LIVE-PAINT: Super-Resolution Microscopy Inside Live Cells Using Reversible Peptide-Protein Interactions

Curran Oi, Zoe Gidden, Louise Holyoake, Owen Kantelberg, Simon Mochrie, Mathew H. Horrocks, Lynne Regan

Posted on: 11 June 2020

Preprint posted on 15 May 2020

Article now published in Communications Biology at http://dx.doi.org/10.1038/s42003-020-01188-6

No photobleaching, no photoconversion – a novel super-resolution technique for live-cell imaging

Selected by Xenia Meshik

Categories: bioengineering, biophysics, cell biology, synthetic biology

Background

Traditional microscopy is limited by the diffraction limit of light and is therefore capable of around 250 nm resolution. As various cellular processes are studied in more and more detail, there has been much interest in developing convenient and affordable super-resolution techniques. Some techniques, like photoactivation localization microscopy (PALM) and stochastic optical reconstruction microscopy (STORM) are based on sequential photoactivation of fluorophores. These techniques are often limited by photobleaching of fluorophores over time. Stimulated emission depletion microscopy (STED) is based on selective deactivation of fluorophores and requires specialized instrumentation, making it impractical for many researchers. Point accumulation for imaging in nanoscale topography (PAINT) is based on transient binding of fluorescent molecules to the protein of interest, and DNA-PAINT increases the specificity of this technique by fusing complementary DNA strands to the fluorophore and the protein of interest. However, DNA-PAINT cannot be used in live cells. In this work, the authors overcome this limitation with LIVE-PAINT, a technique that allows for super-resolution imaging of dynamic proteins in live cells.

Key findings

The authors demonstrate the effectiveness of the LIVE-PAINT, which is based on the reversible binding of a non-functional protein, which is fused to a fluorescent protein (FP), and a small peptide, which is fused to the protein of interest. The protein-FP complex diffuses freely in the cytosol and is temporarily immobilized as it binds to the peptide-protein of interest complex. These localization events result in transient fluorescence spikes as the cell is imaged continuously in the plane of interest. The sum of acquired images results in a super-resolution image. The expression level of the components can be tuned based on the abundance of the protein of interest in the cell to achieve the optimal resolution.

LIVE-PAINT has a number of advantages over other super-resolution techniques. It does not rely on sequential photoswitching of fluorophores, like PALM and STORM, and does not require special instrumentation, like STED. It is not affected as much by photobleaching, since any bleached FPs diffuse away upon dissociation and are rapidly replaced by unbleached FPs. This allows for the sample to be imaged for longer periods of time. Because the protein of interest is not fused directly to a large FP, it more closely resembles its endogenous state. The authors showed that the peptide-binding protein can be fused to as many as 3 FP to achieve a better signal-to-noise ratio.

Using this technique, the authors were able to image the proteins Cdc12, actin, and cofilin in live yeast with a resolution as low as 50 nm and observe these proteins’ dynamic movement.

What I like about the paper

I like this paper because it presents a practical super-resolution technique that seems easily achievable for any lab that has the capability to perform confocal or TIRF imaging. It does not require the purchase of specialized equipment or specialized fluorescent proteins, and the necessary peptide-protein and protein-FP fusions can be achieved with standard molecular biology techniques. It seems particularly well-suited for live imaging of dynamic cellular processes.

Future directions and questions

It would be interesting to see how well this technique works with transient transfections, where controlling the expression levels of various constructs in a cell is more difficult.
One potential disadvantage of this technique is that it is not possible to visually confirm the expression of the protein of interest in the cell, since it is not tagged with a FP. This may make it difficult to troubleshoot the experiment if it does not immediately work. Do the authors anticipate this being a problem?
I wonder how much initial optimization would be required when first using LIVE-PAINT to image a particular protein. A protein of interest that moves rapidly throughout the cell would necessitate a protein-peptide pair that associates and dissociates very rapidly (high Kd value). Do the authors anticipate that one protein-peptide pair would be fairly well-suited for most applications, or would the user need to try several pairs?

Tags: microscopy, super-resolution

doi: https://doi.org/10.1242/prelights.21865

Read preprint

(No Ratings Yet)

Author's response

The author team shared

Many thanks for your excellent summary of our method.

We are currently setting up LIVE-PAINT in mammalian cells, working with the UK Centre for Mammalian Synthetic Biology, at the University of Edinburgh, who are experts in genome manipulation. Our approach is to integrate all the components into the chromosome, because of the issue you mention with transient transfections. Chromosomal integration will reduce variability and the cell line with the integrations will be available for many future experiments.

We do not anticipate that the protein-of-interest not being directly fused to an FP will be an issue. If we cannot detect the protein-of-interest via interaction with the labelling FP, using traditional fluorescence microscopy, we could investigate further to determine if it’s an issue with expression or with peptide tag accessibility. Note that when we fuse a peptide to an essential protein, replacing the endogenous protein, the existence of viable cells confirms expression of the fusion protein and its incorporation into functional complexes.

We are currently investigating how different interaction pairs and different expression levels of the FP are best suited to labelling endogenous proteins of different abundance and behaviour. In the manuscript, we show how the protein CDC12 can be imaged using different peptide-protein pairs and different FP expression levels. All combinations work, but some give better signal:background than others.

Going forward, we will publish and make available the peptide-protein pairs and expression levels that are best suited for imaging different abundance cellular proteins. Thus, other researchers will be able to choose the best starting combination and conditions for their application, in either yeast or mammalian cells, and we anticipate that minimal additional tweaking will be necessary.