Branching out of the intein active site in protein splicing.

Inteins perform a macromolecular vanishing act that continues to impress as its secrets are revealed. Dispersed through all domains of life, these autocatalytic domains exit the folds of ostensibly unrelated proteins soon after translation. Inteins bust at the seams, severing peptide bonds that immediately precede and follow their sequence. Critically, they also sew up the damage; the two polypeptide segments that first flanked the intein, known as exteins, are concurrently joined. Extein ligation renders the transformation traceless while adding a significant exception to the one gene/one protein rule. The act proceeds without assistance from cofactors, accessory proteins, or energy source—just a small catcher’s mitt structure of ∼130 amino acids (Fig. 1A). Since the discovery of inteins (1, 2), efforts to understand their protein splicing activity have inspired not only mechanistic insight but biotechnology applications galore. Following a proposal put forward by Perler, Paulus, and coworkers more than 20 y ago, the chemical mechanism for protein splicing has three catalyzed steps (3) (Fig. 1B). The sequence begins when the intein’s first residue, typically cysteine or serine, attacks the peptide bond at the upstream splice junction, yielding a (thio)ester known as the linear intermediate (LI) (4). Next, the activated extein residue, along with all extein residues preceding it in the precursor, migrate to the extein residue at the downstream splice junction. The recipient extein residue is also typically a serine or cysteine. This transesterification event gives rise to the so-called branched intermediate (BI), a term coined by Xu et al., who provided the first experimental evidence for its existence (5). The branched intermediate was identified as a relatively slowly traveling band on SDS/ PAGE that gave rise to two N-terminal amino acids when subjected to protein sequencing. With the exteins now linked, the intein cuts free in the final catalyzed step as its last residue, commonly asparagine, cleaves at the adjacent downstream splice junction, yielding an intein-succinimide and a new protein. The ligated and newly liberated exteins undergo rapid, spontaneous ester-to-amide rearrangement (6), finish folding, and proceed to function normally as if nothing happened. However, how did that happen? In PNAS, Liu et al. (7) focus on the final catalyzed step of protein splicing: resolution of the branched intermediate via succinimide formation (below). Three central questions have dogged the field: how is asparagine activated as a nucleophile; how is the backbone peptide bond rendered scissile; and finally, how does the intein coordinate this chemical step with the preceding steps? Liu et al. pursue what would seem to be the most straightforward, and daring, experimental endgame: direct observation of the BI. In almost any multistep reaction, characterizing intermediates proves to be informative, yet such studies often stumble because intermediates are by nature fleeting. Protein splicing poses additional challenges. Inteins are autocatalytic, thus bond breaking begins as soon as the intein folds into its native structure, and chemical intermediates generated will be processed before high-resolution structures can be trapped—or sometimes before the precursor is even purified! In lieu of antagonists or effective trapping strategies (8), the authors turned to expressed protein ligation (EPL). EPL is an intein-based method where engineered proteins with unnatural modifications can be prepared semisynthetically in a controlled stepwise manner (9). Two polypeptide fragments comprising the branched intermediate were thus prepared separately and then reassembled covalently in vitro by EPL. The larger of the two fragments was obtained using standard recombinant techniques; the intein