Correlative confocal and scanning electron microscopy of cultured cells without using dedicated equipment

Protocol Correlative confocal and scanning electron microscopy of cultured cells without using dedicated equipment This protocol enables correlative light and electron microscopy (CLEM) imaging of cell surface features without using dedicated equipment. Cells are cultured and fixed on transparent substrates for confocal microscopy imaging. No conductive coating is employed in the scanning electron microscopy workflow, providing a clean cell surface observation, with fiducial markers assisting alignment of optical and topographical images. This protocol describes CLEM imaging for midbody remnants in MDCK cells but can also be applied to different cell types and surface features. Javier Casares-Arias, Miguel A. Alonso, A ´lvaro San Paulo, Marı ´aUjue´Gonza´lez [email protected]. ch (J.C.-A.) maria-ujue.gonzalez@ csic.es (M.U.G.) Highlights A CLEM protocol without dedicated equipment requirements Confocal and SEM data sets are acquired on independent setups Large-scale sample features are used for initial correlation and navigation Gold nanobeads are used as fiducial markers during final image alignment Casares-Arias et al., STAR Protocols 2, 100727 September 17, 2021 ª2021 The Author(s). https://doi.org/10.1016/ j.xpro.2021.100727 ll OPEN ACCESS Protocol Correlative confocal and scanning electron microscopy of cultured cells without using dedicated equipment Javier Casares-Arias, 1,3,5,6, *Miguel A. Alonso, 1 A ´lvaro San Paulo, 2,4 and Marı ´a Ujue ´Gonza ´lez 2,4,5, * 1 Centro de Biologı ´a Molecular Severo Ochoa, Consejo Superior de Investigaciones Cientı ´ficas and Universidad Auto ´noma de Madrid, 28049 Madrid, Spain 2 Instituto de Micro y Nanotecnologı ´a, IMN-CNM, CSIC (CEI UAM+CSIC), Tres Cantos, Madrid 28760, Spain 3 Present address: Eidgeno ¨ssische Technische Hochschule (ETH) Zurich, Department of Biosystems Science and Engineering, 4058 Basel, Switzerland 4 These authors contributed equally 5 Technical contact 6 Lead contact *Correspondence: [email protected] (J.C.-A.), maria-ujue.[email protected] (M.U.G.) https://doi.org/10.1016/j.xpro.2021.100727 SUMMARY This protocol enables correlative light and electron microscopy (CLEM) imaging of cell surface features without using dedicated equipment. Cells are cultured and fixed on transparent substrates for confocal microscopy imaging. No conductive coating is employed in the scanning electron microscopy workflow, providing a clean cell surface observation, with fiducial markers assisting alignment of optical and topographical images. This protocol describes CLEM imaging for midbody remnants in MDCK cells but can also be applied to different cell types and surface features. For complete details on the use and execution of this protocol, please refer to Casares-Arias et al. (2020). BEFORE YOU BEGIN To be identified under the confocal microscope, the biological structures must be labeled with fluorescent molecules prior to imaging. Immunofluorescence labeling protocols, widely used in the field, usually include a permeabilization step that leads to ultrastructural detail loss, rendering the use of antibodies unsuitable for a CLEM approach. Therefore, in order to localize and analyze the structures of interest, fluorescent fusion proteins or similar approaches, such as click chemistry (SNAP/HALO-tag), must be used. This protocol can be applied to different cell types and surface features. In the examples included to illustrate the protocol, a MDCK cell line stably expressing two different fluorescent fusion proteins has been used. For further details, please refer to (Casares-Arias et al., 2020). KEY RESOURCES TABLE REAGENT or RESOURCE SOURCE IDENTIFIER Chemicals, peptides, and recombinant proteins Minimum Essential Media (MEM) Thermo Fisher Scientific Cat#31095-029 Trypsin Thermo Fisher Scientific Cat#27250-018 EDTA (Tritiplex III) Merck Cat#108418.0250 (Continued on next page) STAR Protocols 2, 100727, September 17, 2021 ª2021 The Author(s). This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). 1 ll OPEN ACCESS MATERIALS AND EQUIPMENT Confocal microscopy Images were captured using a 603water objective (NA 1.2) on a Nikon A1R+ confocal microscope. 