Synthesis of Oligonucleotide Conjugates Employing a Diselenide Linker Authors: Shivam Tikoo, Anne M Noronha, Christopher J Wilds Published: 2026 ABSTRACT ABSTRACTIn this study, we report the synthesis of an alkylene linker containing a terminal 2‐cyanoethyl protected selenium functionality for coupling at the 5'‐end of an oligonucleotide by solid phase synthesis for the preparation of oligonucleotide conjugates. Upon deprotection and cleavage from the solid support, quantitative formation of a 5'‐diselenide cross‐linked oligonucleotide dimer occurred, which was verified by denaturing polyacrylamide gel electrophoresis, ion‐exchange (IEX) high performance liquid chromatography and mass spectrometry. Using dithiothreitol a reducing agent, the diselenide linkage can be converted into the corresponding monomeric selenol, which then reacts with an alkyl halide to form the desired conjugate. The efficient conjugation of various moieties demonstrates the adaptability of the platform to support oligonucleotide‐based applications in chemical biology.Keywords: chemically modified oligonucleotides, diselenide chemistry, nucleic acid chemistry, oligonucleotide conjugate, selenium Introduction Nucleic acids, DNA, and RNA, play crucial roles in numerous biological processes and in the storage of genetic information. Many diseases, including cancers and genetic disorders, arise from the over‐ or underexpression of specific proteins, making them important therapeutic targets [1]. While protein targeted therapies exist, they often lack specificity and may cause off‐target side effects [2, 3, 4]. Over the past few years, tremendous therapeutic potential has been realized by targeting sequences in RNA with complementary oligonucleotides by Watson‐Crick‐Franklin base pairing to achieve gene silencing [3, 5]. These oligonucleotides are typically 20–25 nucleotides in length that can be chemically synthesized using solid‐phase synthesis (SPS). However, these molecules can face challenges in achieving their gene silencing activity including susceptibility to enzymatic degradation and short retention times in the body [6]. To overcome this, various chemical modifications to the nucleic acid scaffold at the nucleobase, sugar, or backbone can be employed to improve stability and pharmacokinetics [7, 8, 9]. An important strategy to enhance therapeutic performance of an oligonucleotide is the attachment of a conjugate, where a small molecule (e.g., cholesterol and folic acid) or larger entities (e.g., peptides and antibodies) are covalently linked to the oligonucleotide [3, 10, 11, 12, 13]. A prominent example is the GalNAc moiety, which targets hepatocytes for liver‐related therapies [14]. In these conjugates, the therapeutic oligonucleotide is attached to the desired ligand through a chemical bond or group with examples including disulfide, thiosuccinimide, triazole, and amide linkages [15, 16, 17, 18]. The synthesis of such conjugates, however, presents challenges, including regioselective attachment of functional groups and chemical incompatibilities between the ligand and oligonucleotide. Thus, while each of these conjugation strategies has found use in various applications, many suffer from specific limitations [19, 20]. To overcome synthetic limitations, alternative chemistries capable of mild, selective ligation are needed. Towards this endeavor, our group has explored selenium‐based linker chemistry for the preparation of oligonucleotide conjugates. Selenium offers unique reactivity not accessible with oxygen or sulfur‐based systems, including enhanced nucleophilicity, rapid exchange kinetics, and a reversible redox cycle between diselenides and selenolates [21, 22]. These properties create a chemically programmable handle for mild, selective, and rapid transformations that is fundamentally distinct from conventional linker chemistry. Selenium's ability to form dynamic diselenide covalent bonds enables versatile applications in synthesizing and functionalizing biomacromolecules [23, 24, 25, 26]. Previous work from our lab demonstrated the utility of an L‐serine methyl ester‐based linker and highlighted the versatility of diselenide chemistry (Figure 1A). We have explored the application of diselenide–selenoester ligation (DSL), inspired by native chemical ligation (NCL), to prepare DNA‐based peptide oligonucleotide conjugates (POCs) [27]. This approach benefits from the enhanced redox susceptibility of diselenides, leading to improved reaction kinetics compared to NCL. Huang et al., have also investigated alkylation of selenium modified nucleic acids, with the multistep preparation of a 5'‐cyanoethyl‐selenium thymidine phosphoramidite which, after incorporation by SPS and deprotection under basic conditions in the presence of an electrophile, can undergo alkylation at the 5'‐terminus (Figure 1B) [28]. Methodologies to prepare alkylselenide nucleic acid adducts. (A) DSL – alkylation [27]. (B) Incorporation of a 5'‐cyanoethyl