Antibody-oligonucleotide conjugate

An AOC is a covalent, three-module biomolecule. A monoclonal antibody (or fragment) supplies antigen-specific homing and a long circulating half-life; a linker tethers the payload and governs when and where it is released; an oligonucleotide, usually a small interfering RNA (siRNA), antisense oligonucleotide (ASO), or phosphorodiamidate morpholino oligomer (PMO), engages a target RNA and changes gene expression.¹

Structurally it is the ADC's nucleic-acid cousin: the same antibody-linker-payload grammar, but the "warhead" is a gene-regulatory sequence rather than a cytotoxin. That single swap makes conjugation, cellular uptake, and manufacturing materially harder, because the payload is large, highly negatively charged, and hydrophilic. Note: for siRNA/ASO the charge is not neutralized during manufacturing. It is managed, and even exploited for analysis (see §13). Only PMO payloads are charge-neutral, and that comes from the backbone chemistry itself, not a process step.

The AOC works in a relay: bind, internalize, escape, act.

• Bind. The antibody recognizes a cell-surface receptor enriched on the target tissue (e.g., transferrin receptor 1 on muscle).
• Internalize. Binding triggers receptor-mediated endocytosis. The receptor's endocytic efficiency is the single best predictor of whether an AOC delivers a useful dose.²
• Escape. The oligonucleotide must exit the endosome into the cytoplasm or nucleus. Endosomal entrapment is a primary efficiency loss across the whole modality, with escape often below 1 to 2%. Adding an endosomal-escape / cell-penetrating peptide (e.g. endosomolytic peptides like INF7, or CPPs) to help the payload "punch out" is an active strategy, but a hard trade-off: unmodified endosomolytic agents (chloroquine, membrane-lytic peptides) tend to be too toxic for systemic use, so the engineering goal is conditional, low-toxicity release (see §15).
• Act. Freed payload engages its RNA target. An siRNA degrades an mRNA (cytoplasm); an ASO or PMO alters splicing or blocks translation (often nuclear).

The idea is twenty years old; the clinical reality is recent.

• 2005. Song et al. show antibody-mediated in vivo siRNA delivery via cell-surface receptors, the conceptual seed.¹
• 2013. Avidity founded; pivots from antibody-delivered siRNA nanoparticles to direct antibody-siRNA conjugation.
• 2015. Genentech's THIOMAB-siRNA study maps how hard antibody-siRNA delivery really is.
• 2021 to 2022. Avidity's AOC 1001 becomes the first AOC in the clinic (DM1); preclinical DMD exon-skipping data published.
• 2023. Malecova et al. publish the foundational AOC delivery paper in Nucleic Acids Research; positive AOC 1001 Phase 1/2 topline.
• Oct 2025. Novartis agrees to acquire Avidity for roughly $12B; early cardiology assets spun out into Atrium Therapeutics; close targeted H1 2026.⁴

Choosing the receptor

The receptor decides the tissue. The dominant target is transferrin receptor 1 (TfR1), abundant on skeletal and cardiac muscle and a fast recycler, which drives efficient uptake. Both Avidity and Dyne use it, which is exactly why TfR1-to-muscle is now crowded rather than defensible.²

Choosing the format

A full monoclonal antibody (Avidity) gives a long half-life and is engineered effector-null so it delivers without triggering immune killing. A smaller Fab fragment (Dyne's FORCE) trades half-life for potentially deeper penetration. Emerging bispecifics combine a targeting arm with a transport arm, a frontier for reaching the brain.²

The linker must stay intact in circulation, then release on cue inside the cell. Too stable and the oligo never frees; too labile and it dumps in the bloodstream. This, plus where on the antibody the payload attaches, is the densest zone of craft and patents.

• Cleavable linkers release on an intracellular trigger: reducible disulfides (–S–S–), acid-labile bonds, or enzyme-cleavable peptides (e.g., Val-Cit).⁷
• Non-cleavable linkers rely on lysosomal degradation of the antibody; often more systemically stable.
• Site-specific conjugation fixes payload position and number, giving a homogeneous product with predictable behavior instead of a messy random mixture.
• Drug-antibody ratio (DAR) is tuned for potency versus stability; Avidity describes an optimized oligo:antibody ratio, and a raised DAR for PMOs to boost muscle delivery.⁸

Conjugation & linker chemistries: patent status

These are the attachment and release chemistries an AOC borrows from the ADC world. As with the payload, "off-patent" refers to the core chemistry; branded site-specific platforms remain proprietary.

