Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Sep 2.
Published in final edited form as: J Am Chem Soc. 2025 Aug 20;147(35):32118–32123. doi: 10.1021/jacs.5c10855

Radical Propagation via σ-Cleavage Mediates Radical SAM Catalyzed Sulfur-for-Oxygen Swapping Reaction during the Biosynthesis of Albomycin δ2

Ziyang Zheng 1, Richiro Ushimaru 2, Conor M Thomas 3, Hung-wen Liu 4
PMCID: PMC12401566  NIHMSID: NIHMS2107348  PMID: 40834352

Abstract

AbmM is a radical S-adenosyl l-methionine (SAM) enzyme that catalyzes a radical initiated sulfur-for-oxygen swapping reaction, transforming the furanose ring of cytidine diphosphate (CDP) to a 4′-hydroxy-4′-thiofuranose product. While the function of AbmM has been demonstrated, the underlying mechanism regarding the formation of the radical intermediates during the reaction pathway remains to be fully established. To gain additional insight into this vital step in the biosynthesis of albomycin δ2, 2′-deoxy-2′-methylidene CDP was synthesized as a mechanistic probe. Upon incubation with AbmM and dithionite, a C1′ radical intermediate is generated from this mechanistic probe in the form of an allylic radical that can be trapped via oxidation to a sulfinate or a sulfenate versus reduction. Moreover, incubation of 2′-deoxy-2′-spirocyclopropryl CDP with AbmM also leads to a C1′ radical intermediate that triggers opening of the cyclopropane ring. In this case, however, the resulting C7′ terminal radical is not directly quenched but instead adds to the C5=C6 double bond of the cytosine base to form a new C7′–C6 bond. Taken together, these studies establish the intermediacy of a C1′ radical species and thus suggest radical propagation from the C4′ radical to the C1′ radical through cleavage of the C1′–O bond prior to the sulfur insertion step during the AbmM-catalyzed reaction.

Graphical Abstract

graphic file with name nihms-2107348-f0001.jpg

INTRODUCTION

Sulfur-containing compounds are widely distributed in nature and many of them have been shown to exhibit potent biological activities.1,2 Moreover, many enzymes have been identified capable of catalyzing the construction of carbon–sulfur bonds in natural products.1,3–5 The majority of these reactions rely on Lewis acid–base chemistry (i.e., paired-electron mechanisms) in which a thiol nucleophile attacks an electrophilic carbon in the substrate to effect C–S bond formation.1,3–5 However, several cases have been reported to involve a radical mechanism wherein a substrate carbon radical is generated prior to coupling with a sulfur donor to generate the C–S bond. Well-known examples of enzymes catalyzing this type of reaction include the nonheme iron enzyme isopenicillin N-synthase,6 the radical S-adenosyl methionine (SAM) enzymes BioB,7,8 LipA,9–11 AlbA,12 RimO,13 MiaB,14,15 MybB,16 and the B12-dependent radical SAM enzyme ThnL.17

The radical SAM enzyme AbmM was recently identified in the biosynthetic pathway of albomycin δ2 and found to catalyze a sulfur-for-oxygen swapping reaction converting the furanose ring of cytidine 5′-diphosphate (CDP, 1) to a sulfur-bridged ribofuranose moiety in 5 (Figure 1).18 AbmM is a twitch radical SAM enzyme having two [Fe4S4] clusters in the active site. The function of the auxiliary cluster remains unclear; however, the radical SAM catalytic cluster (the RS cluster) appears to play a pivotal role in providing the sulfur introduced into the thiofuranose core in addition to its canonical function mediating the reductive cleavage of SAM during the AbmM-catalyzed reaction.18 Although nonradical-mediated C–S bond formation is known in thiosugar biosynthesis,19,20 the discovery of AbmM demonstrated that radical-mediated mechanisms of thiosugar biosynthesis also exist in nature.

Figure 1.

Figure 1.

Open in a new tab

Proposed mechanism for the AbmM-catalyzed reaction. RS cluster: the radical SAM catalytic [Fe4S4] cluster is responsible for not only mediating the reductive cleavage of SAM but also providing the sulfur incorporated into the thiofuranose core.

