protocol
Optimized sample preparation for single-molecule localization-based superresolution microscopy in yeast Charlotte Kaplan1,3 & Helge Ewers1–3 1Institute of
Biochemistry, ETH Zurich, Zurich, Switzerland. 2Randall Division of Cell and Molecular Biophysics, King’s College London, London, UK. 3Present addresses: Department of Molecular and Cell Biology, University of California, Berkeley, California, USA (C.K.); Institute for Chemistry and Biochemistry, Freie Universität Berlin, Berlin, Germany (H.E.). Correspondence should be addressed to H.E. ([email protected]).
© 2015 Nature America, Inc. All rights reserved.
Published online 11 June 2015; doi:10.1038/nprot.2015.060
Single-molecule localization-based superresolution microscopy methods allow the resolution of cellular structures in the range of tens of nanometers. However, these techniques are of limited use in current yeast labeling protocols, owing to problems with structural preservation. Here we describe an optimized sample preparation protocol that enables single-molecule localization microscopy at high resolution combined with improved structural preservation in Saccharomyces cerevisiae. The protocol uses small binders called nanobodies and an enzymatic labeling strategy to deliver organic dyes to the target protein. These small binders readily penetrate through the yeast cell wall and thus eliminate the requirement for its prior degradation, and they allow structural preservation. In addition, the small size of the binders reduces the distance of the dye to the target protein, and thus it reduces the localization error. The preparation of S. cerevisiae cells for superresolution imaging takes 2–4 h to perform. Researchers should have skills in yeast molecular biology, immunolabeling techniques and access to a microscope equipped for single-molecule imaging.
INTRODUCTION Single-molecule localization microscopy (SMLM) techniques currently allow a resolution down to 10–20 nm. The different approaches to SMLM, photoactivation localization microscopy (PALM1), stochastic optical reconstruction microscopy (STORM2 or dSTORM3) and ground-state depletion and single-molecule return (GSDIM4) all apply the same concept to circumvent the diffraction limit of light. Instead of imaging all molecules in a structure at the same time, every molecule is imaged individually over time and localized with nanometer accuracy. From the refined positions, an image is then assembled. In this way, the organizations of proteins in functional cellular structures such as focal adhesions, nuclear pores and the actin-spectrin cytoskeleton have been resolved on the nanometer scale5–7. In one 3D SMLM study, the protein layers of a focal adhesion could be resolved down to approx. ~20 nm in all three dimensions5. Such studies clearly demonstrate the potential of optical superresolution methods, which allow a nanometer resolution that was previously only attainable with electron microscopy. Current sample preparation protocols in superresolution imaging are generally optimized for the mammalian cell system, with only a small number of papers available for the model organism budding yeast8–11. For a general overview of SMLM methods in yeast, we recommend the review by Mund et al.8. There are two main aspects that distinguish yeast cells from mammalian cells with respect to the application of SMLM imaging. First, mammalian cells in culture generally attach to the coverslip surface via their extracellular matrix proteins and are, in that way, immobilized. An immobilized sample is important for avoiding artifacts in the resulting SMLM data caused by cell movement. The yeast cell wall does not provide naturally adherent molecules that attach to a glass surface, and appropriate surface coating is required for effective immobilization. Second, organic dyes (as they are preferably applied in SMLM) are delivered by antibodies to the targeted structure. Classical immunofluorescence staining in yeast requires degradation of the cell wall, as the immunoglobulins are too large to penetrate the cell wall
by diffusion. This approach can affect the intrinsic protein and organelle organization of yeast, and it prevents optimal structural preservation, which is crucial when imaging structures with nanometer spatial resolution. Here we provide a solution to both problems and a detailed, step-by-step protocol describing the application of SMLM in S. cerevisiae within 2–4 h. SMLM labeling techniques The main limitations in SMLM lie in the brightness (photon yield) of the fluorophore (the greater the fluorophore brightness, the more accurately the emitter’s position can be determined) and the labeling density of the sample. Optimal labeling density is needed to capture all important information and to minimize the risk of misinterpreting the obtained results12. If the labeling density is too low, the spatial resolution will suffer, but if the labeling density is too high nonspecific background binding is likely to cause problems. Genetically encoded photoswitchable fluorescent proteins (FPs) are useful tools for SMLM in yeast8,11. FPs offer key advantages in live-cell superresolution imaging, for example, by providing high specificity for the target structure and low levels of nonspecific background. However, even the most recently developed FPs13 yield markedly fewer photons than organic dyes, resulting in much lower localization precision. An added advantage to using organic dyes for SMLM is that their brightness and blinking properties can be controlled fairly well in situ14,15. A major disadvantage of organic dyes is the difficulty in delivering them specifically, accurately and directly to cellular molecules in yeast cells. Traditional methods use dye-conjugated antibodies to label cell structures in spheroplasts formed by enzymatic digestion of the cell wall combined with permeabilization of the plasma membrane16 (Fig. 1). The use of spheroplasts is necessary for the large (~150 kDa) antibody complex to penetrate the cell and bind to its target. However, spheroplasts are fragile structures that have a shape different from healthy cells and lack their organizational features (Fig. 1a,c,d). These changes in the 3D protein nature protocols | VOL.10 NO.7 | 2015 | 1007
protocol
© 2015 Nature America, Inc. All rights reserved.
FWHM (nm)
Figure 1 | The use of small binders conjugated with organic fluorophores GFP Traditional yeast c d SMLM nanobodies a immunofluorescence renders cell wall digestion unnecessary in yeast immunocytochemistry. (a) Conventionally, yeast immunofluorescence is accompanied by enzymatic Digest cell wall, digestion of the cell wall, as, owing to their size, immunoglobulins cannot permeabilize penetrate it. Cell wall digestion, however, leads to a collapse of cell shape and possibly also of intracellular structures. (b) The smaller epitope-binding domains from the single heavy-chain immunoglobulins termed nanobodies can penetrate the yeast cell wall. This makes cell wall digestion unnecessary. GFP b e f SMLM nanobodies Nanobody staining Nanobodies against GFP are commercially available for delivery of organic dyes to any GFP-tagged molecules from the yeast GFP library. (c–f) Strain YBR127C digest cell wall, permeabilize from the genomic GFP yeast library expressing the vacuolar H+-ATPase pump Vma2 coupled to GFP labeled with AF647-conjugated anti-GFP nanobodies to demonstrate the effect of cell wall digestion on the structure of cellular organelles. (c,d) Digestion of the cell wall leads to structural defects in the vacuole. Vma2 staining is visible as discrete and dispersed aggregates g h 60 Organic dye Organic dye (d, arrowheads). (e,f) Cells stained without cell wall digestion. The reconstructed 50 *** Second SMLM image shows a defined vacuolar outline. (g) The dimensions of a Nanobody 40 antibody nanobody compared with first and second antibody staining of a target Target 30 protein (classical sandwich immunofluorescence). Owing to the size of the antibodies, the dye is delivered at a significant distance from the target First 20 2 nm antibody structure, resulting in additional localization error. The nanobody delivers the αTub Nanobody Target dye to within a few nanometers of the target structure. (h) Quantification of microtubule SMLM by classical sandwich immunolabeling compared with nanobody labeling. Clearly, the full width at half-maximum (FWHM) of the microtubules stained with the anti-GFP nanobody coupled to AF647 is much closer to the actual thickness of a microtubule (green dashed line)17. ***P = 3.8 × 1052 (α-tubulin (αTub) antibody against anti-GFP nanobody) by Mann–Whitney test. Scale bars, 1 µm. All reconstructed superresolution images were visualized with the US National Institutes of Health (NIH) ImageJ look-up table: red-hot.
organization inside the cell can lead to loss or misinterpretation of information in fluorescence imaging. Changes to the structural and spatial protein organization have a major effect on superresolution imaging, as seen in fluorescence superresolution images of the vacuole in yeast cells, with or without cell wall degradation (Fig. 1c–f). As such, traditional immunofluorescence protocols should be avoided for SMLM in yeast. Protocol overview In this protocol, we provide a detailed description of our previously published method for performing the SMLM method dSTORM3,17 using organic dyes, and we make suggestions for achieving optimal preservation and labeling density of each protein or cytoplasmatic structure in budding yeast. The procedure includes a description of the immobilization of yeast cells on coverslips via the lectin concanavalin A (Con A). The proper immobilization of yeast cells is crucial for SMLM, as even small movements of the cell can result in artifacts. After immobilization, the yeast cells are chemically fixed to avoid intrinsic movements and to ensure that the position of intracellular structures is fixed. The yeast cells are then permeabilized and treated with blocking reagents to avoid nonspecific binding of the labeling reagents. Nonspecific background binding can lower the signal-to-noise ratio, reducing localization precision. To avoid the need for cell wall digestion and its associated problems for superresolution imaging, we have optimized two alternative methods for labeling target proteins in yeast cells: nanobody-based labeling and enzymatic labeling. Depending on the protein tag, labeling is performed with either GFP nanobodies or benzylguanine derivatives coupled to an Alexa Fluor 647 (AF647) dye. Nanobodies used in nanobody-based labeling are ~13-kDa hypervariable domains of camelid antibodies18 that can be coupled to fluorescent dyes by N-hydroxysuccinimidyl ester linkage of amino groups (usually on lysines)17 using commercially available reagents. Dye-conjugated nanobodies can penetrate the 1008 | VOL.10 NO.7 | 2015 | nature protocols
cell wall of fixed membrane-permeabilized yeast cells (Fig. 1b,e,f). A very high affinity (0.59 nM) nanobody against GFP (and its derivatives, see list on http://www.chromotek.com) is commercially available that can be used to label any GFP-coupled target protein with high specificity and efficiency. When used in conjunction with the yeast-GFP clone collection19, which contains GFP-tagged open reading frames (ORFs) for ~75% of yeast genes, almost any yeast protein can be studied in cells at nanoscopic detail using fluorescence superresolution microscopy. For enzymatic labeling strategies, a genetically encoded fusion protein of an enzyme and the target protein is generated (see Experimental design section), and a substrate coupled to an organic dye is then conjugated to the enzyme. The SNAP-tag is an engineered version of a bacterial DNA-repair protein called O6-alkylguanine-DNA alkyltransferase. These engineered alkyltransferases interact covalently with O6-benzylguanin (SNAP-tag) derivatives that can be linked to organic dyes 20. We recommend using the SNAPf-tag in our procedure, as it provides the highest labeling efficiency of the enzymatic engineered tags that we have tested (SNAP-tag21, CLIP-tag20 and Halo-tag22). The SNAP-tag labeling approach has been successfully applied in budding yeast assays, including epifluorescence microscopy in live cells. We expanded the method to label membrane-permeabilized fixed cells with the highly efficient AF647 benzylguanine derivative that is not cell permeable for live cells. To label live yeast cells, either benzylguanine derivatives with membrane-permeable dyes have to be applied, or dyes such as AF647 can be transferred into yeast cells via electroporation methods9. In both cases, those yeast cells have to be mutated such that they lack functional plasma efflux transporters in order to keep the dyes inside23. Advantages Maintenance of an intact yeast cell wall (Fig. 1c–f) is essential for retaining structural integrity and avoiding artifacts in superresolution imaging. In addition, imaging of the cell wall can provide
© 2015 Nature America, Inc. All rights reserved.
