Researchers found an artificial RNA molecule capable of silencing or turning off a targeted gene that blocks ALS degeneration


Writing in Nature Medicine, an international team headed by researchers at University of California San Diego School of Medicine describe a new way to effectively deliver a gene-silencing vector to adult amyotrophic lateral sclerosis (ALS) mice, resulting in long-term suppression of the degenerative motor neuron disorder if treatment vector is delivered prior to disease onset, and blockage of disease progression in adult animals if treatment is initiated when symptoms have already appeared.

The findings are published in the December 23, 2019 online issue of the journal Nature Medicine.

Martin Marsala, MD, professor in the Department of Anesthesiology at UC San Diego School of Medicine and a member of the Sanford Consortium for Regenerative Medicine, is senior author of the study.

ALS is a neurodegenerative disease that affects nerve cells in the brain and spinal cord. Motor neurons responsible for communicating movement are specifically harmed, with subsequent, progressive loss of muscle control affecting the ability to speak, eat, move and breathe.

More than 5,000 Americans are diagnosed with ALS each year, with an estimated 30,000 persons currently living with the disease. While there are symptomatic treatments for ALS, there is currently no cure. The majority of patients succumb to the disease two to five years after diagnosis.

There are two types of ALS, sporadic and familial. Sporadic is the most common form, accounting for 90 to 95 percent of all cases. It may affect anyone. Familial ALS accounts for 5 to 10 percent of all cases in the United States, and is inherited. Previous studies show that at least 200 mutations of a gene called SOD1 are linked to ALS.

The SOD1 gene normally serves to provide instructions for making an enzyme called superoxide dismutase, which is widely used to break down superoxide radicals – toxic oxygen molecules produced as a byproduct of normal cell processes.

Previous research has suggested that SOD1 gene mutations may result in ineffective removal of superoxide radicals or create other toxicities that cause motor neuron cell death, resulting in ALS.

The new approach involves injecting shRNA – an artificial RNA molecule capable of silencing or turning off a targeted gene – that is delivered to cells via a harmless adeno-associated virus.

In the new research, single injections of the shRNA-carrying virus were placed at two sites in the spinal cord of adult mice expressing an ALS-causing mutation of the SOD1 gene, either just before disease onset or when the animals had begun showing symptoms.

Earlier efforts elsewhere had involved introducing the silencing vector intravenously or into cerebrospinal fluid in early symptomatic mice, but disease progression, while delayed, continued and the mice soon died.

In the new study, the single subpial injection (delivered below the pia matter, the delicate innermost membrane enveloping the brain and spinal cord) markedly mitigated neurodegeneration in pre-symptomatic mice, which displayed normal neurological function with no detectable disease onset.

The functional effect corresponded with near-complete protection of motor neurons and other cells, including the junctions between neurons and muscle fibers.

In adult mice already displaying ALS-like symptoms, the injection effectively blocked further disease progression and degeneration of motor neurons.

In both approaches, the affected mice lived without negative side effects for the length of the study.

“At present, this therapeutic approach provides the most potent therapy ever demonstrated in mouse models of mutated SOD1 gene-linked ALS,” said senior author Martin Marsala, MD, professor in the Department of Anesthesiology at UC San Diego School of Medicine.

“In addition, effective spinal cord delivery of AAV9 vector in adult animals suggests that the use of this new delivery method will likely be effective in treatment of other hereditary forms of ALS or other spinal neurodegenerative disorders that require spinal parenchymal delivery of therapeutic gene(s) or mutated-gene silencing machinery, such as in C9orf72 gene mutation-linked ALS or in some forms of lysosomal storage disease.”

The research team also tested the injection approach in adult pigs, whose spinal cord dimensions are similar to humans, for safety and efficacy. Using an injection device developed for use in adult humans, they found the procedure could be performed reliably and without surgical complications.

Marsala said next steps involve additional safety studies with a large animal model to determine the optimal, safe dosage of treatment vector.

“While no detectable side effects related to treatment were seen in mice more than one year after treatment, the definition of safety in large animal species more similar to humans is a critical step in advancing this treatment approach toward clinical testing.”

Amyotrophic lateral sclerosis (ALS) is a progressive and fatal neurodegenerative disorder of motor function.