As a radical SAM enzyme, AbmM catalysis proceeds via formation of a 5′-deoxyadenosyl radical (5′-dA•, 6) equivalent from the reductive cleavage of SAM that abstracts the C4′ hydrogen atom of CDP (1), which has been demonstrated using [4′−2H]-1 previously.18 While direct addition of sulfur at C4′ is possible (see Figure S1), it would lead to exchange of the C4′ hydroxyl group with solvent H2O in the product, which was ruled out by earlier experiments conducted in buffered H218O.18 Therefore, the putative C4′ radical in 2 was proposed to undergo homolytic cleavage of the C1′–O bond to yield a C1′ radical intermediate with concomitant formation of a C4′ keto group (2 → 3, Figure 1). Further experiments revealed that the C1′ radical may add to a sulfur in the radical SAM (RS) cluster of AbmM to generate the C1′–S bond (3 → 4, Figure 1).18 Subsequent intramolecular cyclization of 4 affords the product 5 as a mixture of two C4′ epimers (4 → 5, Figure 1).

Despite indirect evidence for the proposed mechanism of AbmM catalysis in Figure 1, there has been relatively little evidence to corroborate the intermediacy of the putative C1′ radical 3. Therefore, in an effort to investigate the mechanism of AbmM in detail, several substrate analogues of CDP (1) bearing strategically incorporated reporting groups for a C1′ radical (3) were prepared and analyzed as mechanistic probes of AbmM catalysis. Reported herein are the syntheses of C2′-chloro (8), C2′-methylidene (15), and C2′-spirocyclopropryl (28) analogs of CDP, the reactions they undergo in the presence of AbmM, and the mechanistic implications of the incubation outcomes. All observations are consistent with a C1′ radical intermediate in the AbmM catalyzed reaction indicating initial C4′ radical formation is followed by propagation to a C1′ radical prior to sulfur incorporation unlike other radical SAM sulfurtransferases where sulfur insertion typically occurs at the site of initial H atom abstraction.

RESULTS AND DISCUSSION

Rapid homolytic cleavage and elimination of a chlorine radical is known to occur from a C–Cl bond located β to a radical center.21,22 Therefore, the 2′-deoxy-2′-chloro-CDP substrate analog 8 was synthesized as a probe to test for the formation of the putative C1′ radical 3. It was expected that formation of the radical intermediate 10 would trigger elimination of a chlorine radical and introduce a double bond between C1′ and C2′ (8 → 9 → 10 → 11, Figure 2A). Accordingly, 0.4 mM 8 was incubated with 30 μM AbmM, 0.6 mM SAM, 2 mM Na2S, and 2 mM sodium dithionite anaerobically at room temperature overnight followed by workup in the presence of calf intestinal phosphatase (CIP) to dephosphorylate any nucleotides in the reaction mixture in order to facilitate product isolation and characterization by HPLC. Compared to control reactions without sodium dithionite, a new peak eluting at 15.2 min was observed in addition to 5′-dA (7) which appeared at 24.2 min (Figure 2B). This new species exhibited a λmax at 260 nm consistent with retention of the adenine chromophore in the structure. ESI-HRMS and NMR analysis (Figures S2–S6 and Table S1) confirmed its identity as the 5′-deoxyadenosyl sulfinic acid adduct 14 (Figure 2C). This species has been observed in reactions of several radical SAM enzymes such as DesII23 and NosL,24 and is believed to arise from recombination of 5′-dA• (6) and the sulfinate anion radical (13) generated via homolysis of dithionite (Figure 2C).23 This adduct was also detected in reactions without 8 (Figure S7). The fact that no products derived from 8 were observed indicated that 8 is not a substrate for AbmM possibly due to the large size of the 2′-chlorine substituent.

Figure 2.

Figure 2.

Open in a new tab

(A) Proposed reaction outcome for the reaction of AbmM with 8. (B) HPLC analysis of the reaction of AbmM with 8 after CIP treatment. (C) Formation of 14..

