Antisense Oligonucleotides: Translation from Mouse Models to Human Neurodegenerative Diseases

Abstract

Multiple neurodegenerative diseases are characterized by single-protein dysfunction and aggregation. Treatment strategies for these diseases have often targeted downstream pathways to ameliorate consequences of protein dysfunction; however, targeting the source of that dysfunction, the affected protein itself, seems most judicious to achieve a highly effective therapeutic outcome. Antisense oligonucleotides (ASOs) are small sequences of DNA able to target RNA transcripts, resulting in reduced or modified protein expression. ASOs are ideal candidates for the treatment of neurodegenerative diseases, given numerous advancements made to their chemical modifications and delivery methods. Successes achieved in both animal models and human clinical trials have proven ASOs both safe and effective. With proper considerations in mind regarding the human applicability of ASOs, we anticipate ongoing in vivo research and clinical trial development of ASOs for the treatment of neurodegenerative diseases.

Keywords: antisense oligonucleotides; clinical trial; in vivo models; neurodegeneration; therapy.

Figure 1. Antisense oligonucleotide (ASO) chemical modifications, properties, and common uses

Decades of synthetic chemical research have produced various ASO structural designs to enhance the pharmacokinetic properties, target affinity, and tolerability profile for ASO application. Common modifications have been made to the phosphodiester backbone to create phosphorothioate (PS) DNA, morpholino, and peptide nucleic acid (PNA) designs, which all confer excellent nuclease resistance for more potent ASO activity. Notably, PS designs exhibit stability, high binding affinity to proteins enabling efficient uptake into cells, and support RNase H-mediated cleavage of the target. Therefore, PS modifications are broadly employed in preclinical studies and are the primary design for human clinical trials for spinal muscular atrophy (SMA), amyotrophic lateral sclerosis (ALS) and Huntington's disease (HD). In combination, 2′ ribose substitutions – 2′-O-methyl (2′-O-Me), 2′-O-methoxyethyl (2′-MOE), and locked nucleic acid (LNA) – greatly enhance target binding, increase resistance to degradation by nucleases, and generally confer less toxicity than unmodified designs. Uniform 2′ modifications support non-degrading mechanisms of action, including splice site modification and translational inhibition; however, introduction of a “gapmer” strategy will permit RNase H-mediated activity. Both 2′-O-methyl and 2′-O-methoxyethyl modifications are commonly used in preclinical research and clinical trials.

Figure 2. Common antisense oligonucleotide (ASOs) mechanisms of action

When administered, ASOs bind to target RNA with base pair complementarity and exert various effects based on the ASO chemical structure and design. Three mechanisms, commonly employed in preclinical models of neurodegenerative disease and human clinical trial development, are shown. These mechanisms include mRNA target degradation via recruitment of the RNase H enzyme, alternative splicing modification to include or exclude exons, and miRNA inhibition to inhibit miRNA binding to its target mRNA.

Figure 3. Progression of select antisense oligonucleotide (ASO) candidates for the treatment of neurodegenerative diseases

Successful translation of ASOs from preclinical rodent models to human clinical trials is evident particularly for the treatment of neurodegenerative diseases spinal muscular atrophy targeting the survival motor neuron 2 (SMN2) gene and familial amyotrophic lateral sclerosis targeting the superoxide dismutase 1 (SOD1) gene. The C9orf72 gene is an additional target in amyotrophic lateral sclerosis with key pathological improvements identified in patient iPSC-derived neurons and novel mouse models. ASOs targeting the huntingtin gene demonstrate notable efficacy in Huntington's disease models and are in human clinical trial in Canada and Europe. ASOs for application in Alzheimer's disease and tauopathies, most notably those targeting amyloid precursor protein (APP) and tau (MAPT), respectively, show promise in preclinical development. ¹Hua, Y. et al., 2010. Genes Dev 24, 1634-1644. ²Williams, J.H. et al., 2009. J Neurosci 29, 7633-7638. ³Porensky, P.N. et al., 2012. Hum Mol Genet 21, 1625-1638. ⁴Passini, M.A. et al., 2011. Sci Transl Med 3, 72ra18. ⁵Chiriboga, C.A. et al., 2016. Neurology 86, 890-897. ⁶Smith, R.A. et al., 2006. J Clin Invest 116, 2290-2296. ⁷Miller, T.M. et al., 2013. Lancet Neurol 12, 435-442. ⁸Stanek, L.M. et al., 2013. J Huntingtons Dis 2, 217-228. ⁹Kordasiewicz, H.B. et al., 2012. Neuron 74, 1031-1044. ¹⁰Donnelly, C.J. et al., 2013. Neuron 80, 415-428. ¹¹Lagier-Tourenne, C. et al., 2013. Proc Natl Acad Sci USA 110, E4530-4539. ¹²Sareen, D. et al., 2013. Sci Transl Med 5, 208ra149. ¹³O'Rourke, J.G. et al., 2016. Science 351, 1324-1329. ¹⁴Jiang J. et al., 2016. Neuron 90, 535-50. ¹⁵Sud, R. et al., 2014. Mol Ther Nucleic Acids 3, e180. ¹⁶Schoch, K.M. et al., 2016. Neuron 90, 941-947. ¹⁷DeVos, S.L. et al., 2017. Sci Trans Med 9. ¹⁸Kumar, V.B. et al., 2000. Peptides 21, 1769-1775. ¹⁹Erickson, M.A. et al., 2012. J Alzheimers Dis 28, 951-960. ²⁰Farr S.A., et al., 2014. J Alzheimers Dis 40, 1005-1016.