Nucleic acid therapeutics

An oligonucleotide drug finds its target by Watson-Crick pairing. That single property inverts the usual difficulty of drug discovery: designing a molecule with the right specificity is a sequence problem solvable at a desk, and an undruggable protein — one with no binding pocket — becomes accessible by silencing its messenger RNA instead. Everything hard about the field is therefore downstream of design.

Three distinct mechanisms

Small interfering RNA is loaded as a duplex into Argonaute 2 within the RNA-induced silencing complex; the passenger strand is discarded and the guide strand directs AGO2 to cleave the complementary mRNA endonucleolytically. This is catalytic — one loaded complex destroys many transcripts, and the loaded complex persists, which is why a single subcutaneous dose can suppress a hepatic target for months.

Antisense oligonucleotides in the gapmer configuration work differently: the DNA-like central window forms an RNA-DNA heteroduplex that recruits RNase H1, an endogenous nuclease that cuts the RNA strand. Unlike RISC, RNase H1 is active in the nucleus, so gapmers can degrade pre-mRNA and nuclear-retained transcripts.

The third mechanism is not degradation at all. A fully modified steric-blocking oligonucleotide binds a splice site or splicing enhancer and redirects the spliceosome. Nusinersen uses this to force inclusion of exon 7 in SMN2 transcripts, converting a non-functional protein into a functional one; exon-skipping oligonucleotides in Duchenne muscular dystrophy restore a reading frame to produce a shortened but partly functional dystrophin.

Chemistry keeps them alive

Unmodified RNA is destroyed by serum and tissue nucleases within minutes and is filtered by the kidney. Three modifications carry the field. The phosphorothioate backbone, in which a non-bridging oxygen is replaced by sulphur, resists nucleases and — as importantly — makes the molecule bind plasma proteins, which prevents immediate renal clearance and enables tissue uptake. Sugar modifications at the 2’ position, methoxy, fluoro, methoxyethyl or a locked bicyclic bridge, raise duplex affinity and further block nuclease attack. Chemical modification also suppresses innate immune recognition by endosomal Toll-like receptors, which unmodified oligonucleotides trigger readily.

Delivery is the constraint

The industry’s decisive tool is a trivalent N-acetylgalactosamine ligand. It binds the asialoglycoprotein receptor, which is expressed on hepatocytes at extremely high density and recycles from the surface every few minutes, giving repeated uptake opportunities. This is why the approved siRNA drugs treat hepatically expressed targets — transthyretin, PCSK9, ALAS1, complement C5 and others: the conjugate solves cell entry for exactly one cell type.

Entry is still not arrival. Endocytosed oligonucleotide is trapped in endosomes, and the fraction that reaches the cytosol where AGO2 or RNase H1 operates is generally estimated in the low single-digit percentages or less. Almost all administered material is degraded or exocytosed. Endosomal escape, not affinity, sets the dose. Extending the platform beyond liver — to muscle, kidney or across the blood-brain barrier — requires either an equivalent receptor-ligand pair or local administration, which is why intrathecal delivery is used for central nervous system targets. Backbone-related toxicities, including thrombocytopenia, complement activation and renal tubular effects, track phosphorothioate protein binding and constrain how far the dose can be pushed.

See also the technology article at /technology/nucleic-acid-therapeutics/.