Compound 15 was next considered as a substrate analog with a more compact 2′-methylidene substituent in order to facilitate correct binding into the AbmM active site. Should 15 be processed by AbmM, the allylic C1′ radical in 17 would be stabilized by resonance (17 ↔ 18) and thus allow for sulfur incorporation at C6′ to afford 19 (Figure 3A). To test this prediction, 15 was synthesized and incubated with AbmM as described above. After 4 h, the reaction mixture was treated with CIP and analyzed by HPLC. Apart from (5′-dA, 7) and the previously observed 5′-deoxyadenosyl sulfinic acid adduct (14), five new peaks were detected with retention times at 4.0, 4.8, 5.0, 6.4, and 10 min (Figure 3B). According to HRMS analysis, the first three of these products are dephosphorylated isomers of each other, each with a mass increase of 64 Da compared to that of the dephosphorylated substrate (20). Likewise, the product eluting at 6.4 min is an isomer of 20, whereas the compound eluting at 10 min is 48 Da heavier than 20 (Figure S8).

Figure 3.

Figure 3.

Open in a new tab

(A) Proposed reaction outcomes for the reaction of AbmM with 15. (B) HPLC analysis of the reaction of AbmM with 15 after CIP treatment. Peaks at 4.0 and 5.0 min are labeled as x and y, respectively, which correspond to isomers of 24. (C) Products 24, 25, and 26.

Since addition of a sulfinate moiety would result in a 64 Da increase in mass, the three isomeric products eluting between 4.0 and 5.0 min are likely sulfinate adducts of radical intermediates such as 16, 17 and 18. Indeed, loss of a 64 Da neutral fragment consistent with SO2 was observed in the MS spectra of these products (Figure S9).25 The incorporation of two 18O atoms into these products when the reaction was run in 46% H218O provided further evidence supporting their assignments as sulfinate adducts (Figure S10), because the two oxygens of dithionite that generates sulfinate anion radicals readily exchange with water.26

The reaction with 15 was subsequently scaled up to isolate the products for structural characterization. In the 1H NMR spectrum of the product eluting at 4.8 min, H3′ is observed as a singlet (δ = 5.10 ppm), and the two proton signals associated with C5′ appear as two doublets (δ = 4.67 and 4.61 ppm, 2JHH = 19 Hz) (Figure S11). Furthermore, both the H3′ and H5′ signals are shifted downfield compared to 20 (Figure S11). These observations are consistent with a structure having a C4′ keto group. Indeed, this product along with its two isomers at 4.0 and 5.0 min could be reduced by NaBH4 with a mass increase of 2 Da (or 3 Da with NaBD4) (Figure S15). It was also noted that the two methylidene protons at C6′ in 20 are shifted upfield in this product as two methylene doublets (δ = 3.27 and 3.09 ppm; 2JHH = 13 Hz). Moreover, the chemical shifts of the H6′ resonances are close to those of H5′ in 14 (δ = 3.08 and 2.71 ppm) and thus suggest a sulfinate group at C6′. Migration of the C2′=C6′ double bond to C1′=C2′ is supported by the downfield shift of H1′ and the C1′ resonance at 128 ppm (Figures S11 and S13). These data along with additional 2D NMR spectra (Figures S11–S13, Table S2) allowed the structure of the product eluting at 4.8 min to be assigned as 24 with a Z-geometry of the C1′=C2′ double bond established via NOE correlation between H1′ and H3′ (Figures 3C and S14). The two isomeric products eluting at 4.0 and 5.0 min could not be firmly characterized due to poor recovery; however, they may correspond to isomers of 24 with the sulfinate at C1′ as both showed an identical UV absorption at 301 nm which is distinct from the 282 nm absorption of 24 (Figure S16).

The 1H NMR spectrum of the product eluting at 6.4 min, which is an isomer of 20, exhibited similar chemical shifts to those of 24; however, both H3′ and H5′ appear as singlets (Figure S17). Another singlet (δ = 1.56 ppm) corresponding to a vinylic methyl group was also detected (Figure S17). Since cross peaks were observed between the latter signal and H1′ in the COSY spectrum (Figure S18), this proton singlet can be ascribed to a C6′ methyl group. Hence, this product was assigned as 25 based on these and additional NMR results (Figures 3C and S17–S20, Table S3). The observation of products 24 and 25 is consistent with radical mediated ring opening of 15 to generate the 17 ↔ 18 resonance pair. This intermediate can then undergo oxidation via radical addition of the sulfinate anion radical (13) to yield 21 and thus 24 following dephosphorylation with CIP (Figure 4).25,27 However, 17/18 is apparently also susceptible to reduction leading to the formation of 22, which upon dephosphorylation produces 25 (Figure 4).

