Introduction
Atopic dermatitis (AD) is the most prevalent chronic inflammatory skin disease worldwide, affecting up to 20% of children and 10% of adults. AD impairs both cutaneous and systemic health, profoundly disrupting sleep, mental health, quality of life, and daily function. It may also serve as an early gateway to IgE-mediated sensitization and the atopic march toward food allergy, asthma and allergic rhinitis, fostering lifelong allergic multimorbidity. The natural history of AD is characterized by heterogeneous trajectories of disease activity. More than 50% of affected children may enter spontaneous, durable, therapy-free remission, whereas others develop persistent, relapsing and treatment-refractory disease. The biological basis for why AD resolves spontaneously in some individuals yet persists lifelong in others remains a clinically important and mechanistically unresolved question with direct implications for prognosis, biomarker discovery, prevention, and drug development. To address this, we aimed to define the immune, stromal, and epigenetic programs underlying AD remission.
Methods/Material
To define cellular and molecular landscape of disease states, we obtained skin punch biopsies from patients with active AD, long-term natural remission, and healthy human donors. These skin biopsies were subjected to spatial transcriptomic profiling using the 10X Genomics’ Visium HD platform. Raw sequencing reads were mapped to the human reference genome, and downstream spatial clustering, cellular deconvolution, differential expression analyses and immune niche analysis were executed using the Seurat and BANKSY packages. In addition, we performed whole-genome bisulfite sequencing (WGBS) on peripheral CD4+ T cells from a subset of donors to explore the epigenetic modifications in peripheral immune compartment. The WGBS data were processed using Bismark for alignment and methylation calling, followed by DMRichR for differentially methylated region (DMR) analysis.
Results
Spatial profiling identified 21 cell types based on canonical marker gene expression. Long-term remission represented a near-homeostatic tissue state that remained distinct from healthy skin and was characterized by selective immune regulation. In long-term remission donors, keratinocytes (KC) normalized expression of the T-cell chemoattractant CCL27 but retained features of barrier dysfunction, including reduced filaggrin expression and increased SPRR expression, indicating persistent alterations in KC differentiation. Remission fibroblasts showed reduced interferon signaling (CXCL12, CXCL14) and restoration of immune homeostasis but only partial stromal normalization, with elevated complement genes expression (C3, CFD) and incomplete recovery of extracellular matrix components. Immune niche analysis performed using BANKSY identified conserved microenvironments shared between active AD and remission, suggesting persistent localized tissue remodeling despite clinical resolution. Epigenetic profiling of CD4+ T cells showed differential promoter methylation of Th2 master regulator GATA3 and Th2-associated genes, including IL25 and IRF4 across conditions.
Discussion
Our preliminary experiments provide the first high-resolution cellular map comparing long-term AD remission skin against both active AD and healthy skin. Our findings indicate that clinical remission is not merely a passive return to baseline, but rather a distinct immunological state. This state is defined by persistent tissue remodeling along with unique cellular and molecular adaptations across immune, stromal and epidermal compartments of the skin. Ongoing work is focused on further defining the cellular, spatial and epigenomic architecture of remission and relapse in human skin and blood to better understand the mechanisms underlying divergent AD trajectories (persistence, remission, and relapse), with the aim of identifying new therapeutic targets and strategies for disease modification and long-term therapy-free disease resolution.