An alpha-helix inside a folded protein is held in place by the rest of the protein. Cut the same twenty residues out and dissolve them in buffer, and most of that structure is gone: the backbone samples a broad ensemble, and the helical population that remains is often small. Peptide stapling is the attempt to put the constraint back chemically — to tie two side chains together across the helix so that the folded state is favored without the parent protein being present. The all-hydrocarbon staple, installed by ring-closing metathesis, is the version of this idea worked on longest, and also the version whose limits have been argued about most publicly. This guide covers the chemistry, the two standard spacings and where their geometry comes from, what a helicity measurement does and does not establish, and why the field's own literature disagrees with itself.

Why an isolated helix comes apart

The problem is entropic before it is anything else. A helix requires a specific set of backbone dihedral angles at every residue, and an unconstrained chain in water pays a large conformational entropy penalty to adopt them. The intramolecular hydrogen bonds that stabilize the helix — each carbonyl oxygen to the amide nitrogen four residues along — compete with hydrogen bonds to solvent water, which are available everywhere and cost nothing. Short peptides are therefore poorly folded in isolation even when the same sequence is fully helical in the parent protein. That framing is standard structural chemistry and rests on no single primary source.

A covalent cross-link changes the arithmetic. If a bridge holds two positions at the separation they would have in a helix, the unfolded states that were previously accessible are removed from the ensemble, and the folded state becomes relatively more populated. A 2014 account from a group with more than a decade in this chemistry describes the all-hydrocarbon staple as conferring alpha-helical structure, protease resistance, cellular penetrance and biological activity upon successful incorporation of a series of design and application principles — a formulation that carries its own condition, since the same account is organized around lessons learned and pitfalls to avoid [1].

Ring-closing metathesis: the reaction that makes the staple

The synthetic route is unusual in that the macrocyclization happens on the solid support, before cleavage. Two olefin-bearing non-natural residues are incorporated at chosen positions during chain assembly, and a ruthenium alkylidene catalyst then joins their terminal alkenes, expelling ethylene and closing the ring. The published protocol identifies two features as crucial: constructing peptide substrates that contain hindered alpha-methyl, alpha-alkenyl amino acids, and performing the ring-closing olefin metathesis on the resulting resin-bound substrates [2].

The alpha-methyl group does structural work, not just occupying space. An alpha,alpha-disubstituted residue has a restricted backbone conformational range, so the building blocks bias toward helix before any cross-link forms. That also makes them harder to couple and to cyclize, which is the trade the chemistry accepts.

Because the metathesis is run on resin, it inherits every property of the support at that point in the synthesis — swelling, chain accessibility, and whatever aggregation the sequence has already developed. An incomplete metathesis leaves the uncyclized bis-olefin peptide, which differs from the stapled product by the mass of ethylene and is close to it in hydrophobicity; route-specific impurities that a routine chromatogram may not resolve are covered in our guide to solid-phase peptide synthesis impurities. We are not aware of a study quantifying how often that impurity survives purification, and state it here as a chemical expectation rather than a measured frequency.

i,i+4 and i,i+7: where the two standard spacings come from

The choice of positions is a geometry problem. Using the idealized alpha-helix parameters of 3.6 residues per turn and a 1.5 angstrom rise per residue, each residue advances 360 divided by 3.6, or 100 degrees, around the axis. From position i:

  • i to i+4 advances 4 times 100 = 400 degrees. Subtracting one full turn of 360 degrees leaves the two side chains 40 degrees apart around the helix — close to the same face — and 4 times 1.5 = 6.0 angstroms apart along the axis.
  • i to i+7 advances 7 times 100 = 700 degrees, which is 20 degrees short of two full turns at 720 degrees. The side chains are again nearly aligned, and 7 times 1.5 = 10.5 angstroms apart along the axis.
  • i to i+3 advances only 300 degrees, leaving a 60 degree offset — a larger angular mismatch than i,i+4 gives — and 3 times 1.5 = 4.5 angstroms along the axis.

Those numbers are geometric consequences of the idealized parameters, not measurements on any particular peptide; a real helix deviates, and the staple itself perturbs the local geometry. But they explain why the protocol literature describes its stapling systems as bridging one or two turns of an alpha-helix [2], and why the bridge for a two-turn span has to be substantially longer than the bridge for one turn.