selenium‐derivatized phosphoramidite by SPS followed by alkylation [28]. (C) Reduction of diselenide cross‐linked DNA and alkylation. CE = cyanoethyl. In the present study, we report a simplified synthetic route for preparing oligonucleotide conjugates by alkylation of a cross‐linked oligonucleotide dimer connected by a diselenide bond (Figure 1C). This approach employs a decanyl linker with a phosphoramidite and cyanoethyl protected selenium at each of the termini. Coupling of this phosphoramidite by SPS to a thymidine decamer (dT10) followed by ammonolysis, results in deprotection and formation of the diselenide cross‐linked oligonucleotide. Using a reducing agent, the dimer can be converted into the corresponding monomeric selenolate, which then reacts with an alkyl halide to form the desired conjugate. This approach provides a robust and straightforward strategy, opening a wide range of synthetic possibilities for producing oligonucleotide conjugates. Results and Discussion Synthesis of the Selenium Modified Linker The simplified approach to prepare nucleic acid conjugates from a diselenide cross‐linked oligonucleotide employs a bifunctional, flexible alkylene linker with a cyanoethyl protected selenide and phosphoramidite group at each terminus amenable to solid‐phase oligonucleotide synthesis (Scheme 1). Starting with 1,10‐decanediol, an acid‐labile dimethoxytrityl group was installed at one of the terminal alcohols. The major product formed was the mono‐dimethoxytritylated compound 1. This reaction also produces a bis‐dimethoxytritylated side product, which is readily separated from compound 1 by flash column chromatography. The other alcohol was converted to an iodide by the Appel reaction to afford 2 for introduction of the protected selenium functionality. To accomplish this di‐2‐cyanoethyl diselenide was prepared, a versatile reagent employed in the preparation of a variety of selenium modified nucleosides [29]. Reduction of this diselenide in situ with sodium borohydride, followed by nucleophilic substitution of the iodo group on 2, afforded the cyanoethyl protected selenide linker 3. In the subsequent step the dimethoxytrityl group was removed using p‐toluenesulfonic acid in dichloromethane/methanol to produce compound 4 with a free alcohol group which was then converted to phosphoramidite 5. The product was characterized by 31P NMR, with a signal at δ 146.9 ppm, diagnostic of a phosphoramidite (see Figures S1–S11 for NMR spectra of the synthesized compounds). Reagents and conditions. (i) 4,4'‐Dimethoxytrityl chloride, pyridine, rt, 13 hrs; (ii) triphenylphosphine, imidazole, iodine, tetrahydrofuran, 0°C to rt, 3 hrs; (iii) di‐2‐cyanoethyl diselenide, sodium borohydride, ethanol, 0°C to rt, 1 hr; (iv) para‐toluenesulfonic acid, dichloromethane:methanol (3:1), rt, 30 min; (v) N,N'‐diisopropylamino cyanoethyl phosphonamidic chloride, diisopropylethylamine, tetrahydrofuran, rt, 30 min. DMT = dimethoxytrityl; CE = cyanoethyl. Oligonucleotide Synthesis and Formation of the Diselenide Cross‐Linked Oligonucleotide In this strategy to prepare the alkylated selenium oligonucleotide conjugate, linker phosphoramidite 5 was first coupled to the 5'‐OH group of a thymidine decamer (dT10) that was assembled by SPS and still bound at the terminal 3'‐O atom to the controlled‐pore glass (CPG) solid support. The coupling of the protected 2‐cyanoethylselanyl phosphoramidite 5 was performed manually with a 0.2 M solution of phosphoramidite 5 and 2‐ethylthiotetrazole as activator in acetonitrile for 10 min. The CPG was then washed with acetonitrile, dried and placed on the oligonucleotide synthesizer to oxidize the phosphite triester to a phosphate triester. The solid‐support was then treated with 10 mL of triethylamine:acetonitrile (50:50) for one hour to remove the cyanoethyl protective groups and minimize acrylonitrile adduct formation [30]. Cleavage from the solid‐support and deprotection of the oligonucleotide was accomplished by ammonolysis. Denaturing PAGE of the crude oligonucleotide product revealed that efficient coupling of linker phosphoramidite 5 occurred with only a trace of the dT10 precursor present (see Figure S12, lane 1). Moreover, the major oligonucleotide product displayed a reduced mobility relative to a dT10 control suggesting formation of the diselenide cross‐linked oligonucleotide. The major oligonucleotide product dT10‐Se‐)2‐ was purified by denaturing PAGE and desalted (see Supporting Information for details). The purity of the diselenide cross‐linked oligonucleotide was demonstrated by analytical PAGE and ion‐exchange (IEX) HPLC (Figure 2) with the expected molecular mass as determined by MS (Figure S13). (A) Denaturing PAGE (20%) of oligonucleotides visualized by UV shadowing. Left to right—Lane 1: Xylene cyanol (XC, 28 nt marker) and bromophenol blue (BPB, 8 nt marker) dyes; Lane 2: dT10; Lane 3: Purified diselenide cross‐linked dT10 (dT10‐Se‐)2‐). (B) dT10 and (C) diselenide cross‐linked