Reagents and antibodies. Detailed information for reagents and antibodies is shown in Supplementary Table 11.
Human samples. The human pathology specimens were purchased from US Biomax, catalog numbers HuFPT072 (normal human kidney cortex), UNC242 (universal control tissue array), BL243 (bladder cancer with normal tissue array) and LY241g (lymph node and tonsil tissue array).
Mouse brain samples. All mouse brain expansion experiments were performed using C57BL/6 mice except where otherwise noted. C57BL/6 mice were deeply anesthetized with ketamine/xylazine before transcardial perfusion with 20 mL 4% paraformaldehyde in 1× PBS. Brains were harvested and post-fixed in 4% PFA in 1x PBS overnight at 4°C. Tissue was cryoprotected by incubating in 30% (w/v) sucrose in 1× PBS at 4°C until the brain sank (usually 16 hours/overnight). Brains were sectioned in 30-µm slices (either coronally or sagittally) using a freezing microtome and stored at 4°C in glycerol solution (30% (v/v) glycerol and 30% (v/v) ethylene glycol in 1× PBS). Figure 4a-d was generated using DAT-cre mice with C57 black background implanted with headposts and injected with two viruses (AAV2/5-hSynapsin1-FLEx-axon-GCaMP6s and AAV5-Syn-FLEX-rc[ChrimsonR-tdTomato]) in the ventral tegmental area. DAT-cre mice were deeply anesthetized with isoflurane before transcardial perfusion with 20 mL 4% paraformaldehyde in 1× PBS. Brains were harvested and post-fixed in 4% PFA in 1× PBS overnight at 4°C. Tissue was cryoprotected by incubating in 30% (w/v) sucrose in 1× PBS at 4°C until the brain sank (usually 16 hours/overnight). Brains were sectioned in 80-µm slices (either coronally or sagittally) using a cryostat and stored at 4°C in glycerol solution (30% (v/v) glycerol and 30% (v/v) ethylene glycol in 1× PBS). Figure 4e-h was generated using SST-cre/Ai3 mouse brain tissue (SST-Cre mouse line, Jax 013044, crossed to Ai3 (Jax 007903)).
Culture of airway basal cells. Normal human bronchial epithelial (NHBE) cells were cultured in 804G-conditioned medium coated culture vessels in bronchial epithelial cell growth medium (BEGM) and supplemented with 1 mM A8301, 5 mM Y27632, 0.2 mM of DMH-1, and 0.5 mM of CHIR9902151 at 37°C with 5% CO2.
Differentiation of airway basal cells into airway organoids. A 96-well tissue culture plate was coated with 40% (v/v) growth factor reduced (GFR) Matrigel in PneumaCult™-ALI Maintenance Medium. The NHBEs were resuspended in 40% (v/v) GFR Matrigel in PneumaCult™-ALI Maintenance Medium and added to the coated wells. 100 µL PneumaCult™-ALI Maintenance Medium was placed in the wells and changed every other day. The cultures were maintained at 37°C with 5% CO2 for 21 days.
Treating airway organoids with Paclitaxel. 21-to-28-day old organoids were treated with paclitaxel, diluted in PneumaCult™-ALI Maintenance Medium to a final concentration of 20 µM, for 24 hours. The control groups were treated with an equal concentration of dimethyl sulfoxide (DMSO) diluted in PneumaCult™-ALI Maintenance Medium.
Tissue section recovery and heat treatment. Formalin-fixed paraffin-embedded (FFPE) clinical samples were washed in the following solutions 2 times for 3 minutes each at room temperature (RT): xylene, 100% ethanol, 95% ethanol, 70% ethanol, 50% ethanol, and doubly deionized water. For samples that were stained prior to gelation, tissue slides were placed in 20 mM sodium citrate solution (pH 8) at 100 ºC. The container was transferred to a 60 ºC incubator for 30 minutes.
