The Risk versus Reward of Targeting Epileptogenic Signaling Pathways via miRNA Replacement Therapy

Commentary

MicroRNAs (miRNAs) are evolutionary conserved, noncoding RNAs that regulate gene expression. miRNA expression is deregulated in CNS disease, and epilepsy can be modified or prevented by targeting miRNA expression (1). Based on these observations, miRNA-based antiepileptogenic therapies may be effective alone or in combination with current targeted therapies. Gene therapy requires DNA plasmids or viral vectors to deliver the encoded protein; and vector size, inefficient delivery, and required nuclear localization represent technical challenges that limit this approach to local administration. Smaller miRNA mimics need only to enter the cytoplasm of target cells and can be delivered systemically. Nonspecific, off-target effects are unlikely as miRNA mimics have the same sequence as the naturally occurring miRNA and will target the same mRNAs as the depleted miRNA. miRNA-based therapies are advantageous because they concurrently target multiple effectors of pathways involved in epileptogenesis. Current understanding of epilepsy suggests it is a “pathway disease” that may only be successfully treated when multiple epileptogenic pathways are simultaneously downregulated. Hypothetically, restoring the expression of miRNAs via synthetic miRNAs will reinstate normal cellular programs and interfere with epileptogenic mechanisms. However, affecting a broad set of targets has far-reaching biological and potentially unpredictable consequences.

Brennan et al., proposed that proinflammatory and neuron restrictive silencing factor (NRSF)- mediated epileptogenic mechanisms could be inhibited via synthetic miR-124 preventing the development of spontaneous seizures in a model of kainate-induced epilepsy. The studies focus on hippocampal-specific miR-124 due to its link to expression of the transcriptional repressor NRSF (2), and because its expression has been reported to reduce the activation state of microglia and macrophages (3).

Hippocampal NRSF expression increases several-fold following seizures (4, 5). Neuron restrictive silencer element (NRSE)-containing genes are preferentially repressed by seizure-provoked increases in hippocampal NRSF levels. NRSE-containing genes (e.g., Calb1, Glra2, Grin2a, Hcn1, Kcnc2, Klf9, Lrp11, Myo5b, and Stmn2) are downregulated following kainate-induced seizures and can be rescued by NRSE-oligodeoxy-nucleotides (ODNs) (6). This group and others have previously demonstrated the critical role of NRSF in epileptogenesis. Specifically, that downregulation of hyperpolarization-activated and cyclic nucleotide-gated (HCN) channels can be restored and short-term seizure frequency reduced following NRSE-ODN treatment in kainate-induced epilepsy (5).

Conflicting data exist in relation to the expression of miR-124 following brain insults and in epileptic tissue. Decreases in miR-124 have been observed 48 hours after pilocarpine-induced SE (7) and in the hippocampus of mesial temporal lobe epilepsy (MTLE) patients (8). In contrast to these findings, an upregulation of miR-124 has been observed acutely and chronically following pilocarpine-induced (SE) in immature rats and in tissues obtained from surgically treated children with MTLE (9), in the hippocampus 24 hours after pilocarpine-induced SE (10) and in hypoxic ischemic brain damage in neonatal rats (11). Conflicting results have also been reported for the use of miR-124 as a therapeutic target. Both inhibition of miR-124 (12, 13) and exogenous miR-124 (2) have been reported to reduce brain injury and functional impairment. Therefore, there is evidence both for and against the use of synthetic miR-124 as a potential antiepileptogenic therapy.

Brennan et al., describes a seizure-induced surtuin 1 (SIRT1) reduction of miR-124 directly augmenting the expression of transcription factor CCAAT/enhancer binding protein (C/EBPα) and subsequent binding of C/EBPα to the promoter of NRSF. Hippocampal SIRT1 mRNA and protein expression does not change following kainate-induced SE, but the authors were able to detect an acute change in nuclear fractions at 1 and 4 hours. Chromatin immunoprecipitation (ChIP) experiments demonstrated increased binding of SIRT1 to the miR-124 promoter and reduced H4K16ac levels. Total miR-124 levels were reduced by 90 minutes after kainate-induced SE and remained low for 48 hours. Levels of mature miR-124 were significantly reduced following seizure-like events in organotypic hippocampal slice cultures and it was in this system that the authors investigated whether SIRT1 mediated seizure-induced reduction in miR-124. SIRT1 inhibition reversed these changes and prevented repression of pri-miR-124-1 and restored levels of mature miR-124.

