNA additive effects can produce strong suppression. There are often sharp, bi-directional changes in miRNA levels within specific cell types adding to the complexity of understanding miRNA effects on epileptogenesis. This had not been factored in any epilepsy studies to date. A key task of EpimiRNA would be to establish biologically meaningful connections and link these to disease processes by generating mathematical models that take account of the behaviour of entire gene pathways. Together, this would allow us to better predict the impact of miRNAs on brain function, seizures and identify novel targets.
Objective 5 Identify novel miRNA-modulating molecules as future therapeutics for epilepsy
MiRNAs represent an entirely new target for the treatment and prevention of epilepsy. Oligonucleotides targeting miRNAs have been tested in humans and these so-called “antagomirs” are highly potent and long-lasting - a single dose inhibits a miRNA for more than one month. Just how many high-value miRNA targets exist for epilepsy was, however, unknown. Reports were few and no study had validated an effect in another model. EpimiRNA would develop and test miRNA inhibitors using antagomir and other approaches, targeting those miRNAs causally important for epileptogenesis or attempting to reverse already-established epilepsy. A further innovation would be to measure miRNAs in biofluids in the animals during the epilepsy monitoring studies and record how gene expression changes in response to miRNA-based interventions to understand mechanisms of action.
1.2.3 Work Strategy
The discovery of miRNA un-locked a previously unknown layer of gene expression control with important implications for epilepsy. Changes in levels of miRNAs or mutations might be a key contributor to the altered gene expression landscape in the epileptic brain. Targeting miRNA could be a way to protect the brain from seizures.
Detecting circulating miRNA in blood or other biofluids could help with diagnosis. The five Objectives were developed into tasks that fell into a series of 11 work packages (WPs). Each WP had a leader, an expert in the main scientific and technical demands of the experiments, along with several other teams bringing expertise to ensure the research was achieved:
WP1: Characterise miRNAs in brain and biofluids from three different animal models to identify the miRNA changes that occur during epileptogenesis.
This work package was led by the Coordinator’s team at RCSI and UNIMAR and UNIVR generated the samples from the different preclinical models. The AU team analysed miRNA in brain tissue whereas RCSI analysed blood miRNA. SDU performed protein analysis.
WP2: Characterise the miRNA profiles in brain tissue from epilepsy patients and miRNA levels in biofluids before and after seizures in patients. Also, develop a new combined intracortical EEG-microelectrode.
Led by Prof Hamer’s group at UKER, the device development was performed with DIXI and collaborations across all clinical sites provided the samples with miRNA analysis and bioinformatics performed by RCSI teams.
WP3: Define miRNA changes following experimental brain stimulation as well as determine whether brain stimulation in patients or volunteers produces a miRNA signature that can be used as a biomarker.
Led by clinical co-coordinator, Prof Rosenow, samples were analysed by AU and a clinical trial performed using Cerbomed’s brain stimulation device.
WP4: Identify the gene targets of the identified miRNAs and assess the impact of manipulating these miRNAs on animal behaviour and brain excitability.
Prof Pasterkamp’s team with InteRNA, UCL’s Prof Schorge, UNIMAR and RCSI studied how various miRNAs control brain structure and function at a cell and subcellular level.
WP5: Determine the genetic variation in miRNAs, their biogenesis enzymes and the hippocampal target genes in epilepsy patients.
Prof Goldstein’s and Prof Lopes-Cendes’ teams performed the genetic study of miRNA variation in epilepsy patients, with DNA provided by multiple partners and collaborators.
WP6: Develop an integrated data storage and retrieval platform for the EpimiRNA datasets and support data mining, hypothesis testing and biomarker identification.
Prof Kjems and team at AU performed RNA-sequencing and developed several technologies including iCLIP. Data storage requirements were provided by partner BCPlatforms.
WP7: Develop computational models that can predict effects of miRNA changes on network excitability which can be fed back for testing and biomarker prediction.
Prof Prehn’s team at RCSI supported the project’s bioinformatics needs, developing new tools to search for targets of miRNAs.
WP8: Perform in vivo manipulations of miRNAs and assess the impact on epileptogenesis in animal models.
Prof Schorge’s team with RCSI tested numerous miRNA inhibitors and other compounds and explored how these new drug types affect seizure susceptibility and epilepsy in models.
WP9: Develop in vitro screening assays to identify natural compounds that display miRNA inhibitory properties that could be developed as anti-seizure drugs.
Dr Lamottke and Bicoll led the investigation of plant-based compounds for epilepsy and developed an anti-seizure small molecule with the teams at UCL and RCSI.
WP10, WP11: Ensure the dissemination of EpimiRNA results, inter-site training and exploitation of discoveries as well as project management.
Coordinated by RCSI with support from ARTTIC, this focused on scientific and communication achievements and the smooth running of this large and complex project.
Project Results:
1.3.1 Overview
The main results of the EpimiRNA project fall into six categories according to the main objectives. These are summarised as:
(1) Identification of novel miRNAs in epilepsy. The EpimiRNA project has generated genome-wide datasets identifying all functioning miRNA in experimental and human epilepsy. This includes every miRNA in the rodent hippocampus and how its level changes during epilepsy development in three different models and every miRNA in the human epileptic brain and its target(s).
(2) Deep mechanistic understanding of how miRNAs change neuronal networks in the brain. Our studies have identified the targets of numerous new miRNAs and mechanisms by which this regulates brain excitability. By locating the miRNA within causal pathways and identifying their target(s) our results explain how miRNA changes influence the epileptogenic process and lead to lasting changes in brain excitability. The data have proven effective for predicting novel targets for anti-epileptogenesis strategies.
(3) Identification of miRNA biomarkers of epileptogenesis and seizures. We have identified miRNAs in blood and cerebrospinal fluid that distinguish TLE patients from controls and other neurological disorders. These findings could be used to develop tests to support diagnosis, help predict disease risk and course/prognosis, direct treatment and avoid adverse side-effects. miRNA biomarkers of non-pharmacological treatments (e.g. brain stimulation) were found and may improve use parameters to help patients better control their seizures with fewer side effects.
(4) Novel therapeutics by targeting miRNAs. We have tested more than 10 different miRNA inhibitors (antagomirs) during our project, of which more than half produced anticonvulsant effects in animal models. Two of the antagomirs prevented spontaneous recurrent seizures. This has led to a patent and preclinical development of a miRNA-based disease-modifying therapy for epilepsy, with the hope to reach clinical development in future.
(5) Investigation of genetic variation in miRNAs. By searching 3500 patient and control genomes, we have discovered that mutations or variations in miRNAs, their biogenesis genes and their main targets in the hippocampus are unlikely to cause human TLE. This resolves a major question in our pursuit of the genetic causes of epilepsy. It could provide potential healthcare savings in terms of genetic testing and refines genetic counselling so patients have a better understanding of the cause and prognosis of their condition.
(6) Technological advances and resources. We have made a number of important technological advances and have delivered training of these, including techniques to characterise miRNAs with small RNAseq, as well as generating a systems biology framework for miRNA. EpimiRNA has delivered advances in oligo/DNA-based and traditional small chemical scaffolds. The biomarker findings have driven development of new technology for rapid, direct, detection of epilepsy-associated miRNA. EpimiRNA has generated research databases and a new data integration platform for RNA-seq, genetics and proteomics.
1.3.2 Objective (1) Identification of novel miRNAs in epilepsy
One of the most ambitious, multi-partner collaborative projects within EpimiRNA was to catalogue every miRNA that was functional in the epileptic brain. We would do this for three common animal models of epilepsy and the human brain. This ambitious goal was fully achieved. We produced a dataset on the relative amounts of every miRNA in the brain of a rat or mouse and whether and when this changes in epilepsy. We have identified the target(s) of every miRNA in the human epileptic brain.
Measuring every functional miRNA in the brain: AGO-Seq
The approach the EpimiRNA project took to measure every functioning miRNA required us to pull down the protein responsible for miRNA function. The protein is called AGO (argonaute). This is done by mixing an antibody raised to detect the AGO protein with the brain sample. This can then be separated and the miRNA extracted and identified. For the identification step, we were able to use RNA sequencing, a powerful technique that identifies all miRNA sequences and can tell you the relative amounts of each. The team at AU developed the technique and, once perfected, began processing tissue for AGO-Seq. Meanwhile, the teams at RCSI, UNIMAR and UNIVR were generating brain tissue samples from each of their animal models. We wanted to know what every miRNA was doing at every important step during the development of epilepsy and we therefore included time points that corresponded to when epilepsy was developing, the day of the first spontaneous seizure and when the animals had active epilepsy. We monitored every animal, so we knew precisely when their epilepsy emerged and how many seizures they were having. The brain tissue samples were finished during the second year of the project and were shipped to AU for analysis. By the end of the third year we had our complete miRNA functional catalogue.
Comprehensive profiles of functional miRNAs in three animal models of epilepsy
Altogether, the AU team generated 240 datasets – over 3 billion data-points - on the relative amount of functional miRNAs in three rodent models, at multiple different time points and three different regions of the hippocampus (the brain region involved in human TLE). One of the first checks we made was to look for miRNAs known to be abundant in the hippocampus. One of these, let7c is especially enriched. When we checked levels we found it to be among the highest of any miRNA (~15 on our scale that ran from 0 to 17). miR-9 is another abundant miRNA in the brain. Levels of miR-9 were also around 15 on our scale. Then we checked for a less abundant miRNA, miR-134-5p (one we had previously linked to epilepsy). Levels of miR-134-5p averaged around 7 on the scale. Finally, we looked up miRNAs that were only expressed outside the brain. One of these is miR-122, a liver-enriched miRNA. We couldn’t detect any miR-122 in any of the brain samples at any time in any model. We also looked at levels of miRNAs between the models. We had two mouse models but the particular strain of mouse was different. We would still expect, however, that the amount of any miRNA should be quite similar. This is exactly what we found. Taking miR-9 for example, in the RCSI mouse the average level was ~15.2 on the scale. In the mice from UNIVR it was 14.9. Another, miR-124-3p, was found at 11.7 in both mice.
We now started to look for miRNA changes during epilepsy development. We first checked for some of the miRNAs that had already been linked to epilepsy. We started with miR-146a and miR-132, which had been reported by multiple teams to be upregulated in epilepsy. We found both to be upregulated in all three models, but the timing was quite different. miR-132 levels went up almost immediately whereas miR-146a was much later.
We also had data on the amounts of different miRNAs within three different parts of the hippocampus. This is useful because the hippocampus of humans with TLE shows patterns of neuronal loss and glial scaring that are not uniform. Within the EpimiRNA data set we found sets of miRNAs that were very evenly expressed across the three regions as well as examples of miRNAs that were at higher levels within one or other part.
In addition to providing a rich dataset on all functional miRNAs in the normal brain and subfield-specific enrichment, the project generated findings on miRNA levels in epilepsy. Data from analysis of the mouse model from RCSI showed there were 476 miRNAs detected. This is close to the approximately 500 known miRNAs in the mouse genome. This confirms the brain as the most miRNA-enriched organ. From among the detected miRNAs in the RCSI model, 225 were altered at one or more time points during epilepsy. Thus, epilepsy induces changes to nearly half of all the brain-expressed miRNAs.