It is characterized by the selective degeneration of the lower and upper motor neurons. Among the symptoms of this disease are progressive muscle weakness and paralysis, swallowing difficulties and breathing impairment due to respiratory muscle weakness that ultimately causes death, usually within 2–5 years following clinical diagnosis (Kiernan et al., 2011).

Though most cases of ALS are sporadic, some families (10%) demonstrate a clinically indistinguishable form of ALS with clear Mendelian inheritance and high penetrance (Pasinelli and Brown, 2006).

Treatments to slow the progression of ALS to date remains riluzole (Bensimon et al., 1994) and edaravone (Abe et al., 2014) but they are only modestly effective. However, in the past couple years, there has been a real encouragement in witnessing potentially efficacious treatments, such as Masitinib and Pimozide (Trias et al., 2016Patten et al., 2017Petrov et al., 2017) claiming to demonstrate clinical benefit.

Furthermore, RNA-targeted therapies are currently intensively being evaluated as potential strategies for treating this ALS (Schoch and Miller, 2017Mathis and Le Masson, 2018).

There is indeed hope to have new and potentially more effective treatment options available for ALS in the near future.

Mutations in over more than 20 genes contribute to the etiology of ALS (Chia et al., 2018) (Table ​(Table1).1). Amongst these genes, the major established causal ALS genes are SOD1 (Cu-Zn superoxide dismutase 1), TARDBP (transactive response DNA Binding protein 43kDa), FUS (fused in sarcoma) and hexanucleotide expansion repeat in Chromosome 9 Open Reading Frame 72 (C9ORF72).

These genetic discoveries have led to the development of animal models (Julien and Kriz, 2006Kabashi et al., 2010Patten et al., 2014Picher-Martel et al., 2016) that permitted the identification of key pathobiological insights.

Currently, RNA dysregulation appears to be a major contributor to ALS pathogenesis.

Indeed, TDP-43 and FUS are deeply involved in RNA processing such as transcription, alternative splicing and microRNA (miRNA) biogenesis (Buratti et al., 20042010Polymenidou et al., 2012).

Mutations in C9ORF72, lead to a toxic mRNA gain of function through RNA foci formation, and the subsequent sequestration in stress granules and altered activity of RNA-binding proteins (Barker et al., 2017).

In addition to the major ALS genes, other ALS genes including ataxin-2 (ATXN2) (Ostrowski et al., 2017), TATA-box binding protein associated factor 15 (TAF15) (Ibrahim et al., 2013), heterogeneous nuclear ribonucleoprotein A1 (hnRNPA1) (Dreyfuss et al., 1993), heterogeneous nuclear ribonucleoprotein A2 B1 (hnRNPA2 B1) (Alarcon et al., 2015), matrin 3 (MATR3) (Coelho et al., 2015), Ewing’s sarcoma breakpoint region 1 (EWSR1) (Duggimpudi et al., 2015), T-cell-restricted intracellular antigen-1 (TIA1) (Forch et al., 2000), senataxin (SETX) and angiogenin (ANG) (Yamasaki et al., 2009), play critical role in RNA processing (Table ​(Table11).

Table 1

ALS genes and their involvement in RNA processing.

GeneProtein encodedRegulation of RNA processing
SOD1Superoxide dismutase 1Yes
TARDBPTar-DNA-binding protein-43Yes
FUSFused in sarcomaYes
TAF15TATA-box binding protein associated factor 15Yes
UBQLN2Ubiquilin 2No
KIF5AKinesin family member 5ANo
hnRNPA1Heterogeneous nuclear ribonucleoprotein A1Yes
hnRNPA2 B1Heterogeneous nuclear ribonucleoprotein A2/B1Yes
MATR3Matrin 3Yes
CHCHD10Coiled-coil-helix-coiled-coil-helix domain containing 10No
EWSR1EWS RNA binding protein 1Yes
TIA1TIA1 cytotoxic granule associated RNA binding proteinYes
CCNFCyclin FNo
NEK1NIMA related kinase 1No
TBK1TANK binding kinase 1No
VCPValosin containing proteinNo
SQSTM1Sequestosome 1No
PFN1Profilin 1No
TUBB4ATubulin beta 4A class IVaNo
CHMP2BCharged multivesicular body protein 2BNo
SPG11Spatacsin vesicle trafficking associatedNo
ALS2Alsin Rho guanine nucleotide exchange factorNo

In this review, we focus on the four major ALS-associated genes (SOD1, TARDBP, FUS, and C9orf72) and present how they play critical roles in various RNA pathways. We particularly highlight recent developments on the dysregulation of RNA pathways (Figure ​(Figure1)1) as a major contributor to ALS pathogenesis and discuss the potential of RNA-targeted therapies for ALS.

