Epidermal resident memory T cell fitness requires antigen encounter in the skin
- Department of Dermatology, University of Pittsburgh, United States
- Department of Immunology, University of Pittsburgh, United States
- Institute for Microbial Diseases, Osaka University, Japan
- Institute for Open and Transdisciplinary Research Initiatives, Osaka University, Japan
- Department of Immunology, Harvard Medical School, United States
- Department of Pathology, Brigham and Women’s Hospital, Harvard Medical School, United States
- Center for Immunology, Department of Microbiology and Immunology, University of Minnesota, United States
- Department of Dermatology, Meyer Cancer Center, Program in Immunology and Microbial Pathogenesis, Weill Cornell Medicine, United States
- Center for Systems Immunology, University of Pittsburgh, United States
- Department of Microbiology and Immunology, The University of Melbourne, The Peter Doherty Institute for Infection and Immunity, Australia
Figures
Figure 1 with 1 supplement
Epidermal TRM are transcriptionally heterogeneous.
(A) Experimental design. Mice were treated with vaccinia virus infection by skin scarification on the left flank on day 0, and then on day 5 post-infection, the right flank was painted with 0.15% DNFB. On day 56, flanks were harvested for either epidermal whole mount, flow cytometry, or cells were sorted for scRNA-seq. (B) Representative images and (C) quantification of epidermal whole mounts of VV and DNFB treated flanks harvested on day 50 post infection stained for CD8a. (D) Representative flow plots gated on CD45+ CD3+ CD8+ CD90.2+ cells isolated from VV or DNFB treated flanks. (E) Representative flow plots and (F) quantification of B8R tetramer binding of CD103+ TRM gated as in (D) isolated from VV or DNFB treated flanks. (G) The integrated Minimal-Distorted Embedding of all 96 experiments from the ImmgenT consortium with annotated clusters. (H) Minimal-Distorted embedding visualization of transcriptional clusters of skin TRM projected over the ImmgenT dataset. (I) The percentage of cells in each transcriptional cluster found in VV or DNFB treated sites. Each symbol represents paired data from the same individual animal (C, F). Data shown to be nonsignificant by paired Student’s t-tests (C, F). Data are representative of 3 separate experiments. Scale bar in (B) represents 50 µm. Panel A was created with BioRender.com.
Figure 1—figure supplement 1
Flow cytometric and gene expression analysis of the TRM used in the Immgen T scRNA-seq dataset.
(A) Quantification of percent CD103+ T cells from VV and DNFB treated flanks, gated on live CD45.2+CD8+CD90.2+. Each bar represents the mean +/- SEM for n=6 animals. (B) Gating strategy for skin single-cell RNA-seq. (C) Flow plots of CD69 and CD103 expression of VV and DNFB treated flanks for single-cell RNA-seq gated as in (B). (D) Heatmap showing the top 10 DEG per cluster when compared with each other, calculated by FindAllMarkers of Seurat V2, percent expressed >0.02, Log2foldchange >2.
Figure 2
Cutaneous antigen is required for complete TRM differentiation.
(A) Signature score analysis of skin T cell clusters calculating the enrichment for previously published T cell state gene sets (N=naïve). Signature score = Mean (average upgenes z-scores) - Mean(average downgenes z-scores). *** P<0.001 by Dunnett’s multiple comparisons of cluster 3 to all other clusters. (B) Pseudobulk analysis comparing signature scores of cells isolated from VV and DNFB sites as in (A). ** P<0.01 and *** P<0.001 by paired Student’s t-tests. (C) Signature score analysis of individual clusters or (D) cells isolated from VV and DNFB sites calculating the enrichment for compared to epidermal TRM isolated at the indicated day post infection. ***P<0.001 by Dunnett’s multiple comparison test of cluster 3 to all other clusters in (C). * P<0.05, ** P<0.01, **** P<0.0001 by paired Student’s t-tests in (D) Gray line represents a randomized control generated by the average enrichment of each group compared to a randomly generated gene set of an equal number of probes. Datapoints and error bars represent mean and 95% confidence interval of the relative enrichment of each set of DEGs.
