To detect the Ras-Raf interaction in the cell-free system, His6-tagged Ras (V12 or N17)-VC210 and VN210-Raf-RBD fusion proteins were purified from Sf9 cells (Figure 3C) and subjected to the in vitro BiFC assay. Similar to the case of the cofilin-actin interaction, incubation of purified VN210-Raf-RBD with RasV12-VC210 increased the fluorescence intensity, but incubation with RasN17-VC210 did not recover the fluorescence (Figure 3D). Thus, the fluorescence recovery reflects the active Ras-dependent interaction between Ras and Raf-RBD.
The Ca2+-dependent interaction between CaM and its target peptide M13 was previously analyzed by BiFC probes (termed “split-pericam”), composed of the N- and C-terminal GFP fragments that were split at 144 and fused to CaM and M13, respectively (26). In this study, we examined whether the pair of VN210 and VC210 fragments of Venus is useful for detecting the CaM-M13 interaction by coexpressing the pair of CaM-VC210 and VN210-M13 in E. coli. Coexpression of this pair of BiFC probes exhibited strong fluorescence intensity in E. coli, whereas no fluorescence was observed by expression of each probe separately (Figure 4A). To examine whether the fluorescence emission is dependent on Ca2+, lysates of E. coli separately expressing CaM-VC210 or VN210-M13 were mixed in the presence or absence of Ca2+. The fluorescence intensity was increased when lysates were mixed in the presence of Ca2+, but no increase was observed in the presence of EDTA (Figure 4B). These results suggest that the fluorescence recovery reflects the Ca2+-dependent interaction between CaM and M13.
In this study, we applied the BiFC method to the detection of the cofilin-actin interaction. Among the 728 pairs of VN and VC fragments fused to cofilin and actin, the pair of actin-VC210 and VN210-cofilin showed the most distinct difference in fluorescence intensity between cells expressing VN210-cofilin(WT) and VN210-cofilin(S3E). Accordingly, this pair is a useful BiFC probe for visualizing the specific interaction between actin and cofilin in living cells. Cotransfection of LIMK1 with actin-VC210 and VN210-cofilin(WT) reduced the fluorescence in cells, which indicates that the fluorescence intensity reflects the level of cofilin phosphorylating and dephosphorylating activities in the cells.
The pair of VN210 and VC210 fragments was successfully applied to detect the active Ras-specific interaction with Raf1 and the Ca2+-dependent interaction between CaM and M13. Previous studies usually have used fluorescent protein fragments that were split near amino acid residues 144, 154, and 171 (1-5)(18, 26-28); however, the combinations of these Venus fragments fused to actin and cofilin caused no apparent difference in fluorescence intensity between cells expressing cofilin(WT) and cofilin(S3E) probes. The pairs of VN154/VC154 and VN171/VC171 produced fluorescence emission spontaneously by self-assembly without fusion of interacting proteins, but the pair of VN210/VC210 did not. Thus, the new combination of Venus fragments, composed of VN210 and VC210, is a useful candidate BiFC probe for visualizing specific protein-protein interactions that are not detectable by fusing the previously reported combinations of Venus fragments. Multiply-mutated GFP fragments split at 214/215 were previously used for protein tagging (30). The C-terminal short fragment was fused to a protein of interest and used as a reporter of protein solubility and aggregation by measuring the fluorescence recovery after the complex formation with the N-terminal fragment (30). In this case, the highly mutated GFP fragments were selected in terms of their ability to self-assemble, rather than the ability to visualize the interaction of fused proteins.
We also developed an in vitro BiFC assay system for detecting the specific interactions between cofilin and actin and between H-Ras and Raf-RBD using the purified VN210- and VC210-fused proteins. Our results indicate that the in vitro and in vivo BiFC assays using the new pair of Venus fragments, VN210 and VC210, are useful for detecting the complex formation of various pairs of interacting proteins with high specificity and under low background fluorescence conditions. This system will be applicable to the screening of specific inhibitors that block protein-protein interaction. Considering the remarkable stability of the reassembled complex of Venus fragments, careful experimental design is required to use the BiFC assay for the inhibitor screening. Additionally, because the cofilin-actin interaction is blocked by LIMK1-mediated cofilin phosphorylation, our BiFC assay system will provide a useful method to screen for inhibitors of LIMK1.