Inhibiting LSD1 unlocks retinoid AP-1 programming to activate epithelial immunity and skin tumor suppression
Nina Kuprasertkul, Alyssa F. Moore, Carina A. D’souza, Julia Chini, Eun-Kyung Ko, Sijia Huang, Shuo Zhang, Ashley S. Anderson, Shaun Egolf, Laura V. Pinheiro, Alison Jaccard, Claudia T. Magahis, Lydia Bao, Yann Aubert, Cyria Olingou, Stephen M. Prouty, Donna Brennan-Crispi, David A. Hill, John T. Seykora, Kathryn E. Wellen, Brian C. Capell
Nina Kuprasertkul, Alyssa F. Moore, Carina A. D’souza, Julia Chini, Eun-Kyung Ko, Sijia Huang, Shuo Zhang, Ashley S. Anderson, Shaun Egolf, Laura V. Pinheiro, Alison Jaccard, Claudia T. Magahis, Lydia Bao, Yann Aubert, Cyria Olingou, Stephen M. Prouty, Donna Brennan-Crispi, David A. Hill, John T. Seykora, Kathryn E. Wellen, Brian C. Capell
View: Text | PDF
Research Article Dermatology Oncology

Inhibiting LSD1 unlocks retinoid AP-1 programming to activate epithelial immunity and skin tumor suppression

Abstract

Lysine-specific demethylase 1 (LSD1; KDM1A) orchestrates context-dependent chromatin programs, yet its role in epithelial immunity remains largely unknown. Here, we identify LSD1 as a central brake on retinoid-driven and activator protein-1–driven (AP-1–driven) enhancer activation in epidermis and demonstrate that its inhibition induces antitumor immunity. Whereas epidermal LSD1 is required during development, acute loss or topical inhibition in adult skin was tolerated and triggered coordinated expression of retinoic acid signaling, lipid remodeling, and chemokine induction pathways. CUT&RUN profiling revealed that LSD1 occupies enhancer regions enriched for AP-1 motifs at retinoid metabolism, lipid homeostasis, and immune genes. LSD1 loss increased H3K4me1/2 and gene activation at these sites, licensing a poised AP-1–retinoid program. Single-cell spatial analyses showed that discrete keratinocyte subsets initiate retinoid signaling to recruit dendritic cells and activate CD4+ T cell responses. Topical LSD1 inhibition suppressed cutaneous squamous cell carcinoma in 2 models while amplifying keratinocyte–immune cell crosstalk. Functional perturbations revealed that retinoid signaling partially contributes to, whereas CD4+ T cells are essential for, tumor control. These findings define LSD1 as a master repressor of epithelial immune competence and nominate LSD1 inhibition as a therapeutic strategy to activate retinoid–AP-1 enhancer circuits and drive CD4-dependent tumor immunity in skin cancer.

Authors

Nina Kuprasertkul, Alyssa F. Moore, Carina A. D’souza, Julia Chini, Eun-Kyung Ko, Sijia Huang, Shuo Zhang, Ashley S. Anderson, Shaun Egolf, Laura V. Pinheiro, Alison Jaccard, Claudia T. Magahis, Lydia Bao, Yann Aubert, Cyria Olingou, Stephen M. Prouty, Donna Brennan-Crispi, David A. Hill, John T. Seykora, Kathryn E. Wellen, Brian C. Capell

×

Figure 1

LSD1 is required for epidermal barrier development and maintenance of adult homeostasis.

Options: View larger image (or click on image) Download as PowerPoint
LSD1 is required for epidermal barrier development and maintenance of ad...
(A) Toluidine blue barrier assay of control and Lsd1-KO (Lsd1-eKO) embryos at E17.5. (B) Tamoxifen induction schematic for inducible Lsd1 deletion in adult mice (LSD1-KO). (C) Gross phenotype of wild-type (WT) and LSD1-KO mice taken posteuthanasia at approximately 12 weeks old. Yellow arrows highlight phenotypes. (D) H&E of WT and LSD1-KO mice with quantification. (E) Immunofluorescence (IF) for KRT10 (green) in WT and LSD1-KO mice. KRT14 (red) and DAPI (blue). (F) Immunohistochemistry (IHC) for LOR (yellow arrows indicate staining) (left), IF for IVL (green) (middle); IF for Ki67 (green) (right) in WT and LSD1-KO mice. KRT14 (red) and DAPI (blue). (G) Topical LSD1 inhibition (LSD1i) schematic. (H) H&E of topical DMSO– or topical LSD1i–treated mice with quantification. Inset depicts sebaceous glands at same scale. All scale bars: 100 μm. Data represented as mean ± SD. Each dot represents an individual mouse: for A, n = 16 control and n = 3 Lsd1-eKO; for C, images representative of 4–5 independent experiments; for D, n = 6 mice per condition; for E and F, n = 3 mice per condition for staining; for H, n = 5–7 mice per condition. Two-tailed Student’s t test unless indicated.