N‐ to C‐Peptide Synthesis, Arguably the Future for Sustainable Production

ABSTRACT

A revolution in peptide production arrived from the innovation of carboxylate to amine C‐ to N‐direction solid‐phase synthesis. This cornerstone of modern peptide science has enabled multiple academic and industrial applications; however, the process of C‐ to N‐solid phase peptide synthesis (C‐N‐SPPS) has extreme process mass intensity and poor atom economy. Notably, C‐N‐SPPS relies upon the use of atom‐intensive protecting groups, such as the fluorenylmethyloxycarbonyl (Fmoc) protection and wasteful excess of protected amino acids and coupling agents. On the other hand, peptide synthesis in the amine to carboxylate N‐ to C‐direction offers potential to minimize protection and may arguably enable more efficient means for manufacturing peptides. For example, efficient amide bond formation in the N‐ to C‐direction has been accomplished using methods employing thioesters, vinyl esters, and transamidation to achieve peptide synthesis with minimal epimerization. This review aims to provide an overview of N‐ to C‐peptide synthesis indicating advantages in taking this avenue for sustainable peptide production.

The advent of epimerization‐free methods for coupling in N‐ to C‐direction synthesis offers improved atom economy for sustainable peptide synthesis.

1. Introduction

Peptides are short poly‐amide chains formed from amino acids. Playing vital roles in various biological processes, peptides function as hormones, neurotransmitters, growth factors, and defensins. Peptide synthesis is critical for exploring such activities and for harnessing the poly‐amide structure to develop therapeutics [1, 2]. As therapeutics, peptides offer attributes including high potency and low toxicity combined with drawbacks such as limited bioavailability and rapid metabolism [1, 2]. Chemical modifications to overcome such limitations can add additional synthetic challenges [3, 4]. The advent of solid‐phase peptide synthesis (SPPS) provided an effective method for procuring peptides due in part to ability to automate a process in which excess reagent could be employed to drive reactions to completion and by‐products could be removed by simple filtration [5, 6, 7].

Solid‐ and liquid‐phase peptide synthesis (SPPS and LPPS) have become widely used platform technologies for product procurement; however, these methods commonly lack atom economy and possess relatively high process mass intensity (PMI) raising environmental concerns [3, 4]. Compared to that of small molecules, the average PMI for SPPS is approximately 40 to 80 times more intensive [3]. Moreover, SPPS has a PMI that is nearly twice as large as that for the making of biopharmaceuticals [3]. Efficient and environmentally friendly methods are required to overcome the high PMI in peptide manufacturing, which is distributed over the synthesis and purification steps to provide pure product. With the growing use of peptides in pharmaceuticals, demand increases for more optimal synthetic and purification manufacturing technologies [3, 7, 8].

Typically, SPPS is performed by linking the carboxylate terminal to the resin and elongation in the C‐ to N‐direction [5, 6]. A key advantage of such an approach has been the ability to minimize racemization through the employment of carbamate nitrogen protecting groups [9]. The advantages of carbamates are however balanced by the waste produced in repetitively using such temporary protection in the synthetic process. In contrast, nature synthesizes peptides in the amine to carboxylate N‐ to C‐direction relying heavily on the ribosome to ensure that amide bond formation proceeds without racemization [10]. In the history of solid‐phase peptide synthesis, the N‐ to C‐direction elongation approach originated at the same time as the more well practiced C‐ to N‐method [5, 11]. A recent renaissance of methods for N‐ to C‐peptide synthesis without racemization has emerged driven in part by desires for more environmentally friendly manufacturing.

This overview is intended to provoke alternative thinking by highlighting developments in methods for N‐ to C‐peptide synthesis. Tolerance is requested for lack of comprehensiveness and for including research performed in 2000 within the twentieth century perspective because of a relationship to earlier related methods. In reviewing historical and recent methods, perspective is offered for reconsidering a different more natural direction for sustainable peptide production.

2. Historical Twentieth Century Background

In 1963, Merrifield introduced C‐ to N‐solid‐phase peptide synthesis (C‐N‐SPPS, Scheme 1) [5]. After nitration of cross‐linked polystyrene 1, tetrapeptide 10 was prepared using sequential carbodiimide couplings and deprotections of N‐benzyloxycarbonyl (Cbz)‐protected amino acids with strong acid. Saponification was finally used to cleave the resin.

