Controlling proteins the intein way

Abstract

Several methods have been described recently that permit researchers to exert post-translational control over gene products. An intein-based method shows it is a strong contender in the field by demonstrating control of protein activity in a whole animal.

Main

Inteins are curious protein elements that can excise themselves out of a protein and reseal the protein behind them as if they had never been there. Although they are described as parasitic, inteins can function in a split form with each half present on the end of a different protein. The split inteins then ligate such proteins together, creating a new protein with novel functions. This ability has made them an attractive tool for protein engineering.

Tom Muir of Rockefeller University and his colleagues have been keenly focused on exploiting inteins for the purposes of targeted protein engineering and control. They previously described the creation of rapamycin-inducible split inteins for regulated protein ligation (Mootz & Muir, 2002). Muir says, “We'd initially thought of this as a binary switch, but as we got into it more it became clear to us that it had the potential to be a rheostat.” As most biological processes result from graded changes in protein function over time, they thought that this system could be a useful method for controlling protein function in vivo in a lifelike manner.

As validation of this idea, Muir and his colleagues split the bioluminescent enzyme luciferase into halves and attached their drug-inducible intein domains (Schwartz et al., 2007; Fig. 1). They expressed the proteins in cells and observed that addition of rapamycin resulted in strong dose-dependent activation of luciferase activity. It worked so well they decided to try it in an animal. Muir says, “It was actually my student Ed Schwartz who bravely took that big leap and moved it into the fly system. We were very happily surprised how straightforward it was.” By feeding the fly food laced with rapamycin they could switch luciferase activity on within minutes.

Figure 1: Inducible protein splicing.

The rapamycin-binding proteins FKBP and FRB are fused to split intein domains VMA^N and VMA^C connected to the N- and C-terminal pieces of a split luciferase. Addition of rapamycin induces protein splicing, removal of the intein and release of full-length luciferase.

It turns out that the rate-determining step in this technology is getting it to work initially in culture by finding the right points in the target protein to split the molecule. Muir suggests that going forward it may be possible to develop some empirical rules to simplify the process, but there will always be an element of trial and error requiring careful testing and validation. To justify this effort you have to have a problem that can only be addressed by this type of method.

What kinds of applications are these new methods appropriate for? Most likely it will be ones that require titration of protein levels, something that is very difficult to achieve with transcriptional regulation. Muir believes this is going to be the 'killer app' for this kind of technology. “Any system where the protein levels oscillate as a function of some sort of intrinsic or extrinsic control [is] worth exploring with this kind of small molecule–regulated system,” says Muir.

“What we'd really like to do with this technology is adapt it such that we get away from small molecules altogether,” says Muir. “One thing we are looking at carefully in the lab is trying to control this with light. That would ultimately provide the best level of control.” Muir is unwilling to make any bets on which of the recent methods developed to regulate proteins at the post-transcriptional level will ultimately prove the most useful but you can bet his intein-based methods will be contenders.