Regulation of Duplex Formation in Response to pH and Complementary RNA: Development of Oligonucleotides Modified with S‑Methylthioimidate-Bridged Nucleic Acid and Evaluation of Their Function

Abstract

Biomaterials such as oligonucleotides that can control their functions in response to pH are highly useful in the fields of biotechnology. In this study, S-methylthioimidate-bridged nucleic acid (Me-TIBNA) was designed to transform 2′-aminothymidine into an amido-bridged nucleic acid (AmNA), allowing for the modulation of oligonucleotide functionality in response to pH changes. Me-TIBNA-modified oligonucleotides were synthesized to investigate alkaline intramolecular cyclization subsequent to the acidic hydrolysis of the thioimidate bridge moiety. The results revealed that Me-TIBNA could be converted to 2′-aminothymidine, followed by the condensation of the amine and thioester under basic conditions to yield AmNA. Interestingly, the presence of a complementary RNA enhanced the alkaline intramolecular cyclization process. Evaluation of the duplex-forming ability of the resulting oligonucleotides showed that Me-TIBNA- and AmNA-modified oligonucleotides exhibited excellent RNA-binding ability. In contrast, oligonucleotides modified with the 2′-aminothymidine derivative, the acid-hydrolyzed form of Me-TIBNA, led to a more pronounced destabilization of duplexes compared to their natural counterparts. These findings highlight the potential of Me-TIBNA as a versatile material for conferring pH and complementary RNA responsiveness to oligonucleotide duplex formation. To our knowledge, this study is the first to demonstrate the control of nucleoside sugar conformation within oligonucleotides by forming a bridged structure between the 2′- and 4′-positions in aqueous solution in response to external stimuli.

Introduction

Oligonucleotides, such as DNA and RNA, exhibit strict recognition of the complementary strand sequence through Watson–Crick base pairing to form duplexes. , Additionally, DNA and RNA play vital roles in vivo, not only in storing genetic information but also in regulating gene expression. For example, riboswitches change their conformation in response to binding small molecule ligands and pH changes, regulating gene expression. − Genetic systems are influenced by changes in the biochemical environment. Therefore, oligonucleotides that adjust their hybridization ability with the complementary strand in response to environmental changes in vivo can be powerful tools for controlling gene expression and sensing environmental changes.

Artificial nucleic acids are crucial for the functionalization of oligonucleotides, with oligonucleotides finding applications in therapeutics, − gene diagnosis, , and nanotechnology. , Various nucleic acids have been engineered to respond to environmental stimuli such as light irradiation, , redox changes, , and pH changes. − In particular, pH changes are significant not only in intracellular vesicles during endocytosis but also in tumors, where extracellular pH is lower than in normal tissue. Consequently, chemically modified oligonucleotides that hybridize with complementary strands in response to pH changes are valuable for regulating DNA and RNA functions based on intracellular pH and for detecting pH fluctuations. However, the hybridization capacity of current pH-responsive oligonucleotides with complementary single strands is comparable to or lower than that of natural oligonucleotides. Therefore, enhancing the affinity of the duplex for complementary strands is essential for developing effective pH-responsive nucleotides.

Oligonucleotides modified with 2′,4′-bridged nucleic acids (2′,4′-BNAs), , exemplified by locked nucleic acid (LNA) − or an amido-bridged nucleic acid (AmNA), , can form highly stable duplexes with complementary RNA compared to natural DNA. This enhanced stability is attributed to the constrained sugar moiety, which adopts the same N-type conformation as RNA in the duplex. Based on these characteristics, AmNA has been applied to therapeutic oligonucleotides. − Conversely, 2′-amino DNA derivatives, considered as a cleaved analog of the amide bridge ring in AmNA, significantly decrease the thermal stability of duplexes. On the other hand, we have demonstrated that a 2′,4′-BNA analog featuring a disulfide bridge, which undergoes reversible changes in response to redox conditions, significantly changes the duplex stability under such conditions. Benzylidene acetal-protected 2′,4′-BNA derivatives lose their duplex-forming capability due to bridge ring cleavage upon photoirradiation. However, there is no precedent, including our previous studies, for artificial nucleic acids incorporated into oligonucleotides forming a bridged structure between the 2′- and 4′-positions in response to external stimuli in an aqueous buffer. Therefore, we have devised a novel artificial nucleic acid, 2′-aminothymidine with 4′-thioester, which forms an AmNA structure in an aqueous buffer in response to pH changes and controls its hybridization ability with complementary strands. However, since this nucleoside derivative was expected to be unstable under various conditions of oligonucleotide synthesis, we developed a thioimidate-bridged nucleic acid (TIBNA) as its precursor (Figure ). The thioimidate moiety in TIBNA can be hydrolyzed under aqueous acidic conditions to yield a 4′-thioester-2′-aminothymidine, which is anticipated to destabilize the duplex based on our previous research. In addition, the acidic hydrolysis of thioimidate requires both a proton source and water; therefore, Me-TIBNA-modified DNA can be synthesized without using an acidic aqueous solution. Furthermore, the hydrolysis product of TIBNA serves as a precursor for forming amide bonds in an aqueous buffer, as the condensation of thioesters and amines is commonly employed in the chemical ligation of peptides and proteins. − Consequently, the resulting 2′-aminothymidine with 4′-thioester can be transformed into AmNA through intramolecular cyclization under basic conditions. In this study, an S-methylated derivative of TIBNA (Me-TIBNA) was synthesized and integrated into oligonucleotides. This was done to investigate alkaline intramolecular cyclization following acid hydrolysis of the thioimidate moiety, with the aim of developing potent pH-responsive oligonucleotides.

1.

Molecular design of Me-TIBNA for developing pH-responsive oligonucleotides with a high binding ability to target RNAs.

Results and Discussion

Synthesis and Intramolecular Cyclization of Me-TIBNA-Modified Oligonucleotides

The hydroxy groups of AmNA 1 were protected with triethylsilyl (TES) groups to yield compound 2, which was then converted to compound 3 using Lawesson’s reagent (Scheme ). The thiocarbonylation of the amido-bridge in 2 was confirmed through the analysis of the heteronuclear multiple bond correlation spectrum of 3 (Figure S1). Deprotection of the TES groups in 3 with hydrogen trifluoride triethylamine complex followed by reaction with 4,4′-dimethoxytrityl (DMTr) chloride led to the protection of the 5′-hydroxy group with a DMTr group, affording compound 4 in two steps with a good yield. Subsequently, S-methylation of the thioamide was performed using methyl iodide and sodium bicarbonate to obtain the monomethylated compound 5 in 62% yield. The site of methylation on the sulfur atom of the thioamide was determined by comparing the ¹H and ¹³C NMR spectra of compounds 4, 5, and thioAmNA (Figure S2). Me-TIBNA-phosphoramidite 6, a crucial building block for oligonucleotide synthesis, was synthesized through the phosphitylation of 5. Subsequent experiments were conducted to confirm the feasibility of amidation posthydrolysis of Me-TIBNA in oligonucleotides. Me-TIBNA 5 was dissolved in a 5% trifluoroacetic acid in aqueous acetonitrile (1:1) solution to yield the 2′-aminothymidine derivative 7 without any noticeable impurities. Compound 7 was then treated with 100 mM Tris buffer (pH 9.0) to regenerate AmNA 1.

