Cyclic Peptides Explained

Cyclic peptides explained: a ring changes everything

About 40 of the 100-odd peptide-based drugs approved globally have a cyclic or partially cyclic structure. That's a striking proportion, and it's not accidental. In this cyclic peptides explained walk-through, the thread running through everything is that the ring does a lot of work: conformational work, metabolic work, receptor-binding work. Understanding why requires backing up a little to the basics.

A peptide is a short chain of amino acids connected by peptide bonds. If you want the chemistry of that linkage in more detail, the article on the peptide bond covers it, and amino acids as the building blocks covers the units themselves. Linear peptides are like an open strand of beads: flexible, somewhat unpredictable in shape, and, critically, exposed at both ends to the enzymes that would rather they didn't exist for long. Cyclic peptides solve that last problem by connecting the two ends of the chain (or connecting internal residues) to form a closed loop. No free N-terminus. No free C-terminus. Fewer loose ends for proteases to grab.

That's the short version of why cyclisation matters. But the structural consequences run deeper than protease resistance, so let's work through them properly.

How cyclic peptides form their rings

There are several ways the ring can close, and the chemistry involved changes the downstream properties considerably. A useful habit when reading about any cyclic peptide is to ask first which atoms are joined to close the loop, because that answer tells you a good deal about stability, shape and how the molecule might be made.

The most straightforward is a head-to-tail cyclisation: the C-terminus of the last residue bonds to the N-terminus of the first, forming what's called a homodetic cyclic peptide. Cyclosporin A, the immunosuppressant, is the most widely cited example of this class, and it's a useful one because it's also full of non-standard amino acids, which we'll come back to.

Then there are heterodetic structures, where the ring closure isn't a standard peptide bond at all. Disulfide bridges between cysteine residues are common here; the cyclic peptide oxytocin uses exactly this arrangement (see oxytocin and vasopressin, and the separate piece on disulfide bonds in peptides). Side-chain-to-backbone lactam bonds are another variation. Some cyclic peptides contain ester linkages rather than amide bonds, making them technically depsipeptides; but that's a nomenclature argument for another day.

Bicyclic and polycyclic architectures exist too, where two or more rings share residues. These tend to be the stiffest, most conformationally locked structures, and they're of considerable interest in drug discovery for precisely that reason. The trade-off is that more rings usually means a harder synthesis, a point I'll return to near the end.

Conformational rigidity and why it matters for binding

Here's where I find myself going slightly nerdy even for a pharmacology piece, so bear with me.

When a linear peptide approaches a receptor, it has to adopt the right three-dimensional shape. That's an entropic cost: the molecule has to give up conformational freedom to bind, which thermodynamically disfavours the interaction. A cyclic peptide, especially a well-designed one, is already pre-organised. The ring constrains the backbone into a narrower range of shapes, so it arrives at the binding site closer to its bioactive conformation. Less entropic penalty. Often a tighter, more selective bind.

This is part of why cyclic peptides can achieve receptor selectivity that's difficult to engineer into smaller linear analogues. The shape is baked in rather than arrived at through chance. For the general ideas behind that, receptor binding, agonists and antagonists and peptide receptor selectivity are good companions, and peptide folding and tertiary structure explains why shape matters so much to begin with.

One common mistake is to assume that rigid always means better. A ring that locks the backbone into the wrong shape can bind worse than a floppy linear relative. Pre-organisation only helps when the frozen shape resembles the one the receptor wants, which is why design is often a matter of trying several ring sizes and closure points and comparing them.

Pharmacokinetics also shifts meaningfully with cyclisation. Oral bioavailability, usually a weakness of peptides, can improve because cyclic structures are less susceptible to gut proteases and, in some cases, are better able to cross biological membranes via passive transcellular diffusion. Cyclosporin A being orally bioavailable is still something of a pharmacological curiosity, given its size; its backbone N-methylation and cyclic structure are the main reasons it works. For the wider context on why oral delivery is normally so difficult, see why most peptides are injectable, not pills.

Metabolic stability: the protease problem

Linear peptides are, from the body's perspective, food. Proteases in the gut, blood, and tissues degrade them efficiently. This is actually the same reason dietary protein works: you need it broken into fragments for absorption, but it's a serious obstacle if you're trying to deliver an intact peptide to a specific target. The topic is covered more fully in enzymatic degradation of peptides.

Cyclic peptides resist enzymatic cleavage through a few overlapping mechanisms. Removing the free termini eliminates the substrate recognition sites that exopeptidases rely on. The conformational constraint also makes it harder for endopeptidases to achieve the geometry required for cleavage. And if the peptide incorporates D-amino acids or N-methylated residues, both common in natural cyclic peptides, enzymatic recognition drops further still, because most mammalian proteases evolved to process L-amino acid substrates in predictable conformations.

I'll be honest: I assumed for years that cyclisation alone was doing most of the stability work. The data suggest it's always a combination, ring closure plus non-standard residues plus sometimes lipophilic modifications, and no single feature is universally sufficient.

