Chemical Tagging of N-Alkylamine-Containing Natural Products and Pharmaceuticals through C(sp³)–H Functionalization

Abstract

The ability to introduce functional groups suitable for bioconjugation into natural products and pharmaceuticals in an efficient, chemo- and site-selective manner is critical for advancing drug discovery, biological chemistry, and targeted therapy. Late-stage C–H functionalization enables direct structural diversification of lead compounds, but broader application is constrained by the scarcity of catalysts that selectively activate relatively inert C–H bonds in polyfunctional drug compounds. Here, we report a photoinduced, flavin-catalyzed method that installs bioconjugation handles on a broad range of bioactive N-alkylamines. The transformation proceeds under mild, aerobic conditions with blue-light irradiation: flavin analogues mediate sequential α- and β-amino C–H bond scission to generate enamines, which undergo inverse electron-demand Diels-Alder reactions with tetrazine- or nitroalkene-based tagging agents. We assessed biological consequences by cytotoxicity and ligand-binding assays, revealing examples in which tagging either preserves or alters bioactivity. The method’s versatility is demonstrated by the efficient synthesis of antibody–drug conjugates derived from the anticancer agent irinotecan.

Graphical Abstract

1. INTRODUCTION

Incorporating bioconjugation handles into natural products and pharmaceuticals is an essential strategy in drug discovery and biological chemistry, enabling the construction of antibody–drug conjugates (ADCs), bifunctional drugs, and molecular probes.^1–14 Representative examples include the arming of DM1, a semi-synthetic maytansinoid analogue bearing a thiol, with an N-hydroxysuccinimide (NHS) ester by thiol-Michael addition en route to trastuzumab emtansine (1a, Figure 1A),⁶ and a dasatinib–diazirine–alkyne probe for photoaffinity chemoproteomics (1b), assembled through carbamate formation from an alcohol unit within dasatinib.⁷ Such derivatizations expand therapeutic modalities by permitting covalent attachment of small molecules to proteins, nucleic acids, or other ligands, thereby enhancing target specificity or enabling identification of drug-binding proteins.^2,4–7,14

Fig. 1. Chemical tagging of bioactive N-alkylamines by late-stage functionalization.

(A) Chemical tagging of bioactive compounds by transformations of polar functional groups. (B) (C) State-of-the-art chemical tagging methods through transformations of C–H bonds positioned α- to a heteroatom (N, S). (D) Chemical tagging of bioactive amines by transformations of α- and/or β-amino C(sp³)–H bonds.

However, conventional tagging strategies, despite their widespread application, predominantly exploit native functional groups (e.g., alcohols, thiols, amines, carboxylic acids).^4,17 This constrains conjugation sites and often requires synthesis of new analogues when those handles are absent in the lead compounds (e.g., DM1, Figure 1A).⁶ Late-stage functionalization of ubiquitous, otherwise-inert C–H bonds in lead compounds offers a complementary solution: it enables direct introduction of bioconjugation handles at positions inaccessible to classical functional-group interconversion (Figure 1B–1D).^8–20 Key advances include triazolinedione-mediated, tyrosine-selective peptide tagging and Minisci-type modification of N-heteroarenes.^8–10 Notable methods targeting C(sp³)–H bonds are L_nRh–nitrenoid insertion for α-amino C–H functionalization of alkaloids (2a + 3a → 4a, Figure 1B), and Giese-type radical addition for methionine-selective peptide modification (2b + 3b → 4b, Figure 1C).^11,12 These approaches exploit the electronic and steric accessibility of α-to-heteroatom C–H bonds: electron donation from nitrogen or sulfur stabilizes intermediates such as a carbon with partial positive charge (I) or an α-thioradical (II).^{11,12,21–37} Despite these advances, extending tagging to more remote, less-activated methylene C(sp³)–H bonds remains a major challenge.^21–53

