Abstract
TAR DNA-binding protein 43 (TDP-43) aggregation driven by stress granule (SG) accumulation is a hallmark of amyotrophic lateral sclerosis (ALS). Our work addresses this through two complementary strategies: antisense oligonucleotides (ASOs) for knocking down proteins implicated in the formation of stress granules and RNA aptamers for limiting TDP-43 aggregation.
We identified three proteins whose knockdown reduces SG assembly. For each, we developed, generated, and validated a panel of ASOs through computational modeling and in vitro testing. ASOs targeting DAZAP1 and FAM98A, two critical stress-granule-associated proteins, demonstrated significant protein knockdown and SG suppression. We also developed a machine learning model to predict ASO efficacy for increased wet-lab efficiency.
In parallel, we generated RNA aptamers targeting TDP-43's C-terminal domain through a computational pipeline; top candidates bound to TDP-43 with high specificity in vitro. Our dual-strategy approach offers a layered therapeutic strategy for ALS and establishes a novel framework for future disease research.
Overview
Amyotrophic lateral sclerosis (ALS) is a rapidly progressive and fatal neurodegenerative disease characterized by the selective loss of upper and lower motor neurons. Among the molecular hallmarks of ALS, cytoplasmic mislocalization and aggregation of TAR DNA-binding protein 43 (TDP-43) is the most prominent, occurring in ~97% of patient cases. [2] TDP-43 is ubiquitously expressed with functions in pre-mRNA splicing, RNA transport, and transcript stability. ASOs provide an upstream strategy by reducing the expression of transcripts that encode proteins like TDP-43, limiting the pool of proteins available for aggregation. By decreasing pathogenic RNA at its source, ASOs can help lower the burden of toxic protein accumulation before stress granule pathology develops.
Key Insight: Under physiological stress, TDP-43 partitions into stress granules. In ALS, chronic or repeated stress promotes the formation of stress granules. Normally, liquid-liquid phase separation allows TDP-43 to transition into liquid-like condensates, preventing it from forming toxic aggregates. However, during pathological demixing, stress granules undergo liquid-to-solid transitions to create insoluble fibrils. [2]
Mislocalized cytoplasmic TDP-43 consequently accumulates, leading to the formation of insoluble aggregates, which is hypothesized to cause a loss of function, resulting in a failure to execute nuclear RNA regulatory roles. The strong association between TDP-43 pathology and ALS makes TDP-43 a candidate biomarker in ALS research. Phase transitions of TDP-43 within stress granules are recognized as a driving force behind ALS, but there are no molecular tools to prevent the demixing that initiates this process. Under chronic stress, stress granules demix, physically separating into compartments, which promotes transition of TDP-43 from liquid to solid state. This solidification of TDP-43 is a marker of aggregation observed in ALS.
Problem and Inspiration
Problem
Although stress granule dynamics and TDP-43 aggregations are increasingly recognized as key drivers of neurodegeneration, there is no current therapeutic platform or tool to modulate demixing and prevent TDP-43 transition into irreversible aggregates. No current therapy modulates early-stage RNA expression to stop toxic proteins from forming before stress granule demixing. Although ASOs are promising, their delivery, specificity, and ability to affect late-stage aggregation remain challenges. Additionally, there are no molecular tools available to directly prevent TDP-43 from undergoing liquid-to-solid transitions inside stress granules. Current interventions cannot modulate demixing or reverse existing pathological condensation. There is a critical need for molecular strategies that can intervene in early-stage stress granule dynamics before pathological TDP-43 condensation occurs, which could also provide a foundation for therapeutic innovation.
Inspiration
Neurodegenerative diseases like ALS remain incurable largely because of the lack of tools capable of targeting disease-driving protein aggregation at early stages. Recent advances in RNA-based therapeutics have inspired new methods to regulate stress granule behavior. The ability of RNA aptamers to bind to proteins with high levels of specificity presents a promising route to intercept TDP-43 before it undergoes pathological condensation. Importantly, aptamers act post-translationally and are reversible, providing the temporal and mechanistic precision to stop TDP-43 condensation at the protein level.
Inspired by these recent advances in RNA-based therapeutics, our team sought to leverage RNA aptamers to directly modulate TDP-43 phase behavior inside stress granules. By designing a novel aptamer to bind to TDP43, we aimed to block protein-protein interactions that cause aggregation and preserve its functional solubility. This mitigation of downstream neurotoxicity highlights aptamers' potential for future therapeutic development.
