Core-Regenerating Atom-Pair Exchange Enables Skeletal Carbon Isotope Incorporation into Pyridines

Abstract

We report a core-regenerating atom pair exchange strategy for skeletal carbon isotope incorporation into pyridines. The two-step sequence proceeds through pyridine dearomatization followed by a [4 + 2]/retro-[4 + 2] cycloaddition cascade with an isotope-labeled sulfonyl cyanide reagent, enabling selective replacement of a pyridine carbon atom with an isotopically labeled counterpart. This transformation addresses a longstanding challenge in the preparation of carbon-isotope-labeled pyridines, which are valuable for medicinal chemistry and ADMET (absorption, distribution, metabolism, excretion, and toxicity) studies. The method exhibits high regioselectivity and broad tolerance of diverse pyridine substitution patterns, providing rapid access to a wide range of ¹³C-labeled pyridine building blocks and bioactive molecules.

Isotope-labeled compounds are indispensable tools for tracking and quantifying molecular behavior in complex biological systems. − Owing to the high prevalence of pyridine among heterocycles in FDA-approved drugs − and agrochemicals, − its isotopologs play critical roles in ADMET (absorption, distribution, metabolism, excretion, and toxicity) and pharmacokinetic studies. − Efficient methods for synthesizing isotope-labeled pyridines are therefore important for pharmaceutical and agrochemical development.

Pyridine contains three elements amenable to isotope labeling including hydrogen, nitrogen, and carbon. Among these, peripheral hydrogen isotope labels are most commonly employed , due to the availability of diverse hydrogen isotope exchange (HIE) methods. − However, isotopic enrichment can vary depending on the substrate and methods used, and the resulting labels are susceptible to exchange with protons under biological conditions, leading to a label loss and compromised fidelity in metabolic studies. Nitrogen isotope labels have broad applications. − However, their utility for radiolabeling is limited because of the lack of long-lived radionuclides. In contrast, carbon isotopes embedded within the pyridine core are metabolically robust and ideally suited for both stable-isotope tracing (¹³C) , and radiolabeling (¹⁴C). ,,, Despite these advantages, access to carbon-isotope-labeled pyridines primarily relies on de novo synthesis from a limited set of isotope-enriched precursors, resulting in costly, inflexible routes. Direct incorporation of carbon isotope labels into pyridine rings, ideally at a late stage, remains a longstanding synthetic challenge. −

Skeletal editing via temporary ring disruption followed by connectivity-retaining reconstruction (i.e., core-regenerating skeletal editing) represents a powerful approach for the isotope labeling of heterocycles. This transformation is not accessible through prevailing editing strategies that modify the intrinsic topology of the parent scaffold (Figure A). − For instance, independent reports from Smith et al., Sarpong et al., McNally et al., and Audisio et al. in 2024 established ring-opening/closing strategies for ¹⁵N labeling of pyridines (Figure B), with the studies by Sarpong et al. and Audisio et al. further extending the scope to pyrimidines and other relevant heterocycles. In parallel, McNally et al. reported in 2025 a deconstruction-reconstruction strategy that enables efficient access to a range of higher-mass pyrimidine isotopologs via ring reconstruction with isotope-enriched amidines (Figure C). Most recently, a C2-selective ¹³C incorporation of indoles was reported by Zeng and Li and co-workers. Despite these advances, skeletal carbon isotope incorporation into pyridines remains underdeveloped.

1.

(A) Core-regenerating skeletal editing strategies. (B) Core-regenerating approaches to ¹⁵N labeling of pyridines. (C) Core-regenerating approach to isotope labeling of pyrimidines. (D) This work: core-regenerating atom-pair exchange for skeletal carbon isotope incorporation into pyridines.

