An introduction into peptide synthesis

A peptide can be defined as a short chain of amino acids (2 – 50) linked by peptide bonds. In most biological systems, a specific peptide is composed of a fixed sequence encoded by the 20 canonical trifunctional amino acids (all of which are L-enantiomers). The peptide linkages between these amino acid residues are formed by a condensation reaction between the carboxylic acid moiety of one amino acid and the functional amine on a second amino acid. These relatively small molecules (compared with proteins) show a surprisingly diverse range of biological functions. For example, peptides are known to function as active hormones, neurotransmitters, neuropeptides, growth factors, signalling molecules, and antibiotics (Made et al., 2014).

Because of their diverse functions, peptides are of great pharmaceutical interest (Made et al., 2014). Numerous peptides have been developed as commercial or pharmaceutical products (Pontiroli, 1998). Notable examples include aspartame, and the clinically relevant hormones oxytocin (Vigneaud et al., 1953), adrenocorticotropic hormone, and calcitonin (Stawikowski & Fields, 2009; Mutenthaler et al, 2020). In addition, peptides can also be used as smart delivery systems for toxophores/ drug molecules (Erak et al, 2018) In the last 10 – 15 years, peptide therapeutics have matured as a market in the pharmaceutical industry (Erak et al, 2018, reaching the multi-billion dollar level in 2008 (Saladin et al., 2009).

The chemical synthesis reactions that are used to combine two amino acid residues to form a peptide are referred to as ‘coupling methods’ (El-Faham & Albercio, 2011). For synthesis in the C to N direction (the most common synthesis protocol used), coupling involves an attack by the amino group of the residue at the N-terminal end of the chain (or the first amino acid) on the carbonyl atom of the carboxy-containing component of the incoming amino acid (El-Faham & Albercio, 2011). To increase the efficiency of the coupling reaction, the carboxylic acid is activated by the introduction of an electron-withdrawing group (El-Faham & Albercio, 2011). In addition, to ensure the specificity of the coupling reaction, the amino group of the incoming amino acid is chemically protected (N^α-protection) during the reaction. To enable peptide elongation, each coupling reaction is followed by removal of the reversible protecting group (N^α-deprotection). Overall, the peptide synthesis procedure can be simplified as repeated cycles of: N^α-deprotection (of the N-terminal residue), activation of the carboxy group to increase electrophilicity (incoming residue), and coupling. After the final residue has been added, the peptide must be purified and analysed. It should be noted that side chain protection is typically required throughout the synthesis process to minimize side reactions.

The history of peptide synthesis has been well chronicled elsewhere and will be only briefly mentioned here. However, peptide synthesis would not be possible without the work of several important pioneers. In 1901, Fischer and Fourneau reported the stepwise assembly of the dipeptide glycylglycine from its amino acid precursors (Fischer and Fourneau, 1901). To improve on this method, Bergmann and Zervas created the first reversible N^α-protecting group for peptide synthesis (Bergmann and Zervas, 1932). In 1953, Du Vigneaud successfully applied these “classical” strategies in the construction of an octapeptide with oxytocin-like activity (Vigneaud et al., 1953). As described in this review, peptide synthesis has now evolved into a rapid, efficient, and reliable methodology for the chemical synthesis of simple (and many complex) peptides. While peptide synthesis reactions were originally performed in homogenous solution (liquid phase peptide synthesis; LPPS), solid-phase peptide synthesis (SPPS) has become the predominant method. The wide-scale adoption of SPPS has enabled opportunities for using synthetic peptides in medicinal applications (Erak et al, 2018). SPPS can also be used to introduce tools into the peptide sequence for analytical investigations (spin labels, radioactive labels, bioactive molecules, fluorescent dyes) (Made et al., 2014).

Liquid phase peptide synthesis (LPPS)

A standard LPPS method involves repeat steps of N^α-deprotection, activation, and coupling. Since its introduction in 1932 (Bergmann and Zervas, 1932), the benzyloxycarbonyl (Cbz or Z) N^α-protecting group has seen widespread use (Isidro-Llobet et al, 2009). Cbz is easy to prepare, is highly stable, and it suppresses racemization (see later). However, complete N^α-deprotection of the Cbz group can only be accomplished by extended incubations (Li et al, 2020). Thus, the free amine is obtained only after prolonged hydrogenolysis (which reveals a terminal carbamic acid that readily decarboxylates). In the late 1950s, Carpino reported the development of an acid-labile urethane, tert-butyloxycarbonyl (Boc), as a new reversible N^α-protecting group for peptide synthesis. In a typical reaction, Boc is used in conjunction with benzyl (Bzl) or benzyl-based side chain protecting groups. TFA (a moderate acid) is used for N^α-deprotection. While the Boc group is labile to TFA, the side-chain protecting Bzl groups are stable in TFA. The side-chain protecting benzyl (Bzl)-based groups are then removed using hydrofluoric acid HF (a strong acid).

However, the use of differential acid lability for the Boc group and the amino acid sidechain protecting groups is a significant problem. The use of strong acid (liquid HF) during side chain protecting group removal is undesirable. In addition, repeated TFA treatment during deprotection steps can lead to premature side chain deprotection and unwanted side products. To overcome this problem, it was recognized that true orthogonality between the N^α-protecting groups and side chain protecting groups was required. In 1970, Carpino described the preparation and chemical properties of a new base-labile urethane amino protecting group, the 9-Fluorenylmethoxycarbonyl (Fmoc) group (Carpino and Han, 1970). While ammonia was initially used for Fmoc cleavage during N^α-deprotection, it was subsequently demonstrated that Fmoc cleavage could be accomplished using primary and secondary amines. Because of the milder conditions used during repeated cycles, there was a lower risk of side-reactions.

Protection of the C-terminal carboxylic acid of the first amino acid may also be required for successful LPPS (Isidro-Llobet et al, 2009; Li et al, 2020). This can be accomplished using several different methods (Isidro-Llobet et al, 2009): reaction of the free amino free acid with an alcohol under acidic conditions (e.g., HCl, pTos-OH); reaction of free amino free acid or N^α-protected amino acid with isobutene (for tert-butyl protection) under acidic conditions (e.g., pTos-OH, H~2~SO4); reaction of N^α-protected amino acid in the presence of base with halide; or reaction of N^α-protected amino acid with DCC in the presence of DMAP and an alcohol derivative of the protecting group (Isidro-Llobet et al, 2009).

In all standard peptide synthesis protocols, an addition cycle involves steps of N^α-deprotection, activation, and coupling. In LPPS, isolation of the intermediate fragments from unwanted by-products is mandatory after every coupling step. These intermediate fragments must be purified through extraction and chromatography. This process is both time consuming and technically demanding (Yeo et al, 2021). In particular, large, complex protected fragments, which tend to be both rigid and insoluble, are difficult to isolate. Nevertheless, LPPS reactions have the potential for higher crude purity, lower reagent consumption, and greater ease of scaling (Yeo et al, 2021). Moreover, the additional isolation steps do provide excellent quality control. While several attempts have been made to address the crucial problem of repeated fragment isolation (Okada et al, 2019; Yeo et al, 2021), LPPS is typically only used for large-scale manufacturing and for specialized laboratory applications.

Solid phase peptide synthesis (SPPS)

Development of SPPS

In the 1960s, Merrifield introduced and developed a new protocol for peptide synthesis on a solid support (Merrifield, 1963; Merrifield, 1964; Merrifield & Stewart, 1965; Merrifield, 1966; Merrifield, Stewart, Jernberg, 1966; Gutte & Merrifield, 1969). Importantly, this process (SPPS) circumvented the need for isolation of the peptide intermediates. In SPPS, the first N^α-protected amino acid is covalently attached (via a linker) to an insoluble support. This solid support is composed of small, polymeric resin beads functionalized with a specific reactive group (e.g., amine or hydroxyl groups) through which the nascent peptide is linked. As with LPPS, the peptide chain is assembled stepwise from C-terminus to N-terminus using N^α-protected amino acids. The reactive side chain moieties of trifunctional amino acids again need to be blocked (Li et al, 2020). However, because the carboxyl group of the C-terminal amino acid is anchored to a solid support, the linker acts as a protecting group (Isidro-Llobet et al, 2009; Li et al, 2020). Thus, protection of the C-terminal carboxyl group differs between SPPS and LPPS (Isidro-Llobet, et al 2009). While the N^α-protecting groups are temporary (removed after each addition step), the side chain protecting groups and resin ensure permanent (though reversible) protection against unwanted reactions.

The peptide is elongated from the support bound residue using conventional addition cycles (N^α-deprotection – activation – coupling). The deprotection and coupling reactions are driven to completion using an excess of soluble reagents. After every deprotection step and coupling step, the excess reagent is removed by washing with organic solvent (e.g., DCM, DMF). Consequently, there are no manipulative losses during SPPS. Upon peptide completion, the crude peptide is released from the solid support (cleavage). In the original study, SPPS of a tetrapeptide was accomplished on a polymeric (polystyrene-based) solid support using Cbz N^α-protected amino acids. Here, N,N‘-dicyclohexylcarbodiimide (DCC) was used for coupling, and HBr was used during cleavage of the peptide from the support. In a subsequent study, Merrifield reported the synthesis of the nonapeptide bradykinin using Boc N^α-protected amino acids (Merrifield, 1964). Later, Hydrogen fluoride (HF) was introduced as a useful reagent for the removal of the peptide from the resin (Sakakibara et al., 1967). Merrifield and others fine-tuned Boc-based SPPS protocols over the next 20 years (Merrifield, 1986).

Merrifield designed the original SPPS process to be amenable to automation (Merrifield & Stewart, 1965), and the first automated machine for SPPS was soon developed (Merrifield & Stewart, 1965; Merrifield, 1966; Merrifield, Stewart, Jernberg, 1966). The automated peptide synthesizer designed by Merrifield and Stewart was eventually able to add six residues per day. This was embraced as a leap forward in peptide synthesis, since traditional LPPS required months (and sometimes years) to complete synthesis of similar sized peptides. Eventually, this SPPS process became known as Merrifield Synthesis. In a landmark project, Merrifield and co-workers reported the synthesis of ribonuclease A (Gutte & Merrifield, 1969; Gutte & Merrifield, 1971). Synthesis of this 124 amino acid protein required 369 chemical reactions and 11, 931 steps by the automatic synthesizer. After subsequent purification, the protein product was demonstrated to possess native-like function.

Advantages of SPPS

The advantages of modern SPPS methodologies are manifold (Made et al, 2014). The most important advantages are:

• All reactions can be performed in a single vessel;
• The unreacted reagents and by-products are easily removed by washing;
• High coupling yields can be obtained using an excess of reactants, facilitating synthesis of difficult sequences and the incorporation of specific modifications;
• The reaction cycles are very short compared with LPPS;
• The simplified peptide synthesis process facilitates automation.

However, successful SPPS is ultimately dependent upon several inter-related factors (Fields, 1997). These include choice of solid support, choice of linker, choice of N^α-protected amino acids, choice of coupling methodology, and choice of cleavage protocol.

Outline of common N^α-protecting schemes used in SPPS

The two principle orthogonal protecting group schemes used for SPPS were first developed for use in LPPS. In the Boc/ Bzl protecting group scheme, the N^α-protection group (Boc) is TFA-labile (a moderate acid), while the side-chain protecting groups (Bzl) and linker are stable in TFA. The side-chain protecting groups (Bzl) and linker are simultaneously removed using anhydrous HF (a strong acid) during the final cleavage step. In the Fmoc/ *^t^Bu protecting group scheme, the N^α-protection group (Fmoc) is base-labile, while the side-chain protecting groups (^t^Bu) and linker are stable in base. The side-chain protecting groups (^t^*Bu) and linker are simultaneously removed using TFA (a moderate acid) during the final cleavage step. Both strategies are frequently used for SPPS and both can be adapted for automated synthesis. However, each method has specific advantages and disadvantages.

The limitations of the Boc/ Bzl protecting group scheme have been outlined extensively elsewhere (see Behrendt et al, 2015). First and foremost, the Boc/ Bzl protecting group scheme is based on ‘relative acidolysis’. Hence, the reaction conditions used for N^α-deprotection and those used for sidechain deprotection and cleavage are not completely differentiated (Behrendt et al, 2015). In practice, small amounts of the sidechain protecting groups may become deprotected at each cycle during repeated TFA acidolysis, and the nascent peptide may be prematurely cleaved from the polymer support (Behrendt et al, 2015). Furthermore, Bzl removal always leads to Boc N^α-deprotection (no orthogonality). Therefore, classical Boc N^α-protecting schemes are generally used only for specialist SPPS applications (Behrendt et al, 2015). For example, Boc N^α-protecting schemes are more suitable for the synthesis of difficult sequences (largely because aggregation of the nascent peptide can be avoided through the use of repetitive TFA treatments).

Two research groups independently reported the first use of the Fmoc N^α-protecting group in SPPS (Atherton et al, 1978; Chang & Meienhofer, 1978; Behrendt et al, 2015). Here, the Fmoc N^α-protecting group (which is stable in acid) was used in conjunction with mildly acid-labile *^t^*Bu sidechain protecting groups and a p-alkoxybenzyl ester resin linkage. Fmoc N^α-deprotection was achieved under relatively mild basic conditions. Therefore, this scheme demonstrated true orthogonality. In modern protocols, Fmoc N^α-deprotection is usually accomplished with piperidine. Fmoc N^α-deprotection also liberates a fluorenyl group. This fluorophore demonstrates strong absorbance in the UV region (λ~max~ 266 nm in DMF), and this has proven useful for monitoring the N^α-deprotection and coupling reactions spectrophotometrically. This characteristic has been widely employed in automatic synthesizers.

For several reasons, Fmoc N^α-protection schemes have replaced Boc N^α-protection schemes as the method of choice for SPPS (Behrendt et al, 2015). The main advantage of the Fmoc N^α-protection scheme is that the chemistry is milder. Hence, more complex SPPS strategies can be attempted. The Fmoc N^α-protection group is also completely stable to treatment with TFA (Li et al, 2020). The Fmoc/ *^t^*Bu protection scheme is also more compatible with the synthesis of peptides that are susceptible to acid-catalysed side reactions. The indole ring of Trp residues is particularly susceptible to modification during Boc-based SPPS (Barany and Merrifield, 1979), and synthesis using the Fmoc N^α-protection scheme can significantly increase yields of Trp-containing peptides (Fields et al., 1989; Fields et al., 1990; King et al., 1990). As already discussed, Fmoc/ *^t^*Bu also offers true orthogonality (Made et al, 2014; Li et al, 2020). Thus, selective removal of protecting groups is possible using completely different chemical reactions and cleavage mechanisms. A further advantage is that Fmoc/ *^t^*Bu does not require the use of special vessels (for toxic HF).

A shift in the chemistry employed by peptide laboratories was noted in a series of studies carried out during the period 1991 – 1996 by the Peptide Synthesis Research Committee (PSRC) of the Association of Biomolecular Resource Facilities (ABRF). A report on these studies revealed that Boc SPPS was largely replaced by Fmoc SPPS during this period (Angeletti et al., 1997; (Stawikowski and Fields, 2009). This change was accompanied by an improvement in overall peptide quality. This improvement was attributed to the increased use of Fmoc chemistry, the use of improved side-chain protecting group strategies that reduce side reactions during N^α-deprotection, and the use of N^α-deprotection protocols designed to minimize side reactions (Stawikowski and Fields, 2009). In the last 25 years, very high quality Fmoc N^α-amino acids have become commercially available at low cost (Behrendt et al, 2015), facilitating the synthesis of longer peptides. In addition, many modified derivatives of Fmoc N^α-amino acids are now commercially available (Behrendt et al, 2015). Hence, simple peptide syntheses are usually accomplished using Fmoc SPPS (Behrendt et al, 2015).

The development of the base-labile Fmoc N^α-protection group and its integration into peptide synthesis protocols is considered a major landmark (Li et al, 2020). This success is demonstrated in the successful syntheses of bovine pancreatic trypsin inhibitor analogues, ubiquitin, yeast actin-binding protein 539–588, human β-chorionic gonadotropin 1–74, mini-collagens, HIV-1 Tat protein, HIV-1 nucleocapsid protein NCp7, and active HIV-1 protease. Nonetheless, Fmoc chemistry demonstrates several disadvantages. These include suboptimum solvation of Fmoc N^α-amino acids and the Fmoc N^α-protected peptide, slow coupling kinetics, and base-catalysed side reactions (Stawikowski and Fields, 2009; Li et al, 2020). The decrease in solubility (increase in hydrophobicity) of the Fmoc N^α-protected peptide may account for the propensity for peptide aggregation during Fmoc SPSS of ‘difficult peptides’ (Li et al, 2020). Improvements in these areas of Fmoc chemistry have been reported (Atherton and Sheppard, 1987; Fields and Noble, 1990; Fields et al., 2001).

Essential steps in peptide synthesis

The essential steps in any successful peptide synthesis are listed below:

1. Overall synthesis strategy (type of N^α-protecting group);
2. Side-chain protection of N^α-protected amino acids;
3. Backbone protection of N^α-protected amino acids;
4. Choice of solid material;
5. Linker (and choice of C-terminal modification);
6. Choice of coupling reagents (for efficient amide bond formation).

These issues are discussed separately in detail below.

Overall synthesis strategy (Type of N^α-Protecting group)

N^α-Protecting groups

The overall synthesis strategy is largely determined by the type of N^α-protecting group chosen. The principal function of N^α-protection is to prevent polymerization of activated amino acids. However, the chemical and physicochemical characteristics of the N^α-protecting group have a crucial impact on peptide synthesis (Isidro-Llobet et al, 2009). The N^α-protected amino acids/ N^α-protected nascent peptide should be soluble in all solvents used during peptide synthesis (Isidro-Llobet et al, 2009). Ideally, the N^α-protecting group should also prevent (or minimize) racemization during coupling (Isidro-Llobet et al, 2009). During N^α-deprotection, the reaction should be fast, efficient, and free of side reactions (Isidro-Llobet et al, 2009). Moreover, the by-products of N^α-deprotection should be easily eliminated (Isidro-Llobet et al, 2009). Ideally, N^α-deprotection conditions and side-chain/ linker deprotection conditions should also be orthogonal. Since N^α-deprotection occurs once during every cycle, N^α-deprotection should preferably be performed under mild conditions (Isidro-Llobet et al, 2009).