Similarly, antisense oligonucleotides (ASOs) offer an upstream, gene-targeted approach to mitigate protein aggregation in ALS. Building on preclinical successes showing that ASOs can reduce the expression of pathogenic RNA-binding proteins such as TDP-43 or SOD1, our team drew inspiration from RNA-based therapeutics and AAV delivery strategies to explore ASOs as a complementary approach.
By designing ASOs that specifically reduce toxic RNA transcripts before pathological protein aggregation occurs, we aim to intervene at the earliest stages of disease progression. ASOs offer the potential for precise modulation of gene expression, which, combined with optimized delivery systems, could prevent the formation of cytoplasmic aggregates, stress granules, and provide a lasting therapeutic effect.
Current Solutions
Antisense Oligonucleotides (ASOs)
▼Antisense oligonucleotides (ASOs) are being developed to reduce the expression of specific RNA-binding proteins implicated in ALS, such as TDP-43 or SOD1. While they have shown promise in preclinical models, ASOs are often nonspecific, act late in the aggregation process, and may not clear aggregates already present in the cytoplasm. Delivery methods such as adeno-associated viruses (AAVs) can increase efficacy but are difficult to reverse and may trigger innate, cellular, and humoral immune responses [3].
AAV vectors are now a leading platform for CNS gene therapy because many severe neurological diseases are caused by single-gene mutations. Newer capsids, like AAV9, efficiently target neurons and can cross the blood–brain barrier when administered intravenously [3]. Despite these advances, delivering enough vector to the brain without toxicity remains challenging, and strategies such as intraparenchymal or intrathecal injections are under investigation [3].
CRISPR-Cas9 Gene Editing
▼Genome editing tools like CRISPR-Cas9 are also being explored to correct ALS-linked mutations or remove pathogenic gene variants. Preclinical studies have shown that CRISPR-mediated disruption of mutant SOD1 or excision of the toxic hexanucleotide repeat in C9orf72 can reduce toxic protein levels, improve neuronal survival, and delay disease progression in animal models [5]. However, these approaches are permanent and carry risks of unintended edits, and delivery to the CNS without triggering immune reactions remains a major challenge [4][5].
RNA Aptamers
▼Aptamers represent a promising alternative therapeutic strategy due to their high specificity and ability to bind directly to target proteins. Preclinical research suggests that aptamers could interfere with early events in TDP-43 aggregation and potentially prevent phase transitions that lead to cytoplasmic inclusions. While they remain largely experimental in ALS, their precision and adaptability make them an attractive option for addressing gaps in current treatments that ASOs and CRISPR-based approaches may not fully resolve. Aptamers could complement existing strategies by targeting the earliest steps of protein aggregation before irreversible cytoplasmic inclusions form.
Our Solution and Goals
Our solution is two-fold: First, we aim to limit stress granule formation through the development of novel ASOs that target proteins implicated in stress granule proteins (coded by genes DAZAP1, FAM98A, and SND1). Next, we strive to bolster this effect by reducing TDP-43 aggregation through the use of custom RNA aptamers designed to target the C-terminal domain of TDP-43. The C-terminal domain of TDP-43 is the most disordered section of the protein, thus allowing it the most binding space for the RNA aptamer.
ASO Design
Generate ASOs targeting stress granule proteins
In Vitro Testing
Test knockdown efficiency in SH-SY5Y cells
Aptamer Development
Design aptamers for TDP-43 C-terminal binding
Validation
Confirm binding and therapeutic potential
ASO Solution
In collaboration with our sponsoring lab, we modified an ASO-generation pipeline to generate various ASOs given the target protein for knockdown. Through this pipeline (see Model page), we generated 10 ASOs for testing in vitro. 8 ASOs were tested against stress granule proteins, while two were designated for control testing (to quantify any adverse effects that cells may encounter due to our ASO therapy).
We tested ASOs by transfecting SH-SY5Y cells and measuring knockdown of stress granule proteins (DAZAP1, FAM98A, and SND1). We checked for reduced transcript expression of these proteins using RT-qPCR. This approach targets transcript reduction upstream, aiming to minimize protein aggregation potential.
