High-throughput Imaging and Analysis Workflow for Evaluating Skin Cell Phenotypes and Proliferation States in Tissue Samples

Summary

The combination of iterative-bleaching-extends-multiplexity (IBEX) and a commercial nucleotide labeling assay (Click-iT EdU) enables the detection and categorization of dividing cell types in highly dynamic processes in fixed frozen murine tissue sections. Furthermore, a novel open-source image processing pipeline provides high-throughput image acquisition and analysis.

Abstract

Skin injuries initiate a complex regenerative cascade that engages various cell types. Amongst those, peripheral glial cells play a critical role in ensuring the success of the repair process. However, our understanding of these processes remains incomplete due to limitations in current technologies. Therefore, the iterative-bleaching-extends-multiplexity (IBEX) method was employed to characterize changes to the cellular composition of the skin using antibodies directed against proteins expressed by key cell types and structures involved in tissue regeneration. Importantly, antibodies directed against cell proliferation markers often underestimate the number of dividing cells in the skin. To overcome this limitation, Click-iT EdU chemistry was combined with the IBEX method, allowing detailed characterization of proliferating cell types. Using a fully automated spinning disk confocal microscope, a high-throughput workflow was developed for iterative imaging of samples sectioned into 24-well plates. Beyond extending the IBEX method to evaluate cell states in situ with Click-iT EdU, a novel open-source image processing pipeline was also introduced to perform image correction, stitching, and registration, and can be used by researchers without extensive programming experience. Moreover, a detailed documentation on processing and visualizing highly multiplexed imaging data (using conventional image analysis software, including Fiji and QuPath) is also provided. In summary, this protocol highlights the versatility of the IBEX protocol and demonstrates its adaptation to an established conventional imaging platform. Notably, combining IBEX with Click-iT EdU labeling enables the detection and categorization of actively dividing cell types in intact tissues and during highly dynamic processes at spatial resolution.

Introduction

The skin is the largest organ of the human body and the primordial barrier to protect the body from the outside environment¹. Consequently, it is exposed to a plethora of external challenges that compromise its barrier function. Therefore, the skin has developed complex mechanisms to restore tissue integrity and functionality following damage. Skin wound healing takes place in three overlapping stages: inflammation, proliferation, and maturation/remodeling. These three stages involve several different cell types, such as dermal fibroblasts and immune cells, that act together to restore the initial biological properties of the injured skin tissue^1,2,3.

The skin is a densely innervated organ, and it is known that nerve innervation is crucial for a successful repair process^4,5,6. Apart from axonal signaling^4,7,8, the peripheral glia, normally ensheathing the axons, participate in successful tissue regeneration^9,10,11. However, the role of underrepresented cell types, such as peripheral glial cells, remains poorly understood in dynamic processes like skin tissue regeneration. To better determine the function of these cells, it is critical to analyze the cellular microenvironment and the neighboring cell types with which they may interact during the regenerative process. Moreover, since cell proliferation is a key stage for tissue regeneration, it is important to accurately detect and classify the proliferating cell types throughout the repair process.

Multiplexed, high-content, optical imaging methods, such as IBEX¹², have been developed to characterize the cellular composition, visualize and quantify cell-cell interactions, and dissect microenvironmental patterns within complex tissues at spatial resolution. Whereas other multiplexing methods require specialized instruments and proprietary reagents, a particular advantage of IBEX is that it can be established at relatively low costs using commercially available antibodies, reagents, and microscopes accessible in a standard research environment. While anti-Ki-67 antibodies, often used in IBEX panels, reliably label proliferating cell types in many tissues^13,14, the number of Ki-67-positive (Ki-67+) cells was found to be underestimated compared to other methodologies in fixed frozen tissue sections of acute murine skin wounds. Therefore, we combined the Click-iT EdU chemistry with IBEX and found that the number of EdU⁺ cells more accurately represents the total number of proliferating cells obtained by other methodologies.