Figure 3 with 1 supplement
Local antigen experienced TRM have improved expansion in a recall response.
(A) Experimental design. Mice were adoptively transferred with Thy1.1+ OT-I T cells on day –1, then infected with OVA-expressing vaccinia virus by scarification on the left flank on day 0. On day 5 post-infection, the right flank was painted with 0.15% DNFB. On day 50, a primary recall response with SIINFEKL peptide in acetone and olive oil was painted on both flanks and harvested on day 56. In some cohorts, mice were allowed to rest for an additional 120 days and then treated with topical SIINFEKL peptide again on day 176 and harvested on day 182. (B) Representative images and (C) quantification of epidermal whole mounts isolated from VV-OVA and DNFB treated flanks stained for anti-Thy1.1 (n=10 animals). (D) Experimental design. Mice were treated as in (A), but for 6 days prior to SIINFEKL treatment, they were given either i.p. FTY720, i.p. PBS, i.v. titrated anti-Thy1.1 or i.v. isotype control. (E) Representative epidermal whole mount images and (F) quantification of skin from VV-OVA treated flanks on day 6 of a primary recall response treated with either FTY720 or PBS. (G) Quantification of total Thy1.1 OT-I cells in epidermal whole mounts on day 6 after a primary recall response in mice treated with either isotype or anti-Thy1.1 depleting antibody. Each symbol represents paired data from an individual same animal (F, G). Data are representative of 3 independent experiments. *P<0.05 by paired Student’s t-tests. Unpaired Student’s t-tests between FTY720 treated (F) or Anti-Thy1.1 treated (G) flanks and PBS treated flanks shows non-significance for both VV-OVA and DNFB. Scale bar represents 50 µm (B, F). Each symbol represents the mean +/- SEM (C). Panel A and D were created with BioRender.com.
Figure 3—figure supplement 1
FTY720 and anti-Thy1.1 deplete circulating OT-I cells withouth affecting the epidermis.
(A) Representative flow plots of blood from Thy1.1 OT-I adoptive transferred mice after 6 days of vehicle or FTY720 treatment, gated on live CD45+. (B) Quantification of total Thy1.1 OT-I in blood after FTY720 treatment, gated as in (A). Each bar represents the mean +/– SEM for n=6 animals. (C) Representative flow plots of blood of CD45+CD3+CD8+ cells after isotype or Thy1.1 depleting antibody treatment. (D) Quantification of Thy1.1 OT-I in blood before and after isotype or anti-Thy1.1 depleting antibody treatment, gated as in (C). (E) Representative epidermal whole mount of Thy1.1+ cells in the epidermis at steady state of VV-OVA or DNFB treated flanks after 6 days of isotype or anti-Thy1.1 depleting antibody treatment. Each symbol represents data from an individual animal. Data is representative of three separate experiments.*p<0.05 by unpaired Student’s t-tests. Scale bar represents 50 µm.
Figure 4
Local antigen experienced TRM have increased proliferation during a recall response.
(A) Experimental scheme. (B) Representative images of epidermal whole mounts of VV-OVA or DNFB treated flanks on day 2 of a recall response. Arrows highlight cell doublets. (C) Quantification showing percent OT-I cells that are dividing in epidermal whole mounts on day 2 after primary recall response, and (D) total number of dividing OT-I cells. (E) Representative flow cytometric plots and (F) quantification showing BrdU incorporation in gated CD45+ CD8+ CD90.1+ CD103+ CD69+ cells isolates from VV-OVA or DNFB treated flanks. (G) Quantification of total numbers BrdU+ OT-I cells combining OT-I numbers with percentage of BrdU incorporation in (F). (H) Representative histograms and (I) quantification of Annexin V expression in OT-I cells isolated from VV-OVA or DNFB treated flanks or OT-I cells heat-treated at 60°C for 1 hr (HK). Data are representative of three separate experiments. *P<0.05, **P<0.01 by Student's paired t-tests (C, D, F and G) or Dunnett’s test (I). Scale bar represents 50 µm. Panel A was created with BioRender.com.
Figure 5
TRM fitness depends on TCR signal strength.