Figure 4.

Figure 4.

Open in a new tab

Proposed mechanism for the formation of 24, 25, and 26.

The 1H NMR spectrum of the remaining product eluting at 10.0 min in Figure 3B showed a close resemblance to that of 24; however, the C5′ proton resonances were missing presumably due to being obscured by the water peak (Figure S21). The chemical shifts of the H6′ signals (δ = 3.20 and 3.00 ppm; 2JHH = 13 Hz) were found slightly upfield of those in 24; but their splitting patterns were essentially identical. Considering the mass of this product is 16 Da less than that of 24, it is likely the sulfenic acid adduct 26 (see Figure 3C). Indeed, derivatization with phenyl vinyl sulfone yielded the corresponding sulfoxide (Figure S25), consistent with the presence of a sulfenic acid moiety.28 Additional NMR spectra (see Figures S21–S24, Table S4) further supported assignment as 26. Since 26 was not formed when 24 was treated with dithionite anaerobically (Figure S26), 26 is unlikely to be a reduction product of 24. Instead, formation of 26 could result from the coupling of 18 with a sulfenate anion radical (27) (Figure 4). A plausible mechanism for the generation of sulfenate anion radical from dithionite is proposed in Figure S27A. Alternatively, thiols are known to react with hydrogen peroxide (H2O2) to produce sulfenic acids.29 Therefore, formation of 26 could also be the result of S-hydroxylation of the proposed thiolated product (19) by H2O2 generated upon reduction of O2 by excess dithionite during workup of the reaction under aerobic conditions (Figures 4 and S27B). Taken together, the product profile from the reaction of AbmM with the 2′-methylidine analog 15 is consistent with the formation of a C1′ radical (17/18) even though the proposed thiolated product 19 was not observed.

The substrate analog 2′-deoxy-2′-spirocyclopropryl CDP (28) was also synthesized to further test for the intermediacy of a C1′ radical. Based on the reaction of AbmM with 15, it was anticipated that formation of the putative intermediate 30 with a C1′ radical β to the 2′-spirocyclopropryl group would induce opening of the cyclopropane ring in 28 to yield 31 with a C7′ radical (Figure 5A).30–32 The terminal radical at C7′ could then be quenched by sulfinate anion radical (13 + 31 → 32) as in the case of 15 (Figure 5A). During incubation of 28 with AbmM under the aforementioned conditions, the formation of both 5′-deoxyadenosine (7) and 5′-deoxyadenosyl sulfinic acid adduct (14) was observed as well as two new HPLC peaks eluting at 3.9 and 18.8 min following workup with CIP (Figure 5B). HRMS analysis showed that the former product had the same exact mass (calcd for C11H14N3O4− [M–H]− 252.0990, found: 252.0992) as that of the dephosphorylated substrate (33), while the exact mass of the latter product (calcd for C11H13N2O5− [M–H]− 253.0830, found: 253.0836) was 1 Da greater than that of 33. The odd change in mass suggests an odd change in the number of nitrogens in the product.

Figure 5.

Figure 5.

Open in a new tab

(A) Proposed reaction outcomes for the reaction of AbmM with 28. (B) HPLC analysis of the reaction of AbmM with 28 after CIP treatment.

The reaction was then scaled up to isolate the two new products from the reaction with 28 for structure characterization; however, the product eluting at 3.9 min was found to be unstable and converted to the second product upon lyophilization (Figure S28). This suggested that only the product eluting at 3.9 min is generated enzymatically and subsequently undergoes nonenzymatic decomposition to the second product. The decomposition product, however, was sufficiently stable to be isolated and characterized by NMR. As shown in the corresponding 1H NMR spectrum (Figure S29), H3′ and H5′ produce singlet (δ = 4.96 ppm) and two doublet (δ = 4.54 and 4.49 ppm, 2JHH = 19 Hz) resonances, respectively. This suggests the absence of a hydrogen at C4′. Furthermore, a signal with a chemical shift of 210.56 ppm was detected by 13C NMR and displayed 1H–13C HMBC correlations respectively with H5′ and H3′ indicative of a keto group at C4′ (Figure S33). Moreover, there exists a double bond between C1′ and C2′ as evidenced by the 13C chemical shifts of C1′ and C2′ at 123.17 and 118.19 ppm, respectively (Figure S30). These features are consistent with the formation of an intermediary C1′ radical that leads to opening of the cyclopropane ring to generate the C1′=C2′ double bond and a C7′ radical (31).

