CIVIL ENGINEERING 365 ALL ABOUT CIVIL ENGINEERING


General comments

Please see Supplementary Tables 13 for an overview of the reagents and materials used in this work.

Coverslip preparation

Before plating HeLa or U-2OS cells, 12-mm round glass coverslips (Electron Microscopy Sciences, catalog no. 72230-01) were cleaned in a sonic bath (Bronson) submerged in 1 M KOH (Macron Fine Chemicals; catalog no. 6984-04) for 15 min and then rinsed with MilliQ water three times. Glass was then sterilized with 100% ethanol, incubated with 5 µg/mL fibronectin (Sigma-Aldrich, catalog no. F1141) for 30 min, and rinsed with sterile phosphate-buffered saline (PBS; Gibco, catalog no. 10010023) before adding media and cells.

Cell culture

HeLa and U-2OS cells were grown in Dulbecco’s modified Eagle medium (DMEM; Gibco, catalog no. 21063029), supplemented with 10% fetal bovine serum (FBS; Gibco, catalog no. 10438026), and 1% mL/L penicillin-streptomycin (Gibco, catalog no. 15140122) at 37 °C with 5% CO2. Cells were passaged twice to three times a week and used between passage number 2 and 20. Passaging was performed using 1× PBS and 0.05% Trypsin-EDTA (Gibco, catalog no. 25300054). Approximately 24 h before fixation, cells were seeded on fibronectin-coated glass coverslips at ~65,000 cells per well.

Plasmids

For labeling the medial Golgi in HeLa cells, GFP-ManII was expressed. GFP-ManII was made from pEGFP-N1 (Takara Bio Inc.) to include amino acids 1-137 of mouse MAN2A1 fused to GFP, such that GFP is expressed in the Golgi lumen. For labeling the ER membrane in HeLa cells, mEmerald-Sec61-C-18 was expressed. mEmerald-Sec61-C-18 was a gift from the late Michael Davidson (Florida State University, Tallahassee, FL; Addgene plasmid # 54249; referred to as GFP-Sec61β).

Transfection

GFP-ManII and GFP-Sec61β expression in HeLa cells used DNA transfection by electroporation1. DNA was introduced into the cells using a NEPA GENE electroporation device. Approximately 1 million cells were rinsed in Opti-MEM (Gibco, catalog no. 31985070) and then resuspended in Opti-MEM with 10 μg DNA in an electroporation cuvette with a 2-mm gap (Bulldog Bio, catalog no. 12358346). Cells were electroporated with a poring pulse of 125 V, 3-ms pulse length, 50-ms pulse interval, 2 pulses, with decay rate of 10% and + polarity; followed by a transfer pulse of 25 V, 50-ms pulse length, 50-ms pulse interval, 5 pulses, with a decay rate of 40% and ± polarity. After electroporation, the cells were seeded on fibronectin-coated coverslips (see Coverslip preparation). Samples were fixed 24–36 h after electroporation.

MitoTracker orange staining

Live HeLa cells were incubated with 0.5 µM of MitoTracker Orange CMTMRos (Invitrogen, catalog no. M7510) for 30 min at 37 °C and 5% CO2. Next, the cells were washed three times with cell media and fixed immediately after.

Cell fixation

Cells were fixed with 3% formaldehyde (FA) and 0.1% glutaraldehyde (GA) (Electron Microscopy Sciences, catalog nos. 15710 and 16019, respectively) in 1× PBS for 15 min at RT. Samples were rinsed three times with 1× PBS and processed according to the pan-ExM protocol immediately after.

