RNA-Based Therapies for Neurodegenerative Diseases

Abstract

Most neurodegenerative disorders afflict the ageing population and are often incurable. Therefore, therapeutic development is a major focus in biomedical research. We highlight a new class of drugs, RNA molecules, to control gene expression and decrease neurotoxicity. Their efficacy is shown in pre-clinical studies, clinical trials and in cases of approved patient treatment. As the number of RNA-based strategies increases, so does the promise of targeting more disease-associated genes through a variety of different mechanisms.

Introduction

Neurodegenerative diseases encompass a heterogenous group of diseases that involve progressive degeneration of the nervous system. They include more common forms such as Alzheimer’s disease (AD) and Parkinson’s disease (PD), as well as less common forms including Huntington’s disease (HD), frontotemporal dementia (FTD), amyotrophic lateral sclerosis (ALS), spinal muscular atrophy (SMA), and others. There are currently few therapeutic options for these neurological disorders and most are incurable. A common, but not universal feature, is aggregation of specific proteins including tau (in AD and FTD), amyloid beta (AD), α-synuclein (PD), huntingtin (HD), and TDP-43 (FTD and ALS). For many of these neurodegenerative diseases, inherited forms exist and multiple gene mutations have been identified. Generally, these mutations result in either gain-of-toxic function or loss-of-function.

At the cellular level, gain-of-toxic function mutations lead to production of aberrant mutant proteins (such as huntingtin with expanded polyglutamine repeats in HD, SOD1 point mutations in ALS, and C9ORF72 with repeat expansions in ALS and FTD), which have detrimental effects. On the other hand, loss-of-function mutations can lead to several different scenarios. Fundamentally, they can result in inadequate/insufficient production of the protein encoded by the mutated gene (such as progranulin or C9ORF72 in FTD). Loss-of-function mutations can also alter protein homeostasis, including protein processing (such as amyloid beta in AD), protein degradation (through autophagy and the ubiquitin proteasome system), and protein trafficking (such as nucleocytoplasmic transport and endolysosomal transport in ALS and FTD). Additionally, loss-of-function mutations can result in changes in RNA metabolism, as is case when the functional levels of the RNA binding protein TDP-43 are reduced in ALS and FTD.

RNA-Based Therapies for Neurodegenerative Diseases

RNA-based therapies present as attractive therapeutic approaches for a number of these neurodegenerative diseases because they enable highly specific control of gene expression – either to reduce levels of target proteins in situations of gain-of-toxic function, or to increase levels of target proteins in situations of loss-of-function mutations. For RNA therapies, RNA molecules are designed to bind specifically to their target RNAs through Watson-Crick base pairing. In this way, RNA therapies can be used to precisely target the molecular causes of diseases and/or key pathogenic mechanisms. A major advantage of RNA-based therapy is that it can directly target any desired RNA, and thus it greatly expands the realm of possible targets beyond traditional druggable targets that are accessible to small molecule-and antibody-based approaches (Figure 1).

Figure 1.

Overview of Different Mechanisms of Action of Different RNA Therapeutics.

(1) Without therapeutic RNA molecules, the translation of a pathogenic protein proceeds without inhibition (shown in the broken line box). (2) ASOs hybridize to the target mRNA, while the (3) siRNA/miRNA mimics utilize the RISC in the RNAi pathway to (4) inhibit translation of target mRNA. (5) Overexpression of a therapeutic protein that counteracts the function of the pathogenic protein can be done by delivering the mRNA of the therapeutic protein. (6) saRNA can be delivered to the cell where it binds to AGO2, is imported to the nucleus, and in turn activates an endogenous gene. (7) A more permanent approach to remove the pathogenic protein is by gene knockout using Cas9 and sgRNA RNPs.

Abbreviations: AGO2, argonaute 2; ASO, antisense oligonucleotide; RISC, RNA-induced silencing complex; RNP, ribonucleoprotein; saRNA, small activating RNA.

