A peptide (red) spliced from FGF-5 (bottom) induces an immune response.


Eukaryotic cells can make several protein variants from one gene via RNA splicing. Now, a chance encounter with an antigen suggests to Ken-ichi Hanada, Jonathan Yewdell, and James Yang (National Institutes of Health, Bethesda, MD) that vertebrate cells might get even more mileage from the genome through protein splicing.

Yang's group shows that protein splicing produces a fragment of the FGF-5 protein that is recognized as antigenic by human T cells. T cell recognition was stimulated by a peptide as short as nine residues, as long as it contained two short sequences normally separated within full-length FGF-5 by 40 amino acids. Production of the fusion is posttranslational, as untreated cells could process longer synthetic peptides into active antigens, but lightly fixed cells (which are unable to do their own proteolytic processing) could present only an already spliced peptide.

Protein splicing probably takes place in the cytoplasm before the fused product is transported to the ER for presentation. The proteasome, which is normally required for antigen presentation and is a highly proteolytic entity, may be responsible for the splicing activity. “Maybe enzymes that cut proteins also ligate them,” says Yang.

The group is putting their findings to use in developing vaccines against cancerous kidney cells, which express a lot of FGF-5. Yang says that other investigators have recently identified additional antigenic epitopes generated by protein splicing, so FGF-5 is not unique. But he is interested in whether spliced proteins have structural or enzymatic functions that differ from the parent proteins. The study of antigen presentation “is the most sensitive way there is of detecting specific peptide sequences,” says Yang. “But it may be a more common phenomenon. Now, we need to know if it has a functional significance.” ▪


Hanada, K.-I., et al.