Biochemistry
Peptide Bond
Amide linkage with partial C=N character — planar, ~40% rotation barrier, trans favored 1000:1 over cis
A peptide bond is the amide linkage joining the α-carboxyl of one amino acid to the α-amino of the next, formed with loss of water. Resonance delocalization gives the C–N bond ~40% double-bond character, locking the six atoms of the peptide unit (Cα, C, O, N, H, Cα) into a plane. The rotation barrier is ~88 kJ/mol and trans is favored over cis by roughly 1000:1 except at proline where the ratio drops to ~4:1. Linus Pauling proved peptide planarity in 1951 from X-ray diffraction of small amides, the structural insight that founded modern protein crystallography.
- Bond typeAmide (CO–NH)
- C–N character~40% double bond
- Rotation barrier~88 kJ/mol
- Trans:cis1000:1 (4:1 at Pro)
- Planarity provedPauling, 1951
- Hydrolysis t½~500 yr at pH 7, 25°C
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
Why the peptide bond matters
- Planarity reduces protein conformation to two angles per residue. Locking ω near 180° leaves only φ and ψ around Cα as free dihedrals. Ramachandran's 1963 plot exploits this to map allowed conformations in two dimensions — the ~88 kJ/mol rotation barrier is what makes protein structure tractable.
- The 40% double-bond character drives every secondary-structure pattern. α-helices and β-sheets rely on C=O and N–H groups projecting predictably out of the planar amide unit; both lose their definition once ω deviates by more than ~15°. Pauling and Corey's 1951 helix prediction worked because they assumed planarity.
- Hydrolysis half-life of ~500 years sets the protein-decay timescale. Spontaneous cleavage at pH 7 and 25 °C is so slow that only enzymes can break peptide bonds within useful time. Proteases enhance the rate by ~1012, the largest catalytic acceleration known in biology.
- Cis-prolines are real and consequential. ~6% of Xaa-Pro bonds in folded proteins are cis, often at active-site turns. Prolyl isomerases (cyclophilin, FKBP, parvulin) catalyze the slow cis-trans interconversion; their substrates include drug targets — cyclosporine and tacrolimus inhibit them therapeutically.
- Ribosomes synthesize peptide bonds at ~50 ms per residue. Translation runs at 15–20 amino acids per second in bacteria, ~5 per second in eukaryotes. Each bond costs 4 high-energy phosphates (~200 kJ/mol input) of which roughly 21 kJ/mol is captured in the bond itself; the rest pays for fidelity.
- Solid-phase peptide synthesis (Merrifield, Nobel 1984) industrialized chemical assembly. Activated carboxyl coupling reagents (HBTU, HATU) plus Fmoc protecting groups let pharmaceutical chemists string up to 50 residues reliably; recent flow chemistry has pushed past 150.
- The amide N–H is the key hydrogen-bond donor of protein structure. α-helix i to i+4 and β-sheet inter-strand bonds are amide N–H to amide C=O. Backbone hydrogen bonds set the energetic scale of folding: each contributes 5–20 kJ/mol depending on solvent exposure.
Common misconceptions
- Calling the peptide bond a single bond. Bond order is between 1 and 2 because of resonance — closer to 1.4. Drawing it as a pure single bond fails to predict planarity, the trans preference, or hydrogen-bonding geometry.
- Free rotation around C–N. The ω angle is locked. Free rotation belongs to the φ (N–Cα) and ψ (Cα–C) bonds adjacent to the peptide unit; the C–N peptide bond itself is essentially fixed.
- Treating cis-trans isomerism as negligible. 1000:1 sounds extreme but at Xaa-Pro the ratio is only ~4:1 (~20% cis at equilibrium), and the slow interconversion (k ~ 10−3 s−1) is rate-limiting in many folding pathways.
- Confusing the peptide unit with a single residue. A residue is the Cα plus its side chain plus its backbone atoms; the planar peptide unit straddles two residues, including the Cα from each. Drawing peptide planes around residues confuses geometry.
- Equating amide stability with ester stability. Amides hydrolyze ~105 times slower than the corresponding esters because the nitrogen lone pair stabilizes the carbonyl much more than oxygen does. Saponification of esters is fast at room temperature; amide hydrolysis requires hours of refluxing acid or base.
- Thinking the ribosome is just a passive RNA scaffold. The 23S rRNA peptidyl transferase center is the catalyst — Steitz, Yonath, Ramakrishnan won the 2009 Nobel for showing the active site is RNA. Protein components are structural, not catalytic.
