Single-shot 20-fold expansion microscopy – Nature Methods

Cell culture preparation

HEK293 cells (Thermo Fisher, R70007) were cultured in 16-well chambered coverglasses (Grace Bio-Labs, 112359) with DMEM supplemented with 1% penicillin–streptomycin, and the cells were incubated at 37 °C in 5% CO₂ to reach ~20% confluency. NUP96::Neon-AID DLD-1 cells (gifted by T. Schwartz, Massachusetts Institute of Technology (MIT)) were cultured in 24-well glass-bottom plates (Cellvis, P24-1.5H-N), with a 12-mm number 2 round glass coverslip at the bottom of each well, in DMEM supplemented with 1% penicillin–streptomycin, and the cells were incubated at 37 °C in 5% CO₂ to reach ~20–40% confluency.

Microtubule staining was performed following previously reported protocols^8,43. All of the following steps were conducted at room temperature (~24 °C), unless otherwise noted. Cells were incubated in extraction buffer (0.5% (wt/vol) Triton X-100, 0.1 M 1,4-piperazinediethanesulfonic acid, 1 mM ethylene glycol-bis(2-aminoethylether)-N,N,N′,N′-tetraacetic acid and 1 mM magnesium chloride (pH 7); 100 µl per well) for 1 min and fixed in tubulin fixation solution (3% formaldehyde, 0.1% glutaraldehyde and 1× PBS; 200 µl per well) for 10 min, followed by incubation in reduction solution (0.1% (wt/vol) sodium borohydride in 1× PBS; 200 µl per well) for 7 min and washing with quenching solution (100 mM glycine in 1× PBS; 200 µl per well) for 10 min. Cells were incubated in blocking buffer (Active Motif, 15252; 60 µl per well) for 2 h and then with rabbit anti-β-tubulin in staining buffer (Active Motif, 15253; 1:100 dilution, 60 µl per well) for 2 h. Samples were then washed in washing buffer (Active Motif, 15254; 100 µl per well) three times for 5 min each. Primary antibody staining and washes were then repeated under the same conditions. Cells were incubated with anti-rabbit secondary antibody diluted in staining buffer (1:100 dilution, 60 µl per well) for 2 h and washed in washing buffer (100 µl per well) three times for 10 min each. Secondary antibody staining and washes were then repeated under the same conditions.

TOM20 mitochondria staining was performed using HEK293 cells and previously reported protocols²³. Cells were fixed in fixation solution (3% formaldehyde, 0.1% glutaraldehyde and 1× PBS; 200 µl per well) for 10 min, incubated in reduction solution (0.1% (wt/vol) sodium borohydride in 1× PBS; 200 µl per well) for 7 min and washed with quenching solution (100 mM glycine in 1× PBS; 200 µl per well) for 10 min. Cells were incubated in blocking buffer (Active Motif, 15252; 60 µl per well) for 2 h and then with rabbit anti-TOM20 diluted in staining buffer (Active Motif, 15253; 1:100 dilution, 60 µl per well) for 2 h. Samples were washed in washing buffer (Active Motif, 15254; 100 µl per well) three times for 5 min each, incubated with anti-rabbit secondary antibody in staining buffer (1:100 dilution, 60 µl per well) for 2 h and washed three times for 10 min each in washing buffer (100 µl per well).

NPC staining was performed on NUP96::Neon-AID DLD-1 cells. Cells were fixed in fixation solution (4% formaldehyde and 1× PBS; 1 ml per well) for 10 min and incubated with quenching solution (100 mM glycine in 1× PBS; 1 ml per well) for 10 min. Cells were incubated in blocking buffer (Active Motif, 15252; 300 µl per well) for 2 h and then with rabbit anti-mNeonGreen diluted in staining buffer (Active Motif, 15253; 1:100 dilution, 300 µl per well) for 2 h. Samples were washed in washing buffer (Active Motif, 15254; 500 µl per well) three times for 5 min each, incubated with anti-mouse secondary antibody diluted in staining buffer (1:100 dilution, 300 µl per well) for 2 h and washed in washing buffer (500 µl per well) three times for 10 min each.

