Multiplex Imaging of Rho GTPase Activities in Living Cells

Abstract

Förster resonance energy transfer (FRET) biosensors are popular and useful for directly observing cellular signaling pathways in living cells. Until recently, multiplex imaging of genetically encoded FRET biosensors to simultaneously monitor several protein activities in one cell was limited due to a lack of spectrally compatible FRET pair of fluorescent proteins. With the recent development of miRFP series of near-infrared (NIR) fluorescent proteins, we are now able to extend the spectrum of FRET biosensors beyond blue-green-yellow into NIR. These new NIR FRET biosensors enable direct multiplex imaging together with commonly used cyan-yellow FRET biosensors. We describe herein a method to produce cell lines harboring two compatible FRET biosensors. We will then discuss how to directly multiplex-image these FRET biosensors in living cells. The approaches described herein are generally applicable to any combinations of genetically encoded, ratiometric FRET biosensors utilizing the cyan-yellow and NIR fluorescence.

Keywords: FRET biosensor; Multiplex imaging; Near-infrared fluorescent protein; Rho GTPases.

Fig. 1

Schematic drawing of the inverted microscope used for live cell, multiplex imaging of Rho GTPase activities. The microscope as depicted is configured for simultaneous CFP-YFP FRET biosensor imaging and DIC/NIR imaging setup. Filterwheel 1: switches the neutral density filters; Filterwheel 2: switches the excitation band-pass filters; Filterwheel 3: switches the emission band-pass filters. Microscope is equipped with Olympus Zero Drift Compensation (ZDC) autofocus mechanism using a 794 nm laser source. The main fluorescence turret of the microscope is equipped with an 80/20 (transmittance/reflection) mirror (Chroma). Details of the construction and specification of this microscope can be found elsewhere [9]

Fig. 2

Multiplexing of NIR FRET Rac1 and CFP-YFP FRET RhoA biosensors, reproduced from Shcherbakova et al. 2018 Nature Chemical Biology [1]. (a) Cartoon of the two biosensors multiplexed in single living cells. The NIR Rac1 biosensor uses miRFP670 [2] and miRFP720 [1] as the FRET pair, modulated by activity of C-terminally attached full length wildtype Rac1 GTPase interacting with a p21-binding domain (PBD) derived from p21 activated kinase 1. Affinity of the PBD toward Rac1 interaction is tuned appropriately by incorporating a dimerization-based autoinhibitory motif through a 2nd PBD containing GTPase-binding deficient mutations [12]. The NIR Rac1 biosensor can be imaged together with any cyan-yellow FRET-based biosensors including that for RhoA also depicted here [3]. (b) The excitation spectra of miRFP670 (FRET donor) and miRFP720 (FRET acceptor) are shown together with the bandpass filter used to excite the cyan-yellow FRET biosensor (magenta shaded region, ET436/20× Chroma Technology). The spectra are normalized to the peaks at ~400 nm at the “Soret” band. (c) Bleedthrough characterization between different channels of the microscope used for multiplex imaging. Top-Left: Purified Venus FP fluorescence. Top-Right: Purified Cerulean FP fluorescence. Bottom-Left: Purified miRFP670 fluorescence. Bottom-Right: Purified miRFP720 fluorescence. In all panels, the first data point (yellow square) is the vehicle control (water) where no FP was included. Intensity range useful for live-cell imaging is shown as a diagonally shaded box

Fig. 3

Western blot showing the inducible expression of RhoA biosensor in MDA-MB231 tet-OFF cell line, detected using anti-RhoA GTPase antibody. Lane 1: Cell lysate from cell culture maintained with 2 μg/mL Dox; Lane 2: Cell lysate from biosensor induced cells that underwent single cycle of trypsinization; Lane 3: Cell lysate from biosensor induced cells that underwent two cycles of trypsinization. Endogenous RhoA expression levels are similar in all cases

Fig. 4

Alignment of multiple imaging channels based on a priori calibration. Top panels: A field of multispectral beads, taken using the two cameras for CFP and cyan-yellow FRET channels (Cameras 1 and 2, as depicted in Fig. 1). The red arrows indicate the point used for the manual alignment of two fields of view. White bar = 20 μm. Middle left: Centroid locations of the calibration beads from the top panels as shown were extracted (Matlab routines incorporating particle-tracking methods by J.C. Crocker and D.G. Grier [24] were used) and overlaid, indicating a significant misalignment as a function of the location within the field. Middle right: The morphed alignment indicates correction applied to the FRET channel field of view to bring the two channels into register. The red arrows show the corresponding positions indicated in the top panels. Bottom panels: Zoomed in views of the data overlay showing the original and the morphed bead images. The zoomed region is indicated in the middle panels. White bar = 20 μm

Fig. 5

Grid-based alignment of side port channel (CFP) to the bottom port views. Comparison of the original unmodified bottom port view (Top left, Camera 3 in Fig. 1) shows significant deviation from the reference image obtained from the CFP channel (Top right) at the side camera (Camera 2 in Fig. 1). Coordinate transformation based on manual determination of control points within the grid image brings the bottom port view partially into register (Bottom left). Additional X-Y linear translation shift (−51 pixels in X and + 3 pixels in Y) (Bottom right) was required to fully register the bottom port view against the reference CFP channel image (Top right). Red bar = 20 μm

Fig. 6

Representative, timelapse panels of RhoA and Rac1 activities imaged in a single living MEF, reproduced from Shcherbakova et al. 2018 Nature Chemical Biology [1]. Top panels: Differential interference contrast. Middle two panels: Rac1 activity (NIR FRET Rac1 biosensor; Upper panels) and RhoA activity (CFP-YFP FRET RhoA biosensor; Lower panels). Bottom panels: Localizations of high Rac1 (yellow) and high RhoA (blue) activities are overlaid, where regions of colocalization are shown in white. Regions of high Rac1 and RhoA activities were defined by intensity thresholding the top 2.5% of pixel ratio values within the image intensity histogram. Regions and features of interest are shown using matching colored arrowheads. White bar = 20 μm. Pseudocolor bar corresponds to ratio limits of 1.0 to 1.55 for Rac1 and 1.0 to 1.32 for RhoA activities (black to red)