Putting a Spotlight on Protein Folding

Failure to fold properly is a major obstacle to protein bioengineering efforts. A rather large fraction of proteins require some help in folding—chaperones, isomerases, and various other helpers—and, lacking that, end up as insoluble masses in bacterial inclusion bodies. While some are working on ways to resolubilize misfolded proteins, or ways to include all the necessary factors to get proper folding, Geoffrey Waldo and co-workers at Los Alamos National Laboratory (Los Alamos, NM) have come up with an assay that allows you literally to see what's going on with your protein. By attaching green fluorescent protein (GFP) to the protein under production, they have shown that following GFP fluorescence will lead you to a correctly folded protein. The work, which appears in the July 1 issue of Nature Biotechnology, was described by Jonathan King and Scott Betts (MIT) in an accompanying editorial as both "surprising and exciting."

Working with a collection of 20 randomly selected proteins from the thermophilic bacteria Pyrobaculum aerophilum, the researchers engineered proteins as N-terminal fusions with GFP. When expressing these fusions in bacteria, they found up to a 50-fold difference in the level of fluorescence, while only a 20% difference, at most, in the amount of protein made. The fluorescence intensity correlated with the amount of expressed protein in the supernatant, and failure of the fluorochrome to form with the presence of the protein in inclusion bodies.

But the pièce de résistance was yet to come. The researchers used a directed evolution strategy in combination with GFP screening to isolate protein variants that fold correctly, are soluble, and retain activity. Starting with an insoluble wild-type ferritin, successive rounds of evolution gave rise to a fully functional form.

This "folding reporter" vector method lends itself to large scale screening, by using fluorescence activated cell sorting, for example, to isolate individual fluorescing cells. In addition, the GFP-fusion approach can be used on completely uncharacterized proteins or proteins with unknown functions—all good news for those in the business of identifying novel genes.

For more information: Geoffrey S. Waldo, Structural Biology Group, MS-M888 Los Alamos National Laboratory, Los Alamos, NM 87545. Tel: 505-667-8161. Fax: 505-665-3024. Email: waldo@telomere.lanl.gov.

By Laura DeFrancesco