Genetic visualization of protein interactions harnessing liquid phase transitions

Protein-protein interactions (PPIs) are essential components of cellular function. Current fluorescence-based technologies to measure PPIs have limited dynamic range and quantitative reproducibility. Here, we describe a genetically-encoded PPI visualization system that harnesses the dynamics of condensed liquid-phase transitions to analyze protein interactions in living cells. The fluorescent protein Azami-Green and p62-PB1 domain when fused to PPI partners triggered a rapid concatenation/oligomerization process that drove the condensation of liquid-phase droplets for real-time analysis of the interaction with unlimited dynamic range in the fluorescence signal. Proof-of-principle studies revealed novel insights on the live cell dynamics of XIAP-Smac and ERK2-dimer interactions. A photoconvertible variant allowed time-resolved optical highlighting for PPI kinetic analysis. Our system, called Fluoppi, demonstrates the unique signal amplification properties of liquid-phase condensation to detect PPIs. The findings introduce a general method for discovery of novel PPIs and modulators of established PPIs.

introduced according to our protocols as described previously 46 to generate FKBP12(F36V) and PB1(D67A, D69R). The T2A sequence encoding GRGSLLTCGDVEENPGP was used for the equimolar expression of two constructs 47 . cDNA cloning. The stony coral Scolymia vitiensis was acquired from the sea off the Aka Island (Okinawa). Total RNA was isolated from the coral by guanidine thiocyanate extraction. Synthesis, amplification of the fragment of interest using degenerate primers, and generation of full-length cDNA were performed as previously described. The degenerate primers used were as follows: 5'-ATCAAGNTNWRYATGGAAGG-3' and 5'-ACVGGDCCATYDGVAAGAAARTT-3' (R = A or G; Y = C or T; V = A, C, or G; and D = A, G, or T). The cDNA encoding the protein-coding region was amplified using primers containing 5'-BamHI and 3'-EcoRI sites. The restricted product was cloned in-frame into the BamHI/EcoRI sites of pRSET B (Thermo Fisher Scientific) for bacterial expression. The 5' end of the gene was modified by PCR to contain a Kozak consensus sequence (CCACCATG) after the BamHI site to promote efficient expression in mammalian cells.   Image acquisition and analysis were performed by using MetaMorph software (version 7.8.9.0) (Molecular Devices).

Dose-response relationship studies (Nutlin-3) by high-content analysis.
Stable transformant cells were seeded into poly-D-lysine black-wall 96-well plates (Corning) at a density of 40,000 cells/well and incubated for 20 hours at 37 °C in a 5% CO 2 atmosphere. After treatment with nutlin-3 for 30 minutes at room temperature, cells were fixed with 4% paraformaldehyde (PFA) for 10 minutes, and then stained with Hoechst33342 (Dojindo) for 30 minutes. Image acquisition was performed using IN Cell Analyzer 2200 (GE Healthcare). A 10× objective lens was used so that each field of view contained > 2,206 cells. For each concentration of nutlin-3, three wells were used. Image analysis was performed using IN Cell Investigator software (GE Healthcare). Green (AG) and blue (Hoechst33342) fluorescence images were used to monitor cytoplasmic puncta and nuclei, respectively (see Fig. 3a); they were segmented automatically using the granular and nuclear segmentation algorithms, respectively. P.I. values were calculated by dividing the total green fluorescence by the number of nuclei (cells). The averaged P.I. values were normalized to that at the lowest concentration and plotted (Fig. 3b). Curve fitting was performed using KaleidaGraph (Synergy Software).

Dose-response relationship analysis (Nutlin-3) using a plate reader.
Stable transformant cells were seeded onto poly-D-lysine black-wall 96-well plates (Corning) at a density of 20,000 cells/well and incubated for 20 hours at 37 °C in a 5% CO 2 atmosphere. After treatment with nutlin-3 for 30 minutes at room temperature, cells were fixed and permeabilized with 0.75% PFA and 2% Triton-X 100 for 15 minutes, and then stained with Hoechst33342 (Dojindo) for 30 minutes. Representative images ( Fig. 3c) were acquired using IX71 (Olympus). Fluorescence intensities of AG (green) and Hoechst33342 (blue) were measured using a Wallac Arvo HTS 1420 Multilabel Counter (Perkin-Elmer). For each concentration of nutlin-3, three wells were used. P.I. values were simply calculated by dividing the green signal intensity by the blue signal intensity. The averaged P.I. values were normalized to that at the lowest concentration and plotted (Fig. 3d). Curve fitting was performed using KaleidaGraph (Synergy Software).

