Most people picture drugs as small chemical tablets or large protein injections. Macrocyclic peptides sit in an interesting middle ground. They are chains of amino acids, like proteins, but they are short and deliberately folded into rings. That ring shape, researchers argue, gives them some of the best properties of both worlds: the precise targeting of a large biological molecule and the durability and body-friendliness of a small chemical drug.
A comprehensive review published in Chemistry examined the synthetic strategies scientists use to build and select these ring-shaped peptides through a technique called bacteriophage display, or phage display for short. The review offers a detailed map of where the field stands today and where it is likely to go, covering everything from well-established chemical tricks to machine-learning tools on the horizon.
Understanding this research matters because macrocyclic peptides are appearing more frequently in early drug-discovery pipelines. The review explains both why that is happening and what technical hurdles still stand in the way.
What macrocyclic peptides are
A peptide is a short chain of amino acids, the same building blocks that make up proteins. Standard peptides have two loose ends, a beginning and an end to the chain. A macrocyclic peptide, by contrast, has those ends connected, forming a closed loop or ring. The word 'macrocyclic' simply means 'large ring' in chemical terms.
This ring structure is not just a design preference. The review explains that cyclization stabilizes the peptide's three-dimensional shape, making it harder for the body's enzymes to chew through it. Straight-chain peptides are often broken down quickly in the bloodstream or the gut. A rigid ring is harder to degrade, which is one reason researchers are so interested in these structures.
Beyond stability, the ring shape also constrains the peptide into a specific conformation, meaning it tends to sit in a defined posture when it approaches a target molecule. That consistency can improve how tightly and selectively the peptide binds to its target, a property researchers call binding affinity.
The role of phage display
Phage display is a biology-based screening method that has been in use since the 1980s. Bacteriophages are viruses that infect bacteria, not humans. Researchers can engineer these viruses to display peptide sequences on their outer surfaces, essentially wearing different peptides like name badges.
By creating enormous libraries of phages, each wearing a slightly different peptide, and then exposing those libraries to a target molecule, scientists can fish out the phages whose peptides bind well to the target. The bound phages are collected, amplified, and screened again. After several rounds, the strongest binders emerge. This iterative process is called biopanning.
The review points out a core limitation: phage display works with the natural amino acid building blocks that biology uses. When a promising peptide binder turns up, it is almost always a straight chain of standard amino acids. Turning that into a stable, ring-shaped drug candidate typically requires significant additional chemistry work after the selection process is complete. That extra work takes time and resources, and it can alter a peptide enough that the original binding properties are lost.
Hybrid macrocyclic peptides as a solution
The central idea the review explores is doing the ring-forming chemistry during the selection process, not after it. If phages can display already-cyclized peptides, then researchers would be selecting from a library of ring-shaped candidates from the start. The resulting hits would already have the stability and shape benefits built in, reducing the optimization work downstream.
These structures are called hybrid macrocyclic peptides because they combine natural amino acid sequences with synthetic, non-peptidic scaffolds or linkers that form the ring. The 'hybrid' label reflects the blend of biological and purely chemical components.
The review surveys a wide range of chemical strategies that make this possible. Some methods target the amino acid cysteine, which carries a sulfur atom that reacts readily with certain chemical groups. Others target lysine, which carries an amine group. More recent approaches use enzymes or exploit the natural spatial proximity of atoms within a growing peptide chain to trigger cyclization. Each approach has tradeoffs in terms of which peptide sequences it works with, how cleanly it reacts, and whether the phage virus tolerates the chemical conditions required.
Key chemical strategies reviewed
Cysteine-targeting chemistry is among the most established methods. Because cysteine has a reactive sulfur atom, chemical crosslinkers can grab two cysteine residues within the same peptide chain and bridge them together, forming a ring. This approach has been refined over decades and the review describes it as a workhorse of the field.
Lysine-targeting chemistry follows a similar logic but uses the amine group on lysine instead of the sulfur on cysteine. This expands the range of peptide sequences that can be cyclized.
The review also covers newer residue-selective methods, which can target less common or modified amino acids, giving chemists finer control over where the ring forms. Chemoenzymatic approaches use enzymes, proteins that act as biological catalysts, to perform the cyclization step in a more controlled and gentle way. This is valuable because harsh chemical conditions can damage the phage and destroy the library.
Finally, the review discusses proximity-driven reactivity. When two parts of a peptide chain are held close together by the phage's three-dimensional structure, certain chemical reactions can occur spontaneously at those sites. Researchers are learning to exploit this spatial logic to trigger cyclization without needing aggressive chemistry.
Design principles for phage-compatible cyclization
Not every ring-forming chemical reaction is suitable for use inside a phage display experiment. The review outlines several requirements that a cyclization strategy must meet to be useful in this setting.
First, the reaction must work in water, because phages live in aqueous buffer solutions. Many organic chemistry reactions require non-water solvents that would destroy the phage. Second, the reaction must be fast and efficient enough to modify the peptides on the phage surface without requiring extreme temperatures or pH levels that would degrade the virus. Third, the chemical groups involved must not react with the many other biological molecules present in the mixture.
The review also emphasizes selectivity. If a crosslinker reacts with unintended sites on the phage itself, the result is noise in the selection, making it harder to identify genuine binders. The best strategies are those that react cleanly and specifically with only the intended amino acid residues.
Future directions and machine learning
The review closes by looking at where the field is heading. One emerging area is the integration of machine learning tools into peptide discovery. By training models on large datasets of peptide sequences and their binding properties, researchers hope to predict which sequences are worth synthesizing and screening before any wet-lab work is done. The review describes this as a promising way to reduce the sheer number of experiments needed.
Enzyme engineering is another frontier. By redesigning enzymes to perform cyclization chemistry that does not exist in nature, researchers could gain access to entirely new ring structures. The review notes that this area is still early but shows considerable potential.
The authors also discuss bioconjugation chemistry, the science of attaching synthetic molecules to biological ones in precise ways. As these techniques improve, the range of non-peptidic scaffolds that can be incorporated into phage-displayed libraries is expected to grow, further expanding the chemical space available during biological selections.
Taken together, the review presents a field in active development. The core challenge, making macrocyclic peptide discovery faster and more reliable, is being approached from multiple chemical and computational angles. Early data points at meaningful progress, though the authors are careful to note that many of these strategies remain at the proof-of-concept stage and have not yet been widely validated in clinical pipelines.