An external file that holds a picture, illustration, etc.
Object name is fgene-09-00712-g001.jpg
RNA dysfunction in amyotrophic lateral sclerosis (ALS). Major ALS mutations may disrupt RNA processing by several mechanisms. For instance, (A) mutations in ALS genes SOD1, TDP-43, FUS and C9orf72 can alter gene expression. (B) The RNA binding proteins TDP-43 and FUS can affect global splicing machinery. Dipeptide repeat proteins from C9orf72 intronic expansion can also alter splicing patterns of specific RNAs. (C) TDP-43, FUS, and dipeptide proteins can also promote microRNA biogenesis as components of the Drosha and Dicer complexes. TDP-43 and FUS also alter mRNA transport (D) and local translation (E)(F) TDP-43 and FUS predominantly reside in the nucleus, but when mutated they are can mislocalization to the cytoplasm where they bind and regulate different sets of RNAs including the export and mislocalization of other transcripts to the cytoplasm. Poly-PR dipeptide can also bind nuclear pores channels blocking the import and export of molecules.

Superoxide Dismutase-1 (SOD1)

Unlike TDP43 and FUS, SOD1 does not contain RNA-binding motifs, however, several reports have demonstrated a potential role of mutant SOD1 in regulating RNA metabolism (Menzies et al., 2002Lu et al., 2007Lu L. et al., 2009Chen et al., 2014).

Particularly, mutant SOD1 can bind mRNA species such as vascular endothelial growth factor (VEGF) and NFL and negatively affects their expression, stabilization and function (Menzies et al., 2002Lu et al., 2007Lu L. et al., 2009Chen et al., 2014). More precisely, mutant SOD1 can directly bind to specific adenylate- and uridylate-rich stability elements (AREs) located in the 3′ UTR of transcripts of VEGF (Lu et al., 2007) and NFL (Chen et al., 2014). It is believed that such a gain of abnormal protein–RNA interactions can be caused by SOD1 misfolding that results in the exposure of polypeptide portions with the ability to bind nucleic acids (Kenan et al., 1991Tiwari et al., 2005).

Binding of mutant SOD1 to the 3′ UTR of the VEGF mRNA results in the sequestration of other ribonucleoproteins such as TIAR and HuR into insoluble aggregates. These interactions, which are specific to mutant SOD1, result in decline levels of VEGF mRNA, impairment of HuR function and ultimately hampering their neuroprotective actions during stress responses (Lu et al., 2007Lu L. et al., 2009).

In motor neuron-like NSC34 cell lines expressing mutant SOD1 (G37R or G93A), the level of NFL mRNA is significantly reduced (Menzies et al., 2002). Reduction in NFL mRNA levels has also been reported in G93A transgenic mice and human spinal motor neurons from SOD1-ALS cases (Menzies et al., 2002).

It is proposed that destabilization NFL mRNA by mutant SOD1, result to altered stoichiometry of neurofilament (NF) subunits and subsequent NF aggregation in motor neurons (Chen et al., 2014). NF inclusion in the soma and proximal axons of spinal motor neurons is a hallmark of ALS pathology (Hirano et al., 1984).

In IPSC-derived model of ALS, a reduction of NFL mRNA level has been reported to result in NF aggregation and neurite degeneration (Chen et al., 2014). Altogether, these studies support a pathogenic role for dysregulation of RNA processing in SOD1-related ALS.

Interestingly, SOD1 has been shown to interact with TDP-43 to modulate NFL mRNA stability (Volkening et al., 2009). As mentioned above, TDP-43 was found to directly interact with the 3′ UTR of NFL mRNA to stabilize it (Strong et al., 2007).