However, in contrast to the products of the reaction between AbmM and 15, the two characteristic signals for H5 and H6 of the cytosine base (e.g., δ = 6.25 and 8.04 ppm, 3JHH = 7.8 Hz in 24) were absent in the 1H NMR spectrum (Figure S29). Instead, H5 (δ = 2.73 and 2.82 ppm) and H6 (δ = 3.87 ppm) in the new product had apparently shifted upfield (Figure S29). Moreover, H5 appears as two doublets of doublets (δ = 2.73 ppm, 2JHH = 17.1 Hz, 3JHH = 13.1 Hz and 2.82 ppm, 2JHH = 17.0 Hz, 3JHH = 4.4 Hz), while H6 is split by 4 nuclei (δ = 3.87 ppm, 3JHH = 13.1, 12.2, 4.4, 3.1 Hz) and demonstrates a correlation with H7′ as shown in the 1H–1H COSY spectrum (Figures S29 and S31). These features indicated that a C–C bond is formed between C6 and C7′. HRMS, 1D and 2D NMR analysis (Figures S29–S38, and Tables S5 and S6) collectively allowed assignment of the decomposition product as 38, which features a C–C linkage between C6 and C7′ along with an iminol at C4 (see Figure 6). The latter was confirmed by 1H–13C HMBC correlations between the iminol hydrogen and C4 as well as C5, respectively when DMSO-d6 was used as the NMR solvent (Figure S38).

Figure 6.

Figure 6.

Open in a new tab

Proposed mechanism for the formation of 36 and 38. HMBC correlations in 38 were recorded in solvent DMSO-d6.

Based on these results, it is proposed that the terminal C7′ radical (31) generated upon opening of the cyclopropane ring in 30 attacks the C5=C6 double bond in the cytosine base to produce a C7′–C6 bond and a C5 radical (34), which is reductively quenched to yield 35 (Figure 6). Dephosphorylation of 35 by CIP would result in 36, which has the same molecular weight as the dephosphorylated substrate 33 and thus the unstable product peak at 3.9 min (Figure 5B). Compound 36 (C11H15N3O4) once formed could then undergo hydration followed by elimination of ammonia to yield the stable decomposition product 38 (C11H14N2O5). The C4 amino group of the cytosine ring seems most likely to be expelled compared to the other two nitrogen atoms in the cytosine base. To ascertain whether the C4 nitrogen is indeed lost from 28 in the reaction with AbmM, [4-15N]-28 was synthesized and assayed with AbmM. Consequently, mass analysis showed that 15N at C4 was retained in the enzymatic product 36 and lost upon decomposition (Figure S39) consistent with the proposed model.

To probe whether the radical quenching at C5 is stereoselective, the reaction was run in buffered D2O tointerrogate the stereochemical course of deuterium incorporation from solvent. Control experiments showed that the C5 hydrogens of 38 do not exchange with solvent (Figure S40). In contrast, the 1H NMR spectrum of 38 isolated from the reaction in buffered D2O reaction showed that the two hydrogens at C5 were 48 and 65% deuterated (Figure S41). This suggests that the protonation at C5 is not strictly stereoselective.

CONCLUSIONS

In summary, existence of a C1′ radical intermediate during the AbmM catalyzed reaction has been probed using a series of substrate analogs (i.e., 8, 15 and 28) that are sensitive to formation of such an intermediate. While the reaction with the 2′-chloro analog 8 did not yield any new products beyond the 5′-sufinate derivative of 5′-deoxyadenosine (14), reaction of the 2′-methylidene analog 15 with AbmM indeed led to a ring-opened C1′ radical (17/18). This intermediary species generated from 15 can undergo a number of different transformations including reduction or oxidation via the addition of either a sulfinate or sulfenate anion radical in the presence of dithionite (Figure 4). The AbmM reaction with the 2′-spirocyclopropryl analog 28 also generated a C1′ radical (30). In contrast to the scenario with 15, however, the cyclopropane ring in 30 opens to yield a C7′ radical that adds to the C5=C6 double bond of the cytosine ring to form a new C7′–C6 bond. The resulting C5 radical is subsequently quenched by one-electron reduction and protonation to form 35 (Figure 6). The resulting product is unstable in aqueous solution such that the dephosphorylated adduct of 35 (i.e., 36) undergoes nonenzymatic hydrolysis of the C4 amidine thereby yielding the C4 iminol compound 38 (Figure 6).

