Defining The Immunomodulatory Effects of Gapmer Antisense oligonucleotides (ASOs) on Toll-like receptor (TLR) 7 and TLR8

Abstract

After decades of development, RNA therapeutics are currently beginning to deliver their potential value as shown with the success of COVID-19 mRNA vaccines. However, this is owed to the chemical modifications adopted, including nucleoside and backbone modifications that provide RNA therapeutics with drug-like properties and limit unwanted immune activation resulting from RNA therapeutics engaging nucleic acid receptors. Nevertheless, how these modifications or combinations of modifications avoid nucleic acid recognition is not fully defined. Here, we set out to define the immunomodulatory effects of gapmer antisense oligonucleotides (ASOs), which are composed of a central DNA region flanked by nucleotides ofْmodifiedْchemistry,ْsuchْasْ2’-O-methylْ(2’Ome),ْ2’-Methoxyethyl (2'MOE), and locked nucleic acid (LNA) in combination with phosphorothioate (PS) backbones, on endosomal nucleic acid sensing by Toll-like receptors (TLRs) 7 and 8, which are crucial innate immune sensors in the defense against viral and bacterial infections.

Afterْ screeningْ ofْ ~200ْ 2’OMe,ْ 76ْ 2’MOE,ْ andْ 93ْ LNAْ gapmerْ ASOs,ْ weْ foundْ thatْ TLR7ْ sensing was broadly inhibited independently of the chemistry used, in a sequence-dependent manner. Importantly, through in-depth sequence analyses and targeted sequence mutations, we were able to characterise some of these activities and identify a few non-inhibitory motifs of TLR7 sensing, such as 2'OMe CUU and LNA CC motifs at the 5'end of gapmer ASOs. This suggests that gapmer ASOs can be rationally designed to limit unintended immunosuppressive effects on TLR7, thereby retaining its protective function against pathogens such as SARS-CoV-2. Conversely, we found that TLR8 sensing was chemistry-dependent, clearly shown in the high potentiation frequency with 2'OMe gapmers, followed by 2'MOE and lastly LNA with no potentiation activities at 100nM. Our data suggest that there is a negative correlation between TLR8 potentiation and gapmer ASO nuclease resistance, since boosted TLR8 sensing could be conferred by degradation products of gapmer ASO binding site 2, sensitizing site 1 to its ligand.

Critically, while the majority of 2'OMe, 2'MOE, and LNA gapmer ASOs inhibited TLR7, TLR8 inhibition was more seen with LNA gapmer ASOs, as we found 28% of the LNA ASOs inhibited TLR8 > 20%. Importantly, our data suggests that inhibitory ASOs could operate through competitive bindings to site 1 of TLR7/8 receptors, hindering the binding of their agonists.

Interestingly, we identified the effectors of 2'OMe-PS ASOs modulating TLR7/8 sensing being as short as 3 bases. As such, we demonstrate that 2'OMe GUX trimer being strong TLR7 inhibitors, while TLR8 sensing is inhibited by 2'OMe GAG trimer. We also identified 2'OMe CGG being the optimal TLR8 potentiator.

Importantly, we provided a proof-of-principle that gapmer ASOs combining gene targeting and TLR8 potentiation could have great therapeutic potential in the context of cancer immunotherapy. Furthermore, we propose that inhibitory trimer oligonucleotides could be co-packaged with mRNA to limit its unwanted immunogenicity, which will have great application in the field of mRNA vaccine development.