pan-ExM reagents

Acrylamide (AAm; catalog no. A9099), N,N′-(1,2-dihydroxyethylene)bisacrylamide (DHEBA; catalog no. 294381), N,N′-Cystaminebisacrylamide (BAC; catalog no. 9809) were purchased from Sigma-Aldrich. Three different batches of sodium acrylate (SA) were used. The first batch (catalog no. 408220, lot no. MKCF0390) was purchased from Sigma-Aldrich. The second and third batches (catalog no. sc-236893C, lot nos. H3019 and L0619) were purchased from Santa Cruz Biotechnology. We noticed significant batch-to-batch variability in SA purity. To verify that SA was of acceptable purity, 38% (w/v) solutions were made in water and checked for quality37. Only solutions that were light yellow were used. Solutions that were yellow and/or had a precipitate were discarded. N,N′-methylenebis(acrylamide) (BIS; catalog no. J66710) was purchased from Alfa Aesar. Ammonium persulfate (APS; catalog no. AB00112), N,N,N′,N′-tetramethylethylenediamine (TEMED; catalog no. AB02020), tris [hyroxymethyl] aminomethane (Tris; catalog no. AB02000), and 20% sodium dodecyl sulfate solution in water (SDS; AB01922) were purchased from American Bio. Sodium chloride (NaCl; catalog no. 3624-01) was purchased from J.T. Baker.

pan-ExM gelation chamber

The gelation chamber was constructed using a glass microscope slide (Sigma-Aldrich, catalog no. S8400) and two spacers, each consisting of a stack of two no. 1.5 22 × 22 mm coverslips (Fisher Scientific, catalog no. 12-541B), were glued with superglue to the microscope slide on both sides of the cell-adhered coverslip, with the cell-adhered coverslip glued in between. A no. 1.5 22 × 22 mm coverslip was used as a lid after adding the gel solution. This geometry yielded an initial gel thickness size of ~170 µm.

First-round of expansion

HeLa and U-2OS cells, previously fixed as described in the Cell fixation section, were incubated in post-fix solution (0.7% FA + 1% AAm (w/v) in 1× PBS) for 6–7 h at 37 °C. Next, the cells were washed twice with 1× PBS for 10 min each on a rocking platform and embedded in the first expansion gel solution (19% (w/v) SA + 10% AAm (w/v) + 0.1% (w/v) DHEBA + 0.25% (v/v) TEMED + 0.25% (w/v) APS in 1× PBS). Gelation proceeded first for 15 min at room temperature (RT) and then 1.5 h at 37 °C in a humidified chamber. Coverslips with hydrogels were then incubated in ~1 mL denaturation buffer (200 mM SDS + 200 mM NaCl + 50 mM Tris in MilliQ water, pH 6.8) in 35 mm dishes for 15 min at RT. Gels were then transferred into denaturation buffer-filled 1.5 mL Eppendorf tubes and incubated at 73 °C for 1 h. Next, the gels were placed in petri dishes filled with MilliQ water for the first expansion. Water was exchanged at least twice every 1 h and then the gels were incubated overnight in MilliQ water. Gels expanded between 3.8× and 4.5× according to SA purity (see Reagents).

Re-embedding in neutral hydrogel

Expanded hydrogels were incubated in a fresh re-embedding neutral gel solution (10% (w/v) AAm + 0.05% (w/v) DHEBA + 0.05% (v/v) TEMED + 0.05% (w/v) APS in 1× PBS) three times for 20 min each on a rocking platform at RT. Immediately after, residual gel solution was removed by extensive but gentle pressing with Kimwipes. The gels were then sandwiched between two pieces of no. 1.5 coverslips and incubated at 37 °C for 1.5 h in a nitrogen-filled humidified chamber. Next, the gels were detached from the coverslip and washed three times with 1× PBS for 30 min each on a rocking platform at RT. Gels were incubated in post-fix solution (0.7% FA + 1% (w/v) AAm in 1× PBS) for 15 min at RT and then for 6–9 h at 37 °C. The gels were subsequently washed three times with 1× PBS for 30 min each on a rocking platform at RT.