The i,i+3 case is instructive rather than useless. A 3-10 helix has roughly 3 residues per turn, so i to i+3 corresponds to almost exactly one full turn there. Consistent with that, a 2024 study using ring-closing metathesis to install 4-atom hydrocarbon staples reported that the staple's configuration is critical for both stabilization and screw sense control of 3-10 helices, with circular dichroism showing that one particular stereochemical arrangement — R at position i, S at position i+3, flanked by methyl groups — strongly induces right-handed 3-10 helices, especially in sequences built from proteinogenic L-amino acids [3]. Short spacings and short bridges do not simply make a worse alpha-helix; they can select a different helix.

Even i,i+1 has been examined. A crystallographic study of stapling between a cis-4-allyloxy-L-proline and various olefin-tethered amino acids reported that the X-ray structure suggested the i,i+1 staple stabilizes the peptide secondary structure to the right-handed alpha-helix — a hedge the authors placed there deliberately, and one worth preserving, since a single crystal structure of a short oligopeptide is not a solution measurement [4].

Length, stereochemistry and direction are variables, not details

It is tempting to treat the staple as an inert brace whose only job is to span the right distance. Several systematic series argue otherwise. In a 2025 study of alkenyl butylcarbamate cross-links formed by ring-closing metathesis — a carbamate-containing bridge, not an all-hydrocarbon one — the authors varied staple length, stereochemistry and directionality independently, identified a 13-atom hex-2-enyl butylcarbamate staple as optimal within that series, and reported that peptide stereochemistry and staple orientation significantly impact both metathesis efficiency and helix stabilization [5]. Their optimized peptides showed a 45-fold increase in resistance to trypsin-mediated degradation relative to the unmodified counterparts — a within-study comparison in one series, not a general figure for stapling.

The staple also does not have to run between two side chains. A ruthenium-mediated macrocyclization tethering the alpha-amino group of one residue to the alpha-carbon of a residue two positions along, forming a 15-membered ring with a hept-4-enoyl staple, was reported to be highly effective at stabilizing helical structures of short peptides — a rigidified mimic of the N-capping box motif found in proteins rather than a conventional side-chain-to-side-chain bridge [6].

Nor is the effect of a given bridge transferable between sequences. A study of amine-bearing hydrocarbon cross-links described their helix-stabilizing effects as context-dependent and influenced by various structural parameters, while singling out a 9-atom version as a robust stabilizer that preserved water solubility; the authors explicitly called for further investigation rather than presenting a finished rule [7].

What the metathesis step leaves behind

Ring-closing metathesis produces a disubstituted alkene inside the macrocycle, and that alkene has a geometry. A 2023 paper introducing an alternative i,i+7 stapling chemistry stated the problem directly as motivation for its own approach: two distinct non-natural amino acids are required to synthesize i,i+7 alkene-stapled peptides, driving production costs up, and low purified yields are obtained because of the cis and trans isomers produced during ring-closing metathesis macrocyclization [8].

Whether that isomer ratio can be controlled appears to depend on the ring being closed. In the i,i+1 study above, E- or Z-selectivity was observed depending on the ring size of the stapled side chains and the structure of the olefin-tethered amino acids: the ratio reached up to 8.5 to 1 favoring E for the larger 14- to 17-membered rings, and up to 1 to 20 in the opposite direction for 13-membered rings [4]. Those are results for one specific stapling system and should not be read as a general rule for hydrocarbon staples; what generalizes is that the olefin geometry is an outcome of the reaction rather than something the designer specifies.

For anyone reading an analytical trace, the consequence is that a stapled peptide can present as two closely eluting species of identical mass. Whether they are separated, and which is in the vial, is a purification question rather than a synthesis one.

What a helicity measurement actually shows

Helicity in this field is almost always reported from circular dichroism, and usually from the mean residue ellipticity at 222 nanometers converted to a percentage using reference values for a fully helical and a fully random chain. Two features of that measurement are frequently underweighted.

The first is that the number is solvent-dependent by design. Trifluoroethanol, a helix-stabilizing solvent, is frequently used as the medium for determining the average alpha-helicity of polypeptides by circular dichroism, and a 1996 study presenting spectra at 10, 15, 20, 50 and 90 percent trifluoroethanol was framed specifically around demonstrating possible variations in the effect of that concentration on secondary structure [9]. A helicity percentage without its buffer and cosolvent is not a comparable quantity.