(A) Experimental scheme. Local antigen experienced and bystander TRM were generated and restimulated as previously described, but on the DNFB-treated flanks, altered peptide ligand variants of SIINFEKL were topically applied to the skin 1 day after recruitment to the skin. (B) Representative images and (C) quantification of epidermal whole mounts (CD90.1 cyan) of VV-OVA and DNFB+ APL treated flanks, at steady state (no recall) or 6 days post-recall response. Each symbol represents data from an individual animal. Data are representative of 5 independent experiments. *p<0.05 by unpaired Student’s t-tests. Scale bar, 50 µm. Panel A created with BioRender.com.
Figure 6
Local antigen experienced TRM have improved persistence mediated by TGFßRIII.
(A) Representative flow cytometric plots of TGFßRIII staining of Thy1.1+ OT-I cells stimulated in vitro for 48 hr with PBS or anti-CD3 anti-CD28. (B) Representative histograms showing TGFßRIII expression in CD45+ CD3+ CD8+ CD90.1+ gated OT-I cells isolated from VV-OVA, DNFB or untreated flanks at least 50 days post infection. (C) Quantification of (B). (D) Schematic demonstrating genetics of Tgfbr3WT and Tgfbr3ΔCD8 mice. (E) Representative histogram and (F) quantification of TGFßRIII expression in huNGFR- or huNGFR+ Tgfbr3∆CD8 T cells harvested 5 days following i.p. treatment with tamoxifen then stimulated in vitro for 48 hr with anti-CD3 and anti-CD28. (G) Experimental scheme. (H) Representative histogram of CD45.2+CD3+CD8+CD90.1+OT I cells isolated from LNs after tamoxifen treatment demonstrating transformation efficiency. (I) Representative epidermal whole mounts showing Thy1.1 staining (green), huNGFR staining (red) or merge (yellow) of VV-OVA or DNFB treated flanks from mice adoptively transferred with either Tgfbr3WT or Tgfbr3∆CD8 cells, treated with tamoxifen i.p., and then given either PBS or i.p. CWHM12 for 10 days. Hair follicles in the sample present as long yellow streaks. (J) Quantification of total huNGFR+ Thy1.1+ OT-I in (I). Each symbol represents data from an individual animal. Black lines represent group means. Data is representative of 3 independent experiments. *P<0.05 by Dunnett’s test (C) or paired Student’s t-tests (F) and (J). Unpaired Student’s t-tests between Tgfbr3WT and Tgfbr3∆CD8 vehicle-treated groups show non-significance for both VV-OVA and DNFB treated flanks. Scale bar represents 50 µm. Panel D and G were created with BioRender.com.
Author response image 1
Pseudobulk analysis by average gene expression of Tgfbr3 in cells isolated from either VV or DNFB treated flanks, divided by the average gene expression of Tgfbr3 in naïve CD8 T cells from the same dataset.
Author response image 2
Representative images of epidermal whole mounts of VV treated flank skin, and an untreated site from the same mouse isolated on day 50 post infection and stained for CD8a.
Author response image 3
Quantification of total number of CD8+ T cells in left and right inguinal lymph nodes.
Each symbol represents paired data from the same individual animal, and this is representative of 3 separate experiments.
Author response image 4
Representative flow cytometric plots showing CD69 and CD103 expression in gated live CD45+CD8+CD90.1+ cells isolates from VV-OVA or DNFB treated flanks.
