Search

Skin aging is characterized by structural and functional changes that contribute to age-associated frailty. This probably depends on synergy between alterations in the local niche and stem cell-intrinsic changes, underscored by proinflammatory microenvironments that drive pleotropic changes. The nature of these age-associated inflammatory cues, or how they affect tissue aging, is unknown. Based on single-cell RNA sequencing of the dermal compartment of mouse skin, we show a skew towards an IL-17-expressing phenotype of T helper cells, γδ T cells and innate lymphoid cells in aged skin. Importantly, in vivo blockade of IL-17 signaling during aging reduces the proinflammatory state of the skin, delaying the appearance of age-related traits. Mechanistically, aberrant IL-17 signals through NF-κB in epidermal cells to impair homeostatic functions while promoting an inflammatory state. Our results indicate that aged skin shows signs of chronic inflammation and that increased IL-17 signaling could be targeted to prevent age-associated skin ailments. Research in the S.A.B. laboratory is supported partially by the European Research Council under the European Union’s Horizon 2020 research and innovation program (grant agreement no. 787041), the Government of Cataluña (SGR grant), the Government of Spain (MINECO) and Foundation Lilliane Bettencourt. This work was partially supported by a Leonardo grant (BBVA Foundation) awarded to G.S. P.S. was awarded a Spanish Ministry of Economy and Development fellowship (MINECO, no. BES-2017-081279). J.B. was awarded a Spanish Ministry of Science and Innovation fellowship (MCI, no. PRE-2021-097118). M.E. and E.M. were supported by Ministerio de Ciencia e Innovación (MCI), Agencia Estatal de Investigación (AEI) and the European Development Regional Fund, ‘A way to make Europe’ ERDF (no. RTI2018-094049-B-I00). H.H. has received funding from Ministerio de Ciencia, Innovación y Universidades (no. SAF2017-89109-P; AEI/FEDER, UE). IRB Barcelona is a Severo Ochoa Center of Excellence (MINECO award no. SEV-2015-0505). The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript. We thank the histopathology, advanced digital microscopy and genomics facilities of IRB Barcelona and the genomics unit at CRG. We thank C. Niessen and E. Wachsmuth for useful tips on TEWL usage. We thank V. Raker for manuscript editing.

Additional file 1. Supplementary figures S1-S15.

Additional file 5. Review history.

Table S1. Overview table of experiments. Table S2. Differential gene expression between fresh and cryopreserved K562 cells. (MARS-Seq; top 40 genes). Table S3. Differential gene expression between fresh and cryopreserved HEK293 cells. (MARS-Seq; top 40 genes). Table S4. Differential gene expression between fresh and cryopreserved NIH3T3 cells. (MARS-Seq; top 40 genes). Table S5. Differential gene expression between fresh and cryopreserved MDCK cells. (MARS-Seq; top 40 genes). Table 6: Differential gene expression between fresh and cryopreserved HEK293 cells. (SMARTseq2; top 40 genes). Table 7: Differential gene expression between fresh and cryopreserved K562 cells. (SMARTseq2; top 40 genes). Table S8: Differential gene expression between fresh and cryopreserved PBMC. (MARS-Seq; top 40 genes). Table S9: Differential gene expression between a fresh and cryopreserved PDOX. (MARSseq; top 40 genes). (PDF 40 kb)