protocol information about the cell cycle stage and the age of the cell, and it can serve as a spatial reference when imaging the cell wall or a protein of interest in dual color. Differential interference contrast images of the yeast cell or staining of the cell wall with a fluorophore can be easily combined with our protocol to obtain this information. Another key advantage of using nanobodies and enzymatic tags is a reduction in localization error resulting from the distance between the fluorescent molecules and the actual target protein. The relative distances using the different labeling schemes are depicted in Figure 1g. In the nanometer resolution range of SMLM, the size of the label or delivery vehicle has a substantial effect on the localization error, as demonstrated in an experiment comparing nanobody labeling with sandwich immunofluorescence labeling of microtubules17 (Fig. 1h). The nanobody delivers the dye into close proximity of the target structure, whereas the antibody (which is ~10 nm in size and requires binding of a secondary antibody) can only bring the dye within ~20 nm of the target structure, resulting in a substantial localization error compared with the localization accuracy of a single dye. The small size of the labels probably also enables them to reach buried epitopes that are inaccessible to bulky antibodies. This increase in labeling efficiency is crucial for obtaining a high spatial resolution in SMLM. The high affinity of the nanobody combined with its highly specific binding (to exactly one site on the endogenously GFP-tagged molecule) may also allow for quantitative imaging, and it certainly increases signal linearity compared with multivalent tags or polyclonal antibodies.
In general, the same limitations apply for our labeling approach as for standard immunolabeling protocols. For example, the fixation methods might not be applicable to all proteins or organelles, and the permeabilization can affect the localization of trans-membrane and membrane-binding proteins. Although the small labels offer improved epitope accessibility, certain proteins still will not be accessible enough to achieve a high labeling density. Applications This labeling procedure allows users to label almost any protein in budding yeast cells for SMLM techniques in a simple and quick manner. 2D SMLM imaging is usually performed in total internal reflection fluorescence (TIRF) illumination to increase the signal-to-noise ratio and thereby the localization precision. The globular shape of budding yeast means that only a small area of the cell periphery can be accessed in TIRF mode (Fig. 2), but this modality is still suitable for imaging the organization of structures at the cortex and endocytotic events24. For a detailed interpretation of biological processes and structures within the cell, epifluorescence imaging is performed. Depending on the structure, this imaging modality can offer high resolution in all three dimensions25,26. Our protocol is compatible with 3D bifocal plane single-molecule imaging17 and can, in principle, be used with any other commercially available 3D SMLM system. Dual-color or multicolor labeling and imaging would be another extended application of this protocol. We have already demonstrated that the outline of the cell wall, labeled with another dye suitable for SMLM, could provide information about the cell cycle stage17. Several commercially available Alexa Fluor and Atto dyes with different excitation and emission spectra can be coupled to nanobodies or linked to derivatives of the enzymatic SNAP-tag
Limitations One fundamental limitation of this approach is that the enzymatic or GFP tags may affect the native localization and function of the tagged proteins. For instance, in the S. cerevisiae genomic GFP Budding yeast preparation library, only 75% of proteins could be labeled without loss of localization and Inoculation of liquid culture function. This implies that, for the other Incubation overnight 25% of proteins, different tagging strategies need to be explored or tagging will not be possible, thus restricting our suggested labeling scheme.
Labeling procedure
Blocking of unspecific binding Optimization Step 9A and B
Labeling of target molecule Optimization Step 10A and B
Next day: inoculate to prepare a fresh culture
Figure 2 | Schematic of sample preparation of budding yeast for single-molecule localization microscopy. A brief overview of the procedure is shown. Cells are grown overnight and inoculated the next day in fresh medium to the appropriate cell density for further sample preparation. Cells are immobilized on Con A–coated coverslip before fixation. Washing steps are used to remove overlaying cells, as this can cause background during imaging. Blocking of nonspecific dye reaction, dye labeling and washing can be optimized according to the advice in the TROUBLESHOOTING table. Budding yeast is globular, and only small areas of the cell surface are accessible in TIRF illumination. Imaging in the TIRF mode can help minimize background fluorescence caused by scattering of light within the cell. Steps in which the procedure can be optimized are highlighted in red.
OD 0.8
Overnight culture
Cell immobilization
Sample washing and mounting for microscopy Optimization Step 10A and B
Single-molecule localization microscopy
Cell wall intact Fixation Optimization Steps 7A, B and 8
200 nm
16 h), we recommend cross-linking Con A to the coverslip8. In a first step, the coverslip is functionalized by silane chemistry. The applied aldehyde alkoxy silane product Bio-Conext can then react covalently with the Con A–exposed amine groups. Note that the composition of cell wall glycans can differ substantially within and between yeast families35, so a different lectin may be needed for immobilization of other yeast cells. Fixation methods. Fixation is crucial for ensuring optimal preservation of the ultrastructure and authentic localization of cellular components to avoid artifacts during superresolution microcopy. A number of approaches can be used to fix cells, including freezing and chemical fixation. Chemical fixation by cross-linking reagents such as formaldehyde and glutaraldehyde is highly compatible with light microscopy, and it is the approach that we use in this protocol. As glutaraldehyde (a dialdehyde) is a potent cross-linker of proteins and phospholipids with free amino groups, it preserves fine structures very well and is recommended for resolving the ultrastructure of complex protein assemblies and cytoskeletal proteins. However, glutaraldehyde penetrates tissue more slowly than formaldehyde, which builds methylene bridges between proteins and forms a matrix to trap other cellular molecules. Other methods include organic solvent fixation and permeabilization, but we do not recommend these approaches, which remove lipids and solvent proteins36. Different concentrations (formaldehyde 1–4% (wt/vol); glutaraldehyde 10) freeze/thaw cycles. 1× BRB80 buffer (80 mM PIPES, 1 mM MgCl2 and 1 mM EGTA) This buffer is used specifically for microtubules51 and septin cytoskeleton preservation. Mix 24.18 g of PIPES, 95.20 mg of MgCl2 and 380 mg of EGTA in 1 liter of ddH2O and adjust the pH to 6.9. Filter-sterilize the buffer and store it at room temperature (RT; 23 °C) for up to 6 months. 1× PBS 1× PBS is 10 mM phosphate buffer, 2.7 mM potassium chloride, 137 mM sodium chloride. Mix 5 PBS tablets in 1 l of ddH2O, pH 7.4. Filter-sterilize the solution and store it at RT for up to 6 months. 200 mM Tris buffer for imaging Prepare 1 M Tris hydrochloride in 1 liter of ddH2O as a stock, and adjust the pH to 8.0. Dilute the buffer to 200 mM Tris in ddH2O. Filter-sterilize the buffer and store it at RT for up to 6 months. 1× yeast peptone dextrose (YPD) medium For 1 liter of medium, mix 20 g of peptone, 10 g of yeast extract and 20 g of glucose. Stir and bring the volume to 1 liter in ddH2O. Autoclave the medium before use and store it at 4 °C for up to 6 months. Blocking reagent, 5% (optional) Add 5 g of BSA, HS or goat serum (GS) to 100 ml of buffer X. Filter-sterilize the reagent and store it at 4 °C for up to 2 years. Con A solution Prepare 2 or 5 mg/ml 500-µl Con A aliquots in ddH2O; store the aliquots at −20 °C for up to 6 months. CRITICAL Thaw the solution freshly before use and avoid too many (>10) freeze/thaw cycles. Formaldehyde solution, 4% (wt/vol) Dissolve 4 g of paraformaldehyde and 2 g of sucrose in 100 ml of the appropriate buffer X. Treat the solution with NaOH pellets at 60 °C on a heating plate with constant stirring until the solution becomes clear. Adjust the pH to the buffer conditions required. nature protocols | VOL.10 NO.7 | 2015 | 1013
© 2015 Nature America, Inc. All rights reserved.
protocol Prepare 10-ml aliquots and store them at −20 °C for up to 1 year. When precipitated paraformaldehyde appears after thawing, it should not be used anymore. ! CAUTION Wear appropriate protective safety equipment and work under a fume hood. CRITICAL Prepare the solution a day in advance, and thaw the aliquot immediately before use. (Optional) If glutaraldehyde is to be added to the formaldehyde solution, thaw it freshly from the −20 °C freezer and use it at the desired concentration % x (vol/vol) in 4% (wt/vol) paraformaldehyde/buffer X. Glutaraldehyde solution (optional) Thaw glutaraldehyde freshly from the −20 °C freezer, and dilute it to the desired concentration % x (vol/vol) in buffer X. CRITICAL Buffer X refers to the appropriate buffer for optimal structural preservation for a particular experiment. Buffer X/50 mM NH4Cl (Only for formaldehyde use) Prepare 2.67 g of NH4Cl in 1 liter of the appropriate buffer X. Filter-sterilize the buffer and store it at RT for up to 6 months. Buffer X/NaBH4 Prepare NaBH4 solution by adding buffer X in the appropriate volume to bring the solid NaBH4 into a 10 mg/ml concentrated solution. ! CAUTION Wear appropriate protective safety equipment and work under a fume hood. CRITICAL Always freshly prepare the solution a few minutes before use. 2-Mercaptoethanol (BME) buffer Prepare a 10-ml 10% (vol/vol) 2-mercaptoethanol and 20% (wt/vol) glucose solution in 200 mM Tris, pH 8.0. Filter-sterilize the buffer and store it in 50-µl aliquots at −20 °C for up to 3 months. ! CAUTION Wear appropriate protective safety equipment and work under a fume hood. GLOX buffer Mix 0.5 mg/ml glucose oxidase and 40 µg/ml catalase in 200 mM Tris and 10% (vol/vol) glycerol to a final volume of 1 ml. Filter-sterilize the buffer, and store the stock at −20 °C in 10-µl aliquots for up to 3 months. ! CAUTION Wear appropriate protective safety equipment and work under a fume hood.