Owners/dates from Abzena, ADC linker literature, and patent records.³⁰

Is the ncAA platform feasible for AOCs? Yes. Genetically encoding a non-canonical amino acid (via an orthogonal amber-suppressor tRNA/synthetase pair) places a chemically unique handle, such as the ketone of p-acetylphenylalanine or the azide of p-azidophenylalanine, at a defined antibody position. That handle then reacts by oxime ligation or click, giving a truly homogeneous, position- and stoichiometry-controlled conjugate with a stable bond and clean PK. It was validated for ADCs by Scripps/Ambrx (>95% coupling, full tumor regression in models) and the same chemistry transfers directly to an oligo payload. Trade-offs: it needs an engineered cell line expressing the tRNA/synthetase pair (more upstream work than plug-and-play lysine/cysteine coupling), and the core platforms are proprietary. So it is technically feasible and attractive for a homogeneous AOC, but it is a build-or-license decision, not a free option.³²

DAR / OAR: is there a consensus?

Yes, and it points low. The field uses DAR (drug-antibody ratio, borrowed from ADCs) and OAR (oligonucleotide-antibody ratio) interchangeably for AOCs. Where ADCs push DAR up toward 4 to 8 to maximize cytotoxin delivered, AOCs do the opposite: the working consensus is a low DAR of about 1 to 3, with 1 to 2 most common. The reason is the payload's mass and dense negative charge.

• Why low. Each siRNA (~13–14 kDa, highly anionic) added to the antibody shifts its isoelectric point (pI), increases hydrophilicity and charge, and accelerates non-specific clearance, which shortens half-life. Excess conjugation also creates steric hindrance that can impair antigen binding. So more payload can mean less delivered to target.
• The evidence. Avidity's systematic structure-activity study (Cochran et al., comparing DAR 1 / 2 / 3 across conjugation sites) found conjugation site and DAR are critical for optimal PK and siRNA delivery, with higher DAR degrading plasma PK. This is the closest thing to a published SAR consensus in the field.³³
• The homogeneity angle. A tight, defined DAR (not an average over a heterogeneous mixture) is itself the goal, which is exactly why site-specific methods (ThioBridge, THIOMAB, ncAA) that lock DAR at a clean 1 or 2 are favored over random lysine/cysteine coupling that yields a DAR distribution.

Bottom line: unlike ADCs, AOCs are payload-limited, not payload-maximizing. The design target is the lowest DAR that delivers enough oligo, kept homogeneous, to preserve antibody-like PK. PMO payloads (charge-neutral) tolerate somewhat higher loading than siRNA, which is why Avidity notes a raised DAR specifically for its PMO programs.⁸

Payload modalities

Chemical modifications

Backbone and sugar chemistry turn a fragile sequence into a drug. Phosphorothioate linkages boost nuclease resistance and half-life; sugar modifications (2′-O-methyl, 2′-MOE, 2′-fluoro) raise affinity and cut off-target effects. Avidity's next-generation siRNA reported up to a roughly 30-fold increase in skeletal-muscle delivery preclinically, proof that payload engineering, not just targeting, moves potency.⁹ See §16 for the full chemical-engineering toolkit, design templates, and frontier chemistries.

The defining claim: gene modulation with an antibody's targeting, aimed at the tissues GalNAc and LNPs leave behind, with potentially fewer off-target effects than a cytotoxic ADC because the payload acts only on a specific RNA sequence.¹⁰

Carrier / targeting format

The carrier is the half that decides tissue reach and pharmacokinetics. Bigger carriers circulate longer; smaller ones penetrate tissue better. Here is how the formats trade off.

Sizes and half-lives are typical class values, not program-specific. Half-life and penetration trade against each other, and the mAb-vs-Fab choice is the clearest example of that tension.¹⁸

Oligonucleotide payload type

The payload is the half that treats the disease. Which one you pick follows the disease mechanism and where in the cell you must act.

