Flow cytometry analysis of the subpopulations of mouse keratinocytes and skin immune cells

Summary Skin is our body’s outermost physical barrier and an immunological interface enriched with various immune and non-immune cells. However, efficient generation of single-cell suspensions for flow cytometry analysis can be challenging. Here, we provide protocols to obtain epidermal and whole skin cell suspensions as well as gating strategies to identify mouse keratinocytes and skin immune cell subsets via flow cytometry. For complete details on the use and execution of this protocol, please refer to Sakamoto et al. (2021).


SUMMARY
Skin is our body's outermost physical barrier and an immunological interface enriched with various immune and non-immune cells. However, efficient generation of single-cell suspensions for flow cytometry analysis can be challenging. Here, we provide protocols to obtain epidermal and whole skin cell suspensions as well as gating strategies to identify mouse keratinocytes and skin immune cell subsets via flow cytometry. For complete details on the use and execution of this protocol, please refer to Sakamoto et al. (2021).

BEFORE YOU BEGIN
Mice were bred and/or maintained in the specific pathogen-free facility in accordance with the Guide for the Care and Use of Laboratory Animals. All experiments were performed at National Institute of Arthritis and Musculoskeletal Skin Diseases under an animal study proposal approved by the NIAMS Animal Care and Use Committee.

Mice
Before working with this protocol, secure sufficient numbers of age matched female C57BL/6 mice for each experiment. We generally use only female mice because adult male mouse skin is relatively thick and cannot be efficiently digested (Azzi et al., 2005). Male mice also tend to have traumatic skin wounds that lead to increased inflammatory cell infiltrates (Kashem and Kaplan, 2018). One might start harvesting 3 to 5 mice at a time and then increase the numbers of mice per harvest after becoming competent in all experimental procedures. Use exactly age-matched mice for control and experimental groups because hair follicle (HF) and immune cell numbers can be affected by the age of mice and the hair cycle, which undergoes bouts of hair growth and regression in an age-dependent manner (Schneider et al., 2009).

Antibody panel preparation
Timing: 0.5-1 h 1. Prepare the flow cytometry antibody panel for keratinocyte, innate lymphoid cell (ILC), and myeloid cell (Tables 1, 2, 3, and 4). Make sure that you have enough antibodies for the total number of samples you want to analyze, which should be confirmed on the prior day. Make an antibody master mix immediately before the staining step. Use a 5% FACS buffer for dilution. Epidermis digestion solution Mix 5 mL of 0.25% Trypsin-1 mM EDTA and 5 mL of 0.05% Trypsin-0.5 mM EDTA to make 0.15% trypsin and 0.75 mM EDTA. This solution should be prepared on the day of sample processing and put on ice until use. Use 10 mL for each skin sample.

Whole skin digestion solution
Dilute Liberase stock to a final concentration of 0.25 mg/mL and DNase stock to a final concentration of 1 mg/mL in RPMI (not supplemented with antibiotics, amino acids or FBS). This solution should be prepared on the day of sample processing and put on ice until use. Use 5 mL for each skin sample.
13 Foxp3 fixation/permeabilization working solution Mix 1 part of Foxp3 Fixation/Permeabilization Concentrate with 3 parts of Foxp3 Fixation/Permeabilization Diluent. For example, mix 1 mL Foxp3 Fixation/Permeabilization Concentrate with 3 mL Foxp3 Fixation/Permeabilization Diluent. This solution should be prepared on the day of sample processing and stored at room temperature until use. Both concentrate and diluent are included in the Bioscienceä Foxp3/Transcription Factor Staining Buffer Set.

permeabilization buffer working solution
Mix 1 part of 103 Permeabilization Buffer with 9 parts of distilled water. For example, mix 5 mL 103 Permeabilization Buffer with 45 mL distilled water. This solution should be prepared on the day of sample processing and stored at room temperature until use. Permeabilization buffer is included in the Bioscienceä Foxp3/Transcription Factor Staining Buffer Set. To detect transcription factors (GATA3 and RORgt), staining is performed after cell surface staining (Table 5).