1. Synthesis of Me-TIBNA-Phosphoramidite 6 .

The phosphoramidite 6 was utilized to synthesize oligonucleotides modified with Me-TIBNA following phosphoramidite chemistry. − Activator 42 was employed as an activator with the coupling time of 6 extended to 5 min (as opposed to 30 s for DNA phosphoramidites) to ensure quantitative introduction of Me-TIBNA into the oligonucleotide. Oligonucleotides were synthesized in the DMTr-Off mode to prevent hydrolysis of the thioimidate moiety associated with deprotection of the DMTr group using an acidic aqueous solution postsynthesis. Furthermore, to address partial hydrolysis of Me-TIBNA during the removal of triethylammonium acetate (TEAA) in high-performance liquid chromatography (HPLC) purification, the eluate postpurification was promptly gel-filtered to eliminate TEAA and obtain highly pure Me-TIBNA-modified ON1–4 (Table and Figure ). These optimized conditions facilitated the successful synthesis of oligonucleotides with Me-TIBNA modification. For the 2′-aminothymidine- and AmNA-modified oligonucleotides (ON5–8), Me-TIBNA in ON1 and ON2 was hydrolyzed in acetate buffer (pH 5.5). The resulting ON5 and ON6 were then treated with Tris buffer (pH 9.0) to yield AmNA-modified ON7 and ON8 through intramolecular amide-bond formation (Scheme ). The formation of ON5–8 was verified through LC/MS of the respective molecular weights. The isolated yields of ON5 and ON6 were moderate, probably due to simultaneous acidic depurination and acidic hydrolysis of Me-TIBNA.

1. Sequences of Modified Oligonucleotides (ON1–12) and Their Molecular Weight Data Obtained through LC/MS.

	(5′ to 3′)	Calcd [M]	Deconvoluted mass [M]
ON1	d(GCGTTXTTTGCT)	3702.6	3702.9
ON2	d(GCGXTXTXTGCT)	3844.6	3844.7
ON3	TTTTTTTTTX	3049.5	3049.7
ON4	TTTTTTTTXT	3049.5	3049.7
ON5	d(GCGTTYTTTGCT)	3720.6	3720.9
ON6	d(GCGYTYTYTGCT)	3898.6	3898.7
ON7	d(GCGTTZTTTGCT)	3672.6	3672.9
ON8	d(GCGZTZTZTGCT)	3754.6	3754.6
ON9	TTTTTTTTTY	3067.5	3067.3
ON10	TTTTTTTTYT	3067.5	3067.3
ON11	TTTTTTTTTZ	3019.5	3019.5
ON12	TTTTTTTTZT	3019.5	3019.6

2.

Chemically modified nucleic acids incorporated into the oligonucleotides shown in Tables –.

2. Synthesis of ON5–8 .

Duplex-Forming Ability of the Synthesized Oligonucleotides

The thermal stability of duplexes formed by the synthesized oligonucleotides (ON1, ON5, and ON7) with complementary RNA (RNA-A) and DNA (DNA-A) was evaluated based on the melting temperature (T _m) determined through UV-melting experiments. The results are summarized in Table , and the UV-melting data are presented in Figure . ON1, with a single Me-TIBNA group, exhibited a T _m value 4 °C higher than that of natural DNA (ON13) and demonstrated excellent binding ability to complementary RNA comparable to that of AmNA (ON7). Conversely, when the complementary strand was DNA, the T _m values of the duplexes formed by Me-TIBNA-modified ON1 and AmNA-modified ON7 were 52 and 53 °C, respectively, comparable to that of natural ON13. In contrast to those of ON1 and ON7, the T _m of the duplex formed by ON5, containing the acidic hydrolysis product of Me-TIBNA with complementary DNA and RNA, was 14 °C lower than that of the natural duplexes. These results align with our findings that 4′-modofied 2′-aminothymidine derivatives significantly decrease duplex stability. Remarkably, when the complementary strand was RNA, the T _m values of the duplexes formed by Me-TIBNA-modified ON1 and AmNA-modified ON7 were 18 °C higher than that of the duplex formed by ON5. This finding strongly suggests that Me-TIBNA is an artificial nucleic acid molecule capable of regulating the hybridizing ability of oligonucleotides through intramolecular cyclization following acidic hydrolysis of the thioimidate bridge in Me-TIBNA. Interestingly, the melting curve of the duplex formed by ON5 and the complementary RNA, as shown in Figure A, exhibits a two-step sigmoid, with the T _m values of 34 °C for the first step and approximately 52 °C for the second step. This is consistent with the T _m value of the duplex formed by the AmNA-modified ON7. In addition, the peak of AmNA-modified ON7 was observed in LC/MS analysis of the duplex formed by ON5 and RNA-A after the T _m measurement. Therefore, we propose that the duplex formation between ON5 and the complementary RNA may facilitate the intramolecular cyclization of 2′-aminothymidines with 4′-thioester, leading to their conversion to AmNA.

2. Evaluation of Thermal Stability of Duplexes Formed with Complementary RNA and DNA .

	(5′ to 3′)	RNA T m (ΔT m ) (°C)	DNA T m (ΔT m ) (°C)
ON13	d(GCGTTTTTTGCT)	48	52
ON1	d(GCGTTXTTTGCT)	52 (+4)	52 (0)
ON5	d(GCGTTYTTTGCT)	34 (−14)	38 (−14)
ON7	d(GCGTTZTTTGCT)	52 (+4)	53 (+1)

^aConditions: 10 mM phosphate buffer (pH 7.2), 100 mM NaCl, and 4 μM of each oligonucleotide. The T _m values are the averages of three measurements, and the measurement error is within ±1 °C. The sequences of complementary RNA (RNA-A) and DNA (DNA-A) are 5′-r­(AGCAAAAAACGC)-3′ and 5′-d­(AGCAAAAAACGC)-3′, respectively.

^bΔT _m: change in T _m value compared to that of natural DNA (ON13).

^cThe UV-melting curve exhibits a two-step transition, as shown in Figure A, and the T _m values for the first transition with lower T _m are shown in the table.

3.

Melting curves of absorbance at 260 nm for (A) DNA–RNA and (B) DNA–DNA duplexes.