Second-round of expansion

Re-embedded hydrogels were incubated in a fresh second hydrogel gel solution (19% (w/v) SA + 10% AAm (w/v) + 0.1% (w/v) BIS + 0.05% (v/v) TEMED + 0.05% (w/v) APS in 1× PBS) four times for 15 min each on a rocking platform on ice. Immediately after, residual gel solution was removed by extensive but gentle pressing with Kimwipes. The gels were then sandwiched between two pieces of no. 1.5 coverslips and incubated at 37 °C for 2 h in a humidified nitrogen-filled chamber. To dissolve DHEBA, gels were incubated in 0.2 M NaOH for 1 h on a rocking platform at RT. Gels were next detached from the coverslip and washed three times with 1× PBS for 30 min each on a rocking platform at RT. Subsequently, the gels were labeled with antibodies and pan-stained with NHS ester dyes. Finally, the gels were placed in petri dishes filled with MilliQ water for the second expansion. Water was exchanged at least twice every 1 h at RT, and then the gels were incubated overnight in MilliQ water. Gels expanded between 3.8× and 4.0× according to SA purity (see Reagents) for a final expansion factor of 13× to 20×. Note that for the ER images shown in Fig. 6b–e, the cleavable crosslinker BAC was used instead of BIS at a concentration of 0.1% (w/v).

Antibody labeling post-expansion

For microtubule samples, gels were incubated for 24 h with monoclonal mouse anti-ɑ-tubulin antibody (DM1α; Sigma-Aldrich, catalog no. T6199) diluted to 1:250 in antibody dilution buffer (2% (w/v) BSA in 1× PBS). For mitochondria samples, gels were incubated for 24 h with rabbit anti-TOM20 antibody (Abcam, catalog no. ab78547) diluted to 1:250 in antibody dilution buffer. For centriole samples, gels were incubated for 24 h with rabbit polyclonal anti-polyglutamate chain (PolyE) antibody (Adipogen, catalog no. AG-25B-0030-C050) in antibody dilution buffer. For both Golgi and ER samples, gels were incubated for 36–40 h with polyclonal rabbit anti-GFP antibody (Invitrogen, catalog no. A11122) diluted to 1:250 in antibody dilution buffer. All primary antibody incubations were performed on a rocking platform at RT. Gels were then washed in PBS-T (0.1% (v/v)Tween 20 in 1× PBS) three times for 20 min each on a rocking platform at RT. Next, microtubule samples were incubated for 24 h with ATTO647N-conjugated anti-mouse antibodies (Sigma-Aldrich, catalog no. 50185) diluted to 1:250 in antibody dilution buffer, while mitochondria, ER, Golgi, and centriole samples were incubated for 24 h at RT with ATTO647N-conjugated anti-rabbit antibodies (Sigma-Aldrich, catalog no. 40839) diluted to 1:250 in antibody dilution buffer. All secondary antibody incubations were performed on a rocking platform at RT. The gels were subsequently washed in PBS-T three times for 20 min each, rinsed one time with 1× PBS, and stored in PBS at RT until subsequent treatments. Note that bovine serum albumin (BSA; catalog no. 001-000-162) was purchased from Jackson ImmunoResearch and Tween 20 (catalog no. P7949) was ordered from Sigma-Aldrich.

NHS ester pan-staining post-expansion

After antibody labeling, gels were incubated for 1.5 h with either 20 µg/mL NHS ester-ATTO594 (Sigma-Aldrich, catalog no. 08741), 20 µg/mL NHS ester-ATTO532 (Sigma-Aldrich, catalog no. 88793) or 200 µM NHS ester-DY634 (Dyomics, catalog no. 634-01 A), dissolved in 100 mM sodium bicarbonate solution (Sigma-Aldrich, catalog no. SLBX3650) on a rocking platform at RT. The gels were subsequently washed three to five times in either 1× PBS or PBS-T for 20 min each on a rocking platform at RT. Note that for the experiment where we compared HeLa cells expanded never, once, or twice (Fig. 1), the same concentration of NHS ester-ATTO594 and labeling conditions were used.