To expand the range of taggable drugs and accessible arming sites, we sought to develop a catalyst/tagging-agent combination capable of functionalizing α- and/or β-C(sp³)–H bonds within saturated, nitrogen-containing heterocycles (Figure 1D). Cyclic tertiary amines are common in FDA-approved small-molecule drugs: among 321 approvals from 2013–2023, 40 contain piperidine, 40 pyrrolidine, and 13 morpholine.⁵⁴ Despite this prevalence, few catalyst systems can efficiently cleave α- or β-amino C(sp³)–H bonds under mild conditions (i.e., without strong oxidants, acids, bases, or elevated temperatures) while tolerating diverse functional groups. This balance is crucial, as most bioactive molecules contain multiple Lewis acid- or base-sensitive functionalities (e.g., alcohols, carboxylic acids, amines, heteroarenes) that can deactivate catalysts or undergo side reactions under harsher conditions.^55–59 In chemical tagging, the added requirement that the catalyst tolerate bioconjugation handles within the tagging agent further compounds this challenge. A particularly demanding aspect is the identification of tagging agents that react selectively with intermediates generated in situ by amino C(sp³)–H bond scission while avoiding undesired reactions with native nucleophilic or electrophilic sites. For example, we previously developed an (F₅C₆)₃B/Brønsted-base catalyst system that converts bioactive amines into enamines; however, this method was limited to amine substrates lacking protic functional groups and was therefore incompatible with introducing bioconjugation handles through α- and/or β-C(sp³)–H bonds in saturated, nitrogen-containing heterocycles.^42–44 These limitations underscore the need for a milder and more broadly functional-group-tolerant approach.

We therefore envisioned a robust platform for chemical tagging of α- and/or β-amino C(sp³)–H bonds in bioactive amines using 1.0–2.5 mol% flavin (FAD) analogues (C1) under blue-LED irradiation, paired with tetrazines (6) or nitroalkenes (8) as selective tagging agents (Figure 1D). Tagging of amine 5 likely proceeds as follows (Figure 2). Photoexcitation of the FAD analogue (e.g., C1) furnishes the corresponding excited state species (C1 → IV). A single-electron transfer (SET) from the tertiary amine to the triplet-state flavin generates a nitrogen-centered radical cation,^60–63 which upon deprotonation and a second SET delivers an iminium species.^62–65 Subsequent deprotonation of the iminium by an anionic form of the catalyst affords the enamine and a reduced FADH₂ analogue (VI → VII + VIII), and reoxidation of the reduced flavin by aerobic O₂ regenerates C1 to close the catalytic cycle. The resulting enamines can undergo selective reactions with different tagging agents: 1,2,4,5-tetrazines (6) engage in inverse-electron-demand Diels-Alder reactions with the enamines (VIII → IX),^66–70 furnishing fused intermediates that extrude N₂ (IX → X) and ring-open to give 1,2-diazines (7a) suitable for bioconjugation. Alternatively, nitroalkenes (8) may add to the enamines (VIII → XI → XII) and, after reduction, deliver tagged analogues (9) while preserving the N-heterocyclic core.^47,71

Fig. 2. Plausible reaction mechanisms.

Photoactivated flavin catalyzes the oxidation of bioactive N-alkylamines. The ensuing union of the resulting enamine with a tetrazine or a nitroalkene affords the tagged products.

Flavin derivatives are known to oxidize N,N-dialkylamines to electrophilic imines (e.g., proline → 1-pyrroline-5-carboxylate) through cleavage of a hydridic α-amino C–H bond and a protic N–H bond.^62,63 However, the use of flavin catalysis to convert tertiary amines into nucleophilic enamines (5 → VIII) has not been demonstrated, particularly in the context of polyfunctional drug molecules. A further practical advantage of our FAD-catalyzed approach is that aerobic O₂ serves as the terminal oxidant to regenerate the catalyst (used in 1.0–2.5 mol %), in contrast to other photoactivated FAD systems that require inert conditions to prevent undesired reactions with O₂-derived species (e.g., Figure 1C; see the Supporting Information).^{12,13,22–25,72} Moreover, while seminal studies by Sauer, Boger, and others have established the reactivity of tetrazines with enamines formed via amine–carbonyl condensation,^66,67 our strategy uniquely extends tetrazine chemistry to enamines generated by the oxidation of bioactive, polyfunctional N-alkylamines. Thus, we introduce a photoexcited flavin-based chemical tagging platform that chemoselectively transforms α- and/or β-amino C(sp³)–H bonds in pharmaceuticals and natural products containing other oxidation-prone moieties.