Altogether, these studies suggest that SOD1 and TDP-43 may act in a possible common action in regulating specific RNA stability. In the case of NFL mRNA, it would be interesting to investigate whether mutant SOD1 dislodges TDP-43 from the NFL mRNA in a manner that would affect its mRNA metabolism and potentially making NF prone to form aggregates.

Furthermore, there have been several transcriptome investigations in SOD1 human samples (D’Erchia et al., 2017), motor neuron-like NSC34 cell culture model (Kirby et al., 2005) and transgenic animals including mice (Lincecum et al., 2010Bandyopadhyay et al., 2013Sun et al., 2015), rat (Hedlund et al., 2010) and drosophila (Kumimoto et al., 2013).

These studies have reported dysregulation of genes involved in pathways related to the neuroinflammatory and immune response, oxidative stress, mitochondria, lipid metabolism, synapse and neurodevelopment (Hedlund et al., 2010Lincecum et al., 2010Bandyopadhyay et al., 2013Kumimoto et al., 2013Sun et al., 2015D’Erchia et al., 2017).

However, in these studies it is not clear whether SOD1 directly or indirectly impact the regulation of the differentially expressed genes. In a recent elegant study, Rotem et al. (2017), compared transcriptome changes in SOD1 and TDP-43 models.

They found that most genes that were altered in the SOD1G93A model were not dysregulated in the TDP-43A315T model, and vice versa (Rotem et al., 2017). There were, however, a few genes whose expressions were altered in both ALS models (Rotem et al., 2017). These findings are consistent with the ALS pathology, which is distinguishable between the ALS-related SOD1 phenotype and the TDP-43 phenotype. Although different cellular pathways are likely activated by SOD1 versus TDP-43, it is very plausible that they ultimately convergence onto common targets to result in similar motor neuron toxicity and ALS phenotype.

RNA-Targeted Therapeutics for ALS

Our understanding of RNA biology has expanded tremendously over the past decades, resulting in new approaches to engage RNA as a therapeutic target.

More precisely, RNA-targeted therapeutics have been developed to mediate the reduction or expression of a given target RNA by employing mechanisms such as RNA cleaving, modulation of RNA splicing, inhibition of mRNA translation into protein, inhibition of miRNA binding sites, increasing translation by targeting upstream open reading frames and disruption of RNA structures regulating RNA stability (Robertson et al., 2010Fellmann and Lowe, 2014Vickers and Crooke, 2014Havens and Hastings, 2016Liang et al., 2016).

Therapeutics that directly target RNAs are promising for a broad spectrum of disorders, including the neurodegenerative diseases (Scoles and Pulst, 2018) and are currently under evaluation as potential strategies for treating ALS.

The RNA therapeutics approaches include RNA interference (RNAi) and ASOs (Figure ​(Figure2),2), both bind to their target nucleic acid via Watson-Crick base pairing and cause degradation of or inactivate the targeted mRNA (Burnett and Rossi, 2012).

Recently, application of innovative drug discovery approaches has showed that targeting RNA with bioactive small molecules is achievable (Disney, 2013Bernat and Disney, 2015).

A few researchers including us are currently exploiting such a new type of RNA-targeted therapeutics to search for RNA-targeted small molecules as C9orf72 ALS therapeutics.

An external file that holds a picture, illustration, etc.
Object name is fgene-09-00712-g002.jpg
RNA-based therapy approaches for potentially treating ALS. (A) SiRNAs operate through RNA interference pathway. After strand unwinding, one siRNA strand binds argonaute proteins as part of the RNA-induced silencing complex (RISC) and is recruited to a target mRNA which is then cleaved. Virus can provide a means of shRNA, which will be cleaved once in the cytoplasm by dicer enzyme into siRNA. This approach has been evaluated to reduce the level of mutant SOD1 protein. (B) Antisense oligonucleotide (ASO) binds to targeted mRNA and induces its degradation by endogenous RNase H or blocks the mRNA translation. This strategy is being exploited as a potential therapeutic avenue in ALS aiming principally to reduce the protein level of SOD1 protein or by targeting of C9orf72 RNA foci. (C) Small molecules can be designed to target and stabilize RNA structures. This approach was particularly tested to stabilize G-quadruplex of C9orf72 GGGGCC repeat RNA. Stabilization of G-quadruplex structure reduces RNA foci formation and blocks repeat translation.