Resonance, geometry, and isomerization
The peptide bond's planar geometry comes from one resonance contributor where the nitrogen lone pair is donated into the carbonyl π* orbital, formally giving N=C+–O−. The contribution of this zwitterionic form is roughly 40%, balancing the dominant neutral N–C=O form. The result is a partial π bond between C and N that demands sp2 hybridization at both atoms — and sp2 means trigonal planar geometry. The six atoms of the peptide unit (the two flanking Cα, the central C and N, the carbonyl O, and the amide H) sit in a plane within ~5°, and rotation around C–N would break the partial π bond, costing roughly 88 kJ/mol versus the 12 kJ/mol of a typical C–C single bond rotation.
Isomerization between trans (ω = 180°) and cis (ω = 0°) goes through this 88 kJ/mol barrier. At physiological temperature the half-life for spontaneous interconversion is tens of seconds — slow on the timescale of folding, fast on the timescale of cell division. Prolyl isomerases catalyze the reaction in active-site pockets that distort the peptide bond by twisting it 90° toward the perpendicular orientation, where the π contribution vanishes. The rate enhancement is ~105. Cyclophilin A, the cellular target of cyclosporine, is the textbook example; FK506-binding protein FKBP12 is the target of tacrolimus and rapamycin.
Pauling's 1951 derivation of the α-helix and β-sheet from peptide planarity alone was the proof-of-concept that chemistry could predict biology. He used X-ray diffraction data on glycine and α-keratin from Astbury, plus bond-length and bond-angle data from small amides like formamide and acetamide. The two structures fit the chemistry without adjustable parameters. Watson and Crick's 1953 DNA model explicitly cited Pauling's amide work as the methodology to follow. Subsequent crystallography of myoglobin (Kendrew 1958), hemoglobin (Perutz 1959), and lysozyme (Phillips 1965) all confirmed peptide planarity at atomic resolution.
Variant comparison: peptide bond versus other linkages
| Linkage | Formula | Bond length (Å) | Rotation barrier | Hydrolysis t1/2 (pH 7, 25°C) | Geometry |
|---|---|---|---|---|---|
| Peptide (amide) | R–CO–NH–R' | 1.32 (C–N) | ~88 kJ/mol | ~500 years | Planar |
| Ester | R–CO–O–R' | 1.34 (C–O) | ~50 kJ/mol | ~150 days | Quasi-planar |
| Thioester | R–CO–S–R' | 1.78 (C–S) | ~25 kJ/mol | ~7 days | Less planar |
| C–C single bond | R–CH2–CH2–R' | 1.54 | ~12 kJ/mol | Stable indefinitely | Free rotation |
| Glycosidic (acetal) | R–O–CH(–)–O–R' | 1.42 (C–O) | ~15 kJ/mol | ~5 million years | Free rotation |
| Phosphodiester | R–O–PO2–O–R' | 1.61 (P–O) | ~10 kJ/mol | ~30 million years | Free rotation |
| Disulfide | R–S–S–R' | 2.05 (S–S) | ~30 kJ/mol | Days to weeks | ~90° dihedral |
Applications and examples
- Solid-phase peptide synthesis. Merrifield's 1963 method couples activated amino acids onto a polystyrene resin one at a time. Standard Fmoc/tBu chemistry now produces peptides up to ~50 residues routinely; vendors ship custom 30-residue peptides in 2 weeks for $200–500.
- Insulin manufacture. Recombinant human insulin (Humulin, 1982) was the first FDA-approved biotech drug; the 51-residue hormone is two chains (A: 21 aa, B: 30 aa) joined by disulfides, all peptide bonds formed by E. coli ribosomes.
- Cyclosporine and tacrolimus. Both bind prolyl isomerases (cyclophilin, FKBP12 respectively) and exploit the cis-trans isomerization machinery. The peptide bond chemistry of immunosuppression is one example of the entire pharmacology built around amide stereochemistry.
- Proteases as drug targets. HIV protease inhibitors (saquinavir, ritonavir) bind the active site of an aspartyl protease and block hydrolysis of the gag-pol polyprotein. The transition-state analog hydroxyethylene mimics the tetrahedral intermediate of peptide hydrolysis.
