Biphasic electrochemical peptide synthesis

The large amount of waste derived from coupling reagents is a serious drawback of peptide synthesis from a green chemistry viewpoint. To overcome this issue, we report an electrochemical peptide synthesis in a biphasic system. Anodic oxidation of triphenylphosphine (Ph3P) generates a phosphine radical cation, which serves as the coupling reagent to activate carboxylic acids, and produces triphenylphosphine oxide (Ph3P 
<svg xmlns="http://www.w3.org/2000/svg" version="1.0" width="13.200000pt" height="16.000000pt" viewBox="0 0 13.200000 16.000000" preserveAspectRatio="xMidYMid meet"><metadata>
Created by potrace 1.16, written by Peter Selinger 2001-2019
</metadata><g transform="translate(1.000000,15.000000) scale(0.017500,-0.017500)" fill="currentColor" stroke="none"><path d="M0 440 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z M0 280 l0 -40 320 0 320 0 0 40 0 40 -320 0 -320 0 0 -40z"/></g></svg>
 O) as a stoichiometric byproduct. In combination with a soluble tag-assisted liquid-phase peptide synthesis, the selective recovery of desired peptides and Ph3PO was achieved. Given that methods to reduce Ph3PO to Ph3P have been reported, Ph3PO could be a recyclable byproduct unlike byproducts from typical coupling reagents. Moreover, a commercial peptide active pharmaceutical ingredient (API), leuprorelin, was successfully synthesized without the use of traditional coupling reagents.


Introduction
Recently, peptides have been recognized as candidates for "medium" molecular medicines, 1a which refers to pharmaceutical compounds whose molecular weights are roughly in the 1000 to 5000 range. This class of medicines has more specicity and fewer side effects than conventional small molecular medicines. 1b In 2018, over 60 peptides were approved as drugs in the United States, Europe and Japan. In addition, over 150 peptides are in active clinical development, and more than 260 have been tested in human clinical trials. 2 Even as peptides are gaining such a great deal of attention, the long-standing problem of the large amount of waste produced has not been resolved. 1 One contributor to waste generation is the use of coupling reagents that enable efficient peptide bond formation. 3 Generally, peptides are synthesized through repetitive deprotection and peptide bond formation reactions. To form peptide bonds between carboxylic acids and amines, stoichiometric amounts of coupling reagents are required (Scheme 1a), which generate stoichiometric byproducts, leading to accumulation of chemical waste. 4 In this regard, the development of greener peptide synthetic processes is identied as a key for green chemistry. 1a,e,g To address this issue, catalytic peptide synthesis has emerged (Scheme 1b, above arrow). Since Yamamoto reported boronic acid catalyzed amide bond formation, 5a various organoboron catalysts have been developed, 5b-g some of which have proven to be effective for oligopeptide synthesis. The mechanism of the reaction has been well studied, which has aided researchers in designing more sophisticated catalysts. In addition, recent studies revealed that several metals act as Lewis acid catalysts. These catalysts form active complexes with in situ generated esters, and promote peptide bond formation while avoiding epimerization. 5h,i We envisioned that development of an efficient peptide synthetic method using potentially recyclable reagents would pave the way to an alternative solution (Scheme 1b, below arrow). For this strategy, the use of phosphine (R 3 P) would be a promising approach. Oxidative activation of R 3 P generates an electrophilic phosphine cation, which serves as the coupling reagent to activate carboxylic acids and facilitate amide bond formation. 6 In 2019, the Arora group reported efficient oligopeptide synthesis using Bu 3 P activated by a diselenide catalyst. Although phosphine oxide (R 3 P]O) is produced as a stoichiometric byproduct, its reduction back to R 3 P has been achieved in various ways. 7 Therefore, R 3 P]O can be a recyclable byproduct, unlike byproducts from typical peptide coupling reagents. Among R 3 P, triphenylphosphine (Ph 3 P) would be most suitable because Ph 3 P is easy to handle and reduction of triphenylphosphine oxide (Ph 3 P]O) to Ph 3 P proceeds more efficiently than the case for alkyl phosphines. Moreover, Sevov reported an electrochemical reduction method applicable on a large scale, 7k and Favre-Réguillon has also succeeded in the conversion of Ph 3 P]O to Ph 3 P on a 100 g scale by the combination of Ti(OiPr) 4 and hydrosiloxane. 7i Therefore, development of peptide synthesis utilizing Ph 3 P and recovery of Ph 3 P]O would demonstrate the potential of Ph 3 P as a recyclable coupling reagent.
In this context, we aimed to develop a peptide synthesis method using Ph 3 P by performing electrochemical amide bond formation as reported by Frontana-Uribe. 8 In this electrochemical method, because electrons themselves act as an oxidant to generate a phosphine cation to facilitate amide bond formation, waste from chemical oxidants can be avoided. 9 To facilitate recovery of Ph 3 P]O, soluble tags as the carboxylic acid protecting groups could be combined with the electrochemical method. 10 Soluble tags are benzyl alcohols bearing long alkyl chains, so peptides protected with soluble tags dissolve in THF, CH 2 Cl 2 and c-Hex, but precipitate in polar solvents like MeCN. Hence, peptide synthesis is conducted in the liquid phase, and purication can be accomplished simply by ltration, which realizes facile separation of peptides and Ph 3 P]O. In this paper, we describe biphasic electrochemical peptide bond formation, leading to the synthesis of a commercial peptide active pharmaceutical ingredient (API), leuprorelin, without the use of traditional coupling reagents and recovery of Ph 3 P]O from the reaction mixture (Fig. 1).