Source: Trends in Pharmacological Sciences

RNA therapies largely fall into two classes – antisense oligonucleotides (ASOs) and RNA interference (RNAi). ASOs are single-stranded RNAs that work by specifically binding to and modulating their target RNAs for translation or turnover. The use of ASOs to lower target protein levels was first reported in 1978 by Mary Stephenson and Paul Zamecnik,¹ but ASOs subsequently faced considerable challenges in the 1980s. Following major advances in ASO chemistry and delivery, ASOs made a strong resurgence in the 1990s. There is currently high enthusiasm for ASOs for neurodegenerative diseases, with the recent FDA approval of nusinersen (Spinraza) for SMA in 2016 and promising results in ongoing clinical trials with ASOs targeting huntingtin and SOD1.² In a more limited way, RNAi uses double stranded RNA molecules for gene silencing. RNAi was first reported in 1998 by Andrew Fire et al.,³ who later shared the Nobel prize in 2006 with Craig Mello for this discovery. Two RNAi-based therapies – patisiran and inotersen – received FDA approval in 2018 for hereditary transthyretin-mediated amyloidosis.

Antisense drugs have been in used to study neurological disorders for more than two decades and improvements in the chemistry have extended the lifetime of these reagents. DNA and specially RNA are highly unstable due to degradation by nucleases present in all organisms. The efficacy of RNA oligonucleotides as drugs is made possible by chemical modifications to increase their stability, improve RNA binding activity and enhance distribution. To protect RNAs from nucleases, modifications are commonly designed to the phosphate backbone and ribose sugar. These and additional alterations have also been applied to increase RNA binding affinity and reduce toxicity.^4,5 The most common modification in therapeutic ASOs has been introduction of a phosphorothioate backbone. Sugar modifications are included at the ends of the oligonucleotides as an additional way of preventing nuclease degradation and enhancing RNA binding. The most common substitutions are to the naturally occurring 2’-O-methyl and the synthetic 2’-O-methoxyethyl (2’-MOE). These types of modifications have also been introduced for the drugs tested in clinical trials and those approved for patient use.

For neurological disorders, ASOs are commonly delivered directly into the cerebral spinal fluid (CSF). Clinically, ASOs are injected intrathecally to the spinal canal to reach the CSF. In pre-clinical animal models, such as rodents, ASOs are typically administered by direct intracerebroventricular injection. These forms of delivery result in widespread distribution of the molecules throughout the central nervous system (CNS).^6–10 Examination of ASO distribution in nonhuman primates via CSF shows distribution throughout the spinal cord and cortex¹¹ and deeper brain regions, such as the hippocampus, pons and amygdala.^9,11,12

RNA-based therapies inherently have great potential as a form of personalized medicine. RNA therapies are versatile and can be designed to target specific genes, and potentially even to target specific patient mutations. In the case of ASOs, the most commonly applied approach is to design ASOs to downregulate protein levels (such as for SOD1, huntingtin, and tau). This type of mechanism is particularly useful for targeting mutant proteins with gain-of-toxic function and key cellular pathways/ proteins involved in disease pathogenesis. ASOs can also be designed to modulate splicing of the target RNA (such as for the SMN2 RNA by nusinersen) to effectively increase the levels of functional SMN protein. Additionally, ASOs can be designed to block microRNAs and degradation of mutant mRNAs through the nonsense-mediated mRNA decay pathway. Below, we present these different ASO-based approaches in more detail in the context of specific neurodegenerative diseases.

ASOs Downregulate Protein Expression and Reduce Disease Toxicity

One of the most common uses for ASOs has been to lower the expression of proteins to reduce the levels of toxic aggregates. This is achieved by inducing targeted degradation of mRNA transcripts, such as in the case of SOD1, huntingtin, and tau. The oligonucleotides are designed to pair with the coding mRNA transcripts and trigger degradation by the enzyme ribonuclease (RNase) H1, which specifically degrades RNA:DNA hybrids. In this case, ASOs are designed with five to seven central DNA nucleotides. This approach has been successfully used to target superoxide dismutase 1 (SOD1), the first gene identified in connection with inherited ALS. Mutations in SOD1 account for 20% of familial ALS cases.¹³ The mechanism by which SOD1 mutations cause disease is unknown, but most evidence suggests a gain-of-toxic function rather than loss of function. ASOs designed to downregulate SOD1 levels in preclinical research proved successful and moved on to testing in clinical trials. This was the first use of ASOs for neurodegenerative disease. The drug, Tofersen, is now in phase III clinical trials.^12,14 The collection of these studies showed efficacy in distribution of ASOs in the CNS using targeted therapy that proved safe and correlated with a measurable decrease in SOD1 levels and a reduction in disease in patients.

