Protein splicing sounds like science fiction—cutting and stitching together pieces of protein like biological Lego. But in labs around the world, split inteins do just that, joining separate protein fragments into seamless wholes. They’re a staple in protein engineering, enabling the assembly of synthetic proteins too complex to make in one piece. The trouble? They often don’t work.

A team led by Henning Mootz and Christoph Humberg at the University of Münster has figured out why. Split inteins, especially a cysteine-less variant called Aes, are prone to misfolding in lab conditions, tanking their efficiency. In their Nature Communications paper, the team writes:
“We identify β-sheet-dominated aggregation of its N-terminal intein fragment as the origin of its low (~30%) splicing efficiency.”

Why Misfolding Matters

Misfolded proteins aren’t just a headache—they’re inactive. For split inteins, which need to fold properly before they can carry out splicing, even partial misfolding renders the whole system useless. The team observed that the Aes intein, when split, formed aggregates that couldn’t engage in protein trans-splicing.

This wasn’t unique to Aes. The authors noted,
“Importantly, we further showed that the explanation for the incomplete splicing efficiency seems generalizable, as other benchmark split inteins also show aggregation tendencies of one of their precursor fragments.”

That aggregation tendency, they explain, likely comes from the delicate structural dance required by split inteins:
“This general phenomenon arises from the intrinsic challenge to maintain the precursor in a partially disordered state while promoting stable folding upon fragment association.”

In other words, split inteins walk a fine line—they must stay flexible enough to bind their other half but stable enough not to collapse into misfolded junk.

Rewriting the Script at the Molecular Level

So, how do you fix a protein that folds wrong? You redesign it. Using a combination of crystallography, bioinformatics, and mutagenesis, the team pinpointed problematic amino acids in the Aes intein and modified them.

The result was a new version called CLm—short for Cysteine-Less and monomeric. This intein doesn’t clump, folds properly, and splices with high efficiency.

As the authors explain,
“The optimized CLm intein retains the wild-type’s ultra-fast reaction rate and serves as an efficient, thiol-independent protein modification tool.”

This matters because cysteine-less inteins are uniquely useful—they can operate in oxidative environments and work with thiol-directed chemistries without interference. But until now, none had shown all the activity traits—high rate, efficiency, and fragment affinity—needed for robust use in vitro and in cells.

“This work illustrates the potential of rational protein design in enabling efficient protein modification technologies, providing a general concept for preventing intein precursor aggregation by stabilization of monomeric intein fragments,” The authors concluded.