Nanobody labeling (optional) Dialyze 50 µl from a 1 mg/ml nanobody (GFP-Trap or RFP-Trap) solution in a mini-dialysis unit (MWCO = 3,500 Da) into 0.2 M NaHCO3, pH 8.2. Add fivefold molar excess of NHS (succinimidyl)–AF647 (or any other fluorophore desired) to the nanobodies and incubate the mixture for 1–2 h in the dark at RT. Run the sample once on a Zeba desalting column (MWCO = 7,000 Da) to remove any excess fluorophore. Add 10% (vol/vol) glycerol to the nanobody solution, and then store it at −20 °C for up to 6 months. The labeling ratio of dye to protein can be calculated as explained in the dye manufacturer’s protocols or in the review by Mund et al.8. Fiducial marker (optional) Couple biotin to 10 mg/ml Con A with the Thermo Scientific biotinylation kit in filtered PBS. Excess biotin can be removed quantitatively by single-use desalting columns. Store the stock solution at 4 °C for up to 6 months. Dilute Con A–biotin to 2 mg/ml in the appropriate buffer for the experimental conditions. Con A labeling (optional) Label 2 mg/ml Con A with a suitable fluorescent dye via NHS ester reaction for 2 h by rotation in the dark in PBS. Remove excess dye by Thermo Scientific single-use desalting columns, as described in the instructions. Store the stock solution at 4 °C. If desired, calculate the labeling ratio of dye to protein as explained in the dye manufacturer’s protocols or in the review by Mund et al.8. EQUIPMENT SETUP Yeast strains Yeast strains with GFP-tagged ORFs can be purchased from the GFP library provided by Life Technologies (Life Technologies, strain background EY0986). Labels such as RFP, SNAPf-tag and other enzymatic labels can either be newly integrated by molecular cloning or alternatively the GFP locus can be exchanged in this way. Single-molecule microscope setup Description of how to assemble a single-molecule microscopy setup for superresolution imaging is beyond the scope of this protocol, and it can be found elsewhere in literature49,50.
PROCEDURE Preparation of Con A–coated coverslips ● TIMING ~1–4 h CRITICAL Perform all steps at RT, unless otherwise specified. CRITICAL Buffer X refers to the appropriate buffer for optimal structural preservation for a particular experiment. For example, to preserve microtubules51 and the septin cytoskeleton, we recommend BRB80, whereas cytoskeletal buffer52 should be used for actin preservation. CRITICAL Filter all solutions with a 0.22-µm filter before use to avoid nonspecific background fluorescence signal during superresolution imaging. CRITICAL Care should be taken to ensure that reagents and equipment do not become contaminated with fluorescent dyes. 1| Coat the coverslip with Con A to immobilize S. cerevisiae according to either option A or option B. Spin coating (option B) is recommended when a thin and isotropic layer of Con A is required. CRITICAL STEP The coverslip should be prepared either on the day before or the day of the experiment. (A) Without spin coating ● TIMING 4 h (i) Rinse the coverslip with ddH2O, transfer it to a coverslip rack and air-dry it. (ii) Transfer the rack to a plasma cleaner and use air plasma for 60 s at high frequency if you are using the Harrick Plasma O2 plasma cleaner (SMARTech) or 3 min with power output 5 if you are using the FEMTO cleaner (Diener Electronic). If no plasma cleaner is available, treat the coverslips with HNO3, as described in Supplementary Methods. (iii) Pipette 20 µl of Con A (2 mg/ml) onto an 18-mm-diameter coverslip or 50 µl (2 mg/ml) onto a 22–25-mm-diameter coverslip; distribute the solution with a pipette and incubate the coverslip for 30 min at RT. (iv) Remove the remaining liquid and blow-dry it with pressurized nitrogen. Alternatively, if the coverslips are prepared the day before the experiment, air-dry them overnight. (B) With spin coating ● TIMING 1 h (i) Clean the coverslip as described in Step 1A(i,ii). (ii) Pipette 20 µl of Con A (5 mg/ml) on an 18-mm-diameter coverslip or 50 µl (5 mg/ml) on a 22–25-mm-diameter coverslip; distribute the solution with a pipette and incubate the coverslip for 30 min at RT. Keep the coverslip in a humid chamber to avoid drying out. (iii) Spin-coat the coverslip for 15 s at 4,000 r.p.m. (iv) Place the coverslip in a desiccator under vacuum for at least 30 min. 1014 | VOL.10 NO.7 | 2015 | nature protocols
protocol S. cerevisiae preparation and growth ● TIMING 16 h (overnight) 2| The day before the experiment, inoculate 5 ml of YPD medium (or synthetic complete medium) in a 50-ml plastic vial with a colony of a freshly streaked yeast strain expressing the protein of interest. Grow the liquid yeast culture overnight in a shaking incubator at 30 °C, 110 r.p.m. Add the appropriate antibiotic for selection when required. 3| On the morning of the experiment, dilute the yeast into the appropriate fresh medium in an Erlenmeyer flask and grow it for 3 h until an optical density (OD) of 0.8 is reached, as measured by spectrophotometry. Approximately 10 ml of culture volume is sufficient to prepare two 25-mm-diameter circular coverslips and four 18-mm-diameter coverslips. Mounting S. cerevisiae cells on coverslips and subsequent fixation ● TIMING 45–60 min CRITICAL For Steps 4–13, use forceps to transfer the coverslips. For an 18-mm circular coverslip, use 1 ml of solution in 12-well plastic plates. For a 22–25-mm circular coverslip, use 3 ml of solution in six-well plastic plates. CRITICAL All washing steps are performed on the lowest setting on an orbital rocker, referred to here as ‘gentle shaking’. 4| Transfer the yeast culture to a 50-ml vial and centrifuge it for 4 min at 2,800g at RT.
© 2015 Nature America, Inc. All rights reserved.
5| Discard the supernatant and resuspend yeast in the required volume of ddH2O for mounting: 200 µl per 18-mm circular coverslip, and 400 µl per 25-mm circular coverslip. 6| Pipette the yeast suspension onto the Con A–coated coverslip and allow the yeast to attach for 15 min. CRITICAL STEP Yeast cells should not be fixed before this step, as the fixative interacts with glycosylated proteins in the cell wall and inhibits binding to the lectin, resulting in poor immobilization and more cell movement during image acquisition. 7| Prepare the appropriate fixation solution in a plastic well plate (Critical note before Step 4 and Reagent Setup) and fix the cells according to option A (if fixing with formaldehyde) or option B (if the fixative contains glutaraldehyde). ! CAUTION When preparing, using and discarding the fixation solution, always work under the fume hood and wear appropriate safety equipment. CRITICAL STEP When exchanging solutions, gently pipette them up and down to remove additional yeast cell layers so that only one layer remains for imaging (additional layers increase background noise). (A) Fixation with 4% (wt/vol) formaldehyde (i) Transfer the coverslip with the mounted yeast cells (from Step 6) into the plastic well plate containing fixation solution (Critical note before Step 4) and incubate it for 15 min. (ii) Discard the fixation solution, replace it with buffer X-50 mM NH4Cl and wash it with gentle shaking for 5 min. (iii) Exchange the solution for fresh buffer X/50 mM NH4Cl (Critical note before Step 4), and then incubate the coverslip for 10 min with gentle shaking. (B) Fixation with 0.2–1% (vol/vol) glutaraldehyde, with or without formaldehyde (i) Transfer the coverslip with the mounted yeast cells (from Step 6) into the plastic well plate containing fixation solution, and then incubate it for 15 min. During the incubation period, prepare a fresh solution of 10 mg/ml NaBH4/buffer X (see note after Step 3 and Reagent Setup). ! CAUTION Execute all NaBH4 work under the fume hood, and wear appropriate safety equipment. (ii) Remove the fixation solution and replace it with NaBH4/buffer X solution (Critical note before Step 4). Wash the sample three times (10 min for each wash) with gentle shaking in NaBH4/buffer X solution (Critical note before Step 4). (iii) Wash the sample three times (5 min for each wash) with gentle shaking in buffer X (Critical note before Step 4). 8| Test the quality of the fixation. When you are working with a target protein that is linked to a FP, use a fluorescence microscope to check whether the overall structure or protein localization in the cell has been preserved. ? TROUBLESHOOTING Reduction of nonspecific staining and simultaneous permeabilization of the S. cerevisiae sample ● TIMING ~40 min CRITICAL Never let the sample dry out; this can affect protein structure and epitope-binding properties (for Steps 9–13). 9| Prepare the appropriate blocking solution and block as described in option A (Image-iT FX) or option B (BSA, HS or GS) for at least 30 min. (A) Blocking with Image-iT FX (i) Prepare a solution of 50% (vol/vol) Image-iT FX/0.25% (vol/vol) Triton X-100 in the required volume of buffer X. (ii) Transfer the coverslip from the multiwell dish onto Parafilm on a transportable pad (with the yeast sample facing up) by using forceps, and then pipette 300 µl per 25-mm coverslip and 100 µl per 18-mm coverslip. Incubate the coverslip for at least 30 min with gentle shaking. nature protocols | VOL.10 NO.7 | 2015 | 1015
protocol
© 2015 Nature America, Inc. All rights reserved.