Based on the observation that the melting curve of the duplex formed by ON5 and the complementary RNA exhibited a two-step sigmoid, we aimed to validate the T _m calculated by UV measurements using another experimental method. To further analyze the kinetics of the duplex formation reaction, the equilibrium constant K _d for ON1, ON5, ON7, and ON13 was determined using the biolayer interferometry (BLI) method (Table , Table , and Figure S3). A strong correlation was observed between K _d and T _m values, with the K _d of the ON5 and complementary RNA duplex being 4.11 × 10^–5, the highest among the measured K _d values. These findings support a T _m of 34 °C for duplex formation by ON5 and complementary RNA, as shown in Table . For RNA targets, the binding constant k _on and dissociation constant k _off were investigated in detail. Me-TIBNA-modified ON1 and AmNA-modified ON7 exhibited similar k _on values to natural ON13 (k _on = 4.11 × 10⁵), while their k _off values were 3- to 4-fold lower than that of ON13 (k _off = 3.70 × 10^–3). The slow dissociation of duplexes with Me-TIBNA and AmNA, RNA analogs with restricted sugar conformations, aligns with previous literature trends. , In contrast, for DNA targets, no significant differences were observed in k _on and k _off among ON1, ON7, and ON13. ON5, modified with an acidic hydrolysis product of Me-TIBNA, exhibited substantially lower k _on values of 1.70 × 10² and 1.15 × 10³ compared to natural ON13 for both RNA (k _on = 4.11 × 10⁵) and DNA targets (k _on = 3.73 × 10⁵). These results suggest that the introduction of 2′-aminothymidines severely hinders the duplex assembly process. The k _off of ON5 was also 2- to 10-fold higher than that of ON13, indicating that 4′-thioester-2′-aminothymidine accelerates duplex dissociation. Considering the T _m measurements, the formation of ON5 duplexes is highly unfavorable, both thermodynamically and kinetically.

3. Kinetic Analysis of Duplex Formation with Complementary RNA Using the BLI Method.

(5′ to 3′)	k on (M–1s–1)	k off (s–1)	K d (M)	T m (°C)
d(GCGTTTTTTGCT) (ON13)	4.11 × 105	3.70 × 10–3	8.99 × 10–9	48
d(GCGTTXTTTGCT) (ON1)	4.51 × 105	1.01 × 10–3	2.23 × 10–9	52
d(GCGTTYTTTGCT) (ON5)	1.70 × 102	6.98 × 10–3	4.11 × 10–5	34
d(GCGTTZTTTGCT) (ON7)	3.36 × 105	8.41 × 10–4	2.50 × 10–9	52

^a T _m values are the same as those shown in Table .

4. Kinetic Analysis of Duplex Formation with Complementary DNA Using the BLI Method.

(5′ to 3′)	k on (M–1s–1)	k off (s–1)	K d (M)	T m (°C)
d(GCGTTTTTTGCT) (ON13)	3.73 × 105	5.76 × 10–4	1.55 × 10–9	52
d(GCGTTXTTTGCT) (ON1)	3.67 × 105	4.96 × 10–4	1.35 × 10–9	52
d(GCGTTYTTTGCT) (ON5)	1.15 × 103	4.28 × 10–3	3.71 × 10–6	38
d(GCGTTZTTTGCT) (ON7)	3.74 × 105	5.71 × 10–4	1.53 × 10–9	53

^a T _m values are the same as those shown in Table .

Kinetic Analysis of Acidic Hydrolysis and Basic Intramolecular Cyclization of Modified Oligonucleotides

Acidic hydrolysis and basic intramolecular cyclization were kinetically analyzed. Initially, the impact of pH and the Me-TIBNA modification position on the ring-opening reaction was evaluated using ON3, where Me-TIBNA was placed at the 3′-end, and ON4, where Me-TIBNA was positioned adjacent to the 3′-end. Oligonucleotides were incubated in 100 mM acetate buffer solutions ranging from pH 4 to 6 at 25 °C, and the reactions were monitored through HPLC. Product concentrations (ON9 and ON10) and remaining starting materials (ON3 and ON4) were determined from peak area ratios. The natural logarithm of these concentrations versus reaction time was plotted (Table S1, Figure S4, and Figure S5). The plot revealed a linear relationship, indicating that acid hydrolysis followed pseudo-first-order kinetics. Therefore, the reaction rate constant k ₁ for hydrolysis at each pH was calculated by using the rate equation for a first-order reaction. The hydrolysis reaction progressed faster as pH decreased (Table S2 and Figure ), with k ₁ approximately 150 times higher at pH 4 compared to pH 6. This indicates a nonlinear correlation between proton ion concentration and k ₁. In general S-alkylthioimidate acid hydrolysis, the reaction rate depends on the pK _a of the conjugate acid of the thioimidate, which is approximately 3–4. Protonation of the thioimidate nitrogen accelerates hydrolysis, with a sigmoidal relationship between pH and the reaction rate constant around the pK _a value. − The pH versus k ₁ correlation depicted in Figure exhibits a sigmoidal pattern, suggesting that the pK _a of the conjugate acid of Me-TIBNA is approximately 4. Notably, DNAs are unstable under highly acidic conditions, and no kinetic analysis of Me-TIBNA acid hydrolysis below pH 4 was performed. Regarding the Me-TIBNA introduction position, hydrolysis was approximately 2 to 3 times faster for ON3 than for ON4. This is probably because of the exposed thioimidate moiety of Me-TIBNA at the 3′-end, rendering it more susceptible to hydrolysis. Additionally, the TIBNA introduced at the 3′-end lacks the 3′-phosphate group, and the acid hydrolysis reaction rate may depend on the presence or absence of the 3′-phosphate group.

4.

Relationship between pH and reaction rate constants k ₁ for the acidic hydrolysis of Me-TIBNA introduced in oligonucleotides.

The dependence of pH and modification position on the ring-closing reaction was evaluated by studying ON9 and ON10 in 100 mM Tris buffer ranging from pH 8 to 9. The concentrations of the resulting products, AmNA-modified oligonucleotides (ON11 and ON12), were calculated by following the same method used for analyzing the acidic hydrolysis reaction. The relationship between the natural logarithm of these concentrations and reaction time was established (Table S3, Figure S6 and Figure S7). The results revealed a linear relationship, indicating that the basic ring-closing reaction is a first-order reaction dependent on the concentrations of ON9 and ON10. The reaction rate constant k ₂ for the basic ring-closing at each pH was calculated from the rate equation for the first-order reaction, suggesting that higher pH values corresponded to faster progression of the ring-closing reaction. However, no direct correlation was observed between k ₂ and the hydroxyanion concentration (Table S4 and Figure ). This could probably be attributed to the increased nucleophilicity of the 2′-amino group, which strongly influences the reaction rate of the basic ring-closure process. Examining the impact of modification positions of the 2′-aminothymidine derivative on the reaction rate, it was found that the reaction rate constant for ON10 exceeded that of ON9, which was modified with hydrolyzed Me-TIBNA at the 3′-end. This difference may be attributed to the hindrance caused by the exposure of the 2′-amine and 4′-thioester in the hydrolyzed Me-TIBNA at the 3′-end. Consequently, the amidation process decelerates possibly due to increased hydration effects compared with the hydrolyzed Me-TIBNA within the sequence. The increased structural freedom resulting from the introduction of hydrolyzed Me-TIBNA at the 3′-end may also have inhibited the formation of intramolecular amide bonds. Notably, intramolecular amide bond formation occurred even under mildly basic conditions at pH 8 to 9, yielding AmNA-modified oligonucleotides (ON11 and ON12). These findings are significant as they demonstrate the successful establishment of covalent bridges in 2′,4′-BNA derivatives in water, which has been extremely difficult to achieve in the past. As shown in Table S3, the cyclization process is expected to require more than 1 day for completion. This indicates that the utilization of Me-TIBNA as a pH-responsive material may necessitate regulating the reaction rate through functional group modifications on the sulfur atom. The amide bond formation by the condensation of thioesters and amines tends to be faster when the pK _a of the leaving thiol is lower. Therefore, TIBNAs modified with benzyl thiols or thiophenols as leaving groups are likely to accelerate intramolecular amide bond formation to give AmNA.