The results from the other two models were similar. From the UNIVR mouse model we detected 481 distinct miRNAs and of these, 291 showed a change in the epilepsy model. In the rat model we detected 419 of which 147 were altered in epilepsy. We were also able to study what happened to different miRNA families - sets of miRNAs produced from the same precursor DNA sequence. We found strong evidence that miRNA families are co-regulated in the epilepsy models.
Finally, we focused on miRNAs that were common to all three models at two main time points. The latent period (the period of time after the insult has occurred but before a first epileptic seizure occurs) and the chronic epilepsy period. We found eight miRNAs were common to all three models in the chronic epilepsy phase, all of which were identical in humans. Some were already known to us, namely miR-132-3p and miR-146a-5p. Then we had six miRNAs that were potentially novel. These were selected for functional testing.
Mapping the brain targets of miRNAs in human epilepsy
In parallel with the above efforts, the UNIMAR and AU teams started a project to map all the miRNAs in the human epileptic brain and their targets. We wanted to know which miRNAs were present in the human epileptic brain and what they targeted. This could not be done in the same way as the previous study and needed a new technique, called iCLIP. The principle is to bind together the miRNA to whatever it is targeting, extract the complex and then identify both. The difficulty was that iCLIP uses a number of complex chemical processing steps and many of the early efforts failed. Eventually, during the third year of the project the AU team fixed the technical problems and it worked. The team then took a set of human brain samples from TLE patients and mapped every miRNA to every targets.
The result is the first ever complete catalogue of miRNAs in human epilepsy with their targets. This unique resource allows EpimiRNA partners (and eventually any researcher) to look up any miRNA of interest and find what it binds to or look up a target of interest and determine which miRNAs are binding to it. The results are also semi-quantitative so more abundant miRNAs tend to have higher numbers of targets linked to them. The dataset can also be used to generate new approaches to drug therapy. For example, you could look up miRNAs that are targeting a particular under-produced protein in epilepsy and then deliver a miRNA inhibitor to the tissue to increase protein levels.
In addition, we used small RNAseq to identify the differentially expressed miRNAs in human resected brain tissue (the iCLIP technique tells you what the miRNA is attached to but is not a suitable method to determine how much of a particular miRNA is present. For this we compared hippocampus from samples with sclerosis versus those without. The results were the identification of numerous differentially expressed miRNAs including known as well as novel miRNAs.
The human miRNA-seq data was subjected to differential expression analysis, showing 85 miRNAs significantly differentially expressed between cases of temporal lobe epilepsy cases with and without hippocampal sclerosis (TLE-HS vs. TLE-nonHS). This included lower levels of miR-22-3p in the sclerotic samples, consistent with a protective role (which we experimentally verified in Jimenez-Mateos et al. (2015)).
Main conclusions and future objectives: EpimiRNA identified and measured all known miRNAs in the rat, mouse and human brain in epilepsy. We identified the targets of many of these miRNAs. The results are the first ever spatiotemporal atlas of functional miRNA changes in epilepsy. This advances knowledge of miRNAs, emphasizes the profound changes to miRNA in epilepsy and can be used to select miRNA to target for seizure control.
1.3.3 Objective (2) Deep mechanistic understanding of how miRNA changes de-stabilise neuronal networks in brain
Introduction
At the time the EpimiRNA project started, a handful of miRNAs had been found to influence seizures in rodents. This was based on observations of the effects of blocking or boosting the amount of the miRNA in the brain and recording a change in severity of seizures. Nobody knew at the time the mechanism by which a given miRNA could exert such an influence. This is important because understanding mechanisms of action can lead us to further drug targets. Even if the miRNA itself is not druggable, perhaps we can design drugs that work on the targets (proteins) instead. However, studying miRNA mechanisms is challenging because each miRNA has many possible targets. A miRNA action in a seizure model may be due to strong effects on one target or weak effects on many. EpimiRNA set itself the challenge of determining the mechanisms of as many epilepsy-associated miRNAs as they could. Overall, the project solved the molecular mechanism of seizure effects for three miRNAs, partially solved the mechanism of several more and for some others we now have strong leads but experimental proof will require further efforts.
miR-22 regulates epileptic seizures by suppressing the P2X7 receptor
Partners in EpimiRNA were the first to solve the molecular mechanism by which an individual miRNA regulated epileptic seizures. The RCSI team had discovered miR-22-3p in part of the brain circuit involved in seizures which normally showed no signs of pathology. We reasoned that miR-22-3p might protect this part of the brain and we were able to prove this. Blocking miR-22-3p using an antagomir resulted in mice developing a more severe form of epilepsy than normal (in the RCSI model, where the agent kainic acid (KA) is injected into the amygdala region of the brain). We also noticed that the brains of mice given the miR-22-3p inhibitor had an increased inflammatory response. If miR-22-3p worked to block brain inflammation, could we find the target(s)? Unfortunately, miR-22-3p has one of the highest numbers of predict targets of any miRNA and is made in every cell type in the brain. It would be impossible to guess which target of miR-22-3p was important - we needed to actually identify what miR-22-3p was binding to in this part of the brain. For this, we turned to our AGO pull-down technique and found one target of miR-22-3p attached to AGO with a known role in inflammation: P2X7 receptor. This protein is found on the outside of certain cells and is normally inactive. But, if enough ATP binds to it (which can happen if nearby cells are damaged), the receptor signals to release powerful inflammatory molecules.
If miR-22-3p normally blocks the P2X7 receptor, this could explain the increase in inflammation in the miR-22-3p-inhibited mice. Promisingly, we found higher levels of the P2X7 receptor in the mice given the miR-22-3p inhibitor. But this did not prove that the P2X7 receptor was the actual culprit. To prove the mechanism, we turned to a genetically-altered mouse that lacked the P2X7 receptor. When we injected the miR-22-3p inhibitor into these mice, we no longer saw the increase in seizures. We also showed that when the miRNA inhibitor was injected along with a drug that blocks the P2X7 receptor, we again didn’t see any change in seizures. So, the mechanism by which blocking miR-22-3p increased epilepsy had to be because of an increase in the P2X7 receptor. A final and unexpected discovery was that inhibiting miR-22-3-p regulated numbers of neurons in the hippocampus and their physical characteristics. We published three papers on these findings:
Jimenez-Mateos et al. microRNA targeting of the P2X7 purinoceptor opposes a contralateral epileptogenic focus in the hippocampus. Sci Rep 5:17486 (2015)
Engel et al. A calcium-sensitive feed-forward loop regulating the expression of the ATP-gated purinergic P2X7 receptor via specificity protein 1 and microRNA-22. Biochim Biophys Acta Mol Cell Res. 1864: 255-266 (2017). Engel et al. MicroRNA-22 controls aberrant neurogenesis and changes in neuronal morphology after status epilepticus. Front Mol Neurosci (accepted)
miR-324-5p regulates seizure initiation by suppressing a potassium channel
Potassium channels are critical for neurons to be able to signal to one another. By allowing movement of the positively charged potassium ion across the cell’s outer membrane, they help generate an electrical charge. This same movement also re-sets the electrical charge after a signal has been sent. Not surprisingly, several forms of epilepsy have been found to be caused by mutations in the genes that code for potassium channels. EpimiRNA partners, together with a team from Emory University led by Dr C Gross, uncovered a key role for a miRNA in setting the levels of one of these channels, called Kv4.2 in the brain. Together, we showed the channel is controlled by miR-324-5p. When the miRNA is removed the amount of the channel on the surface of brain cells increases. We found that during a seizure, miR-324-5p blocks enough Kv4.2 from getting to the surface. As a result, seizures are worse in mice in which miR-324-5p is inhibited. Blocking miR-324-5p had no effect in mice lacking the gene that coded for Kv4.2 (Kcnd2), solving the mechanism. Researchers are now exploring methods for delivering miR-324-5p as a potential approach to seizure suppression. The study was published in the fourth year of the project:
Gross et al. MicroRNA-mediated down-regulation of the potassium channel Kv4.2 contributes to seizure onset. Cell Rep 17, 37 – 45 (2017).
Seizure-suppressive effects of antagomirs targeting miR-134-5p are mediated via Limk1
The most potent anti-seizure effects we have observed for a miRNA inhibitor are those that block miR-134-5p. As we worked on answering key remaining questions about our antagomir targeting miR-134-5p (Ant-134), it was important to determine the mechanism of the seizure-suppressive effects. We already had a good idea of the probable target. Lim kinase (Limk1) is a key enzyme that controls the shape of contact points between neurons, where signals pass from one cell to another. The mouse and rodent Lim kinase have a predicted binding site for miR-134-5p and earlier studies in cell models had showed that when Limk1 is blocked we lost the protective effects of Ant-134. But was this how Ant-134 worked in the intact mouse brain? This was difficult to prove because we did not have a genetic model of Limk1 deficiency and there were no drugs selective for the enzyme. EpimiRNA therefore took a new approach. We designed a stretch of DNA-like molecules (Gapmers) that would stick perfectly to the Limk1 message. The normal reaction of a cell to the formation of such a duplex is to degrade the targeted message. We injected our Limk1-targeting molecule into mice along with Ant-134. When we did this we lost the anti-seizure effects of Ant-134. This proved the anti-seizure effects of Ant-134 depended on protecting Limk1. These findings are currently being prepared for publication.
The human Limk1 molecule does not contain such a good binding site for miR-134-5p. This could mean that other targets of miR-134-5p will be important in human tissue. We explored this idea by analysing the protein changes in the hippocampus of mice given Ant-134. This revealed a select number of proteins change after Ant-134. These could also be important to the mechanism of Ant-134 or relevant to whether or not the therapy will translate to humans.
miR-129-5p is a novel target for seizure control via a calcium pump and RNA binding protein
So far, it was mostly single gene targets that underlay the mechanism by which many of our miRNA inhibitors produced seizure-suppressive effects. This has changed thinking in the field about how miRNAs work, at least in the brain. However, this turned out not to be the case for the final miRNA for which EpimiRNA dedicated major efforts to elucidate the molecular mechanism. We had found that miR-129-5p was one of the most consistently upregulated miRNAs in our animal models. It was a neuron-expressed miRNA and, like miR-134-5p, was important for shaping the physical structure of neurons. No other team had yet looked at this miRNA. We could show that blocking miR-129-5p in mice strongly reduced seizures in the mouse IAKA model - we had a new epilepsy miRNA. The UNIMAR team led efforts to identify the targets of miR-129-5p. They found several. The first was a calcium pump, encoded by the Atp2b4 gene. This functions to extrude calcium from cells, a critical job since too much calcium can interfere with metabolism or cause cell death. Since miR-129-5p is increased in epileptic brain tissue (and Atp2b4 is down) this could be compromising how neurons handle calcium. Another target was an RNA binding protein called Rbfox. This class of proteins is critical for the distribution of RNA around a cell. While there is growing evidence that disturbances to RNA binding proteins cause neurodegenerative diseases, no-one had looked in epilepsy. The EpimiRNA project helped elucidate the mechanism by which this miRNA controls excitability and found two exciting new targets for seizure control. The findings were published in the fourth year of the project:
Rajman et al. A microRNA-129-5p/Rbfox crosstalk coordinates homeostatic downscaling of excitatory synapses. EMBO Journal 36:1770-1787 (2017)
Other miRNAs and their potential targets-mechanisms
The EpimiRNA project has made significant progress in understanding the mechanisms of at least four other miRNAs. This includes miR-135a, which was identified as dysregulated in human epileptic brain by the UMCU team. Within the EpimiRNA project, we conducted experiments that showed antagomirs targeting this miRNA can suppress epileptic seizures. The UMCU team has identified two important targets. One is a transcription factor (TF), which controls the activity of multiple genes, allowing this miRNA to shape an extensive gene expression landscape. The UMCU team also showed the miRNA functioned in the control of axon guidance, which may be altered following epileptic injury. Findings have been published and others are in preparation:
van Battum w et al. An image-based miRNA screen identifies miRNA-135s as regulators of CNS axon growth and regeneration by targeting Krüppel-like factor 4. J Neurosci 38:613-630 (2018).