Payload mass and dense negative charge cap the drug-antibody ratio (typically 1–3): over-loading shifts the antibody's isoelectric point, speeds clearance, and can impair antigen binding.¹⁸

Delivery, not sequence design, is the field's binding constraint. GalNAc and LNPs solved the liver because hepatocytes have fenestrated vasculature, high endocytic activity, and one dominant ligand-friendly receptor (ASGPR). Every other tissue is harder, and each new destination is a defensible franchise.¹¹

• Muscle. The first extrahepatic tissue AOCs cracked, via TfR1, and where the clinical proof now lives.
• CNS. Gated by the blood-brain barrier; approaches use BBB-transcytosis receptors and routes like ENT2. Huge unmet need (Huntington's, ALS, Alzheimer's).
• Kidney. Hard because of rapid renal clearance; selective proximal-tubule targeting is the prize.
• Immune cells. The lane Tallac exploits for oncology.

Avidity Biosciences Pioneer · acquired

The category-definer and most clinically advanced. Full mAb to TfR1, effector-null, carrying siRNA or PMO. Three lead "delpacibart" programs:

• del-zota (DMD, exon-44): about 25% of normal dystrophin on average (total up to 58%), creatine kinase down more than 80%, functional gains vs natural history. Breakthrough + Fast Track + Rare Pediatric; BLA planned 2026.⁴
• del-desiran (DM1): fully enrolled global Phase 3 HARBOR; 54-week readout expected 2026; prior data suggested reversal of DM1 progression.
• del-brax (FSHD): Phase 1/2 FORTITUDE + confirmatory FORTITUDE-3; met its primary biomarker endpoint under Novartis.¹³

The Novartis deal: roughly $12B (Oct 2025). The early precision-cardiology assets (AOC 1086, AOC 1072) plus BMS/Lilly collaborations were spun into "SpinCo," now the independent, Nasdaq-listed company Atrium Therapeutics (see below); $450M / $600M reciprocal termination fees; close targeted H1 2026.

Atrium Therapeutics Avidity spin-out · cardiac

The precision-cardiology company carved out of Avidity in the Novartis transaction, and a pure-play AOC developer aimed at the heart. It runs on the same clinically validated AOC platform Avidity built. Atrium describes its technology as combining "the tissue selectivity of monoclonal antibodies and other targeted delivery ligands with the precision of oligonucleotides," delivering siRNA directly into cardiac muscle and building on Avidity's skeletal-muscle delivery to reach the heart, while sidestepping the cytotoxicity, immunogenicity, and re-dosing limits of other delivery systems. That platform description makes it a true AOC, not an adjacent RNA modality. Two lead candidates (renamed from their AOC-10xx designations):

• ATR 1072 (formerly AOC 1072), for PRKAG2 syndrome, with an IND planned for H2 2026.
• ATR 1086 (formerly AOC 1086), for PLN (phospholamban) cardiomyopathy, with an IND expected to follow in 2027.
• Plus two undisclosed research-stage rare-cardiomyopathy targets. Led by CEO Kath Gallagher, launched with roughly $270M, and trades on Nasdaq (RNAM).³⁷

Strategically, Atrium is the clearest test of the AOC cardiac thesis. It takes TfR1-to-muscle delivery, which is proven in skeletal muscle, and pushes it into the heart, targeting genetic cardiomyopathies (PLN, PRKAG2) that today have only symptom management, pacemakers or ICDs, or transplant.

Avidity / Novartis precision cardiology The other cardiac-AOC engine

Atrium is not the only cardiac-AOC source. Avidity itself kept a Precision Cardiology program and struck cardiovascular collaborations with Bristol Myers Squibb (up to five cardiovascular and other targets) and Eli Lilly, all of which now sit inside Novartis after the acquisition. So the cardiac-AOC field is effectively two related efforts: the independent Atrium, and the Avidity cardiac pipeline plus partnered targets now owned by Novartis. Both draw on the same TfR1-to-muscle delivery platform.

