U2OS cell line
This knock-in cell line, which expresses the nuclear pore complex 96 (NUP96) tagged with mEGFP (U-2 OS-CRISPR-NUP96-mEGFP clone #195), was purchased from the CLS cell lines service (Eppelheim, Germany). Cells were maintained at 37°C, 5% CO2 in a humified incubator in Dulbecco’s Modified Eagle Medium (DMEM #D5671, Merck, Darmstadt, Germany) supplemented with 10% Fetal Calf Serum (FCS, #S0615, Merck), 4 mM L-glutamine (#25030-024, ThermoFisher Scientific, Waltham, USA) and 1% penicillin/streptomycin (ThermoFisher Scientific), and split twice a week. For experimental usage, cells were plated on poly-L-lysine (PLL, #P2658, Merck) -coated 18 mm glass coverslips and allowed to grow for 24h.
BHK cells
Baby hamster kidney (BHK) fibroblasts were purchased from CLS cell lines service and cultured in DMEM, containing 10% tryptose phosphate broth solution (#T8159, Sigma-Aldrich now Merck), 5% FCS, 2 mM L-glutamine, 60 U/ml penicillin and 60 U/ml streptomycin. Cells were split twice a week, for experiments seeded on PLL coated coverslips and fixed after ~ 24 h with 4% PFA in PBS for 30 minutes.
Primary neuronal culture
Primary hippocampal neurons were obtained from newborn Wistar rats (P0-P1), bred in the animal facility of the University Medical Center Göttingen. All experiments were carried out according to the regulations of the Ethics Committee of the University Medical Center Göttingen, of the local ethic committee, the Lower Saxony State Authority for Consumer Protection and Food Safety (Niedersächsisches Landesamt für Verbraucherschutz und Lebensmittelsicherheit, LAVES), and licensed by a Tötungsversuch (T09/08). All procedures were performed in accordance to the ARRIVE guidelines.
Pubs were decapitated, brains were extracted and hippocampi dissected from both hemispheres. After removing the meninges, hippocampi were washed in 4°C cold Hank´s Balanced Salt Solution (HBSS, #14175-053, Invitrogen, Waltham, MA, USA), and afterwards incubated for 1 h at 37°C in a slowly rotated (~ 30 rpm) enzyme solution, containing 15 U/ml papain (#LS003126, Worthington, Lakewood, USA), 0.5 mg/ml L-cysteine (#30090, Merck), 1 mM CaCl2 (#102382, Merck), and 0.5 mM EDTA (#108418, Merck) in DMEM, which was previously equilibrated with CO2 for 10 minutes. Inactivation of enzyme activity was achieved by transferring the hippocampi to a pre-warmed enzyme inactivation solution, containing 5 mg/ml bovine serum albumin (BSA, #A1391, Applichem, Darmstadt, Germany) and 10% FCS in DMEM. After 15 minutes incubation at 37°C, the inactivating solution was replaced by pre-warmed plating medium (10% horse serum (#S900-500, VWR International GmbH, Darmstadt, Germany), 1.8 mM glutamine, 0.6 mg/ml glucose (#108342, Merck) in MEM (#51200046, ThermoFisher Scientific) and hippocampi were washed 3-4x with plating medium. Neurons were isolated by gentle trituration with a 10 ml serological pipette in 6 ml of plating medium, before they were centrifuged at 800 rpm for 8 minutes. Afterwards, supernatant was removed and cells were resuspended in 6 ml plating medium, before they were seeded at a density of 80k on PLL coated coverslips. Neurons could adhere to coverslips for 2–3 h in the incubator, then plating medium was replaced by Neurobasal culture medium (0.2% B27-supplement, #17504-044; 2 mM GlutaMAX #35050-038, in Neurobasal-A medium #10888-022, all ThermoFisher Scientific), and neurons were kept at 37°C and 5% CO2 for minimum 14 days, before they were fixed with 4% Paraformaldehyde (PFA #30525894, Merck) in PBS for 30 minutes at RT.