488 and 561 nm laser lines were used to image the fluorescent fusion proteins, 640 nm laser line was 13fixing solution Reagent Final concentration Amount Paraformaldehyde (8% in water) 2% 2.5 mL Glutaraldehyde (8% in water) 2% 2.5 mL Phosphate Buffer 1M pH 7.4 0.1 M 1 mL ddH 2 O n/a 4 mL Total n/a 10 mL 13Fixing solution can be stored at room temperature (20C–22oC) for up to a month. 23fixing solution Reagent Final concentration Amount Paraformaldehyde (16% in water) 4% 2.5 mL Glutaraldehyde (16% in water) 4% 2.5 mL Phosphate Buffer 1M pH 7.4 0.2 M 2 mL ddH 2 O n/a 3 mL Total n/a 10 mL 23Fixing solution can be stored at room temperature (20C–22oC) for up to a month. Continued REAGENT or RESOURCE SOURCE IDENTIFIER Poly-L-Lysine Merck Cat#P1524 G418 Santa Cruz Cat#Sc-29065B Phosphate Buffer Solution 1 M pH 7.4 Merck Cat#P3619 Paraformaldehyde Merck Cat#30525-89-4 Glutaraldehyde EMS Cat#16220 Ethanol Merck Cat#1.00983 Acetone Merck Cat#24201-M Hexamethyldisilazane (HMDS) Merck Cat#440191 250 nm Gold nanoparticles BBI Solutions Cat#em.gc250 Experimental models: Cell lines MDCK cell line ATCC Cat#CRL2936 Recombinant DNA mCherry-Tubulin Takara Bio N/A pNG72-GFP-L-CHMP4B Juan Martı ´n Serrano, King’s College London (Ventimiglia et al., 2018) Software and algorithms FIJI-ImageJ2 (Rueden et al., 2017)imagej.net/ImageJ2 TrakEM2 (Cardona et al., 2012)https://www.ini.uzh. ch/acardona/trakem2.html Other 35 mm glass-bottom dishes MatTek Cat#P35G-1.5-20-C A1R+ confocal microscope Nikon N/A 3MCopper Conductive Tape, Single Adhesive Surface Ted Pella, Inc. Cat#16072 3MXYZ Axis Tape, Electrically Conductive, Double Sided, 9712 Ted Pella, Inc. Cat#16081 Large Sample Stub for SEM, B32 mm Ted Pella, Inc. Cat#16148 Field Emission SEM Verios 460 FEI (now Thermo Scientific) N/A ll OPEN ACCESS 2STAR Protocols 2, 100727, September 17, 2021 Protocol used on reflection mode to image the gold nanoparticles (Au NPs). Specific acquisition parameters for each image set are detailed on the step-by-step method details section. Note: Additional tracks, such a transmitted light, can also be acquired during confocal imaging. Scanning electron microscopy Scanning electron microscopy (SEM) images were acquired with a Field Emission SEM FEI (now Thermo Scientific) Verios 460. High-end recent SEMs, as this model, are able to provide high resolution in the very low voltage (VLV) range (%1 nm at 1 kV). This has allowed us to image uncoated biological samples on glass substrates. Note: Though we strongly recommend the use of VLV-SEM to avoid the need of coating the samples with conductive layers, the protocol can be adapted to allow the utilization of more conventional SEM equipment (see troubleshooting 1: Using Alternative SEM without VLV capabilities). The imaging of uncoated biological samples has the advantage of reducing possible artifacts in the observed morphology due to the deposited layer. However, the reduced conductivity of the samples induces charging effects that degrade the image quality and therefore have to be minimized. The use of VLV is a first strategy to achieve this. Additional methods include the use of very short dwell times (50–100 ns). To maximize the signal to noise ratio in these conditions, frame integration scanning modes need to be used (around 100–200 frames per image are recommended), with software-based drift correction option activated (if this option is not available, charge-originated drift maydestroythequalityofimageafterframeintegration). Particular acquisition parameters for each image set are specified on the step-by-step method details section. Most SEM systems providing high resolution at very low voltages have implemented the option of applying beam deceleration (Zarraoa et al., 2019). This method consists in applying a negative bias voltage (typically 0.5–4 kV) to the sample holder in order to decelerate the incident beam before reaching the sample surface. As a consequence, the effective beam landing energy equals the column acceleration voltage (which is kept relatively high from 2 to 5 kVfor optimum column performance and resolution) minus the sample bias. The resulting electric field between the pole piece and the sample acts as an additional electrostatic lens that reduces the beam diameter and enhances secondary electron (SE) collection at the detector placed inside the column –also called in-lens detector– improving spatial resolution. The SE and back-scattered electrons are also spatially redistributed by