dT10 (dT10‐Se‐)2‐) IEX HPLC chromatographs. The trace shows dT10 with a tR = 6.40 min and dT10‐Se‐)2‐ with a tR = 8.76 min. The column was monitored at 260 nm and eluted at rt using a linear gradient of 0%–60% buffer B over 10.5 min, 60%–80% buffer B over 10.50 to 12.50 min, followed followed by 80%–0% buffer B from 12.50 to 14.55 min (buffer A: 100 mM Tris HCl, pH 7.5, 10% acetonitrile; buffer B: 100 mM Tris HCl, pH 7.5, 10% acetonitrile, 1 M sodium chloride). Reduction and Conjugate Formation With the Diselenide Cross‐Linked Oligonucleotide Oligonucleotide conjugation strategies have been extensively explored using a variety of linkages, ranging from reversible disulfide bridges [15] to stable triazole rings formed via click chemistry [17] and amide bonds [31, 32, 33]. These diverse conjugation approaches demonstrate a high propensity for forming functional biomolecular conjugates. However, despite their significant therapeutic potential, each linker possesses distinct chemical and practical limitations. For instance, standard copper‐catalyzed click chemistry requires Cu(I) catalysts that can cause detrimental, oxidative strand degradation to the oligonucleotide backbone [34]. Post‐synthetic amide bond formation requires complex orthogonal protection group schemes and specialized activation reagents to prevent cross‐reactivity between functional groups [33, 35]. Hence, we wanted to develop a versatile, functional linker platform by exploiting the unique reactivity of selenium chemistry. The diselenide bond can be easily reduced using simple reducing agents under aqueous, metal‐free conditions and at relatively low temperatures. To assess the utility of selenium‐based alkylation as a conjugation strategy, we first investigated the redox activation for functionalization of the diselenide linked oligonucleotide. In this approach, reduction of the diselenide bond leads to the generation of a selenolate ion which, given its high nucleophilicity, can be exploited for direct alkylation with brominated small molecules (Figure 3). Conjugation can be accomplished by alkylation at the selenium atom under mild, aqueous metal‐free conditions without the need for protection groups. A variety of alkyl bromides differing in chain length, aromaticity, and hydrophilicity were investigated. Synthetic strategy employing the diselenide cross‐linked DNA to prepare conjugates via reduction followed by alkylation at selenium. To identify the optimal conditions for the generation of the nucleophile, we investigated three reducing agents which include dithiothreitol (DTT), tris(2‐carboxyethyl) phosphine (TCEP), and reduced glutathione (GSH) [36]. Each of these reducing agents was evaluated under identical reaction conditions (750 equivalents with respect to monomer oligonucleotide) to determine their suitability for the subsequent alkylation step. Optimal reduction and formation of the nucleophilic selenolate ion occurred upon treatment of dT10‐Se‐)2‐ with DTT, which is converted to a stable six‐membered ring upon oxidation [37]. In the absence of alkylating agent, after desalting of the product with a C18‐cartridge, oxidation of the monomer oligonucleotide occurred with efficient reversion to the diselenide cross‐linked oligonucleotide without degradation as observed by IEX HPLC and MS (see Figures S14 and S15). With TCEP, deselenization of the oligonucleotide was observed resulting in the formation of a 5'‐O‐decyl adduct (see Figures S16 and S17) as the major product, consistent with phosphite induced reduction of selenium and sulfur containing molecules [38]. Minor side products that were observed include the dealkylated oligonucleotide and a product where replacement of the selenium atom with a hydroxyl group occurred. The latter has been observed previously for the deselenization of selenocysteine with TCEP to produce serine [39]. Employing GSH, the predominant product was a conjugate of oligonucleotide and GSH linked by a selenosulfide bond (see Figures S18 and S19) with a minor secondary product which corresponds to the oligonucleotide adduct with a GSH‐glutamyl fragment. Preliminary attempts to alkylate the diselenide linked oligonucleotide dimer in the presence of TCEP or GSH did not result in significant alkylation as determined by MS (Figures S20–S23). When alkylation was attempted with TCEP and 1‐bromopentane, the major product observed was the deselenized 5'‐O‐decyl adduct [38]. With GSH and 1‐bromopentane the major product was the selenosulfide linked GSH oligonucleotide conjugate. Thus, DTT was utilized to generate the selenolate nucleophile for alkylation of the diselenide linked oligonucleotide dimer. To evaluate the scope of the alkylation step, a series of structurally diverse brominated molecules were examined including linear alkyl chains (1‐bromopentane and 1‐bromodecane), benzyl bromide, a fluorescent probe (4‐bromomethyl‐6,7‐dimethoxycoumarin), and polydispersed methoxy‐PEG (1K) bromide. All alkylation