A barcode-free way to find permeable macrocycles
A Nature Communications paper introduces CycloSEL, a mass spectrometry driven workflow that screens fully synthetic macrocycle libraries and reports an intracellular-target hit later optimized for permeability.
Macrocycles sit in an awkward middle ground. They can be big enough to grab difficult protein surfaces, but small enough to feel “drug-like” if you can get the chemistry right.
The catch is that macrocycles often live beyond Lipinski’s Rule of Five, which means the usual tricks for small molecules stop being reliable. A macrocycle can look too polar on paper and still cross a membrane if it can temporarily hide its hydrogen bonds. Or it can look promising and still fail because it cannot adopt the right shapes.
That reality has pushed the field toward a practical question: how do you search enough chemical space to find macrocycles that bind hard targets, and still have a fighting chance at properties like permeability and stability?
A new Nature Communications paper describes one answer: an end-to-end workflow called CycloSEL (Cyclic Self-Encoded Libraries) that screens fully synthetic macrocycle libraries using affinity selection and identifies hits by tandem mass spectrometry, without requiring a genetic barcode (PubMed).
The bottleneck CycloSEL is trying to remove
There are already ways to screen vast numbers of macrocycles. But many high-throughput platforms lean on genetically encoded libraries (for example, display technologies) where “library size” is easy, while chemical flexibility is constrained.
CycloSEL is pitched as a way to keep the scale while widening the chemistry. The authors build libraries enriched for features medicinal chemists care about in this space: the messy, shape-shifting territory where permeability and binding can coexist.
In other words, instead of asking a library format to behave like a drug, they try to build “drug-like macrocycle” biases into the library itself.
What they built (at headline level)
The core claim is architectural: CycloSEL can run affinity selection → hit identification on a fully synthetic macrocycle library at high scale.
In the paper’s main demonstration, they report constructing a 16 million member macrocycle library and then using tandem mass spectrometry (MS/MS) to identify enriched binders after selection.
Two targets matter for how you read the story:
- Carbonic anhydrase IX (CAIX), used as a validation target.
- WDR5, a protein involved in chromatin regulation, used as a tougher test case.
The WDR5 piece is where CycloSEL starts to look like more than a library trick. The authors report a macrocycle with sub-nanomolar affinity and inhibition of the WDR5–MLL1 interaction, then describe follow-on modifications that produced a more “chameleonic” molecule with passive membrane permeability, serum stability, and anti-proliferative activity in leukemia cells.
Why the “permeable macrocycle” claim is the point
For years, macrocycles have been sold on the idea that they can reach targets small molecules struggle with. But intracellular targets raise the bar: binding is not enough if your compound cannot reach the cytosol.
So the interesting arc here is not simply “screen bigger.” It is: screen big, then show a credible path from selection hit to an intracellularly relevant molecule.
This also lines up with what broader macrocycle surveys have been saying: many successful macrocycles behave like drug candidates because they exploit conformation and hydrogen bonding in ways that simple property filters miss (PubMed).
The skeptical read, and what would make this a real shift
A platform paper can look great on a hand-picked target. The hard test is generality.
If CycloSEL is going to change how macrocycles are discovered for intracellular biology, you would want to see it work repeatedly on targets that are:
- not already “library-friendly”,
- not dependent on unusually forgiving binding sites,
- and not rescued by unusually heavy optimization.
The best next proof is boring but decisive: multiple independent targets where CycloSEL yields hits that can be tuned into molecules with permeability and stability without years of bespoke chemistry.