this electric field and the amount of SE more pathologically affected by charge effects that are collected by the detector is reduced. Therefore, the use of beam deceleration can improve both the image resolution and the charge management. STEP-BY-STEP METHOD DETAILS Coating and addition of gold fiducial markers to the coverslip Timing: 2 h 1. Cover the surface of the inset of a glass-bottomed 35 mm Petri dish with 650 mLof0.1%(wt/vol) poly-L-lysine solution (dissolved in water) and incubate for 30 min at 37C(Figure 1A). Note: On the presented example, poly-L-lysine has been used only to ensure correct adsorption of the Au NPs to the glass surface, since MDCK cells do not require any substrate coating to grow. Though this possibility has not been tested, alternative coatings required to ensure cell attachment should not interfere with the protocol, provided that they allow the correct visualization of the Au NPS over the substrate. ll OPEN ACCESS STAR Protocols 2, 100727, September 17, 2021 3 Protocol 2. Rinse with deionized water and air-dry. Note: Though sterile working conditions are optional during this phase, drying steps should be performed in a fume hood to keep the coverslip surface as clean as possible. 3. Prepare a suspension of Au NPs (in deionized water) to achieve a final density along the coverslip surface of around 1–5 310 4 particles mm 2 . a. Thoroughly vortex the Au NPs stock before pipetting the required volume on a separate tube. b. Sonicate the suspension in a bath sonicator for 3 cycles of 30 s at the highest output setting to break up NP aggregates. c. Dilute the suspension to the desired concentration in deionized water. Note: In our case, 250 nm spherical gold nanoparticles were used, as they can be easily observed both under the confocal and scanning electron microscopes. Au NPs with smaller sizes or different shapes can also be used, though final density may need to be adjusted to ensure that individual NPs can be observed near the structures of interest. CRITICAL: The presence of salts in the Au NPs suspension could cause undesired aggregation. 4. Add 650 mLofAuNPssuspensiontotheinset,incubatefor2hat37 C(Figure 1B). Note: You may need to optimize the final density of fiducial markers by adjusting Au NPs concentration and/or incubation time. Au NPs suspension volume may vary for different dish sizes. 5. Retire the remaining solution and air-dry (Figure 1C). 6. Mark the bottom of the coverslip with an ethanol-resistant marker to orient the sample during subsequent image acquisition steps. Two marks, defining ‘‘north’’ and ‘‘east’’ of the sample are typically enough (Figure 1D). Cell culture Timing: 2–3 days Figure 1. Coverslip coating and fiducial markers attachment procedure First, the coverslip is coated with Poly-L-lysine (A), followed by incubation with a Au NPs suspension (B) and then airdried (C). Once the surface is ready for cell culture, reference marks are added to the bottom of the coverslip for further alignment (D). ll OPEN ACCESS 4STAR Protocols 2, 100727, September 17, 2021 Protocol CRITICAL: These steps should be carried out under sterile working conditions. Note: This protocol has been optimized for the imaging of midbody remnants on renal epithelial MDCK cells, other cell types or structures of interest may require additional steps or different conditions. 7. Sterilize the dish previously coated with Au NPs by exposing it to the germicide lamp of a cell culture hood for 15–30 min. 8. Add 2 mL of MDCK cells expressing the desired fluorescent fusion proteins (7.5 310 4 mL 1 )suspended in MEM to the plate. CRITICAL: The initial number of cells must be optimized according to the cell line and the duration of the culturing phase, so that the culture is not completely confluent by the time of fixing. Note: Culture media may vary for different cell types. 9. Culture the cells for 48–72 h at 37Cinanatmosphereof5%CO 2 . CRITICAL: For the alignment strategy to work, culture must be subconfluent at the time of fixing. For this, the initial number of cells and duration of the culturing phase must be adjusted. The goal is not to achieve a specific percentage of confluence, but to ensure that the substrate and Au NPs are exposed in some areas (see Figure 2). If the use of subconfluent cultures represents a limitation, see troubleshooting 2: Use of confluent cultures. Fixation Timing: 2–10 h CRITICAL: Only electron microscopy-rated fixatives must be used, common-use aldehydes might include traces of organic solvents, such as methanol, that can damage the cell surface. 