Dose-response relationship studies (AT-406) by high-content analysis.
Stable transformant cells were seeded onto poly-D-lysine black-wall 96-well plates (Corning) at a density of 40,000 cells/well and incubated for 20 hours at 37 °C in a 5% CO 2 atmosphere. After treatment with AT-406 for 15 minutes at room temperature, cells were fixed with 4% PFA for 10 minutes, and then stained with Hoechst33342 (Dojindo) for 30 minutes. Image acquisition was performed using Cell Voyager CV7000 (Yokogawa Electric Corporation). A 20× objective lens was used so that each field of view contained > 2,679 cells. For each concentration of AT-406, four wells were used.
Image analysis was performed using Cell Voyager analytical software (Yokogawa Electric Corporation). Green (AG) and blue (Hoechst33342) fluorescence images were used to monitor cytoplasmic puncta and nuclei, respectively (see Fig. 3f); they were segmented automatically using the granular and nuclear segmentation algorithms, respectively. P.I. values were calculated by dividing the total green fluorescence by the number of nuclei (cells). The averaged P.I. values were normalized to that at the lowest concentration and plotted (Fig. 3g). Curve fitting was performed using KaleidaGraph (Synergy Software).  (a) Temporal montage of rapamycin-induced development of fluorescent puncta in HeLa cells coexpressing PB1-FKBP and FRB-AG. Arrowheads indicate two puncta that collided with each other and then fused. Scale bar, 1 mm. (b) Visualization of rapamycin-induced association between FRB and FKBP in HeLa cells required the homo-oligomerizing capability of both AG and PB1. One day posttransfection of PB1-FKBP/FRB-AG (top), PB1-FKBP/FRB-mAG1 (middle), and mPB1 (monomeric PB1 with two mutations: D67A and D69R)-FKBP/FRB-AG (bottom), cells were imaged before (-5 min) and after (5 min) the addition of 100 nM rapamycin. The fraction of punctum-carrying cells in examined cells for the above three transfections was calculated. Data are shown as mean ± SEM (n = 4 or 5). Scale bars, 10 mm. (c) Segmentation of fluorescent puncta for calculation of P.I. Segmentation was performed using 'Spot detector' of the ICY open bioimage informatics platform. Scale bar, 10 mm. (d) Application of Fluoppi to PPI occurring inside the nucleus. Representative fluorescence images of HeLa (left) and Cos-7 (right) cells expressing PB1-p50/AG-p65 (upper) and PB1/AG-p65 (lower). Images were acquired one day post-transfection. Punctum formation was observed inside the nucleus depending on the association between p50 and p65. Scale bars, 10 mm. (e) Application of Fluoppi to PPI occurring beneath the plasma membrane. Representative fluorescence images of HEK293 cells expressing PB1-KRas(wild type)/AG-cRaf (upper), PB1-KRas(S17N)/AG-cRaf (middle), and PB1-KRas(G12D)/AG-cRaf (bottom). Cells were imaged before (-0.5 min) and after (8 min) the addition of 50 ng/mL epidermal growth factor (EGF). Punctum formation was observed on the plasma membrane depending on the EGF-induced association between KRas and cRaf. The fraction of punctum-carrying cells in examined cells under the five conditions was calculated. Approximately 20 cells were observed in quintuplicate, and data are shown as mean ± s.e.m (n = 5 or 9). Statistical significance (*p < 1E-9) was examined by Bonferoni's multiple comparison test. Scale bars, 10 mm.      Observation time (min:sec). Rapamycin (20 nM) was added at t = 00:00. Fig. 2b is a temporal montage made from this movie.  Fig. 5c is a temporal montage made from this movie.

Supplementary
Supplementary Video 8 | Phase separation due to association between HRas and cRaf. Visualization of 2D phase separation on the plasma membrane. Micron-sized structures appeared depending on the association between HRas and cRaf induced by EGF stimulation and PB1polymerization. Cos-7 cells co-expressing PB1-HRas and AG-cRaf (left) and mPB1-HRas and AG-cRaf (right) were time-lapse imaged by TIRF microscopy. Imaged every approximately 10 seconds. Observation time (min:sec). EGF (50 ng/mL) was added at t = 00:00. Fig. 5d and Fig. 5e show representative images made from these two movies (left and right, respectively).