Pre-expansion immunostaining of FFPE samples. After heat treatment, samples were blocked with SuperBlock Blocking Buffer in PBS for 1 hour at 37 ºC followed by incubation with primary antibodies diluted to approximately 1 µg/mL in staining buffer (9x PBS/10% TritonX/10mg/L heparin) overnight at RT or for 1 hour at 37 ºC. Samples were washed at least three times with washing buffer (1×PBS/0.1% TritonX-100) at RT for at least 10 minutes. Samples were then incubated in staining buffer with the corresponding secondary antibodies diluted to approximately 1 µg/mL together with 300 nM DAPI for at least 3 hours at RT or for 1 hour at 37 ºC. Samples were then washed at least 3 times with washing buffer for at least 10 minutes each. After washing, samples used for pre-expansion imaging for protein retention estimation were stained with NHS-ATTO-488 in 1×PBS for 30 minutes at RT and then washed with 1× PBS prior to imaging. NHS-stained samples were not gelled.
Pre-expansion immunostaining of mouse brain samples. Mouse brain samples were taken from glycerol solution and washed three times for 10 minutes each in 1× PBS at RT before permeabilization in 1% TritonX-100 in 1× PBS at RT. Samples were then incubated with primary antibodies in 1% TritonX-100 for 15 minutes in 1× PBS overnight at 33 ºC. Samples were washed three times for at least 20 minutes each in 1× PBS at RT before incubation in 1% TritonX-100 in 1× PBS with the corresponding secondary antibodies for 3 hours at 37 ºC and were washed with 1× PBS prior to imaging. Samples used for total protein content estimation were stained with NHS-ATTO-488 or NHS-ATTO-532 diluted to 1:25 − 1:150 for 3 hours at RT in 1× PBS. Samples used for total lipid content estimation were stained with DiO or DiD for 72–96 hours in 0.1% TritonX-100 at RT and washed generously.
In situ polymer synthesis of FFPE samples with MAGNIFY. A monomer solution made of 4% DMAA (v/v), 34% SA (w/v), 10% AA (w/v), 0.01% Bis (w/v), 1% NaCl (w/v), and 1× PBS was prepared and stored at 4 ºC prior to synthesis. Immediately prior to gelation, the chemicals 4HT, APS, TEMED, and methacrolein were added to a final concentration of 0.2–0.25% (w/v) APS, 0-0.25% (v/v) TEMED, 0.001% 4HT (w/v), and 0.1–0.25% (v/v) methacrolein. The solution was vortexed, and tissue slides were incubated with the gelling solution for 30 minutes at 4 ºC to allow the monomer solution to diffuse into the tissue while preventing premature gelation. A gelling chamber was then constructed around the tissue, consisting of spacers cut from #1.5 cover glass and a glass slide, placed backside down, on top. The samples were incubated overnight in a humidified container at 37 ºC to complete gelation.
In situ polymer synthesis of mouse brain samples with MAGNIFY. MAGNIFY gel monomer solution was prepared as described above. Prior to gelation, brain slices were placed into custom gelling chamber consisting of two spacers cut from #1.5 cover glass adhered to the uncoated back of a microscope slide. Excess PBS around the tissue was absorbed with a Kimwipe and sections were allowed to air dry partially on the slide. Immediately prior to gelation, the chemicals 4HT, TEMED, APS, and methacrolein were added to the gel monomer solution to a final concentration of 0.25% (w/v) APS, 0.001% 4HT (w/v), 0.04% TEMED (v/v), and 0.1% (v/v) methacrolein, adding TEMED and APS last to prevent premature gelation. Tissue was incubated in gelling solution for 30 minutes at 4 ºC to allow the monomer solution to diffuse into the tissue. A glass microscope slide was placed backside down over the gelling chamber and the samples were incubated overnight in a humidified container at 37 ºC to complete gelation.
In situ polymer synthesis of human stem-cell derived organoids with MAGNIFY. MAGNIFY gel monomer solution was prepared as described above. Prior to gelation, PFA fixed organoid samples were placed into custom gelling chamber consisting of two spacers cut from #1.5 cover glass adhered to the uncoated back of a microscope slide. Excess PBS around the tissue was absorbed with a Kimwipe and sections were allowed to air dry partially on the slide. Immediately prior to gelation, the chemicals 4HT, APS, and methacrolein were added to a final concentration of 0.25% (w/v) APS, 0.001% 4HT (w/v), and 0.1% (v/v) methacrolein. The solution was vortexed, and tissue slides were incubated with the gelling solution for 30 minutes at 4 ºC to allow the monomer solution to diffuse into the tissue while preventing premature gelation. A gelling chamber was then constructed around the tissue, consisting of spacers cut from #1.5 cover glass and a glass slide on top. The samples were incubated overnight in a humidified container at 37 ºC to complete gelation.