Tubulin extraction and immunocytochemistry
Unfixed U2OS cells were incubated in a prewarmed solution of 0.2% Saponin (#47036, Sigma-Aldrich now Merck) in Cytoskeleton Buffer (CB, 10 mM 4-Morpholineethanesulfonic acid,2-(N-Morpholino)ethanesulfonic acid (MES #M3671, Merck), 138 mM KCl (#6781, Carl Roth, Karlsruhe, Germany), 3 mM MgCl2 (#105833, Merck), 2 mM EGTA (Triplex®VI #108435, Merck), 320 mM sucrose (#107651, Merck), pH:6.1) for 1 minute and were subsequently fixed with prewarmed 4% PFA and 0.1% Glutaraldehyde (#A3166, PanReac, Darmstadt, Germany) in CB. Quenching of unreactive aldehydes was achieved by 0.1% NaBH4 (#71320, Sigma-Aldrich now Merck) for 7 minutes, followed by incubation in 0.1 M glycine (#3187, Carl Roth) in PBS for 10 minutes. Immunocytochemistry was performed after premixing the primary anti-tubulin antibodies (Ab, #T6199 Sigma-Aldrich, #302211 Synaptic Systems, Göttingen, Germany, #302203 Synaptic Systems, and #ab18251 Abcam, Cambridge, UK), using AbberiorStar635P (AS635P) labeled secondary nanobodies (#N1202-Ab635P-S and #N2402-Ab635P-S both from NanoTag Biotechnologies, Göttingen, Germany) for 30 minutes at room temperature in a molar ratio of 1:5 (Ab:Nb). In the meanwhile, samples were blocked and permeabilized with 2% BSA and 0.1% Triton X-100 (#9036-19-5, Sigma-Aldrich now Merck) in PBS for 30 minutes at RT. The AB-Nb-mix was diluted in blocking/permeabilization buffer and was added to the cells for 1 h at room temperature. Samples were washed with PBS for 3x ten minutes and then further processed by expansion protocols.
Vesicle plating
Synaptic vesicles were obtained from rat brains and isolated as described before [44,45]. For vesicle plating, clean and sterilized 18 mm glass coverslips were placed into a 12 well plate and coated with sterile-filtered 5% BSA in PBS at 37°C overnight. Afterwards, coverslips were washed 3x five minutes with PBS and 100–150 µl of vesicle solution was added to the coverslips in 500 µl PBS. The plate was centrifuged at 4000 rpm for 30 minutes at room temperature, coverslips were gently washed with PBS for five minutes, and then vesicles were fixed with 4% PFA in PBS for 30 minutes at room temperature. Afterwards, coverslips were quenched in NH4Cl (#101145100, Merck) for 20–30 minutes at RT and were used for immunocytochemistry.
Tissue supply and immunohistochemistry of brain slices
Rat brains were obtained from P0-P1 old Wistar rat pups. Rats were decapitated, brains were removed and directly fixed with 4% PFA for 20 h. Brains were embedded in 4% Agarose (#9012366, VWR Life Science, Hannover, Germany) and were cut into 100, 150 or 200 µm thick slices with a vibratome. After quenching with 50 mM glycine in PBS, slices were gently washed with PBS 3x for five minutes and blocked in 2.5% BSA in 0.3% Triton-PBS (PBS-T) for 2 h. Primary antibodies against Bassoon (#ADI-VAM-PS003-F, Enzo Life Sciences GmbH, Lörrach, Germany) and Homer1 (#160003, Synaptic Systems) were diluted 1:500 in 2.5% BSA in PBS and slices were incubated overnight at 4°C. After washing the slices for 3x five minutes in PBS, they were incubated in secondary antibodies AS635P (#ST635P-1001, Abberior, Göttingen, Germany), for the detection of Bassoon, and Cy3 (#711-165-152, Dianova, Hamburg, Germany) for the detection of Homer1, in a dilution of 1:1000 in 2.5% BSA in PBS-T for 3 h at room temperature. Finally, the slices were washed 5x for five minutes in 2.5% BSA in PBS-T and afterwards 2x for five minutes in PBS. Brain slices were then used for starting the expansion procedure.