Herein, we report a core-regenerating atom-pair exchange strategy that enables pyridine skeletal carbon isotope incorporation (Figure D). As outlined in Figure D, the strategy proceeds via a two-step sequence: (1) dearomatization of a pyridine (1) through nucleophilic 1,2-addition to generate a dihydropyridine intermediate (2) and (2) [4 + 2] cycloaddition with an isotope-enriched sulfonyl cyanide (3), followed by tandem retro-[4 + 2] cycloaddition that yields a C2-sulfonylated, isotope-incorporated product (4). During this process, the native C–N atom-pair is replaced by the incoming cyano group and expelled as an imine byproduct (5). The resulting C2-sulfonyl groups are important pharmacophores in numerous bioactive pyridines. They also serve as versatile functional handles that can be diversified via S_NAr or cross-coupling reactions , to access various functionalized pyridines, or removed reductively to enable traceless carbon isotope labeling. Although dihydropyridines have been used for cycloaddition with diverse dienophiles such as alkenes, alkynes, , nitroso, and azo compounds, their reactivity with sulfonyl cyanide for pyridine synthesis has not been reported. We hypothesized that electronic complementarity between the diene and sulfonyl cyanide would dictate the regioselectivity required to preserve the native pyridine connectivity. Additionally, we anticipated that the substituent R in dihydropyridine 2 would play a key role in modulating the driving force for the retro-[4 + 2] cycloaddition, with greater π-stabilization of the expelled imine 5 rendering the process more favorable.

Our investigation began with identifying optimal conditions for the [4 + 2] cycloaddition cascade using unlabeled sulfonyl cyanides. 4-(Trifluoromethyl)­pyridine (6) was selected as the model substrate and converted to dihydropyridines 7–10 under the corresponding dearomatization conditions (see Supporting Information). After subjecting to p-toluenesulfonyl cyanide (TsCN) at 110 °C in toluene for 24 h, 7 and 8 afforded the desired C2-sulfonylated product 11 in 62% and 47% yield, respectively, whereas 9 and 10 gave significantly lower yields (Figure , entries 1–4). This substituent dependence supports our hypothesis that π-stabilization of the expelled imine is critical for effective retro-[4 + 2] cycloaddition.

2.

Reaction development. Reactions were performed on a 0.1 mmol scale based on the limiting reagent. Concentration refers to the limiting reagent. ^a Yields refer to the C2-sulfonylated pyridines derived from the corresponding sulfonyl cyanides (TsCN and 12–16) and were determined by ¹H NMR analysis using nitrobenzene as an internal standard. Yields were calculated based on the limiting reagent. See Supporting Information for full experimental details.

Dihydropyridine 7 was used for further reaction optimization. A series of aryl and alkyl sulfonyl cyanides (12–16) was examined, affording corresponding C2-sulfonylated pyridines in yields comparable to TsCN (Figure , entries 5–9). These results indicate that the reaction is insensitive to the electronic and steric properties of the sulfonyl substituent. TsCN was therefore selected for the subsequent studies due to its commercial availability. Increasing or decreasing the reaction temperature resulted in lower yields (Figure , entries 10 and 11). Increasing the loading of TsCN did not affect the reaction outcome (Figure , entry 12). In contrast, increasing the amount of dihydropyridine to 1.5 and 2.0 equiv improved the yield to 82% and 86%, respectively, with TsCN as the limiting reagent (Figure , entries 13 and 14). We attribute this result to competing reactions between the dihydropyridine and the expelled imine byproduct, which consumes a portion of the dihydropyridine and necessitates its use in excess to achieve full conversion of TsCN. Notably, phenylpropiolaldehyde (not shown), derived from the expelled imine byproduct, was isolated from the crude reaction mixture after acid treatment. This observation is consistent with the proposed [4 + 2]/retro-[4 + 2] cycloaddition mechanism (see Supporting Information section 5.2). Balancing product yield and material economy, 1.5 equiv of dihydropyridine was selected as the standard condition.

With suitable conditions established, we evaluated the substrate scope across pyridines bearing diverse substitution patterns. Cu- and Zn-mediated alkynylation was employed as the primary dearomatization strategy; when it failed to deliver dihydropyridine intermediates that afforded satisfactory yields in the subsequent cycloaddition, Zn-mediated phenylation was employed instead. Notably, organozinc reagents consistently provide higher yields in the 1,2-addition of pyridine compared with commonly employed Grignard reagents.

A range of C4-substituted pyridines were converted to the corresponding dihydropyridines via Cu- and Zn-mediated alkynylation, which were subjected to TsCN after purification. A variety of electron-withdrawing, electron-neutral, and electron-donating groups was well tolerated, affording the corresponding C2-sulfonylated products (11, 17–27) in moderate to good yields (Figure A).