10. Add a volume of 23fixing solution (4% paraformaldehyde + 4% glutaraldehyde in phosphate buffer 0.2 M) equal to the volume of culture medium (MEM) in the dish. Incubate for 10 min at room temperature (RT; 20C–22C). 11. Remove most of the liquid and add fresh 13fixing solution (2% paraformaldehyde + 2% glutaraldehyde in phosphate buffer 0.1 M). Incubate for 2 h at 20C–22Cor8–10hat4 C. Pause point: Samples can be stored in fixing solution at 4Cforuptoaweek. Alternative: Different storage solutions, such as PBS, can also be used. In such case, the addition of antibacterial agents (Penicilin-Streptomycin or sodium azide) is advised. Confocal imaging Timing: 4–6 h During this phase, the localization of the surface features of interest labeled with fluorescent proteins and the Au NPs present in the sample is determined by confocal fluorescence and reflection microscopy, respectively. Candidate structures for CLEM analysis are selected, and all the relevant confocal images are acquired. ll OPEN ACCESS STAR Protocols 2, 100727, September 17, 2021 5 Protocol 12. Substitute fixing solution with phosphate buffer 0.1 M. 13. Place the dish on the sample stage, using the marks added during step 6 as a reference. Perform the imaging at the center of the coverslip if possible. 14. Acquire a large field-of-view (FOV) (2 32 mm) image for alignment and navigation purposes, including the fluorescence signals and the reflection channel showing the Au NPs. In our case, these images were generated by acquiring a 10 310 tilescan z-stack (5 mmtotalwith 250 nm steps) with a resonant scanner, 83averaging and pixel size of 200 nm (Figure 2A). Note: Acquisition settings must be optimized depending on the abundance of the structure of interest, the proportion of cells expressing the fluorescent protein and signal strength. CRITICAL: We strongly recommend performing all the confocal imaging in one session to guarantee a fixed sample orientation. If this is not possible, the rest of the images can be Figure 2. Confocal dataset of the CLEM procedure (A) Confocal large-FOV image (contrast enhanced). (B) Confocal medium-FOV image. (C) Look-up image showing the localization of the medium-FOV images and structures of interest (squares and circles respectively). Arrowheads in (B) show the position of the Au NPs that are closer to the structure of interest (circle). Magenta box in C corresponds to the imaging area shown in B. ll OPEN ACCESS 6STAR Protocols 2, 100727, September 17, 2021 Protocol acquired on a different session, though alignment might be an issue and require special attention. 15. Localize candidate surface features and acquire medium-FOV (100 3100 mm) images, including the fluorescence signals and the reflection channel showing the Au NPs. We acquired these images with a unidirectional galvano scanner, no averaging and 50 nm pixel size (Figure 2B). Note: Thecloseracandidatestructureistoanareawithexposedsubstrate,theeasierwillbe to localize it under the scanning electron microscope (SEM). CRITICAL: When selecting the imaging region, be sure to include an area with exposed substrate including Au NPs near the structure(s) of interest (see arrowheads on Figure 2B). 16. As medium-FOV images are recorded, keep track on their position relative to the large-FOV image. This information is needed for the elaboration of the look-up map (see next step). A convenient way of doing this is by using the option that most microscope manufacturers offer of re-centering the stage to a given position of a previously acquired image when a motorized stage is available. During this process, the position of each medium-FOV image is generally annotated on an additional copy of the large-FOV image. If this capability is not available, see ‘‘troubleshooting 3: Confocal dataset tracing’’ for alternatives. 