Muscle secretion of toxic factors,regulated by miR126-5p, facilitates motor neuron degeneration in amyotrophic lateral sclerosis

Amyotrophic lateral sclerosis (ALS) is a lethal neurodegenerative disease characterized by neuromuscular junction (NMJ) disruption, motor neuron (MN) axon degeneration, and neuronal death.Unfortunately, there is currently no effective treatment available for ALS and consequently, most patients die several years post diagnosis. The neurodegeneration that occurs in ALS is considered to be a non-cell autonomous process involving interactions between the motor neuron and its diverse extracellular microenvironments via an unknown mechanism. Distal Axonopathy is one of the early disease signs; however, the involvement and contribution of neighboring tissues and specifically, the muscle environment to the disease pathology remain controversial. Few works have concluded that muscles play a minor role or none at all in ALS pathology. For example, reducing hSODG93Adirectly in the muscles of the SODG93Amouse model, as well as crossing lipoxygenase (LOX) SODG37Rwith the Cre coding sequence under the control of the muscle creatine kinase (MCK) promoter, or performing manipulations using follistatin did not affect the disease's onset and survival (Miller et al.,2006). Moreover, application of muscle condition media (CM)from SODG93A-expressing muscles on healthy spinal cord neurons or embryonic stem cell-derived motor neurons in vitro resulted in no appreciable effect (Nagai et al., 2007). In contrast with these findings, overexpressing SODG93A protein specifically in healthy skeletal muscle results in severe muscle atrophy and induces an ALS phenotype. In addition, expressing hSOD1 with G37A and G93A gene variants only in skeletal muscles led to limb weakness,NMJ abnormalities, MN axon degeneration, and cell death, suggesting a direct role for muscles in ALS physiology (Dobrowolny et al., 2008). Recently, we have characterized a mechanism by which the diseased muscles contribute to the motor neuron degeneration observed in ALS (Maimon et al., 2018). Using a simplified microfluidic chamber (MFC) for studying muscle and motor neuron interactions, we demonstrated that ALS-mutant muscles affect MN axons. Our results show that ALS-mutant muscles facilitate a delay in axon growth towards the muscle compartment, axon degeneration, and NMJ disruption. However, eventually the connections between axons and muscles are established. Thus, at least in our system, apparently the non-cell autonomous contributions of the muscle are insufficient to recapitulate all the toxic effects observed in ALS. Interestingly, once the MNs also carries an ALS mutation,the axons are more susceptible to degeneration by mutated muscle CM. Therefore, apparently although the muscles have contribution to ALS progression, MNs are key in ALS physiology.

In order to survive, MNs have to respond accurately in both space and time to intra- and extracellular cues. These responses involve different mechanisms including local protein synthesis,ligand receptor interactions, and axonal transport machinery. Alterations in these mechanisms can lead to cellular dysfunction and disease. Axonal transport machinery, which supplies the distal synapse with newly synthesized protein and lipids, and signals the cell body to initiate activity in distal axons, was found to be dramatically altered in ALS disease. Importantly, alterations in transport can induce neurodegeneration and neuronal cell death (Perlson et al.,2010). Moreover, there is an emerging consensus that extracellular cues can induce retrograde death signals under disease conditions.An elegant example was recently published demonstrating a mechanism by which a death signal is formed and moves along dorsal root ganglion (DRG) axons in an ALS model (Pathak et al., 2018).Taking this into account, we speculate that there are undiscovered retrograde death pathways specifically in ALS-diseased MNs that cause the normal MNs to be more vulnerable to its toxic distal environment. Thus, this basic mechanism, of cell bodies respond to distal stress in health and disease has to be deeply characterized in order to progress toward possible future treatment for ALS (Figure 1).

Our results further indicate that the expression of ALS-causative mutations results in the secretion of multiple toxic factors. At least one of them is Semaphorin3A (Sema3A). Sema3A protein is well known for its ability to act as a repellent guidance molecule (Luo et al., 1993) that interacts with the neuropilin1 (NRP1) co-receptor binding component for its proper functioning Sema3A was found to be up-regulated in several neurodegenerative diseases (Kaneko et al., 2006). Importantly, it was found to be overexpressed in terminal Schwann cells (TSCs) of the SOD1G93Atransgenic mouse model for ALS and in the motor cortex of ALS patients (De Winter et al., 2006; Körner et al., 2016), suggesting that it plays a toxic role in disease pathology. Interestingly, injecting specific ab's against NRP1 resulted in improvement in disease progression in vivo(Venkova et al., 2014). Nevertheless, anti-NRP1 blocking antibody has only a modest effect. In contrast to these results, a recent study demonstrated that crossing mice expressing a truncated form of Sema3A with SOD1G93Amice does not result in any rescue effect(Moloney et al., 2017). An explanation for this contradiction could be the possibility that secreted factors such as Sema3A plays a more complex role in the biology of MNs (Zahavi et al., 2017). Indeed,Sema3A was shown to increase the survival of MNs when added to MN cultures, but it induced degeneration when introduced to the MN axons (Molofsky et al., 2014; Maimon et al., 2018). Consistent with this, deletion of the Sema3A gene specifically in spinal astrocytes resulted in a gradual loss of spinal MNs (Molofsky et al., 2014). Interestingly, the levels of Sema3A in the spinal cord of post-mortem ALS patients is down-regulated, compared with healthy controls (Körner et al., 2016). Thus, assuming that Sema3A total knock out will rescue ALS phenotype make no sense. Apparently, Sema3A may play a dual spatial role in MNs: When secreted from muscles and it targets distal axons at NMJs, it mediates their destabilization; however, when it is secreted by spinal astrocytes and targets MN soma, it acts as a survival factor. Following this idea, a better approach for treating ALS in this regard would involve down-regulation of Sema3A in distal muscle tissue but over-expressing it in spinal cord astrocytes near the MN soma (Figure 1).