SCHEME 1.

Merrifield C‐ to N‐direction solid‐phase peptide synthesis.

After much refinement, C‐N‐SPPS remains the contemporary paradigm for peptide production [6, 7], because sequential elongation enables use of N‐protected amino acids without the need for purification after each coupling and deprotection step. Immobilization of the growing peptide chain on an insoluble polymer resin simplified the purification process by eliminating the isolation of intermediates and paved the way for the construction of longer and more complex peptides [6, 7].

In 1963, Letsinger and Kornet introduced N‐ to C‐solid‐phase peptide synthesis (N‐C‐SPPS) [11], stating that the “recent publication by Merrifield describing the synthesis of a tetrapeptide on polymer beads prompt us to report at this time experiments carried out independently which demonstrate the feasibility of using a modified ‘popcorn' polymer as a supporting matrix in repetitive‐step syntheses.” The “popcorn” polymer was, in fact, a variant of the reticulated polystyrene that Merrifield had employed except that it was converted into chloroformate 12 for introducing the first residue as an amino ester (Scheme 2). In contrast to the tetrapeptide prepared by C‐N‐SPPS [5], a dipeptide was made in the N‐ to C‐direction using ester saponification, mixed anhydride coupling, and resin cleavage with strong acid [11].

SCHEME 2.

Letsinger N‐ to C‐direction solid‐phase peptide synthesis.

Commenting on the N‐C‐SPPS in 1965, Merrifield stated that although “a dipeptide was synthesized in this way, the danger of racemization might limit the usefulness of this procedure for longer peptides” [12]. The so‐called “danger of racemization” was known at the time based on various pioneering studies. For example, as reviewed by Neuberger [13], Bergmann and Zervas hypothesized in 1929 that amide protected amino acids could undergo racemization by way of oxazolone intermediates [14]. By the early sixties, various researchers, including Bodanszky [15], Goodman [16, 17], and Williams and Young [18, 19], all had confirmed the problem of racemization of oxazolone intermediates in carboxylate activation of peptides and noted faster rates of epimerization compared with carbamate protected amino acid counterparts. For example, 84‐fold higher rates of racemization were observed in the coupling to valine methyl ester of N‐Cbz‐glycyl‐L‐methionine pentafluorophenyl ester compared to N‐Cbz‐L‐methionine counterpart in tetrahydrofuran (THF) using triethylamine as base [20].

The reason for the difference in racemization rates was elegantly examined by Benoiton, who in the 1980s, reported on the optical integrity of the 2‐alkoxy‐5(4H)‐oxazolones, which were used in coupling reactions without racemization [21]. In head‐to‐head comparisons, “the chiral stability” of the 2‐alkoxy‐5(4H)‐oxazolones (e.g., S‐19a and S‐19b) contrasted with that of the 2‐alkyl‐5(4H)‐oxazolones (e.g., S‐19c and S‐19d) which could not “be coupled without extensive racemization” [21]. In other words, both carbamate and amide protected amino acids underwent oxazolone formation; however, the 2‐alkoxy‐ was less prone than the 2‐alkyl‐5(4H)‐oxazolone to form the hydroxy‐oxazole intermediate (e.g., 20) and epimerize prior to coupling (Figure 1). Notably, under basic conditions, both oxazolones were susceptible to epimerization but at different rates [21].

FIGURE 1.

Oxazolone intermediate lability in peptide synthesis (Bn = benzyl, Cbz = benzyloxycarbonyl).

Inspired by reports that alternative activation methods could avoid racemization of amide protected amino acids [19], especially the conditions of Honzl and Rudinger for preparing and coupling N‐protected amino azides without side product from Curtius rearrangement [22], Felix and Merrifield examined N‐C‐SPPS in 1970 using peptide azides (Scheme 3) [23]. In a proof of concept synthesis, tetrapeptide 10 (H‐Leu‐Ala‐Gly‐Val‐OH) was prepared using N‐C‐SPPS with acyl azide activation and shown to possess identical properties as material made by C‐N‐SPPS [23].

SCHEME 3.

Azide activation in N‐C‐SPPS after N′‐protected hydrazide couplings and deprotections (Boc = tert‐butyloxycarbonyl).