Key Finding: Our experiments found that our ASOs were effective in knocking down DAZAP1 and FAM98A expression within SH-SY5Y cells, a promising indicator of reduced stress granule expression in ALS.
Aptamer Solution
We also generated RNA aptamers: short, sequence-specific RNA molecules to bind to TDP-43 and regulate its behavior inside stress granules to prevent demixing and aggregation. Aptamers function to prevent the TDP-43 from segregating into tightly bound protein clusters that promote solidification to maintain a liquid-like phase. Our long-term goal is to develop RNA-based therapeutic tools to prevent or reverse disease-driving protein aggregation in neurodegenerative disorders.
We used our novel computational pipeline and molecular dynamics simulations (see Model page) to identify RNA aptamer sequences that bind to the TDP-43 C-terminus, and we used machine learning to select the top binding candidates. We aimed to block TDP-43 interactions in stress granules, as previous research suggests TDP-43 interactions in stress granules increase aggregation. The aptamer predicted to bind with the most specificity and effectiveness was selected for testing in vitro.
We validated our novel TDP-43 aptamer using immunofluorescence. To do this, we fixed and stained SH-SY5Y cells to see if the aptamer is able to bind to TDP-43. We compared the aptamer staining to TDP-43 antibody staining as a positive control. Our results showed binding of the aptamer to the TDP-43 protein, though specificity and effect on aggregation is yet to be quantified.
Thus, our novel approach combines ASO transcript knockdown with aptamer-based protein regulation, targeting both the production and pathological behavior of TDP-43.
Implementation
Our primary intervention uses antisense oligonucleotides (ASOs) to reduce the expression of pathogenic RNA-binding proteins like TDP-43 at the transcript level. By lowering the production of toxic proteins before they accumulate, ASOs act upstream to slow or prevent the initial formation of cytoplasmic aggregates. While effective at reducing overall protein burden, ASOs do not fully dissolve existing stress granules or reverse TDP-43 condensates. Therefore, they serve as the first line of defense, reducing the pathological drive while creating an opportunity for a secondary, post-translational intervention.
To complement the ASO strategy, we introduce a synthetic RNA-based aptamer as a secondary or post-translational mechanism. Our synthetic RNA-based circuit integrates real-time sensing of TDP-43 with targeted rescue, allowing aptamers to bind directly to cytoplasmic TDP-43 under pathologic conditions. This prevents further pathological phase transitions, stabilizes functional protein solubility, and minimizes downstream neurotoxicity. Unlike permanent interventions such as AAV delivery or CRISPR editing, aptamers provide temporal control and reversibility. Acting post-translationally, they directly regulate TDP-43 behavior rather than indirectly modulating transcription or translation, offering a precise, two-tiered approach to counter ALS-associated protein aggregation.
References
- Suk, Terry R, Rousseaux, Maxime W.C. "The Role of TDP-43 Mislocalization in Amyotrophic Lateral Sclerosis." Molecular Neurodegeneration vol. 15, article 45. 15 Aug. 2020, https://molecularneurodegeneration.biomedcentral.com/articles/10.1186/s13024-020-00397-1
- Song, Jianxing. "Molecular Mechanisms of Phase Separation and Amyloidosis of ALS/FTD-linked FUS and TDP-43." Aging and Disease vol. 15, issue 5, pg. 2084-2112. 18, Nov. 2023, https://www.aginganddisease.org/EN/10.14336/AD.2023.1118
- Daci, R., & Flotte, T. R. (2024). Delivery of Adeno-Associated Virus Vectors to the Central Nervous System for Correction of Single Gene Disorders. International Journal of Molecular Sciences, 25(2), 1050. https://doi.org/10.3390/ijms25021050
- Salomonsson, S. E., & Clelland, C. D. (2024). Building CRISPR Gene Therapies for the Central Nervous System: A Review. JAMA neurology, 81(3), 283–290. https://doi.org/10.1001/jamaneurol.2023.4983
- Shi, Y., Zhao, Y., Lu, L., Gao, Q., Yu, D., & Sun, M. (2023). CRISPR/Cas9: implication for modeling and therapy of amyotrophic lateral sclerosis. Frontiers in neuroscience, 17, 1223777. https://doi.org/10.3389/fnins.2023.1223777