Cyclic peptides cross cell membranes by switching between open and closed conformations depending on the polarity of their surroundings. Molecular dynamics work on a ten-residue cyclic peptide traces a four-step route: membrane anchoring, insertion, an open-to-closed conformational transition, and…
Cyclic peptides cross hydrophobic cell membranes by changing shape. In water, a cyclic peptide holds an open conformation with its backbone hydrogen-bonding groups exposed to solvent. In a lipid membrane, it folds closed, burying that polar surface in intramolecular hydrogen bonds and presenting a non-polar face to the lipid. This chameleon behavior is the reason some cyclic peptides can be given by mouth, and it plays out through a four-step mechanism: membrane anchoring, insertion, an open-to-closed conformational transition, and diffusion across the bilayer.
The stakes are immediate. Oral delivery of peptides achieved a historic breakthrough in 1983 with cyclosporine A, a natural cyclic peptide with high oral bioavailability that was developed as an oral formulation. Its chemical structure is shown in Fig. 1. But the door cyclosporine A opened has not swung fully open. Researchers have still not presented a complete, general solution for oral peptide bioavailability. The hurdles are the same three: stability in the gut, resistance to enzymatic degradation, and crossing the intestinal cell membrane.
Linear peptides fail at the membrane. They are too flexible, their backbone amides are left unsatisfied in the lipid, and the energetic cost of desolvation is prohibitive. Cyclic peptides are different because those in the 700 to 2000 dalton range can bind flat protein-protein interaction interfaces with antibody-like affinity and specificity PMID 28388463 . That property has made them a serious modality for intracellular targets. Permeability is the bottleneck that decides whether such molecules are useful as drugs or merely as biochemical tools.
The name describes the behavior. A flexible cyclic peptide holds two families of conformations, and the polarity of the surrounding environment selects between them. In a polar environment, such as blood or cytoplasm, the peptide adopts an Open conformation: its hydrogen-bonded atoms are exposed and form hydrogen bonds with polar solvent molecules or with counterparts in solution. In a non-polar environment, such as the interior of a lipid bilayer, the peptide adopts a Closed conformation. It wraps up through intramolecular hydrogen bonds, encapsulates its polar groups inside the molecule, and exposes its non-polar structure at the surface.
Two measurable quantities follow from this switch. First, the polar surface area of the Closed conformation is smaller, because the polar backbone is sequestered. Second, the desolvation energy falls: the energy needed to strip the surrounding solvent from the molecule is lower when less polar surface is exposed. Both changes favor entry into a lipid membrane, and both are paid back on the far side when the peptide reaches polar cytoplasm again and reopens.
This is not a curiosity; it is a solubility strategy. A drug given orally must dissolve in the aqueous contents of the gut, survive transit, cross the lipid membrane of the intestinal epithelium, and then dissolve again in blood. A purely hydrophilic peptide never enters the membrane. A purely lipophilic one never dissolves in the gut in the first place. The chameleon property balances good water solubility and good fat solubility, and that balance is what favors cell permeability. The Closed state is generally regarded as the predominant permeability state of cyclic peptides, with the open-to-closed dynamic controlled by factors such as sequence, molecular size, and hydrophobic surface area.
The biology behind this is subtle. A linear peptide's backbone amides are all engaged in the same way, so the molecule has no reason to fold one way rather than another. A cyclic peptide's backbone amides can be turned inward or outward as a group, because the ring constrains the set of available conformations. The consequence is that the entire molecule can oscillate between two very different shapes, and the environment nudges the equilibrium. This is the chameleon effect, and everything that follows, including oral bioavailability, depends on it.
The most detailed mechanistic account comes from a molecular dynamics study by Stephanie M. Linker and colleagues, published in 2023 in the Journal of Medicinal Chemistry 66 4 :2773-2788 , titled "Lessons for Oral bioavailability: how Conformationally flexible cyclic peptides enter and cross lipid membranes." The researchers simulated the membrane crossing of a flexible decacyclic peptide, a ten-residue cyclic peptide, whose structure appears in Fig. 2 of the study. The simulation resolved the passive crossing into four steps, shown in Fig. 3 Linker, Stephanie M., et al., 2023 .
The four steps are:
| Step | Event | Conformational state | Key detail |
|---|---|---|---|
| 1 | Anchoring | Open or Closed | Side-chain residues establish the first contact with the membrane; anchor residues differ between conformations because residue contact degree differs |
| 2 | Insertion | Open or Closed | An open peptide inserts in either of two orientations, A and B, with A preferred; a closed, pre-folded peptide inserts only in orientation B |
| 3 | Open-to-Closed transition | Open to Closed | The critical switch; only orientation B favors it, and it is required for subsequent diffusion |
| 4 | Diffusion and flip | Closed | The peptide crosses from one polar/non-polar interface through the non-polar leaflet to the other side, then flips direction |
Step 1 is anchoring. Before a cyclic peptide can enter the membrane, specific side-chain residues must make initial contact with the membrane surface. These residues act as molecular anchors. Which residues anchor depends on the peptide's conformation, because the Open and Closed states expose different residues to the membrane, and the degree of contact for a given residue differs between the two conformations. A residue buried in the Closed fold cannot anchor; a residue protruding from the Open fold can.