Cardiac RNA that is not AOC Adjacent, not a competitor

Several well-funded companies target heart disease with RNA, but are not antibody conjugates, so they do not belong in the AOC company list. The distinction is the same one that separates Atrium from the crowd: a true cardiac AOC uses an antibody to carry the oligo into heart muscle, whereas these players use liver-targeted or naked RNA chemistry.

• Corsera Health (founded by John Maraganore and Clive Meanwell). Its lead, COR-1004, is a once-yearly subcutaneous siRNA against PCSK9 and AGT. This is RNAi delivered to the liver with no antibody, aimed at cardiovascular risk rather than genetic cardiomyopathy.
• Kardigan. Its tonlamarsen is an Ionis-discovered ASO that silences AGT mRNA in the liver for hypertension. Again a liver-targeted oligo, not an antibody conjugate.

Why they are listed here and not above: both treat the heart's drivers but reach the liver, so they are adjacent RNA-for-cardiovascular players. If either later adds an antibody-targeting arm to reach cardiomyocytes directly, they would join the future-entrants watch list. The academic case for the whole cardiac-AOC thesis is set out in a 2024 review on cardiac delivery of RNA therapeutics using AOCs for genetic cardiomyopathies.³⁸

Dyne Therapeutics · FORCE Closest rival

Differentiates on format and payload breadth. Fab fragment to TfR1; wider payload range including PMO, ASO, even whole proteins (DYNE-401 carries the GAA enzyme for Pompe). DYNE-101 (DM1) and DYNE-251 (DMD exon 51) in Phase 1/2.¹⁴

Tallac Therapeutics · TRAAC Boundary case

A definitional gray zone, not a core AOC. Tallac markets itself as a "next-generation ADC" company, though its own filings also describe TAC-001 as an "antibody-oligonucleotide conjugate." The payload is a CpG oligonucleotide acting as a TLR9 immune agonist. It stimulates immunity rather than silencing or splicing a target RNA the way a therapeutic siRNA/ASO/PMO does. So structurally it is an oligo conjugate, but mechanistically it is an immunostimulatory ADC, distinct from the extrahepatic gene-modulation thesis Avidity and Dyne embody. TAC-001 (CD22-targeted, solid tumors) is in Phase 1/2 and received FDA Fast Track (Oct 2025); ALTA-002 (SIRPα, with ALX Oncology) is preclinical.¹⁵

Others & adjacent players

• Gennao Bio / Yale. Patents on the 3E10 cell-penetrating antibody (enters via ENT2, localizes to the nucleus), a distinct mechanism for oncology.⁷
• Denali Therapeutics. Cited among AOC players, leveraging BBB transport-vehicle expertise.
• AstraZeneca, Ionis, Silence. Appear in AOC analyses largely via partnerships and platform extensions, not approved AOCs.¹⁷

Dosing frequency & the TPP case

Current AOCs dose roughly every 6 to 8 weeks IV, and the interval is set by the antibody, not the payload, so there is real room to stretch it. This matters commercially: a longer interval is a materially better target product profile (TPP), meaning fewer infusions, less clinic burden, and a friendlier proposition for chronic rare-disease patients.

The key point is that the pharmacodynamic effect far outlasts the dose. CK stays near-normal for a year, and NHP knockdown runs beyond 12 weeks, so the roughly Q6–8W cadence is limited by how long the conjugate circulates (antibody PK), not by the payload wearing off.³⁴ Levers that would push toward Q12W / quarterly dosing: (1) more durable payload chemistry (Advanced-ESC-type stabilization), (2) higher delivery per dose (Avidity's reported ~30× muscle-delivery gain), and (3) longer conjugate half-life. Note the format tension: a full mAb (Avidity) favors longer intervals via FcRn recycling, whereas a Fab (Dyne) has a shorter half-life and trends toward more frequent dosing, so the mAb-vs-Fab choice (see §7) directly sets the achievable frequency and thus the TPP ceiling.