STEP-BY-STEP METHOD DETAILS Skin collection
Timing: 10 min/mouse 1. Euthanize mice (up to 5 mice) by CO 2 inhalation. 2. Collect skin samples from the chest area immediately after euthanization. a. Cut off an approximately 3 cm 3 4 cm area from mouse chest skin using ophthalmic scissors. The chest area does not require prior shaving. b. Float each skin sample in a petri dish containing 10 mL of PBS on ice while other samples are being processed ( Figure 1A). c. If more than 5 mice need to be harvested, repeat step 1-2b. d. Transfer skin samples (one sample at a time) surface-side down onto the lid of petri dishes and scrape off subcutaneous tissue using two forceps ( Figure 1B, Method video S1). e. Transfer back the skin into each petri dish and float on PBS.

Generating cell suspension Generating epidermal cell suspensions
Timing: 1.5 h 3. Float skin surface-side up on 10 mL of epidermis digestion solution (0.15% trypsin and 0.75 mM EDTA) in 100 mm 3 15 mm petri dish. Remove any air bubbles under the skin samples using two curved forceps. 4. Incubate for 45 min at 37 C in a cell culture incubator. CO 2 is not mandatory. 5. Briefly transfer the skin onto the lid of the petri dish and discard epidermis digestion solution by decanting. Put the skin back onto the petri dish and pour 20 mL of 4 C 5% FACS buffer into each petri dish. 6. Scrape off the epidermis from the dermis gently with two curved forceps. If enzymatic digestion is effective, the epidermal component should detach without exerting excessive force. Discard dermis ( Figure 1C, Method video S2). 7. The epidermal cells are further mechanically dissociated with a 50 mL syringe (Covidien). Pump 8 to 10 times without applying too much pressure. Applying excessive force may affect cell viability ( Figure 1D). 8. Filter the cell suspensions through sterile 100 mm Falconâ Cell Strainers (Corning) placed on 50 mL conical tubes. Centrifuge at 400 3 g for 5 min at 4 C and remove the supernatant. 9. Break the cell pellets and resuspend in 10 mL of 5% FACS buffer. 10. Filter the solution through sterile 40 mm Falconâ Cell Strainers (Corning) placed on new 50 mL conical tubes. 11. Centrifuge at 400 3 g for 5 min at 4 C. Remove the supernatant and resuspend with 300 mL of 5% FACS buffer ( Figure 1E).

Generating whole skin cell suspension
Timing: 3 h 12. Add 5 mL of whole skin digestion solution into each well of 6-well plates. Keep on ice. 13. Place the skin sample on Kim wipe very briefly to remove excess PBS. 14. Transfer each skin sample into 35 mm petri dishes and mince well using ophthalmic scissors. 15. Transfer 1 mL of whole skin digestion solution from the 6-well plate and add it to the minced tissue. Further mince until paste-like textures are obtained ( Figure 1F). Mincing process should be done on ice. 16. Pour the minced skin into 6-well plates with a whole skin digestion solution. 17. Incubate for 2 h at 37 C in a cell culture incubator. CO 2 is not mandatory. 18. Add 1 mL of 0.25% Trypsin-1 mM EDTA into each well for the last 10 min of incubation (Figure 1G). 19. Deactivate the enzymatic reaction by adding 4 mL of 4 C 5% FACS buffer into each well. 20. Mechanically dissociate with a 12 mL syringe. Pump 8 to 10 times. Excessive force should not be applied. 21. Filter the cell suspension through sterile 100 mm Falconâ Cell Strainers (Corning) placed on 50 mL conical tubes. Centrifuge at 400 3 g for 8 min at 4 C and remove the supernatant (Figure 1H). In this final centrifugation step for whole skin, we recommend 8 min instead of 5 min to avoid cell loss since pellets may be loose compared to those obtained from epidermal cell suspensions. 22. Break the cell pellets and resuspend in 10 mL of 5% FACS buffer. 23. Filter again with 40 mm cell strainers placed on new 50 mL conical tubes. 24. Centrifuge at 400 3 g for 8 min at 4 C. Remove supernatant and resuspend with 300 mL of 5% FACS buffer.  28. Centrifuge at 400 3 g for 3 min at 4 C and discard the supernatant. 29. Dilute anti-mouse CD16/32 antibody (Fc block) at a dilution of 1:200 with 5% FACS buffer. 30. Resuspend the cells with 100 mL Fc block solution and incubate the cells on ice for 5 min. 31. During incubation, make a master mix solution of indicated antibodies (Table 1) in Eppendorf tubes or 15 mL conical tubes, depending on the total volume required. 32. After incubation with Fc block, centrifuge at 400 3 g for 3 min at 4 C and remove the supernatant. 33. Add 100 mL of antibody master mix and incubate the cells on ice, in the dark, for 25 min.
CRITICAL: For the ILC panel, all antibodies except for two antibodies (GATA3 and RORgt) should be added at this step. Anti-GATA3 and -RORgt antibodies will be used in step 39. d. Centrifuge at 400 3 g for 3 min at 4 C and discard the supernatant. e. During the centrifuge step, prepare 13 Permeabilization Buffer working solution with 2% normal rat serum. Prepare 100 mL/sample. f. Resuspend cells with 100 mL of 13 Permeabilization Buffer working solution with 2% normal rat serum and incubate the cells at room temperature, in the dark, for 15 min. g. Prepare the antibody master mix for transcription factors GATA3 and RORgt in 13 Permeabilization Buffer working solution with 2% normal rat serum to achieve indicated final dilutions (Table 5). h. Add antibody master mix to each sample and incubate the cells at 4 C, in the dark, overnight (approximately 12 h). i. Centrifuge at 400 3 g for 3 min at 4 C and discard the supernatant. j. Wash the cells with 200 mL 13 Permeabilization Buffer working solution. k. Centrifuge at 400 3 g for 3 min at 4 C and discard the supernatant. l. Repeat steps (j) and (k). m. Resuspend the cells with 200 mL 13 Permeabilization Buffer working solution. Keep cells on ice until data collection.