5.

Relationship between pH and reaction rate constants k ₂ for basic intramolecular cyclization.

Finally, the ring-closing reaction rate of hydrolyzed Me-TIBNA was investigated using ON5 and its target DNAs and RNAs. The T _m analysis in Figure suggested that the ring-closing reaction might be accelerated by complementary RNA. In this experiment, ON5, duplexes formed by ON5 with target DNAs, and those with target RNAs were dissolved in 10 mM sodium phosphate buffer (pH 7.2 or pH 8.0) containing 100 mM NaCl. The reaction was incubated at 25 or 37 °C and monitored through HPLC analysis. To test if the ring-closing reaction is influenced by the complementary strand, target RNAs and DNAs (RNA-G, RNA-C, RNA-U, DNA-G, DNA-C, and DNA-T) forming mismatched base pairs with ON5 were also evaluated. Plotting the natural logarithm of the concentration of the product ON7 against reaction time revealed that the ring-closing is a first-order reaction dependent on the concentration of ON5, similar to the basic ring-closing reactions of ON9 and ON10 (Tables S5–S13 and Figures S9–S17). The relative rate constants for the ring-closing reaction (k ₂) compared to those of the reaction without the target strand (25 °C, pH 7.2) were then calculated (Figure and Table S14). The relative rate constants were all greater than 1, indicating that the ring-closing reaction was accelerated by the presence of target DNAs and RNAs. Given the possibility that hydration strongly influences the ring closure of hydrolyzed Me-TIBNA modified at the 3′-end, this enhancement might be due to dehydration near the 2′-amino and 4′-thioester groups of hydrolyzed Me-TIBNA, induced by duplex formation. The acceleration effect of the ring-closing reaction was more pronounced for RNA than for target DNA. For instance, at 25 °C and pH 7.2, the relative reaction rate constant of RNA-A was approximately 4 times higher than that of DNA-A with the same sequence. This difference can be attributed to the conformation of the nucleoside sugar moiety in the duplex. The sugar conformation of nucleic acids is generally S-type in the DNA duplex and N-type in the RNA duplex. In addition, the sugar configuration of nucleic acids can be analyzed from the coupling constant (J _1′,2′ value) in the ¹H NMR spectrum. The J _1′,2′ value for compound 7, the acidic hydrolysis product of Me-TIBNA, is 9.0 Hz, indicating that the sugar conformation of 7 exists in the S-type. Since the sugar moiety of BNAs, such as AmNA, is fixed in the N-type conformation, the N-type conformation of the sugar moiety of the hydrolyzed Me-TIBNA is crucial for the ring-closing reaction to occur. When ON7 forms a duplex with complementary RNA, the N-type conformation brings the 2′-amino group and thioester in the sugar moiety into proximity (Figure ), potentially aiding the ring-closing reaction.

6.

Acceleration effect of intramolecular cyclization by the addition of target DNA or RNA.

7.

Factor hypothesis that intramolecular cyclization is accelerated by duplex formation with RNA.

Comparing the reaction rates of full-match RNA-A and DNA-A with target RNA and DNA containing single-base mismatches revealed that the presence of mismatched base pairs (T-G, T-C, T-U, and T-T) significantly reduced the reaction rates. This outcome was anticipated to be associated with the stability of the mismatched duplexes, which was evaluated through UV-melting experiments to determine the T _m values of these duplexes. The results, depicted in Figure , reveal an intriguing observation: a higher T _m value did not necessarily correspond to a faster reaction. For instance, RNA-G, with a relatively high T _m value among the target RNAs and DNAs containing mismatches, exhibited a faster reaction in ring closure compared to target RNAs and DNAs containing other mismatch base pairs (T-C, T-U, and T-T), but it was approximately three times slower than RNA-A, which shared the same T _m value as RNA-G. 2′-Carbamoyluridine, an analog of 2′-aminothymidine, stabilizes the U-G mismatch base pair by establishing a hydrogen bond between the carbamoyl group and the amino group at the 2-position of guanine. In addition, when the thermal stability of mismatched duplexes formed by ON1, ON5, and ON7 was evaluated and the obtained T _m values were compared, only ON5 showed no destabilizing effect of base U-G mismatch base pair formation on the duplexes (Table S15). Therefore, the amino group of 2′-aminothymidines may also contribute to stabilizing T-G mismatch base pairing, potentially reducing the nucleophilicity of the amino group at the 2′-position through hydrogen bonding with guanine. These findings suggest that the ring-closing reaction is significantly accelerated when ON7 hybridizes with fully complementary RNA-A. Notably, the relative reaction rate constant in the presence of RNA-A was approximately 1700 at pH 8.0 and 37 °C, indicating that the ring-closing reaction proceeds rapidly through duplex formation with fully matched RNA. Moreover, these results suggest that fully complementary RNAs can induce Me-TIBNA-based oligonucleotide duplex formation switching. The reaction is enhanced by over 500-fold in the presence of RNA-A, even under conditions of pH 7.2 and 37 °C, which closely resemble the in vivo environment, with over 80% of the 4′-thioester-2′-aminothymidine in ON5 converted to AmNA within 6 h (Table S9). This suggests the potential for future control of oligonucleotide function by inducing the conversion of 4′-thioester-2′-aminothymidine to AmNA through the presence of noncoding RNAs such as miRNA, snRNA, and mRNA within the cell.

