Making Every Phosphorus Atom Pick a Side The Chiral Auxiliary Claim Behind a Stereopure Oligonucleotide Process

This application covers a way of building synthetic DNA and RNA strands in which the stereochemistry of every backbone linkage is chosen rather than left to chance. Despite a title that sounds like it could cover almost anything, claim 1 is narrow and specific: it is a process claim, a cycle of coupling, capping, modification and de-blocking steps in which the coupling reagent carries a chiral auxiliary group of a defined shape. The applicant is Wave Life Sciences Ltd., and the application is a divisional of Australian application 2020241559, with eleven named inventors. The specification claims priority to United States provisional application 62/821,423, filed 20 March 2019.

The Problem

A synthetic oligonucleotide is assembled one nucleoside at a time on a solid support, and most therapeutic oligonucleotides replace one of the oxygen atoms on each backbone phosphorus with sulfur to resist nuclease attack. That substitution has a consequence that the industry lived with for thirty years: the phosphorothioate phosphorus becomes a stereocentre. Standard phosphoramidite chemistry gives no control over which face the sulfur ends up on, so each linkage is a coin toss between the Rp and Sp configuration.

The specification puts numbers on how close to a coin toss it really is. In one run reported in the examples, a conventional 2′-fluoro G amidite produced a linkage with an Rp to Sp ratio of 47.3 to 52.7. A conventional 2′-fluoro C amidite gave 44.6 to 53.4. A drug with twenty such linkages is therefore not one molecule but a mixture of an enormous number of diastereomers, each with its own binding, stability and nuclease profile, sold and dosed as if it were a single substance.

Controlling the stereochemistry is not new in itself. The recognised route is to attach a chiral auxiliary to the coupling reagent so that the incoming phosphorus is forced to react from one face, then strip the auxiliary off afterwards. The specification’s argument is about what that costs in practice. Adding an auxiliary introduces a new reactive group into the growing chain, typically a secondary amino group liberated when the auxiliary ring opens, and the specification identifies this as a source of impurities if it is not capped.

Worse, the standard fix makes things worse. Conventional capping solutions in oligonucleotide synthesis contain a strong nucleophilic catalyst, and the specification states that these can run at 5 to 15 per cent N-methylimidazole by volume of the capping solution. The document sets out the recognition that a nucleophilic agent used in a capping step placed after coupling and before the sulfurisation step can itself degrade the oligonucleotide and lower the performance of the next step, so the very reagent meant to protect crude purity is eating into it. That is why chirally controlled synthesis has historically come with a crude purity penalty.

What This Invention Does

Claim 1 is a method for preparing an oligonucleotide, comprising one or more cycles, each of which independently comprises five steps: a coupling step, optionally a pre-modification capping step, a modification step, optionally a post-modification capping step, and optionally a de-blocking step. Three of the five are optional, so the mandatory skeleton is only coupling plus modification. The narrowing language comes at the end. In at least one cycle, the coupling step must react a free hydroxyl group of an oligonucleotide or a nucleoside with a coupling partner compound carrying a chiral auxiliary group of one of two drawn structures, which are the two stereochemical variants of the same 1,2 amino alcohol motif bearing substituents R2, R4, R5 and R6.

The single most important restriction sits in one short line: R2 comprises an electron-withdrawing group. Everything after that in claim 1 is definitional boilerplate setting out what the linker, ring and substituent variables may be, and it is expansive, running to C1-30 aliphatic and heteroaliphatic groups and a long list of divalent linkages. The claim is not about a named compound. It is about a class of auxiliary defined by where the electron-withdrawing group sits.

