What is a peptide bond? Formation, structure, hydrolysis and laboratory synthesis

A peptide bond is the link that holds amino acids together in every peptide and protein. Chemically it is an amide: the carboxyl carbon of one amino acid bonded to the amino nitrogen of the next. That single bond explains a great deal, from why proteins fold the way they do to why making a 39-residue peptide in a factory takes dozens of carefully controlled chemical steps.

This guide covers what a peptide bond is, how it forms, why it is flat, the cis and trans forms, how it is broken, how ribosomes make it, and how chemists make it with coupling reagents and protecting groups. For the bigger picture of what peptides are, see what is a peptide?

What is a peptide bond?

IUPAC defines a peptide as any compound produced by amide formation between a carboxyl group of one amino acid and an amino group of another, and states that “the amide bonds in peptides may be called peptide bonds” [1]. The name goes back to Emil Fischer, who first proposed this structure and called the amide link a peptide bond [2].

In standard peptides and proteins the bond joins the alpha-carboxyl carbon (C-1) of one residue to the alpha-amino nitrogen of the next. IUPAC’s definition is wider: the word peptide also covers compounds whose residues are linked by other amide bonds [1], such as the side-chain (gamma) link in glutathione.

Feature	Peptide bond
Bond type	Amide (C(=O)-NH), covalent [1]
Atoms joined	Carbonyl carbon of residue n and amide nitrogen of residue n+1
Formed by	Condensation (dehydration synthesis), releasing one H2O [3]
Broken by	Hydrolysis (adding H2O back) [4]
Geometry	Planar peptide group; N–H oriented 180° to C=O [5]
Preferred configuration	Trans for most residues [6]
Made in cells by	Peptidyl transferase, an RNA-based catalytic centre of the ribosome [7] [8]
Made in the laboratory by	Activating the carboxyl group with a coupling reagent [9]

How is a peptide bond formed?

Peptide bond formation is a dehydration synthesis reaction. In the words of OpenStax’s biology text, one amino acid’s carboxyl group and the incoming amino acid’s amino group combine, releasing a water molecule [3]. Chemists call the same process a condensation reaction.

Peptide bond formation diagram

  Amino acid 1              Amino acid 2

        R1                        R2
        |                         |
  H2N───C───COOH      +     H2N───C───COOH
        |                         |
        H                         H

                     │  condensation
                     │  (one H2O released)
                     ▼

        R1    O                R2
        |     ‖                |
  H2N───C─────C─────N──────────C───COOH     +     H2O
        |           |          |
        H           H          H
              └──┬──┘
                 │
          peptide bond, C(=O)─N(H)

  N-terminus (free NH2) on the left, C-terminus (free COOH) on the right

The product is a dipeptide. It still has a free amino group at one end, the N-terminus, and a free carboxyl group at the other, the C-terminus [3], so the chain can keep growing. By convention, peptides are written with the N-terminal amino acid on the left [5]. A chain of n residues contains n minus 1 peptide bonds and has released n minus 1 molecules of water.

Why mixing two amino acids is not enough

Simply mixing a carboxylic acid with an amine is a poor way to make an amide. Amines are bases that convert acidic carboxyl groups into unreactive carboxylate anions, so amides are difficult to prepare by direct reaction [9]. Cells and chemists both solve this the same way in principle: they first convert the carboxyl group into a more reactive, “activated” form.

Water also matters. A vapour-phase analysis found that the peptide bond is stabilised by solvent water to an extreme degree among uncharged functional groups, and that the equilibria of peptide bond synthesis and hydrolysis in water are largely set by differences in solvation free energy between reactants and products [10].

Peptide bond structure: planarity and resonance

The amide nitrogen is not basic, because its lone pair of electrons is delocalised by resonance with the carbonyl group [5]. The result is that the C–N bond behaves partly like a double bond. OpenStax states it directly: the amide bond is therefore planar, and the N–H is oriented 180° to the C=O [5].

  Resonance forms of the peptide group

     O                         O(−)
     ‖                         |
  ───C─────N───     <──>   ───C═════N(+)───
           |                         |
           H                         H

  Partial C═N double-bond character → restricted rotation → planar group

Quantum-chemical calculations support this resonance picture. A 2007 study computed C–N rotation barriers for acetamide and eight related compounds, found that the barriers span nearly 50 kcal/mol, and showed that they correlate well with the calculated weight of the amide resonance structure [11]. In other words, the more resonance, the harder the bond is to twist.

Planarity is the reason protein backbones have a limited set of shapes. Because the peptide group is rigid, a chain can only rotate freely around the two single bonds either side of each alpha carbon, conventionally called phi (ϕ) and psi (ψ). Backbone geometry also underlies the classic secondary structures: in 1951 Pauling, Corey and Branson proposed two hydrogen-bonded helical configurations of the polypeptide chain [12].

Real peptide bonds are not perfectly flat. An analysis of atomic-resolution protein structures found that trans peptide groups can vary by more than 25° from planarity, that the expected value shifts by about ±8° as a systematic function of the ϕ,ψ angles, and that most of the variation depends on tertiary structure [13].