NRSF mRNA and protein levels are increased 4 hours after and reach peak expression levels 48 hours after SE mediating the repression of target genes including HCN1, GRIN2A, and KCC2 48 hours after SE. miR-124 agomirs administered directly after SE block the increase of NRSF mRNA, increases in nuclear NRSF protein, and seizure-induced repression of GRIN2A in the hippocampus at 48 hours. HCN1 and KCC2 mRNA levels were not measured or not reported after miR-124 administration. Hippocampal C/EBPα mRNA and protein levels were significantly increased following SE and remained high for up to 48 hours. miR-124 agomir administration following SE prevented increases in C/EBPα mRNA levels, but protein levels are not reported. ChIP analysis demonstrated increased binding of C/EBPα protein to the NRSF promoter following SE, but we do not learn whether miR-124 reverses or blocks this association. Instead, treatment with decoy ODNs composed of the C/EBPα binding recognition site were used to demonstrate decreased C/EBPα binding to the NRSF promoter and NRSF gene transcription following seizure-like events in vitro (i.e., organotypic hippocampal slice cultures treated with kainate). These results in hand, the authors conclude that C/EBPα is a “seizure”-dependent and miR-124–dependent regulator of NRSF expression in hippocampus.

The authors describe an increase in proinflammatory mediators (i.e., increased Cd11b, interleukin (IL)-1β, IL-1r1, IL-6, and tumor necrosis factor (TNF)-α mRNA and microglia activation) concurrently with the increased NRSF signaling following SE. miR-124 has been shown to be expressed in the microglia and synthetic expression of miR-124 has been shown to repress microglia activation and inflammatory molecular sequelae (2). Counter to these findings, Brennan et al., reports miR-124 is exclusively expressed in neurons and not in microglia. In contrast to previously published reports, administration of miR-124 agomirs induced inflammatory signaling cascades as evidenced by increased IL-1β, TNF-α, IL-6, and IL-1 receptor, and CD11b mRNA in both control and kainate groups when compared with scrambled agomir administration and miR-124 increased the activation of microglia (defined as a transformation of IBA1 cells from thin and ramified into globular cells). The mechanism underlying miR-124–induced inflammation remains unclear, but the authors suggest that this promotes epileptogenesis without actually demonstrating a change in seizure frequency. Evidence of inflammation was noted in all groups injected with miR-124, and if miR-124 was pro-epileptic it may have induced seizures in the control animals. Control animals were video-EEG monitored, but the authors note the miR-124 did not affect basal EEG in controls or kainate-treated rats, and control rats did not develop seizures. The measured difference in inflammation (i.e., the potential epileptogenic-inducing mechanism) in the miR-124 treated group relative to the scrambled group cannot be interpreted as sufficient to cause a change in the development of spontaneous seizures, the progression of seizure frequency, or the seizure severity based on the lack of any difference in these measures. Concurrent induction of inflammatory mediators via miR-124 in parallel with the blockage of NRSF increases and GRIN2A decreases following SE is suggested as an opposing pathway that may be interfering with the robust antiepileptogenic potential of NRSF-mediated changes. However, a different interpretation might be reached based on these results as they have been presented. If the miR-124 agomirs are capable of significantly altering nuclear NRSF and downstream GRIN2A, but spontaneous seizures are not changed in any way, then those SE-induced molecular mechanisms may not be important to the process of epileptogenesis and may not be the best therapeutic targets.

The authors concede that reduced miR-124 expression following SE may be an adaptive response despite the extensive work devoted to demonstrating that miR-124 replacement therapy would attenuate epileptogenesis via NRSF. They should be commended for publishing their negative results that highlight the complex and potentially opposing roles of miR-124 that should act as a cautionary tale for 1) therapeutic targeting seizure-induced pathways molecularly characterized in part or entirely in vitro, and 2) the use of synthetic miRNAs in epilepsy and other “pathway diseases.” The failure in this antiepileptogenic approach is less likely due to the tool and instead due to the target (i.e., miR-124) chosen based on inconsistent studies and results that could not be reproduced in the context of kainate-induced epilepsy.