FFPE sample digestion and expansion with MAGNIFY. After gelation, the glass slide cover was removed from the gelling chamber, blank gel surrounding the tissue was trimmed from the samples, and the tissue was cut into smaller pieces. Samples were then incubated in homogenization buffer (1–10% w/v SDS, 8M Urea, 25 mM EDTA, 2× PBS, pH 7.5 at RT) for 24-72h at 80 ºC with shaking. Homogenized samples were then washed 3 times with 1× PBS at RT, followed by at least 3 washes in 1% decaethylene glycol monododecyl ether (C12E10)/1×PBS or 1%PBST at RT or 60 ºC to remove remaining SDS.
Mouse brain sample digestion and expansion with MAGNIFY. After gelation, the glass slide cover was removed from the gelling chamber and blank gel surrounding the tissue was trimmed from the samples. Samples were then incubated in homogenization buffer (10% w/v SDS, 8M Urea, 25 mM EDTA, 2x PBS, pH 7.5 at RT) for 4-8h at 80 ºC with shaking. Homogenized samples were then washed 3 times for at least 10 minutes each with 1× PBS at RT, followed by incubation in 1% decaethylene glycol monododecyl ether (C12E10)/1× PBS at 60 ºC for 1h to remove remaining SDS. Samples were finally washed an additional 3 times for at least 10 minutes each with 1× PBS at RT and stored in 1× PBS containing 0.02% sodium azide at 4 ºC.
Human lung organoid sample homogenization and expansion with MAGNIFY. After gelation, the glass slide cover was removed from the gelling chamber, and blank gel surrounding the tissue was trimmed from the samples. Samples were then incubated in homogenization buffer (10% w/v SDS, 8M Urea, 25 mM EDTA, 2× PBS, pH 7.5 at RT) for 8h at 80 ºC with shaking. Homogenized samples were then washed 3 times with 1× PBS at RT, followed by at least 3 washes in 1% decaethylene glycol monododecyl ether (C12E10)/1×PBS or 1%PBST at RT or 60 ºC to remove remaining SDS.
Protease digestion of gel embedded mouse brain samples. After gelation, blank gel surrounding the tissue was trimmed from the samples. Samples were then incubated in the Expansion Pathology homogenization buffer52 (50 mM Tris [pH 8], 25 mM EDTA, 0.5% [w/v] TritonX, 0.8M NaCl) with Proteinase K diluted by 1:400 (final concentration 2 units/mL). Mouse brain samples were then homogenized at RT for 1-2h. Homogenized samples were then washed 3 times with 1× PBS at RT.
Post-expansion immunostaining of mouse brain samples. Expanded samples were taken from storage in 1× PBS at 4 ºC and washed 3 times for 10 minutes each in 1× PBS at RT. Samples were then incubated with primary antibodies in 1% TritonX-100 in 1× PBS overnight at RT. Samples were washed three times for at least 20 minutes each in 1× PBS at RT before incubation in 1% TritonX-100 in 1× PBS with the corresponding secondary antibodies for 3 hours at RT. Prior to imaging, samples were washed with 1× PBS. Samples used for total protein content estimation were shrunk in polyethylene glycol (PEG 200) 3 times for 10 minutes each before staining with NHS-ATTO-488 or NHS-ATTO-532 diluted to 1:25 − 1:150 for 3 hours at RT in 1× PBS. Samples used for total lipid content estimation were stained with DiO or DiD diluted to 1:200 for 72–96 hours in 0.1% TritonX-100 at RT and washed generously. After staining, samples were washed in water for at least 10 minutes. This was repeated until the sample was fully expanded, at least three exchanges of water.
