Polypeptide structure is described in four levels: primary, secondary, tertiary and quaternary. The first is the order of the amino acids. The other three describe how that chain sits in space. This guide explains each level, what holds it together, and why the sequence sets the shape.
Quick answer: What are the four levels of polypeptide structure? Primary: the sequence of amino acids. Secondary: local shapes of the backbone, such as the alpha helix and beta sheet. Tertiary: the overall three-dimensional fold of a single chain. Quaternary: the arrangement of two or more separate chains. Covalent peptide bonds fix the first level. Mostly weaker interactions, plus disulfide bonds, shape the other three.
The four levels at a glance
The four terms go back to Linderstrøm-Lang in 1952. The IUPAC-IUB rules on polypeptide conformation say these concepts “cannot be defined with precision,” but they give working definitions, which this guide follows.
The four levels of polypeptide structure
| Level | What it describes | What holds it together | Example |
|---|---|---|---|
| Primary | The amino acid sequence, read from the N-terminus to the C-terminus | Covalent peptide (amide) bonds | Position 6 of the hemoglobin beta chain is glutamate |
| Secondary | The local arrangement of the backbone atoms | Hydrogen bonds between backbone N-H and C=O groups | Alpha helix; parallel and antiparallel beta sheets |
| Tertiary | The position of every atom in one chain | Hydrophobic interactions, hydrogen bonds, salt bridges, disulfide bonds | Myoglobin, a single chain of 153 residues |
| Quaternary | How separate subunits are arranged and touch | Noncovalent contacts between subunits | Hemoglobin: two alpha and two beta chains |
Primary structure: the sequence
IUPAC defines primary structure as the amino acid sequence of the chain, without regard to its arrangement in space. By convention, a sequence is written with the N-terminal residue, the one with a free amino group, on the left and the C-terminal residue on the right. Each link is a peptide bond, formed between the carboxyl group of one amino acid and the amino group of the next. Our guide to peptide bond formation covers that reaction.
IUPAC’s definition leaves out the positions of disulfide bonds, so primary structure is not quite the same as “covalent structure.” A disulfide bond forms between two cysteine side chains. Within one chain it closes a loop. Between two chains it links them together.
The backbone: why the peptide bond is flat
The shape of a chain starts with the peptide bond itself. The nitrogen’s lone pair is shared with the neighboring carbonyl group by resonance. That gives the C-N bond partial double-bond character and restricts rotation around it, so the amide group is planar.
IUPAC names three backbone torsion angles. Phi (φ) is rotation about the N-Cα bond, psi (ψ) is rotation about Cα-C, and omega (ω) is rotation about the peptide bond. Omega normally sits near 0° or 180°, and about 180°, the trans form, is what is generally found. That leaves phi and psi as the main free rotations at each residue.
Secondary structure: helices and sheets
Secondary structure is the local arrangement of the backbone atoms, without regard to the side chains. In 1951, Pauling, Corey and Branson described two hydrogen-bonded helical configurations of the polypeptide chain. In the same year, Pauling and Corey described pleated-sheet configurations. These became the two classic patterns.
- Alpha helix. The chain coils into a right-handed helix with 3.6 residues per turn and 5.4 Å between coils. The N-H of each residue is hydrogen bonded to the C=O of the residue four positions back. The side chains point outward from the helix.
- Beta sheet. Stretches of chain lie side by side, with hydrogen bonds between residues in adjacent strands. Neighboring strands can run in the same direction (parallel) or in opposite directions (antiparallel).
Tertiary structure: the whole chain in three dimensions
IUPAC defines tertiary structure as the arrangement of all the atoms of one protein molecule, or one subunit, in space. Four kinds of interaction hold it in place:
- Hydrophobic interactions. Nonpolar side chains gather in the interior of a folded globular protein, away from water.
- Hydrogen bonds between nearby residues.
- Salt bridges. Ionic attractions between positively and negatively charged side chains. Charged side chains tend to sit on the outside of the protein.
- Disulfide bonds between cysteine residues, the only covalent cross-links on this list.
Myoglobin is a standard example: a small globular protein made of a single chain of 153 residues.