Based on the promising results for SOD1 in preclinical trials and in patient trials, the use of ASOs has been extended for another gene linked to familial and sporadic cases of ALS and frontotemporal dementia. The repeat expansion in the first intron of the gene C9ORF72 is linked to ALS and FTD (C9-ALS/FTD), resulting in a combination of gain and loss of function mechanisms.^15,16 This mutation is the most common genetic link in inherited ALS and FTD (approximately 40% of familial and 8–10% of sporadic ALS cases). Carriers of this repeat expansion accumulate RNA foci of the repeat transcript along with an increase in dipeptide repeats translated from the C9ORF72 mutation via repeat associated non-canonical (RAN) translation.¹⁷ In addition, the repeat expansion causes haploinsufficiency as seen by reduced C9ORF72 protein levels in patients. The main pathogenic process contributing to disease in C9-ALS/ FTD remains under debate, and it is likely that all these defects synergize during disease. ASOs targeting the pathological C9ORF72 repeat expansion promote degradation of the noncoding region and reduction of RNA foci in patient-derived fibroblast and induced pluripotent stem neurons. Treatment of C9-ALS/ FTD mouse models also showed a decrease in the toxic polypeptides derived from the repeat expansion and improved behavioral phenotype.¹⁸ Importantly, an ASO with these characteristics does not affect C9ORF72 protein synthesis from the unaffected allele and is currently in patient clinical trials. The success of ASOs directed to SOD1 and C9ORF72 show the promise of ASO therapy for ALS and FTD, which may be applied to other targets associated with sporadic disease cases, which account for the majority of ALS and FTD patients.

Several laboratories within the Henry and Amelia Nasrallah Center for Neuroscience are currently carrying out research relating to RNA-based therapies in the context of different neurodegenerative diseases. Susan Farr’s group has focused on using ASOs to target key proteins implicated in the pathogenesis of AD. Their studies have demonstrated that ASOs that decrease levels of amyloid precursor protein (APP), glycogen synthase kinase (GSK)-3β, and presenilin-1 improve learning and memory in mouse models of AD.^19–24 More recently, her laboratory is studying traumatic brain injury (TBI) and has found that post-injury administration of ASOs targeting GSK-3β prevents cognitive deficits.²⁵ These results strongly suggest that ASOs targeting these proteins could have similar beneficial cognitive effects in people with AD and TBI.

ASOs that Restore Gene Runction by Modulating mRNA Splicing

One of the most exciting developments in the ASO field has been the development of the RNA molecule nusinersen (Spinraza) to treat spinal muscular atrophy (SMA), which is the first FDA-approved drug for this disorder. SMA is a genetic disease involving death of motor neurons that had been the most common genetic cause of death in children before development of the drug. Life expectancy in the more severe cases was two years of age.^26,27 This defect is caused by decreased production of survival motor neuron (smn) protein of which there are two paralogous genes in humans, SMN1 and SMN2. The two genes are nearly identical; however, SMN2 is expressed at much lower levels than SMN1 because of a single site substitution in exon 7 in SMN2, which destroys an exonic splicing enhancer (ESE) sequence. Exon 7 skipping results in a truncated mRNA transcript and protein, which becomes degraded and generates loss of function. This results in 80% reduction in smn levels produced from SMN2. The downstream intronic region in SMN1 and SMN2 contains an intronic splicing silencer (ISS) sequence that binds the trans regulatory proteins hnRNP A1 and A2. Therefore, inhibition by ISS is counteracted by the ESE element in SMN1, resulting in sufficient inclusion/ splicing of exon 7 and production of smn protein. SMA results from mutations in SMN1 that disrupt smn protein levels. Starting in the 1990s, the team of Dr. Adrian Krainer at Cold Spring Harbor Laboratories investigated the possibility of manipulating SMN2 splicing to enhance exon 7 inclusion and compensate for the loss of SMN1. The team tested the capability of ASOs to bind the SMN2 transcript and increase exon 7 inclusion. Their results showed that binding to the ISS was able to significantly enhance exon 7 inclusion by preventing the binding of hnRNP A1 and A2.^28,29 The research carried out by Krainer and others determined the molecular mechanisms of exon 7 splicing regulation using splicing reporters and other in vitro tools followed by the characterization of SMA mouse models to study the process in vivo. The ASO molecules designed based on the collection of the research worked in all pre-clinical trials in cells and in animal models. These also proved successful in clinical trials and warranted FDA approval. The use of nusinersen has already made a difference in the lives of many patients and their families. Similar methods have been used to control RNA splicing by ASOs for the treatment of Duchenne muscular dystrophy. In this case, the drugs promote exon skipping to restore the correct reading frame and generate the functional protein dystrophin, albeit smaller than normal.³⁰ The ASOs are approved for patients who benefit from having restored functional dystrophin at significantly levels. The results of these novel drugs show that ASOs may be used to modify RNA processing, not only to modulate transcript degradation, and that their use may be safe in patients.