Furthermore, by employing a fully automated spinning disk confocal microscope, a high-throughput workflow was established for iterative imaging of tissue sections placed into a 24-well plate. The resulting images are then processed by combining multiple open-source Python tools into a novel image processing pipeline that corrects the images for uneven illumination, performs the stitching of the individual tiles, and ultimately, image registration. In detail, the BaSiCPy library¹⁵ was used for illumination correction, the m2stitch library for stitching the images, and phase-cross-correlation for cycle registration. The registered images are saved as OME-TIFFs and can then be loaded into conventional image analysis software, such as Fiji¹⁷ and QuPath¹⁸, where the images are then further processed, and cell type categorization and other measurements, like intensity and distance calculations, can be performed. Taken together, the present protocol allowed us to perform a detailed, highly efficient characterization of the main cell types of the regenerating skin, with a particular focus on the cellular microenvironment surrounding the peripheral glial cell population.

Protocol

​All animal breeding, housing, and experimentation were conducted according to the guidelines of the veterinary office of the Canton of Zurich, Switzerland.

1. Full-thickness skin wounds (Day 0)

1. Induce 8-10 weeks old mice with 5% isoflurane in 70% O₂ and maintain it using 2.5% isoflurane.
2. Shave the back skin and clean it thoroughly.
3. Disinfect prior to the surgery using antibacterial soap and 70% EtOH solution.
4. Apply preoperative analgesia using a subcutaneous injection of buprenorphine (30 µg∙mg∙kg⁻¹).
5. Generate four circular full-thickness excisional wounds of 5 mm in diameter on the lower back skin of each animal, two on each side, 1 cm from the midline of the animal, and roughly 2 cm apart from each other¹⁹.
6. Perform postoperative analgesia with a 3 day treatment of buprenorphine through drinking water (9 µg/mL with 2% sucrose).
7. Let the wounds heal without dressing.
8. Collect the wounds at defined time points for histological examination and flow cytometry analysis.
9. For the IBEX protocol, dissect and fix the skin tissues with 4% paraformaldehyde (PFA) solution at 4 °C overnight. Use a commercially available Biopsy Capsule to prevent curling of the samples during fixation. Cryoprotect the 4% PFA-fixed tissues in 30% sucrose in PBS overnight at 4 °C. Transfer the tissues to a 1:1 mixture of 30% sucrose with O.C.T compound for 2 h at RT and finally embed in O.C.T before snap-freezing in isopentane.

2. Sample preparation (Day 1)

1. Coat the 24-well plate with 200 µL of chrome alum gelatin for 15 min on a rocker shaker. Then, completely remove the gelatin and dry the coated wells for 60 min in a 40 °C oven. Coat double the wells needed as a backup. Add 2 mL of PBS per well.
   NOTE: Ensure that the coating is completely dry before placing the tissue. Wells can also be coated one day before and dried overnight at RT. Once coated, do not store the coated wells in the fridge or freezer, and for best results, use the coated 24-well plate within 1-2 days.
2. Cut the O.C.T-embedded tissue to 15-30 µm thickness at the cryostat and transfer it quickly into a coated well containing 2 mL of PBS.
   NOTE: The tissue can be transferred using a metal spatula or forceps.
3. Carefully remove the PBS using a 1 mL pipette, aiming to keep the tissue section centered in the well. To do this effectively, slowly aspirate the PBS from the edges of the well, adjusting the pipette position as needed to prevent the tissue from shifting.
   NOTE: Precise placement of the tissue is essential for the success of the protocol. Once the tissue adheres to the coated surface, it cannot be repositioned.
4. Dry the tissue sections for 1 h in a 37 °C oven.