Protein analyses provide evidence of strong influence of miRNAs during epilepsy development
By mid-way through the EpimiRNA project, we knew what happened to levels of every miRNA in the different models but it was uncertain whether many of the novel miRNAs were actually important. We chose to look at protein changes within the brain. Ultimately, if a miRNA is functional it should lower levels of proteins. We undertook an ambitious plan to look at all protein changes over the time-course of epilepsy, selecting the rat model from UNIMAR for this special analysis because of the larger size of the brain. The SDU team led the work to profile the protein changes using mass spectrometry technology. A total of 24 samples were processed, 9149 peptides were identified and 580 protein regulations were determined. Each time point had a unique set of protein changes and the amounts of protein that changed differed during the course of epilepsy development. The earliest time point showed only few changes, increasing thereafter. Remarkably, one of the time points showing the most changes was the day after the first epileptic seizure occurred.
To compare the proteomics data with the miRNA profiling data we also did a gene set enrichment analysis (GSEA). This will indicate if there is a significant enrichment of a group of proteins being targets of a specific miRNA among highly changing proteins. Essentially, you are taking a set of altered proteins and asking what miRNAs could explain what you are seeing. You then look if those are miRNA regulated in the model. The result was remarkable, identifying multiple key miRNAs we knew were implicated in epilepsy plus a set of miRNAs that have not yet been interrogated (not disclosed). The data are strong support for an important influence of miRNAs on the proteins in the hippocampus.
Main conclusions and future objectives Collectively, the EpimiRNA project solved the molecular mechanism by which multiple miRNAs control seizures. We also defined all protein changes during epilepsy development and show a strong predicted influence of miRNAs in shaping the protein landscape in the epileptic brain. The discoveries have re-shaped our thinking about the targeting preferences of these molecules. They have led us to new categories of gene that may be involved in epilepsy such as RNA binding proteins. Finally, we may be able to design drugs to the identified miRNA targets as anti-seizure or disease-modifying therapies of the future.
1.3.4 Objective (3) Identification of miRNA biomarkers of epileptogenesis and seizures
Introduction
A major focus and achievement of the EpimiRNA project was to identify miRNAs in biofluids that may be biomarkers of the epileptogenic process or a means to diagnose epilepsy. The project approached this challenge by analyzing miRNAs in biofluids from the three animal models of epilepsy and from human volunteers and patients. We collected blood samples over multiple time points from the animal models and from patients before and after they had a seizure. We also collected cerebrospinal fluid (CSF) samples which, being in direct contact with brain tissue, may be a rich source of biomarkers.
Choosing a miRNA screening technology
Our main hypothesis was that brain-specific miRNAs would be found in blood samples in the epilepsy models. Nevertheless, we approached the identification of miRNAs with an open mind and opted to initially screen for all known miRNAs before closing in on one or two that had the best biomarker potential. The main technology in the field for screening miRNAs at the start of the EpimiRNA project used a method that began by extracting small RNAs from a sample and mixed with reagents that include special “templates” that detect and amplify the miRNA resulting in a semi-quantitative measure of the amount of the miRNA. The EpimiRNA project required a screening platform that would work for the type of ultra-low amounts of RNA and small volumes we would have from the rodents and CSF. We selected a platform that was designed precisely for ultra-low abundance samples; OpenArray (OA). The OA technology provided us genome-wide coverage of miRNAs on an etched glass microchip-like plate into which the sample is added by robotic pipette. The semi-automated workflow then delivers the chips to a specialized PCR machine (12K Flex) for reading. An added advantage of the OA technology was speed, enabling us to screen multiple samples within a few hours.
Optimising and trouble-shooting the miRNA screening technology and other methodology
One of the first decisions we had to make was whether to use serum or plasma. Serum had the advantage that it required minimal processing (a blood sample is left to clot). However, there were concerns that the clotting process would alter miRNA profiles and introduce variability. We therefore opted for plasma and developed a standard operating procedure (SOP) for partners which we later made available to other researchers.
During the first year of the project, while the teams at RCSI, UNIMAR and UNIVR were generating the brain and blood samples from the animal models, the team at RCSI who would be responsible for biofluid work began tests of the OA platform. We first wanted to be certain of the platform’s reproducibility. We took a sample of plasma and split it, then ran it on two different OA chips, with a gap of three months in between. The OA results from the two chips were almost identical, confirming high reproducibility of the platform.
We also investigated whether the implanted telemetry unit, which we would be using to record brain activity and count seizures in mice, affected miRNA profiles in blood. This required a minor surgical procedure on the mice to place the device under the skin. Our tests showed only 1 miRNA out of all those screened differed in plasma from implanted mice. Thus, the OA platform was reproducible and our monitoring method would not bias results.
EpimiRNA’s rodent blood sampling technique
Another technical challenge for the project was how to obtain repeated samples of blood from mice. While the tail vein of the rat is large enough for repeat blood sampling, the mouse tail is much smaller and repeated sampling is not recommended. We adapted a technique in which blood is taken from a vein that runs down the cheek of the mouse (submandibular vein). A tiny needle is inserted into the cheek of the anesthetized mouse which results in a few drops of blood being released which can then be collected. We found this to be highly reproducible and a training video was generated so that all teams performed the procedure the same way. We now had the ideal plasma technique for the mouse OA studies.
Final tests of the OA platform
Our final tests were to investigate how many miRNAs were detectable in normal plasma and check if these were similar to what had been reported before. We ran a set of mouse plasma samples and could detect up to ~200 miRNAs in a sample. This was at least as good as other platforms and within the detected miRNAs we found plenty that were known to be abundant in plasma such as miR-16, miR-19b and miR-223. Equally important, we found very low or undetectable levels of the known brain-enriched miRNAs. When we ran human plasma samples we obtained very similar results. That is, we could often detect ~200 miRNAs in a given sample and the most abundant miRNAs were those that others had reported previously. For the actual experiments we applied strict and consistent criteria for what miRNAs were called present in a sample. This reduced the numbers of miRNAs called present to ~100 per plasma sample.
Plasma miRNA profile in healthy human volunteers
With the collection of human samples from epilepsy patients growing during the first year, we took the opportunity to understand the profile of miRNAs in healthy control plasma. We looked at several key variables: (1) morning versus afternoon sampling, (2) day-to-day differences and (3) male versus female miRNA profiles. We found no differentially expressed miRNAs between the morning or afternoon samples, no significant changes in miRNA levels from the same subject over a period of a month and no sex differences.
We concluded that, overall, plasma miRNA levels were very stable and that our platform compared well to others while having the advantage of speed and low input requirements. The results of these studies were published:
Mooney et al. High Throughput qPCR expression profiling of circulating microRNAs reveals minimal sex and sample timing-related variation in plasma of healthy volunteers. PLoS ONE 10:e0145316 (2015).
Identification of miRNAs in plasma in experimental epileptogenesis
By the third year of the project, the three different teams leading the animal model work had generated enough complete sets of blood samples that we could begin to profile these using the OA platform. We ran a time-series from each model, including a baseline sample (pre-insult), a sample taken after the animals had been subjected to status epilepticus but had not yet developed spontaneous recurrent seizures, and then finally a sample taken during the phase where active epilepsy was established.
We looked first at the baseline samples taken from the rodents prior to experimental manipulation. We detected an average of 103 miRNAs in the mouse IAKA model, 93 in the mouse PILO model and 88 in the rat PPS model. Focusing on the top 20 most abundant miRNAs, we found that 13 were common to all three models. We also found strong overlap between the most abundant miRNAs in rodent plasma and the most abundant in human plasma. Specifically, 7 of the 12 most abundant miRNAs in human plasma were found in the rodents (miR-19b-3p, miR-24-3p, miR-92a-3p, miR-146a-5p, miR-20a-5p, miR-223-3p and miR-191-5p). This suggests discoveries in the rodent models will translate to humans.
When we looked at all miRNAs in the baseline plasma samples from the three animal models we found extensive overlap, with 66 of the same miRNAs detected in all three models. This means that the circulating pool of miRNAs is very similar between the two rodent species. When we looked at the miRNAs present in samples from epileptic animals, we found 74 miRNAs were common between the models. This means that many of the same plasma miRNAs are affected the same way regardless of species or specific model. But we also found a small number of miRNAs that were unique to each model.
Comparing the lists – those found in all models at baseline and those in samples from epileptic animals - we identified 16 miRNAs unique to the epilepsy samples. When we checked databases on tissue-specific miRNAs we found many of these were brain-restricted. This strongly supported our hypothesis that brain-enriched miRNAs are released into the circulation in epilepsy and thus have biomarker potential. When we applied statistical tests to the lists of miRNAs from each model we identified several miRNAs that were altered at different stages of epilepsy in the different models.
We next began validating the OA-identified miRNAs. Since the amount of plasma from the animal models was limited we had to pick just a few miRNAs to focus on. We selected two miRNAs that were different in epilepsy versus baseline samples and three from among those that changed within a given model at a specific time point. We now switched from the OA platform to test each miRNA using individual miRNA assay kits. Two of the five did not show the expected change in at least two of the models whereas expression of three others were all validated as potential epilepsy biomarkers.
To extend this validation, we were interested in whether plasma levels of the miRNAs were sensitive to any common anti-epileptic drug (AED) therapies. This is important to know, since a miRNA that is strongly affected by medication maybe unreliable for diagnosis. For these experiments, mice were dosed daily with one of two common AEDs and plasma levels of the miRNAs were measured afterwards. We found no significant effect of AEDs on the tested miRNAs (Figure 14) suggesting they are stable biomarkers.
We were also interested in whether treating mice with a disease-modifying therapy would “normalise” (restore) the altered miRNA levels in plasma. If the miRNAs are biomarkers of the disease their levels should be corrected by a disease-modifying therapy. For these experiments, mice were injected with the miRNA inhibitor that we had in pre-clinical development (antagomir-134; Ant-134). This reduces spontaneous seizures by ~90% in the RCSI mouse model. Plasma levels of the validated miRNAs changed in mice given Ant-134 indicating the miRNA biomarkers respond to an anti-epileptogenic therapy.
Our final tests of the rodent-identified miRNAs was to look for them in plasma from TLE patients in adults. We had collected a number of plasma samples from epilepsy patients and we used a sub-set of these. Our results showed plasma levels of three of our miRNAs were altered in patient plasma, either at baseline or after seizure. Levels of another also showed a trend to lower levels, in line with the animal data, in samples after seizure. We used a statistical technique called receiver operating characteristics (ROC). to assess the ability of the miRNAs to distinguish between controls and patients with TLE. As an example, we found that post-seizure sample levels of one of our miRNAs could identify TLE patients with 86% confidence. This is a high-quality result for a single molecule biomarker.