17. Prior to sample removal, record the objective position, and thus imaging area, relative to the coverslip. For this, a picture of the stage adaptor during the imaging session can be taken. Additionally, once the backside of the coverslip has been cleaned, make an additional ethanol-resistant mark on the same position to frame the imaged area. Note: The mark indicating the area imaged by confocal microscopy is used as starting point for SEM imaging, and thus should be as accurate as possible. Pause point: Samples can be stored for up to a week at 4C upon substitution of the phosphate buffer by 13fixing solution. Alternative: Different storage solutions, such as PBS, can also be used. In such case, the addition of antibacterial agents (Penicilin-Streptomycin or sodium azide) is advised. Look-up map elaboration Timing: 30 min In order to facilitate the correlation process, it is recommended to elaborate a look-up map prior to the SEM imaging session. This consists of the large-FOV confocal image of the full area where the position of each medium-FOV imaged is indicated. This will be helpful both for navigation and identification of individual structures in the sample along the SEM session. 18. Open the large-FOV image in ImageJ. Given the file size, performing a z-projection and creating a compressed jpeg version is advised. 19. Highlight and identify the regions that correspond to each medium-FOV image, repeat the process with the candidate structures present inside them. This will allow to identify each structure, keeping track of its corresponding medium and large-FOV images (Figure 2C). Coverslip extraction Timing: 5 min ll OPEN ACCESS STAR Protocols 2, 100727, September 17, 2021 7 Protocol CRITICAL: This process involves removing most of the liquid from the dish, it must be carriedoutasfastaspossibleto avoid uncontrolled drying. Note: This phase must be carried out on a fume hood. 20. Remove most of the liquid from the dish, turn it upside-down and add some drops of acetone around the perimeter of the coverslip. Incubate for 5–10 s (Figures 3A and 3B), then and clean any remaining acetone. 21. Now that the adhesive is soft, apply pressure homogeneously along the coverslip perimeter from the inside (sample side). This can be done by pushing the dish against a hollow cylinder with a slightly smaller diameter than the inset (Figures 3C and 3D). Note: This coverslip detaching technique involves touching the periphery of the sample surface, and must be carried out with extreme care. Alternative: If a hollow cylinder of the appropriate size is not available, or the nature of the sample is incompatible with the aforementioned technique, the coverslip can also be detached with the help of a scalpel. For this, Introduce a scalpel in between the coverslip and the plastic and gently rotate the dish to liberate the coverslip (Figures 3E and 3F). This technique, however, is prone to coverslip cracking and must be carried out with extreme care. 22. Once the coverslip is detached, place it (sample facing up) quickly on a 6-well plate with phosphate buffer and remove any remaining adhesive if needed. Figure 3. Coverslip extraction procedure (A) First, imaged region is marked on the bottom of the coverslip. (B–F) (B) Following buffer removal, the dish is turned upside-down, and acetone is added around the perimeter of the coverslip. Once the adhesive has softened, coverslip is extracted by uniformly applying pressure from the sample-side on the coverslip with a hollow cylinder (C and D) or by using a scalpel (E and F). ll OPEN ACCESS 8STAR Protocols 2, 100727, September 17, 2021 Protocol (steps 23–25). It has been previously reported that drying of biological samples, regardless of the methodology used, decreases cell volume (Katsen-Globa et al., 2016). On a CLEM workflow, this results in a slight misalignment of the resulting images. Though fiducial markers over the substrate remain immotile, making alignment possible, the small movements of the cell surface due to shrinking could modify to some extent the localization of the structures of interest (see troubleshooting 5: Image alignment). TROUBLESHOOTING Problem 1 Using Alternative SEM without VLV capabilities In this protocol, a scanning electron microscope capable of achieving subnanometric resolution under very low voltage (VLV) conditions has been used. This allows the analysis of uncoated biological samples on glass substrates, reducing sample manipulation and delivering the highest possible resolution. If a conventional SEM microscope without VLV capabilities is used, charging effects will appear during imaging (steps 27 and 28). Potential solution Coating of the sample with a conductive material (such as gold, platinum or carbon) would