Combining single-molecule and expansion microscopy in fission yeast to visualize protein structures at the nanostructural level

Abstract

In this work, we have developed an expansion microscopy (ExM) protocol that combines ExM with photoactivated localization microscopy (ExPALM) for yeast cell imaging, and report a robust protocol for single-molecule and expansion microscopy of fission yeast, abbreviated as SExY. Our optimized SExY protocol retains about 50% of the fluorescent protein signal, doubling the amount obtained compared to the original protein retention ExM (proExM) protocol. It allows for a fivefold, highly isotropic expansion of fission yeast cells, which we carefully controlled while optimizing protein yield. We demonstrate the SExY method on several exemplary molecular targets and explicitly introduce low-abundant protein targets (e.g. nuclear proteins such as cbp1 and mis16, and the centromere-specific histone protein cnp1). The SExY protocol optimizations increasing protein yield could be beneficial for many studies, when targeting low abundance proteins, or for studies that rely on genetic labelling for various reasons (e.g. for proteins that cannot be easily targeted by extrinsic staining or in case artefacts introduced by unspecific staining interfere with data quality).

Keywords: Schizosaccharomyces pombe; correlative expansion microscopy; expansion microscopy; photoactivated localization microscopy; protein retention yield; single-molecule localization microscopy.

Figure 1.

Principle and performance of the SExY method. (a) Principle of SExY sample preparation, a correlative ExPALM protocol for expanding yeast. Shown are cell wall (black), cell membrane (grey), inner and outer nuclear membrane (magenta), target protein (blue), FP marker (red) and the PAA gel mesh (orange). The acrylate group of the protein anchoring agent AcX is highlighted in orange. (b) Fission yeast cells expand approximately fivefold using the SExY protocol. (c) SExY provides a macroscale isotropic expansion as determined by two different samples, measuring (i) cytosol widths and (ii) nuclear diameter (Nup132) of non-expanded and expanded cells imaged over a total of nine different experiments. The expansion factor was determined to be 4.9 (s.e. 0.2, s.d. 0.8) (cytosol width) and 4.9 (s.e. 0.2, s.d. 0.9) (nuclear diameter), respectively. Note that nuclear widths were estimated from diffraction limited epifluorescence data and cannot be compared directly to cytosolic widths derived from SMLM data. (d) Additionally, the isotropic microscale expansion of the SExY protocol was probed investigating cell borders. Here, cells expanded with the SExY protocol show no mEos2 localizations being pulled outside of the cell area and hence an isotropic expansion on the microscale. (e) SExY retains 46% (s.e. 3.1%, s.d. 21.8%) of FP signal retention, whereas the original proExM protocol retained 22% (s.e. 2.0%, s.d. 12.4%).

Figure 2.

Protocol optimizations for SExY (a) Examples of incomplete cell wall digestion using zymolyase, snail enzyme, a combination of zymolyase and lysing enzyme and lyticase. All tested conditions lead to a partial and non-isotropic expansion constrained by remaining cell wall. (b) The proteinase activity of proteinase K (red) in comparison to different cell wall digesting enzymes measured by the release of tyrosine during casein digestion. Zymolyase (blue), β-glucoronidase (green), lysing enzyme (yellow) and Lallzyme MMX (magenta). Adding proteinase inhibitor PMSF successfully suppressed tyrosine release. (c) sad1-mScarlet-I signal shows an increase of 27% (s.e. 12.8%, s.d. 85%) in FP retention when adding 0.05% glutathione during gelation. Gel rigidity decreases with increasing glutathione concentration. Thus, no gel formation when adding 0.4% glutathione or higher was achieved.

Figure 3.

Examples of SExY microscopy SExY imaging of high (a–c) and low (d–f) abundant proteins in fission yeast. Super-resolved targets are marked in italic, targets imaged by conventional epifluorescence in regular script. (a) mEos2 expressed in the cytosol, (b) nuclear DNA binding protein cbp1, (c) nuclear protein mis16 and DNA (TO-PRO-3 Iodide), (d) nuclea pore protein Nup132, (e,f) centromeric histone protein cnp1 relative to the spindle pole body protein sad1; combined with visualizing DNA (TO-PRO-3 Iodide) (e, inset i)).