Verification of the Application of Me-TIBNA-Modified Oligonucleotides in Controlling pH-Responsive Duplex Formation

To demonstrate the utility of Me-TIBNA-modified oligonucleotides in regulating pH-responsive duplex formation, we conducted an experiment using them as fluorescent probes to identify complementary RNA strands. The experiment involved the utilization of ON2 with Me-TIBNA modifications at three sites, complementary RNA-A with TAMRA labeling at the 3′-end, and ON6 and ON8 synthesized following the procedure outlined in Scheme . Additionally, ON14, modified with 2′,4′-BNA, served as a positive control for forming stable duplexes. The fluorescence emission spectra of single-stranded RNA-A and RNA-A mixed with the modified oligonucleotides (ON2, ON6, ON8, and ON14) were measured, as depicted in Figure . Given that TAMRA fluorescence is quenched by guanine at the 5′-end, − duplex formation was detected by changes in fluorescence intensity. Initially, the fluorescence spectra of duplexes formed by Me-TIBNA-modified oligonucleotide (ON2) and TAMRA-labeled RNA-A were measured in neutral phosphate buffer (pH 7.2). The spectrum of the ON2-formed duplex closely resembled that of the ON14-formed duplex, with weaker fluorescence intensity compared to single-stranded RNA-A (Figure A). Conversely, the spectrum of the ON6 and TAMRA-labeled RNA-A mixture in pH 6 phosphate buffer closely resembled that of single-stranded RNA-A (Figure B), suggesting no duplex formation under this condition. Furthermore, the fluorescence emission of the ON8-formed duplex with TAMRA-labeled RNA-A in pH 8 buffer was notably reduced compared to that of single-stranded RNA-A (Figure C). These findings suggest that Me-TIBNA-modified oligonucleotides could serve as responsive probes for detecting target RNA. This is based on their altered duplex formation ability, which results from acid hydrolysis, followed by structural conversion to AmNA through basic intramolecular cyclization. However, it is noteworthy that ON6 and AmNA-modified ON8 used in this experiment were isolated and purified after a 5-day incubation in buffer solution, as shown in Scheme . Therefore, although TIBNAs show promise in regulating oligonucleotide duplex formation in response to pH changes, further structural optimization may be required to enhance the efficiency of acid hydrolysis and basic ring closure in the future.

8.

Fluorescence emission spectra of duplexes formed by oligonucleotides (ON2, ON6, ON8, and ON14) and TAMRA-modified RNA.

Conclusions

In this study, an S-methylated derivative of TIBNA (Me-TIBNA) was developed as an artificial nucleic acid to regulate the duplex-forming capability of oligonucleotides in response to pH changes. Me-TIBNA phosphoramidite was synthesized from AmNA in six steps and then integrated into oligonucleotides. The thioimidate group in Me-TIBNA can be hydrolyzed under mildly acidic conditions to yield 2′-aminothymidine modified with a 4′-thioester. In particular, the resultant 2′-aminothymidine derivative was effectively transformed into AmNA under mild basic conditions, demonstrating the challenging formation of covalent bridges in 2′,4′-BNAs in aqueous environments. The stability of duplexes formed by the modified oligonucleotides was evaluated by using T _m measurements and the BLI method. While Me-TIBNA- and AmNA-modified oligonucleotides can establish stable duplexes, particularly with complementary RNA, the duplex formation of oligonucleotides modified with the hydrolysis product of Me-TIBNA is significantly disfavored both thermodynamically and kinetically. Kinetic analysis of the hydrolysis of Me-TIBNA incorporated into oligonucleotides and the formation of intramolecular amide bonds revealed variations in reaction rates based on the position of the Me-TIBNA modification. Specifically, acidic hydrolysis occurred more rapidly when Me-TIBNA was introduced at the 3′-end rather than within the sequence, whereas basic intramolecular cyclization proceeded faster when Me-TIBNA was modified within the sequence as opposed to at the 3′-end. Surprisingly, the presence of a complementary strand, particularly complementary RNA, was found to accelerate intramolecular cyclization. Notably, the acceleration effect of the complementary strand was highly dependent on its sequence, with fully complementary RNA enhancing intramolecular cyclization by approximately 1700-fold. This suggests that fully complementary RNAs can also act as switches that regulate the duplex-forming capability of Me-TIBNA-modified oligonucleotides. Furthermore, Me-TIBNA-modified oligonucleotides were validated as pH-responsive probes for detecting fluorescently labeled RNA. The findings of this study strongly support the notion that Me-TIBNA-modified oligonucleotides exhibit both excellent duplex-forming ability and pH responsiveness, characteristics that have been challenging to achieve with traditional pH-responsive oligonucleotides. It should be emphasized that we succeeded in forming a bridge ring between the 2′- and 4′-positions of nucleic acids incorporated into the oligonucleotide in response to external stimuli in an aqueous buffer solution. While structural optimization of Me-TIBNA is probably essential for practical applications, TIBNAs represent a potent tool for modulating the hybridization capacity of oligonucleotides in response to pH changes and the presence of target RNA.

Experimental Section

General

All moisture-sensitive reactions were carried out in well-dried glassware under a N₂ atmosphere. The progress of the reactions was monitored by analytical thin-layer chromatography (TLC) on glass plates (TLC silica gel 60 F254) and the products were visualized by UV light. For column chromatography, silica gel PSQ-100B (Fuji Silysia Chemical Ltd.) was used. ¹H, ¹³C, and ³¹P NMR spectra were recorded on JNM-ECS400 and JNM-ECA500 spectrometers. The chemical shift values were reported in parts per million (ppm), relative to internal tetramethylsilane (δ: 0.00 ppm), residual CHCl₃ (δ: 7.26 ppm), CHD₂OD (δ: 3.31 ppm), or CHD₂CN (δ: 1.94 ppm) for ¹H NMR. The chemical shift values were reported in ppm, relative to CDCl₃ (δ = 77.0 ppm), CD₃OD (δ: 49.0 ppm), or CD₃CN (δ: 1.32 ppm) for ¹³C NMR, chemical shift values are reported in ppm, relative to 85% H₃PO₄ (δ: 0.0 ppm) as an external standard for ³¹P NMR. MALDI-TOF mass spectra of new compounds except for compound 4 were recorded on a JEOL SpiralTOF JMS-S3000. FAB mass spectra of compound 4 were recorded on a JEOL JMS-700 MStation. The synthesis of oligonucleotides was performed on a 0.2- and 1.0-μmol scale by using an automated DNA synthesizer (Gene Design nS-8). For high-performance liquid chromatography (HPLC), SHIMADZU CBM-20A, DGU-20A3, LC-20AT, CTO-20A, SPD-20A, and FRC-10A were utilized. Waters XBridge OST C18 2.5 μm (10 × 50 mm) was used for preparative HPLC, and Waters XBridge OST C18 2.5 μm (4.6 × 50 mm) was used for analytical HPLC. For UV absorbance measurement, a SHIMADZU UV-1800 spectrometer was utilized. For liquid chromatography–mass spectrometry (LC/MS) of all new oligonucleotides, a Waters ACQUITY RDa Detector coupled to the ACQUITY UPLC H-Class System and TUV were used, and ACQUITY UPLC oligonucleotide BEH C18 1.7 μm (2.1 × 100 mm) was used. ESI-TOF-MS spectra of all new oligonucleotides were recorded on a Waters ACQUITY RDa Detector. UV-melting experiments were carried out on a SHIMADZU UV-1800 spectrometer equipped with T _m analysis accessory and quartz cuvettes of 1 cm optical path length. Biolayer interferometry (BLI) experiments were performed with an Octet Red96 instrument (Fortebio). Fluorescent spectra were measured using an FP-8500 spectrometer (JASCO) with a quartz cuvette.