2. RESULTS and DISCUSSION

We describe the development of a catalytic method for C–H functionalization of bioactive N-alkylamines, exemplified by irinotecan (5c), an essential medication for colon and small-cell lung cancers (Table 1).⁷³ To identify an effective catalyst, we evaluated various candidates using 5c and a N-hydroxysuccinimide (NHS)-ester-substituted tetrazine (6c) to obtain the tagged product (7c). Naturally occurring flavin analogues, such as lumiflavin (C2) and riboflavin (C3), proved ineffective mainly due to their limited solubility in CH₂Cl₂. In contrast, the N-(2-methoxyethoxy)ethyl-substituted flavin (C4) successfully catalyzed the reaction, affording 7c in 46% yield. Further enhancements were achieved with catalysts containing more electron-deficient arene moieties. Halo-substituted catalysts such as Cl-(C5), Br-(C6), and I-(C7) provided 7c in 60%, 75%, and 50% yields, respectively. These yield improvements likely stem from the heavy atom effect and the increased oxidation potentials of these catalysts (C5–C7: E_1/2(PC*/PC) = 1.88–1.89 V vs. C4: E_1/2(PC*/PC) = 1.65 V).⁷⁴ The beneficial role of electron-withdrawing groups is further supported by the stark contrast in yields of 7c between the para-methoxyphenyl-substituted catalyst (C8, <3% yield) and the para-cyanophenyl-substituted catalyst (C9, 37% yield). Control experiments demonstrated that both photoirradiation and the catalyst were essential for the reaction, as the absence of either resulted in the complete recovery of 5c (see the Supporting Information).

Table 1.

Screening Studies to Identify an Effective Catalyst^a

^aConditions: irinotecan (5c, 0.05 mmol), tetrazine (6c, 0.10 mmol), catalyst (C2–C9, 1.0 mol %), CH₂Cl₂ (7 mL), under irradiation with blue LED strip lights (465 nm), under air, 22 °C, 24 h. See the Supporting Information for details on the determination of redox potential values.

The abovementioned catalytic process can be performed with 6-methyl-3-aryl tetrazines (6c–6i) featuring a wide variety of bioconjugation handles as the tagging agent (Table 2). This includes compounds bearing a N-hydroxysuccinimide ester or N-Boc-amine, applicable to amine-selective crosslinking (7c, 75% yield; 7i, 51% yield), a ketone that can be utilized for oxime ligation (7d, 70% yield), an azide or an alkyne for CuAAC (7e, 64% yield; 7f, 54% yield), as well as an aryl–boronic ester and an aryl–iodide suitable for cross-coupling (7g, 55% yield; 7h, 58% yield). Notably, C–H functionalization of 5c occurred exclusively at the N-alkyl piperidine unit, leaving intact the aryl piperidine-1-carboxylate group, which is prone to oxidation in other photocatalytic systems.^25,30,32,39 This selectivity aligns with the exclusive engagement of 1,2,4,5-tetrazine (6) in inverse electron demand Diels-Alder cycloadditions with the electron-rich enamine derived from 5c. The reaction proceeds efficiently in air and under ambient temperature. Under such milder conditions, the reaction demonstrated broad functional group tolerance, accommodating alcohol, ester, amide, and pyridine moieties present in 5c, as well as various bioconjugation handles that constitute the tagging agents (6c–6i), underscoring the robustness and versatility of the catalytic system.

Table 2.

Incorporation of Different Bioconjugation Handles^a

^aConditions: irinotecan (5c, 0.05 mmol), tetrazine (6c–6i, 0.10 mmol), C6 (1.0 mol %), CH₂Cl₂ (7 mL), under irradiation with blue LED strip lights (465 nm), under air, 22 °C, 24 h.

We sought an alternative class of tagging agents capable of retaining the piperidine ring of irinotecan (5c) after chemical tagging (Table 3). Nitroalkenes bearing a ketone, an azide, a silyl ether, or a maleimide group (8a–8d) reacted exclusively with the enamine intermediate derived from 5c. With riboflavin 2’,3’,4’,5’-tetraacetate (C10, RFTA; 2.5 mol %) as the photocatalyst, the reactions of 5c with 8a–8d afforded enamine adducts (XIIa–XIId), which were isolated after silica gel chromatography in 32–52 % yield (see the Supporting Information for details). Subsequent borohydride reduction of XIIa–XIId furnished the corresponding tagged analogues 9a–9d.

Table 3.

Arming of Irinotecan with Nitroalkene-Based Tagging Agents^a,b

^aConditions: irinotecan (5c, 0.05 mmol), nitroalkene (8a—8d, 0.10 mmol), C10 (2.5 mol %), CH₂Cl₂ (1 mL), under irradiation with blue LED strip lights (465 nm), under air, 22 °C, 3 h. Resulting enamine (XIIa—XIId) was purified by flash silica gel column chromatography.