Tables
Appendix 1—key resources table
| Reagent type (species) or resource | Designation | Source or reference | Identifiers | Additional information |
|---|---|---|---|---|
| Antibody | Brilliant Violet 605 anti-mouse CD8a (Rat monoclonal, 53–6.7) | BioLegend | Cat# 100744, RRID:AB_2562609 | 1:200 |
| Antibody | Alexa Fluor 647 anti-mouse CD90.1 (mouse monoclonal, OX-7) | BioLegend | Cat# 202508, RRID:AB_492884 | 1:200 |
| Antibody | Alexa Fluor 647 anti-human CD271 (NGFR) (mouse monoclonal, ME20.4) | BioLegend | Cat# 345114, RRID:AB_2572059 | 1:200 |
| Antibody | Anti-mouse Thy1.1 (mouse monoclonal, HIS51) | Thermofisher | Cat# 14-0900-85, RRID:AB_467374 | 1:100 |
| Antibody | PerCP/Cyanine5.5 anti-mouse/human CD44 (rat monoclonal, IM7) | BioLegend | Cat# 103032, RRID:AB_2076204 | 1:200 |
| Antibody | Alexa Fluor 700 anti-mouse CD3 (rat monoclonal, 17A2) | BioLegend | Cat# 100216, RRID:AB_493697 | 1:200 |
| Antibody | FITC anti-mouse CD90.2 (rat monoclonal, 30-H12) | BioLegend | Cat# 105305, RRID:AB_313176 | 1:200 |
| Antibody | BUV737 anti-mouse CD69 (Armenian hamster monoclonal, H1.2F3) | BD Biosciences | Cat# 612793, RRID:AB_2870120 | 1:200 |
| Antibody | Brilliant Violet 510 anti-mouse CD103 (Armenian hamster monoclonal, 2E7) | BioLegend | Cat# 121423, RRID:AB_2562713 | 1:200 |
| Antibody | FITC anti-BrdU (mouse monoclonal, 3D4) | BioLegend | Cat# 364103, RRID:AB_2564480 | 1:200 |
| Antibody | APC anti-mouse CD45.2 (mouse monoclonal, 104) | BioLegend | Cat# 109841, RRID:AB_2563485 | 1:200 |
| Antibody | FITC anti-mouse CD45.2 (mouse monoclonal, 104) | BioLegend | Cat# 109805, RRID:AB_313442 | 1:200 |
| Antibody | Alexa Fluor 700 anti-mouse I-A/I-E (rat monoclonal, M5/144.15.2) | BioLegend | Cat# 107621, RRID:AB_493726 | 1:200 |
| Antibody | Anti-TGFßRIII (1C5H11) | Novus | Cat# NBP2-37418, RRID:AB_3296314 | 1:200 |
| Chemical compound, drug | DNFB | Sigma-Aldrich | D1529 | 1-fluoro-2,4-dinitrobenzene |
| Chemical compound, drug | Olive oil | Sigma-Aldrich | O1514 | |
| Chemical compound, drug | DMSO | Sigma-Aldrich | D2650 | Dimethyl sulfoxide |
| Chemical compound, drug | CWHM12 | Indalo therapeutics | N/A | |
| Chemical compound, drug | Tamoxifen | Sigma-Aldrich | T5648 | |
| Chemical compound, drug | Corn oil | Sigma-Aldrich | C8267 | |
| Chemical compound, drug | BrdU | Sigma-Aldrich | B5002 | |
| Chemical compound, drug | ProLong Gold Antifade Mount with DNA stains DAPI | Thermofisher | P36931 | |
| Chemical compound, drug | FTY720 | Fisher/Cayan Chemical | Cat# 10006292 | |
| Chemical compound, drug | Collagenase XI | Sigma-Aldrich | Cat# 9001-12-1 | |
| Chemical compound, drug | DNase | Sigma-Aldrich | Cat# 04536282001 | |
| Chemical compound, drug | 4-(2-hydroxyethyl)–1-piperazineethanesulfonic acid | Sigma-Aldrich | Cat# 7365-45-9 | |
| Chemical compound, drug | Percoll | VWR | Cat# 89428–524 | |
| Chemical compound, drug | Collagenase D | Sigma-Aldrich | Cat# 11088866001 | |
| Chemical compound, drug | Heparin | Sigma-Aldrich | Cat# H3393 | |
| Chemical compound, drug | Red blood cell lysing buffer | Sigma-Aldrich | Cat# R7757 | |
| Chemical compound, drug | 2.4G2 culture supernatant | ATCC | HB-197 | |
| Commercial assay or kit | MojoSort Mouse CD8 T Cell Isolation Kit | BioLegend | 48007 | |
| Commercial assay or kit | Chromium Next GEM Single Cell 5’ Reagent Kits V2 | 10x Genomics | Cat# CG000330 | |