Synthesis of Me-TIBNA Phosphoramidite 6

1-[(1R,3R,4R,7S)-6-Oxo-7-((triethylsilyl)­oxy)-1-(((triethylsilyl)­oxy)­methyl)-2-oxa-5-azabicyclo­[2.2.1]­heptan-3-yl]­thymine (2)

Chlorotriethylsilane (3.5 mL, 21 mmol) and imidazole (2.02 g, 29.7 mmol) were added to a solution of AmNA 1 (964 mg, 3.40 mmol) in anhydrous DMF (30 mL) under a N₂ atmosphere at room temperature. The reaction mixture was stirred for 12 h at room temperature. After quenching with methanol at room temperature, the whole mixture was extracted with ethyl acetate. The organic layer was washed with water and brine, dried over Na₂SO₄, and concentrated in vacuo. The residue (1.95 g) was purified by column chromatography (silica gel 50 g, n-hexane/ethyl acetate = 1:1 to 1:2) to afford compound 2 as a white foam (1.22 g, 70%). ¹H NMR (400 MHz, CDCI₃) δ: 0.51–0.61 (6H, m), 0.61–0.71 (6H, m), 0.89 (9H, t, J = 7.9 Hz), 0.99 (9H, t, J = 7.9 Hz), 1.94 (3H, s), 3.96 (1H, d, J = 12.5 Hz), 4.09 (1H, s), 4.14 (1H, d, J = 12.5 Hz), 4.31 (1H, s), 5.42 (1H, s), 6.19 (1H, s), 7.71 (1H, s), 8.61 (1H, s). ¹³C NMR (101 MHz, CDCI₃) δ: 0.1, 4.3, 4.6, 6.6, 6.8, 12.5, 14.3, 55.7, 61.2, 73.1, 86.4, 88.1, 111.0, 135.2, 150.6, 163.9, 173.3. HRMS (MALDI) Calcd for C₂₃H₄₁N₃O₆Si₂ [M + Na]⁺: 534.2432, Found: 534.2426.

1-[(1R,3R,4R,7S)-6-Thioxo-7-((triethylsilyl)­oxy)-1-(((triethylsilyl)­oxy)­methyl)-2-oxa-5-azabicyclo­[2.2.1]­heptan-3-yl]­thymine (3)

Lawesson’s reagent (742 mg, 1.83 mmol) was added to a solution of compound 2 (1.57 mg, 3.06 mmol) in anhydrous THF (10.0 mL) under an N₂ atmosphere at room temperature. The reaction mixture was stirred for 6 h at room temperature. The resulting mixture was concentrated in vacuo and the residue (2.55 g) was purified by column chromatography (silica gel 20 g, n-hexane/ethyl acetate = 1:1) to give compound 3 (754 mg, 47%) as a white foam. ¹H NMR (400 MHz, CDCI₃) δ: 0.50–0.62 (6H, m), 0.64–0.79 (6H, m), 0.88 (9H, t, J = 8.0 Hz), 1.01 (9H, t, J = 8.0 Hz), 1.97 (3H, s), 4.01 (1H, d, J = 12.5 Hz), 4.19 (1H, d, J = 12.5 Hz), 4.32 (1H, s), 4.40 (1H, s), 5.45 (1H, s), 7.75 (1H, s), 8.91 (1H, s), 10.33 (1H, s). ¹³C NMR (126 MHz, CDCI₃) δ: 0.1, 4.4, 4.6, 6.7, 6.9, 12.6, 14.3, 57.2, 65.6, 75.3, 85.7, 92.3, 111.3, 135.2, 150.7, 164.1, 202.4. HRMS (MALDI) Calcd for C₂₃H₄₁N₃O₅SSi₂ [M + Na]⁺: 550.2203, Found: 550.2198.

1-[(1R,3R,4R,7S)-1-(4,4′-Dimethoxytrityloxy)­methyl-7-hydroxy-6-thioxo-2-oxa-5-azabicyclo­[2.2.1]­heptan-3-yl]­thymine (4)

Triethylamine trihydrofluoride (59 μL, 0.36 mmol) was added to a solution of compound 3 (174 mg, 0.329 mmol) in anhydrous THF (3.0 mL) under an N₂ atmosphere at room temperature. The reaction mixture was stirred for 12 h at room temperature. The resulting mixture was concentrated in vacuo and washed with ethyl acetate. The obtained residue (124 mg) was dissolved in anhydrous pyridine (3.0 mL) under a N₂ atmosphere, and 4,4′-dimethoxytrityl chloride (DMTrCl, 172 mg, 0.508 mmol) was added to the solution at room temperature. The reaction mixture was stirred for 5 h at room temperature. After quenching with MeOH, the whole mixture was diluted with ethyl acetate. The solution was washed with water and brine, dried over Na₂SO₄, and concentrated in vacuo. The residue (225 mg) was purified by column chromatography (silica gel 10 g, CHCl₃/methanol = 15:1) to afford compound 4 (149 mg, 75%, 2 steps) as a white foam. ¹H NMR (500 MHz, CD₃OD) δ: 1.45 (3H, s), 3.63 (1H, d, J = 11.7 Hz), 3.70 (1H, d, J = 11.7 Hz), 3.77 (6H, s), 4.31 (1H, s), 4.68 (1H, s), 5.34 (1H, s), 6.83–6.90 (4H, m), 7.20–7.38 (7H, m), 7.47 (2H, d, J = 7.3 Hz), 7.87 (1H, s). ¹³C NMR (76 MHz, CDCI₃) δ: 11.5, 54.4, 58.1, 65.3, 75.2, 85.9, 86.7, 91.7, 109.8, 112.9, 126.8, 127.7, 128.1, 130.1, 135.4, 135.5, 135.7, 144.8, 150.7, 159.0, 159.0, 165.2, 202.4. MS (FAB) Calcd for C₃₂H₃₁N₃O₇S [M + Na]⁺: 624.1780, Found: 624.1799.

