Electro-induced C-H/S-H cross-coupling for the functionalization/macrocyclization of cysteine-containing peptides

Abstract

Herein, an electro-induced umpolung approach that enables the efficient functionalization/macrocyclization of cysteine-containing peptides is reported. Notably, this method utilizes simple halogen source and takes metal-mediated halogen atom transfer as the main pathway to enable the in-situ polarity reversal, highlighting the unique possibilities associated with electrochemical activation methods. Under simple and mild conditions, cysteine residue can be well-labelled with high chemo-selectivity and excellent conversion. This transformation can tolerate a wide range of valuable enamines, azoles, and peptides partners, and can also be utilized as a macrocyclization tactic for cyclic peptide synthesis and other areas.

In comparison to the well-developed electrochemical transformations of thiophenols, the electrochemical modification of peptides and proteins through cysteine remains rather underdeveloped. Here the authors report an electro-induced umpolung approach that enables the efficient functionalization/macrocyclization of cysteine-containing peptides.

Introduction

Chemical modification of peptides and proteins has emerged as an efficient strategy to gain insights into the structure-function relationship of peptides and proteins and is represented as an important tool for probing natural systems, creating therapeutic conjugates, and generating protein constructs^1–5. The modification of peptides and proteins must be carefully designed to avoid interference with protein function. Therefore, there is a demand for site-selective methods that produce a uniform product, as they maximize the likelihood of success in the intended application. The bioorthogonal reactivities of the side chain functional groups of specific amino acid residues serve as the basis for most protein chemical modifications^6–12. Among the 20 proteinogenic amino acids, cysteine is one of the most frequently targeted residues for chemical modification of peptides and proteins due to its unique thiol group and its relatively low natural abundance^13–17. Unsurprisingly, the majority of contemporary bioconjugation strategies utilizing Cys predominantly leverage the conventional nucleophilic reactivity of thiol groups (Fig. 1A). Many such transformations, including nucleophilic substitutions^18–27, Michael additions^28–32, metal-mediated reactions^33–36, and visible-light mediated reactions^37–39 have demonstrated applications in chemical biology, medicinal chemistry and organic synthesis. Over the past years, significant advancements have been attained in leveraging the nucleophilicity of cysteine to facilitate the reaction between the cysteine side chain and electrophiles. Over time, the associated methodologies have gradually matured. Nevertheless, there is a notable scarcity of reports regarding the modification of cysteine via the polarity inversion strategy to accomplish the coupling reaction between the cysteine side chain and nucleophiles⁴⁰. Despite this, due to the limited availability of naturally occurring protein side chains with electrophilic properties, using electrophilic equivalents of cysteine offers clear advantages for developing new chemo- and regioselective transformations. Undoubtedly, umpolung-based side chain modification strategies hold significant potential for development; however, the field is still very much in its infancy.

Fig. 1. Approaches for the functionalization of cysteine-containing peptides.

A Summary of the traditional methods for cysteine modification. B Electrochemical strategy for cysteine modification. C This work, an electro-induced umpolung approach for cysteine modification.

Currently, reports on the successful modification of cysteine via the Cys umpolung strategy predominantly entail the pre-activation of thiol groups. This is accomplished by forming disulfide bonds^41–46 or sulfur-halogen bonds^47,48, followed by reacting with diverse nucleophiles through reactions like electrophilic aromatic substitution to achieve the modification of cysteine. Recently, Otaka and co-workers also presented a successful instance of the polarity reversal of cysteine⁴⁹. They synthesized an S-p-methoxybenzyl Cys sulfoxide intermediate, and then generated a Cys-derived sulfenyl chloride in the presence of Guanidine HCl under acidic reaction conditions to achieve the polarity reversal of cysteine. However, these umpolung strategies often require an excess of additives such as toxic activation reagents and oxidants, multiple reaction steps, and a limited scope of substrates. Therefore, the advancement of direct, convenient, and varied bonding patterns facilitated by the umpolung strategy for chemical modification of peptides and proteins is highly appealing.