Palmitate pan-staining

Live 80%-confluent HeLa cells were incubated with 50 µM azide-functionalized palmitate (Thermofisher, catalog no. C10265) diluted in delipidated medium (DMEM + 10% charcoal-stripped FBS; Thermofisher, catalog no. A3382101) for 5 h at 37 °C and 5% CO2. Next, the cells were fixed with 3% FA + 0.1% GA in 1× PBS for 15 min at RT and processed according to pan-ExM protocols. Prior to NHS ester staining, CuAAC (Copper(I)-catalyzed Azide-Alkyne Cycloaddition) was performed using the Click-iT Protein Reaction Buffer Kit (Thermo Fisher, catalog no. C10276) according to manufacturer instructions. Alkyne-functionalized ATTO590 dye (Sigma-Aldrich, catalog no. 93990) was used at a concentration of 5 µM. After CuAAC, the gels were washed three times with 2% (w/v) delipidated BSA (Sigma Aldrich, catalog no. A4612) in 1× PBS for 20 min each on a rocking platform at RT.

Maleimide pan-staining post-expansion

pan-ExM processed gels were reduced with 50 mM Tris(2-carboxyethyl)phosphine hydrochloride solution (TCEP) (Sigma-Aldrich, catalog no. 646547) in 1× PBS for 30 min at RT and subsequently incubated for 1.5 h in an inert environment with 20 µg/mL maleimide-ATTO594 (Sigma-Aldrich, catalog no. 08717) dissolved in deoxygenated 150 mM Tris-Cl pH 7.4 solution. The gels were then washed three times in either 1× PBS or PBS-T for 20 min each on a rocking platform at RT.

SYTOX Green staining post-expansion

pan-ExM processed gels were incubated with SYTOX Green (Invitrogen, catalog no. S7020) diluted 1:3,000 in calcium- and magnesium-free HBSS buffer (Gibco, catalog no. 14170112) for 30 min on a rocking platform at RT. The gels were then washed three times with PBS-T for 20 min each on a rocking platform at RT.

pan-ExM sample mounting

After expansion, the gels were mounted on Poly-L-Lysine-coated glass-bottom dishes (35 mm; no. 1.5; MatTek). A clean 18-millimeter diameter Poly-L-Lysine-coated coverslip (Marienfeld, catalog no. 0117580) was put on top of the gels after draining excess water using Kimwipes. The samples were then sealed with a two-component silicone glue (Picodent Twinsil, Picodent, Wipperfürth, Germany). After the silicone mold hardened (typically 15–20 min), the samples were stored in the dark at RT until they were imaged. Note that gels imaged with an oil objective were incubated overnight in 30% glycerol (Teknova, catalog no. G1797) prior to mounting.

Image acquisition

Confocal and STED images were acquired using a Leica SP8 STED 3X equipped with a SuperK Extreme EXW-12 (NKT Photonics) pulsed white light laser as an excitation source and a Onefive Katana-08HP pulsed laser as depletion light source (775-nm wavelength). All images were acquired using either a HC PL APO 63×/1.2 water objective, HC PL APO 86×/1.2 water CS2 objective, HC PL APO 63×/1.40-0.60 oil objective, or a HC PL APO 100×/1.40 NA oil immersion CS2 objective. Application Suite X software (LAS X; Leica Microsystems) was used to control imaging parameters. ATTO532 was imaged with 532-nm excitation. ATTO594 was imaged with 585-nm excitation and 775-nm depletion wavelengths. ATTO647N was imaged with 647-nm excitation and 775-nm depletion wavelengths. DY634 was imaged with 634-nm excitation. SYTOX Green and MitoTracker Orange were excited by 488-nm and 555-nm excitation light, respectively.

Widefield images to measure protein retention were obtained with a Leica tissue culture microscope (DM IL LED FLUO) equipped with a 10×/0.22 NA air objective. An Andor Clara CCD camera operated by MicroManager was used to record images.