Immunocytochemistry of neurons and isolated vesicles
After fixation all neurons and isolated vesicles were quenched with 100 mM NH4Cl for 20 min at room temperature and blocked and permeabilized with 2.5% BSA in PBS-T (PBS + 0.1% Triton-X100, blocking/permeabilization solution, BSA-T) 3x for five minutes. Primary antibodies and nanobodies were diluted in BSA-T and samples were incubated in the anti-/nanobody solution for 1 h at room temperature. The primary antibodies used were anti synaptotagmin1 (SYT1, #105011 Synaptic Systems), anti synaptophysin1 (SYP1, #101004 Synaptic Systems), and anti SHANK2 (#162204 Synaptic Systems). Primary nanobodies were anti VGLUT1-biotinylated and anti SYT1 (#N1605-Biotin, #N2305-Biotin, both from NanoTag Biotechnologies, Göttingen). Anti VGLUT1-ALFA-tag and anti VGLUT1-AF488 (#N1605-ALFA, #N1605-AF488 were custom made by NanoTag Biotechnologies). After incubation with primary anti-/nanobodies, samples were washed 3x with BSA-T for 30 minutes at room temperature. The immunostainings could be combined with any of the amplification systems described later. Incubation in secondary anti-/nanobodies, without being amplified, was performed with either antibodies conjugated to Alexa Fluor 488 (AF488, #706-545-148, Dianova), Alexa Fluor 546 (AF546, #A11030, Invitrogen, Waltham, USA), Abberior Star 580 (AS580 #ST580-1006, Abberior) or a nanobody that was conjugated in house to Alexa Fluor 546 (see the following paragraph). After all anti-/nanobody incubations, samples were washed 3x for 10 minutes with PBS and were postfixed with 4% PFA in PBS for 15–20 minutes at room temperature. Afterwards, samples were again quenched with 100 mM NH4Cl for 10 minutes at room temperature and kept in PBS until further processing.
Nanobody conjugation to Alexa Fluor546
Unconjugated single-domain antibody anti-mouse IgG1 was obtained from NanoTag Biotechnologies (#N2005). The nanobody was reduced using 10 mM Tris(2-carboxyethyl)phosphin -hydrochlorid (TCEP, #51805-45-9, Merck) on ice for 1h. The excess of TCEP was removed using a NAP5 column (Cytiva, Washington, USA) equilibrated with ice-cold PBS buffer, pH7.4, and mixed immediately with 3 molar excess of maleimide Alexa Fluor546 (#A10258, Thermo Scientific), followed by incubation at room temperature for another 2h. The conjugated nanobody was separated from free fluorophore using a Superdex 75 Increase 10/300 GL column (Cytiva) equilibrated with 2x PBS pH7.4. Conjugated fractions were pooled, the concentration was measured and the sample was diluted with glycerol (#2039, Chemsolute by Th. Greyer, Renningen, Germany) to reach 50% glycerol in 1x PBS, enabling the long-term storage in aliquots at -80°C.
Application of the AF488 amplification system
Samples containing labeling with AF488 could be further incubated with an antibody against AF488 (#A-11094, ThermoFisher Scientific) in BSA-T for 30 minutes at room temperature. A 3x wash with BSA-T was performed before samples were incubated in the final secondary antibody in BSA-T for 30 minutes at room temperature. Antibodies used here were either AF488 (#711-545-152, Dianova) or AS635P (#ST635P-1002, Abberior). Afterwards, samples were washed with PBS and postfixed, as described above.
Biotin amplification system
The immunostaining with primary nanobodies conjugated to biotin (BT) was performed as described above, relying on 30 minutes of incubation at room temperature with an antibody against BT (#31852, ThermoFisher Scientific), diluted in BSA-T. Afterwards, samples were washed 3x for five minutes with BSA-T, before the incubation with the last antibody carrying AS635P (#ST635P-1055, Abberior), AS580 (#ST580-1002, Abberior) or AF488 (#705-545-147, Dianova). This antibody was applied for 30 minutes at room temperature. Final washes with PBS and postfixation were performed as described above.