Chloroformate resin 12 was reacted with L‐leucine N′‐Boc‐hydrazide and triethylamine in chloroform, then capped on treatment with anhydrous dialkylamine. The Boc protection was removed using 4 M HCl in dioxane, and the resulting hydrazide 23 was treated with n‐butyl nitrite in tetrahydrofuran (THF) at −30 °C to provide acyl azide 24. Notably, THF gave more reproducible results than dimethylformamide (DMF). After a sequence of couplings of acyl azides with amino N′‐Boc‐hydrazides in THF, acid mediated Boc removal, and nitrite activation, the protected peptide was terminated using H‐Val‐Ot‐Bu. The final peptide, H‐Leu‐Ala‐Gly‐Val‐OH (10) was cleaved from the resin using HBr in trifluoroacetic acid (TFA) and purified on a column of Dowex 50‐X4 resin eluting with a pyridine acetate buffer at pH 4.0. In concluding the study, the authors stated that the “overall yields are presently marginal and we do not recommend this new approach as a general replacement for solid phase peptide synthesis. However, it may be a valuable supplement to peptide synthesis …” [23].

Subsequently, the Mukaiyama laboratory reported N‐C‐SPPS using thiopyridyl ester activation (Scheme 4) [24, 25]. In a one‐pot process, activation of N‐Cbz‐leucine with 2,2′‐dipyridyl disulfide (PySSPy) and triphenylphosphine gave the corresponding N‐Cbz‐amino thiopyridyl ester which reacted with ethyl glycinate to provide the corresponding dipeptide Cbz‐Leu‐Gly‐OEt in a 96:4 enantiomeric ratio [26]. The thiopyridyl ester activation strategy was applied in preparations of luteinizing hormone‐releasing hormone (LHRH) and adrenocorticotropin‐(1–24) by N‐C‐SPPS. The final peptides were claimed to exhibit comparable activity as the native peptides [24, 25].

SCHEME 4.

Application of thiopyridyl esters in N‐C‐SPPS.

In the using thiopyridyl ester activation strategies, peptide fragments of typically four residues or less were coupled typically to C‐terminal glycine and proline residues to minimize potential for epimerization. Fragment couplings were typically performed after preparation of the thiopyridyl ester (e.g., 30) using PySSPy, triphenylphosphine, and 2‐mercaptopyridine, all in 30‐fold excess in dichloromethane. The C‐terminal carboxylates were protected as tert‐butyl esters (e.g., 28 and 31). The side chains were protected respectively as benzyl ethers and carbamates as well as nitro and N‐toluenesulfonyl guanidine groups to enable selective tert‐butyl ester removal and liberation of the C‐terminal carboxylate using TFA. Finally, resin and protecting group removal were accomplished using hydrofluoric acid. In addition, replacement of 2,2′‐dipyridyl disulfide with the corresponding N‐oxide, 2,2′‐dithiopyridine‐1,1′‐dioxide was found to enhance yields and minimize racemization in the so‐called oxidation–reduction condensation coupling reaction [27].

At the end of the twentieth century, strategies were investigated employing 2‐chloro trityl anchors to avoid the strong acid conditions that were used in the cleavage of the cross‐linked polystyrene carbamate resin (Scheme 5) [28, 29]. Two different amino esters were independently examined. Allyl esters (e.g., 36) were removed using catalytic tetrakis (triphenylphosphine) palladium(0) and phenylsilane in dichloromethane (DCM) [30]. Fluorenyl methyl (Fm) esters (e.g., 37) were taken off using 20% piperidine in DMF [31]. The application of the 2‐chloro trityl anchor enabled convenient cleavage for examining carboxylate activation employing coupling agents commonly employed in C‐N‐SPPS: diisopropylcarbodiimide (DIC), 2‐(1H‐benzotriazole‐1‐yl)‐1,1,3,3‐tetramethylaminium tetrafluoroborate (TBTU), and azabenzotriazole counterpart HATU (2‐(1H‐7‐azabenzotriazole‐1‐yl)‐1,1,3,3‐tetramethylaminium hexafluorophosphate).

SCHEME 5.

N‐C‐SPPS on 2‐chloro trityl resins.