China players Emerging

As of early 2026, China has disclosed preclinical AOC candidates but no clinical-stage AOC yet. The Western developers still lead the clinic (Avidity, Dyne, Tallac, Denali), but domestic AOC programs have now surfaced in the literature, so the picture is shifting from pure infrastructure toward an actual (early) pipeline.³⁵

• Domestic AOC programs (preclinical). ChainGen Bio (Shanghai), CGB-1001 and Shanghai Minwei Biotech, MWN108 (metabolic/endocrine) are named AOC candidates in preclinical development, the first disclosed China-origin AOCs.³⁵
• CDMO / manufacturing muscle. WuXi AppTec (DMPK and integrated conjugate services) and ChemExpress (linker/payload, oligonucleotide and bioconjugate process development) are active AOC enablers, building the exact CMC capability the modality is bottlenecked on.¹⁸
• ADC leadership as a springboard. China now originates a large share of global ADC innovation (Hengrui, Kelun-Biotech, LaNova, DualityBio, Innovent, and the first approved bispecific ADC from Baili/BMS). AOCs reuse the ADC antibody-linker-payload grammar, so this base is the most credible on-ramp to domestic AOC programs.¹⁹
• siRNA / oligonucleotide depth. A deep domestic RNAi and oligonucleotide sector (the natural payload source), including Argo, Rona, Sirius, SiranBio and others (see future entrants), plus heavy conference attention, suggests more domestic AOC candidates are a matter of timing, not capability.¹⁸

Caveat: reflects publicly disclosed programs current to early 2026; the China oligo/AOC space is moving fast, so additional undisclosed domestic candidates likely exist. (This section was updated after CGB-1001 and MWN108 surfaced. An earlier draft stated no China AOC existed, which is no longer accurate.)

Potential future entrants Watch list

The next wave of AOC competition will come from today's siRNA/ASO houses. These companies already own the two hardest pieces of an AOC: a validated oligonucleotide payload and hard-won extrahepatic-delivery chemistry. Bolting on a targeting antibody is a smaller leap than building an oligo platform from scratch, so over the next ~5 years expect several to add antibody-conjugate programs. None of the below is a disclosed AOC developer today; they are listed as convergence candidates, not current players.

• Ionis. The ASO pioneer; already appears in AOC patent filings (anti-CD29 conjugates), the most explicit crossover signal among the incumbents.¹⁸
• Alnylam. The RNAi/GalNAc leader; owns the deepest siRNA chemistry and the strongest incentive to break past the liver.
• Arrowhead, Silence Therapeutics, Dicerna/Novo (legacy). siRNA-conjugate platforms explicitly chasing extrahepatic tissues (muscle, CNS, lung); antibody targeting is the logical next ligand class.
• Wave Life Sciences, Sarepta. ASO/PMO and exon-skipping expertise in the exact muscle indications where AOCs are proving out; conjugation is a natural potency upgrade.
• SiranBio (时安生物). Suzhou siRNA company whose STORK extrahepatic and adipose-targeting delivery platform is a credible AOC on-ramp; its ALK7 siRNA SA030 was licensed to GSK in May 2026 (~$55M upfront, up to ~$1.005B). Note: SA030 is a siRNA conjugate for metabolic disease, not an AOC, so SiranBio is here as an adjacent capability, not an AOC developer.²⁰
• China siRNA houses (Argo, Rona, Sirius). A fast-rising domestic RNAi cohort with exactly the payload plus extrahepatic-delivery assets AOCs build on: Argo Biopharma (靖因), a clinical-stage RADS platform with 7+ molecules in the clinic including a TTR siRNA (BW-50218) and CNS programs, plus major Novartis deals; Rona Therapeutics (大睿), whose APOLLO extrahepatic-CNS platform (ligand-siRNA) reached ~87% CNS knockdown in NHP with Q6M dosing; and Sirius Therapeutics, with long-acting siRNA (e.g. Factor XI for thrombosis). All are siRNA rather than AOC today, but sit one antibody-arm away from the modality.³⁶

Why this is a forecast, not a fact: the antibody + linker + oligo grammar is shared across ADCs, siRNA conjugates, and AOCs, so the barrier between these categories is low. The thesis, that payload owners converge on antibody targeting to escape the liver, is analyst reasoning, not a set of announced AOC programs.

Where each company's programs sit across disease areas and how far each has advanced. Muscle diseases dominate; oncology and cardiology are the expansion frontiers. Cell color encodes the furthest development stage; the stage is also written in each cell, so status never rides on color alone.

