The Science of Cyclopeptides

What Makes a Peptide Cyclic?

All proteins are made of amino acid chains. In a normal (linear) peptide, the chain has two free ends: an N-terminus (amino end) and a C-terminus (carboxyl end). These ends are attack points — digestive enzymes called exoproteases latch onto them and start unravelling the protein from either end, one amino acid at a time. This is why essentially all proteins you consume in food are fully degraded before reaching your bloodstream.

Cyclopeptides have no ends. Their chain is joined head-to-tail by a covalent peptide bond, forming a closed ring. Without free termini, exoproteases have nothing to grip. Endoproteases (which cut in the middle of chains) can still theoretically attack — but in cyclotides, a second structural feature stops them too.

This is not a subtle chemical difference. The absence of free termini fundamentally changes the molecule's relationship with the biological environment it moves through. A cyclic peptide can circulate in the bloodstream, survive the gut, and pass through tissues without being dismantled by the protein-digesting machinery that destroys virtually everything else.

The Cyclic Cystine Knot (CCK) — Nature's Molecular Armour

Cyclotides — the most biomedically important subclass of cyclopeptides — take protection further with a structure called the Cyclic Cystine Knot (CCK). The knot is formed by six conserved cysteine residues in the cyclic backbone, which pair up to create three interlocking disulfide bonds (S–S bonds). These three bonds are arranged so that one pair threads through a ring formed by the other two — creating a true topological knot at the molecular level.

The cyclotide backbone is divided into six loops (Loop 1 through Loop 6) by the positions of the six cysteine residues. The CCK is formed by disulfide bonds connecting Cys I–IV, Cys II–V, and Cys III–VI. The Cys III–VI bond passes through the ring formed by the Cys I–IV and Cys II–V bonds and the intervening peptide backbone — an arrangement that can only exist in a cyclic molecule.

The result is a structure that:

• Cannot be unfolded by heat (stable to boiling at 100°C)
• Resists acid degradation (survives the stomach's pH ~2 environment)
• Cannot be cut by trypsin, chymotrypsin, pepsin, elastase, or carboxypeptidases
• Maintains its 3D shape for years at room temperature
• Retains biological activity after lyophilisation and long-term storage

This is not merely chemical stability — it is topological stability. The molecule would have to be chemically broken (covalent bonds cleaved) to change its shape. Unlike most proteins, which can be denatured simply by disrupting hydrogen bonds and hydrophobic interactions, the CCK scaffold requires the physical destruction of disulfide bridges to unfold. No enzyme in the human body performs this reaction.

The Three Cyclotide Subfamilies

Cyclotides are divided into three structural subfamilies:

Möbius Cyclotides

Möbius cyclotides contain a cis-proline residue in loop 5 that introduces a conceptual "twist" into the backbone — analogous to the half-twist of a Möbius strip. This gives Möbius cyclotides a distinctive shape in which a notional vector traversing the loop changes orientation. Kalata B1 — the first discovered and best-characterised cyclotide — is the archetypal Möbius cyclotide. Other Möbius cyclotides include kalata B2, kalata B7, and varv E.

Bracelet Cyclotides

Bracelet cyclotides lack the cis-proline twist. Their backbone traces a simple ring without the topological flip, making them structurally more like a bracelet than a Möbius strip. Bracelet cyclotides tend to be larger and more structurally diverse in their loop sequences. Cycloviolins (from violet species) are typical bracelet cyclotides, as are the cycloviolacins found across Violaceae species.

Both subfamilies share the CCK motif and exhibit exceptional stability, but their different backbone geometries give them distinct surface properties, membrane interaction profiles, and biological activities. Bracelet cyclotides are the larger subfamily (around two-thirds of all known natural cyclotides) and generally show stronger cytotoxic, membrane-disrupting, and anti-HIV activity — the original anti-HIV cyclotides (the circulins and cycloviolins) are bracelets. Möbius cyclotides are most strongly associated with uterotonic activity, reflecting the discovery of the first cyclotide, kalata B1, in a traditional childbirth preparation.

Trypsin Inhibitor Cyclotides

A third, smaller subfamily — the trypsin inhibitor cyclotides (also called cyclic knottins) — has a looser cystine knot than the Möbius or bracelet subfamilies and a more limited range of bioactivities, primarily trypsin inhibition. The best-known members are MCoTI-I and MCoTI-II, isolated from the seeds of Momordica cochinchinensis (Cucurbitaceae). Despite their limited native activity, MCoTI-II is one of the three most important cyclotide scaffolds used in peptide grafting for drug design, alongside kalata B1 and MCoTI-I, because it is efficient to fold, tolerant to sequence changes, and cell-penetrating.

