Researchers have identified a fundamental genetic mechanism that enables skin cells to retain long-term inflammatory memory, providing new insight into how chronic inflammatory diseases such as psoriasis persist and recur.
The study, led by scientists at The Rockefeller University and published in Science, combines experimental biology with deep learning to uncover how specific DNA sequence features determine the longevity of inflammatory memory in skin stem cells.
Skin stem cells encode long-term inflammatory memory
It has long been observed that inflammatory skin conditions often recur at the same anatomical locations. Previous work demonstrated that skin stem cells can retain “memories” of past inflammatory events, enabling faster responses to future injury—but also contributing to chronic disease when dysregulated.
However, the molecular basis underlying the persistence of these memories over months or years remained unclear.
The new study identifies distinct genomic regions, termed “memory domains,” that remain epigenetically primed following inflammation. While many such regions revert to baseline over time, a subset persists long-term and appears to drive chronic disease states.
As described in the study, “distinct genetic sequences propel a handful of critical memories into the years-long timeframes that underpin chronic disease.”
CpG density encodes memory persistence
To understand what distinguishes short-lived from long-lasting memory domains, the researchers developed a deep learning model called PersistNet. This model analyzed DNA sequence features associated with memory longevity.
The key determinant was CpG dinucleotide density, short DNA motifs consisting of cytosine followed by guanine.
Memory domains with higher CpG density were significantly more likely to persist long-term. The findings suggest that CpG content effectively encodes a “timer” within the genome, dictating how long inflammatory memory is retained.
Experimental validation confirmed that CpG density alone could predict memory persistence across approximately 1,000 identified memory domains.
Epigenetic stabilization sustains memory across cell generations
The study further revealed the mechanistic basis of this persistence. CpG-rich regions undergo coordinated epigenetic modifications, including DNA demethylation, recruitment of transcription factors favoring open chromatin and incorporation of the histone variant H2A.Z.
Together, these changes stabilize chromatin accessibility and maintain gene priming long after the initial inflammatory stimulus has resolved.
This epigenetic state can be inherited across cell divisions, allowing inflammatory memory to persist over the lifespan of the organism.
In mouse models, approximately 10–15% of memory domains remained active for up to two years, effectively the animal’s lifetime, highlighting their potential relevance to chronic human disease.
Resolving a long-standing paradox in inflammation biology
The findings address a central paradox in the field: how epigenetic signals, which are expected to dilute over successive cell divisions, can nonetheless persist long enough to drive chronic disease.
By demonstrating that intrinsic DNA sequence features, rather than transient regulatory inputs, govern memory stability, the study provides a mechanistic link between acute inflammation and long-term disease recurrence.
Therapeutic implications beyond dermatology
The identification of CpG-driven inflammatory memory opens new possibilities for therapeutic intervention. Targeting the epigenetic machinery that stabilizes these memory domains could help disrupt maladaptive inflammatory cycles.
While the current work focuses on skin, the authors suggest that similar mechanisms may operate in other conditions where cellular memory contributes to pathology, including cancer, chronic pain, and metabolic disorders.
Future research will aim to distinguish beneficial inflammatory memory, such as enhanced wound healing, from maladaptive forms that drive chronic disease.
As senior author Elaine Fuchs, PhD, notes, understanding these differences may be key to “break[ing] the cycle of inflammatory disease.”
Toward precision targeting of inflammatory memory
By linking DNA sequence architecture to long-term epigenetic memory, this study establishes a new conceptual framework for chronic inflammation. Rather than focusing solely on transient signaling pathways, it highlights the importance of stable genomic features in disease persistence.
These insights lay the groundwork for next-generation therapies aimed at selectively erasing harmful cellular memories while preserving beneficial ones—an approach that could redefine treatment strategies for chronic inflammatory diseases.