Data collection
Timing: 2-3 h 40. Collect data with LSR ll or LSR Fortessa (BD Biosciences) and analyze by using FlowJo software (FlowJo, LLC). Before acquiring samples, set appropriate PMT voltage and compensation by using compensation beads (Cat# 552845, BD) and a zombie aqua single-stained sample. Note that the compensation beads utilized here are specific for rat and hamster IgGs. If the antibody panels are modified to include antibodies generated in other host species, consider using other compensation bead products. Adjust compensation using the same set of antibodies from each panel.

EXPECTED OUTCOMES
Keratinocyte panel: Doublets are gated out first by FSC-H versus FSC-W and then by SSC-H versus SSC-W gates. Dead cells are subsequently excluded with Zombie Aqua staining. Because keratinocytes continuously turn over, it is normal to see up to 50% of epidermal cells to be positive for Zombie Aqua. Then, a broad gate for SSC-A versus FSC-A is made because the keratinocyte profile in this plot is broad. SSC-A low FSC-A low events represent debris and are excluded. CD45cells are gated to include all epidermal keratinocytes and to exclude immune cells. Further plotting for Ep-CAM and CD200 enables the distinction of keratinocytes from the interfollicular epidermis or HFs. CD200 + cells contain all HF subsets. The vast majority of CD200cells are Sca-1 + keratinocytes from the interfollicular epidermis. Among CD200 + HF cells, the CD34 + population represents the bulge cells (the stem cell area), and CD34cells represent the upper HFs, which are further divided into EpCAM + Sca1 + infundibulum (the HF opening) and EpCAM + Sca1isthmus (narrowing portion of the HFs below the infundibulum and above the bulge) (Sakamoto et al., 2021). This staining strategy for keratinocyte subsets is useful for analysis in non-inflamed skin (Figure 2A). In inflamed skin, the epidermal cell suspensions may be challenging to prepare. Thus, the same antibody panel may be applied to whole skin suspensions. In this case, while CD34 + bulge population is discernible, its separation is not as good as those observed in epidermal cell suspensions. Additionally, because CD31 + endothelial cells express CD200 (Ko et al., 2009), they need to be excluded by including a CD31 antibody in the antibody panel ( Figure 2B).
(B) The gating strategy of interfollicular (EpCAM + CD200 -Sca-1 + ) and HF keratinocytes (EpCAM + CD200 + ) in whole skin cell suspensions. The gating strategy is the same as in epidermal cell suspensions except for the exclusion of CD31 + vascular endothelial cells after gating for CD45cells.

QUANTIFICATION AND STATISTICAL ANALYSIS
FlowJo (FlowJo, LLC https://www.flowjo.com/solutions/flowjo) was used for data analysis. To avoid potential data variability among experiments or among different experimentalists, total collected  events were fixed among samples. The total number of cells that are acquired can be optimized depending on the abundance of the specific skin cell subsets (Table 6).