In the case of couplings using DIC activation in the presence of 1‐hydroxybenzotriazole (HOBt) as additive, impurities from failed couplings were identified by mass spectrometric analyses, which revealed molecular ions corresponding to products from intramolecular cyclization: diketopiperazine and oxazolone [31]. Epimerization and yield varied contingent on coupling agent. The DIC/HOBt conditions provide product of better configurational integrity but poorer yield compared to those featuring TBTU with N‐methyl morpholine (NMM) as base, which gave product in higher yield but with increased degrees of epimerization [30, 31]. The employment of bis(1‐hydroxybenzotriazole) copper complex [Cu(OBt)₂] diminished epimerization significantly in the DIC coupling protocol [30]. In the synthesis of Leu‐enkephalin (39, H‐Tyr‐Gly‐Gly‐Phe‐Leu‐NH₂), the combination of Cu‐(OBt)₂ and DIC afforded product of higher purity (57% vs 21% HPLC purity) without racemization compared to that obtained using HATU and diisopropylethylamine (DIEA) which gave 5% of a diastereomer [30].

In a related approach employing a photo‐cleavable carbamate linker on a TentaGel™ resin, five tripeptides were synthesized with minimal epimerization by N‐C‐SPPS using amino acid tri‐tert‐butoxysilyl esters, HATU, and 2,4,6‐trimethylpyridine (TMP, collidine) as base without pre‐activation [32]. A notably higher level of epimerization occurred in the coupling of the silyl ester of isoleucine to the resin‐bound Asp(Fm)‐Ser(t‐Bu)‐OH dipeptide. Cupric chloride was found to suppress epimerization at the expense of causing resin cleavage and lower loading. Suppression of epimerization was suggested to be due to CuCl₂ induced ring opening of oxazolone intermediates. The amino acid tri‐tert‐butoxysilyl esters proved stable as the hydrochloride salts and the corresponding peptide silyl esters could be cleaved effectively with 5% trifluoroacetic acid in dichloromethane without loss of acid labile tert‐butyl (t‐Bu) ether functionality.

In the research from the twentieth century, a solid foundation was set for the development of N‐C‐SPPS. Systematic analysis of peptide quality during synthesis was facilitated by employing the 2‐chloro trityl anchor which could be cleaved under mild conditions [30, 31]. Key observations were made for minimizing racemization. Pioneering research on N‐C‐SPPS was not focused on improving PMI; however, the shift of protection from the amine to the carboxylate side of the amino acid building block offered in principle means to improve atom economy. Significant excess of activating agent and amino ester were however employed in methods using thiopyridyl esters, and DIC, TBTU, and HATU couplings [24, 30, 31]. In addition, such chemistry was often performed using toxic solvents that should be avoided today [33].

3. N‐C‐SPPS in the Twenty‐First Century

Developments of N‐C‐SPPS over the past quarter century have been driven in part by interest to access peptides with C‐terminal modifications and to illustrate utility of novel coupling strategies. The changes in allowable solvents for peptide manufacturing during this time period have been similarly acknowledged in certain research programs and are noted below. Many of the mentioned methods have yet to be demonstrated on solid phase but are included because of potential utility in peptide production.

3.1. N‐C‐SPPS in C‐Terminal Modification

Towards C‐terminal modification, tripeptide hydroxamic acids (e.g., 43), as well as tetrapeptide boronic acid and trifluoromethyl ketone derivatives (e.g., 44 and 47), all were respectively synthesized by N‐C‐SPPS commencing with chloroformate resin 12 (Scheme 6) [34]. Sequential elongation was performed using amino tert‐butyl esters, ester solvolysis with 50% TFA in DCM, and HATU/collidine couplings with minimal (<2%) racemization. After elongation of the tripeptide acids, HATU/collidine couplings of O‐tert‐butyl hydroxylamine, 2‐amino‐1‐(trifluoromethyl)‐1‐propanol, and 1‐aminoethylboronate pinacol ester introduced the respective C‐terminal modifications. Pfitzner–Moffat oxidation of the trifluoromethyl alcohol resins (e.g., 46) gave the corresponding ketones. Resin cleavage with concomitant removal of borate protection was performed using 10% trifluoromethanesulfonic acid in TFA.

SCHEME 6.

Application of N‐C‐SPPS for peptide C‐terminal modifications (DCC = N,N′‐dicyclohexylcarbodiimide).