Step 2 is insertion. Once anchored, the peptide enters the bilayer. An Open-conformation peptide can take two possible orientations, A and B, with orientation A preferred. A Closed, pre-folded peptide takes only orientation B. The distinction matters because only orientation B favors the conversion from Open to Closed. A peptide that inserts in orientation A is in the wrong geometry for crossing and must reorient before it can proceed.
Step 3 is the Open-to-Closed conformational transition. This is the critical step for membrane penetration. Open-state peptides cannot easily achieve diffusion through the membrane. The transition changes the spatial structure of the peptide and significantly affects its interaction with the membrane environment, which in turn determines the molecule's movement mode and diffusion efficiency. The simulation suggests that crossing a bilayer is not a matter of passively dissolving in the lipid; it is a matter of changing shape at the right point in the crossing.
Step 4 is diffusion and the flip. Only Closed-state molecules can diffuse from the polar/non-polar interface, through the non-polar lipid membrane leaflet, to the polar/non-polar interface on the other side. The final stage of crossing involves amino acid residues anchoring to the polar heads of the lower leaflet, a reversal of the peptide's direction, and a change in the direction of its hydrophobic moment. The molecule enters one leaflet in one orientation and leaves the far leaflet in the opposite orientation, an anchoring-and-flipping mechanism rather than simple diffusion of a rigid solute.
This four-step description is the product of molecular dynamics simulation, not direct experimental measurement. It says what a flexible cyclic peptide is likely to do in a lipid bilayer under the forces encoded in the model. It does not prove that every cyclic peptide crosses membranes this way, and it does not by itself predict what a given peptide will do in the human intestine. The goal of the simulation work is explicit: to improve peptide bioavailability, especially oral bioavailability, so that more peptide drugs can be delivered by mouth rather than by injection. That goal also marks the limit of the study, because bioavailability includes more than membrane crossing. The Open-to-Closed dynamic is controlled by multiple factors, including sequence, molecular size, and hydrophobic surface area, so the mechanism may not apply uniformly across all cyclic peptides.
The simulation resists simple residue-level rules. The effects of individual amino acids on membrane permeation, for example leucine, phenylalanine, and proline, were highly dependent on the peptide sequence, on the position of the residue within the peptide, and on the other residues present. These results argue against rules of thumb of the form "more hydrophobic residues means more permeable." The same residue can promote entry in one scaffold and hinder it in another, depending on where it sits in the ring and what it exposes in the Open and Closed states.
The broader cyclic peptide literature supplies context but does not close the gap between simulation and oral delivery. Much recent work has focused on biosynthesis and structural diversity. One genome-mining study identified 11 novel ribosomally synthesized and post-translationally modified peptides RiPPs with diverse aromatic cross-linking patterns, substantially expanding the known structural diversity of these cyclic peptides PMID 38069901 . A second study showed that the graspetide synthetase ThfB can install amide and, notably, thioester cross-links in prefuscimiditide, and that thioester-linked peptides can undergo native chemical ligation to form head-to-tail cyclic peptides, providing a recombinant route to such structures PMID 38634647 . These studies expand the chemical space available for permeability screening, including macrocycle size and cross-linking chemistry, but neither reports permeability or oral bioavailability data.
On the biological side, an experimental study identified a cyclic peptide, HiP-8, that selectively binds the active two-chain form of hepatocyte growth factor, locks its conformational dynamics, prevents MET binding, and enables detection of active HGF in cancer tissues and by PET, while cross-linked MET-binding cyclic peptides acted as synthetic HGF-mimicking agonists PMID 33121208 . The work demonstrates the targeting power of cyclic peptides, but it does not address oral delivery; the peptides functioned as biologics and imaging agents, not oral drugs.
A diagnostic study illustrates the same point from another direction. Using sera from 608 rheumatoid arthritis patients, researchers compared cyclic peptide-based assays with protein-based assays for detecting anti-modified protein antibodies. The antigen backbone influenced the detection of cross-reactivity: cyclic peptide-based assays detected broader multi-post-translational-modification reactivity, 61.2% for citrulline/carbamyl double reactivity, versus 54.0% for modified fetal calf serum-based assays, which captured fewer cross-reactive antibodies PMID 40624574 . This is evidence about assay performance, not about membrane permeation, but it is a useful reminder that "cyclic peptide" names a chemical class, not a pharmacological outcome. A cyclic peptide that is useful in an immunoassay tells nothing, by itself, about its ability to cross a membrane.