For SPPS, the Fmoc and Boc N^α-protecting groups are both extensively used (Isidro-Llobet et al, 2009). In general, these are used in conventional Fmoc/ *^t^*Bu and Boc/ Bzl strategies, respectively. However, many other N^α-protecting groups (and corresponding protection strategies) have been developed (Isidro-Llobet et al, 2009; Li et al, 2020). In particular, several useful derivatives of the Fmoc N^α-protection group have been reported (Li et al, 2020). Examples include Climoc (which is more base-labile), Sulmoc (which is stable to HF and pyridine but removed by anhydrous morpholine or piperidine), and Fmoc* (the N^α-protected amino acids of which show enhanced solubility) (Isidro-Llobet et al, 2009; Li et al, 2020). Unrelated base-labile N^α-protecting groups include Nsc (Li et al, 2020). For LPPS, the Cbz, Nps (2-nitrophenylsulenyl), and Bpoc (2-(4-biphenyl)isopropoxycarbonyl) N^α-protecting groups are widely used (Isidro-Llobet et al, 2009). In general, strategies involving these N^α-protecting groups use *^t^*Bu-type side chain protection (Isidro-Llobet et al, 2009). However, the Boc/ Bzl strategy can also be used.

N^α–protection

To accomplish N^α-protection, the free amino acid is separately reacted with the acyl halide or carbonate of the protecting group under Schotten-Baumann conditions (Isidro-Llobet et al, 2009). Alternatively, each free amino acid can be separately reacted with the corresponding halide in organic solvents (Isidro-Llobet et al, 2009). Since a free carboxylic acid group may interfere with the N^α-protection reaction, a carboxylic acid-protecting group may also be required (Isidro-Llobet et al, 2009; Behrendt et al, 2015). For example, silylation can protect the carboxylic acid group and help prevent amino acid oligomerization (Behrendt et al, 2015). In the case of Lys, the free amino acid should first be side chain protected.

The Cbz, Fmoc, and Boc N^α-protecting groups can all suppress racemization during activation and coupling (Behrendt et al, 2015). However, the many advantages of Fmoc N^α-protection reactions have yet to be surpassed by other N^α-protection reactions (Li et al, 2020). These advantages include both rapidity and efficiency of the Fmoc N^α-protection reaction. In addition, the Fmoc N^α-protected amino acids are easy to prepare, and the preparation yield is typically high (Li et al, 2020). Fmoc is commonly introduced to an amino acid using Fmoc-succinimidyl carbonate (Li et al, 2020). During Fmoc N^α-protection, a number of side reactions may occur at the N^α-amine of the amino acid. Most notable is Lossen-type rearrangement. For example, Fmoc N^α-protection of Ala can result in the formation of Fmoc-β-Ala-OH, potentially leading to its incorporation as Fmoc-β-Ala-Xaa-OH (Behrendt et al, 2015). The Fmoc N^α-deprotection conditions should be optimized to minimize these reactions.

Inexpensive N^α-protected amino acids are now commercially available (Behrendt et al, 2015). Moreover, the commercially-available N^α-protected amino acid derivatives are industrially regulated (Behrendt et al, 2015). Several known impurities in these products can inhibit peptide synthesis. For example, acetic acid impurities can cause permanent capping (Behrendt et al, 2015). The International Conference on Harmonisation for standards of active pharmaceutical ingredient production (Q11) requires important properties of N^α-protected amino acid derivatives to be specified. These include optical purity, acetic acid content (< 0.02%), and free amine content. Most Fmoc-protected amino acid derivatives are now available as high grade (> 99% pure by RP-HPLC) chemicals (Behrendt et al, 2015). The enantiomeric purity of N^α-protected amino acid derivatives can be quantified by GC-MS (Behrendt et al, 2015).

N^α-deprotection

While stable in acid, the Fmoc group of Fmoc N^α-protected amino acids is base-labile. The Fmoc N^α-protection group is rapidly removed by primary amines (e.g., ethanolamine) and secondary amines (e.g., piperidine, piperazine) (Isidro-Llobet et al, 2009; Li et al, 2020). In practice, liquid NH~3~ (v. slow; 10 h), morpholine, piperidine (very fast; min), and polymeric secondary amines (piperazine, piperidine) have all been used for Fmoc N^α-deprotection (Isidro-Llobet et al, 2009; Li et al, 2020). For comparison, the half-life for Fmoc-Val is approximately 1 min for morpholine (in DMF), approximately 30 s for Piperazine (in DMF), and approximately 6 s for piperidine (in DMF). The Fmoc N^α-deprotection reaction is also much faster in polar solvents (e.g., DMF, NMP) than in apolar solvents (e.g., DCM) (Isidro-Llobet et al, 2009; Li et al, 2020). The stability of Fmoc N^α-protected amino acids in both DMF and NMP has also been confirmed (Li et al, 2020). It should be noted that the free amino group of the resin bound amino acid does not have a significant effect on the Fmoc group of the incoming amino acid (Li et al, 2020).

The Fmoc N^α-deprotection reaction involves abstraction of the acidic proton at the 9-position of the fluorene ring and subsequent β-elimination (Li et al, 2020). The initial product of this reaction is a reactive dibenzofulvene (DBF) intermediate. This DBF by-product can react (and form a stable adduct) with added nucleophiles (including piperidine) and with the free amine group (Behrendt et al, 2015; Li et al, 2020). However, secondary amines are more efficient at capturing this DBF by-product (Isidro-Llobet et al, 2009). Indeed, piperazine, piperidine, and morpholine are all observed to undergo rapid adduct formation (Li et al, 2020). In SPPS, removal of DBF is helped by in-column washing procedures (Behrendt et al, 2015). After Fmoc N^α-deprotection, the exposed amine has a neutral charge. In prone sequences, the addition of HOBt to the piperidine solution (0.1 M) can reduce aspartimide formation (Isidro-Llobet et al, 2009).

The most utilized Fmoc N^α-deprotection condition is 20% piperidine in DMF (Li et al, 2020). However, since piperidine is a controlled substance, the use of replacement secondary amines is becoming commonplace (Li et al, 2020). In particular, 4-methylpiperidine is widely used as a direct substitute (Li et al, 2020). However, while the performance of 4-methylpiperidine matches that of piperidine, the concentration of 4-methylpiperidine in DMF is usually reduced to 2.5% (because of toxicity). Piperazine is another widely used substitute (Li et al, 2020). However, while piperazine is more cost-effective, Fmoc N^α-deprotection is noticeably slower (Li et al, 2020). Li et al (2020) list several alternative Fmoc N^α-deprotection conditions (Li et al, 2020). Of these, DBU (2% in DMF) is reported to rapidly and efficiently remove Fmoc N^α-protecting groups. Furthermore, base-induced racemization is substantially reduced using DBU (Li et al, 2020). For SPPS, the optimal Fmoc N^α-deprotection conditions are 20% piperidine in DMF, 1 – 5% DBU in DMF, or 2% HOBt. For LPPS, the Fmoc N^α-protecting group can be removed with triethylamine (a cheap and readily available base) and an imidazolium-based ionic liquid (Li et al, 2020).

Although it is stable to base and nucleophile attack, the Boc group of Boc N^α-protected amino acids is acid-labile. The Boc N^α-protection group is rapidly removed by TFA (Isidro-Llobet et al, 2009). In the presence of TFA, the Boc group of Boc N^α-protected amino acids forms a positively charged amino group, which needs to be neutralized before coupling. This neutralization procedure can occur either prior to coupling or in situ during the basic coupling reaction. While the product of Fmoc N^α-deprotection (DBF) is reactive and can reattach (Behrendt et al, 2015), the product of Boc N^α-deprotection (butylene) is volatile. The optimal Boc N^α-deprotection condition is 25 – 50% TFA in DCM (Isidro-Llobet et al 2009). However, 4 M HCl in dioxane can also be used (Isidro-Llobet et al 2009).

Protection/ Deprotection of miscellaneous N^α-Protecting groups

Several other N^α-protection groups are commonly used, including Trityl (Trt), Cbz, Nsc, Fmoc*, Bpoc, and Nps (Isidro-Llobet et al, 2009). Trt N^α-protected amino acids are effectively protected from racemization. The bulkiness of the Trt group protects the α-proton from base abstraction. However, the coupling yields of Trt N^α-protected amino acids are lower than those of other N^α-protected amino acids (Isidro-Llobet et al, 2009). The optimal Trt N^α-deprotection conditions are 1% TFA in DCM or 0.1 M HOBt in TFE (Isidro-Llobet et al, 2009). The Cbz N^α-protected amino acids are commonly used in LPPS (and occasionally in SPPS) (Isidro-Llobet et al, 2009). Cbz N^α-protected amino acids are easy to prepare, demonstrate high stability, and racemization is supressed (Isidro-Llobet et al, 2009). Cbz N^α-deprotection is achieved with HF. The Nsc N^α-protection group is the most promising alternative to Fmoc (Isidro-Llobet et al, 2009). Nsc N^α-protected amino acids are more soluble and more stable to base (Isidro-Llobet et al, 2009). Nsc N^α-protected Pro are especially useful in the synthesis of polyproline peptides, where premature Fmoc N^α-deprotection can cause deletions (Isidro-Llobet et al, 2009). Because N^α-protected amino acids solutions are frequently stored for a long time in automated synthesizers, Nsc N^α-protected amino acids are also useful for automated SPPS.

Side Chain Protection Of N^α-Protected Amino Acids

Peptides can contain up to eight distinct functional groups (not including indole and imidazole rings) (Isidro-Llobet et al., 2009). To block the formation of unwanted bonds (and protect against non-specific side reactions), side chains containing these functional groups are blocked using side chain protection (Isidro-Llobet et al., 2009; Made et al., 2014). Side chain protecting groups should be easy to incorporate, stable, and simple to remove (Isidro-Llobet et al., 2009; Made et al., 2014). Ideally, the various side chain protecting groups should also demonstrate orthogonality with the other categories of protecting groups (Nα-protecting groups, backbone protecting groups) and with the linker (Isidro-Llobet et al., 2009). At a minimum, the side chain protecting groups and Nα-protecting group should belong to independent group classes that can be removed by distinct mechanisms (Isidro-Llobet et al., 2009). In some applications, side chain protecting groups should also be stable to removal of the peptide from the column. Preferably, it should also be possible to remove the different categories of protecting groups in any order in the presence of the other groups (selective deprotection).

In most peptide synthesis reactions, protection of specific side chain functional groups is mandatory (Isidro-Llobet et al., 2009; Made et al., 2014). A diverse range of side chain protecting groups are available for protecting the main functional groups (amino, carboxyl, hydroxy, thio, pyrrolidinyl, imidazolyl, guanidinyl, amido, and indoyl) in trifunctional amino acids (Isidro-Llobet et al., 2009; Made et al., 2014). While side chain protection may be required for specific trifunctional amino acids (Cys, Asp, Glu, His, Lys, Asn, Gln, Arg, Ser, Thr, Trp, and Tyr) depending on the peptide sequence and synthesis strategy, side chain protection is not required for the other (Ala, Gly, Ile, Leu, Met, Phe, Pro, and Val) amino acids (Made et al., 2014). Made et al (2014) provide an excellent description of Fmoc/ _t_Bu and Boc/ Bn SPPS orthogonal protecting group strategies (Made et al., 2014). While some of these protecting groups can be used across strategies, others are specifically used for Fmoc/ _t_Bu strategies or for Boc/ Bn strategies (Isidro-Llobet et al., 2009).

In general, the orthogonal protecting group schemes used for Fmoc SPPS require base (e.g., piperidine) for removal of the Nα-protecting group and acid (e.g., high TFA concentrations) for removal of the side-chain protecting groups (and liberation of the peptide from the resin) (Made et al., 2014). Side chain protecting groups like Boc and _t_Bu can both be removed under (relatively mild) acidic conditions (TFA). Likewise, the anchoring linkages commonly used in Fmoc chemistry are also cleaved by TFA. In this scheme, the peptide is simultaneously deprotected (of side chain groups) and cleaved. Made et al (2014) provide a list of commonly used protecting groups for Fmoc/ _t_Bu SPPS (Made et al., 2014). Many of these protecting groups are now commercially available (Made et al., 2014).

Frequently used protecting groups for Fmoc SPPS include Trt (for Asn, Cys, Gln, and His), _t_Bu (for Asp, Glu, Ser, Thr, and Tyr), OtBu (for Asp and Glu), Boc (for Lys and Trp), and Pbf (for Arg) (Made et al., 2014; Behrendt et al., 2015). Several other protecting groups are cleaved under specific conditions. These include Dde (hydrazine), Mmt (low acid concentrations), Alloc (palladium-catalyzed cleavage), and Nvoc (UV) (Isidro-Llobet et al., 2009). While the sidechain protecting groups used for routine Fmoc/_t_Bu chemistry have remained largely unchanged since the 1990s (Behrendt et al., 2015), the choice for several (Arg, Asn, Asp, His, and Cys) can be considered suboptimal (Behrendt et al., 2015). Nonetheless, the adoption of improved protecting groups has been slow (Behrendt et al., 2015). This is largely because high purity standard side chain protecting group reagents and protected amino acids are commercially available and inexpensive (Behrendt et al., 2015).

In Boc-based strategies, the Nα-protecting group is removed by mild acidolysis in TFA, and the anchoring linkage is removed by strong acidolysis in HF. Typically, the side chain protecting groups are also removed by strong acidolysis. The benzyl or benzyl-based side chain protecting groups are ideal for Boc-based strategies. Thus, while benzyl (Bn)-based side chain protecting groups are stable in TFA, these groups are labile in HF. Bn can also be removed with TFMSA in Boc/ Bn strategies (Isidro-Llobet et al., 2009). For certain applications (e.g., preventing aspartimide formation), the cyclohexyl (cHx) side chain protecting group is superior to the Bn group (Isidro-Llobet et al., 2009). Although the cHx side chain protecting group can also be removed with HF or TFMSA (Boc/ Bn), it is more resistant to acid. The cHx side chain protecting group is also considered more suitable for the synthesis of long peptides (Isidro-Llobet et al., 2009).

Side chain and linker deprotection reactions can release reactive species that may modify the side chains of susceptible residues (e.g., Trp, Tyr, and Met). However, these non-specific side reactions can be minimized during TFA cleavage by including scavengers. To quench these reactive species and preserve amino acid integrity, several reagents have been formulated. These include: TFA-phenolthioanisole-1,2-ethanedithiol-H2O (82.5:5:5:2.5:5) (reagent K) (King et al., 1990); TFA-thioanisole-1,2-ethanedithiol-anisole (90:5:3:2) (reagent R) (Albericio et al., 1990); and TFA-phenol-H2O-triisopropylsilane (88:5:5:2) (reagent B) (Solé and Barany, 1992). In addition, Boc sidechain protection of Trp may be required to inhibit alkylation by Pmc or 2,2,4,6,7-pentamethyldihydro-benzofuran-5-sulfonyl (Pbf) groups.

Protection of Asp/ Glu

The side chain carboxylic groups of Asp and Glu amino acids should be protected to prevent activation during peptide synthesis (Isidro-Llobet et al., 2009). In the case of Asp, side chain protecting groups also minimize the formation of aspartimide (Isidro-Llobet et al 2009). Selective side chain carboxylic group protection can be achieved by acid-catalyzed esterification of the free amino acid (Isidro-Llobet et al., 2009). In this simple case, protonation of the amino group under acidic conditions makes the α-carboxylic acid group less reactive. Selective carboxylic acid protection can also be achieved through the formation of an intramolecular anhydride between electrophilic α-carboxylic acid groups (Isidro-Llobet et al., 2009).

Aspartimide formation occurs as a result of the exposure of Asp to strong base (Behrendt et al., 2015). Subsequent hydrolysis of the aspartimide ring yields a 3:1 ratio of (D/L) -peptide and α-peptide (Isidro-Llobet et al., 2009; Behrendt et al., 2015). In addition, piperidine-dependent ring opening yields a mix of (D/L) α-piperidide and (D/L) -piperidide (Isidro-Llobet et al., 2009; Behrendt et al., 2015). Amino group-dependent ring opening can also lead to the formation of dipeptides and cyclic peptides (Behrendt et al., 2015). After epimerization, restoration of the original α-aspartyl is not a viable procedure (Behrendt et al., 2015). Aspartimide formation can be a major problem during the synthesis of long peptides and peptides containing multiple Asp residues (Behrendt et al., 2015). This is largely because the by-products can co-elute with the target peptide. Aspartimide formation is highly dependent on the amino acid following the aspartyl residue (Behrendt et al., 2015). Thus, Asp-Gly, Asp-Asp, Asp-Asn, Asp-Arg, Asp-Thr, Asp-Ser, and Asp-Cys sequences are particularly prone (Lauer et al., 1995; Behrendt et al., 2015). While Asp-Gly is by far the most susceptible, the Asp sidechain in Asp-Gly, Asp-Ser, and Asp-Thr sequences is usually sidechain protected.

Aspartimide formation is considered the most serious side reaction that occurs during Fmoc chemistry (Behrendt et al., 2015). Asp residues are particularly prone to aspartimide formation because of repeated exposure to base (piperidine) during Fmoc Nα-deprotection (Isidro-Llobet et al., 2009; Behrendt et al., 2015). The _t_Bu side chain protecting group used in the Fmoc/ _t_Bu strategy is particularly useful for Asp side chain protection. This is because the _t_Bu group can minimize aspartimide formation through steric hindrance. During Fmoc/ _t_Bu SPPS, _t_Bu-deprotection is completed using 90% TFA in DCM. In Boc-based strategies, the cHx group is superior to the Bn group at preventing acid-catalyzed aspartimide formation (Isidro-Llobet et al., 2009). Bn-deprotection and cHx-deprotection are both completed with HF or TFMSA. In LPPS, the _t_Bu side chain protecting group is used for Asp side chain protection, and _t_Bu-deprotection is again completed with 90% TFA in DCM (Isidro-Llobet et al., 2009).