3.2. N‐C‐SPPS With Acyl Azides, Revisited

In homage to the acyl azide approach developed by Felix and Merrifield [23], a strategy for N‐C‐SPPS was developed based on benzyl ester hydrazinolysis and activation of the corresponding hydrazides using sodium nitrite (Scheme 7) [35]. The effectiveness of the method was exemplified by the synthesis of Ac‐Phe‐Phe‐Gly‐Asn‐Ser‐Leu‐Val‐OH (53). Elongation of the peptide was performed in both directions from an asparagine residue which was attached by coupling Fmoc‐Asp‐OBn to Rink amide resin.

SCHEME 7.

Acyl azide approach based on benzyl ester hydrazinolysis.

Initially, the Fmoc group was removed and C‐ to N‐direction extension of the peptide was performed to create N‐acetyl tetrapeptide resin 48. Hydrazinolysis of benzyl ester 48 using 20% hydrazine hydrate in DMF gave hydrazide 49, which was transformed to acyl azide 50 using acidified sodium nitrite. Reaction with H‐Ser‐OBn gave N‐acetyl peptide benzyl ester 51, which was elongated by sequential hydrazinolysis, followed by activation and aminolysis of acyl azide intermediates. Finally, resin cleavage and side chain protecting group removal were performed using a cocktail of TFA/H₂O/triisopropylsilane (TIPS). Peptide 52 was claimed to be prepared by this route with similar purity and yield as product from C‐N‐SPPS [35]. In a related activation method, dipeptides were synthesized without epimerization by way of acyl nitroso intermediates which were generated from the oxidation of N‐protected hydroxamates in the presence of amino esters using iodine in dimethyl sulfoxide (DMSO) [36].

3.3. Peptide Synthesis in the N‐ To C‐Direction Using Thioacids

Since the pioneering studies in the 1950s by Wieland [37], who was inspired by the biosynthetic use of S‐acetyl‐coenzyme A, thiocarboxylic acid derivatives have served in various approaches for peptide bond formation [38, 39], most notably, native chemical ligation [40]. For example, dipeptides have been synthesized in a two‐step process entailing thioacid synthesis using thioacetic acid and sodium hydrogen sulfide [41]. Oxidative dimerization of the thioacids to the diacyl disulfide followed by amino acylation was shown to give dipeptides without racemization [41].

Mild conditions for the synthesis of amino thioacid and peptide thiocarboxylates (e.g., 55) were developed employing potassium thioacetate (AcSK) as a cost‐effective source of sulphur and catalytic diacetylsulfide (20 mol% Ac₂S) in DMF (Figure 2) [42]. In the proposed mechanism, mixed anhydride 58 is formed on reaction of the peptide with Ac₂S and undergoes anhydride exchange with AcSK to yield the mixed diacylsulfide 59. Regeneration of the Ac₂S catalyst upon thioacid–anhydride exchange generates the potassium salt of the peptide thioacid (e.g., 57). Excess AcSK drives the equilibria at a rate that surpasses potential oxazolone formation and concomitant epimerization.

FIGURE 2.

Synthesis of peptide thiocarboxylate.

Peptide thioacids were employed in liquid‐phase N‐ to C‐peptide synthesis without protection of carboxylic acid, alcohol, imidazole, and indole side chains (Figure 3) [43]. N‐Hydroxy‐2‐pyridinone methyl and 3,7,11,15‐tetramethylhexadecan‐1‐yl esters (61) were employed respectively as additives in the coupling reaction. Air oxidation of thioacid 55 in a DMSO/toluene mixture affords diacyl disulfide 62 which may be intercepted by the N‐hydroxy‐2‐pyridinone additive to provide active ester 63 and acyl disulfide 64. In principle, diacyl disulfide 62, active ester 63, and acyl disulfide 64, all may react with the amine nucleophile to provide the desired amide 60 and respectively disulfide 64, additive 61, and hydrogen disulfide. Acyl disulfide 64 may also react with the N‐hydroxy‐2‐pyridinone additive to give active ester 63 and H₂S₂, which is further oxidized to molecular sulfur S₈ and water. Employing optimized conditions, amino acid and peptide fragment coupling occurred in high yield with minimal epimerization.

FIGURE 3.

N‐ to C‐synthesis employing peptide thiocarboxylate.

