Protein biochemistry: Finding the Perfect Pair

signaling update home

SOURCE:

A new assay for the generation and screening of fragmented proteins promises to considerably accelerate the process of developing novel split-protein sensors for the in vivo characterization of protein interaction.

The venerable two-hybrid assay has proven to be a powerful, if flawed, means for identifying and confirming interactions between pairs of proteins in vivo. In this screen, a transcription factor is broken into two pieces—the DNA-binding domain and the transcriptional activation domain—each of which is then fused to a protein of interest and expressed in yeast. If the two proteins interact, functionality is restored to the transcription factor, activating expression of a gene that enables clone selection.

Unfortunately, this assay has a number of limitations, including the possibility of unexpected transactivation by the peptide linked to the DNA-binding domain, even in the absence of interaction with a partner. Furthermore, the two-hybrid screen can only be used to identify interactions that can be recreated within the environment of the yeast nucleus.

Kai Johnsson and his colleagues at the École Polytechnique Fédérale de Lausanne (Lausanne, Switzerland) and the Forschungszentrum Karlsruhe (Karlsruhe, Germany) were interested in developing alternative ‘split-protein’-based interaction sensors with fewer inherent restrictions. Previous work from a decade ago identified a split ubiquitin protein that could monitor protein interactions in the cytoplasm and at the membrane (see Johnsson and Varshavsky); when the proteins linked to the two fragments interact, the ubiquitin can properly assemble and functionality is restored. Although this sensor pair was effective, the process of derivation was technically demanding, and Johnsson’s team now sought to simplify this process, developing a practical method for the generation and characterization of new and effective sensor pairs.

They used Trp1p, a small (25 kDa) monomeric enzyme involved in the biosynthesis of tryptophan, to develop their screen. Trp1p structure has been well studied, and previous work has demonstrated that protein function can be maintained even after sectioning the enzyme into two fragments. In addition, the use of tryptophan selection increases the ease with which interacting clones can be identified, as a number of trp-deficient yeast strains exist that would be suitable as a background for a split-Trp1p sensor.

Johnsson’s team began with the permutation of the Trp1p sequence, using DNase to randomly cleave Trp1p sequences that had either been circularized or concatamerized. Following size selection, the fragments were cloned into a vector that appended two polypeptides, C1 and C2, onto the newly generated N and C termini of the permutated Trp1p sequence. C1 and C2 are antiparallel coiled-coil sequences designed to mediate interaction. Homologous recombination was then used to insert an additional sequence containing a promoter between the original N and C terminal ends of the gene. Each resulting clone generated thus contains the means to express two peptides: a C-terminal Trp1p fragment with the C1 sequence at its N terminus, and an N-terminal Trp1p fragment with the C2 sequence at its C terminus.

The resulting library was transformed into a Trp-deficient yeast strain, EGY48, yielding an estimated 1600 clones, of which five contained all necessary sequences in-frame and proved capable of complementing tryptophan auxotrophy. Three of these clones were found to require the presence of both fragments and of the interacting C1/C2 polypeptide sequences in order to restore yeast growth, indicating that they might yield useful split-Trp sensors.

Johnsson’s group proceeded to use these split-Trp clones to characterize the interactions of the yeast membrane proteins Sec62p and Sec63p, which are known to associatein vivo. Using the same auxotrophic yeast strain, growth in the absence of tryptophan was restored when the two proteins were attached to the two fragments of split-Trp⁴⁴, one of the clones that had been previously identified as a suitable sensor. No growth was observed when an alternative Trp1p fragment pair was used, or when split-Trp⁴⁴ was fused to Sec62p and Ste14p, which do not interact.

The authors are encouraged both by the potential of these newly identified Trp1p split-protein sensors, and by the apparent effectiveness of this new method for developing additional sensors. Theoretically, the authors indicate, "the introduced combinatorial approach should be able to generate split-protein sensors of almost any protein, thereby yielding tailor-made sensors for different applications."

Michael Eisenstein

References

Tafelmeyer, P. , Johnsson, N. and Johnsson, K. Transforming a (/)₈-Barrel Enzyme into a Split-Protein Sensor Through Directed Evolution. Chem. Biol. 11, 681–689 (2004). | PubMed |
Johnsson, N. and Varshavsky, A. , Split Ubiquiting as a Sensor of Protein Interactions In Vivo . Proc. Natl. Acad. Sci., USA 91, 10340–10344 (1994). | PubMed |

Links

Johnsson Lab Homepage