^bConditions for reduction: enamine (XIIa—XIId), ZnCl₂ (1.0 equiv.), NaCNBH₃ (1.5 equiv.), methanol (0.5 mL), under air, 22 °C, 4 h.

The applicability of our strategy is highlighted by the efficient synthesis of antibody–drug conjugates from irinotecan (5c; Figure 3). As shown in Table 2, our tagging protocol enables incorporation of diverse modifiable units into irinotecan (5c → 7c–7i), facilitating rapid evaluation and selection of chemo- and site-selective bioconjugation methods. Using a photoactivated flavin-catalyzed reaction, we directly tagged 5c with β-lactam (6j) to afford linker–payload 7j in 63 % yield (Figure 3). The resulting 7j is primed for site-specific conjugation via a dual-variable–domain (DVD) monoclonal antibody (mAb) platform, enabling modular “plug-and-play” targeting.^75–79 As a proof-of-concept, HER2- and CD79b-targeting DVDs were each conjugated to 7j, yielding ADCs 10j and 11j with drug-to-antibody ratios of 1.9 and 2.0, respectively, as determined by catalytic-methodol assay.⁸⁰ Their identities were confirmed by intact mass spectrometry (see Figures S7–2, S7–3, and S8–3 in the Supporting Information for details regarding the characterization and preliminary biological evaluation of the ADCs). The latter single-step conversion of 5c to 7j exemplifies our chemical tagging strategy’s efficiency and functional-group tolerance, offering a robust route to ADCs against diverse antigens and laying the groundwork for future biological evaluation and development.

Fig. 3. Synthesis of an antibody–drug conjugate through sequential chemical tagging and bioconjugation.

Various bioactive N-alkylamine-based natural products and pharmaceuticals (5k–5y; Tables 4–5, and Figure 4) were efficiently modified by the C6-catalyzed process, affording tagged analogues 7k–7z. Saturated N-heterocyclic fragments, including piperidines (5k, 5l, 5o–5q, 5u–5w, 5y), pyrrolidines (5m, 5n, 5t), and morpholines (5r, 5s, 5w), served as suitable attachment sites. In addition to trialkylamines (5k–5s), we prepared tagged derivatives arising from oxidation of N-aryl pyrrolidine and piperidine moieties (7t–7v, 7x–7z).

Table 4.

Tagging of Different Bioactive N-Alkylamines^a

^aConditions: amine (5k-5p, 0.05 mmol), tetrazine (6c, 6e, 6l, 6n, 0.10–0.15 mmol), catalyst (C6, 1.0–2.5 mol %), CH₂Cl₂ (5–7 mL), under irradiation with blue LED strip lights (465 nm), under air, 22 °C, 12–36 h. See the Supporting Information for details.

Table 5.

Tagging of Different Bioactive Tertiary Amines^a

^aConditions: amine (5q-5v, 0.05 mmol), tetrazine (6l, 6q, 6u, 0.10–0.15 mmol), catalyst (C6, 1.0–2.5 mol %), CH₂Cl₂ (5–7 mL), under irradiation with blue LED strip lights (465 nm), under air, 22 °C, 12–36 h. See the Supporting Information for details.

Fig. 4. Chemical Tagging of Y-320 and c-Fms-In-13.

Conditions: amine (5w, 5y, 0.05 mmol), tetrazine (6q, 6u, 0.10 mmol), catalyst (C6, 2.5 mol %), 10% DMSO in CH₂Cl₂ (5 mL) or CH₂Cl₂ (5 mL), under irradiation with blue LED strip lights (465 nm), under air, 22 °C, 24 h. See the Supporting Information for details.

For bemcentinib (5n), a diazirine–alkyne tag, widely used as a photoaffinity label in proteomics, was readily incorporated to give 7n in 88% yield.⁷ For the tagging of 5l and 5o–5y, we employed pyridin-2-yl-substituted tetrazines, which undergo particularly facile cycloaddition relative to phenyl-substituted tetrazines; this enhanced reactivity reflects the electron-withdrawing character of the heteroaryl substituent and an intramolecular N–N repulsion.^69,70 Reaction of elcubragistat (5t) with silyl-ether-containing 3,6-di(pyridin-2-yl)-1,2,4,5-tetrazine (6q) furnished an analogue that retained its cyclic amine unit (7t). This product is likely formed through oxidation of a 4,5-dihydropyridazine intermediate (see X → 7b, Figures 2 and 4B).

