2Department of Biology, Faculty of Science, Tohoku University, Sendai, Miyagi, Japan
The bimolecular fluorescence complementation (BiFC) assay is a method for visualizing protein-protein interactions in living cells. To visualize the cofilin-actin interaction in living cells, a series of combinations of the N- and C-terminal fragments of Venus fused upstream or downstream of cofilin and actin were screened systematically. A new pair of split Venus fragments, Venus (1–210) fused upstream of cofilin and Venus (210–238) fused downstream of actin, was the most effective combination for visualizing the specific interaction between cofilin and actin in living cells. This pair of Venus fragments was also effective for detecting the active Ras-dependent interaction between H-Ras and Raf1 and the Ca2+-dependent interaction between calmodulin and its target M13 peptide. In vitro BiFC assays using the pair of purified BiFC probes provided the means to detect the specific interactions between cofilin and actin and between H-Ras and Raf1. In vivo and in vitro BiFC assays using the newly identified pair of Venus fragments will serve as a useful tool for measuring protein-protein interactions with high specificity and low background fluorescence and could be applied to the screening of inhibitors that block protein-protein interactions.
Protein-protein interactions play fundamental roles in coordinating and executing numerous biological processes. To investigate the functional roles and the regulatory mechanisms of protein-protein interactions, several methods have been developed for visualizing protein interactions in living cells, including bimolecular fluorescence complementation (BiFC) (1-5), fluorescence resonance energy transfer (FRET) (6,7), and fluorescence cross-correlation spectroscopy (FCCS) (8). These methods can be used for the screening of the small molecule inhibitors of protein-protein interactions that are candidates for therapeutic drugs for various diseases.
The BiFC method is based on the reassembly of a fluorescent protein from its two complementary nonfluorescent fragments, whose association is facilitated by the interaction between two proteins fused to each fragment (1-5). The BiFC method can be used for visualizing protein-protein interactions in living cells and investigating the subcellular localization and regulation of the protein complex. Compared with FRET, the BiFC method is more sensitive and selective for visualizing protein interactions in living cells. It also can efficiently detect even weak protein-protein interactions because of the remarkable stability of the protein complex after the reassembly of the fluorescent protein fragments (1-5). However, because of this stability and the long time required for fluorophore maturation, the use of the BiFC method is limited in terms of visualizing the real-time dynamics of protein-protein interactions in living cells (1-5).
Cofilin is an actin binding protein with the ability to stimulate severance and depolymerization of actin filaments (9). It binds to both F-actin and G-actin. The actin binding activity of cofilin is inhibited by phosphorylation at Ser-3 by LIM-kinase (LIMK) (10,11) or testicular protein kinase (TESK) (12) and is recovered by dephosphorylation by Slingshot family phosphatases (13). Previous studies have shown changes in the level of cofilin phosphorylation after cell stimulation with growth factors or chemokines and during cell cycle progression (14-16). However, the interaction between cofilin and G-actin has not yet been visualized in living cells.
In this study, we investigated whether the cofilin-actin interaction could be detected in cells by the BiFC method using fragments of Venus, a variant of yellow fluorescent protein (17,18). A new pair of Venus fragments split at 210 enabled the BiFC assay to detect the cofilin-actin interaction in living cells and in cell-free assays. Moreover, this new pair of Venus fragments was applicable to BiFC assays for detecting the specific interactions between H-Ras and Raf1 (19) and between calmodulin (CaM) and M13, a CaM binding peptide derived from myosin light chain kinase (20).Materials and methods Plasmid construction
The expression plasmid coding for Venus was provided by Dr. A. Miyawaki(Riken, Wako, Japan) (17). Plasmids for β-actin, cofilin(WT or S3E), and cyan fluorescent protein (CFP)-tagged LIMK1 were constructed as described previously (10,16,21). The cDNAs for human H-Ras (ID no. BC006499; GenBank), Raf1 (X03484), CaM (BC003354), and M13 (a 26 amino acid CaM binding peptide derived from skeletal muscle myosin light chain kinase, AF325549) (20) were PCR-amplified with a human placenta cDNA library and specific primers. Plasmids for H-RasV12 and H-RasN17 mutants were generated by PCR-based mutagenesis. Plasmids coding for the N-terminal (VN) and C-terminal (VC) fragments of Venus were constructed by PCR amplification. The plasmid cDNAs were substituted for the GFP cDNA in a pEGFPC1 mammalian expression vector (Clontech, Palo Alto, CA, USA) or inserted into a pET17b bacterial expression vector (Novagen, Darmstadt, Germany). To construct plasmids encoding the VN- and VC-fusion proteins, the cDNAs of the interacting proteins (cofilin, actin, H-Ras, Raf1, etc.) were fused upstream and downstream of the cDNA encoding VN or VC fragments by using linker sequences TCTAGA encoding Ser-Lys and GGGAATTCN encoding Gly-Asn-Ser, respectively. For the in vitro BiFC assays, cDNAs for His6-tagged actin, cofilin, H-Ras, and Raf1-RBD were constructed by PCR amplification and subcloned into the pFastBac1 baculovirus expression vector (Invitrogen, Carlsbad, CA, USA). To coexpress VN210-M13 and CaM-VC210 in Escherichia coli, the PCR-amplified cDNAs were subcloned into a pETDuet-1 vector (Novagen).BiFC assay in living cells
HeLa and COS-7 cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. About 5×104 cells were plated on 35-mm glass-bottom dishes, cotransfected with 0.5 µg of each plasmid encoding VN- and VC-fused protein by Lipofectamine 2000 (Invitrogen), and cultured for 24 h. Cells were examined with an inverted fluorescence microscope (model DMIRBE; Leica Microsystems, Wetzlar, Germany) that was equipped with a Plan Apochromat 63×oil immersion objective lens (NA, 1.32) and a yellow fluorescent protein (YFP)-optimized filter set (Omega Optical, Brattleboro, VT, USA). Fluorescence images were captured with an exposure time for 1 min, using a Coolsnap HQ-cooled charge-coupled device (CCD) camera (Roper Scientific, Ottobrunn, Germany) driven by a Q550FW Imaging Software (Leica Microsystems). Screening was performed by dividing the pairs of the plasmids into eight groups as shown in Supplementary Table S1. Each group was analyzed under the same culture and microscopic conditions. Fluorescence intensities of the cells were analyzed by visual observations by eye and categorized to four classes (no, weak, moderate, and strong fluorescence). Experiments were repeated at least twice.