ALFA and Spaghetti Monster amplification system
After immunostaining the samples with primary nanobodies carrying an ALFA-tag, two possible procedures were followed to reveal the ALFA-tag in our X10ht methodology. For the condition termed “BEFORE AC”, the samples were incubated for 30 minutes at room temperature with a NbALFA (#N1502, NanoTag Biotechnologies) as it is performed for general immunostainings directly after a 3x wash with BSA-T. The NbALFA was produced fused to a protein called Spaghetti Monster (SpaMo, [33]), which is an engineered GFP containing 7 FLAG-tags (custom made by NanoTag Biotechnologies). Following the incubation with the SpaMo, samples were washed 3x five minutes in BSA-T and incubated with an antibody against the FLAG-tag (#F1804 Sigma-Aldrich, now Merck, or #14793 Cell signaling, Leiden, the Netherlands), for 30 minutes at room temperature. After repeated washing with BSA-T, samples were incubated with fluorophore-conjugated secondary antibodies (AS635P #ST635P-1001; AS580 #ST580-1001, both Abberior; Cy3 #715-165-150, Dianova) or nanobodies (AS635P, #N1202-Ab635P, NanoTag Biotechnologies) for 30 minutes at room temperature, before washing with PBS and post-fixation, as described above.
Alternatively, we labelled the ALFA-tagged nanobodies during the expansion procedure (for better understanding see scheme in Fig. 5A). For the “AFTER AC” immunostaining, only the primary nanobody carrying the ALFA-tag was applied during the general immunocytochemistry (pre-expansion). The samples underwent the expansion procedure (which is described in detail below), stopped after the samples were autoclaved (AC). At this time point, the samples were incubated with the NbALFA-SpaMo in BSA-T overnight at room temperature on a slowly rotating shaker. The next day, samples were washed carefully in PBS (5x, 1–2 hours) and incubated in anti-FLAG-tag antibody in BSA-T overnight at room temperature on a shaker. Unbound antibody was washed away with PBS 5x for 1–2 h, before the last antibody immunostaining was performed, with fluorophore-conjugated secondary anti- or nanobodies. This final step was performed in BSA-T overnight at room temperature, on a slow rotating shaker. Alternatively, using the secondary AS635P nanobody, a premixing of the anti-FLAG-tag antibody with the respective nanobody (30 minutes at room temperature in a 1:5 molar ratio antibody:nanobody), which helps to reduce the incubation step to one overnight incubation, instead of two. The next day, samples were ready to continue with the last steps of the expansion procedure (see later paragraphs).
Expansion procedure
Immunostained samples were processed as described in Truckenbrodt et al. 2019, with some alterations of the protocol, which is termed X10ht in the following paragraphs [8]. For maximal anchoring of the proteins and labels to the gel, we used 0.3 mg/ml Acryloyl-X (SE; #A-20770, ThermoFisher Scientific) in 150 mM NaHCO3 buffer (Merck, #144-55-8) with a pH of 11. Anchoring was performed overnight on a shaker at room temperature. Over the next day samples were washed 3x with PBS for five min, while the monomer solution was prepared as described before [25]. Briefly, 1.335 g N,N-dimethylacrylamide (DMAA, #274135, Sigma-Aldrich now Merck) and 0.32 g sodium acrylate (SA, #408220, Sigma-Aldrich now Merck) were dissolved in 2.850 g ddH2O and purged with N2 for 40 minutes. Afterwards, 300 µl of a 0.36 g/ml potassium persulfate (KPS, #379824, Sigma-Aldrich now Merck) stock solution was added to 2700 µl of the DMAA/SA solution and further purged with N2 on ice for 15 minutes. Immediately before usage, 4 µl of N,N,N’,N’-Tetramethylethylendiamine (TEMED, #612-103-00-3, Sigma-Aldrich now Merck) was added to 1 ml of the monomer solution, was vortexed and drops were pipetted on parafilm, on which the specimen was placed. Coverslips with cells were flipped to incubate cells in the solution, while drops were applied directly on the tissue slices and covered with a coverslip. Gels polymerized for 6–20 h at a stable temperature of 23°C in a humidified chamber.
The homogenization with 8 U/ml proteinase K (PK, #P4850 Sigma-Aldrich now Merck) in digestion buffer (50 mM TRIS (#AE15.2, Carl Roth GmbH), 800 mM guanidine HCl (#G3272, Sigma-Aldrich now Merck), 2 mM CaCl2 and 0.5% Triton X-100) was performed as described before [8,25]. This procedure was only performed for the gels used for images labeled “X10” in our Figures, and was not combined with heating-based homogenization.
