RNA Interference (RNAi) Mechanism and Therapeutic Applications

RNA interference is one of the most consequential discoveries of late-20th-century biology — a natural gene-silencing pathway that we’ve turned into a drug modality. As of 2025, at least six siRNA therapeutics have full regulatory approvals, targeting diseases from transthyretin amyloidosis to hypercholesterolemia. This article covers the molecular mechanism and the therapeutic engineering that got us here.

1. The endogenous pathway

RNAi is a conserved defense and gene-regulatory mechanism found in plants, fungi, invertebrates, and mammals. The core pathway:

dsRNA precursor
     │
     ▼  (Dicer, RNase III family)
21–23 nt siRNA duplex with 2 nt 3' overhangs
     │
     ▼  (loading onto AGO2 within RISC)
Guide-strand-loaded RISC
     │
     ▼  (target recognition via Watson-Crick base pairing)
Endonucleolytic cleavage of target mRNA between positions 10-11 relative to the guide 5' end
     │
     ▼
Target mRNA degradation → gene silencing

Argonaute-2 (AGO2) is the only mammalian Argonaute with slicer activity, which is why therapeutic siRNAs are designed to load into AGO2 and cleave, not merely repress, their target.

2. Design principles for therapeutic siRNA

Designing a functional siRNA is not “reverse complement the mRNA and hope for the best.” Key rules:

• Duplex asymmetry. The strand with the less-stable 5’ end (lower Tm base) preferentially loads as the guide. Design the intended antisense strand’s 5’ end with A or U to bias correct loading.
• Length. 19-nt duplex with 2-nt 3’ overhangs (typical) or 21-mer blunt formats.
• Seed region (guide positions 2–8). Perfect complementarity to the target here maximizes on-target activity but also maximizes miRNA-like off-target repression of any transcript with a matching seed. Screen candidates for low complementarity to the 3’ UTR-ome.
• Chemistry. Every commercial therapeutic siRNA is chemically modified:
  • 2’-OMe and 2’-F on ribose to resist nucleases and reduce immune sensing.
  • Phosphorothioate (PS) backbone at the ends for exonuclease resistance.
  • Occasional glycol nucleic acid (GNA) or abasic substitutions to mute off-target activity.

3. Delivery — the actual bottleneck

Naked siRNA is degraded in minutes and can’t cross cell membranes efficiently. Three main delivery strategies dominate today:

GalNAc conjugation (Alnylam’s platform). Trivalent GalNAc binds the asialoglycoprotein receptor (ASGPR) on hepatocytes with high affinity. ASGPR internalizes the ligand and, importantly, releases siRNA into the cytoplasm at sufficient rates to load AGO2. Result: hepatocyte-selective, subcutaneous, quarterly dosing for liver-expressed targets. This is what makes drugs like inclisiran feasible.

Lipid nanoparticles (LNPs). Same basic architecture as mRNA vaccine LNPs: ionizable lipid, cholesterol, PEG-lipid, structural phospholipid. Patisiran (Onpattro), the first FDA-approved siRNA drug, uses an LNP.

Antibody-siRNA and ligand-siRNA conjugates. Emerging strategies to extend RNAi beyond liver: anti-integrin siRNA for muscle, ANGPTL3-conjugated siRNA for adipose, and various targeting ligands for tumors.

4. Approved siRNA drugs

Drug	Sponsor	Target	Indication	Year
Patisiran (Onpattro)	Alnylam	TTR	hATTR amyloidosis	2018
Givosiran (Givlaari)	Alnylam	ALAS1	Acute hepatic porphyria	2019
Lumasiran (Oxlumo)	Alnylam	HAO1	Primary hyperoxaluria type 1	2020
Inclisiran (Leqvio)	Novartis / Alnylam	PCSK9	Hypercholesterolemia	2021 (US 2021)
Vutrisiran (Amvuttra)	Alnylam	TTR	hATTR amyloidosis	2022
Nedosiran (Rivfloza)	Novo Nordisk	LDHA	Primary hyperoxaluria	2023

Two observations from this list:

1. Every approval targets a liver-expressed gene. GalNAc-siRNA-to-hepatocyte is the modality’s sweet spot.
2. Chronic conditions with a single, well-validated driver gene ( PCSK9, TTR, ALAS1) are the winners — those are the “sniper rifle” indications for RNAi.

5. Comparison with adjacent modalities

Property	siRNA	ASO (antisense oligo)	CRISPR-Cas9 (in vivo)
Mode of action	AGO2-mediated cleavage or translational repression	RNase H1 cleavage, splice modulation, translation blocking	DNA double-strand break + repair
Duration per dose	Weeks to months (LNP) or up to quarterly (GalNAc)	Weeks (2’-MOE), months (GalNAc-conjugated)	Permanent (edited allele persists)
Reversibility	Yes — stop dosing	Yes	Largely no
Off-target profile	Seed-mediated miRNA-like effects	RNase H1 non-specific cleavage	Off-target editing (permanent)
Regulatory precedent	6+ approvals	10+ approvals	1 approval (Casgevy)

For a deeper dive see our siRNA vs ASO comparison.

6. Design tools and bioinformatics

If you’re supporting an RNAi program, the core computational tasks are:

• Guide design. siDesign Center (Horizon Discovery), Dharmacon’s siDesign, custom scripts using Reynolds/Ui-Tei/Amarzguioui rules. Rank candidates on efficacy, thermodynamic asymmetry, GC content (30–52 %), and lack of internal secondary structure.
• Off-target prediction. BLAST or Bowtie2 the guide (positions 2–8 and full length) against the transcriptome; count seed matches in 3’ UTRs. Tools like siSPOTR and TargetScan (retooled for siRNA seeds) are standard.
• Chemistry design. Position-dependent 2’-OMe/2’-F patterning to minimize immunogenicity and off-target while preserving on-target potency. Alnylam’s ESC-GalNAc pattern is the industry reference.
• PK/PD modeling. Simple biphasic decay models fit siRNA hepatocyte concentrations and downstream target knockdown well.

7. What’s next

• CNS delivery via intrathecal siRNA (analogous to nusinersen for ASO) is in trials for Huntington’s disease and ALS.
• Extrahepatic siRNA to muscle, adipose, and tumor is the frontier — several conjugation platforms are in Phase I.
• Combination with mRNA/CRISPR. siRNA to transiently knock down repair pathways during a CRISPR edit, or to silence a mutant allele while an mRNA replaces the wild-type protein.

Bottom line

RNAi went from mechanism-of-the-year (Nature 1998) to Nobel (2006) to approved drugs (2018-2024) in one working generation. The technology is not a curiosity — it is a mature therapeutic modality with a well-defined design space, a clear delivery paradigm for liver, and an active push into other tissues.

Related reading: mRNA vaccine technology, siRNA vs ASO, and circular RNA as the next therapeutic frontier.