- MALDI mass spectrometry of peptides. Tryptic digests cleave specifically C-terminal to Lys and Arg; the resulting peptides are 5–25 residues with predictable masses, enabling peptide-mass-fingerprint protein identification routine since 1996.
Frequently asked questions
Why is the peptide bond planar?
Resonance delocalizes the nitrogen lone pair into the carbonyl π system, giving the C–N bond roughly 40 percent double-bond character. The result is a partial π bond that demands six atoms — Cα of residue i, the carbonyl C, the carbonyl O, the amide N, the amide H, and Cα of residue i+1 — sit in a plane within about 5 degrees. Rotating around C–N would break the partial π bond and costs roughly 88 kJ/mol, far higher than the 12 kJ/mol barrier for a true single bond. Linus Pauling deduced this planarity in 1951 from electron density maps of glycine and small amides; it is the foundational constraint that made the alpha helix and beta sheet predictable from chemistry alone.
Why is trans favored over cis 1000 to 1?
In trans the two Cα substituents sit on opposite sides of the planar peptide unit, minimizing steric clash between residue side chains. In cis they sit on the same side and the Cα-H of one residue collides with the Cβ-CH3 (or larger) of the adjacent residue. The energy penalty is roughly 8 kJ/mol per cis bond, giving a Boltzmann ratio near 1000:1 trans:cis at body temperature. Proline is the exception: its pyrrolidine ring forces its α-nitrogen substituent into a position where cis and trans clash about equally, so the ratio drops to roughly 4:1 trans:cis. Approximately 6 percent of proline-containing peptide bonds in folded proteins are cis, and prolyl isomerases (PPIases, peptidyl-prolyl cis-trans isomerases) explicitly catalyze the otherwise slow interconversion.
How does the peptide bond form biologically?
On the ribosome, the α-amino group of an aminoacyl-tRNA in the A site nucleophilically attacks the carbonyl of the peptidyl-tRNA in the P site. The transition state is stabilized by RNA active-site residues (the ribosome is a ribozyme — peptidyl transferase activity comes from 23S rRNA, not protein). Water is released, the new peptide bond is formed in roughly 50 milliseconds per residue, and the deacylated tRNA exits to the E site. Each peptide bond formed costs four high-energy phosphates (two ATP for amino acid activation, two GTP for tRNA binding and translocation) — roughly 200 kJ/mol of metabolic energy per residue, of which only about 21 kJ/mol is captured in the bond itself.
How does peptide bond formation differ in chemistry versus biology?
Direct condensation of a free carboxylic acid and a free amine to form an amide is unfavorable in water — the equilibrium lies on the side of the reactants because water is the solvent and the leaving group is itself water. Synthetic peptide chemistry sidesteps this by activating the carboxyl as a more reactive intermediate: an acid chloride, an active ester (NHS, HOBt), or a phosphonium-uranium adduct (HBTU, HATU). The Merrifield solid-phase peptide synthesis (Nobel 1984) builds chains residue-by-residue with a coupling and deprotection cycle. Biology solves the same problem by activating the amino acid as an aminoacyl-AMP and then as an aminoacyl-tRNA — the activation energy is paid up front in ATP.
What does the omega angle measure?
Omega is the dihedral angle around the C–N peptide bond defined by atoms Cα(i) – C(i) – N(i+1) – Cα(i+1). Trans peptides have ω near 180°, cis have ω near 0°. Because of the partial double-bond character, ω is locked within about ±10° of these two values; intermediate angles cost 80 kJ/mol or more. The Ramachandran plot accordingly fixes ω and plots only the freely rotating φ and ψ angles around the Cα — that fixed-ω assumption reduces protein conformation from a 3-angle torsion problem to a 2-angle one and made the entire field of protein structural analysis tractable in 1963.
Why is hydrolysis of a peptide bond so slow?
The same resonance that makes the peptide bond planar also delocalizes electron density off the carbonyl carbon, making it a poor electrophile. Spontaneous hydrolysis at neutral pH and 25 °C has a half-life around 350 to 600 years per peptide bond. Proteases overcome this with specific catalytic strategies: serine proteases (chymotrypsin, trypsin) use a Ser-His-Asp triad to form a covalent acyl-enzyme intermediate; metalloproteases use a Zn-bound water as nucleophile; aspartyl proteases (HIV protease, pepsin) use two Asp residues with one activated water. Catalytic rate enhancements reach 1012, the largest in enzymology, exactly because the uncatalyzed reaction is so slow.