All cells were incubated in AX solution (N-acryloxysuccinimide; Thermo Scientific, 400300010; dilution of 10 mg ml^–1 DMSO stock in 1× PBS, 1:2,000; 60 µl per well for 16-well chambered coverglass or 300 µl per well for 24-well glass-bottom plates) at room temperature (~24 °C) overnight (12–20 h). The cells were then washed in 1× PBS for 10 min.

Tissue preparation

All procedures involving mice (Thy1-YFP-H, 6–8 weeks of age from The Jackson Laboratory, used without regard to sex and maintained under standard housing conditions on a 12-h light/12-h dark cycle at an ambient temperature and humidity) were performed in accordance with the US National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the MIT Committee on Animal Care. Mice were deeply anesthetized with isoflurane and perfused with 30 ml of 1× PBS, followed by 30 ml of 4 °C fixative solution (4% paraformaldehyde in 1× PBS). Brains, kidneys and spleens were then removed and stored in the same fixative at 4 °C overnight (12–18 h). Fixed brains, kidneys and spleens were transferred to 100 mM glycine at 4 °C for 6 h and sectioned to 50-μm-thick coronal slices with a vibrating microtome (Leica, VT1000S). The slices were stored in 1× PBS at 4 °C.

Before expansion, each brain slice was incubated in AX solution (Thermo Scientific, 400300010; dilution of 10 mg ml^–1 DMSO stock in 100 mM MES and 150 mM NaCl (pH 6) buffer, 1:200, 1 ml) at 4 °C overnight (12–20 h). The slices were then washed with 1 ml of 1× PBS for 10 min at room temperature (~24 °C).

Before expansion, kidney and spleen slices were microdissected into ~1 mm × 1 mm sections. Each section was incubated in AX solution (Thermo Scientific, 400300010; dilution of 10 mg ml^–1 DMSO stock in 100 mM MES and 150 mM NaCl (pH 6) buffer, 1:200, 50 µl) at 4 °C overnight (12–20 h). The sections were then washed with 50 µl of 1× PBS for 10 min at room temperature (~24 °C).

Expansion of cell culture and tissue slices

See Supplementary Note 1 for a step-by-step protocol.

To generate hydrophobic glass for the gelation chamber, glass slides and coverslips were immersed in 0.2% (vol/vol) trichloro(octadecyl)silane (Fisher Scientific, AC147400250) in hexane for 90 s. The coverslips were rinsed with 70% isopropanol and double-distilled water sequentially. The glass was dried at 37 °C and wiped with a dry Kimwipe to clear residual white solid. Parafilm strips were cut to ~4.5 cm × 0.2 cm and were wrapped around the glass slide to construct a gelation chamber with a 0.4-cm gap for cell culture or 0.1-cm gap for brain tissue (Extended Data Fig. 3a,e).

The gelation solution was prepared by dissolving 0.522 g of SA (AK Scientific, R624) in 1 ml of acidified Tris buffer (10% (vol/vol) 1 M Tris-HCl (pH 8) buffer and 20% (vol/vol) 1.2 M HCl in double-distilled water), followed by the addition of 7.5 µl of 10% (vol/vol) tetramethylethylenediamine (Sigma, T7024) in double-distilled water and 900 µl of DMAA (Sigma, 274135). The mixture was vortexed, yielding a colorless and noncloudy solution. The gelation solution was then placed on ice and bubbled with a gas dispersion tube (ChemGlass, CG-203-04) connected to a compressed nitrogen cylinder tank at a minimal flow rate for 50 s (Extended Data Fig. 2a). The gelation solution was removed from ice and allowed to return to room temperature (~24 °C). All of the following gelation steps were conducted at room temperature. Gelation solution, initiator solution (potassium persulfate, 45 mg ml^–1 in double-distilled water), cell culture or brain tissue, pipettes (P1000, P200 and P20), pipette tips, humidified chamber, hydrophobic glass slides and coverslips, tweezers, a transfer pipette and empty 1.5-ml centrifuge tubes were moved into a glove bag (GlasCol, 108D X-17-17HG) connected to a compressed nitrogen cylinder tank (Extended Data Fig. 2b).