reactions were performed in PBS buffer (pH 7) containing DTT to maintain the selenium in its reduced state (Figure 3). The brominated electrophiles were dissolved in DMF and then added for conjugate formation which was performed at 37°C. The reactions were analyzed by 20% denaturing PAGE to monitor product formation (Figure 4). For all the electrophiles, near quantitative conversion of the diselenide linked oligonucleotide dimer to the alkylated conjugate was observed after 24 h with formation of a single, faster migrating product. Reaction progress was also monitored by IEX HPLC and the products characterized by MS (see Figures S24–S33). HPLC analysis indicated clear formation of one major product for the pentyl, benzyl and 4‐methylene‐6,7‐dimethoxycoumarin adducts. For the longer chain adducts, decanyl and methoxy‐PEG (1K), consumption of the diselenide linked oligonucleotide dimer was observed. However, a faster eluting product was not observed for these conjugates, presumably due to formation of an aggregated structure given the hydrophobicity of the adducts. Composition of all the conjugates prepared was confirmed by MS analysis by correlation between the expected and determined molecular masses. Denaturing PAGE (20%) for the alkylation of diselenide cross‐linked dT10 (dT10‐Se‐)2‐) treated with DTT. The gel was stained with SYBR Gold Nucleic Acid Gel Stain and visualized with an Amersham Typhoon 5 Biomolecular Imager. Lanes left to right: Diselenide cross‐linked dT10 (dT10‐Se‐)2‐); 1‐bromopentane (BP) 24 h; benzyl bromide (BB) 24 h; bromodecane (BD) 24 h; 4‐bromomethyl‐6,7‐dimethoxycoumarin (DMC) 1 h; Polyethylene glycol (PEG, 1K) 24 h. Conclusion In this work, a strategy to prepare oligonucleotide conjugates by alkylation of selenium connected by an alkylene linker to the 5'‐terminus of an oligonucleotide was developed. In this approach, a phosphoramidite derivative of an alkylene linker containing a 2‐cyanoethylselanyl group was prepared from 1,10‐decanediol in a five‐step process with moderate yield. Coupling of this linker at the 5'‐OH of a decathymidine oligonucleotide by SPS followed by deprotection resulted in quantitative formation of a diselenide cross‐linked oligonucleotide dimer. To generate the selenolate nucleophile various reducing agents were investigated including TCEP and GSH. However, these agents resulted in the formation of deselenized and a selenosulflide linked GSH‐oligonucleotide adduct, respectively. With DTT, the selenolate ion was generated facilitating successful conjugation with various alkyl halides ranging from small aliphatic chains to a polydisperse 1K methoxyPEG. The incorporation of hydrophobic alkyl chains, such as decane, highlights the potential of this strategy to modulate physicochemical properties relevant to therapeutic delivery, including membrane interaction and biodistribution. At the same time, the compatibility with PEG and fluorescent moieties demonstrates the adaptability of the platform for applications in imaging and chemical biology [40]. Importantly, the modular nature of this approach enables rapid diversification from a single diselenide precursor, offering a practical route to structure property exploration in oligonucleotide conjugate design. Collectively, these findings establish selenium‐mediated alkylation as a robust and versatile alternative to existing oligonucleotide conjugation strategies. By combining mild reaction conditions, high selectivity, and broad functional group tolerance, this approach expands the chemical toolbox available for nucleic acid modification. Future studies will expand this approach to the functionalization of RNA with moieties to support applications such as delivery and explore the biological consequences of these conjugates to further extend the scope of selenium‐based transformations while also understanding their anti‐cancerous and anti‐oxidative properties in tandem with oligonucleotide therapeutics [41, 42]. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Supporting file: All experimental details, HPLC chromatographs, NMR and MS spectra are found in the Supporting Information. Acknowledgments The work in this research was supported by a grant from the Natural Sciences and Engineering Research Council of Canada (NSERC, RGPIN‐2023‐04821) to C.J.W. The authors are grateful to Concordia University and the NSERC Collaborative Research and Training Experience Program in Programmed Molecules for Therapeutics, Sensing and Diagnostics (PROMOTE) for postgraduate fellowships to S.T. This manuscript is dedicated to the memory of Dr. Frank Seela. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request. References Associated Data Supplementary Materials Supporting file: All experimental details, HPLC chromatographs, NMR and MS spectra are found in the Supporting Information. Data Availability Statement The data that support the findings of this study are available from the corresponding author upon reasonable request.