Protein retention assay

HeLa cells were transfected with either GFP-ManII or GFP-Sec61β and plated at 75,000 cells per 12-mm fibronectin-coated glass coverslip. Non-expanded cells from the same experiment were stored in 1× PBS at 4 °C after fixation and were permeabilized with 0.1% Triton X-100 (Sigma-Aldrich, catalog no. T8787) in 1× PBS prior to antibody labeling. All non-expanded, once- and twice-expanded samples were subjected to the same antibody labeling scheme described in Antibody labeling post-expansion. However, ATTO594-conjugated anti-rabbit antibodies (1:250; Sigma-Aldrich, catalog no. 77671) were used instead of ATTO647N-conjugated antibodies. Additionally, samples were stained with SYTOX Green as described in SYTOX Green staining post-expansion. All samples were imaged in deionized water with a widefield Leica tissue culture microscope (DM IL LED FLUO) using a 10×/0.22 NA air objective (see Image Acquisition) and the same LED light intensity.

To measure protein retention in the first expansion step, we compared the total fluorescence signal (TFS) of ATTO594 between non-expanded and once expanded cells expressing GFP-Sec61β and immunolabeled against GFP. Since the ER spreads throughout the whole cytoplasm, TFS for these samples was quantified by measuring the total background-corrected mean fluorescence signal per field of view and dividing it by the number of cells in the field of view. Cells were counted based on the SYTOX Green nuclear staining using FIJI’s 3D Objects counter. Cell numbers were manually corrected for nuclei which were so close that the automatic segmentation merged them into single objects (typically the case for 5–20% of the nuclei). Background levels were determined by averaging the signal determined in a manually selected area containing no cells. 5 fields of view containing between 419 and 653 cells each were analyzed for non-expanded samples; 10 fields of view containing 21–58 cells each were analyzed for samples expanded once.

Because signal-to-background levels dropped to values too low to provide reliable results using the described method for the ER staining in samples expanded twice, we chose a different approach using more localized labeling of the Golgi complex for these samples. To measure protein retention in the second expansion step, we compared the TFS of ATTO594 between once-expanded and twice-expanded cells expressing GFP-ManII and immunolabeled against GFP. TFS for these samples was quantified by multiplying the manually-identified total area occupied by stained Golgi stacks in every cell with the background-corrected mean fluorescence signal in that area. Background levels were determined by averaging the signal determined in a manually selected area within a cell containing no GFP-ManII signal. 60 cells in 10 fields of view were analyzed for samples expanded once and 67 cells in 45 fields of view were analyzed for samples expanded twice. For all measurements, TFS was corrected for the different camera exposure times used. Images were processed using FIJI/ImageJ software. Results are summarized in Supplementary Fig. 18.

Image processing

Images were visualized, smoothed, and contrast-adjusted using FIJI/ImageJ or Imspector software. STED and confocal images were smoothed for display with a 0.5 to 1-pixel sigma Gaussian blur. Minimum and maximum brightness were adjusted linearly for optimal contrast. The TOM20 data set (Fig. 2d–f) was corrected for bleedthrough of the NHS ester channel by subtracting a constant fraction of the latter from the former using Imspector. The confocal ManII (Fig. 7) and mitotic cell NHS ester (Supplementary Fig. 9) data sets were corrected for bleedthrough of the NHS ester and SYTOX Green channels respectively by subtracting a constant fraction of the latter from the former using the Image Expression Parser tool in FIJI.

Cristae and Golgi inter-cisternal distance measurements were performed using FIJI. 10-pixel thick line profiles were taken approximately perpendicular to the cristae and Golgi stack orientations and peak-to-peak distances were extracted from the profiles. For mitochondrial cristae, 123 line profiles were drawn from 4 independent experiments. For Golgi inter-cisternal distance measurements, 193 line profiles were drawn from 3 independent experiments. The diameter of ER tubules was determined in FIJI using the Point Tool by manually measuring two positions arranged perpendicular to the orientation of clearly discernible tubules and located at the crests of the signal denoting each side of the tubule. The Euclidean distance between them was used as a measure of the ER tubule diameter. For this measurement, 142 ER tubule widths were extracted from 2 cells in 1 sample. Results are summarized in Fig. 6i–k.