3. Tissue blocking, EdU Click-iT reaction, and antibody immunolabeling, IBEX cycle 1 (Day 1)

1. Rehydrate with 1 mL of PBS for 5 min.
2. Permeabilize the tissue with 500 µL of 0.5% Triton in PBS for 30 min at RT.
3. Wash the tissue briefly 2x with PBS.
4. Perform Click-iT reaction, according to the manufacturer's instructions, for 30 min at RT.
   NOTE: Keep the samples protected from light.
5. Wash the tissue briefly 2x with PBS.
6. Block the tissue for 1 h at room temperature (RT) using a blocking buffer.
   NOTE: Here, 10% donkey serum in PBS was used for this experiment, but this can be adjusted as needed; any blocking buffer compatible with standard immunolabeling protocols is suitable.
7. Apply primary antibodies (diluted in blocking buffer; see the Table of Materials) for the 1st cycle overnight at 4 °C.
   ​NOTE: Blocking and immunolabeling can also be performed for ~1 h using a non-heating microwave, as described in Radtke et al.¹². However, for skin wounds, the best signal-to-noise ratio is obtained when primary antibodies are incubated overnight. Immunolabeled tissue sections can be stored in PBS for several days prior to imaging. Imaging is recommended the following day for optimal signal quality.

4. Secondary antibody immunolabeling and nuclear counterstain (Day 2)

1. Wash the wells with 3 x 1 mL of PBS.
2. Incubate the tissue section with the anti-rabbit secondary antibody staining (1:300) in blocking solution, together with nuclear counterstain Hoechst 1:1,000, for 1 h at RT in the dark.
3. Wash 3 x 5 min with 1 mL of PBS at RT.
4. Leave ~500 µL of PBS in each well and bring the plate to the microscope.
   ​NOTE: This can be a pause point: The samples are stable for several days at 4 °C, but the signal-to-noise ratio is the best when imaged within 24 h of immunolabeling.

