Innovative Improvements in P(V) Oligonucleotide Synthesis Platform, Empowering Oligonucleotide Drug Development!

01 Challenges

Control of Diastereomers and Drug Stability

Large-scale oligonucleotide synthesis has long relied on phosphoramidite P(III) chemistry and the four-step synthesis cycle, including deprotection, coupling, oxidation, and capping. However, with the advancement of oligonucleotide therapeutics, the traditional P(III) four-step cycle faces increasing challenges. One of the most representative challenges is the control of phosphorothioate (PS) bond stereochemistry in oligonucleotides.

The P(III) four-step cycle cannot control the stereoconfiguration of PS bonds, resulting in a mixture of both R (Rectus, clockwise) and S (Sinister, counterclockwise) absolute configurations at each PS linkage. Taking the antisense oligonucleotide (ASO) drug Spinraza as an example, all 18 nucleotides in this drug are connected by a PS-modified backbone, theoretically possessing 2¹⁷ = 131,072 possible diastereomers. This vast number of isomers introduces significant unpredictability into the development and quality control of PS oligonucleotide drugs.

02 Opportunity

The P(V) Synthesis Approach for PS Stereocontrol: The Ψ Reagent

In 2018, Scripps Research and Bristol-Myers Squibb (BMS) published in Science a novel technology using pentavalent phosphorus "P(V)" chemistry for PS oligonucleotide synthesis: the Ψ reagent (Ψ = PSI, i.e., Phosphorus−Sulfur Incorporation).

The Ψ reagent consists of three parts: a chiral scaffold (oxidized limonene), a P(V)-S center, and a leaving group (pentafluorothiophenol). This reagent brought two improvements to solid-phase PS oligonucleotide synthesis:

1. (1) The chiral scaffold enables control over the absolute configuration of the PS bond.
2. (2) Using P(V) chemistry instead of P(III) phosphoramidite avoids the oxidation/sulfurization step, thereby simplifying solid-phase synthesis.

However, several challenges remained for implementing the Ψ reagent in practical, large-scale production:

1. (1) High material cost of the Ψ reagent. The leaving group of the Ψ reagent is very expensive (approximately $3000/kg).
2. (2) The strong organic base DBU (1,8-Diazabicyclo[5.4.0]undec-7-ene) used in the coupling step can cause cleavage of the solid-support linker, reducing yield and purity.
3. (3) The Ψ reagent required non-commercially available nucleobase protecting groups, making it difficult to adapt for large-scale production.

03 Optimization and Upgrade

A Scalable Oligonucleotide Synthesis Platform: Ψ^Br

To address the aforementioned issues, the researchers systematically investigated the leaving group, solid-support linker, and nucleobase protecting groups of the Ψ reagent, leading to the development of an improved synthesis platform: Ψ^Br.

As shown in Figure 1A, the leaving group of the Ψ reagent was first improved by replacing the expensive pentafluorothiophenol with the more readily available and cheaper 4-bromothiophenol (the latter costs less than one-tenth of the former). Furthermore, changing this leaving group did not affect nucleoside monomer synthesis, and like the original Ψ reagent, stereocontrol is achieved via the chiral scaffold (Figure 1, B).

Figure 1. Lower cost and controlled stereochemistry of the Ψ^Br reagent.
(A) Comparison of the leaving groups and material costs between the existing Ψ reagent and the Ψ^Br reagent. The oxidized limonene chiral scaffold is marked in black, the P(V)-S center in yellow, the Ψ reagent leaving group in reddish-brown, and the Ψ^Br reagent leaving group in green.
(B) X-Ray crystal structures of the two stereoisomers of the Ψ^Br reagent. (+) represents the dextrorotatory isomer, (-) represents the levorotatory isomer. The purple stick center represents the P(V) atom.

Secondly, the researchers improved the solid-support linker to enhance its stability in the DBU environment.

As shown in Figure 2, A, the standard succinate linker (#12) is cleaved by DBU during the coupling step, reducing production efficiency and purity. To establish the Ψ^Br synthesis platform, the researchers compared linker #12 with three other linkers (Figure 2, B): #11 (Sar-Glu), #13 (Q), and #14 (PEG4). Under the DBU conditions required for coupling, over 80% of linker #12 was cleaved after just 20 coupling cycles (Figure 2, C). In contrast, #11 and #14 demonstrated much better stability, with over 80% of the linker remaining intact even after 120 coupling cycles. Therefore, linker #11 (Sar-Glu) was ultimately selected as the solid-support linker for the Ψ^Br synthesis platform.

Figure 2. Comparison of four solid-support linkers.
(A) Cleavage of the solid-support linker by the strong organic base DBU. The gray rod represents the solid-support UnyLinker, the dark red circle represents the nucleoside monomer, and the arrow indicates the electron transfer pathway (amide deprotonation).
(B) The four linkers selected for comparison: #11 (Sar-Glu), #12 (succinate), #13 (Q), and #14 (PEG4). The bond marked in gray represents the linker portion.
(C) Stability of the four linkers over 120 coupling cycles. The vertical axis represents the proportion of intact linker, and the horizontal axis represents the number of coupling cycles.

Finally, the researchers screened nucleobase protecting groups for the A, C, G, and T bases suitable for the Ψ^Br synthesis platform (Figure 3).

The existing Ψ reagent required POM and Pya nucleobase protecting groups. These protecting groups not only needed custom synthesis (not yet scaled up) but also required protection of all four bases, including thymidine (T), which lacks a reactive amino group. Screening based on the new Ψ^Br reagent and Sar-Glu linker revealed that conventional protecting groups such as iBu (for G), Bz (for C), and dma (for A) could achieve the same results, with thymidine (T) requiring no protection. These three protecting groups have already been produced on a multi-ton scale, making them more favorable for industrial scale-up.

As seen in the chromatograms in Figure 3, the stereocontrolled PS oligonucleotides (Rp-19 and Sp-19) synthesized using the new Ψ^Br reagent, Sar-Glu linker, and the three common protecting groups exhibited purity identical to those obtained with the existing Ψ reagent. This result demonstrates the feasibility of the Ψ^Br synthesis platform.

Figure 3. Conventional nucleobase protecting groups used in the Ψ^Br synthesis platform. The dashed box shows the custom protecting groups POM and Pya (pivaloyloxymethyl and 1-methylpyrrolidin-2-iminomethyl, reddish-brown) paired with the Ψ reagent. The Ψ^Br synthesis platform employs common protecting groups: iBu (isobutyryl, pink) for G, Bz (benzoyl, blue) for C, and dma (dimethylacetamidate, green) for A. Thymidine (T) requires no protection (black). The reddish-brown upper trace represents the R-configured PS (Rp), and the dark blue lower trace represents the S-configured PS (Sp). The left panel below shows the fully PS-modified sequence synthesized with the Ψ reagent paired with POM and Pya protecting groups, incorporating selective control over both R and S configurations. The right panel shows the same sequence synthesized using the Ψ^Br synthesis platform with iBu, Bz, and dma.

04 Conclusion

P(V) synthesis based on the Ψ reagent offers advantages over P(III) chemistry, including PS stereocontrol and avoidance of the oxidation step (three-step cycle). The Ψ^Br synthesis platform further achieves: (1) lower raw material costs; (2) more stable coupling; and (3) simpler nucleobase protection. Ψ^Br holds promise as an easily scalable synthesis platform with higher yield for stereocontrolled PS oligonucleotide synthesis, providing a new and powerful tool for oligonucleotide drug manufacturing!