SFTI-1: The Sunflower Exception

Not all plant cyclopeptides are cyclotides. SFTI-1 (Sunflower Trypsin Inhibitor-1) is a 14-amino-acid cyclic peptide isolated from sunflower seeds (Helianthus annuus). It is the smallest, most potent naturally occurring Bowman-Birk protease inhibitor and the only naturally occurring cyclic member of its inhibitor class.

Although SFTI-1 is much smaller than cyclotides and lacks the CCK (it has only a single disulfide bond), its cyclic backbone still confers remarkable protease resistance and structural rigidity. It inhibits trypsin with an inhibition constant (Ki) of approximately 0.1 nM — making it one of the most potent serine protease inhibitors in nature.

SFTI-1 is extensively studied as a drug scaffold. Because its single binding loop can be exchanged for sequences that inhibit medically relevant proteases (such as matriptase, kallikrein, and cathepsin G), researchers have used SFTI-1 as a template to engineer cyclic inhibitors with potential applications in cancer, coagulation disorders, and inflammatory disease.

Orbitides: Cyclopeptides Without Disulfide Bonds

Orbitides (also called ribosomally synthesised plant cyclopeptides, or RPCs) are head-to-tail cyclic peptides that gain stability purely from their closed backbone — they contain no disulfide bonds. They are found in flax (Linum usitatissimum), Annonaceae, Caryophyllaceae, Euphorbiaceae, and other plant families.

Because they lack the CCK's additional stabilisation, orbitides are somewhat less resistant to proteolysis than cyclotides. However, their cyclic backbone still confers significantly greater stability than equivalent linear peptides, and hundreds of orbitides have now been structurally characterised. Several are being investigated for cytotoxic activity against cancer cell lines and anti-inflammatory properties. The segetalins from Vaccaria hispanica are among the best-studied orbitides.

How Cyclopeptides Are Found in Plants

Cyclotides naturally accumulate in plant tissues at concentrations of up to 2 mg per gram of dry plant material — remarkably high for a pharmaceutical compound. A single plant species can produce dozens to hundreds of related cyclotide variants simultaneously, each with slightly different amino acid sequences in the variable loops but the same conserved CCK scaffold.

The plant Oldenlandia affinis was the source of the first isolated cyclotide. Today, cyclotides and cyclotide-like peptides have been confirmed in six unrelated plant families:

• Rubiaceae (coffee family) — the original source; estimated to contain 50,000+ variants across the family
• Violaceae (violet family) — European and tropical violets; extensively studied bracelet cyclotides
• Cucurbitaceae (cucumber and squash family) — source of the trypsin inhibitor cyclotides MCoTI-I and MCoTI-II from Momordica cochinchinensis
• Fabaceae (legume family) — including butterfly pea (Clitoria ternatea), source of Sero-X and the peptide ligase butelase-1
• Solanaceae (potato and petunia family) — cyclotides and acyclotides identified in Petunia × hybrida; notable because Nicotiana benthamiana, widely used in plant molecular farming, belongs to this family
• Poaceae (grass family) — linear cyclotide-like peptides (panitides) characterised in Panicum laxum; cyclotide-like gene sequences identified in rice, maize, and wheat

The presence of cyclotides across such distantly related families — separated by hundreds of millions of years of evolutionary history — suggests convergent evolution. Nature independently discovered this same highly stable structural motif multiple times, presumably because the CCK confers a strong survival advantage to the host plant.

Why Plants Make Cyclotides

The primary natural role of cyclotides appears to be host defence against herbivorous predators. Cyclotides are toxic to insect larvae, nematodes, and molluscs at concentrations well within the range found in plant tissue — while remaining largely harmless to vertebrates at comparable concentrations.

The mechanism is membrane disruption. Cyclotides are amphipathic — they have a hydrophobic face and a hydrophilic face. When they encounter the lipid bilayer of a cell membrane, they insert into the membrane and disrupt its integrity, causing ion leakage and eventually cell lysis. In insects, this causes paralysis and death. The selectivity for invertebrate membranes over vertebrate membranes is not completely understood but appears to relate to differences in membrane composition and curvature.

This insecticidal mechanism is exploited in the world's first commercial cyclotide product: Sero-X, a biopesticide derived from butterfly pea extract and commercially released in Australia in 2016 following APVMA approval of the active constituent in late 2015. It is registered for cotton, macadamia, and vegetable crops, and is non-toxic to bees, beneficial insects, and vertebrates at agricultural use rates.

Secondary roles for cyclotides may include anti-microbial activity against plant pathogens and regulation of plant hormonal signalling, but these are less well characterised than the insecticidal role.