(iii) Use forceps to transfer the coverslip back into fresh buffer X in the multiwell dish, and then wash it three times (3 min for each wash). (B) If using BSA, GS or HS (i) Bring Triton X-100 to a concentration of 0.25% (vol/vol) in the blocking reagent solution, and then pipette the appropriate amount of solution into a well dish. For an 18-mm circular coverslip, use 1 ml of solution in 12-well plastic plates. For a 22–25-mm circular coverslip, use 3 ml of solution in six-well plastic plates. (ii) Transfer the coverslip into blocking solution using forceps, and then shake it gently for at least 30 min. Immunofluorescence staining of S. cerevisiae for superresolution imaging ● TIMING ~30 min–2 h CRITICAL Steps 10–12 must be performed in the dark to avoid bleaching of the organic fluorophore. 10| Apply anti-GFP nanobody coupled to AF647 (or another suitable fluorophore) to label FPs such as GFP (and GFP derivatives), or use the benzylguanine derivate coupled to AF647 (or alternative fluorophore) for SNAP-tag labeling. (A) Detecting FP-tagged molecules with dye-conjugated anti-GFP nanobodies (i) Dilute nanobodies to 10 µM in Buffer X/BSA containing 0.25% (vol/vol) Triton X-100. Prepare 300 µl per 25-mm coverslip and 100 µl per 18-mm coverslip. (ii) Transfer the coverslip onto Parafilm and pipette the appropriate volume of staining solution onto it; incubate for 90 min. CRITICAL STEP For uniform immunostaining of cells, we recommend turning the coverslip onto the Parafilm and pipetting the solution onto it. (iii) Prepare the appropriate volume of buffer X in a new well of the plastic dish for washing the sample. (iv) Transfer the coverslip into the wash solution and incubate it at RT. Wash it three times (3 min for each wash) with gentle shaking. (B) Detection of SNAP-tagged molecules with dye-conjugated benzylguanine (i) Prepare a staining solution with 1.25 µM benzylguanine–AF647, 0.25% (vol/vol) Triton X-100 and 1 mM DTT in buffer X. Prepare 300 µl per 25-mm coverslip and 100 µl per 18-mm coverslip. (ii) Transfer the coverslip onto Parafilm and pipette the appropriate volume of staining solution onto it. (iii) Incubate the sample for 30 min at 37 °C. (iv) Prepare the appropriate volume of buffer X in a new well of the plastic dish for washing the sample. (v) Transfer the coverslip into the wash solution (Step 9B(i)) and wash it three times (3 min for each wash) with gentle shaking. PAUSE POINT The sample can be stored at RT for up to 4 h, or for longer (>4 h) at 4 °C. After 16–24 h, Con A can peel off the coverslip and cells can become loose, so prolonged storage is not advised. (Optional) Fiducial marker labeling of the cell wall with Con A ● TIMING ~45 min 11| Remove the remaining liquid from the washed sample by briefly blotting the coverslip on a tissue before placing it on Parafilm with the sample side facing up. 12| Add 50 µl (18-mm-diameter circular coverslip) or 100 µl (22–25-mm-diameter circular coverslip) of Con A–biotin (2 mg/ml) solution in buffer X to the sample and incubate it for 30 min. 13| Vortex the original gold nanorods stock (as described in the manufacturer’s instructions) and prepare the required gold nanorods dilution in buffer X (Supplementary Fig. 5). 14| Remove the Con A–biotin solution from the sample and apply 100 µl (18-mm circular coverslip) or 200 µl (22–25-mm circular coverslip) of gold nanorods solution onto the sample. Incubate the sample for 30 min at RT. ? TROUBLESHOOTING 15| Prepare the appropriate volume of buffer X in a new well of the plastic dish for washing the sample (Step 9B(i)). 16| Transfer the coverslip into the wash solution and wash it three times (3 min for each wash) with gentle shaking. (Optional) Second color labeling of the cell wall with Con A ● TIMING ~45 min 17| Dilute the Con A–fluorophore solution to 70 nM in buffer X. 18| Transfer the coverslips to Parafilm and apply 100 µl (18-mm circular coverslip) or 200 µl (22–25-mm circular coverslip) of solution to the sample. Incubate the sample for 30 min at RT. 19| Prepare the appropriate volume of buffer X in a new well of the plastic dish for washing the sample (Step 9B(i)). 1016 | VOL.10 NO.7 | 2015 | nature protocols
protocol
© 2015 Nature America, Inc. All rights reserved.
20| Transfer the coverslip into the wash solution and wash it three times (3 min for each wash) with gentle shaking. Mounting the sample for superresolution imaging ● TIMING ~5–20 min ! CAUTION Wear safety equipment and follow safety rules while working with 2-mercaptoethanol and epoxy hardener. 21| To mount the sample in an open chamber, follow option A. Use option B if you are using a sealed chamber. (A) Mounting in an open imaging chamber (i) Bring 45 µl of BME from the stock solution and 5 µl of GLOX from the stock solution to a volume of 200 µl in 200 mM Tris, pH 8. When the volume of the mounting chamber is greater, adjust the amount of buffer mixture to that volume. (ii) Mount the sample into the chamber and pipette the imaging buffer from Step 21A(i) immediately onto the sample. (B) Mounting in a sealed chamber (i) Clean the coverslips (15-mm circular coverslip for 18-mm round coverslip; 18-mm circular coverslip for 22–25-mm circular coverslip) either by plasma cleaning (as described in Step 1A(iii)) or in 70% (vol/vol) ethanol, and blow-dry them. (ii) Dilute 1 M Tris (pH 8.0) solution into 100% glycerol to prepare a 200 mM Tris (pH 8.0)/80% (vol/vol) glycerol mix. (iii) Prepare imaging buffer with glycerol, by mixing 150 µl of 200 mM Tris (pH 8.0)/80% (vol/vol) glycerol with 45 µl of BME and 5 µl of GLOX. Prepare 200 µl per coverslip. (iv) Let the washing solution flow gently from the coverslip by tilting it on a tissue. Lay the coverslip flat and add 100 µl of imaging buffer with glycerol and incubate it at RT for 5 min. (v) Let the imaging glycerol buffer flow gently off the coverslip as before, and then add 50 µl of fresh imaging glycerol buffer to it. (vi) Mix the epoxy hardener and place small drops around the edge of a clean coverslip. (vii) Cover the sample with the coverslip, and gently dab it with a tissue to remove remaining imaging buffer. CRITICAL STEP Avoid air bubbles when applying the coverslip. (viii) Wait for 5 min until the epoxy is solid, and then seal the sample with nail polish. Single-molecule imaging ● TIMING ~5 min–1 h 22| Image the sample on a custom-built microscopy setup equipped for single-molecule imaging, and control the acquisition with the appropriate software (e.g., MicroManager)53. Alternatively, use a commercial superresolution microscopy setup, which should come with implemented imaging and data analysis software. 23| Adjust the desired imaging setting (TIRF, semi-TIRF or epifluorescence) and use the appropriate imaging laser line to transfer the AF647 molecules into the dark state by exposing them to high irradiation laser power (5–30 kW/cm2). Depending on the abundance of the target protein and the number of cells, the exposure should last for a few seconds for abundant proteins such as Cdc11 (Fig. 4b) or much less time for less abundant proteins such as Spc42 (Supplementary Fig. 4). Start the acquisition as soon as almost all dyes are residing in the dark state with a constant excitation laser irradiation (1–5 kW/cm2). AF647 is activated into the fluorescent on-state by the 405-nm laser line and, if the microscopy set up is equipped with that feature, the 405-nm laser power will be automatically adjusted to keep the number of molecules in the fluorescent state low and constant during the acquisition. Detailed protocols and troubleshooting for the imaging process are available elsewhere15,43,50. CRITICAL STEP When you are mounting the sample holder on the stage, clean the bottom of the coverslip thoroughly with 70% (vol/vol) ethanol, always wiping in the same direction and carefully drying it afterward. This prevents unwanted material from coming in contact with the immersion oil or objective, which can result in background fluorescence, damage to the objective or distortion of the optical path. ? TROUBLESHOOTING 24| Perform data analysis using appropriate commercial or open-source software packages44–48, either software provided with the commercial microscopy setups or custom-written software. Basic parameters such as the signal-to-noise ratio (SNR) cutoff should be adjusted to assign the fluorescence signal from the single or multiple emitters acquired in one microscopy frame to a single data point or several data points (such that the x and y positions and their error of localization are computationally analyzed). Multiple data points can potentially be grouped together by space and time; for example, if they come from one single emitter that enters a fluorescent on-state several times during the acquisition. Detailed instructions on fitting single-molecule images for superresolution imaging and generating superresolution images is beyond the scope of this protocol and can be found elsewhere15,50,54. ? TROUBLESHOOTING Troubleshooting advice can be found in Table 1. nature protocols | VOL.10 NO.7 | 2015 | 1017
protocol Table 1 | Troubleshooting table. Step
Problem
Possible reason
Solution
8
Poor preservation of protein localization or structure
Inappropriate growth medium
Use synthetic complete medium to test the growth requirements58
Inappropriate buffer solution
Try another buffer solution; see Wheatley and Wang52 and Kellogg et al.51 or MATERIALS section
Centrifugation of yeast cells
Change the speed of centrifugation, and if the structure is very sensitive, let them settle down by gravity. Note that this adds more time to the PROCEDURE
Suboptimal fixation time
Try longer fixation times
Suboptimal fixation reagents
Vary concentration and fixation reagents in solutions
Con A is not dry enough
Dry Con A for longer in air or in a desiccator in Step 1A and B
Con A concentration is too low
Increase the Con A concentration in Step 1A and B
Rotation of the shaker is too strong
Lower the rotation speed of the shaker if possible
Inadequate sample washing
Add more washing steps
© 2015 Nature America, Inc. All rights reserved.
8, 14
14
Cells are not immobilized or incompletely immobilized
Background fluorescence is too high
Vary the duration of the washing steps Ineffective blocking
Check whether permeabilization reagent was added in the blocking step Increase the blocking time up to 1 h; longer blocking is not recommended, as this can shield epitopes and hinder nanobody binding Try another blocking reagent
Fluorescent probe is too hydrophobic Shorten the dye labeling time and generates background Excess glutaraldehyde
Lower the concentration in fixation
Too many cell layers on the coverslip Remove excess cell layers Poor structural resolution owing to low fluorophore density
Concentration of labeling reagents is too low
Increase the concentration of labeling reagents
Increase the labeling time Fixation time is too long
Shorten the fixation time
Insufficient permeabilization reagent
Check whether permeabilization reagent was added in the blocking step Increase the concentration of the permeabilization reagent
Poor preservation of protein localization or structure
Epitope is inaccessible
Apply enzymatic labeling (SNAPf-tag/CLIP-tag/Halo-tag), as fluorophore derivatives are very small
Insufficient fixation time
Try longer fixation times
Suboptimal fixation reagents
Vary concentration and fixation reagents in solutions (continued)
1018 | VOL.10 NO.7 | 2015 | nature protocols
protocol Table 1 | Troubleshooting table (continued). Step
Problem
14
Possible reason
Solution
Too much permeabilization reagent
Try to reduce the use of permeabilization reagent in the blocking and staining steps. Use lower concentrations or use a separate permeabilization step Try out a different permeabilization reagent, such as saponin or digitonin
© 2015 Nature America, Inc. All rights reserved.
23
Inconsistent results between experiments
Sample has dried out
Change buffers more rapidly Incubate in a humid chamber
While imaging fluorophores do not recover, there is photobleaching or no photoswitching, or there is low fluorescence intensity
Buffer conditions and imaging conditions are not optimal
See van de Linde et al.15 for detailed troubleshooting
● TIMING Step 1, preparation of Con A–coated coverslips: ~1–4 h, depending on the equipment used (Step1A: 1 h with plasma cleaning coverslips and 4 h with HNO3 cleaning of coverslips) Steps 2 and 3, S. cerevisiae preparation and growth: 16 h (overnight) Steps 4–8, mounting S. cerevisiae cells on the coverslip and subsequent fixation: ~45–60 min Step 9, reduction of nonspecific staining and simultaneous permeabilization of the S. cerevisiae sample: ~40 min Step 10, immunofluorescence staining of S. cerevisiae for superresolution imaging: ~30 min–2 h Steps 11–16, (optional) fiducial marker labeling of the cell wall with Con A: ~45 min Steps 17–20, (optional) second color labeling of the cell wall with Con A: ~45 min Step 21, mounting the sample for superresolution imaging: ~5–20 min Steps 22–24, single-molecule imaging: ~5 min–1 h ANTICIPATED RESULTS This protocol provides a labeling strategy for SMLM in the model organism budding yeast. By using this approach, it is now possible to image cellular structures of sizes beyond the diffraction limit of conventional fluorescence microscopy (
Optimized sample preparation for single-molecule localization-based superresolution microscopy in yeast Charlotte Kaplan1,3 & Helge Ewers1–3 1Institute of
Biochemistry, ETH Zurich, Zurich, Switzerland. 2Randall Division of Cell and Molecular Biophysics, King’s College London, London, UK. 3Present addresses: Department of Molecular and Cell Biology, University of California, Berkeley, California, USA (C.K.); Institute for Chemistry and Biochemistry, Freie Universität Berlin, Berlin, Germany (H.E.). Correspondence should be addressed to H.E. ([email protected]).