Hover any cell for the specific program. Stages reflect publicly disclosed status current to early 2026; blank cells mean no disclosed program, not impossibility.
*Tallac is a boundary case: its CpG/TLR9-agonist payload makes it structurally an oligo conjugate but mechanistically an immunostimulatory ADC (see §9).
Not charted: China-origin preclinical AOCs CGB-1001 (ChainGen) and MWN108 (Minwei, metabolic). Their target indications aren't cleanly disclosed, so they're covered in §9 rather than forced into a disease column.

Two axes that separate the players. Horizontal: how far the delivery target sits from the crowded TfR1-to-muscle lane, where left is the commodity center and right is novel tissue/mechanism. Vertical: clinical maturity, from preclinical (bottom) to filing-stage (top). The upper-left is validated-but-crowded; the lower-right is differentiated-but-unproven.

Read it as strategy: Avidity and Dyne compete head-on for muscle with proven delivery; Atrium (Avidity's cardiac spin-out) extends the same TfR1 approach from skeletal muscle into the heart; Tallac, Gennao, and Denali trade clinical maturity for a differentiated target no one else owns. The durable moat lives in the lower-right, if it can climb. The dashed band marks expected future entrants: siRNA/ASO houses (Alnylam, Ionis, Arrowhead, Wave, Sarepta, SiranBio) that own payload and delivery today and could add antibody targeting within ~5 years. They are a forecast, not disclosed AOC programs.

The AOC opportunity map is a delivery map. An AOC earns its complexity only where the liver route (GalNAc, LNP) can't reach, so the addressable indications are organized by target organ, the receptor that gets you in, and the RNA target/marker you act on once inside. The table below is a forward-looking whitespace map: proven lanes, active frontiers, and open thesis territory. Rows marked whitespace have no disclosed AOC program; they are where the modality logically extends.

ATTR: the instructive case

ATTR shows why "big market" does not equal "AOC opportunity." Transthyretin is made almost entirely in the liver, the tissue GalNAc and LNP already own. So liver-TTR silencing is a crowded, solved lane: vutrisiran (GalNAc siRNA), eplontersen (ASO), patisiran (LNP), stabilizers (tafamidis, acoramidis), and even one-shot CRISPR (NTLA-2001, ~87% knockdown).²² A generic AOC adds cost and complexity with no delivery advantage there. The real AOC opening is the non-hepatic TTR reservoirs the liver drugs can't reach: the choroid plexus (CNS/leptomeningeal amyloid) and retinal epithelium (ocular). Shutting off TTR production in those cells is something neither the liver drugs nor an anti-amyloid antibody can do.

A caution on the "de-risked antibody" argument. It is tempting to say the existing anti-ATTR antibodies (PRX004, coramitug/NI006) de-risk the antibody half of an ATTR AOC, but that reasoning is flawed. Those are naked antibodies that bind extracellular misfolded ATTR / amyloid fibrils to clear existing deposits. An AOC needs the opposite: an antibody against an internalizing cell-surface receptor that carries the oligo into the TTR-producing cell. Binding a fibril is not a delivery portal, so these programs de-risk none of the AOC's delivery problem. If anything they cut the other way: a naked antibody that clears cardiac amyloid is a simpler competing modality in the same space, weakening the case for a cardiac AOC. The durable AOC rationale in ATTR is therefore mechanistic and anatomical (stop production in the CNS/eye reservoirs), not "the antibody is already de-risked."²²

The strategic-buyer angle

Failed or crowded liver programs create acquisition incentive, not the reverse. AstraZeneca is deeply committed to ATTR through eplontersen (with Ionis) and the large CARDIO-TTRansform program; a player that heavily invested in TTR, especially one facing a Phase 3 setback or a crowded liver field, has strong reason to buy its way into the next ATTR frontier (CNS/eye/cardiac) rather than concede it. An AOC or transcytosis platform reaching those reservoirs would be a natural bolt-on to protect and extend a franchise. The same logic explains the Novartis/Avidity deal: big pharma buys validated delivery platforms rather than rebuild them.