Post-expansion immunostaining of FFPE samples and expansion. After homogenization and washing, kidney samples used for comparing ExM protocols (MAP, X10, ExPath) were stained with 1:1000 DAPI and 1:250 WGA 640 in 1×PBS for 3h at RT and washed 3 times in 1× PBS followed by staining NHS-ATTO-488 in 1×PBS diluted to 1:250 for 30 minutes. MAGNIFY processed samples were stained with respective primary antibodies diluted to approximately 1 µg/mL in staining buffer (9x PBS/10% TritonX/10mg/L heparin) overnight at RT. Samples were then washed 3 times with washing buffer (1×PBS/0.1% TritonX-100) at RT for at least 10 minutes. Samples were then incubated in staining buffer with the corresponding secondary antibodies diluted to approximately 1 µg/mL together with 300 nM DAPI for at least 3 hours at RT. Samples were then washed at least 3 times with washing buffer for at least 10 minutes. After staining, samples were washed in water for at least 10 minutes. This was repeated until the sample was fully expanded, at least three exchanges of water.
Post-expansion FISH. For MAGNIFY samples being processed for DNA FISH probing, digested gel samples were placed in hybridization buffer made of 1× PBS, 5% (v/v) Triton-X, 10% (v/v) ethylene carbonate containing 10 pM of FISH probes against telomere TelC (/5ATTO550N/CCCTAACCCTAACCCTAACCCTAACCC, Integrated DNA Technologies) and CENP-B box motif (/5ATTO647N/ATT CGTTGGAAACGGGA, Integrated DNA Technologies) and 1µg/ml wheat germ agglutinin conjugated with Alexa Fluor 488. The mixtures were then incubated at 37°C for 2 hours. The samples were washed with stringency wash buffer made of 1× SSC (150 mM NaCl, 15 mM sodium citrate, pH 7.0) at 37°C for 15 min, followed by washes with 2× SSC at 37°C three times for 10 min each. Finally, the gel samples were washed with 1× PBS multiple times at RT (5 min each) prior to imaging.
Imaging. Fluorescence imaging was performed using a Nikon Eclipse Ti2 epifluorescence microscope equipped with a CSU-W1 spinning disk confocal module and an Andor 4.2 Zyla sCMOS camera. The system was controlled by NIS-Elements AR 5.21.03 64-bit software. Images were taken using the following Nikon objectives: CFI Plan Apo Lambda 4× (0.2 NA), CFI Plan Apo Lambda 10× (0.45 NA), CFI Apo LWD Lambda S 20×WI (0.95 NA), CFI Apo LWD Lambda S 40×WI (1.15 NA), CFI Plan Apo Lambda 60×Oil (1.4 NA).
SOFI imaging. Samples were fully expanded in ddH2O in a glass-bottom 6-well plate or custom imaging dish for larger samples. To prevent drift, samples were covered with plastic wrap. Alternatively, imaging glass may be coated in poly-L-lysine solution prior to placement of the gel. SOFI images were taken with either a CFI Apo LWD Lambda S 40×WI (1.15 NA) or CFI Plan Apo Lambda 60×Oil (1.4 NA) objective, with an optional 1.5x magnification. Each SOFI image consisted of 50–100 frames per z-plane with 50–200 ms exposure time per frame.
SOFI image processing. SOFI images were processed using a custom MATLAB code. Images were corrected for drift and intensity, cropped, and deconvolved (Lucy-Richardson method) after 3-dimensional cross-correlation SOFI. For more information refer to Supplementary Note 3.
Measurements of expansion factor for FFPE samples. Expansion factors for comparison of ExPath, X10, and MAGNIFY protocols were estimated using average nuclear surface area of kidney samples. Images of DAPI stained kidney samples were obtained at 10× magnification prior to gelation and after homogenization and expansion. Nuclear surface areas were determined using the Analyze Particles tool in FIJI/ImageJ after image thresholding and binarization. To calculate the linear expansion factor, the square root of the ratio of the average post-expansion nuclear surface area to average pre-expansion surface area was calculated. For specimens with pre-expansion images, immunostained FFPE samples were imaged at 4× and 10× magnification. After gelling and homogenization, expanded tissue pieces were imaged at 4× and 10× magnification. Regions of interest in post-expanded images were matched to pre-expansion regions of interest and the distance measurement tool in NIS Elements or FIJI/ImageJ was used to measure feature sizes in both pre- and post-expansion images.
