Abstract:
Current therapeutic developments are shifting their focus from the protein level to the
informational oligonucleotide level. This includes the development of site-directed RNA
editing (SDRE) approaches, which allow selective recoding of RNA nucleobases. Some
SDRE-catalyzing enzymes are expressed endogenously in humans, including ADAR,
which functionally recodes adenines to guanines. SDRE systems based on these
enzymes are a potentially safer alternative to genome editing systems like CRISPR/Cas,
which can induce permanent off-target effects. Several such SDRE approaches have
already been developed, but these are often less effective, dependent on viral
transduction, or require the expression of an exogenous fusion protein. Chemically
modified oligonucleotides (ON) are an alternative that directly steer endogenous ADAR
to the desired RNA site without additional components. However, there are no
systematic studies on the requirements necessary for efficient and nuclease-resistant
ONs. This thesis aims to uncover these requirements through stepwise development of
the RESTORE 2.0 platform – an improvement of the previously published RESTORE
system (recruiting endogenous ADAR to specific transcripts for oligonucleotide mediated
RNA editing). Stabilized RESTORE 2.0 ON designs contained the commercially
available 2’-fluoro, 2’-O-methyl, 2’-H (DNA) modifications and typically retained most of
the editing yields from the corresponding unstable RNA-based designs. Fully modified
RESTORE 2.0 ONs resisted serum and lysosomal nucleases for up to 31 days. Dense
phosphorothioate modifications were essential to recruit endogenous ADAR. The
modification principles were also applicable to the CLUSTER system, another guide RNA
design platform. RESTORE 2.0 ONs could be shortened to just 32 nt and displayed twice
the efficiency without requiring interferon-α supplementation compared to the 95nt long
RESTORE 1.0 ONs. Efficient ON symmetries and favored placements of specific
chemical modifications also aligned with known ADAR dsRNA binding patterns. Further,
fully modified RESTORE 2.0 ONs were efficiently administered via clinically relevant
delivery routes, such as GalNAc-mediated uptake into primary hepatocytes. Most
importantly, optimized RESTORE 2.0 ON designs corrected multiple disease-relevant
point mutations. This included a proof-of-concept treatment of an α-1-antitrypsin
deficiency mouse model, where up to 42% correction on the RNA level was achieved,
restoring therapeutically relevant levels of the corrected protein. Altogether, the
RESTORE 2.0 ON platform provides simple guidelines to further develop chemically
modified ONs for therapeutical purposes.
RNA