So the answer to what claim 1 covers is the process, not the molecule. The reagent and the product are claimed separately, and it is worth noticing how. Claim 20 covers the phosphoramidite reagent itself, drawn as a cyclic oxazaphospholidine formed from the same auxiliary. Claim 18 covers an oligonucleotide containing an internucleotidic linkage designated *PS or *PR, which on the drawings is a phosphorothioate triester with the auxiliary chain still attached to the phosphorus, in other words the protected intermediate rather than the finished drug. Claim 19 then covers a composition of such oligonucleotides in which at least ((DS) to the power Nc times 100) per cent share the same configuration at each chirally controlled linkage, with DS at least 80 to 99 per cent and Nc the number of controlled linkages. The claim set therefore tracks the auxiliary through reagent, process and intermediate.

The specification names two working auxiliaries. DPSE is defined in the abbreviations table as (S)-2-(methyldiphenylsilyl)-1-((S)-pyrrolidin-2-yl)-ethan-1-ol, and because it is silicon based it is removed with fluoride, using triethylamine trihydrofluoride or tetrabutylammonium fluoride. PSM is the base-labile alternative, carrying a sulfonyl electron-withdrawing group and removed by contact with base under mild conditions. The stereoselectivity numbers are the argument. Against the near-racemic 47.3 to 52.7 for a regular 2′-fluoro G amidite, the L-DPSE version of the same amidite gave 0.6 to 99.4 in favour of Sp at 83.4 per cent yield. The L-PSM 2′-fluoro C amidite gave 99.4 to 0.6 the other way at 90.8 per cent yield, and one L-PSM methyl-C amidite reached better than 99.9 to 0.1. The specification also reports the mirror image auxiliaries, with D-PSM versions selecting the opposite configuration at 92.5 to 7.5 and 93.1 to 6.9, which is the point of a chiral auxiliary rather than a chiral catalyst: you pick the enantiomer of the auxiliary to pick the configuration of the linkage.

Crude purity is where the capping change shows up. A worked example prepares a twenty-unit oligonucleotide designated WV-3473, built from 2′-fluoro and 2′-O-methyl sugars with a mixture of stereodefined phosphorothioate and natural phosphate linkages, on scales of 0.315 to 0.430 mmol. Four batches are reported. Batch B6 on a polystyrene support came out at 34.0 per cent full-length product. B19 reached 47.6 per cent, B56 on a different support 44.3 per cent, and B110 on controlled pore glass 75.2 per cent, with a second figure of 59.0 per cent given in brackets. The specification does not isolate which change caused which gain, but the cycle table shows B6 running a single capping step while B19, B56 and B110 all run a post-thiolation cap instead.

Key Features

• A chiral auxiliary defined by an electron-withdrawing group. Claim 1 does not name a compound. It requires a coupling partner bearing a 1,2 amino alcohol auxiliary in one of two drawn configurations, in which the R2 substituent comprises an electron-withdrawing group and is defined as a linker attached to a wide class of residues.
• Two capping steps in a single cycle. The claimed cycle allows a pre-modification capping step after coupling and a post-modification capping step after sulfurisation or oxidation, separated by the modification step, so that the amino group exposed by the auxiliary and any unreacted hydroxyl can be dealt with at different points with different chemistry.
• Capping without a strong nucleophile. The specification identifies the nucleophilic catalyst in conventional capping solutions, present at 5 to 15 per cent N-methylimidazole by volume, as a cause of degradation in stereoselective synthesis, and describes capping conditions reduced to no more than 0.01 to 5 per cent by volume or no more than roughly one equivalent relative to the acylating agent.
• Two removal chemistries for the auxiliary. DPSE, a methyldiphenylsilyl auxiliary removed by fluoride, and PSM, a sulfonyl-bearing auxiliary removed by mild base, let the auxiliary be stripped under conditions orthogonal to the rest of the protecting group scheme.
• Reagent, intermediate and composition claimed alongside the process. Claim 20 covers the oxazaphospholidine phosphoramidite, claim 18 the oligonucleotide bearing the auxiliary-derived linkage, and claim 19 a composition in which a stated fraction of molecules share the same configuration at every controlled linkage.
• Stated stereoselectivity thresholds. The disclosure sets out formation of a chiral linkage with a stereoselectivity of at least 80:20 rising through 95:5 and 98:2 to 99:1, favouring either Rp or Sp, against the roughly 50:50 the same sugar chemistry gives without an auxiliary.