The second is that circular dichroism reports a bulk average over the whole peptide, not where the helix is. A 2018 study made this its central point: it noted that the three-dimensional solution structure of stapled peptides is rarely explored beyond acquiring circular dichroism data to quantify bulk helicity, leaving the detailed backbone structure typically undefined. The authors built diastereomeric stapled peptides based on helical sections of three proteins, quantified overall helicity by circular dichroism, then generated solution models for the most helical member of each series using NMR-derived restraints, and compared those models with deposited structures. Their conclusion was that circular dichroism alone is not sufficient to assess a stapled peptide's structural fidelity [10].

Neither observation makes circular dichroism a bad measurement. They mean a helicity percentage answers a narrow question — how much helical signal is present, in this solvent, averaged over these residues — and not the question a designer usually cares about, which is whether the residues that contact a partner sit where they do in the parent protein.

The disagreement in the literature

The clearest public dispute in this field concerns a stapled peptide derived from a helical BH3 segment. A 2013 study reported that stabilizing that helix did not necessarily enhance affinity or biological activity: the modified peptides had reduced affinity for their protein partners, an outcome the authors attributed to disruption of a network of stabilizing intramolecular interactions present in the bound state of the native peptide, noting that altering this network may compromise binding affinity. They further reported that cells exposed to these peptides did not readily undergo the expected response, which they took as strongly indicating that the construct is not inherently cell permeable [11]. The hedges are the authors' own and matter: they wrote may compromise, not does compromise, and framed the affinity loss as attributed rather than demonstrated.

A 2014 letter from the originating group, published as a comment on that study, argued that the label refers to a family rather than a single reagent. Iterating peptide sequence, staple type and sites of insertion generates discrete compositions suited to different biochemical, structural, proteomic, cellular and drug development applications, and the authors used the occasion to reinforce design considerations for avoiding pitfalls [12]. Read together, the two papers do not resolve into one being right. They describe a real problem — that nominally the same construct can be reported with different properties by different laboratories — and disagree about whether the cause is the technology or the specific composition tested.

A 2016 study from the same originating group is, in some ways, the more useful document, because it concedes the underlying weakness. It stated plainly that the design principles for generating cell-permeable stapled peptides are empiric, then used an unbiased statistical approach across stapled-peptide libraries to identify which biophysical parameters govern uptake. It reported that staple placement at the amphipathic boundary, combined with optimal hydrophobic and helical content, are the key drivers of cellular uptake, whereas excess hydrophobicity and positive charge at isolated positions can trigger membrane lysis at elevated peptide concentrations [13]. That last clause is the important one for interpreting older reports: a construct that permeabilizes membranes can produce cellular readouts that look like uptake and activity without either being on-mechanism.

What remains unsettled

Several things in this area are genuinely open, and it is more useful to name them than to paper over them.

  • There is no validated general rule mapping sequence to a productive staple position. The 2016 analysis frames its own findings as a design roadmap derived statistically, after describing the prior state of the art as empiric [13].
  • The relationship between measured helicity and target affinity is not monotonic. At least one well-documented case reports increased constraint with reduced affinity, attributed to disturbing interactions the native peptide makes in its bound state [11].
  • Bulk helicity and structural fidelity are different properties, and the second is measured far less often than the first [10].
  • Optimal bridge length and stereochemistry are reported per series, not universally — a 13-atom carbamate bridge in one study [5], a 9-atom amine-bearing bridge in another [7], a 4-atom bridge selecting a 3-10 helix in a third [3]. These are not competing answers to one question; they are answers to different questions.
  • Olefin isomer control in the macrocyclization is system-specific, with selectivity reported to invert with ring size in at least one stapling system [4].

The honest summary is that stapling most consistently does one thing — raise the helical population of a peptide that would otherwise be disordered — that even this effect is reported as context-dependent [7], and that everything downstream of it, from affinity to permeability, has to be established for each construct rather than assumed from the staple. Terms used above are defined in the research glossary.

All material discussed here is for laboratory research use only. It is not for human or veterinary use.