A reduction in aspartimide formation during Fmoc SPPS can also be achieved by the addition of certain modifiers (Behrendt et al., 2015). For example, aspartimide formation can be minimized by the addition of 0.1 M HOBt in the piperidine solution (Lauer et al., 1995). Likewise, 1 M ethyl cyano(hydroxyimino) acetate (Oxyma) Pure in 20% piperidine (in DMF) significantly reduces aspartimide related impurities (Behrendt et al., 2015). Aspartimide formation is also dependent on peptide conformation (Behrendt et al., 2015), and this can be exploited to protect against aspartimide formation (See section on backbone-protecting groups). Thus, protection against aspartimide formation can be achieved using an amide backbone protecting group (i.e., 2-hydroxy-4-methoxybenzyl) for residue X in the Asp-X sequence (Quibell et al., 1994).

Protection of Arg

The standard side chain protecting group for Arg amino acids was originally 4-methoxy-2,3,6-methylbenzenesulfonyl (Behrendt et al., 2015). However, long deprotection times were required when using this protection group (Behrendt et al., 2015). Improved deprotection times could be achieved using 2,2,5,7,8-pentamethylchromanyl-6-sulfonyl (Pmc) (Behrendt et al., 2015). Currently, the standard protection group for Arg side chains is Pbf (which is related to Pmc) (Behrendt et al., 2015). Pbf-deprotection can usually be achieved in only a few hours. However, extended incubations are still required for peptides with multiple Arg residues. The subsequently introduced MIS (1,2-dimethylindole-3-sulfonyl) protecting group can be removed in only 30 min (with 1:1 TFA/ DCM) Thus, Fmoc-Arg(MIS)-OH side chain protected amino acids with improved deprotection kinetics (compared with Fmoc-Arg(Pbf)-OH) are now an option (Behrendt et al., 2015).

Protection of Cys

Cys racemization is a problem when using base-mediated activation methods (e.g., methods using phosphonium or aminium reagents). This problem is exacerbated when microwave heating is used during peptide synthesis and when using pre-activated amino acids (Behrendt et al., 2015). Several strategies for reducing racemization of cysteine residues (including careful choice of side chain protecting group) are available. Cys racemization can be reduced by avoiding pre-activation, using a weaker base, switching the solvent mixture to DMF-dichloromethane (DCM) (1:1), and by adding HOBt, 6-Cl-HOBt, or HOAt. Replacement of the Fmoc-Cys derivative by its pentafluorophenyl ester will also reduce Cys racemization. An alternative strategy involves replacement of the standard trityl side chain protecting group (Behrendt et al., 2015). For example, racemization is suppressed following replacement of trityl with the MBom (4-methoxybenzyl-oxymethyl) side chain protecting group. It should be noted that Cys racemization can be avoided altogether by using carbodiimide activation. Nonetheless, epimerization and -elimination of C-terminal cysteine residues anchored to a resin via a benzyl ester is likely to remain a significant problem (Behrendt et al., 2015).

Protection of His

His amino acids are especially prone to racemization (Behrendt et al., 2015). Conventionally, His amino acids are introduced using Fmoc-His(1-Trt)-OH (Behrendt et al., 2015). However, Fmoc-His(1-Trt)-OH derivatives can undergo significant racemization during standard coupling reactions. His racemization is especially high when the coupling is base-mediated, and when the coupling reaction is slow (Behrendt et al., 2015). During His racemization, the imidazole -nitrogen promotes the enolization of histidine active esters (Behrendt et al., 2015). Thus, the most effective approach to reducing His racemization is to employ imidazole -nitrogen protection (Behrendt et al., 2015). For Boc chemistry, N-benzyloxymethyl (Bom) protection is generally used (Behrendt et al., 2015). For Fmoc chemistry, TFA-labile protecting groups such as t-butoxymethyl (Bum) are typically used (Behrendt et al., 2015). Hence, His amino acids are introduced using Fmoc-His(Bum)-OH to suppress racemization (Behrendt et al., 2015). His amino acids can also be introduced using Fmoc-His(MBom)-OH, which requires only a five minute pre-activation (Behrendt et al., 2015). However, Fmoc-His(MBom)-OH is expensive to prepare (Behrendt et al., 2015). Adding HOBt, 6-Cl-HOBt, or HOAt can also suppress His racemization. In general, His stereochemistry can be maintained using acidic coupling reactions (e.g., DIC/ HOBt) at ambient temperature (Behrendt et al., 2015). Nonetheless, His racemization remains a risk with the current range of derivatives used (Behrendt et al., 2015).

Backbone Protection of N^α-Protected Amino Acids

The NH backbone is typically unprotected during simple peptide synthesis (Isidro-Llobet et al., 2009). However, NH backbone protection is essential under specific circumstances (Isidro-Llobet et al., 2009). Firstly, when synthesizing peptide chains which are prone to intra-molecular and inter-molecular interactions. Secondly, when synthesizing sequences prone to aspartimide formation. As noted earlier, aspartimide formation is particularly severe in Fmoc-based strategies (and when synthesizing an Asp-Gly sequence) (Isidro-Llobet et al., 2009; Li et al., 2020). Several backbone protecting groups can provide protection against both of these events. However, choice of backbone protection is also dictated by cost (Behrendt et al., 2015). As a rule, a backbone protection group should be easy to acylate and easy to remove (Behrendt et al., 2015).

Certain peptide chains are especially prone to aggregation behavior during Fmoc-based SPPS. The intra-molecular and inter-molecular interactions underlying aggregation can significantly decrease coupling and deprotection yields (Isidro-Llobet et al., 2009). These interactions are predominantly associated with hydrophobic sequences and may be anticipated (Behrendt et al., 2015). Peptide insolubility is one of the most intractable problems encountered in modern peptide chemistry (Behrendt et al., 2015). Poor peptide solubility on the resin during SPPS is the basis of the ‘difficult sequence’ problem (Behrendt et al., 2015). In general, aggregation is a sequence-dependent, peptide length-dependent, solvent-dependent, and time-dependent process. However, aggregation is also dependent on the extent of derivatization of the solid support (i.e., aggregation is also peptide concentration-dependent). The ‘difficult sequence’ problem may be addressed through a combination of backbone protection and native chemical ligation (see later). This two-pronged complementary approach has been labelled the future of synthetic protein chemistry (Behrendt et al., 2015).

During backbone protection, a substituent is added to the amide bond, blocking its participation in intermolecular H-bonding, thus interfering with interchain aggregation (Behrendt et al., 2015). NH backbone protection also minimizes peptide chain aggregation by inhibiting the formation of secondary structure, by reducing hydrophobic interactions, and by steric hindrance (Isidro-Llobet et al., 2009; Behrendt et al., 2015; Erak et al., 2018). Overall, backbone protection is an important methodology for increasing the solubility of long/ difficult peptide sequences, and thus enhancing the efficiency of their synthesis during Fmoc SPPS (Isidro-Llobet et al., 2009; Behrendt et al., 2015; Erak et al., 2018). Nowadays, peptide backbone modification methods allow a systematic approach to the problem of peptide insolubility (Behrendt et al., 2015). For example, commercially available building blocks can be used to circumvent secondary structure formation during synthesis on solid support (Erak et al., 2018).

The most common approach to backbone protection uses 2,2- dimethyloxazolidine analogs of Ser or Thr (‘pseudoproline’ dipeptides) (Behrendt et al., 2015). Mutter hypothesized that pseudoprolines (ΨPro) kink the peptide backbone, thus breaking the secondary structure (Behrendt et al., 2015). As a result, aggregation is minimized, and peptide solubility is enhanced. No special conditions are required for pseudoproline dipeptide incorporation, and deprotection by TFA is uncomplicated (Behrendt et al., 2015). In addition, pseudoproline dipeptides are simply prepared by Reaction of Fmoc-X-S or Fmoc-X-T with 2,2-dimethoxypropane (Isidro-Llobet et al., 2009). The main disadvantage is that backbone protection using pseudoproline dipeptides is limited to sequences containing Ser or Thr at a convenient position (Behrendt et al., 2015). The introduction of pseudoprolines has facilitated the synthesis of several long peptides, e.g., fas, ubiquitin (Behrendt et al., 2015). The success of modern Fmoc-based SPPS is largely a consequence of applying pseudoprolines (and other back bone protection) to the ‘Difficult sequence problem’ (Behrendt et al., 2015).

Another approach to overcome aggregation via non-productive peptide interactions is the O-acyl isopeptide method (Behrendt et al., 2015). In this backbone protection method, the peptide is first synthesized as a ‘depsipeptide’ derived from Ser and Thr residue (Behrendt et al., 2015). The depsipeptide analogues of aggregation-prone peptides are more soluble and more easily purified (Behrendt et al., 2015). The depsipeptide is subsequently converted to the native form by adjusting the pH to mildly basic conditions. At pH 7.4, O-acyl to N-acyl migration occurs spontaneously (Behrendt et al., 2015).

Reversible backbone protecting groups such as 2,4-dimethoxybenzyl (Dmb) can also be used to temporarily increase peptide solubility (Weygand, 1966; Behrendt et al., 2015). Counterintuitively, adding Dmb (or other hydrophobic groups) to the backbone can lead to an increase in solubility (Behrendt et al., 2015). Dmb (and similar) backbone modifications have found widespread use against difficult peptide sequences and/ or aspartimide formation (Behrendt et al., 2015). The optimal residue separation for backbone protection using Dmb is every six residues (Behrendt et al., 2015). Dmb-deprotection is accomplished using high concentrations of TFA (Isidro-Llobet et al., 2009). Because of Dmb group bulkiness, Dmb-protection is generally limited to Gly. Commercial Dmb-Gly dipeptide building blocks are extensively employed (Behrendt et al., 2015). However, other Dmb-protected dipeptides (e.g., Fmoc-AA2-(Dmb)AA1-OH) are available (Isidro-Llobet et al., 2009). To avoid the problem of steric hindrance encountered with Dmb, other backbone protection groups (e.g., Hmb) have been developed (Johnson et al., 1993; Behrendt et al., 2015).

Johnson et al (1993) introduced the Hmb (2-hydroxy-4-methoxybenzyl) group to use in the synthesis of ‘difficult peptides’ (Behrendt et al., 2015). Because Hmb protected amino acids are easier to couple, Hmb backbone protection is not restricted to Gly amino acids (Isidro-Llobet et al., 2009). Hmb backbone protection is currently widely used (Behrendt et al., 2015). On-resin reduction of Hmb onto the peptide resin simplifies the introduction of backbone protection, thus enabling automation (Johnson et al., 1993; Behrendt et al., 2015). Hmb groups are mainly introduced as Fmoc-(FmocHmb)AA1-OH, Fmoc-AA2-(Hmb)AA1-OH, or Fmoc-(Hmb)AA1-OH amino acid derivatives (Isidro-Llobet et al., 2009). Because acylation of Hmb protected amines differs between amino acids, the coupling reaction can require a lengthy incubation in DCM (Behrendt et al., 2015). Hmb deprotection is generally accomplished using TFA (Isidro-Llobet et al., 2009).

The coupling reaction incubation times can be minimized with the Hmsb (2-hydroxy-4-methoxy-5-methylsulfinyl benzyl) backbone protection group (Behrendt et al., 2015). As an alternative, the photolabile Hnb (2-hydroxy-6-nitrobenzyl) group can be employed (Behrendt et al., 2015). The MIM (1-methyl-3-indolylmethyl) and EDOTn (3,4-ethylenedioxy-2-thenyl) backbone protection groups have also been developed for use in Fmoc/ _t_Bu strategies (Isidro-Llobet et al., 2008). These are both more acid labile than Dmb. MIM and EDOTn can both be removed with TFA-DCM-H2O (95/ 2.5/ 2.5) in 1 h. Because EDOTn is less sterically hindered, coupling reactions involving these derivatives are faster (Isidro-Llobet et al., 2009). Hmb, Dmb, EDOTn, and MIM can be used with all amino acids. However, in practice, these backbone protection groups are mainly used with Gly amino acids (Isidro-Llobet et al., 2009). These backbone protection groups are introduced by reductive amination of the aldehyde of the protecting group by the amine of the corresponding amino acid (Isidro-Llobet et al., 2009). This can be followed by either α-amino acid protection or dipeptide formation. The main limitation of these benzyl type protection groups is that benzyl cations formed during deprotection can react with the Trp side chains (Behrendt et al., 2015).

The pseudoproline, Hmb, Dmb, EDOTn, and MIM backbone protection methods are all widely employed in Fmoc/ _t_Bu strategies (Isidro-Llobet et al., 2009). These methods all induce a different geometry in the backbone and are efficient tools for disrupting secondary structure (Erak et al., 2018). Thus, many of the applications of pseudoprolines are the same as those of Dmb-derived backbone protection. These include cyclization, aspartimide suppression, and epimerization-free segment coupling (Behrendt et al., 2015). In addition, these backbone protection methods are all TFA-labile (Erak et al., 2018). Hence, although these methods are useful for improving SPPS of ‘difficult sequence’ peptides, low peptide solubility after cleavage from the resin and during purification remains a problem (Erak et al., 2018). Hmb backbone protection is resistant to TFA treatment when the 2-OH is acetylated (Behrendt et al., 2015). This modification can be exploited for the purification of ‘difficult peptides’ (Behrendt et al., 2015). Thus, Hmb backbone protection can be maintained after TFA treatment. Hmsb and Hnb can also be retained on sidechain deprotected peptides by this modification (Behrendt et al., 2015).

In Fmoc/ _t_Bu strategies, aspartimide formation during the synthesis of sensitive peptide sequences can be significant (Isidro-Llobet et al., 2009; Li et al., 2020). During Fmoc Nα-deprotection, piperidine can induce aspartimide formation in Asp(OtBu)-X peptides when X = Arg, Asn, Asp, Cys, Gly, Ser, and Thr (Lauer et al., 1995). Aspartimide formation can also be conformation-dependent (Lauer et al., 1995). The known dependence of the reaction rate of aspartimide formation on peptide backbone conformation can be exploited to minimize aspartimide formation (Lauer et al., 1995; Behrendt et al., 2015). Thus, because backbone protection induces a different geometry in the backbone, backbone protection methods can minimize aspartimide formation (Behrendt et al., 2015; Erak et al., 2018). Again, the most popular approach involves the use of pseudoproline dipeptides (Behrendt et al., 2015). Due to steric hindrance in the protected amino acids, these protecting groups are generally incorporated as dipeptides (Isidro-Llobet et al., 2009). It should be noted that pseudoproline use is limited to Ser, Thr, and Cys amino acids (Isidro-Llobet et al., 2009).

Choice of Solid Material

Choice of solid-support resin (and matching linker) is vital for successful SPPS (Made et al., 2014). Solid-support resins are small, polymeric resin beads upon which peptide synthesis is performed. These solid-phase resins are functionalized with reactive groups (e.g., amine or hydroxyl groups) that can be linked to the nascent peptide chain via a reversible chemical linker (Made et al., 2014). The main advantage of SPPS is that excess reagents and side products can be removed by washing and filtration. This can be accomplished batch-wise or by packing the resin beads in columns. In batch mode, the resin beads are reacted and washed batch-wise under agitation. The resin beads are then filtered (using either suction or positive nitrogen pressure). In column mode, the resin beads are packed in columns for continuous-flow applications. The resin beads are then washed by pumping reagents and solvents through the column (Lukas et al., 1981).

Ideally, the solid-phase resin chosen should be inert to reagents and solvents (Made et al., 2014). For example, the solid-phase resin should withstand repeated use of TFA during deprotection. Because peptide synthesis occurs within the interior of the solid matrix, the solid-phase resin should also be mechanically stable (Made et al., 2014). For efficient peptide synthesis, several additional properties of the solid-phase resin need to be optimal (Made et al., 2014). These include resin solvation, resin cross linking, and resin swelling (Made et al., 2014). Solid-phase resin/ nascent peptide solvation is crucial for efficient peptide synthesis. Under proper solvation conditions, accessibility of the constituent resin chains to reagents is almost as high as that observed in solution (Albericio et al., 1989; Ford and Balakrishnan, 1981; Live and Kent, 1982; Ludwick et al., 1986; Manatt et al., 1980). To facilitate accessibility, solid-phase resin cross linking should generally be low. The solid-phase resin should also swell in all the solvents used (thus ensuring penetration of the reagents).

Peptide synthesis is also affected by peptide loading capacity (which itself depends on solvation, cross-linking, swelling, and functionalization/ linker derivatization). To avoid interchain crowding (and improve the synthesis of ‘Difficult sequences’), low levels of resin functionalization and/ or linker derivatization should be used (Tam and Lu, 1995). ‘Difficult sequences’ are defined as peptides which are poorly solvated while attached to the solid support (on-resin) (Mueller et al., 2020). Poor peptide solvation can prevent complete deprotection and inhibit subsequent coupling steps. The on-resin aggregation observed with ‘Difficult sequences’ typically involves sequences that encode -sheets and α-helices (Mueller et al., 2020). Low coupling efficiencies may also be enhanced by adding polar solvents and/ or chaotropic agents (Fields and Fields, 1994). However, successful synthesis of ‘Difficult sequences’ often requires some manipulation of the solid support. In particular, exchanging the solid support and reducing the loading capacity may improve the quality of the crude peptide (Mueller et al., 2020).