In recent years, there has been rapid advancement in the strategy of selectively modifying sensitive amino acids and peptides using electrochemical techniques, resulting in the discovery of unique research findings^50–62. The synergy of sustainability, biocompatibility, and precise kinetic control underscores the growing popularity of electrochemically-driven transformations in the modification of biomolecules. As early as the 1980s, chemists had already begun to study the electroredox properties of cysteine^63,64. However, it has been primarily employed in an analytical capacity. In comparison to the well-developed electrochemical transformation of thiophenols, the electrochemical modification of peptides and proteins through cysteine remains rather underdeveloped. The principal challenge associated with the electrochemical modification of cysteine resides in the fact that cysteine has a propensity to form disulfide bonds or undergo desulfurization under electrochemical circumstances (Fig. 1B)^65,66. Recent reports have also emerged regarding traditional cross-coupling assisted by electrochemistry involving cysteine, which provides more possibilities for exploring electrochemically-driven modifications of peptides⁶⁷. Based on the above-mentioned electrophilic sulfenyl chloride obtained by the reaction of sulfuryl chloride with Cys, we hypothesized that a similar sulfenyl halide intermediate could be obtained through the reaction of Cys with a simple halogen source via electrochemical oxidation conditions, thus achieving in situ polarity reversal. Subsequently, the umpolung Cys could undergo electrophilic substitution with the nucleophilic Sp² C-H bond to achieve the corresponding S − H/C − H cross dehydrogenative coupling for Cys selective modification. To the best of our knowledge, no procedure reporting the bioconjugation of cysteine-containing peptides via S − H/C − H cross-dehydrogenative coupling has been put forward. This approach is expected to be more direct than traditional methods and can overcome the previous limitations. Herein, we demonstrate that a simple electrochemical-enabled in-situ umpolung approach can be utilized for both inter- and intramolecular reactions to produce Cys-containing peptides as well as cyclic peptides (Fig. 1C). The reaction demonstrates a wide range of valuable substrates, including various enamines, azoles, and peptide partners. It employs simple and mild conditions to effectively furnish highly S-functionalized peptides.

Results and Discussions

Investigation of reaction conditions and substrate scope

To explore this electrochemical umpolung concept, we selected the coupling of Boc-DL-Cys-OMe 1a with 3-aminocrotononitrile 2a as a model study under various reaction conditions (Fig. 2). The reaction was initially investigated under the conditions of constant-current electrolysis at 12 mA in the presence of ⁿBu₄NI, ⁿBu₄NBr or ⁿBu₄NCl. We are contemplating whether the electrophilic sulfenyl halide species can be directly obtained under anodic oxidation conditions; however, only trace products can be obtained. To our delight, when various chlorine sources are added, the desired products can be obtained in low yield, among which chloroform has the relatively most favorable effect. This encouraging result impelled us to re-optimize the reaction conditions. Based on the previous reports that Cl-bound transition metal species could serve as atom-transfer reagents, we next explored the influence of metal catalysts that are frequently used in atom-transfer radical reactions. Interestingly, the yield of 3a could be increased up to good yield when different metal catalysts were used in the reaction, such as Mn(OTf)₂, Ferrocene, Ni(acac)₂ and Ni(cod)₂. The introduction of NiBr₂•diglyme created a highly reactive system in which 3a was isolated in 80% yield. Replacing ⁿBu₄NBr with other electrolytes, such as ⁿBu₄NBF_4, ⁿBu₄NCl_, and ⁿBu₄NPF₆ somewhat decreased the product yield. The effect of electrode materials was also explored. Using a carbon rod as the anode or Ni as the cathode decreased product yield, to some extent. Control experiments confirmed that the optimal current was 12 mA and no product was produced in the absence of electric current. We also examined the reaction under various other conditions, including different current, solvent and reaction time. More detailed information can be found in the supporting information about the investigation of the reaction conditions. To further illustrate the impact of the newly designed electrochemical umpolung modification strategy, we next conducted a horizontal comparison with other reported conditions, which designed for thiophenol functionalization^68–71. It is known from previous reports that hypervalent iodine reagents, peroxides, and iodine-mediated electrolysis have all been verified to be effective approaches for achieving the thiophenols S − H/C − H cross dehydrogenative coupling. A horizontal comparison of these reaction conditions was conducted with ours. The result showed that only our strategy could obtain the desired S-functionalized peptides under the optimized reaction conditions.