For AX-treated cell cultures, the coverglass from the cell culture well was separated using a coverglass removal tool (Grace Bio-Lab, 103259; Extended Data Fig. 3d). Parafilm strips on the glass slide were adjusted to match with the positions of the wells to be expanded. The remaining rubber was carefully removed from the coverglass with tweezers. The coverglass was placed on top of the parafilm strips with the cells facing up, and 1× PBS was added to keep the cell culture hydrated (Extended Data Fig. 3e). All samples, solutions and tools were moved into the glove bag. The glove bag was purged three times by repeatedly filling the bag with nitrogen and pushing down on the bag to expel most of the accumulated gas. The bag was then sealed and filled with nitrogen. If, in rare occurrences, the bag was leaky and slowly deflated when sealed without nitrogen flow added, a small flow of nitrogen was provided to keep the bag inflated. Inside the glove bag, 20 µl of initiator solution was added to 411 µl of gelation solution in a 1.5-ml centrifuge tube. The tube was flipped upside down five times to mix. The 1× PBS was removed from the cell culture coverglass with a transfer pipette, and 50 µl of the activated gelation solution was added to each well of cell culture. The coverglass was then flipped upside down with tweezers and placed on the parafilm strips to form the gelation chamber (Extended Data Fig. 3f). The gelation chamber was placed in an airtight humidified chamber, taken out of the glove bag and incubated at room temperature (~24 °C) in the dark for 2 h. After incubation, the portion of gel containing cell culture was cut out from the chamber and incubated in digestion buffer (20 µg of LysC/trypsin proteinase in 1 ml of 100 µM Tris-HCl (pH 8) buffer per gel) at 37 °C overnight (12–16 h). Digested gels were washed in PBS two times for 15 min each before proceeding to immunostaining.

For AX-treated mouse brain slices, the brain slices were microdissected to acquire somatosensory cortex as previously reported¹⁰. All microdissected brain, kidney or spleen slices were placed on a glass slide immersed in 1× PBS (Extended Data Fig. 3a,b). All samples, solutions and tools were moved into the nitrogen gas-filled glove bag, followed by three purges as described above. Inside the glove bag, 4 µl of initiator solution was added to 411 µl of gelation solution in a 1.5-ml centrifuge tube. Please note that we added 4 µl of initiator solution for tissue but 20 µl for cell culture. We optimized initiator concentration and gelation time for the tissue protocol to ensure ample time for monomer solution to diffuse into the brain slice. The tube was flipped upside down five times for mixing. The 1× PBS immersing the tissue was removed with a transfer pipette, and 50 µl of the solution was added to incubate the tissue for 15 min in a humidified chamber; the gelation chamber was then constructed by placing a coverslip on top (Extended Data Fig. 3c). The gelation chamber was placed in an airtight humidified chamber, taken out of the glove bag and incubated at room temperature (~24 °C) in the dark overnight (16–20 h).

After incubation, a portion of the gel containing the brain tissue was cut out from the chamber and incubated in denaturation buffer (1 ml; 5% (vol/vol) SDS, 200 mM NaCl, 50 mM Tris (pH 8) and 10 mg ml^–1 DTT) for 1 h at 95 °C. Denatured gels were washed in 1× PBS two times for 15 min each before proceeding to immunostaining.

The gel containing kidney or spleen tissue was cut out from the chamber and incubated in digestion buffer (20 µg of LysC/trypsin proteinase (Thermo Fisher, A41007) in 1 ml of digestion buffer (1 mM EDTA, 50 mM Tris-HCl (pH 8) and 0.1 M NaCl)) at 37 °C overnight (16–24 h), as previously reported²⁵. Digested gels were washed in 1× PBS two times for 15 min each before proceeding to staining.

For blank gels without embedded biological specimens, 20 µl of initiator solution was added to 411 µl of gelation solution in a 1.5-ml centrifuge tube inside the glove bag. The tube was flipped upside down five times for mixing. The activated gelation solution was added to a constructed gelation chamber (Extended Data Fig. 3e). The gelation chamber was placed in an airtight humidified chamber, taken out of the glove bag and incubated at room temperature (~24 °C) in the dark for 1 h (Extended Data Fig. 4a) or for various durations of time (Extended Data Fig. 4b). Gels were cut into ~0.5 × 0.5 cm shapes and expanded by washes in double-distilled water five times for 5 min each.

Immunostaining and imaging of expanded cell culture and tissue slices

All of the following steps were performed without shaking, unless otherwise noted. Gels containing brain tissue or cell culture were incubated in blocking solution (0.5% Triton X-100 and 5% normal donkey serum (Jackson ImmunoResearch, 017-000-121) in 1× PBS) for 2 h at room temperature (~24 °C). Gels containing brain tissue or cell culture were then incubated with primary or tertiary antibodies, respectively (see Supplementary Table 3), in staining buffer (0.25% Triton X-100 and 5% normal donkey serum in 1× PBS) at 4 °C overnight (12–24 h). Gels were washed in washing buffer (0.1% Triton X-100 in 1× PBS) four times for 30 min each on a shaker at 40 rpm at room temperature (~24 °C). Gels containing brain tissue were then incubated with secondary antibodies diluted in staining solution at 4 °C overnight (12–24 h) and washed in washing buffer two times for 30 min each on a shaker at 40 rpm at room temperature (~24 °C). Immunostained gels were fully expanded via three to five 20-min washes with 10 ml of double-distilled water in an imaging plate (MatTek, P384G-1.5-10872-C). DAPI staining was performed during the first expansion wash (Thermo Fisher, D1306; dilution of 10 mg ml^–1 DMSO stock in double-distilled water, 1:1,000, 10 ml).

Gels containing kidney or spleen tissue were incubated in NHS staining solution (Alexa Fluor 488 NHS Ester; Thermo Scientific, A20000; dilution of 10 mg ml^–1 DMSO stock in 1× PBS, 1:50, 1 ml) at 4 °C overnight (12–24 h) and washed in 1× PBS three times (20 min each) on a shaker at 40 rpm at room temperature (~24 °C). NHS-stained gels were fully expanded via three to five 20-min washes with 10 ml of double-distilled water on an imaging plate (MatTek, P384G-1.5-10872-C).

20ExM-processed sample images were acquired using a Nikon CSU-W1 confocal microscope with a ×4/0.2-NA air objective, a ×10/0.45-NA air objective or a ×40/1.15-NA water-immersion objective, 100% laser power and 300–500 ms exposure time.

The confocal images in Fig. 2a were collapsed to two dimensions using maximum intensity projection, and contrast was adjusted with Fiji’s autoscaling function. Confocal images in Fig. 2d,i were adjusted with Fiji’s autoscaling function. Confocal images in Fig. 3b were background subtracted using Fiji’s rolling ball algorithm with a radius of 50 pixels, collapsed to two dimensions using maximum intensity projection and passed through a two-dimensional Gaussian filter (σ = 1). The confocal images in Extended Data Fig. 5a,e–h were collapsed to two dimensions using maximum intensity projection, and contrast was adjusted with Fiji’s autoscaling function and manually adjusted to improve contrast for the stained structures of interest.

Expansion factor and resolution measurement

Expansion factors for each sample were determined by imaging whole specimens (tissues and cultured cells) with a confocal microscope before and after the expansion. The expansion factor was determined by measuring the distance between two landmarks in the specimens (Supplementary Table 2)⁴⁴. For samples described in Supplementary Note 2, we also measured the physical gel size with a ruler immediately after gelation and after full expansion.

Resolutions for confocal images in Fig. 2i,j were determined by performing block-wise FRC on a pair of two images that captured the same region with Fiji plugin NanoJ-SQUIRREL’s Calculate FRC-Map function¹⁴.

Peak-to-peak distance measurement

For microtubule analysis, the cross-section line intensity profile was measured over a box area, with the long axis perpendicular to the microtubule and the short axis covering ~185 nm in biological length, using Fiji’s line selection tool. The intensity was averaged along the long axis, and the line intensity profile was fitted with a double Gaussian function to detect the two peaks in fluorescence intensity in Python (source code is available at github.com/shiwei-w/20ExM). The distance between the two peaks was measured as the peak-to-peak distance of the microtubule sidewalls.

r.m.s. error measurement

r.m.s. error measurement was performed similar to as described in previous studies¹⁶. For xy plane analysis, postexpansion confocal images were passed through a Gaussian filter (σ = 4), background subtracted using Fiji’s rolling ball algorithm with a radius of 50 pixels and collapsed to two dimensions using maximum intensity projection. Pre-expansion images and processed postexpansion confocal images were registered using rigid body registration in Fiji (TurboReg → Scaled Rotation/Accurate/Manual). The images were then nonrigidly registered, and deformation vector fields were calculated in MATLAB (source code is available at github.com/shiwei-w/20ExM).

For analysis in the xz or yz plane, confocal image z stacks of the same brain region were collected and projected onto xz and yz planes using Fiji’s orthogonal view tool and passed through a Gaussian filter (σ = 4). Both pre- and postexpansion confocal images were registered using rigid body registration and nonrigidly registered in MATLAB in the same fashion as the xy plane analysis.

Autocorrelation and protein enrichment analysis of synaptic nanocolumn

The synaptic nanoarchitecture analysis used in this study was based on previously described methods, specifically autocorrelation (g_a(r)) and protein enrichment analysis^10,15,18. Source code is available at github.com/shiwei-w/20ExM.

For autocorrelation, synapses were identified manually by observing the juxtaposition of presynaptic and postsynaptic clusters¹⁰. Postexpansion 20ExM images were background subtracted using Fiji’s rolling ball algorithm with a radius of 50 pixels, as previously described¹⁰. The autocorrelation function (g_a(r)) in three dimensions measured the likelihood of finding a similar signal at a distance (r) from a given signal. This function quantified the heterogeneity of the measured signal within a given volume. To normalize the autocorrelation of each synaptic cluster, the synaptic cluster was compared to an object with the same shape and volume but a homogeneous voxel intensity, which was set to the average intensity of the synaptic cluster. Consequently, a synaptic cluster with uniform intensity would exhibit baseline g_a(r) values at all radii, whereas local intensity peaks within a synaptic cluster would result in higher g_a(r) values over a radius corresponding to the size of the high-intensity region, which then decayed outside of that radius.

For protein enrichment analysis, a cross-enrichment analysis was performed to analyze the distribution of two different protein clusters in relation to each other. This involved measuring the average voxel intensity of one protein cluster (referred to as the ‘measured cluster’) at various distances from the point of peak intensity in the other protein cluster (referred to as the ‘reference cluster’, which was shifted in space as previously defined¹⁰). The measured cluster’s intensity values were normalized by comparing them to the average intensity at corresponding distances from the peak intensity point in the reference cluster. To establish this baseline, an object with the same shape and volume as the measured cluster was used, and its voxel intensities were set to the average intensity of the measured cluster. Regions within the measured cluster that exhibited high local intensity would result in normalized intensity values greater than 1.

Quantification of NPCs

We performed 20ExM with intact NUP96::Neon-AID DLD-1 cells and imaged the NPCs on the top and bottom of the nuclei, tangential to the imaging z plane. We manually identified NPCs in seven cells from two culture batches based on the characteristic ring structure with at least four visible corners in top view. To measure the radius of individual NPCs, we used Fiji’s radial profile plot plugin to acquire radial intensity distribution and take the peak of the distribution as the radius, as in a previous study⁶. To quantify the number of corners per NPC, we used a previously reported ‘Counting Corners’ algorithm²⁰ (α = 0.93, threshold = 0.6) that divides each NPC into eight sectors and counts how many sectors contain signals above a given threshold. We then measured the distance between adjacent corners, as determined by the Counting Corners algorithm using Fiji’s line selection tool. The line intensity profile was plotted, and the distance between the two peaks was measured as the corner-to-corner distance.

SNR quantification

We adopted the method for quantifying SNR from a previous study¹⁰ and applied it to the dataset used for synaptic nanocolumn analysis (Fig. 3b). In summary, the images were background subtracted using Fiji’s rolling ball algorithm with a radius of 50 pixels. Subsequently, we binarized the image using a threshold calculated as seven times the standard deviation of the average intensity of manually identified background regions, selected every 10–15th slice of the z stack. Synapses were identified by selecting the largest three-dimensional connected components¹⁰. Finally, SNR was determined by dividing the signal intensity by the standard deviation of the background.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

• SEO Powered Content & PR Distribution. Get Amplified Today.
• PlatoData.Network Vertical Generative Ai. Empower Yourself. Access Here.
• PlatoAiStream. Web3 Intelligence. Knowledge Amplified. Access Here.
• PlatoESG. Carbon, CleanTech, Energy, Environment, Solar, Waste Management. Access Here.
• PlatoHealth. Biotech and Clinical Trials Intelligence. Access Here.
• Source: https://www.nature.com/articles/s41592-024-02454-9