The choice of solid support is vital for successful peptide synthesis. The solid supports available (Albericio F 2000, Solid-phase synthesis: A practical guide. P.848) include gel-type, surface-type, and composite. Ultimately, choice of solid support (and linker) is largely dependent on the chosen SPPS reaction conditions (Made et al., 2014). However, choice of resin might also consider the need for a C-terminal carboxylic acid or C-terminal amide (e.g., Wang resin yields a C-terminal carboxylic acid). Specific resins are also required for the synthesis of protected peptides (e.g., Sieber amide resin) (Made et al., 2014). Finally, aggregation of ‘Difficult sequences’ is more likely on polar resins (Mueller et al., 2020). The list of resins commercially available for complex peptide syntheses is large and growing. Commonly used resins, and their swelling and loading (defined in equivalents of amino acids) properties, have been listed and discussed (Shelton et al., 2013; Made et al., 2014). The three main classes of solid support can be summarized as: Polystyrene (PS), PEG-functionalized PS, and pure PEG-based resins (e.g., PEGA), and ChemMatrix (Made et al., 2014). The most used solid supports are cross-linked polystyrene. These were first introduced by Merrifield in 1963 (Made et al., 2014).

Under proper solvation conditions, linear polystyrene chains in the solid support are nearly as accessible to reagents as polystyrene chains in solution (Albericio et al., 1989; Ford and Balakrishnan, 1981; Live and Kent, 1982; Ludwick et al., 1986; Manatt et al., 1980). NMR studies of other resins (e.g., Pepsyn; copolymerized dimethylacrylamide, N,N’-bisacryloylethylenediamine, and acryloylsarcosine methyl ester) have also revealed relatively high mobilities at resin-reactive sites. Solid supports created by grafting polyethylene glycol (polyoxyethylene) onto polystyrene combine the advantages of liquid-phase synthesis (i.e., a homogeneous reaction environment) and with those of solid-phase synthesis (an insoluble support). NMR measurements reveal that the polyoxyethylene (POE) chains of POE-PS are more mobile than the polystyrene matrix. Other supports that show improved solvation properties include polyethylene glycol polyacrylamide (PEGA), cross-linked acrylate ethoxylate resin (CLEAR), and augmented surface polyethylene prepared by chemical transformation (ASPECT).

During difficult syntheses, deprotection of the Fmoc group can proceed slowly. By monitoring deprotection with a spectrophotometer, problems can be detected during synthesis and the solvation conditions can be adjusted. Likewise, base-deprotection times can be extended once a problem is identified. Hou et al (2017) have reviewed resins commonly used for Fmoc-based SPPS (Hou et al., 2017). These include Rink amide, Pal, Sieber, Wang, Sasrin, HMPB, Trityl chloride, and 2-chlorotrityl chloride resins (Hou et al., 2017). Rink amide, Pal, and Sieber resins are required for Fmoc-based SPPS yielding C-terminal amides (Hou et al., 2017). Wang amide, Sasrin, HMPB, Trityl chloride, and 2-chlorotrityl chloride resins are required for Fmoc-based SPPS yielding C-terminal acids (Hou et al., 2017). For Boc-based SPPS, Merrifield resins yield a C-terminal acid, while MBHA resins yield a C-terminal amide (Hou et al., 2017).

Linker (and Choice of C-Terminal Modification)

A linker (or ‘handler’) is a bifunctional chemical moiety used to attach the nascent peptide to the solid-phase support during SPPS (Gongora-Benitez et al., 2013). The linker determines the separation between resin and C-terminal amino acid, the cleavage conditions utilized for liberation of the peptide following synthesis, and the C-terminal functionality of the released peptide (Made et al., 2014). Unless the peptide is alternatively linked to the solid-phase support (e.g., through a side chain), the linker also acts as a C-terminal carboxylic acid protecting group (Isidro-Llobet et al 2009; Gongora-Benitez et al., 2013). Thus, protection of the C-terminal carboxylic acid is different between SPPS and solution synthesis (Isidro-Llobet et al 2009). Linker development and utilization in SPPS has been extensively reviewed by Gongora-Benitez and colleagues (2013).

Choice of an orthogonal/ compatible set of protecting groups and linkers is essential for efficient SPPS of protected peptides (Gongora-Benitez et al., 2013). Ideally, the linker group should remain inert under all conditions encountered during peptide synthesis. Most importantly, premature cleavage of the linker group (which may occur during the coupling and deprotection cycles) should be minimized. At a pre-determined end-point (usually after peptide synthesis is complete), the linker should be cleavable under experimentally specified conditions, consequently releasing the peptide from the solid-phase support (Gongora-Benitez et al., 2013). Thus, the underlying goals are efficient peptide synthesis and downstream release of an undegraded peptide (Gongora-Benitez et al., 2013). Most peptides are released from the solid-phase support as C-terminal acids or amides. However, various linkers are available for the synthesis of C terminal modified peptides (Guillier et al., 2000).

A wide assortment of linkers has been developed, many designed with functionalities for use under specific circumstances. Linkers in this assorted toolkit can be cleaved by a variety of different mechanisms. Thus, linkers have been developed which are susceptible to acid, base, nucleophiles, fluoride, enzymes, light, reducing agents, oxidizing agents, and palladium complexes (Gongora-Benitez et al., 2013). Although the choice of available linkers is extensive, acid-labile linkers are commonly used for most standard syntheses (Gongora-Benitez et al., 2013). The acid lability of acid-labile linkers is dependent on the stability of the carbocation (Gongora-Benitez et al., 2013). Thus, acid lability can be modulated by incorporating substituents within the aromatic rings of the linker (Gongora-Benitez et al., 2013). Following acid cleavage (conventionally TFA), acid-labile linkers yield an unprotected C-terminus (Gongora-Benitez et al., 2013).

In most cases, acid-labile linkers are based on benzyl, benzhydryl, and trityl systems (Gongora-Benitez et al., 2013). Commonly-used linkers/ linker-derivatized resins based on the benzyl moiety include: SASRIN resin; 4-(4- hydroxymethyl-3-methoxyphenoxy)butyric acid (HMPB) linker; Sheppard linker; and the hypersensitive acid-labile (HAL) linker (Gongora-Benitez et al., 2013). Commonly-used linkers/ linker-derivatized resins based on the benzhydryl moiety include Rink acid resin and Rink chloride analog (Gongora-Benitez et al., 2013). Commonly-used linkers/ linker-derivatized resins based on the trityl moiety include 2-chlorotrityl chloride (2-CTC) resin and various trityl system-based analogs (Gongora-Benitez et al., 2013). As a rule, carbocations of trityl structures are more stable than carbocations of benzhydryl moieties, and carbocations of benzhydryl moieties are more stable than carbocations of benzylic structures (Gongora-Benitez et al., 2013). Several other acid-labile linkers are additionally available. These include the thiophene acid labile (THAL) linker (based on EDOT), and a family of highly acid-labile linkers/ linker-derivatized resins based on the xanthenyl moiety (e.g., Sieber amide resin, XAL linkers, and Sieber-based analog) (Gongora-Benitez et al., 2013).

The most widely used acid-labile resin is 2-CTC (Gongora-Benitez et al., 2013). This is mainly because of its extensive availability and low cost. However, 2-CTC resin (and the trityl-based linkers in general) offer other advantages in peptide synthesis. For example, cleavage of the protected peptide from the 2-CTC resin can be achieved using either 1 − 2% TFA in CH2Cl2 (i.e., a low-acidic solution) or using a trifluoroethanol (TFE)−CH2Cl2 mixture (i.e., under milder conditions) (Gongora-Benitez et al., 2013). In addition, epimerization (racemization) of the C-terminal residue is minimized (compared with benzyl alcohol-based supports) when using the 2-chlorotrityl linker (Gongora-Benitez et al., 2013). Thus, the attachment of susceptible Fmoc-amino acids (especially Fmoc Nα-protected Cys and Fmoc Nα-protected His) to trityl-based resins is largely free from epimerization (because of steric hindrance by the trityl groups).

Another advantage afforded by 2-CTC resin is minimization of diketopiperazine (DKP) formation (again because of steric hindrance by the trityl groups) (Gongora-Benitez et al., 2013). DKP formation is one of the most problematic deletion reactions in peptide synthesis (Isidro-Llobet et al., 2009; El-Faham & Albercio, 2011). Cyclic DKP can be released following base deprotection of the Fmoc group on the C-terminal amino acid, leaving a hydroxymethyl-handle group on the resin. Piperidine has been found to be a particularly efficient catalyst of this intramolecular aminolysis reaction. Trityl-based resins are particularly useful in the synthesis of peptides with a C-terminal Pro amino acid (as the bulk of the trityl linker helps to prevent DKP formation). However, DKP formation can be significant in other residues that can form cis peptide bonds (e.g., Gly, N-methylamino acids, or D-amino acids) in either the first or second position during synthesis.

Peptide cleavage can also be achieved under essentially neutral conditions using palladium-labile acid-stable (and base-stable) allylic linkers (Gongora-Benitez et al., 2013). Typically, palladium(0)-catalyzed allyl transfer to scavenger nucleophiles is employed to cleave allylic linkers (e.g., HYCRAM, HYCRON) (Gongora-Benitez et al., 2013). Under these cleavage conditions, Boc-groups, _t_Bu-esters, and _t_Bu-ethers remain intact (Gongora-Benitez et al., 2013). Silicon-based linkers have also been developed to exploit the stability of Fmoc SPPS side chain protecting groups to fluoridolysis (Gongora-Benitez et al., 2013). Again, protected peptides are released under basic or neutral conditions after cleavage of these fluoride-labile linkers (Gongora-Benitez et al., 2013). For example, cleavage of “silico Wang” linker Pbs can be achieved with tetrabutylammonium fluoride in DMF (in the presence of the appropriate scavengers). Similarly, cleavage of the SAC linker can be achieved with fluoride ions (or 1% TFA in CH2Cl2) (Gongora-Benitez et al., 2013).

C-terminal-modified peptides can be simply prepared using the backbone amide linker (BAL) method. Here, the peptide is anchored to the solid-support through a backbone nitrogen (Gongora-Benitez et al., 2013). After SPPS, the completed peptide can be released by cleavage of the BAL linker with 1 − 5% TFA. A panel of BAL linkers (e.g., tetraalkoxynaphthalene NAL-4) have been designed based on alkoxynaphthalene core structures (Gongora-Benitez et al., 2013). A novel thiophene backbone amide linker (T-BAL) has also been designed based on the EDOT scaffold (Gongora-Benitez et al., 2013). T-BAL exhibits very high acid-lability. For side chain attachment of C-terminal Asp/ Glu amino acids, an amide linker can be used (converting Asp/ Glu to Asn/ Gln). In this case, the α-carboxyl group must be protected as an allyl ester (Stawikowski and Cudic, 2006).

Advances in linker technology (e.g., the ‘safety-catch’ and ‘cyclorelease’ strategies) have been developed to facilitate additional synthesis reactions (Gongora-Benitez et al., 2013). Safety-catch linkers anchor the nascent peptide to the solid-support resin and are stable under SPPS conditions. The linkage remains stable until the safety-catch system is activated by a specific chemical reaction (Gongora-Benitez et al., 2013). After this additional activation step, release of a C-terminally modified peptide from the solid support is achieved under mild cleavage conditions (Gongora-Benitez et al., 2013). This system allows the employment of conditions that would otherwise cleave the peptide from the solid support.

Several linkers based on the safety-catch concept have been developed (Gongora-Benitez et al., 2013). Pascal et al. reported the use of a Dpr(Phoc) safety-catch linker for the synthesis of protected peptides using Fmoc chemistry (Gongora-Benitez et al., 2013). The associated cyclic acylurea is stable under conditions encountered during typical Fmoc/ _t_Bu chemistry (Gongora-Benitez et al., 2013). After peptide synthesis is completed, the protected peptide acid is released using NaOH and CaCl2 (in iProOH−water) (Gongora-Benitez et al., 2013). Alternatively, a protected peptide amide can be released using NH3 (Gongora-Benitez et al., 2013). Ellman and co-workers reported the development of an alkanesulfonamide safety-catch linker (4-Sulfamylbutyryl linker) that is stable under basic and strongly nucleophilic reaction conditions (Gongora-Benitez et al., 2013). Activation with iodoacetonitrile yields the N-cyanomethyl derivative. The completed peptide is subsequently released following cleavage under (comparatively mild) nucleophilic reaction conditions (Gongora-Benitez et al., 2013).

The cyclorelease strategy (or cyclative cleavage) relies on an intramolecular nucleophilic displacement reaction within the linker, which ultimately generates a peptide with an attached cyclized product (e.g., a DKP moiety) (Gongora-Benitez et al., 2013). Using this approach, a versatile linker affording protected peptides has been described. A dipeptidyl moiety is first incorporated onto hydroxymethyl resin, and this is followed by the assembly of a bifunctional linker. Following peptide synthesis, the Nα- protecting group on the dipeptidyl moiety is selectively removed. Piperidine treatment promotes DKP formation, intramolecular cyclization, and subsequent release of the protected peptide. The DKP moiety is maintained as a C-terminal protecting group (and selectively removed when required).

Another novel strategy for the synthesis of protected peptides is the use of photolabile linkers (Gongora-Benitez et al., 2013). Giese et al. reported the development of a photolabile linker based on the 2-pivaloylglycerol moiety for SPPS of protected peptides. The protected peptide is subsequently cleaved from this linker by irradiation (in THF) at 320 − 340 nm (Gongora-Benitez et al., 2013). Copley et al. reported the development of a photolabile linker based on the 3′,5′-dimethoxybenzoin moiety which is compatible with Fmoc/ _t_Bu SPPS (Gongora-Benitez et al., 2013). This photolabile linker can be quantitatively removed using 3% TFA (in CH2Cl2) followed by irradiation at 350 nm (in aqueous or partially aqueous media).

Several miscellaneous strategies for the synthesis of protected peptides have also been reported (Gongora-Benitez et al., 2013). A protected peptide can be cleaved from the resin under neutral conditions using an (azidomethyl)benzamide linker (Gongora-Benitez et al., 2013). Using this strategy, the _t_Bu, Boc, and Fmoc protecting groups can all be preserved (Gongora-Benitez et al., 2013). A wide variety of C-terminal-modified peptides can be prepared using the acid- and base-stable aryl hydrazine linker. This linker is subsequently cleaved under mild oxidative conditions in the presence of nucleophiles and basic media (Gongora-Benitez et al., 2013). This hydrazine linker is oxidized to acyldiazene by either O2 (in the presence of Cu(II) salts) and a nucleophile or, when oxidative-sensitive residues are present (e.g., Trp, Tyr, and Cys), by N-bromo-succinimide (NBS) followed by nucleophilic cleavage (Gongora-Benitez et al., 2013). Finally, a Dde-based linker can be employed in the Fmoc/ _t_Bu strategy (Gongora-Benitez et al., 2013). While this linker is stable under standard basic conditions, it can easily be cleaved in 2% hydrazine monohydrate (Gongora-Benitez et al., 2013).

As noted above, Gongora-Benitez and colleagues have provided an excellent review of linker strategies (Gongora-Benitez et al., 2013). The ultimate choice of linker is largely determined by the chosen Nα-protection/ side chain-protection orthogonal scheme, and the chosen resin (Made et al., 2014). It is also important to remember that peptide loading capacity is determined by levels of derivatization of the resin with the chosen linker. Therefore, to improve the synthesis of ‘Difficult sequences’), low levels of resin functionalization and/ or linker derivatization should be used to avoid interchain crowding (Tam and Lu, 1995). PEG-functionalized linkers can also be used to elevate solubility and decrease aggregation (and thus improve peptide synthesis yields) (Made et al., 2014).

Choice Of Coupling Reagents (For Efficient Amide Bond Formation)

During peptide synthesis in the C to N direction, a peptide bond is generated between the incoming amino acid and the N-terminal residue in the growing peptide chain. The term ‘coupling’ is used to describe the attack on the carbonyl atom of the carboxy group of the incoming amino acid by the Nα-amino group (of the N-terminal residue) (El-Faham & Albercio, 2011). The free carboxy terminus of the incoming amino acid must first be activated by the introduction of an electron-withdrawing group (El-Faham & Albercio, 2011; Made et al., 2014). The use of ‘coupling reagents’ to activate the free carboxy terminus is required because peptide bond formation is inherently slow. In addition, to ensure highly efficient peptide bond formation (which is often a requirement for successful peptide synthesis), an excess of each amino acid is added (2 – 10-fold). Under these favorable reaction conditions, the yield of the coupling step is typically extremely high. However, to prevent side-products (i.e., homopolymers of the activated amino acid), the Nα-amino group of the incoming amino acid must remain protected (Li et al., 2020).

The activated form of the incoming amino acid can be a shelf-stable reagent (e.g., some active esters) (El-Faham & Albercio, 2011). Alternatively, the activated form of the incoming amino acid may be a compound of intermediate stability (e.g., acyl halide, azide, or mixed/ symmetrical anhydride) or a transient (neither isolatable nor detectable) intermediate (e.g., containing O-acyl urea, acyl phosphonium, or acyl uranium groups) (El-Faham & Albercio, 2011). In some cases, the transient intermediate undergoes immediate aminolysis to the peptide (El-Faham & Albercio, 2011). In other cases, the transient intermediate may react with a second nucleophile that originates from the reactants (or that was added for the purpose) to yield a more stable active ester or symmetrical anhydride (El-Faham & Albercio, 2011). Subsequent aminolysis of this product then generates the peptide (El-Faham & Albercio, 2011).

Loss of Chiral Integrity

Although activation of the carboxyl residue is a necessary step, activation and subsequent coupling may induce a loss of chiral integrity (El-Faham & Albercio, 2011). Loss of chiral integrity may be caused by either direct enolization or 5(4H)-oxazolone formation (see Scheme 2 in El-Faham & Albercio, 2011). In either case, loss of configuration is catalyzed by base (El-Faham & Albercio, 2011). To avoid epimerization in the final peptide product, amino acid racemization should be minimized during coupling. Several strategies are used to minimize racemization during peptide coupling reactions. The most important consideration is choice of suitable Nα-protection group (El-Faham & Albercio, 2011). Thus, while carbamate decreases the likelihood of oxazolone formation, groups containing electron-withdrawing moieties are more prone to enolization (El-Faham & Albercio, 2011). The basicity and purity of the tertiary amines used during the coupling reaction are another issue (El-Faham & Albercio, 2011). Thus, H abstraction may be compromised in tertiary amines that are critically hindered (El-Faham & Albercio, 2011).