5. Microscope setup and imaging of first IBEX cycle (Day 2)

1. Log in to the microscope and load the software, and sign in to the user profile.
2. Enter the acquisition setup (open via Screening | Acquisition Setup) for protocol configuration and image acquisition.
3. Use the Eject button to load the plate into the microscope.
4. Clean the bottom of the 24-well plate with 70% ethanol just before placing the plate in the microscope. Click the Load Plate button to load the plate into the microscope.
5. Select the desired magnification and acquisition mode under Objective and Camera. As the acquisition mode, select confocal 51 µm slit mode.
   NOTE: Here, a 20x water immersion objective with a numerical aperture of 0.95, and a spinning disk with 50 µm pinholes, spaced at 250 µm was used for the experiment.
6. Under the Plate tab, select the used plate model.
7. Go to the Plate/Sites to Visit tab. Under the Site options, click Fixed number of sites. Choose the number of Columns and Rows such that roughly the entire well is covered. Click Overlap sites 10%.
   NOTE: This button needs to be activated to ensure stitching of the individual tiles.
8. Select only the first filled well for acquisition by right-clicking the corresponding well on the well map. Selected wells are green and deselected wells are grey.
9. Move the stage to a position where tissue should be by left-clicking on the first filled well in the plate-map and on a site near the center in the sitemap. Look for a darker color of the well and site indicating the current position.
10. Go to the Acquisition tab. Select the options Enable laser-based focusing and Acquire Z Series.
11. Go to the Acquisition/Autofocus tab. For Well to well autofocus, select the Focus on plate and well bottom option, and for Site Autofocus, select the All sites option.
12. Go to the Wavelengths tab and choose the number of dyes to be imaged in the current cycle. Look for the corresponding number of new tabs that open below the Wavelengths tab.
13. Go to the first wavelength tab. Under Illumination, select the filter settings best suited for the fluorophore of interest (e.g., choose DAPI if Hoechst was used).
14. Choose Laser with z-offset for the z-positioning and click the Calculate Offset button to set the correct imaging height. A series of images will be acquired at different z-positions. Choose the one where the sample is best in focus.
    NOTE: Ensure that the stage position is set to a location with tissue. If no tissue is visible at this position, select a different site and try again (see step 4.9).
15. Press the Focus button and wait for an in-focus image of the tissue to be shown.
16. For Illumination Power, choose 50%. Enter 2000 for Target max intensity and press Auto Expose to adjust the exposure time automatically.
    NOTE: To speed up acquisition, the laser power can be increased, which should reduce the exposure time. However, this might lead to more photo-bleaching, which will lead to image artefacts where the individual tiles overlap.
17. For Z Series, choose Z-Series and 2D Projection Image, and for Shading Correction, choose FL Shading only.
18. Repeat steps 5.13-5.17 for the other wavelengths, but instead of Laser with z-offset, choose Z-offset from W1 and choose 0.
19. Go to the Z Series tab. Enter the desired Step Size (for the 20x objective, use a step size of 0.5 µm).
    NOTE: Acquisition speed can be increased by using a step size that is 2-3x the recommended value.
20. Click the Focus button and wait for an arrow labeled current position to appear.
21. Click on Start Live and wait for a new window with live images to appear. Drag the current position arrow until the bottom of the tissue is reached and click the Set B button. Drag the current position arrow until the top of the tissue is reached and click the Set T button. Click on F2: Stop.
    NOTE: One can choose the wavelength used for live imaging under Active Wavelength.
22. Select the sites to acquire; first, ensure that the stage is at the first well by left-clicking on it and press Start Live again. Go to different positions inside the well by left-clicking on different sites in the site map. Find the border of the tissue and mark all sites containing tissue for acquisition by right-clicking. Selected sites will be green and deselected sites gray.
    NOTE: For now, do this only for the first well and make sure only the first well is selected for acquisition.
23. Go to the Run tab. For the Plate Name, name it "Cycle{i}_Well{w}_{description}"; replace {i}, {w} and {description} with the corresponding values/description.
24. Click on Save Protocol and save it, again including "Cycle{i}_Well{w}" in the name.
25. Set up the acquisition regions for all the other wells. For this, repeat steps 5.22-5.24.
    NOTE: For each well to acquire, an individual protocol file needs to be saved. Ensure that the correct well is activated prior to drawing the new ROI. The result should be one saved protocol per well to acquire. This is necessary since the tissues have different shapes from well to well. In case the ROI and z-stack are identical among the wells, repeating steps 5.22-5.24 can be skipped, and steps 5.26-5.29 can also be skipped. In that case, simply select all wells to be imaged (see step 5.22), go to the Run tab and press Acquire Plate.
26. Click Control | Journal | Start recording and then immediately Control | Journal | Stop recording. Save the created Journal and open it again via Control | Journal | Edit Journal.
27. In the left panel, find and double-click on Plate Acquisition - Load Protocol From File and on Plate Acquisition to add those steps to the journal; look for them in the right panel. Double-click on the added Plate Acquisition - Load Protocol From File and choose the protocol saved for the first well (see step 5.24).
28. Repeat step 5.27 for each well to be acquired, always loading the respective protocol.
29. Click on Save and finally on Run Journal to start the acquisition (requires approximately 3 h to acquire 180 tiles with a z-stack of 15 µm).

6. Fluorophore bleaching (Day 2)

1. Approximately 30 min prior to the end of the journal, prepare the bleaching reagent. Dissolve 10 mg of Lithium Borohydride in 10 mL of ddH₂O and pass through a 0.22 µm filter. Leave for 10 min, waiting for bubbles to form. To ensure efficient bleaching, use the bleaching reagent within 4 h after preparation.
   CAUTION: Work under a chemical hood when handling the Lithium Borohydride, as it can react with air and may produce hydrogen gas which is flammable.
2. Bleach the fluorophores with 1 mL of 1 mg/mL of Lithium Borohydride for 15 min whilst exposed to standard ambient lighting. Look for small bubbles on the tissue during the bleaching procedure.
   NOTE: When working under a chemical hood, ensure that the lighting is not automatically switched off during the bleaching procedure.
3. Repeat bleaching with another 1 mL of 1 mg/mL of Lithium Borohydride (prepared from the same solution) for 15 min.
4. Wash the well for 3 x 5 min with 1 mL of PBS.
   NOTE: Do not shorten washing times, as incomplete removal of the bleaching reagent could damage the next panel of antibodies.