All line profiles were extracted from the images using the Plot Profile tool in FIJI/ImageJ.

Expansion factor calculation

Images of HeLa cell nuclei in non-expanded and pan-ExM expanded samples stained with SYTOX Green (1:3,000) were acquired with a Leica SP8 STED 3X microscope using a HCX PL Fluotar 10×/0.30 dry objective. Average nuclear cross-sectional areas were determined using FIJI/ImageJ software. To calculate the expansion factor, the average nuclear cross-sectional area in pan-ExM samples was divided by the average nuclear cross-sectional area of non-expanded samples. The square root of this ratio represents an estimate of the linear expansion factor. Results are summarized in Supplementary Fig. 11c.

Expansion homogeneity calculation

To compare nuclei, mitochondria, and microtubules pre- and post-expansion, U-2OS cells were cultured as specified in Cell culture and live-labeled with MitoTracker Orange as described in MitoTracker Orange staining. After fixation with 3% FA + 0.1% GA in 1× PBS for 5 min at RT, cells were incubated in post-fix solution (0.7% FA + 1% (w/v) AAm in 1× PBS) for 7 h at 37 °C, permeabilized with 0.1% Triton X-100 in 1× PBS for 5 min on a rocking platform at RT, and labeled with a mouse monoclonal anti-ɑ-tubulin antibody (MT antibody 1; Sigma-Aldrich, catalog no. T5168) diluted in cell antibody dilution buffer (1% (w/v) BSA + 0.2% TX-100 in 1× PBS) for 1 h on a rocking platform at RT. They were then washed three times with wash buffer (0.05% TX-100 in 1× PBS) for 5 min each, labeled with ATTO647N-conjugated anti-mouse antibodies (Sigma-Aldrich, catalog no. 50185) diluted to 1:1,000 in cell antibody dilution buffer for 1 h on a rocking platform at RT, and stained with Hoechst (abcam, catalog no. ab228551) diluted to 1:10,000 in 1× PBS for 20 min on a rocking platform at RT. The samples were next washed three times with wash buffer for 5 min each and rinsed with 1× PBS.

Pre-expansion image stacks of nuclei (Hoechst), mitochondria (MitoTracker Orange), and microtubules (MT antibody 1) were acquired. The cells were then immediately processed with pan-ExM protocol. For post-expansion microtubule labeling, gels were labeled with a different mouse monoclonal anti-ɑ-tubulin antibody (MT antibody 2; Sigma-Aldrich, catalog no. T6199) diluted 1:250 in antibody dilution buffer (2% (w/v) BSA in 1× PBS) for 6 h at 37 °C and 6 h on a rocking platform at RT, washed three times with PBS-T for 20 min each, and labeled with the same ATTO647N-conjugated anti-mouse antibodies as above. Gels were then stained with SYTOX Green dye diluted 1:3,000 in calcium- and magnesium-free HBSS buffer for 45 min on a rocking platform at RT, washed three times with PBS-T for 30 min each at 37 °C, and expanded in MilliQ water as described above. Post-expansion image stacks of nuclei (SYTOX Green), mitochondria (MitoTracker Orange), and microtubules (MT antibody 2) were acquired. Maximum projection images of corresponding pre- and post-expansion image stacks were generated with the FIJI/ImageJ z projection tool. Additionally, post-expansion images of mitochondria and microtubules were despeckled and masks were created manually to exclude regions of no features for microtubule samples.