Side Reactions

There are also several side reactions intrinsically associated with the coupling step (El-Faham & Albercio, 2011). These include N-carboxyanhydride formation (when the Nα-protection group is a carbamate) and diketopiperazine (DKP) formation (when at least one dipeptide is present). These two reactions are strongly favored by the presence of a leaving group in the carboxyl function (the C-terminal one in the case of DKP formation). In addition, when an aminium/ uranium salt is used as a coupling reagent, a guanidine side product may arise. This is usually due to slow pre-activation of the carboxylic acid or to the use of excess uranium reagent. See Scheme 4 (El-Faham & Albercio, 2011).

Coupling reagents

Coupling reagents (‘activators’) are required to transform the free carboxy terminus into an active, more electrophilic species (Made et al., 2014). Nα-protected amino acid halides have a long history of use as coupling reagents in solution synthesis. Emil Fisher originally proposed the use of acid chlorides for peptide synthesis (1903). In later years, amino acid halides were extensively applied as amino acid coupling reagents in SPPS. Although acid halides had all but fallen out of fashion, there use is again on the rise (El-Faham & Albercio, 2011).The most conventional approach to amino acid activation and coupling involves the use of carbodiimides (El-Faham & Albercio, 2011). Sheehan and Hess introduced the use of dicyclohexylcarbodiimide (DCC) as a coupling reagent for the preparation of amide bonds in 1955. DCC has been used extensively in SPPS since its original introduction. These and other coupling reagents are now commercially available (Made et al., 2014). Almost all coupling reagents can be stored in DMF under N2 atmosphere (El-Faham & Albercio, 2011).

Carbodiimide coupling reagents

Carbodiimide coupling reagents such as DCC continue to be used extensively as activators (Rich and Singh, 1979; El-Faham & Albercio, 2011; Made et al., 2014). In general, carbodiimide-mediated couplings are achieved by pre-activation of the protected amino acid in dichloromethane (DCM) (El-Faham & Albercio, 2011). This activation reaction between weakly alkaline N atoms of the carbodiimide and the carboxylic acid group generates an O-acylisourea active intermediate (see Scheme 4 of El-Faham & Albercio, 2011). The highly reactive O-acylisourea species then undergoes rapid aminolysis in the presence of amine to yield the peptide (El-Faham & Albercio, 2011). This process occurs much faster in non-polar solvents such as DCM. However, while pure DCM is optimal for carbodiimide coupling, N,N-dimethylformamide (DMF) mixtures may be used for poorly soluble Fmoc Nα-protected amino acids (El-Faham & Albercio, 2011). A more polar solvent may also inhibit secondary structure formation and subsequent aggregation (El-Faham & Albercio, 2011).

The O-acylisourea may also undergo attack by a second molecule of carboxylic acid (when carboxylic acid is in excess) to yield the symmetrical anhydride (see Scheme 4 of El-Faham & Albercio, 2011). Subsequent aminolysis of the symmetrical anhydride again yields the peptide. O-acylisourea may also cyclize to yield the oxazolone (see Scheme 4 of El-Faham & Albercio, 2011). Again, subsequent aminolysis yields the peptide. Because the oxazolone is less reactive than O-acylisourea, the risk of racemization is higher. A final reaction is the rearrangement of O-acylisourea to N-acylurea, which is a ‘dead-end’ product (does not generate peptide). Although this irreversible reaction is slow in DCM, it proceeds much faster in DMF mixtures (El-Faham & Albercio, 2011).

In all reactions involving DCC,an N,N-dicyclohexylurea (DCU) precipitate is formed. Because DCU is only soluble in TFA, DCC is only compatible with Boc Nα-protection schemes (El-Faham & Albercio, 2011). However, several other carbodiimides are soluble in DCM and therefore suitable for use with Fmoc Nα-protection schemes. These include N,N′-Diisopropylcarbodiimide (DIC), N-ethyl-N‘-(3-dimethylaminopropyl)carbodiimide (EDC), and N-cyclohexyl-N‘-isopropylcarbodiimide (CIC) (El-Faham & Albercio, 2011). Because the urea product of DIC is easily washed away, DIC is especially useful for SPPS. In contrast, EDC is commonly used for LPPS.

As noted above, amino acid activation with carbodiimides is susceptible to racemization (especially when activation proceeds via the oxazolone intermediate) (El-Faham & Albercio, 2011; Made et al., 2014). To minimize the loss of chiral integrity, numerous additives capable of suppressing racemization have been developed (El-Faham & Albercio, 2011; Made et al., 2014). While the active esters formed using additives such as 1-hydroxybenzotriazole (HOBt) or 1-hydroxy-7-azabenzotriazole (HOAt) are less reactive than O-acylisourea, the overall efficiency of carbodiimide mediated reactions is increased (El-Faham & Albercio, 2011). N-hydroxy derivatives (HOXt) achieve this increase in efficiency by protonating and trapping O-acylisourea (El-Faham & Albercio, 2011). The result is a suppression of N-acylurea formation and a consequent decrease in racemization. Active ester formation is also promoted by the presence of a tertiary amine (El-Faham & Albercio, 2011). In terms of both yield and extent of racemization, the active esters formed by HOAt are considered superior (compared with HOBt) (El-Faham & Albercio, 2011). Nonetheless, the additive 6-Cl-HOBt is regarded as a good compromise between reactivity and price. In 2009, El-Faham & Albercio introduced a safe, highly efficient additive for use with carbodiimides (El-Faham & Albercio, 2009; El-Faham & Albercio, 2011). This additive, ethyl 2-cyano-(hydroxyimino)acetate (Oxyma), suppresses racemization and has a comparable coupling efficiency to HOAt (El-Faham & Albercio, 2011).

Anhydride-mediated coupling

For the symmetric anhydride-mediated coupling reaction, the incoming amino acid is allowed to react with carbodiimide in the absence of N-nucleophile (El-Faham & Albercio, 2011). The basic nitrogen atom of the O-acylisourea is protonated by the amino acid, and then the activated carbonyl of the acyl group is attacked by the carboxylate anion (El-Faham & Albercio, 2011). The resulting symmetrical anhydride is sufficiently stable to be isolated (El-Faham & Albercio, 2011). This reaction is generally carried out in DCM. Before the symmetric anhydride solution can be used for coupling, DCU must be removed (El-Faham & Albercio, 2011). The peptide is formed from the symmetric anhydride by aminolysis (El-Faham & Albercio, 2011). It should be noted that symmetrical anhydrides are less reactive and more selective than O-acylisourea. A mixed anhydride-mediated coupling reaction is also possible. However, this method is not regioselective (El-Faham & Albercio, 2011).

Active ester-mediated coupling

Active esters are mixed anhydrides formed from a carboxylic acid and a phenolic or hydroxamic acid (El-Faham & Albercio, 2011). The incoming amino acid is converted to an active ester by reacting it with a substituted phenol or a substituted hydroxylamine (HOXt) in the presence of carbodiimide (El-Faham & Albercio, 2011). The resulting N-alkoxycarbonylamino acid is stable enough to be stored. Nonetheless, this stable derivative is reactive enough to combine with an Nα-amino group during coupling. Indeed, the substituents render the carbonyl of the acyl moiety susceptible to nucleophilic attack by an amine at room temperature (El-Faham & Albercio, 2011). Many different types of active esters are currently available.

Acid-halide coupling reagents

A straight-forward approach for activation of the carboxyl group of an amino acid at room temperature involves the acid halide (El-Faham & Albercio, 2011). As noted above, this method was first introduced by Fisher in 1903 (El-Faham & Albercio, 2011). However, the activated species is over-active and prone to numerous side chain reactions, including loss of chiral integrity (El-Faham & Albercio, 2011). Moreover, the efficiency of this coupling reaction is significantly compromised when using carbamate-based Nα-groups (El-Faham & Albercio, 2011). It should be noted that acid-sensitive side chains (including _t_Bu-protected groups) are not compatible with this scheme (El-Faham & Albercio, 2011). Amino acid fluorides are more stable to hydrolysis, and their preparation is compatible with acid-sensitive side chains (El-Faham & Albercio, 2011). Thus, amino acid fluorides can be used for Fmoc-based SPPS (El-Faham & Albercio, 2011). This method has found widespread use in the coupling of hindered amino acids (El-Faham & Albercio, 2011). Nα-amino-protected amino acid fluorides are also useful for LPPS (El-Faham & Albercio, 2011).

Phosphonium salt coupling reagents

After pioneering studies by Castro and Coste, phosphonium salts such as CloP were widely adopted as coupling reagents (El-Faham & Albericio, 2011). However, racemization can be significant using this approach (El-Faham & Albericio, 2011). Because phosphonium salts react directly with the carboxylate, an equivalent of base is essential (El-Faham & Albericio, 2011). The salt is very reactive and reacts immediately with carboxylate ions. When phosphonium salts containing nucleophilic derivatives are used, the active species is known to be an active ester. While these reagents form the same active ester species as carbodiimide activation, they differ in the rate of the initial activation step. These couplings are carried out using an excess of base in the presence of a hydroxylamine derivative (e.g., HOBt, HOAt) (El-Faham & Albercio, 2011). HOBt phosphonium salt derivatives include the BOP coupling reagent. Although BOP coupling demonstrates several advantages, a toxic compound (HMPA) is generated during this reaction. To address this issue, coupling reagents which do not generate HMPA were introduced (e.g., PyBOP). HOAt phosphonium salt derivatives include AOP and PyAOP. These coupling reagents are generally more efficient than BOP or PyBOP. More recently, Oxyma-based phosphonium salt derivatives have been introduced (El-Faham & Albercio). The phosphonium salt of Oxyma (PyOxm) is an efficient, racemization-suppressing coupling reagent, and it is especially useful for the assembly of hindered peptides (El-Faham & Albercio, 2011). PyOxm also renders a high percentage of cyclic peptides.

Aminium salt coupling reagents

Aminium salts (also called uranium salts) bear a positive carbon atom instead of the phosphonium group. Mechanistically, aminium salts (e.g., N-HBTU, N-HBTU, etc.) function in a similar way to their phosphonium analogs. Thus, in the presence of tertiary base, an active ester is generated during the formation of carboxyl aminium salts. HOXt can be used to accelerate coupling and reduce the loss of configuration (El-Faham & Albercio, 2011). There are numerous aminium salt coupling reagents, and many are commercially available (El-Faham & Albercio, 2011). COMU is a third generation uranium salt derived by El-Faham & Albercio (El-Faham & Albericio, 2009). Here, Oxyma is used as a leaving group to provide a superior and safe coupling reagent for amide formation (El-Faham & Albercio, 2011). COMU is also compatible with microwave assisted peptide synthesizers. One disadvantage is that aminium salts can react directly with the Nα-amino group of the amino acid (especially during slow pre-activation of the carboxylic acid or in the presence of an excess of aminium). After coupling, the resulting guanidine side product induces termination of the peptide chain (El-Faham & Albercio, 2011). In contrast, phosphonium salts do not react with the Nα-amino group.

Miscellaneous coupling reagents

Several other types of coupling reagents have also been developed (see El-Faham & Albericio, 2011). When using carbonyl diimidazole (CDI), the activated species (acylimidazole) is pre-formed for 1 h (El-Faham & Albercio, 2011). A peptide bond is subsequently formed upon reaction with the Nα-amino group of an amino acid. This method is commonly used during large scale peptide synthesis. N-(protected-α-aminoacyl) benzotriazoles facilitate rapid synthesis under mild reaction conditions (El-Faham & Albercio, 2011). High yields and purity can be achieved without loss of chirality when using this method (El-Faham & Albercio, 2011). Thus, complex peptides and peptide conjugates can be generated using simple LPPS and SPPS strategies (El-Faham & Albercio, 2011). Although acyl azide methods largely preserve chiral integrity, they are considered cumbersome and unsuitable for repetitive peptide synthesis strategies (El-Faham & Albercio, 2011). Nonetheless, these methods are useful for peptide-bond formation between peptide fragments (El-Faham & Albercio, 2011). Phosphinic and phosphoric acid derivatives (e.g., DPPA, MPTA) have also been developed as coupling reagents. Using these organophosphorus reagents, coupling proceeds through a carboxylic-phosphinic mixed anhydride (El-Faham & Albercio, 2011). Lastly, the triazine coupling reagent 2-chloro-4,6-dimethyl-1,3,5-triazine (DMCT) is a stable commercially available compound (El-Faham & Albercio, 2011). This method requires the presence of a tertiary amine in the reaction medium.

Polymer-Supported Coupling Reagents

Several immobilized coupling reagents have been developed for use in LPPS (El-Faham & Albercio, 2011). These include polymer-bound carbodiimides. For example, PS-EDC can be synthesized using Merrifield resins. Likewise, polymer-bound DCC and polymer-bound DIC have also been synthesized. However, epimerization remains a disadvantage of these immobilized coupling reagents.

Practical aspects of Coupling

No single coupling reagent is suitable for all coupling reactions (El-Faham & Albericio, 2011). Indeed, several different coupling reagents may be used during the synthesis of a single peptide. For each coupling reaction, numerous factors need to be considered. These include price, synthesis chemistry (e.g., whether LPPS or SPPS is used), synthesis process (e.g., whether manual or automatic), whether an excess of coupling reagents can be used, and whether other functional groups are present. The choice of coupling agent also depends on the bulkiness of the amino acid to be coupled, its solubility, and its stability (Made et al., 2014). The synthesis of Oxa-thiocoraline provides an excellent illustration of the multifaceted process by which a peptide is synthesized (El-Faham & Albercio, 2011).

For the synthesis of the most challenging peptides, an appropriate combination of coupling reagent, Nα-amino protecting group, solid support, solvent temperature, and other experimental conditions must be found (El-Faham & Albercio, 2011). To increase the chances of a successful synthesis, reactivity should be enhanced, and epimerization minimized (Made et al., 2014). An increase in reaction temperature can speed up assembly and reduce synthesis time (El-Faham & Albercio, 2011; Behrendt et al., 2015). Heating can be applied by conventional means, by microwave, or by IR radiation (Behrendt et al., 2015). However, an increase in reaction temperature may lead to loss of peptide loading during peptide assembly, cysteine and histidine racemization, aspartimide formation, etc. (Behrendt et al., 2015).

Peptide Cleavage

In SPPS, the crude peptide must be released from the solid support (“cleaved”) following completion of peptide synthesis. The cleavage conditions utilized are largely dependent on the linker employed. As mentioned above, linkers have been designed which can be cleaved with acid, base, nucleophiles, fluoride, enzymes, light, reducing agents, oxidizing agents, and palladium complexes (Gongora-Benitez et al., 2013). The cleavage conditions must also be compatible with the chosen orthogonal/ compatible synthesis scheme (Gongora-Benitez et al., 2013). In general, cleavage conditions in orthogonal protection schemes are usually milder. This is because peptide cleavage is governed by alternative cleavage mechanisms and not by rates (Isidro-Llobet et al., 2009). In the series of studies carried out by the Peptide Synthesis Research Committee (PSRC) of the Association of Biomolecular Resource Facilities (ABRF), peptide cleavage conditions were reported to have an impact on peptide synthesis success.Thus, the peptide cleavage protocol used is essential to a successful outcome.

In the original SPPS study by Merrifield, after peptide synthesis using Cbz Nα-protected amino acids, the peptide was cleaved from the support using HBr. HF was introduced as a reagent for peptide cleavage in a subsequent study (Sakakibara et al., 1967). During peptide cleavage from the resin (and simultaneous cleavage of side chain protecting groups), by-products may be released (Made et al., 2014). Therefore, these final steps are frequently performed in the presence of scavengers (Made et al., 2014). For example, water and triisopropylosilane (TIPS) can be added during final cleavage to prevent side chain reactions. These by-products can also be removed during subsequent HPLC purification of the cleaved peptide.

In the Boc/ Bzl protecting group scheme, TFA (a moderate acid) is used for deprotection of the Boc Nα-protection group. Therefore, the chosen linker should be (relatively) stable to TFA treatment. The semi-permanent side-chain protecting groups (Bzl) are also stable under Nα-deprotection conditions. During the final peptide cleavage step, the linker and side-chain protecting groups are simultaneously removed by hydrolytic cleavage with anhydrous HF (a strong acid). Because the Boc/ Bzl protecting group scheme is based on ‘relative acidolysis’, the reaction conditions used for Nα-deprotection are not completely differentiated from those used for sidechain deprotection and peptide cleavage (Li et al., 2020). Indeed, premature cleavage of the nascent peptide from the polymer support during TFA-induced Nα-deprotection is a major limitation of Boc-based SPPS (Behrendt et al., 2015). This progressive loss of peptide from the polymer support is likely to produce unwanted side products. Another disadvantage is the potential for degradation of peptide by HF. Cresol can be added as a scavenger to HF to prevent reactive t-butyl cations from generating undesired products. It should also be noted that liquid HF is toxic. Thus, the use of strong acid (liquid HF) is less than ideal. However, the final product of Boc/ Bzl SPPS is a HF salt, which is easy to solubilize.

In the Fmoc/ _t_Bu protecting group scheme, base is used for deprotection of the Fmoc Nα-protecting group. Therefore, the chosen linker should be stable to base treatment. The semi-permanent side-chain protecting groups are also stable under deprotection conditions. During the final peptide cleavage step, the linker and side-chain protecting groups commonly used in Fmoc chemistry (e.g., _t_Bu) are simultaneously removed using TFA (a moderate acid). TFA cleavage of the linkers and side-chain protecting groups results in the liberation of reactive species that can modify susceptible residues (e.g., Trp, Tyr, and Met). These modifications can be minimized by utilizing effective scavengers. As detailed above, efficient cleavage cocktails that can quench reactive species and preserve amino acid integrity are available. For most standard syntheses, acid-labile linkers are used (Gongora-Benitez et al., 2013). Following acid cleavage (conventionally with TFA), acid-labile linkers yield an unprotected C-terminus (Gongora-Benitez et al., 2013). The crude peptide is obtained as a TFA salt, which is more difficult to solubilize than the HF salt obtained from Boc SPPS.