Taken together, the results herein reveal that the AbmM catalyzed sulfur-for-oxygen swapping into CDP follows a mechanism involving radical initiation at C4′ followed by ring-opening to generate the C1′ radical where addition of sulfur takes place (Figure 1). Cyclization of the thiolated product then yields the thiofuranose ring. Thus, the sites of H atom abstraction (C4′) and sulfur addition (C1′) are neither the same nor linked via π-conjugation. Instead, propagation of the radical from C4′ to C1′ requires cleavage of a σ-bond. This feature of AbmM catalysis distinguishes it from other radical SAM sulfurtransferases, where thiolation generally proceeds in a rebound-like manner with sulfur or a thiol adding to the carbon radical generated immediately upon H atom abstraction. Moreover, the thiofuranose product of AbmM catalysis is hydroxylated at C4′ indicative of the oxidative nature of the transformation. This also places the product in the correct oxidation state for subsequent coupling with l-theronine catalyzed by AbmH;33 however, dephosphorylation and rearrangement to the C5′ aldehyde is required prior to this coupling and a current subject of investigation. Overall, the biosynthesis of albomycin has proven to be rich in interesting and unique chemistry with several questions still open to be fully explained including the precise origin of the sulfur introduced during the AbmM catalyzed reaction.

Supplementary Material

Supporting information

Supporting Information

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacs.5c10855.

Additional experimental details, materials, and methods, including chemical synthesis, in vitro enzymatic assays, supplementary tables and figures, and NMR spectra of synthetic compounds (PDF)

ACKNOWLEDGMENTS

We are grateful to an anonymous reviewer for suggesting an alternative mechanism regarding the formation of 26. (18 → 19 → 23 → 26, Figure 4). We thank Dr. Mark Ruszczycky for his invaluable comments on this manuscript. We also thank Dr. Yu-Hsuan Lee for her assistance in the NMR analysis of compound 38. This work was supported by the National Institutes of Health (GM035906 to H.-w.L.).

Footnotes

Complete contact information is available at: https://pubs.acs.org/10.1021/jacs.5c10855

The authors declare no competing financial interest.

Contributor Information

Ziyang Zheng, Department of Chemistry, University of Texas at Austin, Austin, Texas 78712, United States.

Richiro Ushimaru, Department of Chemistry, University of Texas at Austin, Austin, Texas 78712, United States; Institute for Advanced Study and Department of Chemistry, Graduate School of Science, Kyushu University, Fukuoka 819-0395, Japan.

Conor M. Thomas, Department of Chemistry, University of Texas at Austin, Austin, Texas 78712, United States

Hung-wen Liu, Department of Chemistry and Division of Chemical Biology and Medicinal Chemistry, College of Pharmacy, University of Texas at Austin, Austin, Texas 78712, United States.