Fig. 2. Investigation of the reaction conditions^a.

a Standard conditions: platinum plate anode, platinum plate cathode (15 × 15 x 0.3 mm), constant current = 12 mA, 1a (0.2 mmol), 2a (2.0 equiv.), ⁿBu₄NBr (1.0 equiv.), NiBr₂•diglyme (20 mol%), 6.0 mL MeCN, 1.0 mL CHCl_3, undivided cell, 2.5 h. Isolated yields were shown. Condition (1): C (Cloth) | Fe (-), 3 mA, ⁿBu₄NBF₄, CH₃CN, rt, 5 h. Condition (2): Pt | Pt, 5 mA, NH₄I (20 mol%), NH₄BF₄, rt, 6 h, CH₃CN.

With these highly encouraging outcomes at hand, the scope of the reaction was next explored with various cysteine-containing dipeptides and polypeptides. As shown in Fig. 3. The reaction exhibited excellent compatibility with a variety of natural aliphatic amino acids such as leucine (Leu), alanine (Ala), valine (Val), glycine (Gly), and glutamic acid (Glu) as well as unnatural medicinal-useful bulky amino acids, affording the dipeptide cross-coupling products in moderate to good yields (3–12). In a gram-scale reaction using the amino group, unprotected cysteine as the substrate, a moderate yield of the corresponding product 4 was obtained, which demonstrates good chemo-selectivity. Meanwhile, when employing cysteine bearing an unprotected carboxylic acid group, we were also able to obtain the target product 5 smoothly in a moderate yield. It should be noted that we investigated some sensitive amino acids such as methionine, tryptophan and phenylalanine, which were highly challenging substrates in electrochemical transformations. Pleasingly, the mercaptomethyl, indolyl, and benzyl groups of the corresponding amino acid residues were not altered during this electrocoupling. These substrates were all compatible under standard conditions, generating corresponding products in medium to good yields (13–15). Meanwhile, the OH-protected amino acids, such as tyrosine and serine are also compatible to produce the desired peptides (16–18). However, in the case of amino acids bearing unprotected hydroxyl groups, arginine and histidine are not compatible with our reaction system. Furthermore, nonproteinogenic amino acids such as propargylglycine, allylglycine and 4-Bromophenylalanine could also be readily converted into the corresponding products in moderate yields (19–21). Interestingly, this reaction can also be used for the synthesis of saccharide derivatives, which highlights the potential of this method for bioconjugation reactions (22). With the aim of further demonstrating the utility of our protocol, we concentrated our attention on conducting this reaction on more complex peptides. The modification of a variety of polypeptides from 6-mers to 8-mers has also been effectively achieved with good to moderate yields (23–25).

Fig. 3. Scope of cysteine-containing peptides and enamines and azoles^a.

^a Standard conditions as shown in Fig. 1. ^b Reaction conditions: platinum plate anode, platinum plate cathode (15 × 15 x 0.3 mm), constant current = 4 mA, polypeptides (0.01 mmol), 2a (2.0 equiv.), ⁿBu₄NBr (2.0 equiv.), NiBr₂•diglyme (20 mol%), MeCN (6.0 mL), CHCl₃ (1.0 mL)_, undivided cell, room temperature, 15 min. Isolated yields (purified by preparative HPLC).