3.

Reaction scope. Dearomatization condition A: pyridine (1.0 equiv), phenylacetylene (1.5 equiv), methyl chloroformate (3.0 equiv), CuI (0.20 equiv), N,N-diisopropylethylamine (3.0 equiv), CH₂Cl₂, 0 °C to rt. Dearomatization condition B: phenylacetylene (1.3 equiv), MeMgBr (1.2 equiv), ZnCl₂ (1.3 equiv), THF, 0 °C, then pyridine (1.0 equiv), methyl chloroformate (1.3 equiv), 0 °C to rt. Dearomatization condition C: PhMgBr (1.2 equiv), ZnCl₂ (1.3 equiv), THF, 0 °C, then pyridine (1.0 equiv), methyl chloroformate (1.3 equiv), 0 °C to rt. [4 + 2]/retro-[4 + 2] cascade conditions: dihydropyridine (1.5 equiv), TsCN (1.0 equiv), toluene, 110 °C, 24 h. Yields are isolated and calculated based on the limiting reagent, TsCN. (A) Substrate scope of C4-substituted pyridines. (B) Substrate scope of C3-substituted pyridines. (C) Substrate scope of C2-substituted and multisubstituted pyridines. ^a Reaction performed using a 1.0:1.1 mixture of 29a and 29b. ^b Yields refer to the combined isolated yield of all regioisomeric products obtained as an inseparable mixture after dearomatization. ^c Reaction performed using the inseparable regioisomeric mixture of dearomatization products. See Supporting Information for full experimental details.

We next examined the substituent effect at the C3 position (Figure B). Cu-mediated alkynylation of nikethamide (28) furnished a 1.0:1.1 mixture of regioisomeric dihydropyridines 29a and 29b, which were separated and evaluated independently in their reaction with TsCN. Isomer 29a decomposed without forming 30 or any other identifiable products, whereas 29b furnished 31 in 59% yield. We rationalize that steric repulsion between the incoming sulfonyl group and the adjacent amide disfavors the formation of 30, as depicted in transition state 32. Interestingly, direct exposure of the unseparated 1.0:1.1 mixture of 29a and 29b to TsCN furnished 31 in 49% yield. This result suggests that although 29a does not undergo productive cycloaddition, it may suppress competing side reactions that divert 29b, thereby improving the effective yield of 31. Similarly, dearomatization of ethyl nicotinate produced an inseparable mixture of regioisomers that, upon treatment with TsCN, delivered 33 in 58% yield. Compounds 34–37 were obtained from the corresponding C3-substituted pyridines via phenylzinc-mediated dearomatization followed by [4 + 2] cycloaddition cascade. For these substrates, the Cu- and Zn-mediated alkynylation exhibited diminished regioselectivity, affording lower proportions of the desired C6-addition products.

Finally, C2-substituted and multisubstituted pyridines were evaluated, affording the corresponding products (38–45) under substrate-optimized dearomatization conditions (Figure C). These results further demonstrate the broad compatibility of the method across diverse substituent patterns.

It should be noted, however, that the success of the atom-pair exchange process depends on the efficient formation of dihydropyridine intermediates capable of undergoing the subsequent [4 + 2] cycloaddition. Representative unsuccessful substrates are summarized in the Supporting Information.

We next leveraged the method to access ¹³C-labeled pyridine derivatives, showcasing its efficiency and broad utility for the rapid preparation of isotope-labeled compounds relevant to ADMET studies. The requisite Ts¹³CN was prepared from K¹³CN via a concise two-step sequence involving thiocyanate formation followed by oxidation mediated by TFAA and urea·H₂O₂ complex (Figure A). Notably, the route also enabled the synthesis of dual-labeled Ts¹³C¹⁵N and may provide a potential approach for the future preparation of radioactive Ts¹⁴CN using the corresponding isotope-enriched KCN.

4.

(A) Synthesis of Ts¹³CN. (B) Access to ¹³C-labeled pyridine building blocks relevant to the synthesis of bioactive molecules. (C) ¹³C-labeling of pyridines via Zn-mediated desulfonylation. (D) ¹³C-labeling of pyridines via ferrioxalate-mediated photochemical desulfonylation. See Supporting Information for full experimental details. TBAOx, tetrabutylammonium oxalate.