Note: the buyer-incentive reasoning is analyst inference from disclosed franchise positions and deal patterns, not a reported transaction or a stated intent from any company.

An AOC is three hard manufacturing problems stacked together. Structural complexity and mechanistic diversity create real quality-control burdens; a delivery platform is increasingly treated as a reusable CMC and regulatory asset, not just a targeting trick.¹

• Oligo synthesis + purification. Phosphoramidite chemistry is mature, but resolving closely related oligo impurities is the bottleneck.
• Conjugation homogeneity. Controlling DAR and site to avoid unpredictable heterogeneous mixtures.
• Characterization. Oligos form higher-order structures and aggregates; standardized analytical guidance is thin.
• Platform reuse. Reusing one validated antibody across programs (Avidity) turns a component into a scalable engine.⁸

Typical conjugation workflow, and the charge question

The oligonucleotide's negative charge is not neutralized during manufacturing. A common misconception is that you strip the charge before or after coupling; you don't. For siRNA/ASO the charge stays throughout. It is managed, and the analytics even exploit it. A representative click-based route (Rady et al.):³¹

1. Oligo → linker. An amine-modified single strand is coupled to a strained-alkyne linker (e.g. BCN- or DBCO-based) at slightly basic pH (7.4–8.0, kept ≤4 h to avoid base-mediated RNA cleavage), then acetone/lithium-perchlorate precipitated.
2. Hybridize the linker-strand with its complementary strand (for siRNA): heat to the melting point, cool slowly.
3. Antibody → azide. Separately, lysines on the antibody are functionalized with an azide handle; purified by gel filtration.
4. Click conjugation (SPAAC). Azide-antibody + alkyne-oligo join by copper-free click, with no charge step involved.
5. Purify & read out charge. Size-exclusion chromatography separates conjugated from free antibody and excess oligo; strong-anion-exchange HPLC (SAX) then quantifies the drug-antibody ratio precisely because each added oligo makes the conjugate more negative and elute later. The charge is the measurement handle, not a defect to remove.

So the answer to "neutralize before or after conjugation?" is neither for siRNA/ASO: the charge is retained and worked around (it is what caps the DAR at ~1–3). PMO is the only charge-neutral payload, and that is intrinsic to its phosphorodiamidate backbone (see §16), not a manufacturing step you add.

Patents concentrate around three claim types: the targeting antibody and its epitope/receptor-binding proteins; the linker and conjugation chemistry; and the oligonucleotide sequence plus modifications. Avidity's foundational patents and TfR1-binding filings anchor the muscle space; Yale/Gennao's 3E10 filings stake out a distinct cell-penetrating mechanism for oncology.⁷

The regulatory-IP wildcard is the FDA's Platform Technology Designation, which rewards reproducible methods reusable across products, but requires a first approval to unlock. No AOC has one yet, so it's a future prize, not a current moat.²¹

The next differentiation wave is architectural. Reviews converge on four frontiers:²

• Bispecific antibodies. One arm to target, one to drive transcytosis, especially across the blood-brain barrier.
• Stimuli-responsive "smart" linkers. Inert in blood, releasing only at a specific pH or enzyme.
• Endosomal-escape engineering. The "let it out" problem. Approaches include appending endosomolytic / cell-penetrating peptides (e.g. INF7-type, or Entrada's EEV cyclic-peptide vehicles), fusogenic lipids, or pH-triggered membrane-disruptors to rescue trapped payload. The catch is the therapeutic window: anything that punches through the endosomal membrane can also damage other membranes, so the whole design problem is making escape conditional and non-toxic rather than just adding a lytic peptide.²⁷
• New payloads & AI-driven design. Gene-editing cargo, and machine learning to screen antibody/linker/oligo combinations.

The honest constraint: every platform is still pre-approval. Activity clusters where delivery is already tractable (muscle), and extrahepatic reliability isn't yet at GalNAc's level for the liver. A durable edge shows up as human efficacy in a tissue no one else can reach, not a cleaner version of the same TfR1-to-muscle molecule.