Results and discussion
First, we optimized the reaction between Fmoc-Asp(O t Bu)-OH (1a) and H-Asp(O t Bu)-OTAG (2a) as models using Ph 3 P (Table  1). In CH 2 Cl 2 , the reaction proceeded efficiently to give the dipeptide (3aa) in 92% yield (entry 1). On the other hand, the yield was signicantly decreased to 19% in THF and most starting materials were recovered even when 4.8 F mol À1 was passed (entry 2). Since THF is more coordinating than CH 2 Cl 2 , it is likely that THF coordinates to Ph 3 Pc + to lower its electrophilicity, which leads to low efficiency of carboxylic acid activation. 11 A biphasic condition (MeCN/c-Hex) was also found to be effective, 12 affording 3aa in excellent yield (entry 3, Fig. S1 and Scheme S1 †). Hydrophobic 2a is localized in the upper c-Hex, while polar 1a and other reagents are selectively dissolved in the lower MeCN, where the electrochemical reaction takes place. The peptide bond formation between the activated 1a and 2a is expected to occur at the surface of the reversed micelle. Since 3aa is also localized in c-Hex, it is noteworthy that its separation from the supporting electrolyte is possible by simple phase separation. Furthermore, in the biphasic condition, since c-Hex is not conductive, undesired oxidation of 2a and/or overoxidation of 3aa could be suppressed. Avoiding the use of hazardous CH 2 Cl 2 is preferred from the green chemistry viewpoint, and therefore, further investigations were carried out using the biphasic condition.
When LiClO 4 or KClO 4 was used instead of Bu 4 NClO 4 , the yields decreased somewhat (entries 4 and 6), while the use of NaClO 4 caused severe gelation of the reaction mixture (entry 5). These observations suggest that tetrabutylammonium ions are more suitable for the reaction than alkali metal cations, presumably due to weaker interactions with the carboxylate anion of 1a. Moreover, 31 P NMR indicated that the  transformation of Ph 3 P into Ph 3 P]O was selective. Aer further investigation of the procedure, we succeeded in the recovery of both Bu 4 NClO 4 (>99%) and Ph 3 P]O (91%) with high purity (>95%) without column purication (detailed procedure and purity assessment are in ESI †).
Having conrmed the potential of Ph 3 P as a recyclable coupling reagent, we explored the feasibility of electrochemical peptide bond formation with other Fmoc-protected amino acids (1b-1u) instead of 1a, using 2a as a reaction partner ( Table 2). Although heating was needed in some cases to avoid gelation of the reaction mixture, all the Fmoc-protected canonical amino acids were demonstrated to be amenable to the reaction. It should be noted that not only amino acids with redox inactive alkyl side chains but also those with redox active moieties, such as Cys and Met, gave the desired products in excellent yields. When Fmoc-His(Trt)-OH was used (3pa), however, partial epimerization occurred, likely due to the basicity of the imidazole moiety. 