RNA Interference (RNAi)

RNAi is an endogenous cellular mechanism to regulate mRNA. It operates sequence specifically and post-transcriptionally via the RISC (Carthew and Sontheimer, 2009). Methods of mediating the RNAi effects are via small interfering RNA (siRNA), short hairpin RNA (shRNA), and artificial miRNA (Fire et al., 1998Moore et al., 2010Chakraborty et al., 2017). These approaches can help to reduce the expression of mutant (toxic) gene and can provide significant therapeutic benefit in treating ALS and other neurodegenerative disease implicating aberrant accumulation of misfolded proteins.

The challenge of using siRNA for treating ALS is that it has to be designed to have the specificity and ability to reduce the aberrant mutant protein while leaving wild-type protein intact. Attempts were made to design siRNA, which could recognize just a single nucleotide alternation to selectively suppress mutant SOD1 (particularly G93A) expression leaving wild-type SOD1 intact (Yokota et al., 2004Wang H. et al., 2008).

The design of siRNA G93A.1 and G93A.2 by Yokota et al. (2004) were found to successfully suppress the expression of approximately 90% of mutant SOD1 G93A. Importantly, both siRNA had virtually little or no effect on wild-type SOD1 expression (Yokota et al., 2004).

To achieve long-term expression of siRNA in cells, the use of viral delivery system has proved powerful to provide a continuous delivery and expression of shRNA in sufficient quantities (Bowers et al., 2011). Indeed, diverse viral vectors have been studied such as adeno-associated virus (AAV), lentivirus (LV), and rabies-glycoprotein-pseudotyped lentivirus (RGP-LV) (Raoul et al., 2005Wu et al., 2009).

Recombinant AAVs are currently the choice of RNAi treatment vehicle for neurological diseases because they are non-pathogenic and safe (Maguire et al., 2014Smith and Agbandje-McKenna, 2018). Several studies have aimed at engineering AAV serotypes with better cell-type and tissue specificities and an improved immune-evasion potential (Gao et al., 2005Weinmann and Grimm, 2017).

AAV9 and AAVrh10 serotypes have been shown to cross the blood–brain barrier and efficiently transduce cells in the CNS, with widespread and sustained transgene expression in the spinal cord and brain even after just a single injection (Thomsen et al., 2014Dirren et al., 2015Borel et al., 2016).

Importantly, they can efficiently target neurons and astrocytes, making them the most applicable delivery systems for treating ALS.

Several researchers have independently use siRNA or shRNA to silence mutant SOD1 expression in vitro and in vivo (Miller et al., 2005Raoul et al., 2005Ralph et al., 2005Foust et al., 2013).

Intramuscular delivery of siRNA targeting mutant SOD1 in SOD1G93A mice delays the onset of motor neuron symptoms and extend their survival (Miller et al., 2005). Similarly, SOD1G93A mice treated with injection of AAV encoding shRNA against human SOD1 mRNA (hSOD1) exhibited delayed diseases onset and significantly increased their survival by 23% (Foust et al., 2013).

The same group later demonstrated the efficacy of this approach in SOD1G93A rats, showing that silencing of hSOD1 expression selectively in the motor cortex also delayed disease onset and prolonged survival (Thomsen et al., 2014). Silencing of SOD1 using an artificial miRNA (miR-SOD1) systemically delivered using the viral vector AAVrh10 in SOD1G93A mice was also found to significantly delayed disease onset, preserved muscle motor functions and extended survival (Borel et al., 2016).

Interestingly, similar findings were observed in non-human primates treated with AAVrh10-miR-SOD1 (Wang et al., 2014Borel et al., 2016). These findings suggest that miRNA silencing strategy warrants further investigations and may offer promise for the development for the treatment of SOD1-related ALS.

Study safely delivers RNAi-based gene therapy for ALS in animal model
More information: Spinal subpial delivery of AAV9 enables widespread gene silencing and blocks motoneuron degeneration in ALS, Nature Medicine (2019).  DOI: 10.1038/s41591-019-0674-1 ,


Please enter your comment!
Please enter your name here

Questo sito usa Akismet per ridurre lo spam. Scopri come i tuoi dati vengono elaborati.