In the present study, we present a heat homogenization step for 10X expansion by the application of a milder protein disruption via autoclaving over 100°C (AC), naming the new protocol X10ht. Polymerized gels were rinsed shortly with 1 M NaCl (#7647145, Merck) and afterwards extensively soaked in disruption buffer, consisting of 100 mM TRIS, 5% Triton-X and 1% SDS (#1057.1, Carl Roth GmbH) in ddH2O with pH 8, for minimal 2 h at room temperature, while buffer was exchanged 4x. Autoclaving was performed for 30 minutes at 110–121°C and gels were allowed to cool down slowly. To enable the overview imaging of the expanded tissue, slices were labeled with NHS-Fluorescein (#46409, ThermoFisher Scientific) over night after being extensively washed with PBS for 5x for 30 minutes. For expanding the gels after homogenization (both PK-digested and autoclaved gels) or NHS-Fluorescein, they were placed in 22 x 22 cm culture dishes and ~ 400 ml ddH2O was added to each gel. Water was exchanged every 30–90 minutes for 4x in total and gels were left in water over night to fully expand.
Investigation of optimal temperature for X10ht
Fixed BHK cells were processed for expansion as described in Truckenbrodt et al., 2018 until homogenization. Instead of applying proteinase K in digestion buffer, gels were shortly rinsed with 1 M NaCl and soaked in disruption buffer for ~ 2h while exchanging the buffer 4x. Afterwards, gels were autoclaved for 30 minutes at either 70, 80, 90, 100, or 110°C. Disruption buffer needed to be extensively washed away before adding 1 ml 1 µM Atto590-NHS ester (NHS-Atto590, #79636, Merck) in PBS to each gel, to label all retained proteins, followed by incubation overnight at room temperature on a slowly rotating shaker. On the next day, gels were expanded as it was described before [8] and imaged.
Image acquisition
For image acquisition, gels were cut into smaller pieces and placed in a selfmade imaging chamber on top of a glass coverslip. Water was removed to the maximum by simply soaking any unnecessary droplets with paper tissues. Most epifluorescence images were obtained with an inverted epifluorescence microscope, Nikon Eclipse Ti (Nikon Corporation), using a Plan Apochromat 60x objective (1.4 NA, oil immersion), a HBO-100W Lamp, and relying on an IXON X3897 Andor camera. Alternatively, we used a second inverted fluorescence microscope, Olympus IX 71 (Olympus, Hamburg, Germany), using a 20× objective (from Olympus). The Olympus setup employed a CCD camera (FView II, Olympus) and CellF software (Olympus). To identify NHS-AF590, a 545/30 HQ excitation filter (AHF, Tübingen, Germany), a 570 LP Q beam splitter and a 610/75 HQ emission filter were used.
An Abberior Expert line setup (Abberior Instruments) using an IX83 microscope (Olympus) with an 100x oil immersion objective (UPLSAPO, 1.4 NA; Olympus) was employed to generate confocal and STED images. Star635P was excited with a 640 nm excitation laser (set to 10–30% of max. power, nominally 1.77 mW, pulsed at 80 MHz). Signals were detected using an avalanche photodiode (APD) that has a preset range of 650–720 nm. A 561 nm excitation laser (40–50% of max. power, 440 µW; 80 MHz) was used to excite AS580 or Cy3, while the emission was detected with an APD, 605–625 nm. Depletion for AS635P, AS580 or Cy3 was achieved by a 775 nm depletion laser (set to 15–40% of max. power of 1.2 W). Excitation of AF488 was achieved by a 485 nm excitation laser (10–15% of max. power), and detected with an APD 525–575 nm. The used depletion laser for AF488 was a pulsed solid-state 595 nm laser (set to ~ 20% of max. power of 2 W). For all images, the pixel size was 20 nm, with a dwell time of 5–10 µs per pixel and a line accumulation of 3–5 for STED.