Quantification of protein retention in FFPE samples. After antigen retrieval, FFPE kidney samples stained with NHS-ATTO-488 were imaged 4× and 10× magnification. Due to the long incubation time necessary to homogenize the kidney samples, a separate set of FFPE kidney samples were gelled without NHS ester staining. Samples were homogenized for 60 hours at 80 ºC, washed in washing buffer, and stained with NHS-ATTO-488. After staining, samples were washed in washing buffer and incubated in 10x PBS to achieve an expansion factor closer to the pre-expansion images. Samples were imaged at 4× and 10× magnification using the same parameters as the pre-expansion images. Mean fluorescent intensity was calculated over a region of interest (ROI) drawn in NIS Elements after background subtraction for both pre- and post-expansion images. Mean intensities were averaged over technical replicates, and post-expansion data was scaled by the cubed linear expansion factor to account for volume differences.
Quantification of protein and lipid retention in mouse brain samples. Mouse brain sections were stained with NHS-ATTO-532 and DiD and were imaged at 10× magnification. One sample was stained with NHS-ATTO-532 and DiD with no homogenization, while the rest of the samples were homogenized for varying times in either ProK or hot surfactant, washed, and stained with NHS-ATTO-532 and DiD, while the other was stained with NHS-ATTO-532 and DiD with no homogenization. Samples were then incubated in 10x PBS to achieve an expansion factor closer to the pre-expansion images and were imaged at 10× magnification using the same parameters as the pre-expansion images. Mean fluorescent intensity was calculated over an ROI drawn in NIS Elements after background subtraction for both pre- and post-expansion images. Mean intensities were averaged over technical replicates, and post-expansion data was scaled by the cubed linear expansion factor to account for volume differences.
Quantification of protein retention in human lung organoid samples. Human lung organoid samples were stained with NHS CF 555 and placed in homogenization buffer for 5 hours and then imaged at 4× magnification. Samples were then homogenized for 9 hours at 80 ºC, washed, and then incubated in 10x PBS to achieve an expansion factor closer to the pre-expansion images and were imaged at 4× magnification using the same parameters as the pre-expansion images. Mean fluorescent intensity was calculated over an ROI drawn in NIS Elements after background subtraction for both pre- and post-expansion images. Mean intensities were averaged over technical replicates, and post-expansion data was scaled by the cubed linear expansion factor to account for volume differences.
Comparison of pre-expansion SOFI images and post-expansion images of FFPE samples. Pre-expansion SOFI images were taken using a CFI Plan Apo Lambda 60×Oil (1.4 NA) objective. Each SOFI image consisted of 50 frames per z-plane with 50–200 ms exposure time per frame. Time series of all channels were taken at each z-plane before moving to the next. SOFI images were processed using custom MATLAB code.
Measurement error quantification. Error was quantified using previously described methods for distortion vector field calculation and root-mean- square (RMS) error calculation.1,7 Briefly, pre-expansion SOFI images were taken at a single z-plane at 60× magnification and multiple z-planes for the same fields of view were obtained post-expansion at 40× magnification as precise matching of z-planes to pre-expansion images can be challenging. To match post-expansion z-planes, scale invariant feature transform (SIFT) key points were generated for all possible combinations of pairs of the pre-expansion images and post-expansion z projections. Because the sample expands along the z axis and different imaging conditions were used, one pre-expansion z plane should correspond to one post-expansion z projection from 8–25 z planes. SIFT key points were generated using the VLFeat open-source library and filtered by random sample consensus (RANSAC) with a geometric model that only permits rotation, translation, and uniform scaling. The pair of pre-expansion and post-expansion images with the most SIFT key points were then used for image registration by rotation, translation, and uniform scaling, as well as calculation of expansion factors and distortion vector fields. By subtracting the resulting vectors at any two points, distance measurement errors could easily be sampled, and the RMS error for such measurements was plotted as a function of measurement length from at least three technical replicates.