3.4. Peptide Synthesis in the N‐ To C‐Direction and N‐C‐SPPS Using Vinyl Esters

Vinyl esters have been generated from carboxylic acid addition to different acetylene derivatives and employed in amide bond formation [44]. Since pioneering studies in the 1950s by Arens, alkoxyacetylenes have served as activating agents for peptide coupling [45, 46, 47]. Moreover, racemization‐free ruthenium‐catalyzed addition of protected dipeptides to alkynes gave the corresponding C‐terminal peptide enol esters which were employed in enzyme‐catalyzed fragment couplings [48].

Inherent challenges in the use of vinyl esters are due in part to the reactivity of the alkyne precursor and the acyl enol ether counterpart in the activation and amino acylation steps, respectively [47, 49]. Innovations from the Zhao lab have overcome such challenges through the respective uses of allenones and ynamides as activating agents in peptide coupling with minimal racemization [47, 50, 51, 52]. For example, 1‐phenylbuta‐2,3‐dien‐1‐one (65) was used in the N‐ to C‐solution‐phase synthesis of the proteasome inhibitor carfilzomib (73, Scheme 8) [50]. Amino acyl activation was typically performed with enone 65 at room temperature in dichloroethane to form the vinyl esters (e.g., 67) which were isolated by chromatography and subsequently coupled to the respective tert‐butyl amino ester at room temperature in DMF. Subsequently, the tert‐butyl ester was solvolyzed with TFA in DCM and the sequence was repeated. By commencing with chloroacetic acid and finishing with chloride displacement by morpholine with potassium iodide in THF, the anticancer peptide drug, carfilzomib (73) was accomplished in a nine‐step linear protocol with 68% overall yield without epimerization [50].

SCHEME 8.

N‐ to C‐solution‐phase synthesis of carfilzomib using allenone activation.

Ynamines derived from various amines were examined as coupling agents for peptide synthesis soon after the use of alkoxyacetylenes [53]; however, epimerization was encountered likely due to the alkaline nature of the reagents [53]. To overcome the issue of racemization and provide an electron deficient acetylene for effective carboxylate activation, the Zhao laboratory has explored N‐(methyl)ynetoluenesulfonamide (74) and N‐(methyl)ynemethylsulfonamide (75) as ynamides in peptide synthesis (Scheme 9) [51, 52]. Sulfonyl ynamides 74 and 75 are stable and have been employed in N‐ to C‐direction peptide synthesis in solution and on solid phase with high yields and minimal epimerization [48]. Mechanistic studies supported by X‐ray analysis of the adduct from pivalic acid and toluenesulfonyl ynamide 74 have indicated that the ester carboxylate bond of the α‐acyloxyenamide is elongated relative to those of vinyl and aryl ester counterparts [51]. The rate‐limiting step in the aminolysis reaction was deduced to be formation of the tetrahedral intermediate, collapse of which is energetically favored due to keto–enol tautomerization of the leaving enolate [51]. α‐Acyloxyenamide synthesis was typically conducted by activation of carboxylic acid 76 with ynamide 74 in dichloromethane. After evaporation of the volatiles, α‐acyloxyenamide 77 was reacted effectively with amine nucleophiles in DMSO/H₂O mixtures, the polarity of which was suggested to favor formation of the rate‐limiting tetrahedral intermediate. Alcohol (Ser and Thr), amide (Gln), and indole (Trp) side chains, all were tolerated in the aminolysis reaction. Alternatively, in situ silylation with bis(trimethylsilyl)acetamide (BSA) was used to provide amino trimethylsilyl esters that reacted with the α‐acyloxyenamide 77 in DMF [51].

SCHEME 9.

N‐ to C‐solution‐phase synthesis using ynamide activation.