Subsequently, we investigated the scope of the e-coupling by exploring various enamines and azoles (Fig. 3). A range of representative enamines, featuring differently substituted aromatic rings (-Me, -OMe, -CN, -Br, -Cl) at meta or para positions, successfully yielded the desired products (26–31). Then, we turned to examine the influence of the N-substituents on the 3-phenylacrylates in this transformation. As we expected, substrates (32–36) bearing methyl, isopropyl, cyclopropyl, and benzyl groups showed good reactivity under the e-coupling protocol. To further illustrate the utility of our protocol, we carried out additional research on the substrate compatibility of enamines bearing bioorthogonal azide groups. Gratifyingly, the corresponding product 37 was obtained in 40% yield. Subsequently, the scope of pyrazol-5-amines was investigated. The substituents on the N1 atom, including -C₆H₄, p-Me-C₆H₄, p-Cl-C₆H₄, -Me, as well as the substituents on the C3 position, such as tert-butyl, phenyl, and methyl, were all found to be suitable for this electrocatalytic C − H/S − H coupling reaction. The corresponding products 38–43 were obtained with yields ranging from 39 to 72%. Notably, the aminopyrazole bearing an alkynyl group can also exhibit excellent compatibility with this reaction, giving the corresponding product 44 in 56% yield. This compatibility further offers opportunities for subsequent bioorthogonal reactions. Meanwhile, the substrate with substituent on the 5-amines, such as N, N-dimethyl-N’-(3-methyl-1-phenyl-1H-pyrazol-5-yl)formimidamide, N’−3-(tert-butyl)−1-phenyl-1H-pyrazol-5-yl-N, N-dimethylformimidamide, N, N-dimethyl-N’-(3-methyl-1-(p-tolyl)−1H-pyrazol-5-yl)formimidamide and N, N-dimethyl-N’-(1-methyl-3-phenyl-1H-pyrazol-5-yl)formimidamide proceeded the reaction smoothly to deliver the product with moderate yields (45–48). It should be noted that other azoles derivatives, such as 3-phenylisoxazol-5-amine and nonsteroidal anti-inflammatory drug Antipyrine derivatives were also compatible under standard conditions, generating corresponding products in medium to good yields (49–50).

With the aim of further demonstrating the utility of our e-coupling, we were next interested in the reactivity of the peptides with double cysteine functions (Fig. 4). We showed that in the presence of Oxytocin precursor 51 and Antipyrine, double cross-dehydrogenative couplings occurred to deliver compound 52 in good conversion. These encouraging outcomes indicate that the method described herein can be applicable in the fields of chemical biology and medicinal chemistry. In recent years, cyclic peptide compounds have emerged as a prevalent topic in drug research and development, attributed to their distinctive physiological activities and structural diversity. Traditional small-molecule drugs frequently encounter issues such as suboptimal targeting and numerous side effects, whereas cyclic peptide compounds typically exhibit remarkable efficacy owing to their superior target specificity and high biological activity. Hence, the artificial chemical synthesis of cyclic peptide compounds to establish a structurally diverse cyclic peptide molecule library holds great significance and constitutes an important direction for future research and development of peptide drugs. To further demonstrate the synthetic value of this versatile technology, we sought to achieve cyclic peptide synthesis under our conditions. The success of the double cross-dehydrogenative couplings applied to a peptide containing two cysteine functions would pave the way to cyclic peptides via a stapling process under electrochemical conditions. This hypothesis was examined by selecting Oxytocin skeleton 51 and diolefin as substrate models. To our delight, the desired stapled peptide 53 was obtained in moderate conversion (Fig. 4). Next, we sought to achieve the intramolecular electrochemical C − H/S − H coupling of peptides carrying a free cysteine and a pyrazole moiety on the peptide backbone (Fig. 5), cyclization of the polypeptides bearing a cysteine at the N terminus proceeded smoothly to give preferentially the desired ring cyclic peptides 54–56 (22–16 membered ring) in 28–85% yields after HPLC purification. Similarly, hexapepides (pI-Phe-Trp-Pro-Cys), pentapepides (pI-Phe-Val-Pro-Cys), (pI-Phe-Val-Pro-Cys) proceeded smoothly to give preferentially the cyclic peptides 57–59 (25, 19-membered ring) in moderate yields (purified by preparative HPLC) even if in this non-optimized assay.