Who Is Behind It

Wave Life Sciences Ltd. is a clinical-stage RNA medicines company incorporated in Singapore with its operating base in Massachusetts. The title page carries no address, so the country of incorporation comes from the company record rather than from this document. Wave’s own description of its PRISM platform sets out exactly the thesis this application serves, that oligonucleotide pharmacology can be tuned through three levers, sequence, chemistry and stereochemistry, and that stereopure oligonucleotides differ from the mixture-based products currently on the market or in development by others.

That makes the manufacturing question existential rather than academic for this applicant, which is why a process patent of this shape exists at all. Wave’s manufacturing page describes a 90,000 square foot current good manufacturing practice facility in Lexington, Massachusetts, and claims oligonucleotide synthesis capacity from high throughput screening scale up to large-scale GMP manufacturing with, in the company’s words, comparable yield and cost of goods to standard stereorandom oligonucleotides. The specification is the technical account of how that parity is supposed to be reached, and the process tables in the examples run from a process development scale of 0.4 to 1 mmol through to a manufacturing scale of 10 mmol on a 10 cm diameter anion exchange column.

Eleven inventors are named, led by Pachamuthu Kandasamy and including Chandra Vargeese, and the spread of names across process chemistry and analytical development matches a filing that is about running a plant rather than discovering a molecule.

On the family position, the specification is unusually explicit. Its cross-reference paragraph states that the application is a divisional of Australian application 2020241559 and claims priority to United States provisional application 62/821,423 filed 20 March 2019. The international publication in the same family is WO 2020/191252, and the Australian parent is AU 2020241559. Priority country is therefore the United States.

Why It Matters

Stereochemistry in oligonucleotide drugs has moved from a curiosity to a live regulatory and commercial question. Approved antisense and small interfering RNA drugs are mixtures of diastereomers, and that has been accepted because there was no practical alternative at manufacturing scale. Once a stereopure version of the same sequence can be made at comparable cost, the comparison stops being theoretical, and the specification’s numbers are aimed squarely at the cost half of that comparison rather than the biology half.

The capping insight is the part worth remembering. It is a small, unglamorous observation, that the catalyst in a routine housekeeping step is quietly destroying product in a cycle that has an extra reactive amine in it, and the response is to remove the catalyst and split the capping into two differently tuned steps. A crude purity moving from 34 per cent to 75 per cent on the same twenty-unit target is the kind of change that decides whether a stereopure antisense drug can be priced like a conventional one.

It is also worth being clear about what the divisional does not claim. Claim 18 is drawn to the oligonucleotide while the auxiliary chain is still on the phosphorus, not to the finished stereopure drug substance after the auxiliary has been cleaved off. Anyone reading this document expecting a composition of matter claim on a marketed stereopure oligonucleotide will not find one here. What is claimed is the manufacturing route, the reagent that makes it work, and the protected intermediate it passes through.

Related Concepts

• Chiral auxiliary – the general synthetic strategy of attaching a removable chiral group to force a reaction to one face, which is exactly what claim 1 requires.
• Solid-phase synthesis – the support-bound cycle of coupling, capping and de-blocking that the claimed method is a variant of.
• Protecting group – the broader class that both the auxiliary and the capping chemistry belong to, and the reason fluoride and base removal need to be orthogonal.
• Chirality in chemistry – the handedness at the backbone phosphorus that conventional phosphoramidite chemistry leaves to chance.
• Nucleoside – the building block whose free hydroxyl group the claimed coupling step reacts with.
• Enantiomeric excess – the way selectivity ratios such as 99.4 to 0.6 are usually expressed in synthetic chemistry.

---

AU 2026202225 was published in the Australian Official Journal of Patents on 9 April 2026 and is open for public inspection. Patent applications represent inventions that are sought to be protected and do not necessarily reflect commercially available products.