With Ts¹³CN in hand, a series of pyridines was converted to the ¹³C-incorporated sulfonylated products in yields comparable to those obtained without isotope label. These intermediates were further transformed via sulfone-mediated S_NAr reactions into valuable ¹³C-labeled analogs of building blocks previously used in the synthesis of bioactive molecules (Figure B). For instance, [ ¹³ C]­11 was prepared from 4-(trifluoromethyl)­pyridine (6) and subsequently subjected to thiophenol 47 to yield the C2-thiolated product [ ¹³ C]­48, a labeled analog of the key intermediate en route to potent monoacylglycerol lipase (MAGL) inhibitor 49. Similarly, [ ¹³ C]­50–[ ¹³ C]­52, which are ¹³C-labeled analogs of intermediates to access other bioactive compounds, were successfully prepared using carbon- oxygen-, and nitrogen-centered nucleophiles in the S_NAr reactions, respectively. These results highlight the utility of the method for the rapid assembly of structurally diverse carbon-isotope-labeled pyridines. They also highlight the synthetic versatility of C2-sulfonylated pyridines.

The challenging traceless skeletal ¹³C labeling of bioactive pyridines was also achieved. The key to success lies in identifying mild conditions for desulfonylation that traditionally required strong reducing reagents, such as sodium–mercury amalgam, which exhibit poor functional group tolerance. We found that Zn metal in the presence of acetic acid smoothly promotes the desulfonylation of pyridines bearing phenyl or ester substituents (Figure C). For instance, pyridines 53 and 54 were converted to the ¹³C-labeled analogues after atom-pair exchange transformation and the Zn-mediated desulfonylation. Furthermore, dual ²H/¹³C labeling was achieved by employing AcOD in the desulfonylation step, as demonstrated by the synthesis of the labeled lipid-lowering reagent etofibrate ([ ² H ¹³ C]­55), which was obtained with 80% deuterium incorporation.

In addition, we found that the ferrioxalate-mediated photoreduction developed by Juliá et al. is highly effective for the desulfonylation of pyridines that are unreactive or decompose under alternative conditions (Figure D). For example, in the sequence for ¹³C-labeling of nikethamide (28), ferrioxalate-mediated conditions provide the desulfonylated product [ ¹³ C]­28 in 74% yield, whereas alternative conditions involving metal Zn, Mg, SmI₂, or sodium–mercury amalgam led to significant decomposition of the desired product. Similarly, ¹³C-labeling of O-protected tropicamide (56) and bisacodyl (57) was achieved using the ferrioxalate-mediated desulfonylation, whereas other reduction conditions failed to afford the desired products. The successful labeling of 58 further demonstrated the compatibility of the photocatalytic desulfonylation with a disubstituted pyridine substrate. Notably, dual ¹³C/¹⁵N labeling of 59, a cholesterol 24-hydroxylase (CH24H) inhibitor, was accomplished using Ts¹³C¹⁵N in the cycloaddition step. All labeled pyridine products were obtained with >99% ¹³C isotopic enrichment without noticeable isotope dilution. Together, these results demonstrate the robustness of the core-regenerating atom-pair exchange strategy for traceless skeletal carbon isotope labeling of pyridines.

In conclusion, we have developed a core-regenerating atom-pair exchange strategy for skeletal carbon isotope incorporation into pyridines. The method proceeds via pyridine dearomatization, followed by [4 + 2] and retro-[4 + 2] cycloadditions with an isotopically labeled sulfonyl cyanide reagent. It exhibits high regioselectivity and broad tolerance of pyridine-substituent patterns. Its synthetic versatility is demonstrated by the rapid synthesis of diverse ¹³C-labeled pyridine building blocks and by the challenging late-stage skeletal ¹³C and ¹³C¹⁵N labeling of drug molecules, enabling access to high-value derivatives for ADMET studies. This work advances core-regenerating atom-pair exchange as a robust platform for aromatic editing and highlights its potential for the future development of radioactive carbon-isotope-labeling strategies for pyridines.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacs.6c11666.

• General methods, synthetic procedures, product isolation and characterization, and NMR spectra (PDF)

All authors have approved the final version of the manuscript.

The authors declare no competing financial interest.