Fig. 4. Applications in S − H/C − H cross dehydrogenative coupling for Cys selective modification.

^c Reaction conditions: platinum plate anode, platinum plate cathode (15 × 15 x 0.3 mm), constant current = 6 mA, 48 (0.005 mmol), Antipyine (2.0 equiv.), ⁿBu₄NBr (2.0 equiv.), NiBr₂•diglyme (20 mol%), MeCN (6.0 mL), CHCl₃ (1.0 mL)_, undivided cell, room temperature, 1 h. ^d Reaction conditions: platinum plate anode, platinum plate cathode (15 × 15 x 0.3 mm), constant current = 6 mA, 48 (0.005 mmol), Bisenamine (4.0 equiv.), ⁿBu₄NBr (2.0 equiv.), NiBr₂•diglyme (20 mol%), MeCN (6.0 mL), CHCl₃ (1.0 mL)_, undivided cell, room temperature, 1.5 h.

Fig. 5. Applications in electrochemical peptides cyclization.

e Reaction conditions: platinum plate anode, platinum plate cathode (15 × 15 x 0.3 mm), constant current = 6 mA, 51 (0.05 mmol), ⁿBu₄NBr (2.0 equiv.), NiBr₂•diglyme (20 mol%), MeCN (6.0 mL), CHCl₃ (1.0 mL), undivided cell, room temperature, 30 min. Isolated yields (purified by preparative HPLC).

To collect more mechanistic information concerning this transformation, some mechanistic and control experiments were performed (Fig. 6). First, ESI-MS was applied to track the reaction process, under the standard reaction conditions, the electrophilic sulfenyl chloride moiety and dichloromethane were detected. When the reaction was carried out in the presence of TEMPO, the S-functionalized product was not detected. Meanwhile, the oxidation peak of cysteine was observed through cyclic voltammetry (CV) experiments (see supporting information). These results indicated that a Cys-S radical may be involved in the reaction process and the path of chlorine atoms throughout the electrochemical reaction process. The CHCl3 is first reduced at the cathode to generate the chloride ion and the dichloromethyl radical. Then, the generated dichloromethyl radical abstracts a hydrogen atom from the solvent to afford dichloromethane. In the meantime, the generated chloride ion combines with cysteine at the anode to generate the electrophilic sulfenyl chloride moiety. In addition, when 2.0 equivalents of sulfonyl chloride (SO₂Cl₂) were used instead of an electric current, the reaction mixture was complicated; however, the corresponding product was indeed produced in low yield (Fig. 6b1). This result further proves the possibility of the generation of electrophilic sulfenyl chloride in the reaction. Given that a considerable quantity of bromide ions is also present in the reaction system, and the distinct oxidation peaks of NiBr₂•diglyme and ⁿBu₄NBr can be observed through CV experiments as well. Although sulfenyl bromide was not detected in the mass spectrometry, the possibility that bromide ions play a role similar to that of chloride ions in the reaction cannot be excluded. To avoid the interference of bromide ions in the CV experiment, we choose to use Ni(cod)₂ instead of NiBr₂•diglyme, and ⁿBu₄NBF₄ instead of ⁿBu₄NBr. The Ni(cod)₂ displayed a dual anodic wave with a peak potential of 0.6 V and 1.3 V, which we attributed to the Ni⁰/Ni^I and Ni^I/Ni^II redox. Thus, a controlled potential electrolysis was carried out. When the potential of the anode was maintained at 1.3 V, the corresponding product was indeed obtained (See supporting information). Theoretically, the attack of thiol on the electrophilic sulfenyl halide intermediate does exist as well, which would result in the formation of the disulfide by-product. Nevertheless, after the reaction, we did not detect the formation of the disulfide product. Therefore, we carried out the reaction using the prepared disulfide cystine, and we found that the product could be obtained smoothly (Fig. 6b2). This result indicated that even if a disulfide bond can be formed during the reaction, the disulfide bond can be reduced to a radical anion at the cathode, and then release sulfur radical to participate in the subsequent reaction. Based on the above studies, a proposed mechanism for this electrochemical S − H/C − H cross dehydrogenative coupling is shown in Fig. 6. In the first step, cysteine is anodically oxidized to the thiyl radical A, which is captured by the Ni(II)-X to generate the Ni(III) species. A reductive elimination will produce the desired sulfenyl halide intermediate B. Subsequently, the reaction between the sulfenyl halide intermediate B and enamines occurs to give intermediate C, which is converted into the desired product via intramolecular hydrogen transfer. In the meantime, we cannot exclude the possibility that the halide ion may also be reduced at the cathode to generate a halogen radical, and then reacts with the thiyl radical A to furnish the intermediate B.