Twenty‐one amino acid peptide 85 was prepared by N‐C‐SPPS using methansulfonyl ynamide 78 (Scheme 10) [51]. Onto Rink amide MBHA AM resin, Boc‐Glu‐OFm was anchored by way of the side chain and converted to carboxylic acid 80 with 20% piperidine in DMF. Subsequently, the carboxylate activations were performed using ynamide 78 in dichloroethane (DCE) or in 7:3 DCE/sulfolane to provide vinyl esters (e.g., 81). Aminolysis reactions employed amino silyl esters in DCE or 2‐methyltetrahydrofuran (2‐MeTHF). The amino silyl esters were prepared using 1‐(trimethylsilyl)pyrrolidone (TMSP) and the resulting peptide silyl esters (e.g., 82) were cleaved using citric acid in methanol to provide carboxylates (e.g., 83) for subsequent activation. Side chain alcohol (Ser), phenol (Tyr) and carboxylate (Glu) functions were respectively protected with tert‐butyl groups and the ε‐amine of lysine was Boc protected. Resin cleavage and side chain liberation was performed on protected peptide 84 using a TFA/H₂O/TIPS cocktail. Peptide 85 was obtained in 50% purity, which was comparable with the 57% purity achieved in the preparation of the same product using conventional C‐N‐SPPS [51].

SCHEME 10.

N‐C‐SPPS using methansulfonyl ynamide activation.

3.5. Peptide Synthesis in the N‐ to C‐Direction by Transamidation

Transamidation chemistry in which amine exchange is performed on the amide bond by transamidase enzymes, such as sortase and transglutaminase, has significant biological importance and potential for environmentally benign synthesis [54]. A thermodynamically neutral transformation of an intrinsically stable poor electrophile, transamidation has also been achieved on ordinary amides under ambient conditions through the employment of transition metal catalysis and under strongly basic conditions [55, 56, 57]. In the context of N‐ to C‐direction synthesis, polyamide mixtures were prepared under so claimed “prebiotic conditions” in which N‐acetamido amides were heated with amino amides in 0.5 M NaOH at 80 °C [58]. Moreover, peptides have been synthesized by ring‐opening of N‐alkyl N‐Boc‐diketopiperazines without epimerization on heating with amino esters in the neat and with peptide esters in toluene with a benzoic acid catalyst [59]. In addition, a number of methods have been conceived for performing overall transamidation chemistry by way of thioester intermediates in native chemical ligation strategies towards peptide products [60, 61, 62].

In two N‐ to C‐peptide synthesis strategies, transamidation has relied on the respective usefulness of the reactivity of N‐alkyl N‐acylsulfonamides [63, 64], and of zinc‐complexes of 2‐amido pyridine intermediates [65, 66, 67]. The switch in reactivity from relatively stable N‐acylsulfonamide (e.g., 88) to reactive N‐alkyl N‐acylsulfonamide (e.g., 89) has been effectively employed in the “safety catch principle” in solid‐phase peptide synthesis [68], and hypothesized to result from amide bond twisting from planarity [64]. Without resonance stabilization, twisted amides exhibit increased carbonyl electrophilicity as indicated by downfield shifted ¹³C, ¹⁵N, and ¹⁷O chemical shifts [69, 70]. Moreover, NMR spectroscopic studies of deuterium exchange have indicated that epimerization during aminolysis is suppressed by an intramolecular hydrogen bond between the N‐alkyl sulfonamide oxygen and the amide NH of the activated amino acyl residue (e.g., 90, Scheme 11) which prevents oxazolone formation [64]. Activation has been achieved by a two‐step method commencing with carboxylic acid (e.g., 86) conversion into the N‐acylsulfonamide (e.g., 87) on reaction with p‐toluenesulfonyl isocyanate (TsNCO) and triethylamine in THF to form a carboxylic acid–carbamic acid mixed anhydride intermediate (e.g., 87) which loses carbon dioxide. Alkylation of N‐acylsulfonamide 88 on heating with pentafluorobenzyl bromide and N,N‐diisopropylethylamine in DMF activates amide 89 which undergoes transamidation with amino esters at room temperature [64].

SCHEME 11.

N‐ to C‐direction synthesis by way of N‐alkyl N‐acylsulfonamide.