To determine spatial sample distortion, post-expansion images were smoothed with a 2-pixel Gaussian blur and first registered to the pre-expansion image of the same field of view with either a similarity transform (uniform scaling, rotation, and translation) or an affine transform (scaling, shear, rotation, and translation). FIJI TurboReg plugin was used for this initial registration. The similarity- and affine-registered post-expansion images were registered again to the pre-expansion images with a B-spline-based non-rigid registration package in Matlab8. The similarity measure (error) was set to squared pixel distance and the penalty of registration was set to 1e-1 (nuclei images) or 1e-2 (mitochondria and microtubule images). Using the deformation vector field from the output B-spline transformation parameters, the root mean square (RMS) error of expansion was calculated across different distance measurements. The deformation field was applied to the coordinates of either a binary outline of the pre-expansion image (nuclei images) or its binary skeleton (mitochondria and microtubule images). The distance between every two pairs of points in the pre-expansion image binary image (d_i) and the corresponding deformed coordinates (d_def) were calculated. The RMS error is the absolute difference of these distance measurements (RMS = abs(d_def – d_i)). RMS error was calculated for every combination of points across distances of 20 µm (nuclei images), 10 µm (mitochondria images), and 8 µm (microtubule images). For nuclei, 5 cells in 5 fields of view were analyzed. For mitochondria, 14 fields of view in 5 cells were analyzed. For microtubules, 5 fields of view in 4 cells were analyzed. Results are summarized in Supplementary Fig. 12.

To quantify the expansion factor of nuclei, mitochondria, and microtubules. Similarity-registered post-expansion images and the corresponding pre-expansion images were cropped using four manually-identified landmark features in both images. To determine the expansion factor, the area of the cropped post-expansion image was divided by the area of the cropped pre-expansion image. The square root of this ratio represents the linear expansion factor. Results are summarized in Supplementary Fig. 11.

pan-ExM modified protocol

To test whether the mechanism of the second expansion is primarily polymer entanglement or chemical crosslinking between the sample and the final expansion hydrogel, a modified pan-ExM protocol was developed. After denaturation (Step 3, Supplementary Fig. 1), possible reactive Schiff base groups on proteins that may react covalently with acrylamide monomers were quenched by incubating the gels in 0.1% sodium borohydride (Sigma-Aldrich, catalog no. 452882) in 1× PBS for 7 min at RT followed by an incubation with 100 mM glycine (Sigma-Aldrich, catalog no. G8898) in 1× PBS for 10 min at RT. Additionally, the second post-fixation step (Step 7, Supplementary Fig. 1) was omitted to prevent crosslinking of the sample to the second expansion hydrogel. The remaining steps are identical to the original protocol.

Centriole roundness quantification

Image stacks of PolyE-labeled U-2OS mature centrioles were acquired by confocal microscopy and axial-view centriole image stacks were selected for quantification. Volume Viewer plugin in FIJI was used to align image stacks that were not perfectly axial. Manually selected images of the distal region were smoothed with a 1-pixel Gaussian blur and converted to binary using automatic thresholding in FIJI. Using the shape descriptor tool in FIJI, centriole roundness was measured for both post-fixed centrioles (standard protocol; n = 7 centrioles) and quenched/no post-fix centrioles (pan-ExM modified protocol; n = 8 centrioles). Results are summarized in Supplementary Fig. 15.

Centriole length-to-width quantification

Image stacks of PolyE-labeled and NHS ester pan-stained mature centrioles were acquired by confocal microscopy and lateral view image stacks were selected for quantification. Similarly as above, Volume Viewer plugin in FIJI was used to align image stacks that were not perfectly lateral. To compare the length-to-width ratio of centrioles, 5-pixel thick line profiles of NHS ester pan-staining along the length of the centrioles were drawn to measure centriolar length, and 5-pixel thick line profiles of PolyE staining along the width of the centriole were drawn to measure centriolar width.The peak-to-peak distances were extracted from these profiles and the ratio is the centriole length-to-width ratio12. This ratio was measured for both post-fixed centrioles (standard protocol; n = 17 centrioles) and quenched/no post-fix centrioles (pan-ExM modified protocol; n = 22 centrioles). Results are summarized in Supplementary Fig. 15.