Fig. 6. Mechanistic experiments and proposed mechanism.

a Employing ESI-MS to monitor the reaction process. b Preliminary mechanism studies. c Proposed mechanism, metal-mediated halogen atom transfer pathway (Path A) and radical-radical coupling pathway (Path B).

Discussion

In conclusion, we report the electro-induced umpolung approach for the functionalization/macrocyclization of cysteine-containing peptides under mild reaction conditions. The designed Ni-catalyzed S − H/C − H cross dehydrogenative coupling obviates the requirement for precious metals, excessive toxic activation reagents and oxidants, as well as multiple reaction steps. This transformation can tolerate a wide range of valuable enamines, azoles, and peptides partners, and can also be used as a macrocyclization tactic for cyclic peptides synthesis and other areas. We anticipate that advances in electrochemically induced bioconjugation will lead to an expanding library of interdisciplinary methodologies.

Methods

General procedure

In an oven-dried undivided three-necked bottle (25 mL) equipped with a stir bar, cysteine residue (0.2 mmol), 3-Aminocrotononitrile (0.4 mmol), ⁿBu₄NBr (0.2 mmol) and NiBr₂•diglyme (20 mol%) were combined and added. Then, CH₃CN (6 mL), CHCl₃ (1 mL) were injected into the tubes via syringes. The bottle was equipped with a platinum plate (15 × 15 ×0.3 mm) as the anode and a platinum plate (15 × 15 ×0.3 mm) as the cathode. The reaction mixture was stirred and electrolysis was performed at a constant current of 12 mA under room temperature for 2.5 h. After completion of the reaction, as indicated by TLC and LC-MS, the pure product was obtained by flash column chromatography on silica gel (eluent: petroleum ether/ethyl acetate = 1:1).

Author contributions

M.G. conceived the project. F.X., X.B., Y.Z., J.Z., and Y.W. performed the experiments, analyzed the data, and discussed the results. M.G. and F.X. co-wrote the manuscript contributed to data analysis and scientific discussion; J.D. supported the polypeptide precursors.

Peer review

Peer review information

Nature Communications thanks Fei Ling, Sandip Murarka, and the other, anonymous, reviewer for their contribution to the peer review of this work. A peer review file is available.

Data availability

The authors declare that the data supporting the findings of this study are available within the paper and its Supplementary Information. Extra data are available from the corresponding author upon request. Crystallographic data for the structures reported in this Article have been deposited at the Cambridge Crystallographic Data Centre, under deposition numbers CCDC 2339717. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.

Competing interests

The authors declare that they have no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s41467-025-64615-4.