Peptides, Bonds and Structure

What a peptide is, precisely

A peptide is a chain of amino acids joined end to end by peptide bonds. That is the entire definition, and every other property follows from it: the identity of the residues, their order, and the chain's length are what distinguish one peptide from another.

Each amino acid in a chain is called a residue, because forming the bond costs it atoms — a molecule of water is removed at every junction. A chain of five residues therefore weighs less than five free amino acids added together, which is why a peptide's mass has to be calculated from residue masses plus one water, not from the amino acids themselves. This is the arithmetic behind every molecular-weight figure on this site.

Chains have direction. One end has a free amino group and the other a free carboxyl group, and by universal convention a sequence is written and numbered from the N-terminus to the C-terminus. The direction is not arbitrary bookkeeping: it is the order in which a ribosome builds the chain, and reversing a sequence gives a different molecule.

The peptide bond, and why it is not an ordinary single bond

The bond forms by condensation: the carboxyl group of one amino acid reacts with the amino group of the next, releasing water and leaving a C–N linkage — chemically an amide. In the cell a ribosome catalyses this; in a flask it needs an activating reagent, because the reaction is not spontaneous in water.

What makes the peptide bond consequential is that it does not behave like a plain single bond. The nitrogen's lone pair delocalises toward the carbonyl oxygen, giving the C–N linkage substantial double-bond character. Three things follow, and together they determine the shape of every peptide and protein: the bond is shorter and stronger than a normal C–N single bond; the six atoms around it are held in one plane; and rotation about it is restricted.

Because the amide unit is rigid and planar, a peptide backbone can only fold at the two bonds on either side of each α-carbon. Those two rotations are the φ and ψ angles, and the entire conformational repertoire of peptides and proteins — helices, sheets, turns — is built from their allowed combinations. A backbone that could rotate freely at every bond would have no stable structure at all.

Planarity also means the bond has two possible arrangements, trans and cis. Trans is strongly preferred almost everywhere, because in the cis form the neighbouring side chains collide. Proline is the exception that proves the rule: its side chain loops back onto the backbone nitrogen, so the two forms are closer in energy and cis-proline appears at meaningful frequency. The interconversion is slow enough to be rate-limiting in protein folding, and its energetics are measurably solvent-dependent (Eberhardt et al., J Am Chem Soc 1992).

Formed by

Condensation — a water molecule is removed as the bond is made. Hydrolysis reverses it, which is why peptides are digested.

Geometry

Planar. Six atoms — Cα, C, O, N, H, Cα — lie in one plane, held there by delocalisation.

Rotation

Restricted about C–N; free about the flanking φ and ψ bonds. Structure comes from that asymmetry.

Preferred form

Trans, overwhelmingly. Cis is common only before proline.

Stability

Kinetically stable in neutral water — half-lives of hundreds of years uncatalysed — but cleaved in minutes by proteases. This gap is why most peptide drugs are injected rather than swallowed.

Peptide versus protein: where the line actually falls

There is no chemical boundary between a peptide and a protein. Both are amino-acid chains joined by peptide bonds; the distinction is one of convention and size, and the convention is soft enough that authoritative sources disagree about individual molecules.

In common use, a chain of up to about 50 residues — roughly 5,000 daltons — is called a peptide, and anything longer is called a protein. Insulin, at 51 residues in two chains, sits exactly on the line and is described both ways in reputable literature. Nothing about the chemistry changes at that threshold; what changes is behaviour. Short chains tend to be flexible and to adopt structure only on binding a partner, while long chains fold into stable, defined three-dimensional shapes with hydrophobic cores.

The practical distinction that does track something real is whether the molecule folds. A protein has a tertiary structure that is essential to its function and can be destroyed by heat or solvent — denaturation. Most peptides have little to denature. That is also why peptides can be made by stepwise chemical synthesis while proteins usually cannot: there is no folding problem to solve at the end.

Oligopeptide

A short chain, conventionally a handful of residues — dipeptide, tripeptide and so on for specific lengths.

Polypeptide

A long chain, used when the emphasis is on the chain itself rather than on a folded functional protein.

Residue

One amino acid within a chain, so called because it has lost the atoms of a water molecule to the bond.

How the body makes peptides

Most peptides in biology are made by the ribosome and then cut down. A gene is translated into a long precursor — a preprohormone — which is trafficked, trimmed of its signal sequence, and then cleaved at specific basic-residue sites by prohormone convertases to release the active peptide. Insulin is the textbook case: proinsulin is a single chain that becomes active insulin only after excision of the connecting C-peptide.

This processing step is not a detail. When a convertase fails, the precursor accumulates and the active peptide never appears, with clinical consequences that demonstrate how much depends on the cut being made correctly (Meier et al., Nat Commun 2022). The same logic has been run in reverse as a discovery method: predicting where convertases will cleave has been used to find previously unrecognised peptide hormones hiding inside known precursors (Coassolo et al., Nature 2025).

A second route bypasses the ribosome entirely. Non-ribosomal peptide synthetases are large modular enzymes, found mainly in bacteria and fungi, that assemble peptides directly. Because they are not constrained by the genetic code they incorporate D-amino acids, unusual residues, and cyclic or branched architectures — which is why so many antibiotics, including vancomycin and the polymyxins, come from this route rather than from translation (Gulick, Nat Prod Rep 2017).

Why peptides make good drugs, and why they are difficult ones

Peptides occupy a middle ground between small molecules and antibodies, and they inherit advantages from both. They are large enough to bind flat, extended protein surfaces that small molecules cannot grip, which makes them selective and potent — and they are small enough to synthesise chemically, characterise exactly, and modify atom by atom. Their breakdown products are amino acids, so they tend not to accumulate.

The difficulties are equally structural. Proteases cleave peptide bonds readily, so an unmodified peptide can survive minutes in circulation. The kidney filters small molecules efficiently, removing what proteases miss. And the gut is a protease-rich, acidic environment that also presents an epithelial barrier a large polar molecule crosses poorly, which is why the overwhelming majority of peptide drugs are injected.

Almost every modification discussed elsewhere on this site is an answer to one of those three problems. Non-canonical residues such as α-aminoisobutyric acid block the protease. Fatty-acid chains bind albumin so the kidney does not filter the molecule. Cyclisation and stapling rigidify the backbone against both. The chemistry of peptide drugs is largely the chemistry of surviving the body long enough to work.