© 2015 Nature America, Inc. All rights reserved.
Published online 11 June 2015; doi:10.1038/nprot.2015.060
Single-molecule localization-based superresolution microscopy methods allow the resolution of cellular structures in the range of tens of nanometers. However, these techniques are of limited use in current yeast labeling protocols, owing to problems with structural preservation. Here we describe an optimized sample preparation protocol that enables single-molecule localization microscopy at high resolution combined with improved structural preservation in Saccharomyces cerevisiae. The protocol uses small binders called nanobodies and an enzymatic labeling strategy to deliver organic dyes to the target protein. These small binders readily penetrate through the yeast cell wall and thus eliminate the requirement for its prior degradation, and they allow structural preservation. In addition, the small size of the binders reduces the distance of the dye to the target protein, and thus it reduces the localization error. The preparation of S. cerevisiae cells for superresolution imaging takes 2–4 h to perform. Researchers should have skills in yeast molecular biology, immunolabeling techniques and access to a microscope equipped for single-molecule imaging.
INTRODUCTION Single-molecule localization microscopy (SMLM) techniques currently allow a resolution down to 10–20 nm. The different approaches to SMLM, photoactivation localization microscopy (PALM1), stochastic optical reconstruction microscopy (STORM2 or dSTORM3) and ground-state depletion and single-molecule return (GSDIM4) all apply the same concept to circumvent the diffraction limit of light. Instead of imaging all molecules in a structure at the same time, every molecule is imaged individually over time and localized with nanometer accuracy. From the refined positions, an image is then assembled. In this way, the organizations of proteins in functional cellular structures such as focal adhesions, nuclear pores and the actin-spectrin cytoskeleton have been resolved on the nanometer scale5–7. In one 3D SMLM study, the protein layers of a focal adhesion could be resolved down to approx. ~20 nm in all three dimensions5. Such studies clearly demonstrate the potential of optical superresolution methods, which allow a nanometer resolution that was previously only attainable with electron microscopy. Current sample preparation protocols in superresolution imaging are generally optimized for the mammalian cell system, with only a small number of papers available for the model organism budding yeast8–11. For a general overview of SMLM methods in yeast, we recommend the review by Mund et al.8. There are two main aspects that distinguish yeast cells from mammalian cells with respect to the application of SMLM imaging. First, mammalian cells in culture generally attach to the coverslip surface via their extracellular matrix proteins and are, in that way, immobilized. An immobilized sample is important for avoiding artifacts in the resulting SMLM data caused by cell movement. The yeast cell wall does not provide naturally adherent molecules that attach to a glass surface, and appropriate surface coating is required for effective immobilization. Second, organic dyes (as they are preferably applied in SMLM) are delivered by antibodies to the targeted structure. Classical immunofluorescence staining in yeast requires degradation of the cell wall, as the immunoglobulins are too large to penetrate the cell wall
by diffusion. This approach can affect the intrinsic protein and organelle organization of yeast, and it prevents optimal structural preservation, which is crucial when imaging structures with nanometer spatial resolution. Here we provide a solution to both problems and a detailed, step-by-step protocol describing the application of SMLM in S. cerevisiae within 2–4 h. SMLM labeling techniques The main limitations in SMLM lie in the brightness (photon yield) of the fluorophore (the greater the fluorophore brightness, the more accurately the emitter’s position can be determined) and the labeling density of the sample. Optimal labeling density is needed to capture all important information and to minimize the risk of misinterpreting the obtained results12. If the labeling density is too low, the spatial resolution will suffer, but if the labeling density is too high nonspecific background binding is likely to cause problems. Genetically encoded photoswitchable fluorescent proteins (FPs) are useful tools for SMLM in yeast8,11. FPs offer key advantages in live-cell superresolution imaging, for example, by providing high specificity for the target structure and low levels of nonspecific background. However, even the most recently developed FPs13 yield markedly fewer photons than organic dyes, resulting in much lower localization precision. An added advantage to using organic dyes for SMLM is that their brightness and blinking properties can be controlled fairly well in situ14,15. A major disadvantage of organic dyes is the difficulty in delivering them specifically, accurately and directly to cellular molecules in yeast cells. Traditional methods use dye-conjugated antibodies to label cell structures in spheroplasts formed by enzymatic digestion of the cell wall combined with permeabilization of the plasma membrane16 (Fig. 1). The use of spheroplasts is necessary for the large (~150 kDa) antibody complex to penetrate the cell and bind to its target. However, spheroplasts are fragile structures that have a shape different from healthy cells and lack their organizational features (Fig. 1a,c,d). These changes in the 3D protein nature protocols | VOL.10 NO.7 | 2015 | 1007
protocol
© 2015 Nature America, Inc. All rights reserved.
FWHM (nm)
Figure 1 | The use of small binders conjugated with organic fluorophores GFP Traditional yeast c d SMLM nanobodies a immunofluorescence renders cell wall digestion unnecessary in yeast immunocytochemistry. (a) Conventionally, yeast immunofluorescence is accompanied by enzymatic Digest cell wall, digestion of the cell wall, as, owing to their size, immunoglobulins cannot permeabilize penetrate it. Cell wall digestion, however, leads to a collapse of cell shape and possibly also of intracellular structures. (b) The smaller epitope-binding domains from the single heavy-chain immunoglobulins termed nanobodies can penetrate the yeast cell wall. This makes cell wall digestion unnecessary. GFP b e f SMLM nanobodies Nanobody staining Nanobodies against GFP are commercially available for delivery of organic dyes to any GFP-tagged molecules from the yeast GFP library. (c–f) Strain YBR127C digest cell wall, permeabilize from the genomic GFP yeast library expressing the vacuolar H+-ATPase pump Vma2 coupled to GFP labeled with AF647-conjugated anti-GFP nanobodies to demonstrate the effect of cell wall digestion on the structure of cellular organelles. (c,d) Digestion of the cell wall leads to structural defects in the vacuole. Vma2 staining is visible as discrete and dispersed aggregates g h 60 Organic dye Organic dye (d, arrowheads). (e,f) Cells stained without cell wall digestion. The reconstructed 50 *** Second SMLM image shows a defined vacuolar outline. (g) The dimensions of a Nanobody 40 antibody nanobody compared with first and second antibody staining of a target Target 30 protein (classical sandwich immunofluorescence). Owing to the size of the antibodies, the dye is delivered at a significant distance from the target First 20 2 nm antibody structure, resulting in additional localization error. The nanobody delivers the αTub Nanobody Target dye to within a few nanometers of the target structure. (h) Quantification of microtubule SMLM by classical sandwich immunolabeling compared with nanobody labeling. Clearly, the full width at half-maximum (FWHM) of the microtubules stained with the anti-GFP nanobody coupled to AF647 is much closer to the actual thickness of a microtubule (green dashed line)17. ***P = 3.8 × 1052 (α-tubulin (αTub) antibody against anti-GFP nanobody) by Mann–Whitney test. Scale bars, 1 µm. All reconstructed superresolution images were visualized with the US National Institutes of Health (NIH) ImageJ look-up table: red-hot.
organization inside the cell can lead to loss or misinterpretation of information in fluorescence imaging. Changes to the structural and spatial protein organization have a major effect on superresolution imaging, as seen in fluorescence superresolution images of the vacuole in yeast cells, with or without cell wall degradation (Fig. 1c–f). As such, traditional immunofluorescence protocols should be avoided for SMLM in yeast. Protocol overview In this protocol, we provide a detailed description of our previously published method for performing the SMLM method dSTORM3,17 using organic dyes, and we make suggestions for achieving optimal preservation and labeling density of each protein or cytoplasmatic structure in budding yeast. The procedure includes a description of the immobilization of yeast cells on coverslips via the lectin concanavalin A (Con A). The proper immobilization of yeast cells is crucial for SMLM, as even small movements of the cell can result in artifacts. After immobilization, the yeast cells are chemically fixed to avoid intrinsic movements and to ensure that the position of intracellular structures is fixed. The yeast cells are then permeabilized and treated with blocking reagents to avoid nonspecific binding of the labeling reagents. Nonspecific background binding can lower the signal-to-noise ratio, reducing localization precision. To avoid the need for cell wall digestion and its associated problems for superresolution imaging, we have optimized two alternative methods for labeling target proteins in yeast cells: nanobody-based labeling and enzymatic labeling. Depending on the protein tag, labeling is performed with either GFP nanobodies or benzylguanine derivatives coupled to an Alexa Fluor 647 (AF647) dye. Nanobodies used in nanobody-based labeling are ~13-kDa hypervariable domains of camelid antibodies18 that can be coupled to fluorescent dyes by N-hydroxysuccinimidyl ester linkage of amino groups (usually on lysines)17 using commercially available reagents. Dye-conjugated nanobodies can penetrate the 1008 | VOL.10 NO.7 | 2015 | nature protocols
cell wall of fixed membrane-permeabilized yeast cells (Fig. 1b,e,f). A very high affinity (0.59 nM) nanobody against GFP (and its derivatives, see list on http://www.chromotek.com) is commercially available that can be used to label any GFP-coupled target protein with high specificity and efficiency. When used in conjunction with the yeast-GFP clone collection19, which contains GFP-tagged open reading frames (ORFs) for ~75% of yeast genes, almost any yeast protein can be studied in cells at nanoscopic detail using fluorescence superresolution microscopy. For enzymatic labeling strategies, a genetically encoded fusion protein of an enzyme and the target protein is generated (see Experimental design section), and a substrate coupled to an organic dye is then conjugated to the enzyme. The SNAP-tag is an engineered version of a bacterial DNA-repair protein called O6-alkylguanine-DNA alkyltransferase. These engineered alkyltransferases interact covalently with O6-benzylguanin (SNAP-tag) derivatives that can be linked to organic dyes 20. We recommend using the SNAPf-tag in our procedure, as it provides the highest labeling efficiency of the enzymatic engineered tags that we have tested (SNAP-tag21, CLIP-tag20 and Halo-tag22). The SNAP-tag labeling approach has been successfully applied in budding yeast assays, including epifluorescence microscopy in live cells. We expanded the method to label membrane-permeabilized fixed cells with the highly efficient AF647 benzylguanine derivative that is not cell permeable for live cells. To label live yeast cells, either benzylguanine derivatives with membrane-permeable dyes have to be applied, or dyes such as AF647 can be transferred into yeast cells via electroporation methods9. In both cases, those yeast cells have to be mutated such that they lack functional plasma efflux transporters in order to keep the dyes inside23. Advantages Maintenance of an intact yeast cell wall (Fig. 1c–f) is essential for retaining structural integrity and avoiding artifacts in superresolution imaging. In addition, imaging of the cell wall can provide
© 2015 Nature America, Inc. All rights reserved.