Depending on the linker chosen, peptide cleavage can also be achieved under non-standard cleavage conditions to yield a protected peptide (Gongora-Benitez et al., 2013). Under neutral conditions, a protected peptide can be released using palladium-labile acid-stable and base-stable allylic linkers (Gongora-Benitez et al., 2013). Typically, palladium(0)-catalyzed allyl transfer to scavenger nucleophiles is employed (Gongora-Benitez et al., 2013). Under basic or neutral conditions, protected peptides can also be released using fluoride-labile linkers (Gongora-Benitez et al., 2013). For example, cleavage of “silico Wang” linker Pbs can be achieved with tetrabutylammonium fluoride in DMF (in the presence of the appropriate scavengers). Under neutral conditions, protected peptides can be released using an (azidomethyl)benzamide linker (Gongora-Benitez et al., 2013). The peptide is cleaved using Bu3P in a mixture of DMF-imidazole buffer at pH7 (Gongora-Benitez et al., 2013). Under these conditions, _t_Bu, Boc, and Fmoc protecting groups remain intact. Under mild oxidative conditions (in the presence of nucleophiles and basic media), the acid- and base-stable aryl hydrazine linker is cleaved to release the protected peptide (Gongora-Benitez et al., 2013). Under standard basic conditions, the Dde-based linker can be cleaved in 2% hydrazine monohydrate. These cleavage conditions are compatible with the Fmoc/ _t_Bu protection scheme (Gongora-Benitez et al., 2013). Cleavage of protected peptides can also be achieved through irradiation of photolabile linkers (Gongora-Benitez et al., 2013).

The “safety-catch” and “cyclorelease” strategies have been developed to facilitate additional synthesis reactions (Gongora-Benitez et al., 2013). Using a safety-catch system, the linkage remains stable until activated by a specific chemical reaction (Gongora-Benitez et al., 2013). Subsequent release of the peptide is achieved under mild cleavage conditions (Gongora-Benitez et al., 2013). This system allows the employment of conditions that would otherwise cleave the peptide from the solid support. The cyclorelease strategy (or cyclative cleavage) relies on an intramolecular nucleophilic displacement reaction within the linker, which ultimately generates a protected peptide with an attached cyclized product (e.g., a DKP moiety) (Gongora-Benitez et al., 2013). Following peptide synthesis, the Nα-protecting group on the dipeptidyl moiety is selectively removed. Piperidine treatment promotes DKP formation, intramolecular cyclization, and subsequent release of the protected peptide.

“Difficult Sequence” Problem

Although a 50 amino acid residue peptide synthesis may be achievable, a successful synthesis cannot be guaranteed, and much shorter sequences can be problematic (Behrendt et al., 2015). Synthesis of “difficult sequences” by SPPS can be a particular challenge (Tickler & Wade, 2007; Mueller et al., 2020). In this context, difficult sequences are defined as peptides which are poorly solvated while attached to the solid support, thus impeding deprotection and coupling steps (Mueller et al., 2020; Mueller et al., 2020). In general, peptide insolubility is considered one of the greatest problems in peptide chemistry (Behrendt et al., 2015). These difficult sequences are highly prone to aggregation and precipitation in conventional solvents (Tickler & Wade, 2007; Mueller et al., 2020). Hence, poor solubility may also interfere with peptide purification and characterization (Behrendt et al., 2015). Well-known examples of difficult sequences include amylin, A1-42, α-synuclein, and BM2(1-51) (Tickler & Wade, 2007; Mueller et al., 2020).

In most cases, difficult sequences show a tendency to form strong inter-molecular and intra-molecular non-covalent interactions, increasing the likelihood of peptide aggregation (Tickler & Wade, 2007; Mueller et al., 2020). A consequent increase in steric hindrance at the N-terminus of the growing peptide hinders Nα-deprotection and acylation (Tickler & Wade, 2007). On-resin aggregation is particularly associated with primary sequences that encode -sheets and α-helices (Tickler & Wade, 2007; Mueller et al., 2020). Since these secondary structures are typically associated with regular patterns of hydrophobic residues, difficult sequences may be anticipated (Behrendt et al., 2015). For example, difficult sequences generally contain high numbers of -branched amino acid residues (L, V, F, or I) (Mueller et al., 2020). Difficult sequences containing both -branched residues and Gly may be capable of -sheet packing.

Five factors that affect the synthesis of difficult peptides have been identified (Tickler & Wade, 2007). 1. High resin loadings can promote intermolecular interactions and consequent SPPS failure. 2. Nonpolar resins such as polystyrene may yield poorer quality peptides. In contrast, polar resins create a polar environment that reduces the tendency for self-aggregation. 3. The use of polar solvents (e.g., DMF) significantly improves SPPS of difficult sequences. However, not all polar solvents/ solvent mixtures are compatible with SPPS. 4. SPPS problems mostly occur during synthesis of residues 5 – 15 (from the C-terminus). 5. Secondary imino acids (e.g., Pro) have intra-chain structure-breaking properties. The presence of these residues can aid in the synthesis of (otherwise) difficult sequences.

To address the difficult sequence problem, protocols should be optimized for each synthesis (Tickler & Wade, 2007; Mueller et al., 2020). Resin type and resin loading capacity both have an influence on the yield of difficult sequences (Tickler & Wade, 2007; Mueller et al., 2020). As mentioned above, the aggregation tendency of difficult sequences on non-polar resins (e.g., polystyrene) is high (Tickler & Wade, 2007; Mueller et al., 2020). Several resins have been developed for use with difficult sequences. ChemMatrix (CM) resin (Matrix Innovation) is a 100% PEG solid-support resin. Even high loading capacities are possible with CM resin. In a standard Aβ1-42 synthesis, substitution with CM resin resulted in a crude peptide purity of approximately 91%, much higher than previous attempts to synthesize this peptide (Tickler et al., 2001). The AminoMethyl SUrface-layered polystyrene REsin (AMSURE) resin is a 1% divinylbenzene (DVB) cross-linked polystyrene resin (BeadTech). Although it is a nonpolar support, it has functionalized linker groups located on the surface. This modification minimizes steric crowding within the resin bead core. The resin may also be functionalized to provide the peptide with H-bonding partners that can participate in H-bonds with resin-bound peptides (interfering with intrachain or interchain peptide H-bonds).

Another important strategy to improve SPPS protocols for difficult sequences is modification of the external conditions (Tickler & Wade, 2007; Mueller et al., 2020). Polar organic solvents like DMF are known for their increased solvation potential, and their ability to inhibit aggregation. Likewise, NMP is known to facilitate peptide solubilization and improve SPPS yields. DMSO is also known to inhibit peptide aggregation and improve acylation efficiency. In addition, chaotropic salts can be used to break up aggregates on resin prior to or during amino acid acylation. A “magic mixture” comprised of DCM, DMF, and NMP (1:1:1) together with 1% Triton X-100 detergent and 2 M ethylenecarbonate at 55°C has been shown to improve yield in difficult acylations. Insufficient Fmoc Nα-deprotection is another contributory factor in the failure of difficult sequence synthesis (Tickler & Wade, 2007; Li et al., 2020). In this respect, DBU + 5% piperidine (to quench DBF) may be beneficial for the synthesis of difficult sequences (Li et al., 2020).

One of the most important developments in the synthesis of difficult sequences is the widespread availability of instruments with microwave irradiation capabilities (Tickler & Wade, 2007; Mueller et al., 2020). In general, higher reaction temperatures improve the final yield of a peptide synthesis. While the benefits of microwave irradiation in organic synthesis have been known for many years (Gedye et al., 1986; Giguere et al., 1986), dedicated microwave peptide synthesizers have only become widely available more recently. Microwave irradiation accelerates the chemical reactions by increasing the kinetic energy of the reagents. The polar N-terminus amine group and peptide backbone of the peptide may also align with the alternating electric field of the microwave, precluding the formation of stable intermolecular or intramolecular H-bonds.

The placement of Pro residues at regular intervals is also likely to increase SPPS yield. However, the introduction of Pro residues may not be compatible with peptide structure/ function. A very successful strategy for overcoming the problem of difficult sequences is backbone protection (Tickler & Wade, 2007; Behrendt et al., 2015; Mueller et al., 2020). Pseudo-proline dipeptides (Ser, Thr, and, Cys) can be incorporated into difficult sequences to disrupt secondary structure formation and thus inhibit aggregation (Wohr and Mutter, 1995; Tickler & Wade, 2007; Mueller et al., 2020). Insertion of a pseudo-proline in the first five residues, within five residues of another Pro, or within five residues of the N-terminus is not necessary. After cleavage from the resin with TFA, pseudo-proline dipeptides are converted into native dipeptide sequences. A range of pseudo-proline dipeptides are now commercially available.

A comparable strategy to the insertion of Pro residues or pseudo-prolines involves the use of N-alkyl amino acid derivatives (e.g., Hmb-modified residues) (Tickler & Wade, 2007; Mueller et al., 2020). Following incorporation, the Hmb group disrupts the formation of secondary structure (Johnson et al., 1995). Thus, it prevents intermolecular/ intramolecular interactions, and frees the N-terminus for further acylation. A problem of inefficient acylation associated with bulky Hmb derivatives has been overcome with the introduction of the Dmb group (Zahariev et al., 2006). The O-acyl isopeptide method is another secondary structure-breaking technique. This method involves the introduction of oxo-esters on Ser or Thr. Again, the result is a kink in the amide backbone which precludes secondary structure formation. The resulting isopeptide is stable until purification in a solution at pH 7.4. Because the above methods all require Ser, Thr, or Cys residues, their use is dependent upon the primary sequence.

Improved yields may also be obtained with Boc SPPS, although this is largely because of the use of TFA as a solvent (Mueller et al., 2020). During Boc SPPS, TFA selectively dissolves the peptide and disrupts the formation of secondary structures (Mueller et al., 2020). In addition, optimization of Boc-based SPPS using in situ neutralization protocols facilitates synthesis of difficult sequences (Mueller et al., 2020). During the in situ protocol, a high concentration of activated amino acid in a polar solvent containing DIEA is added directly (Mueller et al., 2020). It should be noted that Boc-based SPPS is not suitable for use with backbone modifications designed for Fmoc-based SPPS (Mueller et al., 2020). This is because of iterative use of TFA and the final cleavage with HF (Mueller et al., 2020). Several photo-labile backbone amide protection groups have been introduced for the synthesis of difficult sequences using Boc-based methods (Johnson and Kent, 2006; Mueller et al., 2020).

Tickler & Wade (2007) provide four recommendations for the synthesis of difficult sequences: 1. Use microwave irradiation; 2. Choose a hydrophilic support and avoid high loading; 3. Use a polar solvent (DMF is the preferred choice); and, 4. Use appropriately selected and positioned pseudoproline and N-alkyl derivatives (if the primary sequence allows). Finally, consideration should be given to synthesizing a difficult sequence in two (or three) smaller segments which can then be ligated (Tickler & Wade, 2007). Mueller et al (2020) have proposed a step-by-step approach to the synthesis of difficult sequences. The steps are: SPPS; analytical characterization; purification; fragment ligation; and post-ligation steps (if needed) (Mueller et al., 2020). During each step, high peptide solubility is essential. Semi-permanent backbone modifications that remain stable through fragment ligation should be used. These semi-permanent backbone modifications can be subsequently cleaved with TFA. Additional strategies include the addition of organic solvents (TFE) or surfactants (OG/ DPC) to the ligation solution (Mueller et al., 2020). Using fragment ligation, the limitations of SPPS in the synthesis of some difficult sequences may be surmounted (Mueller et al., 2020).

Automated and Multiple Syntheses

The first automated peptide synthesis instrument was built by Merrifield, Stewart, and Jernberg (Merrifield et al., 1966). This instrument was built to automate the Boc SPPS process. Commercially available instruments for peptide synthesis are now commonplace (Table 18.1.1). Indeed, automated peptide synthesizers can be purchased from more than 15 companies (Made et al., 2014). A review of instruments commonly used for fully automated single and parallel SPPS has been published (Pedersen & Jensen, 2013). These peptide synthesizers normally have different capabilities (Made et al., 2014). Notable differences can include: the type of solution transfer; the mixing process; the synthesis scale; the use of automated monitoring; the availability of microwave heating; the employment of an inert atmosphere; and the availability of automated peptide cleavage (Made et al., 2014). In general, synthesizers may be run in batch-wise or continuous-flow modes (Made et al., 2014). Ultimately, the choice of peptide synthesizer depends on the intended application.

In many automated SPPS instruments, peptide synthesis can be performed in parallel in different reaction vessels (Made et al., 2014). Thus, multiple, different peptide sequences can be synthesized at the same time. While the same solution can be used for Fmoc Nα-deprotection and washing, the coupling steps are usually performed individually under optimized conditions (Made et al., 2014). Because excess amounts of reagents are used, high yields of synthesized peptides are possible (Made et al., 2014). However, problems may be encountered when synthesizing difficult sequences and larger peptides, including the presence of impurities, synthesis termination, and low yields. To avoid racemization and side reactions, temperature optimization is essential (Made et al., 2014). It should be noted that the reaction temperature is influenced by solvents, reactants, volumes, and mixing mode. To improve overall crude peptide quality, microwave-assisted synthesis can also be incorporated into automated peptide synthesis instruments (Made et al., 2014). Microwave-assisted SPPS has been shown to enhance reaction rates and thus facilitate the synthesis of difficult sequences (Made et al., 2014). Moreover, the synthesis of longer peptide sequences at high yield and low degrees of racemization is now a possibility.

Peptide Modifications

Purification and Analysis of Synthesized Peptides

The by-products that accumulate during peptide synthesis and upon cleavage can be removed during peptide purification by high-performance liquid chromatography (HPLC). Efficient characterization of synthetic peptides is best achieved using a combination of reverse-phase HPLC (RP-HPLC) and mass spectrometry (MS). However, the homogeneity of synthetic materials should be checked by at least two chromatographic or electrophoretic techniques (e.g., RP-HPLC, ion-exchange HPLC, capillary zone electrophoresis). In many cases, sequencing information may be required. Edman degradation sequence analysis and tandem MS can both be used to check the sequence. Using these methods, the positions of modifications and deletions can be identified. To obtain proof of structure, the corresponding molecular ion should be determined by MS using soft ionization techniques (e.g., MALDI-TOF, ESI). It should be noted that poor solubility can still complicate or prevent peptide characterization (Behrendt et al., 2015).

Chemical Ligation

SPPS of longer peptides is limited by the synthesis efficiency at each cycle (Gongora-Benitez et al., 2013). As an example, consider the synthesis of a 50-residue peptide. With a hypothetical efficiency at each synthesis cycle of 99.9%, the final peptide yield is 95%. With a hypothetical efficiency at each synthesis cycle of 95%, the final peptide yield decreases to 8%. However, for a 100-residue peptide, the final yields are 90% (efficiency of cycle, 99.9%) and 0.6% (efficiency of cycle, 95%). Because the final yield decreases as peptide length increases, there is a length limitation on peptide synthesis (even when the synthesis efficiency at each cycle is high). Furthermore, the tendency of growing peptide chains to aggregate also increases as peptide length increases (Gongora-Benitez et al., 2013). After cleavage, longer peptides (especially those with difficult sequences) may also show poor solubility. Thus, SPPS of large peptides is inherently a challenging process.

Despite these limitations, the demand for longer peptides is increasing, and ligation strategies have been developed to facilitate the synthesis of long peptide sequences (El-Faham & Albercio, 2011; Gongora-Benitez et al., 2013; Hou et al., 2017). In general, fragment ligation strategies rely on the preparation (synthesis, purification, and characterization) of multiple protected peptide fragments, and their subsequent assembly (condensation) (Gongora-Benitez et al., 2013; Made et al., 2014). These convergent synthesis strategies often offer the only possibility for manufacturing peptides larger than 50 amino acids (Made et al., 2014). Using these ligation strategies, many of the limitations of SPPS can be overcome (Gongora-Benitez et al., 2013; Mueller et al., 2020). In particular, the cumulative effects of stepwise synthetic errors are minimized.

For conventional fragment assembly, the protected fragments should be purified (to ensure fragment integrity) before being covalently linked to produce the final polypeptide (Albericio et al., 1997; Gongora-Benitez et al., 2013). The C-terminal-protected fragment (acyl-acceptor-peptide fragment) and N-terminal-protected fragment (acyl-donor-peptide fragment) can be coupled on resin (solid-phase fragment coupling) or in solution (hybrid strategy) (Gongora-Benitez et al., 2013). The protecting groups are then removed to yield the finished polypeptide (Gongora-Benitez et al., 2013). While N-terminal fragments may be directly used in the hybrid strategy, the C-terminal carboxylic acid of C-terminal fragments should be protected before fragment coupling (Gongora-Benitez et al., 2013). The efficiency of the subsequent fragment coupling reaction is markedly lower than the efficiency of SPPS coupling reactions (using activated amino acids), and racemization may occur at the C-terminus.

A remaining obstacle is the insolubility of the fragments/ assembled polypeptide in conventional buffers. Strategies to increase fragment solubility and minimize aggregation (and thus increase the efficiency of ligation protocols) involve modifications of the external conditions (e.g., polar organic solvents, fluorinated alcohols, chaotropic agents, surfactants, ionic liquids) and internal modifications of the sidechains or backbone. Any semi-permanent backbone modifications may be removed after assembly. If necessary, solubilizing tags (e.g., poly-R, poly-K) can be incorporated at the N-terminus, C-terminus, or on side chains. Efficient assembly of protected N-terminal and C-terminal fragments requires handles, and the handle cleavage conditions should be compatible with maintaining side chain protecting groups (Gongora-Benitez et al., 2013). Numerous handles have been applied in the synthesis of protected peptides for use in the fragment condensation approach. Handles that are compatible with Fmoc/ _t_Bu fragment synthesis for subsequent assembly include diluted-acid-labile, palladium-labile, silicon-based, BAL approach, photolabile, and safety-catch linkers (Gongora-Benitez et al., 2013). When using these handles, semi-permanent backbone modifications can be subsequently cleaved with TFA.