1-[(1R,3R,4R,7S)-1-(4,4′-Dimethoxytrityloxy)­methyl-7-hydroxy-6-(methylthio)-2-oxa-5-azabicyclo­[2.2.1]­hept-5-en-3-yl]­thymine (5)

Iodomethane (37 μL, 0.60 mmol) and NaHCO₃ (110 mg, 1.31 mmol) were added to a solution of compound 4 (300 mg, 0.499 mmol) in anhydrous DMF (5.0 mL) under a N₂ atmosphere at room temperature. The reaction mixture was stirred at room temperature for 20 h. After quenching with sat. NH₄Cl at room temperature, the whole mixture was extracted with ethyl acetate. The organic layer was washed with water and brine, dried over Na₂SO₄, and concentrated in vacuo. The residue (410 mg) was purified by column chromatography (silica gel, 15 g, CHCl₃/methanol = 10:1) to give compound 5 (190 mg, 62%) as a white foam. ¹H NMR (500 MHz, CDCl₃) δ: 1.72 (3H, s), 2.44 (3H, s), 2.62 (1H, s), 3.61 (1H, d, J = 11.3 Hz), 3.73 (1H, d, J = 11.3 Hz), 3.79 (6H, s), 4.37 (1H, d, J = 5.9 Hz), 4.71 (1H, s), 5.07 (1H, s), 6.80–6.91 (4H, m), 7.18–7.42 (7H, m), 7.47 (2H, d, J = 8.2 Hz), 7.68 (1H, s), 8.68 (1H, s). ¹³C NMR (126 MHz, CDCl₃) δ 12.7, 12.9, 29.8, 55.3, 57.0, 73.0, 79.4, 84.9, 87.1, 92.6, 110.4, 113.2, 113.4, 127.2, 127.9, 128.1, 128.2, 129.2, 130.2, 130.3, 135.2, 135.3, 135.6, 144.4, 150.2, 158.9, 164.6, 181.5. HRMS (MALDI) Calcd for C₃₃H₃₃N₃O₇S [M + Na]⁺: 638.1937, Found: 638.1938.

1-[(1R,3R,4R,7S)-7-[2-Cyanoethoxy­(diisopropylamino)­phosphinoxy]­1-(4,4′-dimethoxytrityloxy)­methyl-6-(methylthio)-2-oxa-5-azabicyclo­[2.2.1]­hept-5-en-3-yl]­thymine (6)

N,N-Diisopropylethylamine (69 μL, 0.40 mmol), 1-methylimidazole (21 μL, 0.27 mmol), and 2-cyanoethyl-N,N-diisopropylphosphoramidochloridite (59 μL, 0.27 mmol) were added to a solution of compound 5 (82 mg, 3.64 mmol) in anhydrous CH₂Cl₂ (2 mL) under an N₂ atmosphere at 0 °C. The reaction mixture was stirred at room temperature for 3 h. After quenching with sat. NaHCO₃ at room temperature, the whole mixture was extracted with CHCl₃. The organic layer was washed with water and brine, dried over Na₂SO₄, and concentrated in vacuo. The residue (110 mg) was purified by column chromatography (silica gel, 5 g, CHCl₃/ethyl acetate/TEA = 1:1:0.2) to give compound 6 (101 mg, 93%) as a white foam. ¹H NMR (500 MHz, CDCl₃) δ: 0.96 (2H, d, J = 6.7 Hz), 1.03–1.15 (6H, m), 1.27 (8H, t, J = 6.9 Hz), 2.30–2.37 (1H, m), 2.44 (3H, s), 2.49–2.56 (1H, m), 2.72–2.79 (1H, m), 3.41–3.57 (4H, m), 3.74–3.84 (6H, m), 4.08–4.25 (1H, m), 4.40–4.51 (1H, m), 4.78–4.94 (1H, m), 5.12–5.16 (1H, m), 6.81–6.89 (4H, m), 7.22–7.38 (7H, m), 7.43–7.47 (1H, m), 7.71–7.77 (1H, m), 8.05 (1H, s). ¹³C NMR (126 MHz, CDCI₃) δ: 11.3, 11.6, 19.0, 19.6, 21.9, 21.9, 23.5, 23.6, 23.6, 43.2, 45.3, 45.4, 54.4, 56.4, 58.5, 59.0, 71.7, 85.0, 87.0, 92.1, 110.0, 113.0, 113.0, 117.4, 117.9, 127.0, 127.8, 128.1, 130.1, 130.2, 134.9, 135.5, 144.4, 150.6, 159.2, 165.1, 181.6. ³¹P NMR (202 MHz, CDCl₃) δ 150.54, 150.16. HRMS (MALDI) Calcd for C₄₂H₅₀N₅O₈PS [M + Na]⁺: 838.3015, Found: 838.2991.

4′-C-Methylthiocarbonyl-2′-aminothymine (7)

Compound 5 (35 mg, 0.057 mmol) was dissolved in a mixture of MeCN and 10% (v/v) aqueous TFA (1:1, 5.0 mL) at room temperature. The reaction mixture was stirred for 18 h at room temperature. The resulting mixture was concentrated in vacuo. Et₂O was added to the obtained residue (40 mg), and the precipitate was collected by filtration and washed three times with Et₂O to afford compound 7 (16 mg, 85%) as a pale orange solid. ¹H NMR (400 MHz, CD₃OD) δ: 1.89 (3H, s), 2.23 (3H, s), 3.72 (1H, d, J = 11.0 Hz), 3.97 (1H, d, J = 11.0 Hz), 4.03 (1H, dd, J = 6.0, 9.0 Hz), 4.59 (1H, d, J = 6.0 Hz), 6.55 (1H, d, J = 9.0 Hz), 7.79 (1H, s). ¹³C NMR (101 MHz, CD₃OD) δ: 9.7, 11.2, 55.2, 64.3, 72.3, 85.6, 97.3, 111.7, 135.2, 151.4, 164.7, 200.3. MS (FAB) Calcd for C₁₂H₁₇N₃NaO₆S [M + Na]⁺: 354.0736, Found: 354.0738.

1-[(1R,3R,4R,7S)-7-Hydroxy-1-(hydroxymethyl)-6-oxo-2-oxa-5-azabicyclo­[2.2.1]­heptan-3-yl]­thymine (1)

Compound 7 (5.0 mg, 0.015 mmol) was dissolved in 100 mM Tris buffer (pH 9.0, 1.0 mL) at room temperature. The reaction mixture was stirred for 18 h at room temperature. The resulting mixture was purified by reversed-phase HPLC (linear gradient mode of 5–40% MeCN in 0.1 M triethylammonium acetate (TEAA) buffer (pH 7.0) for 30 min) to afford compound 1 (4.5 mg, 58%) as a white solid. The spectral data were found to be consistent with those presented in the literature.

Synthesis of Oligonucleotides

Me-TIBNA-phosphoramidite 6, dT-phosphoramidite (Sigma), Bz-dC-phosphoramidite (Sigma), and iBu-dG-phosphoramidite (Sigma) were dissolved in anhydrous MeCN to a final concentration of 0.1 M. The synthesis of oligonucleotides (ON1–4) was performed on a 0.2-μmol scale in DMTr-off mode by using 0.25 M Activator 42 in MeCN (Sigma) as an activator. 0.1 M CSO in MeCN (Glen Research) was used for the oxidation (oxidation wait time: 3 min). The phosphoramidite 6 was incorporated into oligonucleotides at a prolonged coupling time of 5 min (cf. 30 s). All protecting groups of oligonucleotides were removed by treatment with 28% aqueous NH₃ at 55 °C for 16 h. Removal of NH₃ was carried out in vacuo, and the obtained crude oligonucleotides were purified by reversed-phase HPLC. The compositions of oligonucleotides were confirmed by LC/MS analysis, and the yields were calculated from peak values recorded at a 260 nm wavelength on a microvolume UV–vis spectrophotometer (DeNovix DS-11). Purification conditions in reversed-phase HPLC are shown below.