Cis and trans peptide bonds

Because rotation around the C–N bond is restricted, each peptide bond is locked into one of two configurations. In the trans form the alpha carbons of the two linked residues lie on opposite sides of the bond; in the cis form they lie on the same side.

Bond type	Usual configuration	What the evidence shows
X-non-Pro (most peptide bonds)	Trans	Trans is generally more favourable for most amino acids [6]. A survey of 571 proteins found only 43 non-proline cis peptide bonds [14]
X-Pro (bond before proline)	Mostly trans, cis common	The free-energy difference between cis and trans is small for proline [6]
Isomerisation catalysts	Peptidyl-prolyl isomerases (cyclophilins, FKBPs, parvulins)	They catalyse cis-trans isomerisation of the peptide bond preceding proline [6]

Non-proline cis bonds, though rare, tend to sit in functionally important regions, often near active sites, and are found more often in high-resolution structures, which suggests some are missed at lower resolution [14]. Cis-trans isomerisation of prolines can be the critical, slow step in the refolding of some proteins, as reported for human creatine kinase [6].

Is a peptide bond strong? Hydrolysis of the peptide bond

Peptide bonds are covalent and kinetically robust. Amides are the least reactive of the common carboxylic acid derivatives, and in the laboratory they are hydrolysed to a carboxylic acid plus an amine by heating in aqueous acid or aqueous base [4].

The rate is not identical for every bond. Model studies extrapolated to neutral solution at 25 °C found that the side-chain thiol of N-acetylcysteine speeds hydrolysis of its amide bond by a factor of 70, and the hydroxyl of N-acetylserine by a factor of 12, compared with N-acetylalanine [15]. The authors attribute this to intramolecular acid-base catalysis and describe bonds to cysteine and serine nitrogens as “weak points” in protein backbones [15].

Enzymatic hydrolysis

In living systems, protease and peptidase enzymes catalyse peptide bond hydrolysis with high selectivity. A pharmacologically important example is dipeptidyl peptidase-4 (DPP-4), which cleaves the N-terminal dipeptide from GLP-1(7-36)amide and GLP-1(7-37), producing low-affinity metabolites; together with renal elimination this gives native GLP-1 a half-life of around 1 to 2 minutes [16]. Drug designers protect such bonds with non-natural residues, as described in our guide to GLP-1 receptor agonists.

Ways a peptide bond is broken	Conditions	Example
Chemical hydrolysis	Heating in aqueous acid or base [4]	Laboratory hydrolysis of amides to acid plus amine
Neighbouring-group assisted hydrolysis	Neutral solution; side-chain -SH or -OH assists [15]	Bonds to Cys or Ser nitrogens
Enzymatic hydrolysis	Physiological; protease or peptidase [16]	DPP-4 cleavage of GLP-1

How ribosomes make peptide bonds

Cells pay for peptide bonds in advance. Aminoacyl-tRNA synthetases first bind and hydrolyse ATP to attach each amino acid to its tRNA, and the energy for each peptide bond comes from that high-energy amino acid to tRNA bond [7]. GTP is spent on delivering each new aminoacyl-tRNA to the ribosome’s A site [7].

The bond is then formed by peptidyl transferase, an RNA-based enzyme [7]. The 2000 atomic structures of the large ribosomal subunit established that the ribosome is a ribozyme: the substrate analogues are contacted only by conserved ribosomal RNA, and no protein side-chain atom lies closer than about 18 ångströms to the peptide bond being synthesised [8]. The authors noted that the mechanism appears to resemble the reverse of the acylation step in serine proteases [8]. In E. coli, the translation machinery adds an amino acid about every 0.05 seconds [7].

How chemists form peptide bonds

Laboratory peptide synthesis has two problems to solve: activating the carboxyl group, and making sure only the intended carboxyl and amino groups react. OpenStax gives a small example: five steps are needed to synthesise a dipeptide such as Ala-Leu [17]: protect, protect, couple, deprotect, deprotect.

Step 1: protecting groups (Fmoc and Boc)

The amino group of the incoming amino acid is temporarily blocked so it cannot react with itself. The two dominant temporary protecting groups are:

Protecting group	Full name	Removed by	Used in
Boc	tert-butyloxycarbonyl	Brief treatment with a strong acid such as trifluoroacetic acid (TFA) [17]	Classical Boc SPPS, now generally used only for specialist applications; Fmoc chemistry avoids anhydrous HF [18]
Fmoc	9-fluorenylmethoxycarbonyl	20% piperidine in dimethylformamide [17]	Fmoc/tBu SPPS, the method of choice today [18]

Fmoc chemistry pairs a base-labile temporary group with acid-labile side-chain groups, giving an orthogonal protection scheme, and the majority of synthetic peptides are now prepared by Fmoc SPPS [18]. Fmoc and Boc both belong to the urethane (carbamate) family of protecting groups, which suppress racemisation during activation and coupling [18]. Side chains carry their own “permanent” protection, such as Pbf on arginine, Trt on asparagine, glutamine, cysteine and histidine, OtBu on aspartate and glutamate, and tBu on serine, threonine and tyrosine [18].

