Secondary Structure of Short Peptides

The main secondary structures

Secondary structure describes the local, regular shapes a backbone adopts, defined by repeating values of the φ and ψ angles introduced in the guide to the peptide bond and stabilised by hydrogen bonds between backbone C=O and N–H groups.

In an α-helix every backbone C=O and N–H inside the helix is hydrogen-bonded, which is why helices are so stable in the hydrophobic interior of proteins and membranes. The four N–H groups at the N-terminal end and the four C=O groups at the C-terminal end have no partners inside the helix. Residues that satisfy them from the side chain, and charges that interact favourably with the helix dipole, are called helix caps; aspartate, serine, threonine and asparagine are common at the N-cap position.

Why short peptides are usually disordered

A protein folds because thousands of weak interactions, above all the burial of hydrophobic side chains, add up to outweigh the loss of conformational freedom. A 10- or 20-residue peptide has too few of these interactions to pay that price, so in water it remains an ensemble of rapidly interconverting shapes.

Helix formation adds a further difficulty: it is cooperative. The first turn is expensive, because three residues must be fixed in helical angles before the first i→i+4 hydrogen bond forms. Once a nucleus exists, each additional residue adds a hydrogen bond at a much smaller cost. In the classic helix–coil theory of Zimm and Bragg, and later Lifson and Roig, this appears as a small nucleation constant and a propagation constant close to one. The practical consequence is that helicity in short peptides rises steeply with length and depends heavily on the residues at the ends.

This is why a peptide sequence taken from a helix in a folded protein rarely forms a helix on its own. It lost the neighbours that held it in place.

Helix propensity

Not all residues favour a helix equally. The most widely used scale, measured by Pace and Scholtz from host–guest experiments in peptides and proteins, expresses each residue's preference as a free-energy cost relative to alanine:

Three features stand out. Alanine is the best helix former, which is why alanine-rich designed peptides were used by Baldwin and colleagues in the late 1980s to show, for the first time, that short peptides can form stable helices in water. Glycine is unfavourable because its flexibility makes the helical conformation entropically costly. And proline is in a class of its own: its ring fixes φ near −65° and its nitrogen lacks the hydrogen needed for an i→i+4 bond, so it is tolerated only in the first turn of a helix and otherwise produces a break or a kink. The kink at proline 14 in melittin is a well-known example, discussed in the hydropathy plot guide.

Side chains also interact with each other along the helix. Residues spaced i and i+4, or i and i+3, sit on the same face, and pairs such as Glu–Lys or Asp–Arg at those spacings can form salt bridges that stabilise the helix. Designed helical peptides use such pairs deliberately; stapled peptides replace them with covalent links.

Structure that depends on the environment

Many biologically important peptides are disordered in buffer and fold only in the right environment. This conditional folding is a feature rather than a defect: it lets a peptide remain soluble and accessible until it reaches its target.

• Membranes. Antimicrobial peptides such as magainin and LL-37 are largely unstructured in water but form amphipathic α-helices on contact with lipid bilayers, where the hydrophobic face inserts into the membrane and the charged face stays in contact with lipid head groups and water. This is covered in the guide to antimicrobial peptides.
• Receptors. Peptide hormones often adopt their active shape only when bound. Substance P, for example, is flexible in water, and its receptor-binding C-terminal region becomes helical in membranes and at its receptor.
• Co-solvents. Trifluoroethanol (TFE) and hexafluoroisopropanol strengthen backbone hydrogen bonds and induce helicity in sequences with any helical tendency. TFE titrations are a common way of measuring that tendency, but helicity in 50% TFE does not mean the peptide is helical in a cell.
• Concentration. Some peptides form helical bundles or β-sheet aggregates only when enough molecules come together. Structure measured at 1 mM may not exist at 1 µM.

β-Hairpins and aggregation

β-Structure in short peptides is harder to achieve in isolation than a helix, because a single strand has nothing to pair with. A β-hairpin solves this by folding the chain back on itself around a tight turn. Designed hairpins use turn sequences that strongly favour the right geometry, notably D-Pro-Gly, together with cross-strand pairs of aromatic residues. The "tryptophan zipper" peptides described by Cochran and colleagues in 2001 fold into stable hairpins of only 12–16 residues.

When strands pair between molecules instead of within one, the result is aggregation. Stretches rich in Val, Ile, Phe, Tyr and Gln are particularly prone to stack into the cross-β structure of amyloid fibrils. This is a major practical concern in peptide handling and formulation, discussed in the guide to stability and storage, and on resin it is one of the main causes of difficult syntheses.

Measuring secondary structure: circular dichroism

Circular dichroism (CD) measures the difference in absorbance of left- and right-circularly polarised light. In the far UV, between about 190 and 250 nm, the signal comes from the peptide bonds, and their regular arrangement in each type of secondary structure produces a characteristic spectrum:

For helices, the signal at 222 nm is commonly converted into an estimate of helical content. The measured ellipticity is first normalised to mean residue ellipticity, [θ], which removes the effects of concentration, path length and number of residues. The fraction of helix is then estimated by comparing [θ]₂₂₂ with the value expected for a fully helical peptide of the same length, which is around −35,000 to −40,000 deg·cm²·dmol⁻¹ for long helices and less negative for short ones, because the frayed ends contribute less.

Two cautions apply. The estimate depends directly on an accurate peptide concentration, which in turn depends on knowing the real peptide content of the sample, as explained in the molecular weight guide. And aromatic side chains, especially tryptophan, contribute their own CD signals in the far UV and can distort the spectrum of short peptides.

Higher-resolution information comes from NMR spectroscopy, which can give residue-by-residue secondary structure and full three-dimensional structures of peptides in solution, micelles or bicelles. Infrared spectroscopy, through the amide I band near 1650 cm⁻¹, is widely used for aggregated and membrane-bound samples.

Predicting structure from sequence

Classical predictors such as the Chou–Fasman method assigned propensities to each residue and looked for stretches favouring helix or strand. For short peptides the problem is different from that of proteins, because the right answer is often "disordered unless bound". Several practical rules follow from the principles above:

1. A stretch of at least 10–12 residues with high helix propensity and no Pro or multiple Gly is a candidate helix.
2. If hydrophobic residues fall every three to four positions, the helix would be amphipathic, and the peptide is likely to fold at membranes. The helical wheel in the antimicrobial peptides guide shows how to check this.
3. Alternating hydrophobic and polar residues suggest a β-strand, and a risk of aggregation.
4. High proline content suggests polyproline II or a flexible, extended peptide.

Modern structure prediction tools based on deep learning are very accurate for folded proteins. For short, conditionally folded peptides their confidence scores should be read carefully: a confident helix may be the bound conformation rather than the state in solution.

Frequently asked questions

References

• Pace CN, Scholtz JM (1998) A helix propensity scale based on experimental studies of peptides and proteins. Biophysical Journal 75:422–427.
• Marqusee S, Robbins VH, Baldwin RL (1989) Unusually stable helix formation in short alanine-based peptides. PNAS 86:5286–5290.
• Cochran AG, Skelton NJ, Starovasnik MA (2001) Tryptophan zippers: stable, monomeric β-hairpins. PNAS 98:5578–5583.
• Greenfield NJ (2006) Using circular dichroism spectra to estimate protein secondary structure. Nature Protocols 1:2876–2890.

Related guides

Browse all 27 guides →