protocol information about the cell cycle stage and the age of the cell, and it can serve as a spatial reference when imaging the cell wall or a protein of interest in dual color. Differential interference contrast images of the yeast cell or staining of the cell wall with a fluorophore can be easily combined with our protocol to obtain this information. Another key advantage of using nanobodies and enzymatic tags is a reduction in localization error resulting from the distance between the fluorescent molecules and the actual target protein. The relative distances using the different labeling schemes are depicted in Figure 1g. In the nanometer resolution range of SMLM, the size of the label or delivery vehicle has a substantial effect on the localization error, as demonstrated in an experiment comparing nanobody labeling with sandwich immunofluorescence labeling of microtubules17 (Fig. 1h). The nanobody delivers the dye into close proximity of the target structure, whereas the antibody (which is ~10 nm in size and requires binding of a secondary antibody) can only bring the dye within ~20 nm of the target structure, resulting in a substantial localization error compared with the localization accuracy of a single dye. The small size of the labels probably also enables them to reach buried epitopes that are inaccessible to bulky antibodies. This increase in labeling efficiency is crucial for obtaining a high spatial resolution in SMLM. The high affinity of the nanobody combined with its highly specific binding (to exactly one site on the endogenously GFP-tagged molecule) may also allow for quantitative imaging, and it certainly increases signal linearity compared with multivalent tags or polyclonal antibodies.
In general, the same limitations apply for our labeling approach as for standard immunolabeling protocols. For example, the fixation methods might not be applicable to all proteins or organelles, and the permeabilization can affect the localization of trans-membrane and membrane-binding proteins. Although the small labels offer improved epitope accessibility, certain proteins still will not be accessible enough to achieve a high labeling density. Applications This labeling procedure allows users to label almost any protein in budding yeast cells for SMLM techniques in a simple and quick manner. 2D SMLM imaging is usually performed in total internal reflection fluorescence (TIRF) illumination to increase the signal-to-noise ratio and thereby the localization precision. The globular shape of budding yeast means that only a small area of the cell periphery can be accessed in TIRF mode (Fig. 2), but this modality is still suitable for imaging the organization of structures at the cortex and endocytotic events24. For a detailed interpretation of biological processes and structures within the cell, epifluorescence imaging is performed. Depending on the structure, this imaging modality can offer high resolution in all three dimensions25,26. Our protocol is compatible with 3D bifocal plane single-molecule imaging17 and can, in principle, be used with any other commercially available 3D SMLM system. Dual-color or multicolor labeling and imaging would be another extended application of this protocol. We have already demonstrated that the outline of the cell wall, labeled with another dye suitable for SMLM, could provide information about the cell cycle stage17. Several commercially available Alexa Fluor and Atto dyes with different excitation and emission spectra can be coupled to nanobodies or linked to derivatives of the enzymatic SNAP-tag
Limitations One fundamental limitation of this approach is that the enzymatic or GFP tags may affect the native localization and function of the tagged proteins. For instance, in the S. cerevisiae genomic GFP Budding yeast preparation library, only 75% of proteins could be labeled without loss of localization and Inoculation of liquid culture function. This implies that, for the other Incubation overnight 25% of proteins, different tagging strategies need to be explored or tagging will not be possible, thus restricting our suggested labeling scheme.
Labeling procedure
Blocking of unspecific binding Optimization Step 9A and B
Labeling of target molecule Optimization Step 10A and B
Next day: inoculate to prepare a fresh culture
Figure 2 | Schematic of sample preparation of budding yeast for single-molecule localization microscopy. A brief overview of the procedure is shown. Cells are grown overnight and inoculated the next day in fresh medium to the appropriate cell density for further sample preparation. Cells are immobilized on Con A–coated coverslip before fixation. Washing steps are used to remove overlaying cells, as this can cause background during imaging. Blocking of nonspecific dye reaction, dye labeling and washing can be optimized according to the advice in the TROUBLESHOOTING table. Budding yeast is globular, and only small areas of the cell surface are accessible in TIRF illumination. Imaging in the TIRF mode can help minimize background fluorescence caused by scattering of light within the cell. Steps in which the procedure can be optimized are highlighted in red.
OD 0.8
Overnight culture
Cell immobilization
Sample washing and mounting for microscopy Optimization Step 10A and B
Single-molecule localization microscopy
Cell wall intact Fixation Optimization Steps 7A, B and 8
200 nm
16 h), we recommend cross-linking Con A to the coverslip8. In a first step, the coverslip is functionalized by silane chemistry. The applied aldehyde alkoxy silane product Bio-Conext can then react covalently with the Con A–exposed amine groups. Note that the composition of cell wall glycans can differ substantially within and between yeast families35, so a different lectin may be needed for immobilization of other yeast cells. Fixation methods. Fixation is crucial for ensuring optimal preservation of the ultrastructure and authentic localization of cellular components to avoid artifacts during superresolution microcopy. A number of approaches can be used to fix cells, including freezing and chemical fixation. Chemical fixation by cross-linking reagents such as formaldehyde and glutaraldehyde is highly compatible with light microscopy, and it is the approach that we use in this protocol. As glutaraldehyde (a dialdehyde) is a potent cross-linker of proteins and phospholipids with free amino groups, it preserves fine structures very well and is recommended for resolving the ultrastructure of complex protein assemblies and cytoskeletal proteins. However, glutaraldehyde penetrates tissue more slowly than formaldehyde, which builds methylene bridges between proteins and forms a matrix to trap other cellular molecules. Other methods include organic solvent fixation and permeabilization, but we do not recommend these approaches, which remove lipids and solvent proteins36. Different concentrations (formaldehyde 1–4% (wt/vol); glutaraldehyde 10) freeze/thaw cycles. 1× BRB80 buffer (80 mM PIPES, 1 mM MgCl2 and 1 mM EGTA) This buffer is used specifically for microtubules51 and septin cytoskeleton preservation. Mix 24.18 g of PIPES, 95.20 mg of MgCl2 and 380 mg of EGTA in 1 liter of ddH2O and adjust the pH to 6.9. Filter-sterilize the buffer and store it at room temperature (RT; 23 °C) for up to 6 months. 1× PBS 1× PBS is 10 mM phosphate buffer, 2.7 mM potassium chloride, 137 mM sodium chloride. Mix 5 PBS tablets in 1 l of ddH2O, pH 7.4. Filter-sterilize the solution and store it at RT for up to 6 months. 200 mM Tris buffer for imaging Prepare 1 M Tris hydrochloride in 1 liter of ddH2O as a stock, and adjust the pH to 8.0. Dilute the buffer to 200 mM Tris in ddH2O. Filter-sterilize the buffer and store it at RT for up to 6 months. 1× yeast peptone dextrose (YPD) medium For 1 liter of medium, mix 20 g of peptone, 10 g of yeast extract and 20 g of glucose. Stir and bring the volume to 1 liter in ddH2O. Autoclave the medium before use and store it at 4 °C for up to 6 months. Blocking reagent, 5% (optional) Add 5 g of BSA, HS or goat serum (GS) to 100 ml of buffer X. Filter-sterilize the reagent and store it at 4 °C for up to 2 years. Con A solution Prepare 2 or 5 mg/ml 500-µl Con A aliquots in ddH2O; store the aliquots at −20 °C for up to 6 months. CRITICAL Thaw the solution freshly before use and avoid too many (>10) freeze/thaw cycles. Formaldehyde solution, 4% (wt/vol) Dissolve 4 g of paraformaldehyde and 2 g of sucrose in 100 ml of the appropriate buffer X. Treat the solution with NaOH pellets at 60 °C on a heating plate with constant stirring until the solution becomes clear. Adjust the pH to the buffer conditions required. nature protocols | VOL.10 NO.7 | 2015 | 1013
© 2015 Nature America, Inc. All rights reserved.
protocol Prepare 10-ml aliquots and store them at −20 °C for up to 1 year. When precipitated paraformaldehyde appears after thawing, it should not be used anymore. ! CAUTION Wear appropriate protective safety equipment and work under a fume hood. CRITICAL Prepare the solution a day in advance, and thaw the aliquot immediately before use. (Optional) If glutaraldehyde is to be added to the formaldehyde solution, thaw it freshly from the −20 °C freezer and use it at the desired concentration % x (vol/vol) in 4% (wt/vol) paraformaldehyde/buffer X. Glutaraldehyde solution (optional) Thaw glutaraldehyde freshly from the −20 °C freezer, and dilute it to the desired concentration % x (vol/vol) in buffer X. CRITICAL Buffer X refers to the appropriate buffer for optimal structural preservation for a particular experiment. Buffer X/50 mM NH4Cl (Only for formaldehyde use) Prepare 2.67 g of NH4Cl in 1 liter of the appropriate buffer X. Filter-sterilize the buffer and store it at RT for up to 6 months. Buffer X/NaBH4 Prepare NaBH4 solution by adding buffer X in the appropriate volume to bring the solid NaBH4 into a 10 mg/ml concentrated solution. ! CAUTION Wear appropriate protective safety equipment and work under a fume hood. CRITICAL Always freshly prepare the solution a few minutes before use. 2-Mercaptoethanol (BME) buffer Prepare a 10-ml 10% (vol/vol) 2-mercaptoethanol and 20% (wt/vol) glucose solution in 200 mM Tris, pH 8.0. Filter-sterilize the buffer and store it in 50-µl aliquots at −20 °C for up to 3 months. ! CAUTION Wear appropriate protective safety equipment and work under a fume hood. GLOX buffer Mix 0.5 mg/ml glucose oxidase and 40 µg/ml catalase in 200 mM Tris and 10% (vol/vol) glycerol to a final volume of 1 ml. Filter-sterilize the buffer, and store the stock at −20 °C in 10-µl aliquots for up to 3 months. ! CAUTION Wear appropriate protective safety equipment and work under a fume hood.