The condensation step can also be achieved using chemo-selective ligation techniques (e.g., Native Chemical Ligation, Expressed Protein Ligation, click reactions) (Made et al., 2014; Mueller et al., 2020). The Native Chemical Ligation (NCL) approach involves reversible trans thioesterification of de-protected fragments and subsequent amide formation (Dawson et al., 1997; Muir et al., 1997; Ayers et al., 1999; El-Faham & Albericio, 2011; Mueller et al., 2020). The initial thiol-thioester exchange step is between an appropriate C-terminal thioester on the N-terminal fragment and the thiol moiety of a Cys situated at the N-terminus of the C-terminal fragment. A subsequent intramolecular nucleophilic attack by the Nα-amino group (of the C-terminal fragment) on the initial thioester product yields a native peptide bond (and regenerates the Cys thiol) (El-Faham & Albercio, 2011). This amide-forming step is irreversible under the reaction conditions (El-Faham & Albercio, 2011). The NCL reaction takes place in aqueous media (containing 6M GdnHCl or 8M urea and reducing reagent) at neutral pH (El-Faham & Albercio, 2011). “Safety-catch” linkers can be used in conjunction with Fmoc chemistry to produce the necessary peptide thioester (Shin et al., 1999). The main limitation of NCL is the mandatory use of a Cys residue at the N-terminus of the C-terminal fragment.

Peptide Drug Development

In the last two decades, the pharmaceutical market for peptide therapeutics has developed enormously (Made et al., 2014; Erak et al., 2018; Muttenthaler et al., 2021). By 2014, the global market for peptide therapeutics was expanding nearly twice as fast as the overall drug market (Made et al., 2014). While the global market for peptide therapeutics was almost $10 billion in 2005 (Made et al., 2014), it had exceeded $40 billion by 2019 (Muttenthaler et al., 2021). As of 2014, 70 peptide drugs had been approved by the FDA (Made et al., 2014). Between 2011 and 2017, the FDA approved 22 new peptide entities (Erak et al., 2018). Currently, there are around 80 peptide drugs on the global pharmaceutical market (Muttenthaler et al., 2021). These account for approximately 5% of the $1.2 trillion in sales. Muttenthaler et al. (2021) provide a list of the current top peptide drugs by sales (see Table 1 of Muttenthaler et al., 2021). In addition, there are approximately 150 peptides in clinical development, and 400 – 600 peptides undergoing preclinical studies (Erak et al., 2018; Muttenthaler et al., 2021). The peptide market is only predicted to continue its expansion (Muttenthaler et al., 2021).

In contrast to small synthetic drugs, which generally demonstrate moderate target potency and selectivity (and consequently many side effects), peptide drugs have the potential to demonstrate high potency and selective binding. As a consequence of their great versatility and inherent high affinity, peptides have found many diverse therapeutic applications (Made et al., 2014; Erak et al., 2018; Muttenthaler et al., 2021). As of 2021, the majority of peptide therapeutics available are agonists with indications in endocrinology (e.g., insulin analogues), metabolism (e.g., dulaglutide, liraglutide), and oncology (e.g., leuprolide) (Muttenthaler et al., 2021). In addition, new peptide targets are continuously being identified using omics techniques. Synthesized peptides can also be used as targeted delivery vectors, image tags, etc. (Erak et al., 2018; Muttenthaler et al., 2021).

The recent success of peptide biologics can be attributed to the maturity of peptide synthesis technology (Erak et al., 2018; Muttenthaler et al., 2021). In addition, advances in methodologies for the assembly of chemically synthesized peptide fragments have facilitated the production of larger therapeutic peptides (Muttenthaler et al., 2021). Together, these techniques allow cost-effective synthesis of complex peptides and peptide conjugates (Muttenthaler et al., 2021). For example, a 50% reduction in Exenatide production costs was observed when these techniques were combined (in comparison with SPPS only) (Muttenthaler et al., 2021). For modified peptides, standard and non-standard syntheses are both often required, possibly necessitating both automatic SPPS and manual SPPS (Made et al., 2014).

As mentioned above, the target potency and selectivity of small synthetic drugs are generally inferior to those of peptide therapeutics. However, small synthetic drugs have many advantages over natural peptides. In general, small synthetic drugs are characterized by their high metabolic stabilities, their (relatively) small molecular size, their comparative membrane permeability (depending on other molecular characteristics), simple production techniques, and low costs (Made et al., 2014). In contrast, natural peptides are highly susceptible to proteolytic degradation and rapid body clearance, are comparatively membrane impermeable, their druggability is restricted by their poor water solubility, and they are relatively expensive to produce (Made et al., 2014). A full list of the many advantages and disadvantages of peptides as therapeutic agents is presented below (adapted from Made et al., 2014; Erak et al., 2018; Muttenthaler et al., 2021). Some of the more important disadvantages of natural peptides can be overcome through peptide modification.

Advantages of Peptides as Therapeutic Agents

Disadvantages of Peptides as Therapeutic Agents

To improve therapeutic peptide properties, peptide modifications have been widely adopted (Erak et al., 2018). One of the most important objectives is to increase the potency and specificity of selective peptide leads. Another important objective is to prolong peptide stability. Thus, the metabolic and circulation half-life of a potent peptide lead may require improvement (Muttenthaler et al., 2021). Another objective is to increase the bioavailability of therapeutic peptides. This may be achieved through modification or through new formulations (e.g., integration into particles, gels, liposomes) (Made et al., 2014). New peptide delivery systems have also been developed (Muttenthaler et al., 2021). These include slow-release injections, intranasal delivery, and hydrophobic depots (Muttenthaler et al., 2021). The above objectives may be addressed using either a rational methodology (lead-structure oriented) or combinatorial methodologies (which require minimal system knowledge) (Made et al., 2014). However, to achieve multiple objectives, multiple different peptide modifications may be required. As an example of a multiply modified peptide, Erak et al (2018) provide a schematic of semaglutide (see Fig. 1 of Erak et al., 2018).

Synthetic strategies from medicinal chemistry are routinely employed to accomplish these improvements in therapeutic peptide potency, pharmacokinetics, and pharmacodynamics (Muttenthaler et al., 2021). To increase bioavailability, chemical modifications are employed to increase peptide stability and reduce clearance (Made et al., 2014; Muttenthaler et al., 2021). For example, amino acid modifications (e.g., N-terminal acetylation, N-terminal methylation, D-amino acids, unnatural amino acids, amide bond mimetics) can be incorporated at identified cleavage sites (Muttenthaler et al., 2021). To improve other pharmacokinetic properties of therapeutic peptides, covalent conjugation methods (e.g., lipidation and PEGylation) can be applied (Made et al., 2014; Muttenthaler et al., 2021). Small peptides may be stabilized through the incorporation of D-amino acids or disulfide bond mimetics, or through N-terminal to C-terminal cyclization (Erak et al., 2018; Muttenthaler et al., 2021). To increase cell penetration and target affinity, α-helical stabilization (yielding ‘stapled’ peptides) can be employed (Muttenthaler et al., 2021). Made et al. (2014) provide a case study of the application of modifications to neuropeptide Y (NPY) to develop this peptide as a biologic (Made et al., 2014).

A short description of some of the major events in peptide therapeutics is provided below. In the 1950s, the hormones oxytocin and vasopressin became the first synthetically produced therapeutic peptides (Muttenthaler et al., 2021). At the time, LPPS of these peptides required months or years (Muttenthaler et al., 2021). Until the 1980’s, only peptide hormone agonists effective at low doses were considered commercially viable (Muttenthaler et al., 2021). To improve the stability of these hormones, peptide modifications (D amino acids, unnatural amino acids, N-terminal capping, deamination, extensions of termini, disulfide bond mimetics) were applied. These modifications ultimately led to the successful production of more stable analogs (e.g., desmopressin, terlipressin, carbetocin, and atosiban).

Somatostatin (growth hormone-inhibiting hormone) regulates the secretion of growth hormone, insulin, and several other hormones. The commercial viability of synthetic somatostatin was markedly improved by the application of medicinal chemistry (Muttenthaler et al., 2021). This 14 residue hormone (with a single disulfide bond) normally has a half-life of three minutes (Muttenthaler et al., 2021). In 1988, a potent analogue (octreotide) incorporating D-amino acids and a C-terminal amino alcohol extension was developed (Muttenthaler et al., 2021). This analog demonstrated increased metabolic stability and was subsequently released on the market for the treatment of tumors. After N-terminal to C-terminal cyclization, an even more stable analogue (pasireotide) was developed (Muttenthaler et al., 2021). This analogue has subsequently been used in peptide-receptor radionuclide therapy.

In the mid – late 1980s, two synthetic gonadotrophin-releasing hormones (GnRHs) were released on the market (Muttenthaler et al., 2021). Leuprolide and Goserelin both behave as superagonists (agonists that desensitize and downregulate receptors producing a therapeutic response similar to antagonist) at the pituitary gonadotrophin-releasing hormone receptor (GnRHR) (Muttenthaler et al., 2021). By blockading GnRH hormonal function, Leuprolide and Goserelin are useful in the treatment of hormone-responsive cancer (Muttenthaler et al., 2021). Subsequently, GnRH analogs incorporating unnatural amino acids that behave as direct antagonists (e.g., Cetrorelix and ganirelix) were synthesized and successfully released on the market (Muttenthaler et al., 2021).

Enfuvirtide is a HIV-entry inhibitor that must be administered twice daily (because of the short half-life). This linear 36-mer is comprised of L-amino acids (only) and has an acetylated N-terminus and a carboxy-terminal amide (Muttenthaler et al., 2021). To achieve the required scale, Enfuvirtide was produced by solution phase fragment condensation from three side chain protected intermediates (synthesized on chlorotrityl resin) (Made et al., 2014; Muttenthaler et al., 2021). In total, this kg-scale production of Enfuvirtide required 106 steps (Muttenthaler et al., 2021). Despite the many steps and high costs, this synthesis resulted in high yields, and was significantly more time efficient (Made et al., 2014; Muttenthaler et al., 2021).

Recently, several ‘stapled’ peptides have been developed for the pharmaceutical market (Muttenthaler et al., 2021). This modification is designed to decrease proteolytic degradation, increase cell penetration, and increase target affinity. The stapled peptides ALRN-6924 and ALRN-5281 are both currently in clinical trials. ALRN-6924 is designed to reactivate p53 in tumors by inhibiting its major negative regulators. ALRN-5281 is designed to activate Gonadotropin-Releasing Hormone Receptor (GHRH) to increase growth hormone release. A list of selected peptides currently on the market/ in clinical trials along with the modifications that made them viable as therapeutics is provided by Erak et al. (see Table 2 of Erak et al., 2018).

REFERENCES

Albericio, F., Pons, M., Pedroso, E., Giralt, E. (1989) Comparative study of supports for solid-phase coupling of protected-peptide segments. The Journal of Organic Chemistry, 54, 360–366.

Albericio, F., Kneib-Cordonier, N., Biancalana, S., Gera, L., Masada, R., Hudson, D., & Bárány, G. (1990). Preparation and application of the 5-(4-(9-fluorenylmethyloxycarbonyl)aminomethyl-3,5-dimethoxyphenoxy)-valeric acid (PAL) handle for the solid-phase synthesis of C-terminal peptide amides under mild conditions. Journal of Organic Chemistry, 55, 3730–3743.

Albericio, F. (Ed.). (2000). Solid-Phase Synthesis: A Practical Guide (1st ed.). CRC Press. https://doi.org/10.1201/9781482270303

Angeletti, R. H., Bonewald, L. F., & Fields, G. B. (1997). Six-year study of peptide synthesis. Methods in enzymology, 289, 697–717. https://doi.org/10.1016/s0076-6879(97)89071-4

Atherton, E., Fox, H., Harkiss, D. Logan, C. J., Sheppard, R.C., & Williams, B. J. (1978). A mild procedure for solid phase peptide synthesis: use of fluorenylmethoxycarbonylamino-acids. Journal of the Chemical Society, Chemical Communications, 537–539.

Atherton, E., & Sheppard, RC. (1987). The fluorenylmethoxycarbonyl amino protecting group. In: Udenfriend, S.; Meienhofer, J., editors. The Peptides. Academic Press; New York p. 1–38.

Ayers, B., Blaschke, U. K., Camarero, J. A., Cotton, G. J., Holford, M., & Muir, T. W. (1999). Introduction of unnatural amino acids into proteins using expressed protein ligation. Biopolymers, 51(5), 343–354.https://doi.org/10.1002/(SICI)1097-0282(1999)51:5%3C343::AID-BIP4%3E3.0.CO;2-W

Barany, G., & Merrifield, RB. (1979) Solid-phase peptide synthesis. In: Gross, E.; Meienhofer, J., editors. The Peptides. Vol. 2. Academic Press; New York. p. 1–284.

Behrendt, R., White, P., & Offer, J. (2016). Advances in Fmoc solid-phase peptide synthesis. Journal of peptide science : an official publication of the European Peptide Society, 22(1), 4–27. https://doi.org/10.1002/psc.2836

Bergmann M., & Zervas L. (1932) Über ein allgemeines Verfahren der Peptid-Synthese. Berichte der deutschen chemischen Gesellschaft (A and B Series), 65, 1192–1201.

Bozovičar, K., & Bratkovič, T. (2019). Evolving a Peptide: Library Platforms and Diversification Strategies. International journal of molecular sciences, 21(1), 215. https://doi.org/10.3390/ijms21010215

Carpino, L.A., & Han, G.Y. (1970). 9-Fluorenylmethoxycarbonyl function, a new base-sensitive amino-protecting group. Journal of the American Chemical Society, 92, 5748–5749. DOI: 10.1021/ja00722a043

Chang, C. D., & Meienhofer, J. (1978). Solid-phase peptide synthesis using mild base cleavage of N alpha-fluorenylmethyloxycarbonylamino acids, exemplified by a synthesis of dihydrosomatostatin. International journal of peptide and protein research, 11(3), 246–249. https://doi.org/10.1111/j.1399-3011.1978.tb02845.x

Chumachenko, N., Novikov, Y., Shoemaker, R. K., & Copley, S. D. (2011). A dimethyl ketal-protected benzoin-based linker suitable for photolytic release of unprotected peptides. The Journal of organic chemistry, 76(22), 9409–9416. https://doi.org/10.1021/jo2017263

Dawson P. E. (1997). Synthesis of chemokines by native chemical ligation. Methods in enzymology, 287, 34–45. https://doi.org/10.1016/s0076-6879(97)87005-x

El-Faham, A., & Albericio, F. (2011). Peptide coupling reagents, more than a letter soup. Chemical reviews, 111(11), 6557–6602. https://doi.org/10.1021/cr100048w

Erak, M., Bellmann-Sickert, K., Els-Heindl, S., & Beck-Sickinger, A. G. (2018). Peptide chemistry toolbox – Transforming natural peptides into peptide therapeutics. Bioorganic & medicinal chemistry, 26(10), 2759–2765. https://doi.org/10.1016/j.bmc.2018.01.012

Felix, A. M., Heimer, E. P., Wang, C. T., Lambros, T. J., Fournier, A., Mowles, T. F., Maines, S., Campbell, R. M., Wegrzynski, B. B., & Toome, V. (1988). Synthesis, biological activity and conformational analysis of cyclic GRF analogs. International journal of peptide and protein research, 32(6), 441–454. https://doi.org/10.1111/j.1399-3011.1988.tb01375.x

Fields, C. G., Fields, G. B., Noble, R. L., & Cross, T. A. (1989). Solid phase peptide synthesis of 15N-gramicidins A, B, and C and high performance liquid chromatographic purification. International journal of peptide and protein research, 33(4), 298–303. https://doi.org/10.1111/j.1399-3011.1989.tb01285.x

Fields, G. B., & Noble, R. L. (1990). Solid phase peptide synthesis utilizing 9-fluorenylmethoxycarbonyl amino acids. International journal of peptide and protein research, 35(3), 161–214. https://doi.org/10.1111/j.1399-3011.1990.tb00939.x

Fields, G.B., Otteson, K.M., Fields, C.G., & Noble, RL. (1990). The versatility of solid phase peptide synthesis. In: Epton, R., Editor. Innovation and Perspectives in Solid Phase Synthesis. Solid Phase Conference Coordination, Ltd.; Birmingham, U.K. p. 241–260.

Fields, C.G., Fields, G.B. (1994). Solvents for solid-phase peptide synthesis. In: Pennington, MW.; Dunn, BM., Editors. Methods in Molecular Biology, 35. Peptide Synthesis Protocols. Humana Press, Inc.. Totowa, NJ. 29–40.

Fields, G.B. (1997). Editor. Methods In Enzymology. Academic Press; Orlando, FL. Solid-phase peptide synthesis. 289.

Fields, G.B., Lauer-Fields, J.L., Liu, R.-q., Barany, G. (2001). Principles and Practice of Solid-Phase Peptide Synthesis. In: Grant, GA., Editor. Synthetic Peptides: A User’s Guide. 2^nd^ Edition. W.H. Freeman & Co.. New York. 93–219.

Fischer, E., & Fourneau, E. (1903). Ueber einige Derivate des Glykocolls. Berichte der deutschen chemischen Gesellschaft, 34, 2868–2877.

Ford W.T., Balakrishnan T. (1981). Carbon-13 nuclear magnetic resonance relaxation in crosslinked polystyrene gels. Macromolecules, 14, 284–288.

Furka, A., Sebestyén, F., Asgedom, M., & Dibó, G. (1991). General method for rapid synthesis of multicomponent peptide mixtures. International journal of peptide and protein research, 37(6), 487–493. https://doi.org/10.1111/j.1399-3011.1991.tb00765.x

Gedye, R., Smith, F., Westaway, K., Ali, H., Baldisera, L., Laberge, L., Rousell, R. (1986). The use of microwave ovens for rapid organic synthesis. Tetrahedron Letters, 27(3), 279–282. https://doi.org/10.1016/S0040-4039(00)83996-9

Geysen, H. M., Meloen R. N., & Barlelin, S. J. (1984). Use of peptide synthesis to probe viral antigens for epitopes to a resolution of a single amino acid. Proc. Natl Acad. Sci. USA, 81, 3998–4002.