Figure S1. Single cell transcriptome sequencing of 670 fresh and 816 cryopreserved cells using the MARS-Seq (a-c) and the Smart-seq2 (d, e) library preparation protocols. Samples included two human (HEK293 and K562), one mouse (NIH3T3) and one canine (MDCK) cell line, peripheral blood mononuclear cells (PBMC), a primary mouse colon sample and orthotopic tumor xenografts (patient-derived orthotopic xenograft, PDOX). Analyses were split by replicate experiments and conditions. Displayed are the total number of reads per cell (a,d), the total number of detected genes per cell (b,e) and the number of detected transcripts (UMI counts) per cell (c). The results are displayed as boxplot indicating median values (black bar) per experiment and condition. Figure S2. (a-d) Cumulative gene counts split by fresh (red) or cryopreserved (blue) cells and analyzed using randomly sampled cells (average of 100 permutations). Displayed are results for K562 (experiment 1, a), HEK293 (experiment 1, b), NIH3T3 (c) and MDCK (d) cell lines. Figure S3. Distribution of sequencing reads numbers per single cell split by conditions (fresh (red); cryopreserved -80ºC: blue; cryopreserved liquid nitrogen: green). Displayed are the distributions for all MARS-Seq experiments (a-h) indicating the median number of reads per single cell (horizontal lines). Experiment types are indicated. Figure S4. MARS-Seq library complexity assessment of fresh (red) or cryopreserved (blue) cells using the number of uniquely detected transcripts. (a-d) Comparative analysis of the number of sequencing reads and detected transcripts using a linear model. The slope of the regression line was calculated separately for fresh and cryopreserved cells. Displayed are results for K562 (experiment 1, a), HEK293 (experiment 1, b), NIH3T3 (c) and MDCK (d) cell lines. Figure S5. MARS-Seq library complexity assessment of fresh (red) or cryopreserved (blue) cells using the total number of detected genes. (a-d) Comparative analysis of the number of sequencing reads and detected transcripts using a linear model. The slope of the regression line was calculated separately for fresh and cryopreserved cells. Displayed are results for K562 (experiment 1, a), HEK293 (experiment 1, b), NIH3T3 (c) and MDCK (d) cell lines. Figure S6. Comparative analyses of single cell transcriptome variance from fresh (red) and cryopreserved (blue) cell lines. Gene expression variances between cells are displayed as principal component analysis (PCA, upper panel a-d) or t-distributed stochastic neighbor embedding (t-SNE, lower panel e-h) using the 100 most variable genes. The displayed experiments include K562 (experiment 1; a,e), HEK293 (experiment 1; b,f), NIH3T3 (c,g) and MDCK (d,h). Figure S7. Joint analyses of single cells from fresh (red), cryopreserved at -80ºC (blue) and cryopreserved in liquid nitrogen (green) HEK293 cells from experiment 1 (circles) and 2 (triangles). Gene expression variances between cells are displayed as principal component analysis (PCA, a) or t-distributed stochastic neighbor embedding (t-SNE, b) using the 100 most variable genes. Figure S8. Comparative analyses of single cells from fresh (red) and cryopreserved (blue) cell lines. Gene expression variances between cells are displayed as t-distributed stochastic neighbor embedding (t-SNE) using the 100 most variable genes and the indicated perplexity parameter values. The displayed experiments include K562 (experiment 1), HEK293 (experiment 1 and 2), NIH3T3 and MDCK. Figure S10. Correlation analysis of gene expression levels split by conditions (fresh (red); cryopreserved -80ºC: blue; cryopreserved liquid nitrogen: green). (a-d) Pearson’s correlation analysis between 20 randomly selected fresh and cryopreserved cells displaying the correlation coefficient (r2). (e-h) Distribution of Pearson’s correlation coefficients (r2) within and between processing conditions. The median coefficients are indicated. (i-k) Linear regression model comparing average gene expression levels of expressed genes. The coefficient of determination (r2) is displayed. Cell lines and experiments are indicated within the figures. Figure S11. Cell subtype analysis of HEK293 (experiment 1) cells split by fresh (a-c) and cryopreserved (d-f) cells. (a,d) Hierarchical clustering of single cells based on transcriptional programs (defined by gene ontology) and correlating genes [21]. Transcriptional programs and gene clusters are summarized in aspects (orange: overrepresented; green: underrepresented). Displayed are the most variable aspects (rows) and their importance (row colors). Cells are assigned to condition (fresh: red; cryopreserved: blue) and clusters. (b, e) A t-distributed stochastic neighbor embedding (t-SNE) representation of similarities between cells using previous defined distances and cluster identity (as in a, d). (c, f) Hierarchical cluster of single cells (as in a,d) displaying the 25 most variable cell cycle genes (G2/M checkpoint). Expression levels of the cell cycle signature are summarized (1st panel; orange: high, green: low) and clusters are indicated. Figure S12. Analyses of Smart-seq2-derived single cell transcriptomes from a cryopreserved patient-derived xenograft (PDOX) tumor. (a) Displayed are the sequencing read distribution following RNA library preparation of full-length transcripts. Each line represents a single cell and transcript sizes are scaled from (0-100). (b,c) Comparative analyses of single cells from fresh (circles) and cryopreserved (triangles) samples. The displayed experiments include single cells from HEK293 cells (blue), K562 cells (orange) and the PDOX (green) sample. Gene expression variances between cells are displayed as principal component analysis (PCA, b) and t-distributed stochastic neighbor embedding (t-SNE, c) using the 100 most variable genes. Figure S13. Cell subtype analysis of HEK293 cells analyzed by Smart-seq2 split by fresh (a) and 5 cryopreserved (b) cells. (a,b) Hierarchical clustering of single cells based on transcriptional programs (defined by gene ontology) and correlating genes [21] displaying the 25 most variable cell cycle genes (G2/M checkpoint). Expression levels of the cell cycle signature are summarized (1st panel; orange: high, green: low) and clusters are indicated. (c) Significantly differentially expressed genes between fresh and cryopreserved HEK293 cells (p < 0.01). Figure S14. Cell subtype analysis of K562 cells analyzed by Smart-seq2 split by fresh (a) and cryopreserved (b) cells. (a,b) Hierarchical clustering of single cells based on transcriptional programs (defined by gene ontology) and correlating genes [21] displaying the 25 most variable cell cycle genes (G2/M checkpoint). Expression levels of the cell cycle signature are summarized (1st panel; orange: high, green: low) and clusters are indicated. (c) Significantly differentially expressed genes between fresh and cryopreserved K562 cells (p < 0.01). Figure S15. Cell subtype analysis of fresh (red) and cryopreserved (blue) peripheral blood mononuclear cells analyzed by MARS-Seq. (a-d ) Hierarchical clustering of single cells based on transcriptional programs (defined by gene ontology) and correlating genes [21] (as defined in Fig. 5f) displaying the 25 most variable genes within the signatures for cytotoxic T-cells (a), memory T-cells (b), B-cells (c) and myeloid cells (d). Expression levels of all signature genes are summarized (1st panel; orange: high, green: low) and conditions (2nd panel: fresh: red; cryopreserved: blue) and clusters are indicated. The lower plots summarize the signature expression levels for each cell (dots) and cluster (box). Inferred cell types are indicated (BC: Bcells, CytoTC: cytotoxic T-cells, MemTC: memory T-cells, Myd: myeloid cells). Figure S16. Hierarchical clustering of single cells from an ovarian tumor PDOX based on the most variable genes. Cells are assigned to conditions (2nd panel; fresh: red; cryopreserved: blue) and clusters (3rd panel). Displayed are the 25 most variable genes implicated in cluster formation. Expression levels of the genes are summarized (1st panel; orange: high, green: low). (PDF 6319 kb)