REFERENCES

  • (1).Dunbar KL; Scharf DH; Litomska A; Hertweck C Enzymatic carbon-sulfur bond formation in natural product biosynthesis. Chem. Rev 2017, 117 (8), 5521–5577. [DOI] [PubMed] [Google Scholar]
  • (2).Hai Y; Wei M-Y; Wang C-Y; Gu Y-C; Shao C-L The intriguing chemistry and biology of sulfur-containing natural products from marine microorganisms (1987–2020). Mar. Life Sci. Technol 2021, 3 (4), 488–518. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (3).Naowarojna N; Cheng R; Chen L; Quill M; Xu M; Zhao C; Liu P Mini-review: ergothioneine and ovothiol biosyntheses, an unprecedented trans-sulfur strategy in natural product biosynthesis. Biochemistry 2018, 57 (24), 3309–3325. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (4).Steele AD; Kiefer AF; Shen B The many facets of sulfur incorporation in natural product biosynthesis. Curr. Opin. Chem. Biol 2023, 76, No. 102366. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (5).Chen X; Li B How nature incorporates sulfur and selenium into bioactive natural products. Curr. Opin. Chem. Biol 2023, 76, No. 102377. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (6).Baldwin JE; Bradley M Isopenicillin N synthase: mechanistic studies. Chem. Rev 1990, 90 (7), 1079–1088. [Google Scholar]
  • (7).Guianvarc’h D; Florentin D; Bui BTS; Nunzi F; Marquet A Biotin synthase, a new member of the family of enzymes which uses S-adenosylmethionine as a source of deoxyadenosyl radical. Biochem. Biophys. Res. Commun 1997, 236 (2), 402–406. [DOI] [PubMed] [Google Scholar]
  • (8).Berkovitch F; Nicolet Y; Wan JT; Jarrett JT; Drennan CL Crystal structure of biotin synthase, an S-adenosylmethionine-dependent radical enzyme. Science 2004, 303 (5654), 76–79. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (9).Ollagnier-de Choudens S; Fontecave M The lipoate synthase from Escherichia coli is an iron-sulfur protein. FEBS Lett. 1999, 453 (1–2), 25–28. [DOI] [PubMed] [Google Scholar]
  • (10).Busby RW; Schelvis JPM; Yu DS; Babcock GT; Marletta MA Lipoic acid biosynthesis: LipA is an iron-sulfur protein. J. Am. Chem. Soc 1999, 121 (19), 4706–4707. [Google Scholar]
  • (11).McLaughlin MI; Lanz ND; Goldman PJ; Lee K-H; Booker SJ; Drennan CL Crystallographic snapshots of sulfur insertion by lipoyl synthase. Proc. Natl. Acad. Sci. U. S. A 2016, 113 (34), 9446–9450. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (12).Flühe E; Knappe TA; Gattner MJ; Schäfer A; Burghaus O; Linne U; Marahiel MA The radical SAM enzyme AlbA catalyzes thioether bond formation in subtilosin A. Nat. Chem. Biol 2012, 8 (4), 350–357. [DOI] [PubMed] [Google Scholar]
  • (13).Forouhar F; Arragain S; Atta M; Gambarelli S; Mouesca JM; Hussain M; Xiao R; Kieffer-Jaquinod S; Seetharaman J; Acton TB; Montelione GT; Mulliez E; Hunt JF; Fontecave M Two Fe-S clusters catalyze sulfur insertion by radical-SAM methylthiotransferases. Nat. Chem. Biol 2013, 9 (5), 333–338. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (14).Zhang B; Arcinas AJ; Radle MI; Silakov A; Booker SJ; Krebs C First step in catalysis of the radical S-adenosylmethionine methylthiotransferase MiaB yields an intermediate with a [3Fe-4S]0-like auxiliary cluster. J. Am. Chem. Soc 2020, 142 (4), 1911–1924. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (15).Esakova OA; Grove TL; Yennawar NH; Arcinas AJ; Wang B; Krebs C; Almo SC; Booker SJ Structural basis for tRNA methylthiolation by the radical SAM enzyme MiaB. Nature 2021, 597 (7877), 566–570. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (16).Zhao H; Bu J; Liu H.-w. Radical S-adenosylmethionine sulfurtransferase MybB catalyzed formation of the 4-thiazolidinone core in mycobacidin represents an intersection between primary and secondary metabolism. J. Am. Chem. Soc 2025, 147 (5), 4180–4187. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (17).Sinner EK; Li RF; Marous DR; Townsend CA ThnL, a B12-dependent radical S-adenosylmethionine enzyme, catalyzes thioether bond formation in carbapenem biosynthesis. Proc. Natl. Acad. Sci. U. S. A 2022, 119 (34), No. e2206494119. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (18).Ushimaru R; Zheng Z; Xiong J; Mori T; Abe I; Guo Y; Liu H.-w. Radical S-adenosyl-l-methionine FeS-cluster implicated as the sulfur donor during albomycin biosynthesis. Nat. Catal 2025, 8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (19).Sasaki E; Zhang X; Sun HG; Lu M-YJ; Liu T.-l.; Ou A; Li J.-y.; Chen Y.-h.; Ealick SE; Liu H.-w. Co-opting sulphur-carrier proteins from primary metabolic pathways for 2-thiosugar biosynthesis. Nature 2014, 510 (7505), 427–431. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (20).Zhao Q; Wang M; Xu D; Zhang Q; Liu W Metabolic coupling of two small-molecule thiols programs the biosynthesis of lincomycin A. Nature 2015, 518 (7537), 115–119. [DOI] [PubMed] [Google Scholar]
  • (21).Wagner PJ; Lindstrom MJ; Sedon JH; Ward DR Photochemistry of δ-halo ketones: anchimeric assistance in triplet-state γ-hydrogen abstraction and β-elimination of halogen atoms from the resulting diradicals. J. Am. Chem. Soc 1981, 103 (13), 3842–3849. [Google Scholar]
  • (22).Joshi S; Mahanta N; Fedoseyenko D; Williams H; Begley TP Aminofutalosine synthase: evidence for captodative and aryl radical intermediates using β-scission and SRN1 trapping reactions. J. Am. Chem. Soc 2017, 139 (32), 10952–10955. [DOI] [PubMed] [Google Scholar]
  • (23).Ko Y; Ruszczycky MW; Choi SH; Liu H.-w. Mechanistic studies of the radical S-adenosylmethionine enzyme DesII with TDP-d-fucose. Angew. Chem., Int. Ed 2015, 54 (3), 860–863. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (24).Bhandari DM; Fedoseyenko D; Begley TP Tryptophan lyase (NosL): a cornucopia of 5′-deoxyadenosyl radical mediated transformations. J. Am. Chem. Soc 2016, 138 (50), 16184–16187. [DOI] [PubMed] [Google Scholar]
  • (25).Joshi S; Fedoseyenko D; Sharma V; Nesbit MA; Britt RD; Begley TP Menaquinone biosynthesis: new strategies to trap radical intermediates in the MqnE-catalyzed reaction. Biochemistry 2021, 60 (21), 1642–1646. [DOI] [PubMed] [Google Scholar]
  • (26).Stolze K; Duling DR; Mason RP Spin adducts formed from nitroso spin traps and dithionite. J. Chem. Soc., Chem. Comm 1988, No. 4, 268–270. [Google Scholar]
  • (27).Lee Y-H; Ren D; Chen R; Liu X; Engstrom A; Liu H.-w. Blasticidin S biosynthesis involves a unique interplay between a radical SAM dehydratase, a transaminase and an ATP-grasp ligase in order to avoid a futile cycle. ACS Catal. 2025, 15 (4), 2742–2749. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (28).Hazra S; Begley TP Alkylcysteine sulfoxide C-S monooxygenase uses a flavin-dependent Pummerer rearrangement. J. Am. Chem. Soc 2023, 145 (22), 11933–11938. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (29).Roscoe JM; Sevier CS Pathways for sensing and responding to hydrogen peroxide at the endoplasmic reticulum. Cells 2020, 9 (10), 2314. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (30).Beckwith ALJ; Bowry VW Kinetics of reactions of cyclopropylcarbinyl radicals and alkoxycarbonyl radicals containing stabilizing substituents: implications for their use as radical clocks. J. Am. Chem. Soc 1994, 116 (7), 2710–2716. [Google Scholar]
  • (31).Huang H; Chang W.-c.; Pai P-J; Romo A; Mansoorabadi SO; Russell DH; Liu H.-w. Evidence for radical-mediated catalysis by HppE: a study using cyclopropyl and methylenecyclopropyl substrate analogues. J. Am. Chem. Soc 2012, 134 (39), 16171–16174. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • (32).Wang Y; Begley TP Mechanistic studies on CysS – a vitamin B12-dependent radical SAM methyltransferase involved in the biosynthesis of the tert-butyl group of cystobactamid. J. Am. Chem. Soc 2020, 142 (22), 9944–9954. [DOI] [PubMed] [Google Scholar]
  • (33).Ushimaru R; Liu H.-w. Biosynthetic origin of the atypical stereochemistry in the thioheptose core of albomycin nucleoside antibiotics. J. Am. Chem. Soc 2019, 141 (6), 2211–2214. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting information
NIHMS2107348-supplement-Supporting_information.pdf (3.8MB, pdf)

RESOURCES