minimize charging effects and allow the usage of higher beam voltages. However, this would be achieved at theexpenseofthelevelofdetailinthefinalimage,sincetheconductivelayerwillevenoutthefinest structural details of the sample, potentially hiding relevant features. The coating should be done after sample dehydration and drying (subsequent to step 25). Deposition of a thin conductive coating, between 2 and 15 nm, by means of a physical vapor deposition technique should be enough. The thickness of the coating needs to be adjusted to allow the visualization of the Au nanoparticles. Problem 2 Use of confluent cultures Figure 7. Expected outcome: correlated light and scanning electron microscopy images (A and B) The pattern formed by the Au NPs on the substrate (arrowheads) has been used to align both images, allowing their superposition (C). The position of the small-FOV image (C, inset) can be traced back to the original confocal dataset, this allows to combine the nanometric-range topographical SEM measurements with the multilabeling protein localization data acquired through confocal microscopy. ll OPEN ACCESS STAR Protocols 2, 100727, September 17, 2021 15 Protocol The proposed methodology for CLEM imaging strongly relies on the existence of large-scale patterns on the sample, such as the overall shape of the cell colonies, to localize the imaging area and perform an initial rough alignment. A confluent cell monolayer after the cell culture phase (step 9) looks homogeneous under low-magnification conditions, preventing the direct application of this protocol. Potential solution In order to introduce a higher degree of physical confinement for the cells, while maintaining orientation cues inside the sample, micro-patterned substrates could be used. Another possibility is the creation of a large-scale pattern on the sample after fixation by scratching out portions of the cell monolayer. This process, however, damages the cells right at the border of the pattern, which is the area where the alignment strategy works best. Problem 3 Confocal dataset tracing Once the confocal large-FOV image has been acquired and the candidate features identified, the proposed methodology relies on the use of a motorized sample stage for repositioning during medium-FOV images acquisition (steps 15 and 16). In the absence of a motorized sample stage, or if the navigation cues present in the sample are too complex, this process might be challenging. Potential solution In this case, the use of gridded coverslips (Mattek, Cat#P35G-1.5-14-CGRD) may be of help. The squared pattern embedded on the substrate, together with the alphanumeric code that identifies each square, allows to localize each medium-FOV image for look-up map elaboration. For this, an additional transmitted-light track has to be acquired during confocal imaging, as to make the pattern visible in the final images. Problem 4 Charging during SEM imaging During SEM imaging (steps 27 and 28) of non-conductive samples, the accumulation of negative charges on the specimen surface causes strong contrast variations in the image or drift, precluding the obtention of good quality images. Potential solution We recommend adapting the SEM imaging conditions as follows, with each step providing a higher level of complexity. 1. Keep the voltage at 1 kV but reduce the beam current. 2. Keep beam voltage at 1 kV and beam current at 13 pA (or lower beam current if desired), and apply beam deceleration (see materials and equipment section). A typical value for the sample bias to employ is 1000 V. Note: Beam deceleration strategies should be applied too if the image quality at 1 kV is too low. CRITICAL: The use of beam deceleration strategies usually prevents the option of tilt imaging. ll OPEN ACCESS 16 STAR Protocols 2, 100727, September 17, 2021 Protocol 3. Reduce beam voltage at 500 V and keep beam current at 13 pA (or lower beam current if desired). Beam deceleration is usually mandatory in this case to obtain high resolution images at this very low voltage. A typical value for the sample bias voltage to be applied is 1500 V. Note: These softer conditions (options 1, 2 and 3) can also be useful for preventing damage of the structures. Problem 5 Image alignment Dehydration and drying procedures carried out between confocal and SEM imaging (steps 23–25) maycausesubtledeformationsofthecellsurfaceduetoshrinking.SincetheAuNPsaredirectly