Nanobody labeling (optional) Dialyze 50 µl from a 1 mg/ml nanobody (GFP-Trap or RFP-Trap) solution in a mini-dialysis unit (MWCO = 3,500 Da) into 0.2 M NaHCO3, pH 8.2. Add fivefold molar excess of NHS (succinimidyl)–AF647 (or any other fluorophore desired) to the nanobodies and incubate the mixture for 1–2 h in the dark at RT. Run the sample once on a Zeba desalting column (MWCO = 7,000 Da) to remove any excess fluorophore. Add 10% (vol/vol) glycerol to the nanobody solution, and then store it at −20 °C for up to 6 months. The labeling ratio of dye to protein can be calculated as explained in the dye manufacturer’s protocols or in the review by Mund et al.8. Fiducial marker (optional) Couple biotin to 10 mg/ml Con A with the Thermo Scientific biotinylation kit in filtered PBS. Excess biotin can be removed quantitatively by single-use desalting columns. Store the stock solution at 4 °C for up to 6 months. Dilute Con A–biotin to 2 mg/ml in the appropriate buffer for the experimental conditions. Con A labeling (optional) Label 2 mg/ml Con A with a suitable fluorescent dye via NHS ester reaction for 2 h by rotation in the dark in PBS. Remove excess dye by Thermo Scientific single-use desalting columns, as described in the instructions. Store the stock solution at 4 °C. If desired, calculate the labeling ratio of dye to protein as explained in the dye manufacturer’s protocols or in the review by Mund et al.8. EQUIPMENT SETUP Yeast strains Yeast strains with GFP-tagged ORFs can be purchased from the GFP library provided by Life Technologies (Life Technologies, strain background EY0986). Labels such as RFP, SNAPf-tag and other enzymatic labels can either be newly integrated by molecular cloning or alternatively the GFP locus can be exchanged in this way. Single-molecule microscope setup Description of how to assemble a single-molecule microscopy setup for superresolution imaging is beyond the scope of this protocol, and it can be found elsewhere in literature49,50.
PROCEDURE Preparation of Con A–coated coverslips ● TIMING ~1–4 h CRITICAL Perform all steps at RT, unless otherwise specified. CRITICAL Buffer X refers to the appropriate buffer for optimal structural preservation for a particular experiment. For example, to preserve microtubules51 and the septin cytoskeleton, we recommend BRB80, whereas cytoskeletal buffer52 should be used for actin preservation. CRITICAL Filter all solutions with a 0.22-µm filter before use to avoid nonspecific background fluorescence signal during superresolution imaging. CRITICAL Care should be taken to ensure that reagents and equipment do not become contaminated with fluorescent dyes. 1| Coat the coverslip with Con A to immobilize S. cerevisiae according to either option A or option B. Spin coating (option B) is recommended when a thin and isotropic layer of Con A is required. CRITICAL STEP The coverslip should be prepared either on the day before or the day of the experiment. (A) Without spin coating ● TIMING 4 h (i) Rinse the coverslip with ddH2O, transfer it to a coverslip rack and air-dry it. (ii) Transfer the rack to a plasma cleaner and use air plasma for 60 s at high frequency if you are using the Harrick Plasma O2 plasma cleaner (SMARTech) or 3 min with power output 5 if you are using the FEMTO cleaner (Diener Electronic). If no plasma cleaner is available, treat the coverslips with HNO3, as described in Supplementary Methods. (iii) Pipette 20 µl of Con A (2 mg/ml) onto an 18-mm-diameter coverslip or 50 µl (2 mg/ml) onto a 22–25-mm-diameter coverslip; distribute the solution with a pipette and incubate the coverslip for 30 min at RT. (iv) Remove the remaining liquid and blow-dry it with pressurized nitrogen. Alternatively, if the coverslips are prepared the day before the experiment, air-dry them overnight. (B) With spin coating ● TIMING 1 h (i) Clean the coverslip as described in Step 1A(i,ii). (ii) Pipette 20 µl of Con A (5 mg/ml) on an 18-mm-diameter coverslip or 50 µl (5 mg/ml) on a 22–25-mm-diameter coverslip; distribute the solution with a pipette and incubate the coverslip for 30 min at RT. Keep the coverslip in a humid chamber to avoid drying out. (iii) Spin-coat the coverslip for 15 s at 4,000 r.p.m. (iv) Place the coverslip in a desiccator under vacuum for at least 30 min. 1014 | VOL.10 NO.7 | 2015 | nature protocols
protocol S. cerevisiae preparation and growth ● TIMING 16 h (overnight) 2| The day before the experiment, inoculate 5 ml of YPD medium (or synthetic complete medium) in a 50-ml plastic vial with a colony of a freshly streaked yeast strain expressing the protein of interest. Grow the liquid yeast culture overnight in a shaking incubator at 30 °C, 110 r.p.m. Add the appropriate antibiotic for selection when required. 3| On the morning of the experiment, dilute the yeast into the appropriate fresh medium in an Erlenmeyer flask and grow it for 3 h until an optical density (OD) of 0.8 is reached, as measured by spectrophotometry. Approximately 10 ml of culture volume is sufficient to prepare two 25-mm-diameter circular coverslips and four 18-mm-diameter coverslips. Mounting S. cerevisiae cells on coverslips and subsequent fixation ● TIMING 45–60 min CRITICAL For Steps 4–13, use forceps to transfer the coverslips. For an 18-mm circular coverslip, use 1 ml of solution in 12-well plastic plates. For a 22–25-mm circular coverslip, use 3 ml of solution in six-well plastic plates. CRITICAL All washing steps are performed on the lowest setting on an orbital rocker, referred to here as ‘gentle shaking’. 4| Transfer the yeast culture to a 50-ml vial and centrifuge it for 4 min at 2,800g at RT.
© 2015 Nature America, Inc. All rights reserved.
5| Discard the supernatant and resuspend yeast in the required volume of ddH2O for mounting: 200 µl per 18-mm circular coverslip, and 400 µl per 25-mm circular coverslip. 6| Pipette the yeast suspension onto the Con A–coated coverslip and allow the yeast to attach for 15 min. CRITICAL STEP Yeast cells should not be fixed before this step, as the fixative interacts with glycosylated proteins in the cell wall and inhibits binding to the lectin, resulting in poor immobilization and more cell movement during image acquisition. 7| Prepare the appropriate fixation solution in a plastic well plate (Critical note before Step 4 and Reagent Setup) and fix the cells according to option A (if fixing with formaldehyde) or option B (if the fixative contains glutaraldehyde). ! CAUTION When preparing, using and discarding the fixation solution, always work under the fume hood and wear appropriate safety equipment. CRITICAL STEP When exchanging solutions, gently pipette them up and down to remove additional yeast cell layers so that only one layer remains for imaging (additional layers increase background noise). (A) Fixation with 4% (wt/vol) formaldehyde (i) Transfer the coverslip with the mounted yeast cells (from Step 6) into the plastic well plate containing fixation solution (Critical note before Step 4) and incubate it for 15 min. (ii) Discard the fixation solution, replace it with buffer X-50 mM NH4Cl and wash it with gentle shaking for 5 min. (iii) Exchange the solution for fresh buffer X/50 mM NH4Cl (Critical note before Step 4), and then incubate the coverslip for 10 min with gentle shaking. (B) Fixation with 0.2–1% (vol/vol) glutaraldehyde, with or without formaldehyde (i) Transfer the coverslip with the mounted yeast cells (from Step 6) into the plastic well plate containing fixation solution, and then incubate it for 15 min. During the incubation period, prepare a fresh solution of 10 mg/ml NaBH4/buffer X (see note after Step 3 and Reagent Setup). ! CAUTION Execute all NaBH4 work under the fume hood, and wear appropriate safety equipment. (ii) Remove the fixation solution and replace it with NaBH4/buffer X solution (Critical note before Step 4). Wash the sample three times (10 min for each wash) with gentle shaking in NaBH4/buffer X solution (Critical note before Step 4). (iii) Wash the sample three times (5 min for each wash) with gentle shaking in buffer X (Critical note before Step 4). 8| Test the quality of the fixation. When you are working with a target protein that is linked to a FP, use a fluorescence microscope to check whether the overall structure or protein localization in the cell has been preserved. ? TROUBLESHOOTING Reduction of nonspecific staining and simultaneous permeabilization of the S. cerevisiae sample ● TIMING ~40 min CRITICAL Never let the sample dry out; this can affect protein structure and epitope-binding properties (for Steps 9–13). 9| Prepare the appropriate blocking solution and block as described in option A (Image-iT FX) or option B (BSA, HS or GS) for at least 30 min. (A) Blocking with Image-iT FX (i) Prepare a solution of 50% (vol/vol) Image-iT FX/0.25% (vol/vol) Triton X-100 in the required volume of buffer X. (ii) Transfer the coverslip from the multiwell dish onto Parafilm on a transportable pad (with the yeast sample facing up) by using forceps, and then pipette 300 µl per 25-mm coverslip and 100 µl per 18-mm coverslip. Incubate the coverslip for at least 30 min with gentle shaking. nature protocols | VOL.10 NO.7 | 2015 | 1015
protocol
© 2015 Nature America, Inc. All rights reserved.