It also helps to remember that "stable against proteases" is not the same as "stays in the body for a long time". Clearance can happen by other routes, and binding to proteins in blood changes how much of a compound is free at any moment, which is the subject of the role of plasma proteins in drug binding.

Natural cyclic peptides: the biosynthetic logic

Nature got to cyclic peptides long before synthetic chemistry did. Fungi, bacteria, plants, and marine organisms have been producing them for hundreds of millions of years, mostly as part of their chemical defence repertoire.

The biosynthetic routes are worth knowing about. Ribosomal cyclic peptides (cyclotides, defensins, some cyanobacterial toxins) are assembled on ribosomes in the usual way and then cyclised post-translationally, often with additional modifications like disulfide formation. Non-ribosomal cyclic peptides are assembled by large enzymatic assembly lines called non-ribosomal peptide synthetases (NRPS), which work more like a molecular conveyor belt than a ribosome, and which are responsible for some of the most clinically significant cyclic peptides: cyclosporin, vancomycin, daptomycin.

The practical difference is flexibility. A ribosome is limited largely to the standard set of amino acids, so any unusual residues in a ribosomal peptide come from later chemical edits. An assembly line can slot in unusual building blocks directly, which helps explain why non-ribosomal products are so rich in D-amino acids and N-methylation.

Vancomycin deserves a mention here. It's a glycopeptide antibiotic with a tricyclic core. The cross-linked ring system is what gives it the rigidity to bind bacterial cell wall precursors so precisely. The binding affinity is a direct consequence of the shape, and the shape is a direct consequence of the cyclisation chemistry. The sugar attachments that put it in the glycopeptide class are a topic of their own; see glycosylation of peptides.

Marine organisms produce some extraordinary cyclic scaffolds too, though many remain at early research stages and haven't translated to the clinic yet, largely due to supply challenges and formulation complexity rather than lack of pharmacological interest.

Synthetic and semi-synthetic cyclic peptides in research

Medicinal chemistry has taken the structural lessons from natural cyclic peptides and run with them. Solid-phase peptide synthesis allows researchers to assemble sequences and then cyclise them on-resin or in solution, incorporating non-natural amino acids, PEG linkers, or fluorescent tags as needed.

One area attracting serious research attention is cyclic peptide macrocycles as "beyond rule-of-five" compounds, molecules that are too large for traditional small-molecule drug rules but can still achieve oral exposure, something that was considered largely impossible for large molecules until relatively recently. The publication record here, including work summarised in Nature Reviews Drug Discovery, suggests this is a genuine structural class rather than a passing trend. For a sense of how these larger molecules sit against conventional drugs, peptides vs small-molecule drugs is a useful comparison.

Australian regulation of peptide-based products

In Australia, peptide-based medicines that reach the clinic are subject to TGA assessment, and their scheduling depends on pharmacology. A peptide used for a specific therapeutic indication may be listed on the Australian Register of Therapeutic Goods (ARTG), see the ARTG glossary entry for how that process works. The Therapeutic Goods Administration is the regulator to check for authoritative information, and how peptides are scheduled in Australia explains the scheduling logic.

Many peptides used in research settings, particularly those marketed online for "performance" or "research" applications without TGA approval, are not approved therapeutic goods, and supplying them in Australia without the appropriate authorisation is illegal under the Therapeutic Goods Act 1989. The TGA has been actively enforcing this. Worth stating plainly, because the online peptide market does not always make this obvious. Related background sits in importation rules for peptides and compounding pharmacies and peptides.

For peptides that do make it to clinical use and require prescribing under controlled conditions, the Special Access Scheme and Authorised Prescriber pathways are the relevant regulatory mechanisms in Australia; the Special Access Scheme for peptides goes into the detail.

What cyclisation doesn't fix

I'd argue that cyclic peptides are sometimes overhyped in popular science writing as a kind of universal solution to peptide drug problems. They're not. Cyclisation can worsen solubility; a rigid, lipophilic cyclic structure might be metabolically stable but essentially insoluble in aqueous formulation vehicles. Getting one property right often means trading off another.

Membrane permeability is also not guaranteed. Some cyclic peptides cross membranes well. Others, particularly larger, more hydrophilic ones, don't, and end up with the same cell-entry limitations as their linear counterparts. And the synthetic accessibility of complex multicyclic structures is still a genuine challenge at scale, even with modern chemistry.

None of that diminishes the structural class. It just means cyclic peptides are a sophisticated tool with specific applications, not a blanket fix for peptide pharmacology. The hard yards in development are still about finding the right ring, for the right target, in the right format.

Between long runs on the Bay Run in Rozelle and whatever true-crime series I'm currently obsessing over, I keep coming back to cyclic peptides because they're a reminder that molecular architecture is almost never incidental. The ring is there for a reason. Usually several reasons at once.

Sources

• Cyclic Peptides: From Bioactive Natural Products to Drugs, NCBI / PMC
• Peptide Drug Overview: Stability and Delivery; NCBI Bookshelf
• Therapeutic Goods Act 1989, TGA, Australian Government
• Oral Bioavailability of Cyclic Peptides, NCBI / PMC

Priya Nandakumar, Pharmacology writer; peptides & PK