The biarsenical dyes FlAsH and ReAsH bind specifically to tetracysteine motifs in proteins and can selectively label proteins in live cells. Recently this labeling strategy has been used to develop sensors for different protein conformations or oligomeric states. We describe the labeling approach and methods to quantitatively analyze binding.
Fluorescent proteins and dyes are essential tools for the study of protein trafficking, localization and function in cells. While fluorescent proteins such as green fluorescence protein (GFP) have been extensively used as fusion partners to proteins to track the properties of a protein of interest1, recent developments with smaller tags enable new functionalities of proteins to be examined in cells such as conformational change and protein-association 2, 3. One small tag system involves a tetracysteine motif (CCXXCC) genetically inserted into a target protein, which binds to biarsenical dyes, ReAsH (red fluorescent) and FlAsH (green fluorescent), with high specificity even in live cells 2. The TC/biarsenical dye system offers far less steric constraints to the host protein than fluorescent proteins which has enabled several new approaches to measure conformational change and protein-protein interactions 4-7. We recently developed a novel application of TC tags as sensors of oligomerization in cells expressing mutant huntingtin, which when mutated aggregates in neurons in Huntington disease 7. Huntingtin was tagged with two fluorescent dyes, one a fluorescent protein to track protein location, and the second a TC tag which only binds biarsenical dyes in monomers. Hence, changes in colocalization between protein and biarsenical dye reactivity enabled submicroscopic oligomer content to be spatially mapped within cells. Here, we describe how to label TC-tagged proteins fused to a fluorescent protein (Cherry, GFP or CFP) with FlAsH or ReAsH in live mammalian cells and how to quantify the two color fluorescence (Cherry/FlAsH, CFP/FlAsH or GFP/ReAsH combinations).
1. Preparation of cells for labeling with ReAsH/FlAsH
Note it is important to use positive and negative controls to assess the extent of specific binding to the TC tags and to assess for bleedthrough of fluorescence between the channels when collecting confocal micrographs. Hence, for two colors (eg FlAsH/Cherry or ReAsH/CFP or ReAsH/GFP combinations), ensure samples are prepared for single colors (eg Fluorescent protein alone or if possible a TC-tagged protein bound to FlAsH/ReAsH but with no fluorescent protein)
It is important to add EDT first before adding FlAsH/ReAsH and make the buffer just prior to adding it to the cells. Incubate for exactly 30 min at 37 °C in tissue culture incubator. In our experience longer incubation times significantly increases the background fluorescence. New constructs should also be optimized for labeling time and FlAsH/ReAsH concentration (0.5-2 μM).
After this wash, the cells may be fixed with paraformaldehyde (15 min with 3.2% solution), although we have found that this increases non-specific biarsenical dye fluorescence. Hence we usually image the cells live at room temperature. (Note that fixation of cells before labeling prevents biarsenical dye binding.)
2. Imaging the cells on a confocal microscope
3. Analysis of the data
4. Representative results:
The success of labeling cells with biarsenical dyes is dependent on a few key parameters. First, the timing of the labeling with dyes is crucial. We have found that extended periods of labeling (more than 30 min) results in a high level of non-specific background staining. Fig 1 shows a typical result for a wild-type form of huntingtin fragment (25Q) fused to the CFP derivative Cerulean containing a TC tag as described previously 7. This sample was stained for 30 min with ReAsH and there is minimal background in the sample lacking the TC tag. We have found that fixing cells with paraformaldehyde increases background while fixing with methanol abrogates the fluorescence of the fluorescent protein tag. Hence where possible we image the cells live. It is important to also note that fixation prior to labeling with biarsenical dyes prevents their binding, presumably due to modifications of the TC motif.
Another critical factor for consistent results is the density of cells. We have found it critical to image cells that are loosely distributed and also that extensive clumping can lead to uneven staining of the biarsenical dyes in different cells.
Figure 1. Tetracysteine tags and ReAsH staining in live cells transfected with huntingtin(exon1-25Q)-Cerulean fusions. The TC tag is located at the junction of the huntingtin-Cerulean fusion (as described in 7). The pixel intensity correlation plot enables an assessment to differences in ReAsH binding throughout the cell and can be used to map changes in ReAsH binding due to conformational change or ligand interactions.
The approach to label protein localization with a fluorescent protein and conformational properties with a second dye offers much potential for mapping where different conformations of proteins accrue in cells and events that change the dynamics of protein conformation. ReAsH/FlAsH was first used as an in-cell sensor for protein folding of the mammalian cellular retinoic acid-binding protein I 4. In this example, the FlAsH bound to a TC tag engineered into cellular retinoic acid-binding protein I had reduced fluorescence yield in the folded form relative to the unfolded form, and the folding could be tracked in E. coli cells. More recently protein folding and self-association was demonstrated in model proteins by bipartite tetracysteine display 6. In this example, distal dicysteine pairs on model peptides were brought into close proximity upon binding of two peptides containing dicysteine pairs, or by the folding of a peptide that reconstitutes a functional tetracysteine motif for binding biarsenical dyes. We took this approach one step further by developing sensors that distinguish monomers from oligomers due to the TC tag becoming occluded from biarsenical dye binding in all oligomeric forms 7.
Despite the potential of TC tags and biarsenical dyes as reporters for conformational changes and association, the methodology suffers from having a comparatively low signal/noise signal compared to fluorescent proteins as a result of baseline background fluorescence 9. Hence in efforts to develop conformation sensors it would be worthwhile to test some of the other newer labeling approaches being developed, such as an engineered fluorophore ligase system that conjugates coumarin derivates to a 13 amino-acid peptide sequence 3.
The analysis presented here can be extended further to examine where within cells differences occur in the binding of ReAsH/FlAsH to the protein of interest. To do this, the images can be divided into subregions by creating regions of interest (ROI) and converting them to masks to filter the different parts of image (this can all be done in ImageJ). Hence, by comparison of the pixel intensity plots it should be possible to statistically evaluate differences between intracellular subregions using the sensors.
The authors have nothing to disclose.
This work was funded by grants to DMH and TDM (NHMRC project grants). DMH is a Grimwade Fellow, funded by the Miegunyah Trust.
Name of the Reagent | Company | Catalogue Number | Comments |
---|---|---|---|
8-well μ-slides | Ibidi | 80826 | We find these chamber slides to be particularly useful for culturing cells for imaging. |
TC-FlAsH II In-cell Tetracysteine Tag Detection Kit *green fluorescence* *for live-cell imaging | Invitrogen | T34561 (FlAsH) or T34562 (ReAsH) | |
Hanks’ Balanced Salt Solution | Invitrogen | 14175-103 | |
2,3-Dimercapto-1-propanol | Sigma-Aldrich | D1129-5ML | |
1,2-Ethanedithiol | Sigma-Aldrich | 02390-25ML |