Clinical studies to identify miRNAs biomarkers of TLE
In parallel with the animal model studies, we began our search for the best plasma miRNAs to diagnose human TLE. The main effort focused on sampling patients within specialist epilepsy monitoring units (EMU). Patients admitted to EMUs typically have complex, drug-resistant epilepsy and may be later offered neurosurgery to remove epileptic brain tissue. After admission, patients are continuously monitored by video and EEG to record epileptic events. After ethical approval was received, we began recruiting patients from the EMUs at epilepsy centres in two countries - Germany and Ireland. The standard protocol was tocollect a baseline sample after admission when the patient had been seizure-free for a few days. The research team would then collect one or more samples at different time points after a seizure occurred. We selected an early time point (1 h) and a later time point (24 h). As we had done with the rodent studies, plasma preparation protocols were harmonized between sites.
Our target was to recruit 60 patients for the profiling study. We reached this number quickly and by mid-way through the project had collected samples from over 150 epilepsy patients. We also started to collect samples from patients with other forms of epilepsy. This would allow us to determine whether a particular miRNA was unique to TLE patients or whether it was a broader biomarker of all epilepsies. We also collected samples from patients who had psychogenic non-epileptic seizures (PNES). These patients are particularly difficult to diagnose outside specialist centres. If the miRNAs we identified in the TLE patients were true biomarkers of epilepsy, their levels should not be altered in a patient with PNES since the seizure events are non-epileptic in origin (typically a manifestation of an underlying psychiatric condition). However, if the levels of the miRNA were similar between TLE and PNES patients then they might instead be reflecting medication or other factors (a PNES patient may be prescribed an AED on the basis that they are thought to have epilepsy).
During the third year of the project we profiled miRNAs in individual samples from epilepsy patients at baseline and after a seizure. In total, 96 plasma samples from the two centres were profiled by the OA platform. A set of miRNAs were identified that were differentially expressed in epilepsy samples compared to controls. Before moving to validate these, however, we decided to add an extra profiling study. The reason for this was that by this point the technology to directly sequence small RNAs from low-abundance sources such as plasma had become more widely available, costs were dropping and a dual approach would give EpimiRNA an advantage over what had come before. “RNA-seq” generates complete and quantitative analysis of all miRNAs and can identify novel miRNAs. We therefore ran RNA-seq, pooling the plasma samples from each centre to reduce costs and compared control to pre- and post-seizure. We found nearly 100 more miRNAs in the RNA-seq data compared to the OA platform. Results were also highly reproducible between Centres. Importantly, we found 11 additional miRNAs that were up- or down-regulated in epilepsy and these were then advanced with the OA-identified miRNAs for validation.
We performed an extensive set of validation experiments. First, we confirmed the profiling results for 15 miRNAs using the original samples, running individual miRNA assays. From these studies we became most interested in three miRNAs. The level of miR-27a-3p and miR-328-3p were consistently changed in samples collected after a seizure. Levels of miR-654-3p were different from controls in both baseline and after-seizure samples. We then tested these in a larger cohort comprising over 100 TLE patients and 100 controls. We also measured their level in plasma from patients with three other forms of seizure disorder and patients with PNES. This confirmed that levels of miR-27a-3p and miR-328-3p were different in TLE patient plasma but revealed they were also altered in patients with genetic generalised epilepsy (GGE). Importantly, the miRNAs were not significantly different to controls in samples from PNES patients. This means that measuring the level of these miRNAs could help diagnose epilepsy and help identify possible PNES cases. Statistical analyses suggested the miRNAs could correctly predict epilepsy with ~70% accuracy. While encouraging, we sought ways to increase the diagnostic accuracy of these biomarkers.
We tested for levels of the three miRNAs in plasma samples from one of our rodent models. Testing in rodent blood means we can exclude certain likely confounders that are impossible to avoid in human studies such as medication. We found that levels of all three miRNAs followed the same pattern in mice as we had seen in the patient samples. We also examined what happened to levels of the three miRNAs in mice given the disease-modifying Ant-134 therapy. Plasma levels of the three miRNAs were all corrected – normalised – in mice given the disease-modifying Ant-134 treatment.
Another technological advance that had appeared since the start of the project was digital PCR (dPCR). This allowed a researcher to count the individual copies of an RNA molecule in a sample. Thus, you could estimate how many copies of the three miRNAs would be present in a drop of blood. We put this technology to the test and ran dPCR analyses on our three miRNAs. The approach worked, and we learned that miR-27a-3p was the most abundant of the three in plasma, with ~6000 copies per sample of plasma in a healthy person. As expected, levels were significantly different in both TLE patients and patients with GGE, with readings of ~2000 copies in an equivalent volume in an epilepsy patient. Interestingly, the count in a PNES patient was ~5000, close to normal.
Making a prototype blood test for epilepsy
Project partners began thinking about how our findings on plasma miRNA could be developed into something resembling a clinical test. One of the partners (RCSI) started working with colleagues on a way to directly measure RNA without the need for template-directed amplification. If this were possible, it would dramatically simplify testing and reduce the sample-to-answer time. A prototype device was eventually developed that comprised a “capture” oligonucleotide that would trigger an electrochemical reaction proportional to the amount of miRNA. Because we had not yet finished screening miRNAs using the OA, we designed the first prototype to detect miR-134-5p. This was one of the miRNAs our project was most focused on because of its therapeutic potential. However, since miR-134-5p was also brain-enriched it could also be a potential biomarker. We tested some of the samples collected in the video-EEG unit using the new device. It gave almost identical results to those when measured by standard PCR but in a fraction of the time. These results were published in 2015: Spain et al. Direct, non-amplified detection of microRNA-134 in plasma from epilepsy patients. RSC Advances 5, 90071-90078 (2015).
We later adapted the device so it could detect the three novel miRNAs. We showed that we could measure all three miRNAs simultaneously with the device, termed “TORNADO” (Theranostic one-step RNA detector). This system is in commercial development.
Superior diagnostic accuracy by analysis of how miRNAs are transported in blood
Despite this progress, we felt we had to improve on the accuracy of our miRNA biomarkers. The breakthrough came from wanting to understand how miRNAs reach the circulation from the brain. There was evidence at the time that miRNAs were released from cells in tiny vesicles called exosomes. Other studies suggested that most circulating miRNA was bound to the AGO protein. Neither mechanism of transport had been explored in epilepsy. We therefore took a set of our plasma samples and split them and then either extracted the AGO protein or prepared exosomes using a special centrifuge. What we found was remarkable. First, there was generally more miRNA within exosomes than in the AGO-bound fraction. More importantly, when we ran the statistical analysis of diagnostic accuracy it noticeably improved our ability to tell patients from controls. This time the miRNAs gave results in the mid-80% range. Using a further statistical method called logistic regression and considering all three miRNAs combined gave a result of 93% confidence of diagnosis. This means that if we first separate this portion from the rest of the plasma we can get a more accurate diagnosis. We are now in discussion to develop a miRNA-based test with industry partners.
Raoof, Bauer, El Naggar et al. Dual-center, dual-platform microRNA profiling identifies potential plasma biomarkers of adult temporal lobe epilepsy. eBiomedicine (2018).
EpimiRNA performs the first analysis of miRNAs in CSF from epilepsy patients
In parallel with the plasma studies, partners in EpimiRNA collaborated to study miRNAs in CSF. As mentioned, we were particularly interested in CSF since it might be an even better source of epilepsy biomarker. Also, we were aware that no other group in the world had studied miRNAs in CSF samples from patients with epilepsy. However, our collection of CSF samples from the original partner hospitals was not sufficient. The clinical co-coordinator of EpimiRNA, Prof Felix Rosenow, reached out to a colleague who specialised in CSF analyses (Dr Peter Körtvelyessy, Magdeburg). We now had enough CSF samples from TLE patients plus samples from patients after status epilepticus. This is an interesting patient group because they have experienced prolonged seizures but may not necessarily have epilepsy. Another group of samples were collected from patients with other neurological diseases for the purposes of comparison. This included patients with Alzheimer’s disease (AD) and Multiple Sclerosis (MS). Last, we had a set of controls. In total we obtained over 100 samples of CSF from different individuals.
We began by using the OA platform to screen 45 samples from controls, TLE patients and SE patients. It was immediately obvious that miRNAs were much less abundant in CSF than in plasma. In fact, we detected only around half as many miRNAs in CSF as we had in plasma. We detected a number of significantly regulated miRNAs. This included five that were different in TLE patients compared to controls (one down and four up regulated) and 15 miRNAs in samples from patients with status epilepticus compared to controls (seven down and eight up regulated). Thus, despite detecting far fewer miRNAs in CSF the numbers of differentially expressed miRNAs remain significant. This is consistent with our hypothesis that CSF may be a rich source of potential biomarkers.
As with our plasma study, we selected a set of three miRNAs to focus on during validation. These were miR-19b-3p, a potential biomarker of TLE and miR-21-5p and miR-451, which were different in patients with status epilepticus. Levels of these were not different to controls in the patients who had other neurologic diseases meaning they were specific for epilepsy and seizure-related disorders. Again, we used dPCR to get an idea of the copy number of these miRNAs in CSF. By far the most abundant was miR-451, with an average of ~200 copies per sample in status epilepticus compared to under 20 for TLE patients and other neurological diseases. Diagnostic accuracy of over 80% was achieved for the miRNAs.
Our final experiments using the CSF were to look at the miRNAs within the exosome and AGO-bound fractions. What we found was that for each neurological disease the proportion of the miRNA in one form or another was quite different.
As an example, miR-19b-3p was most abundant in the AGO-fraction in TLE patients but was more highly expressed in the exosome fraction in MS patients. These findings support the development of a diagnostic test for epilepsy based on CSF miRNAs. Indeed, we tested TORNADO and confirmed that the device could determine the concentration of a miRNA in CSF. The device estimated the concentration of miR-134-5p in CSF at between 100 nM and 10 mM in patients with TLE or status epilepticus. This level is higher than in plasma, suggesting CSF is a unique source of miRNA biomarkers for diagnosis. The findings have been published:
Raoof et al. Cerebrospinal fluid microRNAs are potential biomarkers of temporal lobe epilepsy and status epilepticus. Sci Rep 7, 3328 (2017)
McArdle et al. “TORNADO” – Theranostic OneStep RNA Detector; microfluidic disc for the direct detection of microRNA-134 in plasma and cerebrospinal fluid. Sci Rep 7, 1750 (2017)
Biomarker profiles of human brain stimulation
Our final biomarker project concerned looking for evidence of miRNA changes in biofluids in response to brain stimulation. Our partner Cerbomed had a transcutaneous VNS stimulator (tVNS) in clinical testing for drug-resistant epilepsy, led by Prof Felix Rosenow. Does tVNS change miRNA profiles? We collected blood from volunteers before and after tVNS to determine if the stimulation altered biofluid levels of miRNAs. We focused on a sub-set of miRNA that had been shown in the earlier projects to be biomarkers of seizures or epilepsy. We also included an analysis of cytokines because both VNS therapy and epilepsy are associated with changes to inflammation. We ran samples from subjects with tVNS stimulation as well as sham-stimulated controls. The main finding was that none of the miRNAs were altered within the plasma samples of the healthy volunteers given active tVNS exposure. However, one miRNA (not disclosed) was found to be significantly upregulated in drug-refractory epilepsy patients given active tVNS. This suggests that plasma miRNA levels can be used as a biomarker of tVNS therapy in drug-resistant epilepsy. Of further interest, two of the cytokines changed in tVNS patients. The clinical use and biomarker potential of tVNS for patients with epilepsy remains uncertain, however. The device may be most useful in predicting who may benefit from implanted VNS.
We published a review on the state-of-the-art of miRNAs as biomarkers in collaboration with the EpiTarget consortium: Enright et al. Discovery and validation of blood microRNAs as molecular biomarkers of epilepsy – ways to close current knowledge gaps. Epilepsia Open (in press)
Main conclusions and future directions: EpimiRNA has generated important insights into the potential use of miRNAs as diagnostic biomarkers. There is good evidence that both plasma and CSF contain unique levels of several miRNAs that are not found in other neurological conditions. The levels of these can be absolutely measured using either dPCR or TORNADO. We cross-validated our animal findings in human samples and human findings with animal model results. We showed for the first time that a miRNA-based disease-modifying therapy altered plasma levels of miRNA biomarkers. We identified a brain-enriched miRNA that is altered in plasma of TLE patients given tVNS therapy. These discoveries could prompt the development of a multi-miRNA biomarker panel for diagnosis of epileptogenesis or epilepsy or to support brain stimulation device use.
1.3.5 Objective (4) Novel therapeutics by targeting miRNAs
Introduction
A major objective achieved by the EpimiRNA project was to discover new molecules that could suppress seizures or have disease-modifying effects and move these towards preclinical development. Foremost, we focused on inhibitors of miRNAs, termed antagomirs. Over the course of the project, we tested more than 10 different miRNA inhibitors in animal models. We split our efforts three ways. We tested inhibitors against a set of miRNAs which EpimiRNA partners had already become interested in due to links to epilepsy. Second, we screened inhibitors targeting the entire set of miRNAs we had discovered to be common-to-all-models. Finally, we made efforts to answer important questions around clinical translation of a miRNA inhibitor. Did our inhibitors work in species besides mice? How important is the structure of the antagomir? How would a patient even take such as drug?
Over the course of the project we identified 8 miRNA inhibitors that suppressed seizures triggered by chemicals or brain stimulation. Two more prevented spontaneous recurrent seizures (ie. epileptic seizures). We answered critical questions on the path to preclinical development. Partners at RCSI secured a US patent for one of the miRNA inhibitors. The EpimiRNA project also discovered an anti-seizure compound from Bicoll and by the end of the project had this accepted for analysis by the Epilepsy Therapy Screening Program run by the National Institutes of Health in the USA.
Effects of miRNA inhibitors targeting five epilepsy-associated miRNAs
Ant-22 We had become interested in miR-22-3p as one of the most upregulated miRNAs in protected brain regions in the mouse model at RCSI. Blocking miR-22-3p using antagomirs caused spontaneous seizure rates to increase in mice, indicating the miRNA had anti-seizure properties. We realised that if we could increase the amount of the miRNA in the brain it could reduce seizures. We showed that injecting miR-22-3p itself into the mouse brain could reduce epileptic seizures. However, the dose-range at which this worked was extremely narrow. We think that while inhibiting a miRNA is largely safe while introducing more of a miRNA could impact on cell functions in difficult to predict ways and trigger an inflammatory response. From this point onward, EpimiRNA focused on inhibiting up-regulated miRNAs. These data were published (Jimenez-Mateos et al. Sci Rep 2015)
Ant-129 miR-129-5p was among a set discovered by the UNIMAR team to control the strength of signalling at synapses and was also up-regulated in our animal models. The RCSI team tested antagomirs against the neuron-enriched miR-129-5p in their mouse model. Mice given the antagomir were profoundly protected against seizures induced by
the chemical kainic acid. Recordings of brain activity showed much-reduced seizure severity and there were fewer signs of brain injury from the seizures in those mice. These data
were published (Rajman et al. EMBO Journal 2017).
Ant-324 miR-324-5p was discovered as targeting the Kv4.2 potassium channel. The RCSI team tested antagomirs against miR-324-5p (Ant-324), again using their kainic acid model.
Seizure onset was delayed in mice given the antagomir and, again, there were few signs of brain injury from the seizures in those mice. These data were published (Gross et al. Cell
Rep 2016).
Ant-135a miR-135a-5p was originally found by the UMCU team to be increased in the hippocampus of patients with TLE. We checked brain sections from our mouse model and found it was strongly increased in neurons. We developed a technique to deliver an antagomir into the brain of an already-epileptic mouse. We implanted a device that recorded seizures and once we had a stable baseline we injected the inhibitor. Within a day we started to see a reduction in epileptic seizures while the control group continued to have seizures.
We waited for two weeks and then studied the results. The Ant-135a inhibitor significantly reduced epileptic seizures by over 60% and seizures were shorter in duration.
Ant-134 At the time of starting the project, we already knew that injecting an inhibitor of miR-134-5p (Ant-134) before or immediately after kainic acid in mice had a dramatic effect on seizures. Within EpimiRNA, we embarked on a comprehensive study to address important remaining questions that might otherwise be a barrier to preclinical development. First, the RCSI team showed that Ant-134 could reduce seizures in the PTZ test, a common test in mice for identifying anticonvulsive drugs. The UNIMAR team showed that injecting Ant-134 after triggering epilepsy using electrical stimulation of the brain dramatically reduced epilepsy in rats. The UCL team showed that Ant-134 reduced epileptic-like activity in brain slices. These results were published in the fourth year of the project: Reschke et al. Potent anti-seizure effects of locked nucleic acid antagomirs targeting miR-134 in multiple mouse and rat models of epilepsy. Mol Ther Nucleic Acids 6:45-56 (2017)
Another question was how important was the exact chemical composition of the Ant-134? Would small differences in the chemistry affect potency? We tested this by ordering Ant-134 from a different company and testing alongside our current batch. Both produced nearly identical seizure suppression. We then turned to safety. Since the target, miR-134-5p controlled the shape of contact points between neurons it was possible that we could be interfering with normal brain function. The UCL team led studies to record neuronal behaviour after injection of Ant-134. All physiological properties were found to be normal, while the molecule nevertheless blocked epileptiform events. Findings were published: Morris et al. Spared CA1 pyramidal neuron function and hippocampal performance following antisense knockdown of microRNA-134. Epilepsia 59:1518-1526 (2018)
Overcoming the delivery challenge
We next thought about delivery. Brain tissue is separated from the contents of the blood by a specialised barrier (blood-brain-barrier). This blocks large molecules from passing through. Ant-134 was too big to pass if injected into the blood. Our first idea was to inject into the spinal fluid using a route similar to how CSF is obtained. We found that Ant-134 can reach deep into the brain when given like this in mice and blocked the target miRNA. Since only a single injection of Ant-134 is needed to stop seizures, this could be acceptable to a patient with drug-resistant TLE. But we had a second idea. What if the blood-brain barrier is open in epilepsy? Then a systemic injection might be enough to get it into the brain. We studied the opening of the blood-brain barrier in the model at RCSI, finding it opened profoundly (Figure 23). With this knowledge, we triggered status epilepticus in mice and gave them a single injection of Ant-134 into the belly of the mouse. We tracked the mice for the next three months to be sure any effects were permanent. Epileptic seizure rates were dramatically reduced by Ant-134 indicating a disease-modifying effect. At the final time point of assessment (3 months), Ant-134-injected mice had 99% less seizures than the control group.
Ant-134 can prevent seizures in already-epileptic mice
The final study with Ant-134 was to test whether it could reduce seizures when injected into already-epileptic mice. This was one of the most clinically realistic scenarios, where a patient has recently presented with epilepsy and you could give them a disease-modifying therapy. We triggered epilepsy in mice and waited two weeks. We then injected Ant-134 or a control molecule and tracked the seizures in the mice. Within two days seizure rates had fallen by over 80% in the mice and stayed this way for two weeks. We continued recordings a month later. During these recordings, now many weeks after the single injection of Ant-134, mice had 90% fewer seizures than the controls. There was no sign of seizure rates recovering. Taken together, the work on Ant-134 suggests this is one of the most potent disease-modifying therapies identified to date. Having secured a patent for this invention we subsequently began collaborations with industry partners toward the preclinical development of this new therapy for epilepsy.
Anti-seizure and neuroprotective effects of novel miRNAs discovered by EpimiRNA
By mid-way through the project we had a list of 8 miRNAs that were upregulated in all three epileptic mice, six of which had not yet been targeted in an animal model. We therefore embarked on a systematic effort to test antagomirs against all six miRNAs. For this we needed a high-throughput model and turned again to the mouse model at RCSI. Experiments were run in batches (a so-called “block design”) with six animals a day each given a different antagomir. Two other mice received either saline or a scrambled version of the antagomir. This was repeated until we had ten sets of experiments. Our results showed three of the antagomirs produced powerful anti-seizure effects in the model. Pilot tests of a “combination” of the three antagomirs was not superior to one alone. Analysis of brain injury revealed five of the six antagomirs protected the hippocampus from damage.As we had done for Ant-134, we were also keen to test the safety of the antagomirs on normal neuronal function. The UCL team again led studies to record neuronal behaviour after injection of the three novel antagomirs. All physiological properties were found to be normal.
Non-miRNA-In addition to the miRNA focus, the EpimiRNA project included an SME, Bicoll that had expertise in drug discovery. During the project they rationally designed a small molecule compound for seizure control. During the fourth and fifth year of the project, teams at UCL and RCSI were able to test the molecule, beginning with in vitro tests in brain slices and moving later to an in vivo mouse model. The compound demonstrated strong anti-epileptic effects in both. This included suppressing epileptic activity in human brain tissue and reducing seizures in the PTZ seizure model in mice. The compound was accepted into pre-clinical screening by the NIH’s epilepsy therapy screening program.
Main conclusions and future objectives
Collectively, the EpimiRNA project identified 10 new molecules with anti-seizure effects, advanced one into preclinical development, obtained a patent, moved another anti-seizure
compound into the USA’s epilepsy therapy screening programme and developed new industry partnerships.
1.3.6 Objective (5) Investigation of genetic variation in miRNAs.
While TLE is largely considered an acquired form of epilepsy there is evidence that there is a genetic contribution. That is, that variation or mutation in certain genes may explain a portion of the TLE population. EpimiRNA undertook the first ever investigation to determine whether genetic variation in miRNAs underlies human TLE. The main focus was on the coding sequences of the miRNAs (over 1500) but we designed the study to also look at the over 40 human miRNA biogenesis pathway genes and we looked for variation in the 3’ UTR regions of all known genes expressed in the hippocampus (~90,000 sequences). Variant or mutationof a miRNA binding site in a target transcript is potentially as important as a mutation in the miRNAs themselves, because the introduced sequence change may reduce (or increase) silencing and alter gene expression, leading to a functional change in a cell.
We aimed to sequence 850 TLE patients with matched number of controls and we were also able to include other epilepsy syndromes. However, this exceeded the total number of DNA samples available within the project teams so we collaborated with colleagues at other European and non-European sites and were able to reach (and exceed) our targets. We
eventually sequenced over 3000 patients and controls from the different partner sites. The sequencing workload was spread between the CU team in the USA and the UNICAMP team in Brazil. Sequencing began in the first year of the project and was finished by the fourth year at both sites. The teams then began the process of carefully checking the quality of the sequencing and the analysis of the data. The teams leading the genetics analysis used different approaches to the analysis. However, we did not identify any single miRNA genes that passed the significance thresholds set in our main analysis. Nor did we identify miRNA biogenesis genes or the targets. This would indicate that, within the parts of the genome sequenced and with the sample size we had used, there are no sites that when burdened with genetic variation confer strong enough epilepsy risk. The results were similar when we combined the results of the European-decent subjects with those of Brazilian decent - no single gene passed significance.
1.3.7 Objective (6) Technological advances and resources.
The EpimiRNA project has delivered a number of important technological advances and overseen training of these. The project has also generated a new and unique biorepository of samples from across the various partners that can support research long into the future.
Technical advance - iCLIP analysis of human epileptic brain. A key objective of the project was to identify with certainty the targets of all known miRNAs in the human brain. The technique for this – iCLIP – depends on being able to selectively pull-down the AGO protein from a sample while maintaining the association between the miRNA and the target. If the procedure is not performed correctly the miRNA or target may be missed. The AU team perfected an iCLIP protocol for human post-surgery tissue. This protocol, once published, will be of value to any researchers in the neuroscience community. The data from the study will be available to the research community
Data base resource iCLIP database of all miRNA-target interactions in the human epileptic brain
The iCLIP study has generated a rich dataset that can we searched either by miRNA or by target mRNA. This database will be made freely available upon publication of a companion research article.
Technical advance Systems modelling frameworks for miRNA research
We established an unbiased, systems-level pipeline to identify and rank miRNA-mRNA target interactions (MTIs) from a curated in-house database. Using a common programming language (R), the pipeline compares miRNA sequences across species, collates miRNA-target interactions from multiple databases, identifies transcription factors among targets, and facilitates prioritisation/filtering of MTIs via a bespoke scoring system (‘MTI score’).
The filtered miRNA-target information can be used to generate network interaction graphs and perform enrichment/pathway and similar analyses and has been used in multiple EpimiRNA publications. We also developed an integrative model which enabled us to deduce the dynamics between miRNA and mRNAs based on their expression profile using Bayesian methods. The network was generated using 190 miRNAs and 198 mRNA containing 2100 edges between miRNAs-mRNAs and mRNAs-mRNAs (Different colour identifies different modules. Text size of each node depicts its influence in connectivity of the network).
Resource - EpimiRBase
EpimiRBase was released in 2015 and is a searchable database of all known epilepsy studies on miRNA. It allows any researcher to look up whether a particular miRNA has been identified, the species, model and whether any functional studies were undertaken. It has already been queried more than 1000 times and is available at www.epimirbase.eu
Resource - EpimiRNA biorepository
Over the course of the EpimiRNA project, blood samples were obtained from over 300 patients and healthy controls. These were processed to plasma and any remaining sample will be retained for future biomarker research on miRNAs. We also collected nearly 100 human CSF samples and over 100 human brain resections from drug-refractory epilepsy cases. From our animal model work we have also collected 100s of blood samples, also processed to plasma, and brain tissue samples. These will be used by partners within the project in the immediate future but will later be made available to other teams, where ethical approval and GDPR allow, to validate their own findings or to develop new hypotheses.
Potential Impact:
1.4.1 Impacts
The scientific, technological and non-scientific outputs from the EpimiRNA project have resulted in important impacts. These cover diverse areas including improved understanding of the causes of epilepsy, advances in technologies to detect and monitor miRNAs, resources (biorepository, databases) for the research community, new therapies in pre-clinical development and impacts on our various SME partners.
Key Impact 1: Improved understanding of the aetiology and mechanisms of epilepsy
A primary objective of the EpimiRNA project was to generate new knowledge about the epileptogenic process. Research by the consortium and our collaborators has delivered the first complete identification of functional, conserved miRNA changes that accompany epileptogenesis in animal models and in the brains of refractory TLE patients. This impact was achieved through EpimiRNA’s innovative project design, which brought together, for the first time, experts in RNA research and the required underpinning technologies (deep sequencing, quantitative proteomics), with leading experimental epilepsy researchers and multi-center clinical teams. Specialists in electrophysiology, cell imaging and computer modelling helped determine how the miRNA changes influence excitability, spanning neuron-to-network level assessment. The findings have had a transformative impact on our understanding of epileptogenesis and chronic epilepsy. We have uncovered the involvement of numerous miRNAs in brain excitability and this has generated new understanding of the epileptogenic process.
Among the highlights, we have generated a molecular catalogue that features abundance data on every miRNA expressed in the mouse and rat hippocampus and how these change as epilepsy develops. The data contain reassuring entries on miRNAs previously linked to epilepsy and miRNAs known to be enriched (or absent), but also many surprises including a set of novel miRNAs common to all animal models that had not been previously linked to epilepsy. The human iCLIP dataset are remarkable in their value, allowing anyone to explore the target(s) of every miRNA made in the human epileptic brain.
Our functional studies have uncovered numerous new miRNAs which can influence epilepsy or seizures and demonstrated the target(s) and mechanisms. This includes a miRNA (miR-22) that works to counter excessive neuroinflammation, a miRNA (miR-324) that restrains ion channel levels in neurons, a miRNA that regulates axon guidance (miR-135a), a miRNA that controls calcium levels and movement of other RNAs within neurons (miR-129). For the first time, we proved the mechanism by which miRNA inhibitors altered seizures. As a result, we now know that blocking miR-22 worsens seizures because it allows excessive build-up of the P2X7 receptor that promotes neuroinflammation. We know that the seizure-suppressive effects of inhibiting miR-134 result from protecting (restoring) levels of an enzyme that regulates the structure of synapses (Limk1). These findings not only offer miRNAs as novel targets but can lead to non-miRNA targets for seizure control. For example, EpimiRNA researchers have now shown that drugs that block the P2X7 receptor can have disease-modifying effects in drug-resistant epilepsy. The set of novel miRNAs that we identified as common-to-all models included several that were non-neuronal. Despite this, interference in their function altered seizures. Thus glial-based miRNAs as well as neuronally-expressed miRNAs are targets for seizure control. Again, the introduction and establishment of these concepts have had a transformative impact on the field, as evidenced by the commissioning of multiple articles on the topic in topic research journals (Lancet Neurology, Epilepsy Currents, Epilepsia Open, Nature Reviews Neurology (invited)).
At the time of the project start, miRNAs were a nascent area and minor curiosity within the epilepsy research field. Over the past five years our project’s research along with findings by other groups around the world, have projected miRNA research to the forefront of the epilepsy research agenda. This is evidenced by the numerous workshops scheduled for and by EpimiRNA partners to discuss this topic at European and international epilepsy congresses. Our consortium‘s approach represented a new scientific model for how to uncover the mechanisms of epileptogenesis. The project has been instrumental in shaping the direction of basic, pre-clinical and clinical epilepsy research. The findings have also served to support diagnostic and drug development.
All our miRNA-target data will eventually be lodged in an open-access repository that will allow researchers from outside the project to take advantage of the discoveries and drive new research in the future. Indeed, while we had time to study the functions of ~10 different miRNAs during the course of our project, it is clear that further data mining is likely to identify other miRNA mechanisms and targets.
Key Impact 2: Biomarkers for epilepsy diagnosis
It is widely acknowledged that the identification of biomarkers of epilepsy and the epileptogenic process would transform discovery of disease-modifying therapies and the diagnosis of epilepsy in the future. Through the EpimiRNA project we have shown that miRNAs fulfil several key requirements for a blood-based molecular biomarker being enriched in the brain, dysregulated in epileptic brain tissue and manipulation of miRNAs can have seizure-suppressive and disease-modifying effects in preclinical models. A second major impact of the EpimiRNA project is the identification and validation of miRNAs in blood and CSF samples in patients with epilepsy. The results of these studies could be used to develop a blood or CSF-based test for epilepsy. This could improve the time-to-diagnosis for new epilepsy patients and reduce mis-diagnosis rates. Importantly, EpimiRNA’s studies compared results in epilepsy patients to results in patients with conditions that can often be mistaken for epilepsy such as psychogenic non-epileptic attack. The miRNA biomarkers discovered by EpimiRNA were not altered in these patients. This significantly increased confidence in the specificity of the miRNAs for epilepsy and makes commercial development more likely in the future. Indeed, the commercial market opportunity may be significant. The global in vitro diagnostics market has sustained compound annual growth rate of over 5% in recent years and is predicted to exceed €60 billion. There are currently no molecular diagnostics on the market to predict epilepsy therefore our programme has strong potential to drive innovative new healthcare products that have commercial value.
The EpimiRNA project was the first to discover that one of the miRNAs we had shown to be a target for seizure control and disease-modification, miR-134, was also found at higher levels in plasma from epilepsy patients. We followed this work with the (unbiased) discovery of a triple-miRNA plasma “signature” unique to epilepsy patients in plasma and miRNAs unique to CSF. Because we could show that levels of the miRNAs change in response to an anti-epileptogenic therapy (in mice) these could also be useful for stratifying patients in any future clinical trial of a disease-modifying therapy. Although initial work suggests miRNAs may not be changed in blood samples from humans exposed to brain stimulation devices, we did find some molecules associated with inflammatory responses change suggesting there may be biomarkers of therapeutic brain stimulation.
EpimiRNA research also made important advances in how we can enhance the diagnostic accuracy of miRNAs as biomarkers. This could influence the technical requirements of any future miRNA-based clinical test for epilepsy. Our research showed that measuring the relative amounts of any given miRNA that are bound to the AGO protein or sealed inside microvesicles called exosomes offers superior diagnostic accuracy relative to measuring the overall amount of a miRNA. It may be preferable to pre-process a biofluid before proceeding to miRNA extraction and detection. Our finding shows that protein-bound miRNA levels change after seizures could also be useful in developing a test for seizures. Such a test might be useful beyond the diagnosis of epilepsy – it could help identify seizures in patients admitted to emergency departments who are unresponsive. Currently this requires lengthy and technically-demanding EEG monitoring or brain imaging. A blood test could rapidly and cheaply be used to determine whether a seizure occurred and could be administered by staff with minimal or no specific neurological training.
Research by EpimiRNA and colleagues also drove the development of a series of prototype point-of-care miRNA detection devices. The central innovation is the generation of a dramatically amplified signal upon nucleic acid hybridisation based on nucleic acid functionalized electrocatalytic nanoparticles. During the project we helped advance this technology from a relatively slow, early prototype that could only detect a single brain-enriched miRNA (miR-134) to a rapid, point-of-care test (“TORNADO”) that featured move of the sample through a multi-chamber microfluidic system to allow rapid (~1.5 hour) detection of three different miRNAs simultaneously. Members of EpimiRNA continue to work on the development of this technology and are in advanced discussion about a spin-off company to commercialise the device. EpimiRNA’s exploitation efforts have identified at least one of the major manufacturers of anti-seizure drugs that would support efforts to commercialize a miRNA-based diagnostic device. Molecular diagnostics is the fastest growing segment in the diagnostics industry. Our pilot studies suggest sensitivity rivals high performing PCR target amplification but far surpasses it in terms of speed (results in seconds), operator simplicity and overall cost.
Having demonstrated that we can monitor a disease-modifying therapy by a blood-based miRNA panel, the EpimiRNA project has accomplished the expected future industry strategy of co-developed drugs and diagnostics that are increasingly important to pharmaceutical development. Our biomarker studies will advance in-step with the microfluidics device development to maximise the potential that both become available at the same time. We foresee our miRNA biomarker findings could lead to investment and our Technology Transfer Offices are in discussion to identify potential commercial partners and to identify, capture and commercialise further intellectual property from the project.
Key Impact 3: Preventing the development of epilepsy after potentially epileptogenic brain insults
Preventing the epileptogenic consequences of brain injury is an urgent therapeutic priority. EpimiRNA addressed this need by focusing on the development of miRNAs as therapeutic targets. In contrast to small molecules that hit just one single target, miRNAs can exert a "pleiotropic effect" as individual miRNAs regulate many targets in a network/pathway. At the time that EpimiRNA began, only a single study had investigated the functional consequences of targeting miRNA on the development of epilepsy. By the end of the project, research by EpimiRNA and others outside the project had tested and reported effects of over 20 different miRNAs on seizures or epilepsy. EpimiRNA took a set of miRNAs that were already disclosed, as well as novel miRNAs from its discovery work, and systematically targeted these in animal models. We answered key questions that would be needed to know before any kind of preclinical development could proceed. Are they safe? Do they work in multiple models? Do they affect normal animal behaviour? How important is the chemical “backbone” of the oligonucleotide antagomir? By what route could they be delivered to patients? What is the mechanism by which the miRNA inhibitor works? Can they stop seizures in already-epileptic animals? The scale and multi-disciplinary expertise within the EpimiRNA project allowed us to address all of these questions. The results will be encouraging to biotechnology or pharmaceutical companies considering this area of therapeutics for epilepsy or other diseases of brain excitability.
Advancing previously-known targets into pre-clinical development The EpimiRNA project made important progress in moving a miR-134-based antagomir toward preclinical development. By pooling resources, we demonstrated that the Ant-134 antagomir suppressed seizures in three other animal models and worked in rats as well as mice. Importantly, we demonstrated that normal firing properties of neurons in the hippocampus are not changed by the treatment – the molecule suppresses pathological brain activity while sparing normal functions. This strongly increases the likelihood that the treatment would be safe. We also showed that you could deliver the antagomir into a region of damaged (ultimately epileptogenic) brain tissue by a single systemic injection. We solved the main mechanism by which the molecule suppressed seizures in mice. Finally, we were able to show that a single injection of the therapy into already-epileptic mice permanently suppresses spontaneous recurrent seizures in mice. This is a critical test because the largest category of patients likely to receive this treatment would be patients where epilepsy is pre-existing but for whom neurosurgical treatment is not possible or for whom brain stimulation was ineffective, Together, the EpimiRNA project achieved all its objectives and strongly increased the translational value of the therapy. Partners at RCSI were awarded a US patent for this technology in 2017. During the project, partners at RCSI undertook collaborative pre-clinical development projects with US and European-based industry partners interested in using the therapy to treat drug-resistant epilepsy. This remains an area of ongoing collaboration. EpimiRNA activities have ensured biotechnology and Pharma companies as well as academic teams working on AED discovery, are provided with a robust rationale and series of already comprehensively pre-clinically tested miRNA manipulations, which will form the basis for future therapeutic approaches to prevent the development of epilepsy after epileptogenic brain insults. We are optimistic that a miRNA-based therapy for epilepsy can emerge from this research.
Identifying new targets for preclinical development: EpimiRNA generated a portfolio of pre-clinical assessments of novel miRNA-targeting antagomirs with potential applications in seizure control, epileptogenesis and disease-modification in epilepsy. In total, 9 additional miRNA-based therapies were tested during the course of the project in one or more animal model. One of these, Ant-135a, also prevented spontaneous recurrent seizures in epileptic animals, although the potency and duration of the effect was not superior to Ant-134. The remaining 8 antagomirs have only been shown to reduce seizures when given as a pre-treatment. It will be necessary, in future research by EpimiRNA partners or others, to explore whether these can reduce spontaneous seizures in already-epileptic animals.
Key Impact 4: Translating molecular and cellular targets into drug discovery
The EpimiRNA project has helped develop a non-miRNA inhibitor for seizure control. This recently entered the National Institute’s of Health’s epilepsy therapy screening program. The achievement was a collaboration between chemists at SME Bicoll and two of the academic partners who showed that the Bicoll molecule, now designated EPM06, has acute seizure-suppressive effects in multiple models and reduces epileptiform activity in the human brain.
Another outcome from EpimiRNA was an in vitro assay that can be applied to drug development that is based on identifying chemicals (either individual or initially as a mix) with miRNA-modulatory effects. The assay, based on a reporter molecule in a cell line that responds to the presence of a miRNA inhibitor, was used to screen a plant-based library from SME Bicoll for effects on miRNA processing. In the future, the assay could be useful directly or following further development by our SMEs, partners or future interests (e.g. Pharma).
Key Impact 5: Additional scientific impacts: miRNA and systems biology
The EpimiRNA project has had an important impact on our understanding of the molecular mechanisms underlying miRNA expression and malfunction. We have discovered how the levels of several miRNAs are controlled and we have generated new thinking about the importance of individual targets versus large-scale effects and shown this is miRNA-specific. EpimiRNA has developed novel molecular tools that should facilitate miRNA research in the future. This includes the development of miRNA reporter mice in which the AGO protein is selectively expressed in different cell types in the brain allowing researchers to assign specific miRNA species to neurons or non-excitable cells such as glia. This should lead to major strengthening of the Europe-based laboratories and SMEs focused on identifying miRNA targets and therapeutic opportunities. EpimiRNA also brought outstanding researchers not previously working on epilepsy to the field of epilepsy research including the teams of Gerhard Schratt (UNIMAR), Jorgen Kjems (AU) and Jens Anderson (SDU).
EpimiRNA has already had an impact on miRNA bioinformatics, generating new data integration infrastructure pipelines and a platform to ensure data from RNA-seq and proteomics are optimally mined and pathways identified and understood. The project has taken advantage of developments in the field of systems biology by developing innovative computational and mathematical modelling tools that explore the impact of miRNA on biological function at a systematic level, and that integrate qualitative and quantitative miRNA profiling, mRNA/protein target profiling, as well as neurophysiological data. These models now allow researchers to explore and explain miRNA effects, predict therapeutic targets, and deliver novel prognostic markers. Modelling of miRNA effects not only at the single neuron level, but also at the neuronal circuit/tissue level is furthermore a significant step towards more ‘holistic’ modelling approaches, and therefore also a major project impact.
Key impact 6: Scientific excellence, training and reputation
The EpimiRNA project has already delivered science of the highest quality and impact during the funding period and several major publications are in late development. This includes publications in Lancet Neurology, discoveries about mechanisms of miRNAs in EMBO journal and Cell Reports, the miRNA database in Bioinformatics and a series of miRNA biomarker studies in Scientific Reports and eBiomedicine. This supports Europe’s scientific and global R&D ranking. High impact publications will also attract the best collaborators from academia and industry to Europe to ensure the partners attract researchers of the highest level. The project helped grow the research teams of the participating partners and to retain existing talent (e.g. emerging investigators such as Dr Tobias Engel and Dr Eva Jimenez-Mateos who since obtained permanent faculty positions at top institutes) as well as attract new foreign and domestic researchers. Multiple members of the work package leader laboratories trained and supervised PhD students to completion who benefitted from the research excellence of the project and opportunities to learn from multi-disciplinary teams across Europe. Our innovative programme contained strong training for researchers including in much sought-after areas including pre-clinical animal models, clinical studies/trials, drug delivery/therapeutics, RNA chemistry and nanotechnology. The multi-disciplinary, inter-sectorial development and applied aspects of our research programme, meant team members came out equipped to secure industry as well as academic or other posts.
Key outputs from the research programme have been the establishment of new genetic models, discoveries on molecular mechanisms, technologies for gene and drug delivery to the brain and other “frontier research”. This will form the basis of more competitive applications by the project teams to the various European (H2020) funding mechanisms where the excellence of the science and record of the applicant group is fundamental to success. The success of the EpimiRNA project already drove the establishment of two epilepsy-focused national research centres. In Ireland, the EpimiRNA Coordinator is the Director of the FutureNeuro research centre. Two other EpimiRNA investigators are principal investigators in the Centre (J Prehn, G Cavalleri). The centre focuses on the development of novel diagnostics and innovative therapeutics enabled by the emerging connected (eHealth) sector. The project secured ~8€Mio in support from the Irish government and a further ~3€Mio in industry-committed investment. The clinical co-coordinator Prof Felix Rosenow established CepTER, the Centre for Personalized Translational Epilepsy Research, which secured ~5€Mio in funding in Germany. These projects build upon discoveries and infrastructure that were established through EpimiRNA.
Key Impact 7: Clinical and patient impacts from EpimiRNA
The biomarker, therapeutic and genetic discoveries made by EpimiRNA may lead to direct benefits to patients from faster diagnosis and new treatments but it may be several years before these will be realised. However, a number of discoveries may enter clinical use more quickly. Brain stimulation (BS) devices are already clinically important non-pharmacological treatments for pharmacoresistant epilepsy. Several are already approved but are currently only moderately effective. EpimiRNA has made important contributions to our scientific understanding of BS for epilepsy. First, we have identified a brain region that, when stimulated at a certain frequency, delays the onset of epilepsy. These animal data could be relatively quickly tried in patients and the EpimiRNA clinical network comprising multiple sites in Germany and Ireland could be involved in a future clinical trial. This could lead to new clinical investigations of alternate brain stimulation sites that could produce more powerful clinical effects or even disease-modification. Our studies found select changes to miRNAs occur in the brain in association with this therapeutic stimulation. While we could not prove that an individual miRNA was responsible, the project has generated several leads for further investigation that could identify seizure-regulatory miRNAs.
During our studies we investigated the potential therapeutic effects of transcutaneous vagal nerve stimulation (tVNS). This was a promising technology that offered a non-invasive
form of brain stimulation that avoids the need for surgery that is necessary for a VNS device. Clinical testing during the project lifetime failed to show a therapeutic effect of tVNS on drug-resistant epilepsy. Nevertheless, the device might still have utility, for example, in helping to predict who will best respond to VNS therapy. We identified a brain-specific miRNA that changes levels in the blood of a drug-resistant epilepsy patient given the therapy.
Human epilepsy genetics has the power to explain disease risk, identify new drug targets and avoid adverse drug reactions. EpimiRNA systematically explored all potential variation in miRNA and the pathway of their biosynthesis and all known targets in the hippocampus. We did not find that variations in miRNA could explain epilepsy, thereby refining our understanding of the genetic contributions to this disease. This may direct future research toward other regions of the human genome or stimulate an ever-larger study size which would be better powered to identify an effect of miRNA variation on disease risk. Regardless, this improves our understanding of epilepsy and can lead to revised standards that potentially reduce time to diagnosis and treatment providing potential healthcare savings and refine genetic counselling so patients have a better understanding of the cause and prognosis of their condition.