13 Boc is an electron-withdrawing alternative to Trt for the protection of imidazole, which may decrease its basicity. Using Boc, we succeeded in preventing the epimerization and the desired product (3qa) was obtained in excellent yield as a pure stereoisomer. Slightly excessive amounts of reagents and Edge Article Chemical Science electricity were needed for the reaction of Fmoc-Pro-OH (3ua), presumably due to steric hindrance. We next turned our attention to explore the compatibility of this method with other tag-protected amino acids (2b-2u) instead of 2a, using 1a as a reaction partner (Table 3). Although heating was also needed in some cases to avoid gelation of the reaction mixture, all the tag-protected canonical amino acids were demonstrated to be amenable to the reaction. In this case, no signicant epimerization was observed for the reaction of H-His(Trt)-OTAG, and the desired product (3ap) was obtained in excellent yield as a pure stereoisomer. The reaction of H-Pro-OTAG (3au) was less efficient than others, requiring slightly higher amounts of reagents and electricity. It should be noted that the protecting groups commonly employed in conventional peptide synthesis were proven to be compatible with the reaction, and therefore, commercially available building blocks could be used directly in this method.
Having conrmed that electrochemical peptide bond formation can be applied to all canonical amino acids, we applied this method to the synthesis of a commercial peptide active pharmaceutical ingredient (API), Leuprorelin, as a model (Scheme 2A, see ESI † for experimental details). 14 Starting from Fmoc-Pro-NEtTAG (4), 10d all 8 peptide bonds were electrochemically formed. Deprotections of Fmoc were carried out under typical conditions using DBU and piperidine. Aer repetitive deprotections and couplings, the protected form of leuprorelin (5) was successfully obtained in 45% yield over 16 steps (56% by weight, and its purity was estimated to be 81% by HPLC, Scheme 2B and Fig. S2 †). The average yield in each step was >95%. Finally, acidic global deprotection, which is commonly used in conventional peptide synthesis, afforded leuprorelin (6) quantitatively (Fig. S3 †). This result demonstrates that electrochemical peptide synthesis has the potential to be an alternative choice for conventional SPPS and/or LPPS.

Conclusions
We succeeded in developing a biphasic electrochemical peptide bond formation reaction using Ph 3 P. Electrochemically generated Ph 3 Pc + promoted efficient peptide bond formation with no signicant epimerization. Protecting groups commonly employed in conventional peptide synthesis were proven to be compatible with the reaction, and therefore, commercially available building blocks could be used directly in this method. In combination with a soluble tag-assisted LPPS, the synthesis of a commercial peptide API, leuprorelin, was achieved without the use of traditional coupling reagents. Although using Ph 3 P as a coupling reagent under electrochemical conditions results in the generation of Ph 3 P] O, the biphasic system enabled selective recovery of the desired peptides and Ph 3 P]O by simple phase separation. Given that methods to reduce Ph 3 P]O to Ph 3 P have been reported, Ph 3 P is a potentially recyclable coupling reagent. Therefore, this soluble tag-assisted electrochemical method can be used to address sustainability challenges in peptide synthesis. Aiming for further improvement, we are exploring novel methods to reduce Ph 3 P]O to Ph 3 P.

Data availability
The datasets supporting this article have been uploaded as part of the ESI. †