Taken together, this evidence base shows active research across cyclic peptide biosynthesis, structure, targeting, and diagnostics, but none of it directly tests the four-step membrane-crossing mechanism or demonstrates orally bioavailable cyclic peptide behavior in vivo. The molecular dynamics account from Linker and colleagues stands as the most detailed mechanistic description available, and it is a simulation-based description.
One generalization does reach beyond peptides. Because the first step of membrane crossing is side-chain anchoring, molecules that are not peptides can be made more membrane-permeable by moderate conjugation of aliphatic chains, which supply anchoring contacts analogous to peptide side chains. The suggestion is mechanistically coherent, but like the parent mechanism, it has not been systematically validated across chemical classes. It is a hypothesis for medicinal chemists to test, not a proven design rule.
Several actionable conclusions follow from the chameleon model, each with the caveat that it is grounded in simulation rather than clinical data.
Characterize the conformational behavior before designing the drug. The main conformations a peptide adopts before it reaches a membrane, and which residues are exposed in each, are the right starting point for optimization. Side-chain anchoring is conformation-dependent: a residue exposed in the Open state may be buried in the Closed state, or the reverse. Designing anchor residues without knowing the conformational ensemble is guesswork.
For macrocyclic peptides with continuous hydrophobic surfaces, favor orientation B. In the simulation, orientation B is the only insertion geometry that enables the Open-to-Closed transition essential for permeation. Design choices that bias the peptide toward that geometry, for example the extent and placement of hydrophobic surface, should be prioritized over choices that merely increase overall lipophilicity.
Balance hydrophilic and lipophilic character. The goal is not maximum hydrophobicity but switchability between the Open and Closed states. A peptide too polar to enter the membrane is as useless as one too lipophilic to dissolve in the gut. The chameleon property, measured as the difference in polar surface area between the two states, is the quantity to optimize.
Test residue effects in context. Leucine, phenylalanine, and proline all influenced permeation in the simulation, but their effects depended on sequence, position, and the other residues present. A residue-level rule that works in one scaffold should be assumed to fail in another until tested. The practical implication is that permeability screening, whether by simulation or experiment, must be done on the actual peptide under design, not on a homologous series of fragments.
Consider the anchoring mechanism for non-peptide molecules. Moderate aliphatic chain conjugation can increase membrane permeability by supplying the same kind of membrane contact that peptide side chains provide. This is the most transferable lesson from the simulation, and it may be the most useful to medicinal chemists working outside peptide space.
None of this guarantees oral bioavailability. The key to generally realizing oral bioavailability of peptides has not yet been completely presented, and the chameleon mechanism, though well supported as a physical model, is only one piece of the puzzle. Stability, efflux transporters, and first-pass metabolism remain separate problems that the permeability model does not address.
Four questions stand out. First, can general design rules be derived that reliably produce cyclic peptides with high oral bioavailability across diverse therapeutic targets? The simulation suggests the shape of such rules, but the dependence of amino acid effects on sequence context makes a simple lookup table unlikely.
Second, what is the full set of determinants of the Open-to-Closed equilibrium in membrane environments? Sequence, molecular size, and hydrophobic surface area are known to matter, but the list is probably incomplete. Solvent properties, lipid composition, and the presence of membrane proteins all plausibly shift the equilibrium, and none is captured in the current rules.
Third, how do molecular dynamics predictions translate into in vivo oral bioavailability? The four-step mechanism is a simulation result. Experimental permeation assays on model membranes and cell lines, followed by pharmacokinetic studies in animals, are the missing evidence. Until those exist, the mechanism is an informed hypothesis about what happens in the human intestine, not a demonstrated fact.
Fourth, can the anchoring-and-flipping mechanism be exploited predictively for macrocyclic drugs beyond peptides? The aliphatic-chain conjugation suggestion is promising, but it needs systematic testing across chemical classes before it becomes a design principle.
The chameleon model is the best current answer to how cyclic peptides cross membranes. It explains the 1983 cyclosporine A success, it rationalizes decades of permeability data, and it gives medicinal chemists a concrete mechanism to design against. It is also incomplete: simulation-based, context-dependent, and unproven in vivo. The next step is empirical, and it will decide whether the four-step mechanism becomes a design rule or a footnote.
Related reading: Peptide-Receptor Systems for Tumor Imaging: A Field Guide, AI-Guided Peptide Library Design: Capabilities and Outcomes, Solid-Phase Peptide Synthesis: Resins and Working Protocols, Enzymatic Routes to Oligopeptide Synthesis: A Technical Overview.