Notably, protic functional groups within the drugs, such as an alcohol (7m, 7r, 7u), a secondary amine (7q), and a carboxylic acid (7o, 7u, 7v), remained intact. Various heteroarenes were also well tolerated, including pyridazine and 1,2,4-triazole (7n), benzo[d]imidazole (7o), pyrimidin-4(3H)-one and benzo[d]isoxazole (7p), pyrazole (7w), and furan (7y). For modification of trifarotene (5u), where the drug or tagging agent had limited solubility in CH₂Cl₂, we used more polar or protic co-solvents (e.g., 10% DMSO or 50% MeOH in CH₂Cl₂).

Reactions of N-alkyl-substituted piperidine- and pyrrolidine-based drugs (5c, 5k–5q) furnished products 7c–7q in 21–75% yield, with >95% consumption of the amine substrates. Derivatization of morpholine-based drugs (5r, 5s) gave 7r (27% yield) and 7s (21% yield), with recovery of starting materials in 24% and 56%, respectively (see the Supporting Information for details). The material loss likely arises from competing photoactivated, flavin-mediated side reactions of the drug-derived enamine intermediates, processes that may include conversion to radical-cation species^81,82 and reaction with singlet oxygen (¹O₂) generated under aerobic conditions.^72,83 The formation of the amino-alcohol product 7r indicates that, after cycloaddition and N₂ extrusion, the morpholine-derived intermediate (X, Figure 2) can undergo ring-opening through C–O bond rupture as opposed to the C–N bond cleavage that occurs en route to 7s.⁶⁷

Alteration of Y-320 (5w, Figure 4A), which caused the enamine derived from its N-alkyl morpholine (5w → XIII) or N-aryl piperidine (5w → XIV) moieties to react with tetrazine 6u, afforded 7w and 7x in 20% and 33% yield, respectively. The arming of c-Fms-In-13 5y which possesses two N-aryl piperidine moieties (Figure 4B) was found to occur exclusively at the less sterically hindered N-heterocycle to give 7y (23% yield) and 7z (34% yield). The formation of the two products indicates that intermediate XVI generated by [4+2] cycloaddition between 5y-derived enamine and 6q (1y → XV → XVI) can either engage in amine elimination to give the ring-opening product (XVI → 7y), or oxidation to afford the 1,2-diazine (XVI → 7z). Notably, these regioisomers (7w and 7x, or 7y and 7z) could be isolated and purified by flash silica gel chromatography.

These results highlight the unique ability of the C6-based catalyst system to selectively oxidize tertiary amine–based drugs in different solvents, despite containing Lewis acidic, basic, and/or oxidation-sensitive moieties, including those especially prone to oxidation (e.g., the primary alcohol in 5u, the secondary amine in 5q). Furthermore, we find that tetrazines can react exclusively with the amine-derived enamine intermediates of varying sizes and electronic profiles. Such an attribute allows for rapid generation of a considerable range of desirable analogues.

3. BIOLOGICAL ASSAYS

We first set out to assess how arming irinotecan (5c) with tetrazine- or nitroalkene-based tagging agents might impact its anti-cancer bioactivity. We treated the parent compound 5c along with its tetrazine-derived analogue 7d and nitroalkene-derived analogue 9b (Figure 5A), in the human breast cancer cell line SK-BR-3 and the cell viability was measured. In regard to tetrazine derivatives, ketone-tagged 7d (IC₅₀ = 0.84 ± 0.25 μM) exhibited potency identical to that of the lead compound 5c (IC₅₀ = 0.86 ± 0.36 μM). In regard to nitroalkene derivatives, the azide-tagged 9b showed an IC₅₀ of 1.8 ± 0.4 μM. These results demonstrate that both tetrazine- and nitroalkene-based tagging leads to irinotecan analogues with bioactivities that are comparable to that of the original entities. It is not surprising that modifications to [1,4’-bipiperidine]-1’-carboxylate unit minimally impact bioactivity. In the context of irinotecan, the latter fragment releases SN-38, the active metabolite of 5c, through hydrolysis of its carbamate linker, an event catalyzed by human carboxylesterase 2.⁷³

Fig. 5. Biological Assay.