7. IBEX cycles (Day 2)

1. Apply primary antibodies for the next immunolabeling cycle (step 3.7) and incubate at 4 °C overnight. Repeat steps 4.4-5.4.
   NOTE: Steps 4.2-4.3 are not repeated as an uncoupled rabbit antibody is only used in the first IBEX cycle. However, uncoupled antibodies raised in less common species such as chicken or guinea pig can also be used in later cycles, as their secondary antibody will not cross-react with the coupled primary antibodies. Exact details of the antibodies used in each cycle can be found in the Table of Materials.
2. On the microscope, load the protocol settings from the previous IBEX cycle for the corresponding well by clicking on load protocol, saved on the local PC as Cycle{i}_Well{w} (step 5.24). The protocol settings will have the ROI (selected tiles to acquire), magnification, autofocus settings and any additional acquisition settings saved from the previous IBEX cycle (steps 5.5-5.25).
3. Select the wavelength tabs and adjust exposure and illumination power for the new panel of antibodies (steps 5.12-5.18). Be sure to image one marker (usually DAPI or Hoechst) throughout all cycles to align the cycles afterwards.
   NOTE: As reported in Radtke et al.¹⁸, commonly used fluorescent proteins, such as GFP and tdTomato, are not sensitive to lithium borohydride bleaching. When working with tissue that contains endogenous fluorescent proteins such as tdTomato, the corresponding wavelength can be omitted from acquisition after cycle 1, if it is not needed for image registration-this will speed up imaging. Wavelengths can be removed in the Overall Wavelength tab by reducing the number of acquisition channels. Moreover, a consistent increase in tissue autofluorescence was not observed over the imaging cycles; however, this is likely to vary depending on the tissue type¹⁸.
4. Check if the z-Stack settings are still reasonable and adjust accordingly (steps 5.19-5.21).
   NOTE: Ensure that the step size of the z-stack is the same for all cycles.
5. Update the Plate Name in the Run tab and re-save the Protocol.
6. Repeat steps 7.2-7.5 for all the wells to acquire.
   NOTE: It is recommended to write down the Exposure setting for all wavelengths used in the first well and use the same setting in all subsequent wells.
7. Edit the journal, select the new protocol files, and save the journal for this cycle.
8. Run the journal.
9. Repeat steps 6.1-6.4 to bleach the fluorophores.
10. Repeat steps 7.1-7.9 until the desired amount of immunolabeling has been achieved.