Transamidation of 2‐amido pyridines has also been used for N‐ to C‐direction peptide synthesis [65, 66, 67]. Considering the mechanisms of action of metalloproteases, zinc catalysis was examined in the esterification and transamidation of peptide N′‐2‐aminopyridine amides (Scheme 12). Different peptide N′‐2‐aminopyridine amides were synthesized by two protocols employing respectively primary amides (e.g., 92a) in palladium‐catalyzed cross‐coupling with tert‐butyl 2‐chloronicotinate [65, 66], as well as amino acids (e.g., 92b) in four‐component Ugi reactions with 2‐pyridyl isocyanides, aldehydes and N‐alkylamines in trifluoroethanol [67]. With peptide N′‐2‐aminopyridine amides in hand, zinc‐catalyzed transamidation reactions were performed using 20 mol% of zinc acetate in THF with sodium acetate as a mild base to avoid epimerization [65]. Examination of other metals (e.g., Co, Ni, and Cu) and ligand metal complexes provided evidence to support the importance of a zinc acetate pyridyl amide complex (e.g., 95) for effective transamidation. Synthesis in the N‐ to C‐direction was in principle achieved by employing hydrochloride salts of amino amides as nucleophiles in the coupling step to enable subsequent Pd‐catalyzed cross‐couplings with tert‐butyl 2‐chloronicotinate and zinc‐catalyzed transamidations [65].

SCHEME 12.

N‐ to C‐direction synthesis by way of peptide N′‐2‐aminopyridine amides.

4. Perspective

In spite the similar starting time, Merrifield C‐N‐SPPS quickly outpaced N‐C‐SPPS, due in great part to challenges of avoiding epimerization in peptide coupling steps. Moreover, linking strategies for amino groups in peptide chemistry have arguably been less well developed than those for the attachment of carboxylic esters. As reviewed on entering the twenty‐first century [71], C‐N‐SPPS by the Fmoc/t‐Bu approach was the most common method for the production of peptides; however, several factors have rendered this approach less desirable including needs to minimize product mass intensity (PMI) and to transition towards environmentally friendly solvents and reagents [3, 4, 33]. As discussed above, several innovative methods have emerged for N‐ to C‐direction peptide synthesis. Notably, fears of the “danger of racemization” have been assuaged by the development of various coupling protocols that minimize epimerization including approaches using thioacids [43], vinyl esters [52], and transamidation [65]. Such coupling strategies have demonstrated potential for minimizing the protection of side chains. In general, carboxylate protection in N‐ to C‐direction peptide synthesis has typically better atom economy than the carbamate protecting groups used in the C‐ to N‐counterpart. In addition, N‐C‐SPPS gives effective access to carboxylate precursors for derivatization of the peptide C‐terminal [34].

In comparing C‐ to N‐ versus N‐ to C‐direction peptide synthesis, the various uses of the former approach in the synthesis of products for academic and industrial purposes cannot be overlooked. For example, the preparation of peptide combinatorial libraries for screening against various biological targets is among the most impressive applications of C‐N‐SPPS [72, 73, 74, 75]. Although N‐C‐SPPS has been shown to be effective for making specific peptide targets [52], innovations with respect to resin anchoring strategies are likely needed to enable combinatorial applications.

On the other hand, N‐ to C‐direction peptide synthesis has strong potential to impact on “green” manufacturing approaches. As de la Torre and Albericio have noted, “A major reduction in atom economy will call for major efforts to be channelled into the development of a totally new concept of peptide synthesis. This concept is likely to consist of minimizing the presence of protecting groups following a similar approach to that used for native chemical ligation” [76]. In this context, activation of the carboxylate as thioacid offers significant potential for minimizing protection and mimicking the advantages of native chemical ligation [38]. Combined with methods, such as mechanochemistry, which reduce solvent use and have shown promise for C‐terminal activation of peptide fragments in couplings without racemization [77], N‐ to C‐direction peptide synthesis offers significant potential for reducing PMI. Improvements in product purification after N‐ to C‐direction peptide synthesis may similarly arise from application of soluble supports for N‐terminal anchoring [78]. The notably improved overall yield compared to the C‐ to N‐approaches in the synthesis of the active pharmaceutical ingredient (API) carfilzomib using vinyl ester activation bears well for the application of N‐ to C‐direction strategies in the future [50]. Further investment and research into enabling strategies for peptide anchoring, purification, and elongation under conditions that minimize the use of environmentally benign solvent and reagents are likely to complement advances in amido acid coupling without epimerization with minimal side chain protection to solidify the advantages of N‐ to C‐direction peptide synthesis. Examining the recent achievements and the potential of this atom economical approach, the arguments indicate strongly that N‐ to C‐peptide synthesis merits future application for sustainable production.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

Data sharing is not applicable to this article as no new data were created or analyzed in this study.