(A) IC₅₀ values for SK-BR-3 cell line with irinotecan 5c and analogues 7d, 9b; (B) IC₅₀ values for cancer cell lines with harrintonine 5m, bemcentinib 5n, gefitinib 5s, and analogues 7m, 7n, 7s. The 7m-treated cells did not reach 0% viability; thus, an IC₅₀ value was estimated from the data; (C) Determination of IC₅₀ values for EGFR kinase with 5s and 7s by the use of ADP-Glo^™ luminescent kinase assay. See the Supporting Information for details on the determination of IC₅₀ values.

We next evaluated the bioactivity of non-prodrug anticancer compounds harringtonine (5m), bemcentinib (5n), and gefitinib (5s), along with their tagged analogues (7m, 7n, 7s), in human breast cancer cell lines MCF-7 or SK-BR-3 cells (Figure 5B). These compounds were selected based on structural analyses and literature docking studies, which allowed us to estimate the likelihood of retaining binding activity after functionalization (See the Supporting Information).^84–87 The mechanism of action of these compounds varies. Harringtonine inhibits protein synthesis by binding to the ribosome⁸³ while bemcentinib inhibits the AXL receptor tyrosine kinase,^84,85 and gefitnib inhibits the epidermal growth factor receptor (EGFR) tyrosine kinase.⁸⁶ The SK-BR-3 and MCF-7 cell lines were chosen since they both express EGFR and AXL. In SK-BR-3 cells, tagging harringtonine (5m, IC₅₀ = 0.36 ± 0.03 μM) produced 7m, which exhibited substantially reduced activity (estimated IC₅₀ ≈ 42 μM). These findings are consistent with the role of harringtonine’s tertiary amine, protonated at physiological pH, in anchoring the drug within the ribosomal A-site cleft through H-bonding and electrostatic interactions.⁸⁴ In MCF-7 cells, the bemcentinib analogue 7n showed an IC₅₀ of 35.5 ± 2.2 μM (vs 14.6 ± 1.2 μM for 5n), indicating decrease of potency upon pyrrolidine modification.^85,86 By contrast, modification of the morpholine ring in gefitinib, which is believed to have primarily a metabolic function (vs directly interacting with the EGFR kinase ATP-binding pocket)⁸⁷ had limited influence on cellular potency (IC₅₀ = 52.5 ± 4.9 μM for 5s vs 64.0 ± 3.3 μM for 7s in MCF-7 cells). To confirm that 7s retains on-target activity, an activity assay for kinase inhibition (Figure 5C) was applied to measure the EGFR kinase activity with compound treatment, both 5s and 7s inhibited EGFR kinase activity in the system and yielded comparable IC₅₀ values (3.4 ± 0.5 nM for 5s and 8.3 ± 2.3 nM for 7s). Collectively, these findings illustrate that our tagging method can preserve lead compound bioactivity when a cyclic amine is inessential to target binding or, in the case of a prodrug, the amine is cleaved prior to target engagement.

4. CONCLUSIONS

We have developed a catalytic strategy for the chemical tagging of bioactive N-alkylamines through chemo- and site-selective transformation of C(sp³)–H bonds. We demonstrate that, through the use of photoactivated flavin analogues, it is possible to effect sequential cleavage of α- and β-amino C–H bonds with assorted bioactive compounds, enabling their conjugation with tetrazine- and nitroalkene-based tagging agents. The catalytic reactions proceed under mild conditions without strong oxidants, acids, or bases; a wide array of functional groups can therefore be tolerated, including Lewis acid- and base-sensitive moieties and heteroarenes. The tagged derivatives serve as versatile intermediates for the rapid synthesis of drug conjugates, as demonstrated by synthesis of two irinotecan-based antibody–drug conjugates; this is a workflow that can be extended to other payloads and adapted to multiple cancer types. Current efforts are directed toward developing an improved tagging strategy that introduces a cleavable linker and tag to potent cytotoxic payloads while minimizing the impact of tagging on their intrinsic activity.

The present study serves as a foundation for the site-selective functionalization of small-molecule amine-containing natural products and drugs, providing a rational framework for transforming complex substrates into α- and β-substituted analogues. Ongoing efforts are focused on expanding the scope of electrophiles that can capture the drug-derived enamines, thereby broadening the chemical space and therapeutic potential accessible through this late-stage functionalization platform.

Supplementary Material

ASSOCIATED CONTENT

Supporting Information The Supporting Information is available free of charge at https://pubs.acs.org/doi/XXXXXXXXXX.

Full experimental procedures including compound characterization data (PDF)