8. Image Processing (Day 2)

1. After finalizing image acquisition of the IBEX cycles, export image data:
   1. Select Screening | Plate data utilities | Export Images.
   2. Click Select Plate(s) and choose all the acquired plates for each cycle and well.
   3. Choose All Z Planes, select an output directory, and press OK to export the data.
      NOTE: The data will be exported as individual tiff files for each tile and plane. To combine all tiles and cycles, the following processing steps were performed.
2. Arrange the exported data in folders according to the following scheme: ensure that each cycle is one folder, each containing again one folder for each well. Name as follows: "\cycle{i}\{well}\*_Plate_*", where {i} and {well} are the corresponding cycle and well, and * is any character. For example: "\cycle1\A01\some_name_Plate_8654"
   NOTE: A correct naming and folder structure is required for the subsequent processing steps.
3. Install necessary python packages: use Miniforge for managing the Python environment. See https://github.com/conda-forge/miniforge for install instructions.
   NOTE: The code for the processing steps is provided in the form of Jupyter Notebooks. They can be found at https://github.com/ZMB-UZH/zmb-ibex-jove or as supplemental files (Supplemental File 1, Supplemental File 2, Supplemental File 3, Supplemental File 4, and Supplemental File 5).
4. When downloading the code from Github, download all files by clicking Code | Download ZIP and follow the installation instructions in the README.md file.
5. Start Jupyter-Lab (Supplemental File 2):
   1. Activate the installed conda environment by entering conda activate zmb-ibex-jove in a terminal.
   2. In the terminal, change to the downloaded directory by entering, for example, cd C:\path\to\zmb-ibex-jove.
   3. Start Jupyter-Lab by entering jupyter-lab in the terminal. A browser window will open with jupyter-lab.
6. Calculate illumination correction:
   1. On the left tab, navigate to the " examples/01_get_flatfield_BaSiC.ipynb" notebook and open it by double-clicking it. This notebook will calculate the illumination correction images, using the BaSiCPy library (Supplemental File 3).
   2. Change the "base_dir" and "save_dir" to the correct paths, where the data is located, and where the illumination correction images should be saved.
   3. To run the code within the notebook, select each cell and press Shift+Enter to execute it. Run the notebook from top to bottom.
7. Perform stitching for each individual cycle (Supplemental File 4):
   1. On the left tab, navigate to the next notebook and open it: "02_stitching_MD.ipynb". In this notebook, for each cycle, the individual tiles are loaded, illumination corrected, and stitched together, resulting in one image file per well and cycle. Use the m2stitch library (https://github.com/yfukai/m2stitch) for stitching.
   2. Change the inputs in the first cell to the correct file paths. Execute the notebook again from top to bottom.
8. Perform image registration between cycles (Supplemental File 5):
   1. On the left tab, navigate to the next notebook and open it: "03_registration_pcc.ipynb". In this notebook, the individual cycles are aligned with each other using only translational transforms, which are calculated with the phase-cross-correlation algorithm from the open-access scikit-image library (https://scikit-image.org/). 
   2. Change the inputs in the first cell to the correct file paths. Choose a reference cycle and reference channel to use (usually DAPI). Execute the notebook again from top to bottom.
      NOTE: Only one of the chapters "Option1: Load and save all channels together" and "Option 2: Load and save channels as individual files" needs to be executed. The first chapter will create one large file, containing all channels of all cycles. The second chapter will create an individual file for each channel. The second option will require less memory to process.

9. Image Analysis (Day 2)

1. Download and open Fiji (https://imagej.net/software/fiji/downloads).
   NOTE: The resulting images will be multi-plane and multi-channel images, which could be analyzed as 3D images. Here, an example analysis workflow is shown for 2D data only, for which a maximum intensity projection is first calculated.
2. Open the file in Fiji by dragging the file onto the Fiji window.
3. Calculate the maximum intensity projection by selecting Image | Stacks | Z Project..., choose Max Intensity, and press ok. Resave the image via File | Save as | OME-TIFF".
4. To analyze the cells, use the open-source software QuPath; download and install it from https://qupath.github.io/.
5. Open QuPath and create a new project by dragging an empty folder onto the window.
6. Import all files into the project by dragging them onto the QuPath window and pressing Import with the standard settings.
7. Double-click on an image on the left panel to open it. Specify it as Fluorescence when prompted.
8. Open the Brightness/Contrast window by pressing Shift+C.
   NOTE: Here, one can select which channels are shown, and adjust the brightness and contrast. One can also change the channel names and colors by double-clicking on the names.
9. Select the Tools | Polygon tool and annotate the tissue-region. Look for a new annotation under the Annotations tab. Set it to a specific class by selecting the annotation, selecting a class (e.g., Other), and pressing Set class.
10. Detect Cells: Go to Analyze | Cell detection | Cell detection.
    1. Choose a nuclear channel.
    2. Leave most settings as they are, but adjust the Threshold by hovering over a nucleus with the mouse and noting the intensity displayed in the bottom right corner. Hover over a background region and note the intensity again. Choose a value halfway between the background and foreground as the threshold.
    3. Click Run.
11. Inspect cells and measurements:
12. By double-clicking on a cell or by selecting it in the Hierarchy tab, inspect its measurements in the bottom left panel. Different shape and intensity measurements should be added for cell, nucleus, and cytoplasm.
13. Classify cells:
    1. To classify the detected cells according to the intensity in a channel, go to Classify | Object classification | Create single measurement classifier.
    2. Select the desired channel under Channel filter and an appropriate measurement under Measurement (e.g., Cell: Channel mean).
    3. Adjust the threshold by again hovering over a positive cell and over a negative cell and choosing a value between the two.
    4. Select, for example, Positive for Above Threshold and Negative for Below Threshold. Check the result by checking Live preview.
    5. Press Apply to classify the cells.
14. Measure distance to annotated objects (e.g., nerve bundles):
    1. Select the Tools | Brush tool.
    2. Draw over regions of interest (e.g., nerve bundles).
    3. Go to the Annotations tab. Select all new annotations, select a corresponding class, and press Set class.
       NOTE: One can create a new class by right-clicking in the class window and selecting Add/Remove | Add class.
    4. Go to Analyze Spatial Analysis | Signed distance to annotations 2D and press Yes when prompted.
    5. Add a new measurement to each cell, indicating the distance to the closest annotation of a given class.
15. Export measurements:
    1. Go to Measure | Show detection measurements and wait for a list of all measurements of all detections to appear.
    2. Export the table as a .txt via the Save button.
    3. Import the data into data processing software of choice to be further analyzed.