Immunocytochemistry, expansion and analysis of Nup96
In order to visualize expanded nuclear pore complexes, U2OS cells expressing mEGFP coupled to NUP96 were fixed, permeabilized and blocked as described above. Cells were incubated in a mixture of two different anti-GFP nanobodies, which were conjugated to Alexa Fluor 488 (#N0302-AF488, custom made; #N0303-AF488, custom made, both NanoTag Biotechnologies) for 1h at room temperature. Afterwards, they were washed 3x for five minutes with blocking solution and the AF488 amplification immunostaining was performed as described above. The last secondary antibody was conjugated to AS635P (#ST635P-1002, Abberior). Cells were washed 3x for five minutes with PBS and post-fixed before the expansion microscopy X10ht protocol was applied, followed by imaging with the Abberior STED setup. STED images were analyzed with a custom-made Matlab script, which measured the distances between all fluorescent localizations. Plot profiles of exemplary line scans were generated with ImageJ and data are plotted using GraphPad Prism 7. To analyze distances in Supplementary Figure S3D, we drew line scans across neighboring intensity peaks, and fitted them with a two-peak Gaussian distribution, which provides the peak positions, thus enabling us to determine the peak-to-peak distance.
Analysis of X10ht expansion factor of U2OS nuclei diameter
Fixed U2OS cells were treated by following either the original X10 protocol with PK digestion [8] or the X10ht protocol with the usage of heat denaturation (AC). After the respective homogenization procedure, gels were washed for around 2–4 h with PBS and incubated in 1 ml 1 µM Alexa Fluor 546 NHS ester (NHS-AF546, #A20002, Merck), overnight at room temperature on a slowly rotating shaker. The next day expansion of gels was performed like described before. For image acquisition, the Nikon Eclipse Ti epifluorescence microscope was used as described above. The measurements of nucleus diameter was performed by drawing vertical lines from the highest to the lowest point of each nucleus, using ImageJ. As control, the same measurements were performed in non-expanded cultures (-Ex). For this procedure, fixed cells were incubated in Hoechst (1:2000 in PBS, #62249, Thermo Scientific) for 5 minutes at room temperature, washed 3x for 10 minutes in PBS and embedded in Mowiol (#0713.2, Carl Roth GmbH).
Image processing, analysis and statistics
The scale bars presented in each figure takes into account the respective expansion factor.
In case of drifting, images were manually drift-corrected with ImageJ (Wayne Rasband, NIH, Bethesda, MD, USA). Signal intensities of images obtained by epifluorescence microscopy (Fig. 2) were analyzed with ImageJ by shrinking the average bin size of the raw file and measuring the total grey values of 5–30 pictures per condition.
Fluorescence intensity evaluation of confocal or STED images was performed on raw image files with a semi-automated custom-written Matlab (The Mathworks, Inc, Natick, MA, USA) routine, analyzing mean grey values per ROI. Line scans and plot profiles of the raw images presented in Figs. 2 and 6 were performed using Matlab and ImageJ.
Spot resolution was analyzed by performing line scans of 0.6 µm on fluorescent events on raw STED image files with Matlab (Fig. 6). FWHM was calculated by plotting the line scans and employing a Lorentzian fit. Another method to estimate the resolution was performed by applying the ImageJ plugin “decorrelation analysis” from Descloux and colleagues [37], which is freely available under doi.org/10.1038/s41592-019-0515-7. Distance analysis of vesicular fluorescent spots were performed with ImageJ on raw STED images.
All scatter plots were generated and statistics were performed using GraphPad Prism 7 (GraphPad Software Inc., La Jolla, CA, USA). All data were analyzed for normal distribution by the Shapiro Wilk-test and were then tested for significance with students-t test (when normally distributed) or the Mann Whitney-U-test (when not normally distributed). Data is presented as individual data points and means ± SD. Differences were considered significant and were indicated by asterisks with *P < 0.05; **P < 0.01; ***P < 0.001, and ****P < 0.0001. Illustrations for Fig. 4 were created with Bio-Render (BioRender.com; Science Suite Inc., Canada) and figures were prepared with Adobe Illustrator CS6 (Adobe Systems Incorporated, Mountain View, USA). For better visualization, images of figures were scaled in brightness and contrast unless stated otherwise.