Measurement of antibody labeling efficiency

To test whether the treatments required to dissolve several common cleavable crosslinkers (Supplementary Fig. 27) play a role in reducing post-expansion antibody labeling efficiency, four U-2OS cell samples live-labeled with MitoTracker Orange as described in MitoTracker Orange staining were expanded once (Steps 1–5, Supplementary Fig. 1) using a hydrogel prepared with 0.1% (w/v) BIS (instead of 0.1% (w/v) DHEBA). After the denaturation step (Step 4, Supplementary Fig. 1), one gel was expanded (CTRL) right after. The second gel was treated with 0.2 M NaOH for 1 h at RT (Treatment 1) and then expanded. The third gel was treated with 25 mM sodium periodate (Sigma-Aldrich, catalog no. 311448) diluted in 100 mM sodium acetate buffer (Sigma-Aldrich, catalog no. S7899), adjusted to pH 6.0, for 1 h at RT (Treatment 2) and then expanded. Finally, the fourth gel was treated with 0.25 M TCEP diluted in 1 M Tris-Cl buffer, adjusted to pH 7.5, for 18 h at RT (Treatment 3) and then expanded. Note that we verified that Treatment 1 can dissolve a gel composed of 19% (w/v) SA + 10% (w/v) AAm + 0.1% (w/v) DHEBA + 0.25% (w/v) APS + 0.25% (v/v) TEMED in 1 h at RT; Treatment 2 can dissolve a gel composed of 19% (w/v) SA + 10% (w/v) AAm + 0.2% (w/v) DATD + 0.25% (w/v) APS + 0.25% (v/v) TEMED in 1 h at RT; and Treatment 3 can dissolve a gel composed of 19% (w/v) SA + 10% (w/v) AAm + 0.1% (w/v) BAC + 0.25% (w/v) APS + 0.25% (v/v) TEMED in 18 h at RT. After expansion, all four gels were immunolabeled with both a mouse monoclonal anti-ɑ-tubulin antibody (1:500; Sigma-Aldrich, catalog no. T5168) and a rabbit polyclonal anti-TOM20 antibody (1:500; Abcam, catalog no. ab78547) diluted in antibody dilution buffer (2% (w/v) BSA in 1× PBS) for 12 h on a rocking platform at RT. The gels were washed three times with PBS-T for 20 min each and then immunolabeled with both ATTO647N-conjugated anti-mouse antibodies (1:500; Sigma-Aldrich, catalog no. 50185) and ATTO594-conjugated anti-rabbit antibodies (1:500, Sigma-Aldrich, catalog no. 77671) diluted in antibody dilution buffer for 6 h on a rocking platform at RT. The gels were washed with PBS-T for 20 min each on a rocking platform at RT and expanded again in MilliQ water to the final expansion factor of ~4.5. 3-color images of microtubules (anti-ɑ-tubulin) and mitochondria (MitoTracker Orange and anti-TOM20) were acquired for each condition.

To measure TOM20 signal, background-corrected total fluorescence signal was calculated for each ROI containing mitochondria. This value was divided by the area occupied by mitochondria as calculated from the area of a mask generated from the corresponding MitoTracker Orange signal. Between 23 to 29 fields of view were quantified in 5 cells for every condition. To measure ɑ-tubulin signal, line profiles of microtubules were drawn from the images using the Plot Profile tool in FIJI/ImageJ and the background-corrected peak value was extracted. Between 104 and 123 profiles were drawn in 5 cells for every condition. Results are summarized in Supplementary Fig. 25.

Statistics and reproducibility

For all quantitative experiments, the number of samples and independent reproductions are listed in the figure legends. An unpaired two-tailed t-test in Graphpad Prism 8 was used to analyze the data presented in Supplementary Figs. 15, 18 and 25.

Reporting summary

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



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *