|Home | About | Journals | Submit | Contact Us | Français|
MicroRNAs are small noncoding RNAs that work posttranscriptionally to negatively regulate protein levels. They influence neuronal and glial structure and function, neuroinflammatory signaling, cell death, neurogenesis, and other processes relevant to epileptogenesis. Functional studies using oligonucleotide inhibitors (antagomirs) and mimics (agomirs) to modulate microRNAs in rat and mouse models of epilepsy show effects on evoked and spontaneous seizures and attendant neuropathology. The present review summarizes recent findings and points to gaps in our knowledge of the underlying mechanisms and directions for the future.
There has been substantial progress in understanding the cellular and molecular mechanisms of epileptogenesis, and this has led to new ideas about how to develop novel disease-modifying treatments (1–3). A compelling idea is to find control points—nodes—that coordinate the expression of one or more networks of genes (4, 5). Targeting these could bring about larger-scale interference in the pathways that drive the development or maintenance of the epileptic state. MicroRNAs offer potential opportunities in this regard. MicroRNAs are small noncoding RNAs that control gene expression at a post-transcriptional level, promoting either degradation of their target mRNAs or translational repression (6). MicroRNAs exert network-level effects on gene expression because each can target multiple transcripts that encode proteins in the same pathway or single targets in multiple pathways. Their effects are particularly important for reducing “noise” and conferring precision to cellular protein levels, particularly for low-abundance targets (7, 8). Levels of numerous microRNAs have been found to be dysregulated in experimental and human epilepsy, and modulation of their biogenesis or targeting of individual microRNAs can produce potent effects on evoked and spontaneous seizures in animals (9). The present review aims to capture the pace of discovery by summarizing recent in vivo studies that manipulated microRNAs in rat and mouse models of epilepsy. The review also points to gaps in our knowledge and suggestions for future directions.
MicroRNAs function at a posttranscriptional level to fine-tune protein levels in cells. They do this by binding to short complementary sequences in mRNAs, often within the 3′ untranslated region of the transcript. This has the effect of promoting degradation or inhibiting translational efficiency, leading to overall lower protein levels in cells (6). MicroRNAs originate from inter-and intragenic sites including introns of protein-coding genes and their own specific gene loci. Transcription generates a primary transcript (pri-microRNA). which goes through a nuclear RNase-mediated processing event mediated by the Drosha microprocessor followed by cytoplasmic cleavage of the pre-microRNA by Dicer, resulting in the mature, double-stranded microRNA (~22 nucleotides) (10, 11). One strand of the mature microRNA is selected by the protein Argonaute-2, forming the so-called RNA-induced silencing complex (RISC), which then begins to search mRNAs for a region of sufficient complementarity upon which stable base-pair binding occurs (12). It requires only a 7 to 8 nucleotide “seed” region match for microRNA to mRNA targeting effects to occur; this is why microRNAs have so many potential targets and is the source of their multi-targeting property. After stable binding, other proteins may be recruited to the RISC, which promotes degradation or translational inhibition (13). MicroRNAs are currently named using a number-based system in order of their discovery (e.g., miR-1, miR-2). The number of microRNAs encoded in the human or rodent genome likely ranges from ~500 to more than 2,000 (14). Figure 1A provides an overview of microRNA biogenesis and function.
Studies in genetically modified mice lacking key biogenesis genes have revealed that microRNAs are essential for normal organism development, including brain maturation (15). More recent work in conditional models has determined that loss of microRNA biogenesis components from the mature brain results in progressive tissue dysmorphogenesis, neurodegeneration, and seizures (16–18). The role of a number of specific microRNAs in brain excitability is understood. MicroRNAs including miR-132 and miR-134 target transcripts encoding proteins crucial to growth and remodeling of dendrites, thereby regulating synaptic strength and plasticity (19–21). MicroRNAs miR-9 and miR-124 target mRNAs involved in neurogenesis and integration of newborn neurons into the circuitry of the mature brain (22, 23). MicroRNAs that control the regenerative capacity of axons during aging have also been identified in a model organism (24). There are also a number of glial-expressed microRNAs with potential relevance to epilepsy. For example, miR-155 regulates a network of genes coordinating microglial responses (25). Together, these studies reveal an important influence of microRNAs on brain function that spans most of the processes implicated in epileptogenesis, from controlling synaptic structure and function, glial morphology, inflammatory responses, and beyond. Thus, microRNAs may be relevant to the pathogenesis and treatment of epilepsy.
The first evidence that seizures could be altered by targeting an individual microRNA was reported for miR-134 (26). Intra-cerebroventricular injection of oligonucleotides designed to bind the microRNA, termed antagomirs (Figure 1B), resulted in suppression of evoked and spontaneous seizures in a mouse kainate model (26). Later, knockout of miR-128 in mice was shown to result in fatal epilepsy, whereas overexpression of miR-128 had acute seizure-suppressive effects (27). The hyperexcitability effect of reduced miR-128 was recently confirmed by an independent team using primary neurons transfected with a “sponge” construct to deplete miR-128 (28). Additional studies found that antagomirs targeting miR-34a (29) and miR-132 (30) could reduce seizure-induced neuronal death in models of status epilepticus. Inhibiting miR-132 after status epilepticus has also been reported to reduce the frequency of spontaneous seizures (31). The number of microRNAs for which functional data are available in in vivo models of epilepsy has been expanding rapidly. Recent examples are reviewed below.
The recent studies on microRNA manipulations in in vivo rat and mouse models have invariably used oligonucleotide-based approaches in two types of experiments: 1) by injecting an antagomir to bind and block the microRNA, or 2) by injecting a microRNA mimic (agomir) to artificially increase the microRNA level (Figure 1B). These may be delivered before inducing status epilepticus, looking at the effects on acute seizures or damage, or injected after status epilepticus, and then following the effects on the emergence or frequency of spontaneous seizures. The first type of study screens for potential acute seizure-suppressive effects that would work much like current anticonvulsive drugs. The second approach has the potential to identify disease-modifying (antiepileptogenic) effects of microRNAs.
Four microRNAs have recently been shown to have potential direct or indirect seizure-suppressive effects. Intracerebroventricular injection of mimics of miR-22 (a microRNA identified as upregulated in nondamaged brain after status epilepticus) into mice during the first few days after status epilepticus, induced by intra-amygdala kainic acid, was shown to transiently reduce spontaneous seizures (32). Conversely, antagomirs targeting miR-22 caused more spontaneous seizures in the model and exacerbated astrogliosis within the hippocampus (32). A key target of this microRNA was shown to be the ATP-gated P2X7 receptor, which is known to promote release of interleukin-1β. Thus, miR-22 may function to suppress neuroinflammatory signaling in epilepsy.
Three other microRNAs with putative seizure-suppressive effects are miR-23b, miR-124, and miR-219. Levels of miR-23b were reported to be downregulated in a mouse kainate model, and delivery of miR-23b mimics into the mouse ventricle reduced post-kainate spiking seen the next day (33). The mechanisms and targets of miR-23b were not explored. Another microRNA downregulated in experimental epilepsy was miR-124. Pretreating rats with miR-124 mimics delayed seizure onset and resulted in less severe seizures in the pilocarpine model (34). Tests in the pentylenetetrazol model produced similar results (34). Potential mechanisms were suggested including targeting of glutamate-related signaling components and the cyclic-AMP response element binding protein (34). Lower levels of miR-219 were reported in experimental and human epilepsy and injection of an miR-219 mimic protected against kainate seizures in mice. Inhibiting miR-219 with antagomirs was, on its own, sufficient to produce epileptiform spiking in mice. The effects of miR-219 were suggested to be mediated by targeting of glutamatergic signaling components (35).
Three new microRNAs have been identified, which appear to promote seizures or epilepsy. A recent study showed that delivery of antagomirs targeting microRNA-155 prior to status epilepticus induced by pilocarpine in mice resulted in a trend toward lower mortality and lower scores in a Racine-type behavioral assessment and suggested brain-derived neurotrophic factor as a possible target. The effects of the antagomir on post–status epilepticus epilepsy or neuropathology were not reported (36).
Levels of miR-199a were found to be increased following pilocarpine-induced status epilepticus in rats (37). Pretreating rats with an antagomir targeting miR-199a prior to status epilepticus reduced acute seizure severity and damage. The authors identified the mechanism of the antagomir effect as partly the result of de-repression of Sirtuin1, a transcriptional silencer (37).
In a profiling study using a mouse model of pilocarpine-induced status epilepticus, miR-203 was identified to be upregulated during the epileptic phase and increased in hippocampi from patients with epilepsy (38). Using intranasal delivery, the authors showed that antagomirs targeting miR-203 reduced the frequency of spontaneous seizures in epileptic mice over a two-week analysis period and identified the glycine receptor B as a potential target (38).
While Brennan and colleagues found no improvement in epilepsy in animals given miR-124 mimics, they observed that injection of the mimics alone evoked inflammatory responses. This may have obviated any potential seizure-suppressive effects of elevating miR-124 in neurons. The finding appears to conflict with the anticonvulsant effects reported for miR-124 mimics given 3 days prior to pilocarpine-induced seizures in another study (34). However, the differences may relate to whether pathology is present in the brain at the time of miR-124 manipulation, such that anticonvulsant effects of miR-124 are only observed when delivered to a normal brain. Antagomirs targeting miR-181a were reported to reduce neuronal death after status epilepticus in rats (39).
The recent findings are encouraging in identifying several new microRNAs that offer opportunities for seizure manipulation and neuroprotection. Despite this progress, a number of questions and concerns remain. Most of the studies tested with only a single model. Do the microRNA manipulations work in other models? Validation by an independent group and tests in more than one species would increase confidence in the value of specific microRNAs as targets. So far, microRNA manipulations have only been explored in chemoconvulsant models. These suffer certain confounders, and tests in models such as kindling or electrical stimulation of the perforant path are warranted. By what mechanism does the microRNA manipulation work? The targets of the microRNAs have often not been directly demonstrated or even investigated. Studies will need to prove colocalization and direct interaction of the microRNA and proposed target(s).
Another question is whether effects of microRNA manipulations are a combination of small effects on multiple targets or strong effects on just a few. Many of the recent functional studies on microRNAs have limitations in experimental design or data reporting. While acute anticonvulsant effects of microRNA manipulations are interesting, it is unlikely that a microRNA-based treatment would be used for this purpose unless it worked through markedly different mechanisms to current anticonvulsants. Rather, their therapeutic utility lies more in an ability to produce sustained disruption or stabilization of gene networks during development or maintenance of the epileptic state. Most of the recent studies, however, do not allow a distinction to be made between 1) an acute seizure-suppressive effect, 2) a chronic, long-lasting seizure-suppressive effect, and 3) a true antiepileptogenic/disease-modifying effect. Distinguishing between possibilities 2 and 3 may primarily depend on the duration of the microRNA manipulation's effects on the target. If the effect on seizures only persists as long as the direct microRNA manipulation effect is present but then wears off, it would be 2; if the effect on seizures persists longer or permanently, it would be 3. With perhaps one exception (31), the recent studies did not track animals long enough to conclude a disease-modifying effect occurred. Last, while perhaps a coincidence, many of the new microRNAs with antiseizure properties were downregulated after status epilepticus, whereas those with proseizure effects were upregulated. This was the case previously for miR-134 and miR-128. This finding may shape decision making around selecting microRNAs to target in the future.
Even if a microRNA manipulation passes the hurdles above, there remain important challenges. First, how will a microRNA-targeting treatment be given to a patient? Oligonucleotide-based molecules are too large to cross an intact blood-brain barrier. Delivery could be timed with injury-induced opening of the blood-brain barrier or via a route such as intrathecal or intranasal (26, 38). Alternatively, therapeutic application might use some kind of cell-penetrating adaptor or lipid-based carrier. The benefits of microRNAs owing to their multi-targeting actions may be offset by difficult-to-predict off-target effects. These could be limited, however, if the number of relevant targets is restricted or where the microRNA is expressed only within the brain region triggering seizures. Biotechnology companies are active in developing RNA-based therapeutics, and both an antagomir (40) and a mimic (41) have now been tested in clinical trials.
In summary, the pace of recent functional studies interrogating microRNAs in animal models is encouraging. This must be accompanied, however, by a period of consolidation in which the reported effects are validated by others, the mechanisms better understood, and delivery options explored. If successful, we may see microRNA-based treatments enter preclinical development and, perhaps one day, reach patients.
The author would like to acknowledge funding from the European Union Seventh Framework Programme (FP7/2007-2013) under grant agreement No. 602130, Science Foundation Ireland (13/IA/1891, 14/ADV/RC2721, 11/TIDA/B1988, 08/IN1/B1875), and the Health Research Board Ireland (HRA-POR-2013-325).
Editor's Note: Authors have a Conflict of Interest disclosure which is posted under the Supplemental Materials link.