| Commercial assay or kit | dual index TN set TN set A | 10x Genomics | part no. 3000510 | |
| Gene (Mus musculus) | Thy1.1 | The Jackson Laboratory | Jax stock #005443, CBy.PL(B6)-Thy1a/ScrJ | |
| Gene (Mus musculus) | OT-I | The Jackson Laboratory | Jax stock #003831, C57BL/6-Tg(TcraTcrb)1100Mjb/J | |
| Gene (Mus musculus) | Rag1-/- | The Jackson Laboratory | Jax stock #002216, Rag1-/-: B6.129S7-Rag1tm1Mom/J | |
| Gene (Mus musculus) | E8I-creERT2 | Dario A.A. Vignali (University of Pittsburgh); Hirai et al., 2019 | ||
| Gene (Mus musculus) | ROSA26.LSL.hNGFR | Dario A.A. Vignali (University of Pittsburgh); Hirai et al., 2019 | ||
| Gene (Mus musculus) | Tgfbr3fl/fl | Herbert Y Lin (Program in Membrane Biology/Nephrology, Department of Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts); Li et al., 2018 | ||
| Other | BZX800 fluorescent microscope | Keyence | RRID:SCR_023617 | Microscope used for I.F. |
| Other | GentleMACS Dissociator | Miltenyi Biotec | RRID:SCR_020272 | See ‘Flow cytometry’ in ‘Materials and methods’ section |
| Other | PE Annexin V | BioLegend | Cat# 640907 | Fluorescent stain for cells undergoing apoptosis |
| Other | Soluble tetrameric B8R20–27/H2-Kb complex | NIH Tetramer Core Facility | For identifying CD8+T cells that respond to B8R20–27 peptide | |
| Other | TotalSeq-C Anti-Mouse Hashtags 1–10 | BioLegend | Cat #155861–155879 | Sample identifiers to allow pooling in scRNAseq |
| Other | FcBlock | Bio X Cell | Cat #BE0307 | To reduce false positives in flow cytometry |
| Other | NovaSeq S2 platform | Illumina | RRID:SCR_024569 | Used for scRNA sequencing |
| Other | ImmGenT TotalSeq-C custom mouse panel | BioLegend | Part no. 900004815 | Custom hashtags for scRNA-seq sample identification |
| Peptide, recombinant protein | OVA, N4, SIINFEKL | Genscript/Fisher | gene synthesis RP10161 | |
| Peptide, recombinant protein | Y3, SIYNFEKL | Genscript/Fisher gene synthesis RP10161 | gene synthesis RP10161 | |
| Software, algorithm | BZ-H4A Advanced Analysis Software | Keyence | RRID:SCR_017375 | |
| Software, algorithm | Adobe Photoshop (v6) | Adobe | RRID:SCR_014199 | |
| Software, algorithm | 10 x Genomics 5' v2 platform with Feature Barcoding for Cell Surface Protein and Immune Receptor Mapping | 10x Genomics | RRID:SCR_019326 | |
| Software, algorithm | ImmgenT single-cell RNA sequencing and processing | ImmgenT | Casey and Zemmour, 2025 | https://www.immgen.org/ImmGenT/ |
| Software, algorithm | CellRanger | 10x Genomics | RRID:SCR_023221 | |
| Software, algorithm | Seurat v4.1 | Satija lab and collaborators | RRID:SCR_007322 | |
| Software, algorithm | Prism v10 | GraphPad | RRID:SCR_002798 | |
| Software, algorithm | FlowJo v10 | FlowJo | RRID:SCR_008520 | |
| Strain, strain background (Vaccinia virus) | VV | Dr. J. Yewdell, National Institute of Allergy and Infectious Diseases | Vaccinia virus-Western Reserve strain | |
| Strain, strain background (Vaccinia virus) | VV-OVA | Dr. J. Yewdell, National Institute of Allergy and Infectious Diseases | Vaccinia virus-Western Reserve strain expressing OVA257-264 |
Additional files
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Epidermal resident memory T cell fitness requires antigen encounter in the skin
eLife 14:RP107096.
https://doi.org/10.7554/eLife.107096.3