Figure 1 miR126-5p dysregulation in ALS disease.

However, since Sema3A is only one factor, out of others that were found to be involved in ALS disease and since blocking its activity in the CM of diseased muscle results in a very mild rescue in our hands, a general gene repression mechanism, such as the microRNA (miRs) system, should also be explored. This assumption is also consistent with the fact that miR alterations are apparent in various neurodegenerative diseases including ALS microRNAs are post-transcriptional regulators that play an important role in many cellular processes including axon growth and retraction. Alterations in RNA metabolism and miRs can contribute to, and also be part of mechanisms that initiate the disease. These results lead to additional attempts to either use or target miRs as therapeutic strategies. In our study, we found that miR126-5p is down-regulated in SODG93Amuscles. Furthermore, alterations in miR126-5p expression profile were identified specifically in axons of ALS models (Rotem et al.,2017). We also show that miR126-5p can regulate both Sema3A and NRP1 expression. Importantly, overexpressing miR126-5p in muscle culture of ALS disease models improves the axons' state in vitro. Injecting LV-miR126-5p-GFP into the gastrocnemius muscles of pre-symptomatic diseased mice results in mild rescue. Thus, we demonstrated not only the involvement of miR126-5p in ALS pathology, we also suggest the possible manipulation of this pathway as a potential treatment for neurodegenerative disease (Figure 1).

The concept of RNAi treatments for neurodegenerative diseases is widely common in the current era For example, recent published papers demonstrated that antisense oligomers is powerful tool for treating huntington disease both in mice and non-human primates as well as could be a possible treatment for ALS C9orf72 mutation(Kordasiewicz et al., 2012). One disadvantage of RNAi treatment is its ability to regulate several pathways, thus it can be unspecific and affect unrelated genes. However, for complex, multi factorial disease like ALS, regulating multiple genes and pathways can be beneficial. For example, aside from targeting Sema3A and NRP1,miR126-5p is thought to regulate other Semaphorin-related genes such as other type 3 Semaphorins, several Plexins, and further downstream signaling molecules such as c-jun n-terminal kinase(JNK) and phosphatase and tensin homologue. Moreover, miR126-5p can regulate a few ALS-related genes such as vascular endothelial growth factor A, SPAST, matrix metalloproteinases, AGRIN,and C9orf72 which are directly involved in ALS. Accordingly, the advantages and disadvantages of using RNAi as a tool should be taken in consideration and adapt for each and every purpose, as reviewed by (Aagaard and Rossi, 2007).

All together, keeping in mind that ALS is a multifactorial disease,and that miRs are thought to regulate a wide range of metabolic and signaling pathways, manipulating their subcellular levels in neurons, muscles, or glia in a spatial manner should be explored in the future as a potential therapeutic strategy for treatment of ALS.This work was supported by the Israel Science Foundation (grant number 561-11), and the European Research Council (grant number 309377) to EP.

Roy Maimon, Eran Perlson＊

Department of Physiology and Pharmacology, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel

＊Correspondence to:Eran Perlson, PhD, eranpe@post.tau.ac.il.

orcid:0000-0001-6047-9613 (Eran Perlson)

Received:August 20, 2018

Accepted:November 29, 2018

doi:10.4103/1673-5374.250571

Copyright license agreement:The Copyright License Agreement has been signed by both authors before publication.

Plagiarism check:Checked twice by iThenticate.

Peer review:Externally peer reviewed.

Open access statement:This is an open access journal, and articles are distributed under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 License, which allows others to remix, tweak, and build upon the work non-commercially, as long as appropriate credit is given and the new creations are licensed under the identical terms.

Open peer reviewer:Carolyn Tallon, Johns Hopkins University, USA.

Additional file:Open peer review report 1.