Giguere, R. J., Bray, T. L., Duncan, S. M., Majetich, G. (1986). Application of commercial microwave ovens to organic synthesis. Tetrahedron Letters, 27(41), 4945–4948. https://doi.org/10.1016/S0040-4039(00)85103-5

Góngora-Benítez, M., Tulla-Puche, J., & Albericio, F. (2013). Handles for Fmoc solid-phase synthesis of protected peptides. ACS combinatorial science, 15(5), 217–228. https://doi.org/10.1021/co300153c

Guillier, F., Orain, D., & Bradley, M. (2000). Linkers and cleavage strategies in solid-phase organic synthesis and combinatorial chemistry. Chemical reviews, 100(6), 2091–2158. https://doi.org/10.1021/cr980040+

Gutte, B., & Merrifield, R. B. (1969). The total synthesis of an enzyme with ribonuclease A activity. Journal of the American Chemical Society, 91(2), 501–502. https://doi.org/10.1021/ja01030a050

Gutte, B., & Merrifield, R. B. (1971). The synthesis of ribonuclease A. The Journal of biological chemistry, 246(6), 1922–1941.

Højlys-Larsen, K. B., & Jensen, K. J. (2013). Solid-phase synthesis of phosphopeptides. Methods in molecular biology, 1047, 191–199. https://doi.org/10.1007/978-1-62703-544-6_13

Hou, W., Zhang, X. & Liu, CF. (2017). Progress in Chemical Synthesis of Peptides and Proteins. Trans. Tianjin Univ. 23, 401–419. https://doi.org/10.1007/s12209-017-0068-8

Houghten, R. A. (1985). General method for the rapid solid-phase synthesis of large numbers of peptides: specificity of antigen-antibody interaction at the level of individual amino acids. Proc. Natl Acad. Sci. USA, 82(15) 5131–5135. https://doi.org/10.1073/pnas.82.15.5131

Houghten R. A. (1994). Combinatorial libraries. Finding the needle in the haystack. Current biology, 4(6), 564–567. https://doi.org/10.1016/s0960-9822(00)00127-5

Isidro-Llobet, A., Just-Baringo, X., Alvarez, M., & Albericio, F. (2008). EDOTn and MIM, new peptide backbone protecting groups. Biopolymers, 90(3), 444–449. https://doi.org/10.1002/bip.20823

Isidro-Llobet, A., Alvarez, M., & Albericio, F. (2009). Amino acid-protecting groups. Chemical reviews, 109(6), 2455–2504. https://doi.org/10.1021/cr800323s

Johnson, T., Quibell, M., Owen, D., & Sheppard, R. C. (1993) A reversible protecting group for the amide bond in peptides. Use in the synthesis of difficult sequences. J Chem Soc Chem Commun, 4, 369–372.

Johnson, T., Quibell, M., & Sheppard, R. C. (1995). N,O-bisFmoc derivatives of N-(2-hydroxy-4-methoxybenzyl)-amino acids: useful intermediates in peptide synthesis. Journal of peptide science, 1(1), 11–25. https://doi.org/10.1002/psc.310010104

Johnson, E. C., & Kent, S. B. (2006). Synthesis, stability and optimized photolytic cleavage of 4-methoxy-2-nitrobenzyl backbone-protected peptides. Chemical communications, 14, 1557–1559. https://doi.org/10.1039/b600304d

King, D. S., Fields, C. G., & Fields, G. B. (1990). A cleavage method which minimizes side reactions following Fmoc solid phase peptide synthesis. International journal of peptide and protein research, 36(3), 255–266. https://doi.org/10.1111/j.1399-3011.1990.tb00976.x

Kitas, E. A., Perich, J. W., Wade, J. D., Johns, R. B., & Tregear, G. W. (1989). FMOC-polyamide solid phase synthesis of an O-phosphotyrosine-containing tridecapeptide. Tetrahedron Letters, 30(45), 6229–6232. https://doi.org/10.1016/S0040-4039(01)93351-9

Lauer, J. L., Fields, C. G. & Fields, G. B. Sequence dependence of aspartimide formation during 9-fluorenylmethoxycarbonyl solid-phase peptide synthesis. (1995). Lett Pept Sci, 1, 197–205. https://doi.org/10.1007/BF00117955

Li, W., O’Brien-Simpson, N. M., Hossain, M. A., & Wade, J. D. (2020). The 9-Fluorenylmethoxycarbonyl (Fmoc) Group in Chemical Peptide Synthesis – Its Past, Present, and Future*. Aust J Chem*, 73, 271–276. https://doi.org/10.1071/CH19427

Liu, C. F., & Tam, J. P. (1994). Peptide segment ligation strategy without use of protecting groups. Proceedings of the National Academy of Sciences of the United States of America, 91(14), 6584–6588. https://doi.org/10.1073/pnas.91.14.6584

Live, D. & Kent, S. B. H. (1982) Fundamental aspects of the chemical applications of cross-linked polymers, in Elastomers and Rubber Elasticity (Mark, J. E., ed), American Chemical Society, Washington, DC, pp. 501–515.

Ludwick, A. G., Jelinski, L. W., Live, D., Kintanar, A., & Dumais, J. J. (1986) Association of peptide chains during Merrifield solid-phase peptide synthesis· a deuterium NMR study J Am Chem Soc, 108, 6493–6496.

Lukas, T. J., Prystowsky, M. B., & Erickson, B. W. (1981). Solid-phase peptide synthesis under continuous-flow conditions. Proceedings of the National Academy of Sciences, 78,2791–2795.

Mäde, V.; Els-Heindl, S.; & Beck-Sickinger, A. G. (2014). Beilstein J Org Chem, 10, 1197–1212. https://doi.org/10.3762/bjoc.10.118

Mamidi, S., Mayer, G., Hurevich, M., & Friedler, A. (2020). Multi phosphorylated peptides: Importance, Synthetic Strategies, and applications for studying biological mechanisms. Organic & Biomolecular Chemistry, 18. 3405–3422.

Manatt, S. L., Amsden, C. F., Bettison, C. A., Frazer, W. T., Gudman, J. T., Lenk, B. E., Lubetich, J. F., McNelly, E. A., Smith, S. C., Templeton, D. J., & Pinnell R. P. (1980). A fluorine-19 NMR approach for studying Merrifield solid-phase peptide syntheses. Tetrahedron Letters, 21, 1397–1400. https://doi.org/10.1016/S0040-4039(00)92729-1.

Merrifield, R. B. (1963). Solid phase peptide synthesis I: Synthesis of a tetrapeptide. J Am Chem Soc, 85, 2149–2154.

Merrifield, R. B. (1964). Solid-phase peptide synthesis. 3. An improved synthesis of bradykinin. Biochemistry, 3, 1385–1390. https://doi.org/10.1021/bi00897a032

Merrifield, R. B., & Stewart, J. M. (1965). Automated peptide synthesis. Nature, 207(996), 522–523. https://doi.org/10.1038/207522a0

Merrifield, R. B., Stewart, J. M., & Jernberg, N. (1966). Instrument for automated synthesis of peptides. Analytical chemistry, 38(13), 1905–1914. https://doi.org/10.1021/ac50155a057

Merrifield, B. (1986). Solid phase synthesis. Science (New York, N.Y.), 232(4748), 341–347. https://doi.org/10.1126/science.3961484

Mueller, L. K., Baumruck, A. C., Zhdanova, H., & Tietze A., A. (2020). Challenges and Perspectives in Chemical Synthesis of Highly Hydrophobic Peptides. Frontiers in Bioengineering and Biotechnology, 8, 162–178. https://doi.org/10.3389/fbioe.2020.00162

Muir, T. W., Dawson, P. E., & Kent, S. B. (1997). Protein synthesis by chemical ligation of unprotected peptides in aqueous solution. Methods in enzymology, 289, 266–298. https://doi.org/10.1016/s0076-6879(97)89052-0

Muttenthaler, M., King, G. F., Adams, D. J., & Alewood, P. F. (2021). Trends in peptide drug discovery. Nature reviews. Drug discovery, 20(4), 309–325. https://doi.org/10.1038/s41573-020-00135-8

Mutter, M., Nefzi, A., Sato, T., Sun, X., Wahl, F., & Wöhr, T. (1995). Pseudo-prolines (psi Pro) for accessing “inaccessible” peptides. Peptide research, 8(3), 145–153.

Okada, Y., Takasawa, R., Kubo, D., Iwanaga, N., Fujita, S., Suzuki, K., Suzuki, H., Kamiya, H., & Chiba, K. (2019). Improved Tag-Assisted Liquid-Phase Peptide Synthesis: Application to the Synthesis of the Bradykinin Receptor Antagonist Icatibant Acetate. Organic Process Research & Development, 23(11), 2576–2581. https:// doi.org/10.1021/acs.oprd.9b00397

Ostresh, J. M., Winkle, J. H., Hamashin, V. T., & Houghten, R. A. (1994). Peptide libraries: determination of relative reaction rates of protected amino acids in competitive couplings. Biopolymers, 34(12), 1681–1689. https://doi.org/10.1002/bip.360341212

Ottinger, E. A., Shekels, L. L., Bernlohr, D. A., & Barany, G. (1993). Synthesis of phosphotyrosine-containing peptides and their use as substrates for protein tyrosine phosphatases. Biochemistry, 32(16), 4354–4361. https://doi.org/10.1021/bi00067a027

Otvos, L., Jr, Elekes, I., & Lee, V. M. (1989). Solid-phase synthesis of phosphopeptides. International journal of peptide and protein research, 34(2), 129–133. https://doi.org/10.1111/j.1399-3011.1989.tb01501.x

Otvos, L., Jr, Wroblewski, K., Kollat, E., Perczel, A., Hollosi, M., Fasman, G. D., Ertl, H. C., & Thurin, J. (1989). Coupling strategies in solid-phase synthesis of glycopeptides. Peptide research, 2(6), 362–366.

Paradís-Bas, M., Tulla-Puche, J., & Albericio, F. (2016). The road to the synthesis of “difficult peptides”. Chemical Society reviews, 45(3), 631–654. https://doi.org/10.1039/c5cs00680e

Pedersen, S. L., & Jensen, K. J. (2013). Instruments for automated peptide synthesis. Methods in molecular biology, 1047, 215–224. https://doi.org/10.1007/978-1-62703-544-6_15

Perich, JW. (1997). Synthesis of phosphopeptides using modern chemical approaches. In: Gregg, BF., editor. Methods in Enzymology. Academic Press, 245–266.

Perich, J. W., & Reynolds, E. C. (1991). Fmoc/solid-phase synthesis of Tyr(P)-containing peptides through t-butyl phosphate protection. International Journal of Peptide and Protein Research. 37, 572–575.Perich, J. W., Ede, N. J., Eagle, S., Bray, A M. (1999). Synthesis of phosphopeptides by the Multipinast method: Evaluation of coupling methods for the incorporation of Fmoc- Tyr(PO3Bzl,H)-OH, FmocSer(PO3Bzl,H)-OH and Fmoc- Thr(PO3Bzl,H)-OH. Letters in Peptide Science, 6, 91–97.

Peukert, S., & Giese, B. (1998). The Pivaloylglycol Anchor Group: A New Platform for a Photolabile Linker in Solid-Phase Synthesis. Journal of Organic Chemistry, 63, 9045–9051.

Plaué S. (1990). Synthesis of cyclic peptides on solid support. Application to analogs of hemagglutinin of influenza virus. International journal of peptide and protein research, 35(6), 510–517. https://doi.org/10.1111/j.1399-3011.1990.tb00255.x

Pontiroli A. E. (1998). Peptide hormones: Review of current and emerging uses by nasal delivery. Advanced drug delivery reviews, 29(1-2), 81–87. https://doi.org/10.1016/s0169-409x(97)00062-8

Quibell, M., Owen, D., Packman, L., & Johnson, T. (1994). Suppression of piperidine-mediated side product formation for Asp(OBut)-containing peptides by the use of N-(2-hydroxy-4-methoxybenzyl)(Hmb) backbone amide protection. Journal of The Chemical Society, Chemical Communications, 2343–2344.

Reily, C., Stewart, T. J., Renfrow, M. B., & Novak, J. (2019). Glycosylation in health and disease. Nature reviews. Nephrology, 15(6), 346–366. https://doi.org/10.1038/s41581-019-0129-4

Saladin, P. M., Zhang, B. D., & Reichert, J. M. (2009). Current trends in the clinical development of peptide therapeutics. IDrugs: the investigational drugs journal, 12(12), 779–784.

Sakakibara, S., Shimonishi, Y., Kishida, Y., Okada, M., & Sugihara, H. (1967). Use of anhydrous hydrogen fluoride in peptide synthesis. I. Behavior of various protective groups in anhydrous hydrogen fluoride. Bulletin of the Chemical Society of Japan, 40(9), 2164–2167. https://doi.org/10.1246/bcsj.40.2164

Sakakibara, S., Kishida, Y., Nishizawa, R., & Shimonishi, Y. (1968). Use of anhydrous hydrogen fluoride in peptide synthesis. II. Procedures for the syntheses of simple peptides. Bulletin of the Chemical Society of Japan, 41(2), 438–441. https://doi.org/10.1246/bcsj.41.438

Shelton, P. T., & Jensen, K. J. (2013). Linkers, resins, and general procedures for solid-phase peptide synthesis. Methods in molecular biology, 1047, 23–41. https://doi.org/10.1007/978-1-62703-544-6_2

Shin, Y., Winans, K. A., Backes, B. J., Kent, S. B. H, Ellman, J. A., & Bertozzi, C. R. (1999). Fmoc-Based Synthesis of Peptide-αThioesters: Application to the Total Chemical Synthesis of a Glycoprotein by Native Chemical Ligation. Journal of the American Chemical Society, 121(50), 11684–11689. https://doi.org/10.1021/ja992881j

Sjölin, P., Elofsson, M., & Kihlberg, J. (1996). Removal of Acyl Protective Groups from Glycopeptides: Base Does Not Epimerize Peptide Stereocenters, and beta-Elimination Is Slow. The Journal of organic chemistry, 61(2), 560–565. https://doi.org/10.1021/jo951817r

Sole, N. A, & Barany, G. (1992). Optimization of solid-phase synthesis of [Ala8]-dynorphin A. The Journal of Organic Chemistry, 57(20), 5399-5403. https://doi.org/10.1021/jo00046a022

Stawikowski, M., & Cudic, P. (2006). A novel strategy for the solid-phase synthesis of cyclic lipodepsipeptides. Tetrahedron letters, 47(48), 8587–8590. https://doi.org/10.1016/j.tetlet.2006.09.116

Stawikowski, M. & Fields, G. B, (2009). Published in final edited form as: Fields G. B. (2002). Introduction to peptide synthesis. Current protocols in protein science, Chapter 18, Unit–18.1. https://doi.org/10.1002/0471140864.ps1801s26

Tam, J. P, Wu, C-R., Liu, W., & Zhang, J-W. (1991). Disulfide bond formation in peptides by dimethyl sulfoxide. Scope and applications. Journal of the American Chemical Society, 113(17), 6657–6662.

Tam, J. P., Lu, Y-A. (1995). Coupling difficulty associated with interchain clustering and phase transition in solid phase peptide synthesis. J Am Chem Soc, 117, 12058–12063.Tam, J. P., Lu, Y. A., Liu, C. F., & Shao, J. (1995). Peptide synthesis using unprotected peptides through orthogonal coupling methods. Proceedings of the National Academy of Sciences of the United States of America, 92(26), 12485–12489. https://doi.org/10.1073/pnas.92.26.12485

Tickler, A. K., Barrow, C. J., & Wade, J. D. (2001). Improved preparation of amyloid-beta peptides using DBU as Nalpha-Fmoc deprotection reagent. Journal of peptide science, 7(9), 488–494. https://doi.org/10.1002/psc.342

Tickler, A. K., & Wade, J. D. (2007). Overview of solid phase synthesis of “difficult peptide” sequences. Current protocols in protein science, Chapter 18. https://doi.org/10.1002/0471140864.ps1808s50

Du Vigneaud, V., Ressler, C., Swan, J. M., Roberts, C. W., Katsoyannis, P. G., Gordon, S. (1953). The synthesis of an octapeptide amide with the hormonal activity of oxytocin, Journal of the American Chemical Society, 75, 4879–4880.

Wakamiya, T., Saruta, K., Yasuoka, J-I., & Kusumoto, S. (1994). An Efficient Procedure for Solid-Phase Synthesis of Phosphopeptides by the Fmoc Strategy. Chemistry Letters, 23(6), 1099–1102. https://doi.org/10.1246/cl.1994.1099

Weygand, F., Prox, A., & König, W. (1966). Racemisierung der vorletzten carboxylendständigen Aminosäure bei Peptidsynthesen [Racemization of the next to last carboxyl terminal amino acid in peptide synthesis]. Chemische Berichte, 99(5), 1146–1150.

Yeo, J., Peeva, L., Chung, S., Gaffney, P., Kim, D., Luciani, C., Tsukanov, S., Seibert, K., Kopach, M., Albericio, F., & Livingston, A. (2021). Liquid Phase Peptide Synthesis via One-Pot Nanostar Sieving (PEPSTAR). Angewandte Chemie (International ed. in English), 60(14), 7786–7795. https://doi.org/10.1002/anie.202014445

Zahariev, S., Guarnaccia, C., Pongor, C. I., Quaroni, L., Cemazar, M., & Pongor, S. (2006). Synthesis of ‘difficult’ peptides free of aspartimide and related products, using peptoid methodology, Tetrahedron Letters, 47(25), 4121–4124. https://doi.org/10.1016/j.tetlet.2006.04.074.

Zardeneta, G., Chen, D., Weintraub, S. T., & Klebe, R. J. (1990). Synthesis of phosphotyrosyl-containing phosphopeptides by solid-phase peptide synthesis. Analytical Biochemistry, 190, 340–347.