LIMITATIONS
Adult mice develop spontaneous anagen (growth phase of the hair cycle) patches after 12 weeks of age (Mü ller-Rö ver et al., 2001). While this is only rarely observed in the chest area, it is commonly observed in back skin. Epidermal cell suspensions may be challenging to prepare from such areas. Whole skin processing of anagen patches require proper mincing of the tissues. Trauma or dermatitis may also induce anagen patches. If unmanipulated mice show extensive anagen patches, researchers should assess the skin surface for the possibility of trauma or inflammation. Such mice may have to be excluded from analysis.
Antibody panels must be optimized for each flow cytometer. Signals can be prominently affected by the types of lasers, voltage settings, and the combination of filters that flow cytometers are equipped with. We recommend that the antibody panels are optimized with single-stained samples as well as ''fluorescence minus one'', or FMO, for all antibodies.
We have included NK1.1 in our lineage channel to exclude NK cells. However, exclusion of NK cells in non-C57BL/6 strains such as Balb/C mice, in which NK1.1 is not expressed (Carlyle et al., 2006), may require the use of other NK cell markers such as CD49b that is detected by the antibody DX5 (Arase et al., 2001).

Potential solution
Make sure that all utilized mice are of the same age and at most 1 week apart to ensure the hair cycle is comparable. We recommend experimental mice to be 8.5-12 weeks of age at the time of harvest to avoid anagen skin. If later ages need to be analyzed, and if hair cycle is not the focus of the experiment, consider excluding anagen patches when harvesting skin. Barbering and fighting can traumatize the skin and trigger anagen, potentially leading to incomparable skin conditions between experimental mice. If trauma is an issue, house mice in individual cages 1 or 2 weeks prior to harvest.

Problem 2
Epidermal cells do not detach well from the dermis after trypsin treatment (step 3).

Potential solution
Make sure that subcutaneous tissue is removed completely. Remaining subcutaneous tissue may interfere with enzymatic digestion. One may test effective digestion by scraping off the epidermis in one of the samples at the end of the 45 min incubation with epidermal digestion solution. If epidermal detachment is poor, further incubate samples for additional 5-10 min. Additional incubation for over 10 min is not recommended because longer incubation could potentially lead to the digestion of cell surface markers. One common mistake is including FBS in the epidermal digestion Because of the low cell numbers of ILCs, we recommend collecting 1 million events if analyzing the skin during steady state. If inflammatory models are analyzed, the number of events to be acquired should be determined for each model.

OPEN ACCESS
solution. Make sure digestion solutions do not contain FBS because it will interfere with enzymatic digestion. If none of the above improve tissue dissociation, make a new batch of epidermal dissociation solution. Epidermal cells will not detach well in inflamed skin. If an inflammatory skin model is being studied, consider using whole skin suspensions.

Problem 3
Cells are not retrieved from whole skin digestion (step 24).

Potential solution
Thorough mincing is critical in generating single cell suspensions from whole skin samples. If cells still cannot be retrieved in sufficient numbers, make sure the skin samples are minced well next time. One common mistake is including FBS in the whole skin digestion solution, which will interfere with enzymatic digestion. If none of the above improve tissue dissociation, try a different lot of Liberase T-Flex.

Problem 4
Up to 50% of epidermal and whole skin cell suspensions may be positive for the viability dye. However, in certain instances, a higher percentage of cells could be non-viable (step 40).

Potential solution
Non-immune cells, in particular keratinocytes, may continuously undergo cell death even if the cell suspensions are kept on ice. Thus, delays in cell processing may lead to increased percentages of dead cells. After the generation of epidermal and whole skin cell suspensions, samples should be immediately stained, and data must be collected in a timely manner. If a timely data collection is not possible, immediately fix the cells after antibody staining is done.

Problem 5
No positive staining of some markers by flow cytometry (step 40).

Potential solution
Prior to the acquisition, ensure proper flow cytometer settings. Acquire a small amount of fullstained samples and check if all targeted populations show up.
Ensure that antibodies have been correctly stored and are not expired.
Ensure that antibodies were added and used at the suitable concentration. If one is unsure if a particular antibody was added, go back to the staining step (33). However, cells will be lost during this extra washing step.
Pair the low expressing antigens with bright fluorochromes such as PE or APC next time.
If none of the above work, try another antibody clone.

RESOURCE AVAILABILITY
Lead contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Keisuke Nagao (keisuke.nagao@nih.gov).

Materials availability
No new materials were generated in this protocol.

Data and code availability
This study did not generate or analyze any datasets.