Mobile phase: 0.1 M TEAA buffer (pH 7.0) and MeCN

Linear Gradient of MeCN in preparative HPLC

ON1 and ON2: 5–25% for 30 min.

ON3 and ON4: 5–20% for 30 min.

Flow rate: 3.0 mL/min.

Column: Waters XBridge OST C18 2.5 μm (10 × 50 mm)

Wavelength of UV for oligonucleotide detection : 260 nm

Temperature of column oven: 50 °C

For preparing ON5–8, the synthesized oligonucleotides (ON1 and ON2, 10 nmol) were dissolved in 100 mM sodium acetate buffer (pH 5.5, 1.0 mL) to give a final concentration of 10 μM for the oligonucleotide. The reaction solutions were incubated at 25 °C for 5 days. The resulting crude oligonucleotides were purified by reversed-phase HPLC to give ON5 (5.9 nmol, 59%) and ON6 (2.2 nmol, 22%), respectively. The obtained ON5 (3.0 nmol) and ON6 (1.0 nmol) were dissolved in 100 mM Tris buffer (pH 9.0, 200 μL) to give a final concentration of 10 μM for the oligonucleotide. The reaction solutions were incubated at 25 °C for 5 days. The resulting crude oligonucleotide was purified by reversed-phase HPLC to give ON7 (2.7 nmol, 91%) and ON8 (0.60 nmol, 60%), respectively. The compositions of oligonucleotides were confirmed by LC/MS analysis, and the yields were calculated from peak values recorded at a 260 nm wavelength on a microvolume UV–vis spectrophotometer (DeNovix DS-11). Purification conditions in reverse-phase HPLC are shown below.

Mobile phase: 0.1 M TEAA buffer (pH 7.0) and MeCN

Linear Gradient of MeCN in preparative HPLC

ON5 and ON6: 5–20% for 30 min.

ON7 and ON8: 8–18% for 30 min.

Flow rate: 1.0 mL/min.

Column: Waters XBridge OST C18 2.5 μm (4.6 × 50 mm)

Wavelength of UV for oligonucleotide detection: 260 nm

Temperature of column oven: 50 °C

UV Melting Experiments

The oligonucleotides (ON1, ON5, ON7, and ON13) and their target DNA or RNA were dissolved in 10 mM sodium phosphate buffer (pH 7.2) containing 100 mM NaCl to obtain a final concentration of 4.0 μM per strand. The melting profiles were recorded at 260 nm from 10 to 90 °C at a scan rate of 0.5 °C/min. The two-point average method was employed to obtain the T _m values, and the final values were determined by averaging three independent measurements, which were accurate within a 1 °C range.

Biolayer Interferometry (BLI) Experiments

BLI experiments were performed with an Octet Red96 instrument (Fortebio). All experiments were performed at 30 °C in a running buffer consisting of 1 × PBSMT (8 mM Na₂HPO₄, 1.46 mM KH₂PO₄, 136 mM NaCl, 2.68 mM KCl, 0.5 mM MgCl₂, 0.05% Tween20). Target oligonucleotides with 5′-Biotin modification (10 nM in 1 × PBSMT) were loaded onto SA sensor tips (Fortebio). The sensor tips were dipped into serially diluted oligonucleotides (ON1, ON5, ON7, and ON13: 2.5, 5, 10, 20, and 40 nM) for association and then transferred into the running buffer for dissociation. The binding time was 300 s, and the dissociation time was 300 s. The dilutions and experiments were performed at 30 °C in the running buffer (1 × PBSMT). All data were analyzed with Fortebio Octet data analysis software using a 1:1 binding model. Oligonucleotides with 5′-Biotin modification were purchased from Ajinomoto Bio-Pharma Services, GeneDesign Inc.

Calculation of Acidic Hydrolysis Reaction Rate Constants

Oligonucleotides (ON3 and ON4) were dissolved in 100 mM sodium acetate buffer (pH 4.0–6.0) to give a final concentration of 10 μM for oligonucleotides. The reaction solutions were incubated at 25 °C. Since the reaction was started, solutions of 10 μL from each sample were picked up, and 0.1 M TEAA buffer (pH 7.0). The amounts of the resulting oligonucleotides (ON9 and ON10) and the remaining intact oligonucleotides (ON3 and ON4) were analyzed and quantified by reverse-phase HPLC.

Calculation of Basic Intramolecular Cyclization Reaction Rate Constants (ON9 and ON10)

Oligonucleotides (ON9 and ON10) were dissolved in 100 mM Tris buffer (pH 8.0–9.0) to give a final concentration of 10 μM for the oligonucleotides. The reaction solutions were incubated at 25 °C. Since the reaction was started, solutions of 10 μL from each sample were picked up and added to 0.1 M TEAA buffer (pH 7.0). The amounts of the resulting oligonucleotides (ON11 and ON12) and the remaining intact oligonucleotides (ON9 and ON10) were analyzed and quantified by reverse-phase HPLC.

Calculation of Intramolecular Cyclization Reaction Rate Constants (Duplex)

ON5 and the target DNA or RNA were dissolved in 10 mM sodium phosphate buffer (pH 7.2, 8.0) containing 100 mM NaCl to give a final concentration of 4.0 μM for each strand. The reaction solutions were incubated at 25 or 37 °C. Since the reaction was started, solutions of 10 μL from each sample were picked up and added to 0.1 M TEAA buffer (pH 7.0). The amounts of the resulting ON7 and the remaining ON5 were analyzed and quantified by reverse-phase HPLC.

Measurement of Fluorescence Spectra of Duplexes Formed by Fluorescently Labeled RNA

For ON2, ON14, and the TAMRA-labeled RNA, oligonucleotides were dissolved in 10 mM sodium phosphate buffer (pH 7.2) containing 5 mM NaCl to obtain a final concentration of 50 nM per strand. For ON6, ON14, and the TAMRA-labeled RNA, oligonucleotides were dissolved in 10 mM sodium phosphate buffer (pH 6.0) containing 5 mM NaCl to obtain a final concentration of 50 nM per strand. For ON8, ON14, and the TAMRA-labeled RNA, oligonucleotides were dissolved in 10 mM sodium phosphate buffer (pH 8.0) containing 5 mM NaCl to obtain a final concentration of 50 nM per strand. The fluorescence emission spectra were measured from 400 to 750 nm at a scan rate of 200 nm/min under light irradiation conditions of 550 nm to excite TAMRA. Figure shows an excerpt of the measured spectrum in the region from 560 to 750 nm.

The data underlying this study are available in the published article and the Supporting Information.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.5c10817.

• NMR spectra of new compounds, LC/MS charts, HPLC charts, Data for BLI experiments (PDF)

T.O. conceived the study and wrote and edited the manuscript. K.M., S.I., and K.U. performed oligonucleotide synthesis and analysis. S.O. supervised the study. S.O. and T.O. provided financial support. All the authors read the manuscript and agreed to its contents.

The authors declare no competing financial interest.