EpimiRNA’s research could - in the longer-term - help reduce health spending by delivering breakthroughs and infrastructure for diagnostics, new treatments and monitoring solutions that improve clinical outcomes. A prompt genetic diagnosis impacts on healthcare costs by speeding up time-to-diagnosis and helping better manage and monitor treatment compliance. EpimiRNA’s identification of molecular factors underlying drug-resistant epilepsy and biomarkers thereof could result in better matching of patients to treatments and fewer demands on healthcare resources creating savings for both hospitals and insurers. A patient with refractory epilepsy is more likely than a controlled patient to have a comorbid condition like depression, twice as likely to be hospitalized and when they are in hospital to stay longer. If a disease-modifying treatment emerged as a result of EpimiRNAs’ research it would have a transformational impact on long-term healthcare costs.
Key Impact 8: Benefit to participating SMEs
EpimiRNA’s core scientific objectives were all achieved and this was supported by excellent participation from our SMEs. EpimiRNA afforded our SME partners several opportunities to develop and protect novel intellectual property in the areas of therapeutics, medical devices for epilepsy, miRNA research tools and data integration platforms.
For Dixi, the project provided an opportunity to work with leading epilepsy surgery centres and develop a prototype combined intracerebral microdialysis probe. Due to challenges in preclinical testing the product did not enter clinical testing during the EpimiRNA project. However, should the company be interested in pursuing the device beyond the end of the project this could help grow their portfolio which includes intracranial electrodes for use in patients. There have been no competitor products that combine recording and microdialysis functions. The device also has the potential to be further developed into a local drug delivery system, resulting in the desired “closed-loop” single diagnostic and therapeutic delivery system for patients. This could also be extended to other pathologies such as paroxysmal pain or brain tumors; very important future markets.
The global epilepsy neurostimulation market is estimated at $0.5 billion. Cerbomed’s tVNS device was trialed during the course of the project for drug-resistant epilepsy but was not found to be efficacious. We did, however, identify a miRNA biomarker that could be useful in the future to predict responders. The company is currently being acquired in a merger. While the future of the tVNS system is therefore uncertain, the project nevertheless delivered clinical data on safety and tolerability and EpimiRNA’s findings may encourage future research of similar devices.
Targeting small non-coding RNA such as miRNA is a growing area of biomedical research with clear therapeutic applications. During the EpimiRNA project, SME InteRNA performed RNA sequencing and miRNA research that has supported their interest in expanding beyond cancer therapeutics to the central nervous system. EpimiRNA provided InteRNA early integration with a new field of research and access to expertise from multiple RNA specialists. With EpimiRNA, InteRNA were able to develop their capabilities in RNA research and therapeutic applications which will broaden and increase their potential market.
Bicoll currently offers small molecule drug-like compounds from natural resources for the global pharmaceutical and biotech industries. Through EpimiRNA, Bicoll was able to expand their focus on medicinal chemistry and custom synthesis and screen compounds from within their catalogue that impact on epilepsy or target novel processes, including miRNAs. As a result of EpimiRNA, Bicoll now has a drug lead candidate accepted into the NIH’s Epilepsy Therapy Screening program. This provides access to a battery of preclinical test models and will evaluate whether their lead molecule has a realistic chance of being developed into a new medicine. If successful, Bicoll would be likely to partner with an existing company with expertise in the epilepsy field.
The global market for genetic data analysis and management software continues to grow, partly as a result of high throughput „omics“ data opening new markets for data management and analysis. EpimiRNA enabled BCPlatforms to move towards the area of managing data efficiently and securely including data from „omics“ hetereogenous biological data which increasingly comes from international multi-institutional groups (their primary market currently). The project resulted in BCPlatforms creating cost effective tools for „omics“ platforms that can merge RNA-seq, ChIP-seq, proteomics and genome sequencing, considerably widening their potential market.
Key impact 9: Formation of a “spin-out” company
Another important achievement of the EpimiRNA project was the development of a new spin-out company (SME) from partner AU. The company, named Omiics, is being led by Dr Morten Veno who was responsible for the RNA sequencing efforts and iCLIP technique that were developed within EpimiRNA. Omiics, will offer bespoke RNA discovery and informatics resources to academic and commercial entities.
Key impact 10: Broader European and societal benefits of EpimiRNA
Partners in EpimiRNA have extensively engaged with patient organisations and the lay public during the course of the project. The focus has been on disseminating findings and reducing the stigma associated with the disease. This includes appearances on national television programmes to explain the disease and the type of research being performed in the project as well as lecture series at host institutes for patients and families. Researchers in EpimiRNA have also performed visits to schools to de-mystify careers in science and introduce concepts about how the brain works and help them understand a disease such as epilepsy. Through the epiXchange conference, members of EpimiRNA engaged more closely with project officers at the EC to provide oversight of S&T developments and influence future funding directions. The EpimiRNA project has also helped deliver on commitments made at a national and European level to increase investment in epilepsy research, which was mandated by the European parliament’s acceptance of the written declaration on epilepsy.
The EpimiRNA project has maintained the leading position internationally into the contribution of miRNAs to epilepsy. At the time of project start, this was driven principally by European groups, including the first functional studies of miRNAs. EpimiRNA and miRNA-focused research within other projects (e.g. EpiTarget) have ensured that half of the published research on miRNAs in epilepsy has come from European labs. The project has enabled this momentum to continue and ensued European-based teams continue to make major breakthroughs. EpimiRNA has also catalyzed greater interactions between EU nationalities. The EpimiRNA consortium’s large-scale, multi-disciplinary focused approach has effectively consolidated the European competitive advantage and attracted investment from outside (for example, the USA’s Regulus therapeutics funding a pilot miRNA therapy project within one of the EpimiRNA teams). This dominance is important because breakthroughs often translate into patent filings and other outputs. Indeed, the Coordinator’s institute RCSI was awarded the US patent for a miRNA-based treatment for epilepsy in 2017.
In addition to the scientific and technical achievements, EpimiRNA was central to the development of the epiXchange initiative – a large collaborative network comprising the major FP7-funded epilepsy projects and a number of other major European epilepsy-associated research projects. This resulted in a high-profile event in Brussels in 2018 (see section 1.4.2 for more details) and has since led to establishment of a network that aims to foster cross-project collaboration on epilepsy and jointly pursue research funding in the future.
The EpimiRNA consortium is distinct from other large project approaches that focus just on biomarkers or groups of genes involved in a single process (e.g. inflammation, extracellular matrix, neuropeptides) that are ongoing both in Europe and in the USA. The EpimiRNA project has established a unique programme that brings distinctiveness to European epilepsy research. EpimiRNA nevertheless complements large European and International research efforts and partners in EpimiRNA have been active in other FP and Horizon-2020-funded programmes. 1.4.2 Measure to maximize impact: Main dissemination activities
EpimiRNA communicated about the project’s main objectives and disseminated its main results using a variety of tools. These communication tools were carefully designed to each target audience from the scientific community to patient organisations and policy-makers. EpimiRNA-related activities were promoted on Twitter using @Epilepsylab Twitter account from 2014 to 2016 and a dedicated project Twitter account (@EpimiRNA) from January 2016 onwards, with 295 and 151 followers respectively, as well as 117 and 143 tweets and retweets respectively.
Promoting miRNA research within the scientific community
At the start of the EpimiRNA project, research on miRNAs was at its infancy and very little was known about their role in epileptogenesis and chronic epilepsy. We contributed to uncover new miRNAs and demonstrate their mechanism of actions, making miRNA a credible approach to treat and diagnose epilepsy. Over the course of EpimiRNA, our research already resulted in more than 20 publications in peer-reviewed journals and many more publications are still to come. In addition to talks during scientific conferences and posters, EpimiRNA partners contributed to and led several workshops on epilepsy-miRNA related research at major European (European Congress on Epileptology (Stockholm 2014; Prague 2016, Vienna 2018)) and International conferences (International Epilepsy Congress (Barcelona 2017)) showing that miRNA in epilepsy is now a credible field of research and triggering interest for this research area.
Targeting patients/ lay audience to inform about state-of-the art of epilepsy research
Given the novelty of the miRNA approach to treat and diagnose epilepsy, the focus of the dissemination materials towards a non-scientific audience was always to explain first what miRNAs are and which role do they play at cellular level, before explaining the project objectives and main results. The main communication tools towards the lay audience have been the EpimiRNA website (updated constantly during the project life-time), the release and distribution of the project flyer (printed and electronic version), as well as the production of 3 short videos , available from the EpimiRNA website https://www.epimirna.eu/media/videos/
From 2015 onwards, several face-to-face events have been organised with patient organisations (e.g. Epilepsy Ireland, Ireland’s epilepsy charity) at the coordinator host institution to demystify epilepsy and explain how research on miRNAs could help tackle the many challenges that epilepsy patients are currently facing.
International and European media coverage consisted in several articles being published in the EU’s Horizon magazine, as well as television, radio and press coverage, especially in Ireland where the host institution of the project coordinator is located.
Working hand in hand with epilepsy lobby organisations and other epilepsy-funded projects to promote an agenda for epilepsy research
EpimiRNA partners worked hand-in-hand with six other FP7-funded large collaborative projects (EpiTarget, Desire, EpiSTOP, Epixchange, Epicare and EpiPGX) and existing epilepsy lobby organisations (such as ILAE) to promote a future research agenda for epilepsy. These joint efforts were particularly helpful to target policy-makers at the European Commission and the European Parliament (especially the members European Advocates for Epilepsy working group). The main joint action consisted in the organisation of the epiXchange 2018 conference, a major international conference that was organised in Brussels (23rd May 2018) and that brought together representatives from the industry, patient organisations, research community and policy-makers. The programme of this conference aimed at showcasing the results of the FP7-funded projects and highlight existing research gaps. A paper is currently being drafted and will highlight the major achievements of the projects, as well as recommendations for future epilepsy research. This paper will be submitted shortly.
Some numbers from the epiXchange conference:
• 170 participants from 18 European and 5 non-European countries
• 250 watching the live-stream on Facebook
• Participation of officials from the EC, the European Brain Council, EMA, patient organisations (e.g. Epilepsy Foundation and Federazione Italiana Epilessie) and pharma (strong presence of UCB Pharma)
• 28 talks and 38 posters divided into 5 main themes
• 800 Tweets & retweets from October 2017 to August 2018 to promote the event
On the epixchange website. www.epixchange2018.eu all material from the conference can be found, including the full live-stream (and special cutouts to dedicated part of the conference) and presentations (as pdf).
Promote uptake of project results by industries outside the consortium
During the project life-time, we maintained continuous dialogue with potential industry partners to support the transition from basic research to the clinic. This has resulted into two active collaboration programmes with a pharmaceutical industry to drive the pre-clinical development of a new therapy epilepsy. At the time of writing this report, RCSI were contacted by a leading company that manufactures anti-epileptic drugs to explore potential development of a blood test for epilepsy diagnosis. Discussions are currently ongoing.
Training and harmonization of procedures to ensure high-quality research
Given the multidisciplinary research carried out in EpimiRNA, a number of training visits were organised between EpimiRNA partners during the project life-time. Among other things, partners were trained to perform high-quality hippocampal microdissections, RNAseq data.analysis in vitro methods for electrophysiology, ia KA model and antimir injections.
SOPs for key techniques used at multiple sites were prepared. These relate to techniques for freezing brain tissue, collecting blood and processing plasma and preparation of CSF.
List of Websites:
www.epimirna.eu; Coordinator: David Henshall, RCSI, dhenshall@rcsi.ie