Step 2: activation with coupling reagents

A coupling reagent turns the free carboxyl group into a reactive intermediate that the amine can attack. The field is famously crowded: reviews describe “a myriad of so-called coupling reagents” and “acronym”-based reagents [19], and a 2011 Chemical Reviews survey is titled “Peptide coupling reagents, more than a letter soup” [20].

Reagent family	Examples	How it works	Notes from the literature
Carbodiimides	DCC, DIC, EDC	The acid adds across the N=C=N group of the carbodiimide; the amine then attacks the activated acyl group, releasing a urea (dicyclohexylurea from DCC) [9]	DCC activation gives a highly reactive O-acylisourea that can form an oxazolone and racemise [2]
Additives used with carbodiimides	HOBt, HOAt, Oxyma	Convert the reactive intermediate into an active ester	HOBt suppresses epimerisation with DCC or EDC [2]; Oxyma inhibits racemisation and was developed because HOBt and HOAt were reported to have explosive properties [21]
Aminium/uronium salts	HBTU, HCTU, HATU	Activate the acid in the presence of a tertiary amine base such as DIPEA [18]	HATU was used for couplings involving N-methyl amino acids in one reviewed synthesis [2]; base-mediated activation can racemise cysteine [18]
Phosphonium salts	BOP, PyBOP	Base-mediated activation via an active ester [2]	Also linked to cysteine racemisation under basic activation [18]
HOAt-derived phosphorus esters	Diethyl and diphenyl phosphoryl esters of HOAt	Fast-acting active esters	Reported to minimise loss of configuration in segment coupling compared with uronium analogues [22]

The carbodiimide method is, as OpenStax puts it, the key step in the laboratory synthesis of small proteins [9].

Step 3: repeat on a solid support

Robert Bruce Merrifield won the 1984 Nobel Prize in Chemistry for developing chemical synthesis on a solid matrix [23]. In solid-phase peptide synthesis (SPPS) the C-terminal residue is anchored to resin, and each cycle removes the Fmoc group, washes, couples the next protected amino acid and washes again, so excess reagent can be used and simply rinsed away. The chain is finally cleaved from the resin and the side chains deprotected, usually with TFA in Fmoc chemistry. Our guide to SPPS vs LPPS vs hybrid compares the manufacturing routes in detail.

Racemisation: the side reaction that matters most

Amino acids in natural peptides have the L-configuration at the alpha carbon. During activation, that stereocentre can be lost. A 2023 review describes two base-catalysed pathways: direct enolisation (path A) and formation of a 5(4H)-oxazolone intermediate (path B), with the oxazolone route regarded as the predominant source of racemisation during peptide synthesis [2]. The resulting epimers have almost identical physical properties, which makes them difficult to remove by purification [2].

Some residues are more at risk than others:

• Cysteine racemises with base-mediated activation such as phosphonium or uronium reagents; in one model peptide, racemisation with the standard trityl group rose from 8.0% under conventional conditions to 26.6% at 80 °C [18].
• Histidine has historically been notoriously racemisation-prone, because its imidazole π-nitrogen promotes enolisation of the active ester [18].
• Coupling reagent choice matters: in one solid-phase study of Gly-Phe-Pro-NH2, epimerisation was 29.8% with EDC/HOAt but 4.2% with DIC/HOAt [2].

The standard defences are urethane protecting groups, racemisation-suppressing additives (HOBt, HOAt, Oxyma), careful control of base, and protecting groups designed for difficult residues [2] [21] [18].

Aspartimide formation

The most serious side reaction in Fmoc chemistry is aspartimide formation at aspartic acid [18]. It can generate nine different by-products, some of which co-elute with the target peptide, and it is worst in Asp-Gly sequences [18]. Adding Oxyma to the piperidine deprotection solution has been shown to reduce aspartimide-related impurities [18]. Impurities like these are why peptide specifications set limits on related substances; see peptide API impurities and specifications.

Peptide bond vs other bonds in peptides

Bond	Between	Covalent?	Role
Peptide (amide) bond	Carbonyl C and amide N of adjacent residues [1]	Yes	Backbone; sets primary structure
Disulfide bond	Two cysteine side chains	Yes	Cross-links chains or loops, as in insulin’s two chains [5]
Hydrogen bond	Backbone C=O and N-H, or side chains	No	Stabilises secondary structure such as helices [12]
Side-chain amide (isopeptide)	Side-chain carboxyl or amino group	Yes	Covered by IUPAC’s wider definition of peptide [1]

How Winhycare Pharma makes peptides

Winhycare Pharma supplies research peptides made by solid-phase synthesis as research-use-only material, with purity of ≥99% as stated in our product brochure, and offers custom peptide synthesis for research sequences. Semaglutide, tirzepatide, retatrutide, cagrilintide and BPC-157 are also supplied as API to licensed manufacturers through our peptide API hub; US-DMF and CTD Module 3 support is available for semaglutide only. To discuss a sequence, use the research peptide enquiry form.

For laboratory research use only. Not for human or veterinary use.