Results

This method allows adequate detection of proliferating cell types by combining EdU Click-iT chemistry with IBEX in fixed frozen murine skin sections. Detection of proliferating cell types was compared using different methodologies, in particular, labeling of proliferating cell types with Ki-67 antibody and detection of cell proliferation using the EdU Click-iT chemistry approach (Figure 1). We observed that Ki-67 antibody labeling underestimates the number of...

Discussion

Cell proliferation is one of the four main stages of skin repair, and the transition between the inflammatory phase and the proliferative phase is critical for a successful repair process²⁰. A well-regulated immune response is essential, as prolonged inflammation can lead to chronic wounds, a significant global health issue^1,21. Proliferation is a key biological process, not only important for tissue repair but also for organism developmen...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Antibodies and reagents	SOURCE	IDENTIFIER	Comments
Antibodies
CD31 Alexa Fluor 647 (Cycle 2)	BioLegend	Cat#102516	1:100
AB_2161029
CD45 Alexa Fluor 488 (Cycle 2)	BioLegend	Cat#103122	1:100
AB_493531
Ki-67 FITC (Cycle 1)	BioLegend	Cat#151212	1:50
AB_2814055
Ki-67 chicken (ck)	LSBio	Cat#LS-C772946	1:800
Ki-67 rat (rt)	BioLegend	Cat#652402	1:100
AB_11204254
Neurofilament H&M (NF-H/NF-M) Alexa Fluor 647 (Cycle 3)	BioLegend	Cat#837710	1:300
AB_2734613
p75NTR (Cycle 1)	Cell Signaling Technology	Cat#8238                       AB_10839265	1:1500
Vimentin Alexa Fluor 488 (Cycle 3)	BioLegend	Cat#699305	1:100
AB_2888889
Alexa Fluor 647 donkey anti rabbit	Jackson ImmunoResearch	Cat#711-605-152
Reagents
Cellvis 24 well Glass Bottom Plate	Cellvis	P24-1.5H-N
Chrome alum gelatin	Newcomer Supply	Cat#1033A
Click-iT EdU kit	Thermo Fisher Scientific	C10637
Hoechst 33342	Sigma-Aldrich	Cat#14533; CAS# 23491-52-3
LiBH4, Lithium Borohydride, 95%	Thermo Fisher Scientific	Cat#10438443
Tissue-Tek O.C.T Compound Sakura	Biosystems Switzerland	Cat#7109-4583
PBS	Thermo Fisher Scientific	Cat#10010-015
Superfrost Plus Adhesion Microscope Slides	Epredia	Cat#J1800AMNZ
Triton X-100	Sigma-Aldrich	Cat#T8787
Equipment
Hair brush	Faust AG	Cat#9172051 and 9172050
Microtome blades	Biosystems Switzerland	Cat#207500011
Microscope	Molecular Devices, ImageXpress Confocal HT.ai
Oven	Thermo Fisher Scientific	Cat#HBMCR4220
Rocker shaker	Edmund Bühler GmbH	Cat#32310015