Quaternary structure: chains working together
Quaternary structure is the arrangement of separate subunits and the contacts between them. Hemoglobin is the textbook case: two alpha and two beta chains of about 150 amino acids each. Weak interactions between the subunits stabilize the whole.
Sources draw the line in slightly different places. IUPAC says a protein with no separable subunits, meaning chains not joined by covalent bonds, has no quaternary structure. Its examples are ribonuclease, with one chain, and chymotrypsin, with three chains. OpenStax Biology 2e describes insulin’s two chains, A and B, as linked by disulfide bonds, and uses insulin as an example in its quaternary-structure section. Under IUPAC’s stricter wording, chains held by covalent disulfides count as one unit.
Why the sequence determines the shape
The link between sequence and shape comes from experiments on one enzyme. Chris Anfinsen’s work at NIH showed that the primary sequence of ribonuclease encoded the three-dimensional structure of the enzyme. The disulfide bonds were reduced and the enzyme lost activity. When they were allowed to re-form, activity returned. What became known as the Anfinsen postulate holds that the tertiary structure of a folded protein is prescribed fully by its sequence. A 2012 paper notes that no contradictions seem to have been found, but also that the original experiments used only a single protein.
Two further facts fit the same picture:
- One change can matter. The hemoglobin beta chain is 147 amino acids long. In normal hemoglobin, position six is glutamate. In sickle cell hemoglobin it is replaced by valine, and that single substitution changes the secondary, tertiary and quaternary structure.
- Unfolding is often reversible. Denaturation removes a protein’s shape but keeps its sequence, which is why it is often reversible. Some proteins fold with help from chaperones.
Dill and MacCallum framed one core question as the physical code by which a sequence dictates a protein’s native structure. In 2021, the AlphaFold paper described structure prediction from sequence alone as an open problem for more than 50 years, and reported highly accurate predictions.
What this means for short peptides
The four levels were written for proteins. A review of bioactive peptides named “the conformational flexibility of most peptides” as a major difficulty, along with how strongly their shape depends on the environment. One response has been to lock part of the shape with covalent bonds between side chains, which closes the chain into a ring. Our guide to types of peptides covers cyclic and stapled designs.
For a short synthetic peptide, then, the primary structure is the property that defines it. The sequence fixes the mass, which is covered in peptide molecular weight. Purity is a separate question, covered in our guide to HPLC purity. For where “peptide” ends and “protein” begins, see what is a polypeptide and peptide vs. protein.

Vitrapep Research Catalog
Research peptides of defined sequence, supplied as laboratory reference compounds, each lot with a third-party certificate of analysis. For laboratory research use only, not for human or veterinary use.
Frequently asked questions
What are the four levels of protein structure?
Primary (the sequence), secondary (local backbone patterns), tertiary (the full fold of one chain) and quaternary (the arrangement of separate subunits).
What holds an alpha helix together?
Hydrogen bonds between backbone groups. The N-H of each residue bonds to the C=O of the residue four positions earlier. The helix has 3.6 residues per turn.
Does every protein have quaternary structure?
No. Under IUPAC’s definition, a protein without separable, noncovalently joined subunits has none. Ribonuclease (one chain) and chymotrypsin (three chains) are IUPAC’s examples.
Does the amino acid sequence determine the shape?
That is the Anfinsen postulate: the sequence fully prescribes the folded structure. It rests on refolding experiments with ribonuclease, and no contradictions seem to have been found.
Do short peptides fold like proteins?
Usually not in the same way. Most peptides are conformationally flexible, and their shape depends strongly on the environment.
What to know now
- Primary: the sequence, joined by covalent peptide bonds. IUPAC leaves disulfide positions out of it.
- Secondary: the alpha helix (3.6 residues per turn) and the beta sheet, both held by backbone hydrogen bonds.
- Tertiary: one chain’s full fold, held by hydrophobic interactions, hydrogen bonds, salt bridges and disulfides.
- Quaternary: separate subunits, such as hemoglobin’s four chains. Sources differ on covalently linked chains.
- Sequence sets shape: the Anfinsen postulate, built on ribonuclease refolding.
- Short peptides: most are flexible; their sequence is what defines them.