Peptide synthesis

A peptide is a short chain of amino acids joined by amide (peptide) bonds, formed in a condensation reaction that loses water between the -\text{COOH} of one amino acid and the -\text{NH}_2 of the next.

Making a single, specific dipeptide is harder than it looks. Heat glycine with alanine and you can get Gly-Ala, Ala-Gly, Gly-Gly, Ala-Ala and longer chains, because both molecules carry an amine and an acid. The fix is protecting groups: temporarily convert the groups you do not want to react into something unreactive.

To make Gly-Ala only:

• protect the amine of glycine (for example as a carbamate);
• protect the acid of alanine (for example as a methyl ester);
• activate glycine's free acid so it reacts with alanine's free amine at room temperature;
• remove both protecting groups.

A carboxylic acid and an amine on their own just form a salt, which is why an activating step is needed. In the lab this might be conversion to an acyl chloride, though real peptide chemists use gentler coupling reagents that do not damage other groups.

All amino acids except glycine are chiral, and natural ones are the L form. A good method must not scramble that stereocentre, otherwise the product is a mixture of diastereomers that may not fold or work.

Hydrolysis reverses the process: boiling a peptide with 6 mol dm^{-3} HCl breaks every amide bond and releases the amino acids, which can then be separated by chromatography and identified.

Activation and coupling

The carboxylic acid must be turned into an activated acyl species. Carbodiimides such as DIC (N,N'-diisopropylcarbodiimide) are attacked by the carboxylate to give an O-acylisourea, a very reactive ester. Left alone it can rearrange to an unreactive N-acylurea, so an additive such as HOBt or ethyl cyano(hydroxyimino)acetate (Oxyma) is added. It intercepts the O-acylisourea to give an active ester that is reactive enough to acylate the amine but less prone to side reactions. Uronium/aminium reagents (HBTU, HATU, COMU) and phosphonium reagents (PyBOP) generate similar active esters directly with a tertiary amine base; HATU, built on HOAt, is the usual choice for hindered couplings.

Racemisation

The main stereochemical danger comes from the activated residue. If its nitrogen carries an acyl group (as it does when a peptide segment is activated), the amide oxygen can attack the activated carbonyl to form a 5(4H)-oxazolone. The α-proton of the oxazolone is acidic because the anion is aromatic, so it is lost and regained from either face and the stereocentre is scrambled. Urethane protecting groups (Boc, Fmoc) make oxazolone formation much less favourable, which is why stepwise chain growth from the C-terminus, one urethane-protected residue at a time, keeps chirality intact. Cysteine and histidine remain the residues most prone to epimerisation.

Solid-phase synthesis (SPPS)

In Merrifield's method the C-terminal residue is attached to a resin (cross-linked polystyrene or PEG-based) and the chain grows towards the N-terminus. Each cycle is: remove the N-terminal protecting group, wash, couple the next protected residue with excess reagent, wash.

• Fmoc/tBu: Fmoc is removed by 20% piperidine in DMF through E1cB loss of dibenzofulvene; side chains carry acid-labile groups (tBu, Boc, trityl, Pbf) that come off with the resin linker in TFA plus scavengers such as triisopropylsilane and water. The two protecting group sets are orthogonal.
• Boc/Bzl: Boc is removed by TFA each cycle; final cleavage needs anhydrous HF. Less convenient, but useful for aggregation-prone sequences.

Yields compound harshly. At 99% per coupling a 30-mer is obtained in about 0.99^{29} \approx 75\%; at 97% it falls to about 41%, before any purification by reverse-phase HPLC.

Stepwise SPPS runs into trouble beyond roughly 50 residues, mainly because growing chains aggregate on the resin through β-sheet hydrogen bonding, so couplings and deprotections stall. Standard counter-measures are pseudoproline dipeptides and backbone protection (Hmb/Dmb), which install temporary kinks, plus elevated temperature or microwave heating. Automated fast-flow synthesisers now couple a residue in well under a minute at high temperature, and have produced chains of over 150 residues in a single run.

For whole proteins, native chemical ligation (introduced by Dawson and Kent in 1994) joins unprotected segments in water. A C-terminal thioester undergoes reversible transthioesterification with the thiol of an N-terminal cysteine on the second segment; an irreversible S-to-N acyl shift then forms a native amide bond. Peptide hydrazides and N-acylurea (Dbz) linkers are now common thioester surrogates, and radical or metal-free desulfurisation converts the ligation cysteine to alanine, so the junction no longer needs a native Cys. Serine/threonine ligation and KAHA ligation extend the choice of junction residues. These methods have delivered mirror-image enzymes, homogeneous glycoproteins and proteins carrying defined post-translational modifications such as site-specific ubiquitination.

The field's sustainability problem is real: SPPS consumes large volumes of DMF and NMP, both now restricted under REACH, so greener solvents (for example N-butylpyrrolidinone and binary mixtures) and reduced-wash protocols are active research topics. Large-scale manufacture of GLP-1 agonists has driven hybrid approaches that combine SPPS fragments, solution-phase coupling and, increasingly, enzymatic ligation with engineered peptide ligases.