attached to the substrate, and thus unaffected by these deformations, there might be a slight shift in the position of the features of interest when confocal and SEM images are compared after alignment (step 41). Additionally, the proposed methodology is optimized for the analysis of features that are relatively close to areas with exposed substrate, since alignment accuracy decreases with distance. Potential solution Image alignment precision can be improved, and its application range extended by applying the Au NPs directly on top of the cell surface prior to fixation (step 10). This way, the Au NPs would be affected by the deformation of the cell surface at the same extent as the features of interest, allowing its compensation. This alternative approach would also allow the correlative analysis of features that are far from exposed substrate areas. Though the vast majority of the protocol would be unaffected by this modification, the final image alignment process should be modified to include the movement of the Au NPs between imaging methods. While the proposed methodology uses only translation, rotation and scaling to achieve alignment (step 36), elastic deformation (‘‘Affine’’ on TrakEM2) of one of the datasets might be needed to compensate the movement of Au NPs placed on top of the cells. RESOURCE AVAILABILITY Lead contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Javier Casares Arias ([email protected]). Materials availability This study did not generate new unique reagents. Data and code availability The published article includes all datasets generated or analyzed during this study. ACKNOWLEDGMENTS The expert technical advice of the Optical and Confocal Microscopy Facility of CBMSO is gratefully acknowledged. We acknowledge the Micro and Nanofabrication Laboratory of the Instituto de Micro y Nanotecnologı ´a (MiNa), which is funded by the Comunidad de Madrid (S2018/NMT-4291 TEC2-SPACE), MICIN (project CSIC13-4E-1794), and EU (FEDER, FSE), for support with SEM infrastructure. Research in the laboratory of MAA was supported by a grant (PGC2018-095643-B-I00) from the Spanish Ministerio de Ciencia e Innovacio´ n (MICINN), Agencia Estatal de Investigacio´n, and the Fondo Europeo de Desarrollo Regional, European Union (MICINN/AEI/FEDER, EU). A contract (FPU14/00295) and a short-term fellowship from EMBO to J.C.A. are also acknowledged. ll OPEN ACCESS STAR Protocols 2, 100727, September 17, 2021 17 Protocol AUTHOR CONTRIBUTIONS Conceptualization, J.C.A., A.S.P. and M.U.G.; investigation, J.C.A., A.S.P., and M.U.G.; writing – original draft, J.C.A. and M.U.G.; writing – review & editing, J.C.A., M.A.A., A.S.P., and M.U.G.; funding acquisition, M.A.A., A.S.P.,andM.U.G.;supervision,M.A.A. DECLARATION OF INTERESTS The authors declare no competing interests. REFERENCES Cardona, A., Saalfeld, S., Schindelin, J., ArgandaCarreras, I., Preibisch, S., Longair, M., Tomancak, P., Hartenstein, V., and Douglas, R.J. (2012). TrakEM2 software for neural circuit reconstruction. PLoS One 7, e38011. Casares-Arias, J., Gonza ´lez, M.U., San Paulo, A., Ventimiglia, L.N., Sadler, J.B.A., Miguez, D.G., Labat-de-Hoz, L., Rubio-Ramos, A., Rangel, L., Bernabe ´-Rubio, M., et al. (2020). Midbody Remnant Inheritance Is Regulated by the ESCRT Subunit CHMP4C. iScience 23, 101244. Katsen-Globa, A., Puetz, N., Gepp, M.M., Neubauer, J.C., and Zimmermann, H. (2016). Study of SEM preparation artefacts with correlative microscopy: Cell shrinkage of adherent cells by HMDS-drying. Scanning 38, 625–633. Rueden, C.T., Schindelin, J., Hiner, M.C., DeZonia, B.E., Walter, A.E., Arena, E.T., and Eliceiri, K.W. (2017). ImageJ2: ImageJ for the next generation of scientific image data. BMC Bioinformatics 18, 529. Ventimiglia, L.N., Cuesta-Geijo, M.A., Martinelli, N., Caballe, A., Macheboeuf, P., Miguet, N., Parnham, I.M., Olmos, Y., Carlton, J.G., Weissenhorn, W., and Martin-Serrano, J. (2018). CC2D1B Coordinates ESCRT-III Activity during the Mitotic Reformation of the Nuclear Envelope. Dev Cell 47, 547– 563.e6. Zarraoa, L., Gonza ´lez, M.U., and San Paulo, A. (2019). Imaging low-dimensional nanostructures by very low voltage scanning electron microscopy: ultra-shallow topography and depth-tunable material contrast. Sci. Rep. 9, 16263. ll OPEN ACCESS 18 STAR Protocols 2, 100727, September 17, 2021 Protocol