(iii) Use forceps to transfer the coverslip back into fresh buffer X in the multiwell dish, and then wash it three times (3 min for each wash). (B) If using BSA, GS or HS (i) Bring Triton X-100 to a concentration of 0.25% (vol/vol) in the blocking reagent solution, and then pipette the appropriate amount of solution into a well dish. For an 18-mm circular coverslip, use 1 ml of solution in 12-well plastic plates. For a 22–25-mm circular coverslip, use 3 ml of solution in six-well plastic plates. (ii) Transfer the coverslip into blocking solution using forceps, and then shake it gently for at least 30 min. Immunofluorescence staining of S. cerevisiae for superresolution imaging ● TIMING ~30 min–2 h CRITICAL Steps 10–12 must be performed in the dark to avoid bleaching of the organic fluorophore. 10| Apply anti-GFP nanobody coupled to AF647 (or another suitable fluorophore) to label FPs such as GFP (and GFP derivatives), or use the benzylguanine derivate coupled to AF647 (or alternative fluorophore) for SNAP-tag labeling. (A) Detecting FP-tagged molecules with dye-conjugated anti-GFP nanobodies (i) Dilute nanobodies to 10 µM in Buffer X/BSA containing 0.25% (vol/vol) Triton X-100. Prepare 300 µl per 25-mm coverslip and 100 µl per 18-mm coverslip. (ii) Transfer the coverslip onto Parafilm and pipette the appropriate volume of staining solution onto it; incubate for 90 min. CRITICAL STEP For uniform immunostaining of cells, we recommend turning the coverslip onto the Parafilm and pipetting the solution onto it. (iii) Prepare the appropriate volume of buffer X in a new well of the plastic dish for washing the sample. (iv) Transfer the coverslip into the wash solution and incubate it at RT. Wash it three times (3 min for each wash) with gentle shaking. (B) Detection of SNAP-tagged molecules with dye-conjugated benzylguanine (i) Prepare a staining solution with 1.25 µM benzylguanine–AF647, 0.25% (vol/vol) Triton X-100 and 1 mM DTT in buffer X. Prepare 300 µl per 25-mm coverslip and 100 µl per 18-mm coverslip. (ii) Transfer the coverslip onto Parafilm and pipette the appropriate volume of staining solution onto it. (iii) Incubate the sample for 30 min at 37 °C. (iv) Prepare the appropriate volume of buffer X in a new well of the plastic dish for washing the sample. (v) Transfer the coverslip into the wash solution (Step 9B(i)) and wash it three times (3 min for each wash) with gentle shaking. PAUSE POINT The sample can be stored at RT for up to 4 h, or for longer (>4 h) at 4 °C. After 16–24 h, Con A can peel off the coverslip and cells can become loose, so prolonged storage is not advised. (Optional) Fiducial marker labeling of the cell wall with Con A ● TIMING ~45 min 11| Remove the remaining liquid from the washed sample by briefly blotting the coverslip on a tissue before placing it on Parafilm with the sample side facing up. 12| Add 50 µl (18-mm-diameter circular coverslip) or 100 µl (22–25-mm-diameter circular coverslip) of Con A–biotin (2 mg/ml) solution in buffer X to the sample and incubate it for 30 min. 13| Vortex the original gold nanorods stock (as described in the manufacturer’s instructions) and prepare the required gold nanorods dilution in buffer X (Supplementary Fig. 5). 14| Remove the Con A–biotin solution from the sample and apply 100 µl (18-mm circular coverslip) or 200 µl (22–25-mm circular coverslip) of gold nanorods solution onto the sample. Incubate the sample for 30 min at RT. ? TROUBLESHOOTING 15| Prepare the appropriate volume of buffer X in a new well of the plastic dish for washing the sample (Step 9B(i)). 16| Transfer the coverslip into the wash solution and wash it three times (3 min for each wash) with gentle shaking. (Optional) Second color labeling of the cell wall with Con A ● TIMING ~45 min 17| Dilute the Con A–fluorophore solution to 70 nM in buffer X. 18| Transfer the coverslips to Parafilm and apply 100 µl (18-mm circular coverslip) or 200 µl (22–25-mm circular coverslip) of solution to the sample. Incubate the sample for 30 min at RT. 19| Prepare the appropriate volume of buffer X in a new well of the plastic dish for washing the sample (Step 9B(i)). 1016 | VOL.10 NO.7 | 2015 | nature protocols
protocol
© 2015 Nature America, Inc. All rights reserved.
20| Transfer the coverslip into the wash solution and wash it three times (3 min for each wash) with gentle shaking. Mounting the sample for superresolution imaging ● TIMING ~5–20 min ! CAUTION Wear safety equipment and follow safety rules while working with 2-mercaptoethanol and epoxy hardener. 21| To mount the sample in an open chamber, follow option A. Use option B if you are using a sealed chamber. (A) Mounting in an open imaging chamber (i) Bring 45 µl of BME from the stock solution and 5 µl of GLOX from the stock solution to a volume of 200 µl in 200 mM Tris, pH 8. When the volume of the mounting chamber is greater, adjust the amount of buffer mixture to that volume. (ii) Mount the sample into the chamber and pipette the imaging buffer from Step 21A(i) immediately onto the sample. (B) Mounting in a sealed chamber (i) Clean the coverslips (15-mm circular coverslip for 18-mm round coverslip; 18-mm circular coverslip for 22–25-mm circular coverslip) either by plasma cleaning (as described in Step 1A(iii)) or in 70% (vol/vol) ethanol, and blow-dry them. (ii) Dilute 1 M Tris (pH 8.0) solution into 100% glycerol to prepare a 200 mM Tris (pH 8.0)/80% (vol/vol) glycerol mix. (iii) Prepare imaging buffer with glycerol, by mixing 150 µl of 200 mM Tris (pH 8.0)/80% (vol/vol) glycerol with 45 µl of BME and 5 µl of GLOX. Prepare 200 µl per coverslip. (iv) Let the washing solution flow gently from the coverslip by tilting it on a tissue. Lay the coverslip flat and add 100 µl of imaging buffer with glycerol and incubate it at RT for 5 min. (v) Let the imaging glycerol buffer flow gently off the coverslip as before, and then add 50 µl of fresh imaging glycerol buffer to it. (vi) Mix the epoxy hardener and place small drops around the edge of a clean coverslip. (vii) Cover the sample with the coverslip, and gently dab it with a tissue to remove remaining imaging buffer. CRITICAL STEP Avoid air bubbles when applying the coverslip. (viii) Wait for 5 min until the epoxy is solid, and then seal the sample with nail polish. Single-molecule imaging ● TIMING ~5 min–1 h 22| Image the sample on a custom-built microscopy setup equipped for single-molecule imaging, and control the acquisition with the appropriate software (e.g., MicroManager)53. Alternatively, use a commercial superresolution microscopy setup, which should come with implemented imaging and data analysis software. 23| Adjust the desired imaging setting (TIRF, semi-TIRF or epifluorescence) and use the appropriate imaging laser line to transfer the AF647 molecules into the dark state by exposing them to high irradiation laser power (5–30 kW/cm2). Depending on the abundance of the target protein and the number of cells, the exposure should last for a few seconds for abundant proteins such as Cdc11 (Fig. 4b) or much less time for less abundant proteins such as Spc42 (Supplementary Fig. 4). Start the acquisition as soon as almost all dyes are residing in the dark state with a constant excitation laser irradiation (1–5 kW/cm2). AF647 is activated into the fluorescent on-state by the 405-nm laser line and, if the microscopy set up is equipped with that feature, the 405-nm laser power will be automatically adjusted to keep the number of molecules in the fluorescent state low and constant during the acquisition. Detailed protocols and troubleshooting for the imaging process are available elsewhere15,43,50. CRITICAL STEP When you are mounting the sample holder on the stage, clean the bottom of the coverslip thoroughly with 70% (vol/vol) ethanol, always wiping in the same direction and carefully drying it afterward. This prevents unwanted material from coming in contact with the immersion oil or objective, which can result in background fluorescence, damage to the objective or distortion of the optical path. ? TROUBLESHOOTING 24| Perform data analysis using appropriate commercial or open-source software packages44–48, either software provided with the commercial microscopy setups or custom-written software. Basic parameters such as the signal-to-noise ratio (SNR) cutoff should be adjusted to assign the fluorescence signal from the single or multiple emitters acquired in one microscopy frame to a single data point or several data points (such that the x and y positions and their error of localization are computationally analyzed). Multiple data points can potentially be grouped together by space and time; for example, if they come from one single emitter that enters a fluorescent on-state several times during the acquisition. Detailed instructions on fitting single-molecule images for superresolution imaging and generating superresolution images is beyond the scope of this protocol and can be found elsewhere15,50,54. ? TROUBLESHOOTING Troubleshooting advice can be found in Table 1. nature protocols | VOL.10 NO.7 | 2015 | 1017
protocol Table 1 | Troubleshooting table. Step
Problem
Possible reason
Solution
8
Poor preservation of protein localization or structure
Inappropriate growth medium
Use synthetic complete medium to test the growth requirements58
Inappropriate buffer solution
Try another buffer solution; see Wheatley and Wang52 and Kellogg et al.51 or MATERIALS section
Centrifugation of yeast cells
Change the speed of centrifugation, and if the structure is very sensitive, let them settle down by gravity. Note that this adds more time to the PROCEDURE
Suboptimal fixation time
Try longer fixation times
Suboptimal fixation reagents
Vary concentration and fixation reagents in solutions
Con A is not dry enough
Dry Con A for longer in air or in a desiccator in Step 1A and B
Con A concentration is too low
Increase the Con A concentration in Step 1A and B
Rotation of the shaker is too strong
Lower the rotation speed of the shaker if possible
Inadequate sample washing
Add more washing steps
© 2015 Nature America, Inc. All rights reserved.
8, 14
14
Cells are not immobilized or incompletely immobilized
Background fluorescence is too high
Vary the duration of the washing steps Ineffective blocking
Check whether permeabilization reagent was added in the blocking step Increase the blocking time up to 1 h; longer blocking is not recommended, as this can shield epitopes and hinder nanobody binding Try another blocking reagent
Fluorescent probe is too hydrophobic Shorten the dye labeling time and generates background Excess glutaraldehyde
Lower the concentration in fixation
Too many cell layers on the coverslip Remove excess cell layers Poor structural resolution owing to low fluorophore density
Concentration of labeling reagents is too low
Increase the concentration of labeling reagents
Increase the labeling time Fixation time is too long
Shorten the fixation time
Insufficient permeabilization reagent
Check whether permeabilization reagent was added in the blocking step Increase the concentration of the permeabilization reagent
Poor preservation of protein localization or structure
Epitope is inaccessible
Apply enzymatic labeling (SNAPf-tag/CLIP-tag/Halo-tag), as fluorophore derivatives are very small
Insufficient fixation time
Try longer fixation times
Suboptimal fixation reagents
Vary concentration and fixation reagents in solutions (continued)
1018 | VOL.10 NO.7 | 2015 | nature protocols
protocol Table 1 | Troubleshooting table (continued). Step
Problem
14
Possible reason
Solution
Too much permeabilization reagent
Try to reduce the use of permeabilization reagent in the blocking and staining steps. Use lower concentrations or use a separate permeabilization step Try out a different permeabilization reagent, such as saponin or digitonin
© 2015 Nature America, Inc. All rights reserved.
23
Inconsistent results between experiments
Sample has dried out
Change buffers more rapidly Incubate in a humid chamber
While imaging fluorophores do not recover, there is photobleaching or no photoswitching, or there is low fluorescence intensity
Buffer conditions and imaging conditions are not optimal
See van de Linde et al.15 for detailed troubleshooting
● TIMING Step 1, preparation of Con A–coated coverslips: ~1–4 h, depending on the equipment used (Step1A: 1 h with plasma cleaning coverslips and 4 h with HNO3 cleaning of coverslips) Steps 2 and 3, S. cerevisiae preparation and growth: 16 h (overnight) Steps 4–8, mounting S. cerevisiae cells on the coverslip and subsequent fixation: ~45–60 min Step 9, reduction of nonspecific staining and simultaneous permeabilization of the S. cerevisiae sample: ~40 min Step 10, immunofluorescence staining of S. cerevisiae for superresolution imaging: ~30 min–2 h Steps 11–16, (optional) fiducial marker labeling of the cell wall with Con A: ~45 min Steps 17–20, (optional) second color labeling of the cell wall with Con A: ~45 min Step 21, mounting the sample for superresolution imaging: ~5–20 min Steps 22–24, single-molecule imaging: ~5 min–1 h ANTICIPATED RESULTS This protocol provides a labeling strategy for SMLM in the model organism budding yeast. By using this approach, it is now possible to image cellular structures of sizes beyond the diffraction limit of conventional fluorescence microscopy (