Naked RNA is not a drug; chemistry makes it one. Unmodified oligonucleotides are degraded by nucleases in minutes, provoke innate immune sensing, and bind their targets weakly. Two decades of medicinal chemistry turned them into durable therapeutics, and that same chemistry is what an AOC's payload inherits. This section covers the modification toolkit, the assembled design "templates," and the frontier chemistries. In an AOC the antibody solves tissue delivery; this chemistry solves stability, potency, and safety of the payload once it arrives.

The three modification sites

Every oligonucleotide can be engineered at three places: the backbone, the sugar, and the base. Modifications are combined in patterns, not applied uniformly.

Modification roles per Egli & Manoharan and the WuXi DMPK review.¹⁸ Patent-status entries reflect core composition-of-matter chemistry (owners/dates from Ionis, Alnylam, and USPTO/Google Patents records).²⁹

Design templates (siRNA)

The real engineering is the pattern, meaning where each modification sits across the duplex. Alnylam's public progression shows how templated chemistry compounded into durability and safety, the property that lets a subcutaneous GalNAc–siRNA silence for months:

• STC (Standard Template Chemistry). First stable-enough design; active but needed high doses.
• ESC (Enhanced Stabilization Chemistry). Added terminal phosphorothioate linkages at the 5′ ends of both strands; a big jump in potency and duration.
• Advanced ESC. Systematic optimization of 2′-F / 2′-OMe placement across both strands; further potency and duration gains.
• ESC+. Introduced a thermally destabilizing GNA in the seed to suppress seed-mediated off-target hepatotoxicity while keeping on-target activity, a specificity upgrade rather than a stability one.²⁶

Durability comes from metabolic stability: chemically stabilized siRNA survives in acidic intracellular compartments and re-loads into newly made Ago2/RISC for weeks, so stability chemistry directly buys duration.²⁷

ASO / PMO gapmer chemistry

ASOs that recruit RNase H use a gapmer design: a central DNA "gap" (which RNase H needs) flanked by high-affinity modified "wings" (2′-MOE or LNA) that protect the ends and boost target binding. Steric-block ASOs and PMOs instead avoid RNase H entirely and work by occupying the RNA, the basis of exon-skipping (the PMO mechanism Avidity's del-zota and Dyne's DYNE-251 use).

Frontier: stereochemistry & novel backbones

Each phosphorothioate linkage is a stereocenter, so a conventional oligo is a mixture of up to millions of diastereomers with differing stability and activity. Stereopure / stereodefined chemistry (e.g., Wave Life Sciences) controls this backbone chirality to tune potency, durability, and safety; newer phosphoryl-guanidine (PN) backbone chemistries are being explored to further shift the stability/immunogenicity balance. Stereodefined PMO-gapmers are an active research direction.²⁸

Why this matters for AOCs

An AOC's efficacy and safety window is co-determined by payload chemistry: the same modifications that give a standalone siRNA/ASO its stability and specificity carry over, but conjugation adds constraints. Payload mass and negative charge cap the drug-antibody ratio (1 to 3), and the oligo must survive the nuclease-rich endosome/lysosome long enough to escape (escape efficiency is often below 1%). So AOC payload design is not "any approved oligo bolted to an antibody"; it is chemistry re-optimized for the conjugate's trafficking.¹⁸

AOC

Antibody-oligonucleotide conjugate: antibody + linker + nucleic-acid payload.

ADC

Antibody-drug conjugate: same architecture, small-molecule cytotoxin payload.

ASGPR

Asialoglycoprotein receptor: the hepatocyte receptor GalNAc exploits.

ASO

Antisense oligonucleotide: single strand blocking translation or altering splicing.

DAR

Drug-antibody ratio: payload molecules per antibody.

ENT2

Equilibrative nucleoside transporter 2: cell-entry route of the 3E10 antibody.

GalNAc

N-acetylgalactosamine: the ligand targeting siRNA to the liver via ASGPR.

LNP

Lipid nanoparticle: encapsulation delivery, mainly hepatic; carries mRNA.

PMO

Phosphorodiamidate morpholino oligomer: neutral-backbone splice-switching oligo.

siRNA

Small interfering RNA: double strand triggering RNAi mRNA degradation.

TfR1

Transferrin receptor 1: the muscle-enriched receptor most AOCs target.