RNA-Based Approach to Prevent Protein Aggregation and Restore Protein Function

One innovative proposal for the use of RNA for therapy is in the modulation of protein stability. This is based on recent observations that RNA is very effective at maintaining RNA binding proteins in a soluble state, preventing them from misfolding and aggregation. Most neurodegenerative disease is characterized by the accumulation of protein aggregates, which may result in gain-of-toxic function. In the case of ALS and frontotemporal lobar degeneration with ubiquitin-positive inclusions (FTLD-U), the characteristic aggregates consist of the RNA binding protein TDP-43 (TAR DNA binding protein).^31,32 TDP-43 pathology is observed in other neurodegenerative diseases, including approximately half of Alzheimer’s disease (AD) cases.^33–36 The prevalence of TDP-43 pathology is even more widespread in Limbic-predominant Age-related TDP-43 Encephalopathy (LATE), an AD-like disorder affecting the oldest population.³⁷ The pathogenesis of ALS and FTD is now widely viewed to be caused by dysfunction of RNA binding proteins and defects in RNA processing mechanisms. RNA binding proteins, like TDP-43, are extremely prone to aggregation. TDP-43 regulates RNA processing and thus controls the expression of hundreds of genes. Aggregation is linked to loss of protein function, which is highly deleterious in all cell types and in the model organisms tested. Therefore, approaches that alter TDP-43 levels, either by up or down-regulation, would negatively impact cell survival. Exciting new findings from different groups, including the Ayala lab, show that RNA acts as chaperone to these proteins under physiological conditions. Therefore, the Ayala lab and others aim to exploit RNA-based molecules to test their efficacy in preventing TDP-43 aggregation and restoring its function. If successful, this approach offers the advantage of preventing aggregation without altering normal levels of functional protein. Promising data suggest that RNA-based approaches may hold therapeutic promise for ALS and FTD, disorders associated with RNA-binding protein dysfunction, by specifically modulating protein function and homeostasis.

ASO-Based Approach to Increase Protein Levels

Another innovative use of ASOs that is being pursued in the Henry and Amelia Nasrallah Center for Neuroscience is for increasing protein levels. The Nguyen lab studies a progranulin-deficient form of FTD, in which one GRN (progranulin) allele is mutated and one intact GRN allele remains. The vast majority of disease-causing GRN mutations are nonsense mutations – which introduce a premature termination codon – and these mutant mRNAs are degraded by the nonsense-mediated mRNA decay (NMD) pathway. In 2016, the Krainer lab reported that ASOs can be designed to block NMD in cells.³⁸ Together with Ionis Pharmaceuticals, our lab developed an ASO-based strategy to block degradation of the mutant progranulin mRNA harboring the GRN^R493X mutation, which is the most common GRN mutation found in individuals with FTD. Our cell-based studies had established that the truncated progranulin protein produced from the GRN^R493X mRNA is functional, and thus we reasoned that increasing levels of the mutant protein would restore sufficient levels of functional progranulin protein. While we found that eight ASOs increased progranulin mRNA and protein levels in cells,³⁹ these ASOs did not increase progranulin levels in mice harboring the Grn^R493X mutation.

The Nguyen lab is currently testing a different ASO-based strategy that targets specific microRNA (miR) binding sites within the GRN mRNA. miRs regulate protein levels of about one third of all proteins in humans, and three microRNAs in particular (miR-29b, miR-107, and miR-659) have been reported to lower progranulin protein levels. We designed ASOs to target the binding sites of two of these miRs – with the rationale that ASOs that partially overlap with the miR binding sites would sterically hinder binding of the cognate miR and thereby increase progranulin levels. Our ongoing studies have identified a number of ASOs that effectively increase progranulin protein levels in human cells. We are excited to next test these ASOs in a humanized GRN mouse model. Overall, we are hopeful that ASOs can be used to increase progranulin levels and potentially lead to a treatment for progranulin-deficient FTD.

Conclusion

RNA-based therapies hold great promise for certain neurodegenerative diseases and may offer a tangible path to personalized medicine. These therapeutic efforts are largely informed by genetic studies that identify the genetic bases of disease and by basic science studies that identify specific molecular targets. As we continue to learn more about neurodegenerative diseases we can design and test additional RNA-based therapies.