Psoriasis is a systemic inflammatory disease that primarily affects the skin (1). Approximately one-third of psoriasis patients also develop joint involvement and are therefore diagnosed with psoriatic arthritis (PsA) (2). In adult-onset disease, PsA typically manifests within 10 years after the onset of psoriasis, but approximately 15% of patients experience arthritis either concurrently or prior to the onset of skin disease (3). As a result, the diagnosis of PsA may be missed, delaying the introduction of effective treatment and resulting in disease progression and damage accrual (4). To date, disease-, activity- and/or outcome-specific biomarkers are missing, impeding disease monitoring and individualized care. Though the pathophysiology of psoriasis/PsA is complex and remains incompletely understood (5), the role of effector CD4⁺ T-cells has been established (6), and pathological activation of the Tumor Necrosis Factor (TNF)/Interleukin (IL-)23/IL-17 cytokine axis contributes to the differentiation and activation of effector T-cells and their accumulation in affected tissues (7–9).

Epigenetic modifications alter gene expression without affecting the underlying genomic sequence. Alterations to the epigenetic landscape have been observed in several autoimmune diseases, including psoriasis (10). The addition of a methyl group to the 5’ position of cytosine in cytosine phosphate guanosine (CpG) dinucleotides through DNA methyltransferases (DNMTs) is a potent epigenetic mechanism inhibiting the recruitment of transcription factors and RNA polymerases (11). Due to its stability in biological samples, DNA methylation is the most commonly investigated epigenetic mark (12). DNA methylation has been linked with the establishment of pathological effector T-cell phenotypes and the expression of inflammatory cytokines (13) in systemic autoimmune/inflammatory diseases, including psoriasis (14).

The aim of this study was to identify disease-associated DNA methylation signatures in CD4⁺ T-cells from patients with psoriasis and PsA that may be used as diagnostic and/or prognostic biomarkers to inform treatment and care.

The prediction of joint disease in psoriasis patients who will develop PsA, as well as early and correct diagnosis of PsA patients in the absence of skin disease are as challenging as crucial. The importance of early diagnosis lies in the destructive nature of joint involvement in PsA and the concept of a therapeutic “window of opportunity” in inflammatory musculoskeletal diseases, in which disease progression and damage may be prevented by the termination of inflammation (35, 36). Failure to initiate appropriate treatment in a timely fashion may result in irreversible joint damage and deformities, leading to chronic pain, disability and decreased quality of life (4, 37).

While some clinical predictors of joint involvement have been proposed, including 3 affected sites or more, high severity of skin disease, onychopathy, scalp lesions and intergluteal/perianal psoriasis (38, 39), no reliable clinical or laboratory biomarkers for the development of PsA are available. This study revealed distinct DNA methylation profiles in CD4⁺ T-cells that distinguish psoriasis patients from healthy individuals and revealed additional differences between patients with skin psoriasis and PsA. These observations underscore the importance of altered epigenetic mechanisms in the complex pathophysiology of psoriasis and PsA (40–42) and promise potential as diagnostic and/or prognostic biomarkers and future treatment targets.

A growing body of literature suggests that epigenetic modifications are involved in the molecular pathogenesis of psoriasis and PsA, including dysregulation of adaptive immune responses (43, 44), T-cell differentiation and activation (45), and keratinocyte dysfunction (46). The development of targeted therapeutic interventions to correct epigenetic signatures, e.g., DNA methylation patterns, may therefore permanently resolve immune dysregulation and improve outcomes in psoriasis, PsA and other T-cell mediated autoimmune/inflammatory diseases (47).

Given the recognized role of effector T-cells in the pathophysiology of psoriasis and PsA (6, 48–50), the here presented study focused on examining DNA methylation patterns in CD4⁺ T-cells. One crucial aspect of CD4⁺ T-cell involvement in psoriasis and PsA is their differentiation into distinct effector subsets, particularly T-helper 17 (Th17) and T-helper 22 (Th22) cells (51, 52). The IL-23/Th17 axis has emerged as a key pathway in chronic inflammatory conditions, including psoriasis and PsA (53). IL-23 promotes the survival and maintenance of Th17 cells, which produce IL-17, IL-22 and other inflammatory effector cytokines involved in psoriasis (54, 55). High expression of IL-23 has been observed in psoriatic skin lesions, and experimental models have demonstrated that IL-23 injection can induce epidermal hyperplasia through keratinocyte proliferation in mice (56). To further support the importance of this pathway, blockade of IL-17 (57–59) or IL-12/IL-23 (60–62) are effective treatment options in psoriasis and PsA. The recently identified Th22 effector T cell subset, characterized by IL-22 and IL-13 production, has also been implicated in psoriasis (63). Similarly to IL-17, IL-22 can stimulate keratinocyte proliferation, while impeding keratinocyte differentiation and inducing neutrophil infiltration in mice (64, 65).

Using Gene Ontology enrichment analysis, this study observed differential DNA methylation between controls and “all” psoriasis patients in promoter regions (one or more DMPs) affecting “regulation of wound healing, spreading of epidermal cells”, “negative regulation of cell-substrate junction organization” and “negative regulation of focal adhesion assembly”. Previous studies reported an involvement of accelerated wound healing in psoriasis (66) underscoring the significance of this process in its pathogenesis. Differential methylation of genes associated with cell-substrate junction organization and focal adhesion assembly has previously been observed in association with Munro micro-abscess formation, a histological hallmark in early psoriatic lesions and during disease flares (67). Lastly, in CD8⁺ T-cells, altered DNA methylation affecting genes associated with “cell junction assembly” was identified in psoriasis patients (skin psoriasis and PsA) when compared to healthy controls (14). These observations suggest that altered DNA methylation may affect biophysical properties of the skin and facilitate the recruitment of neutrophils to the epidermis.

To further explore differences between “all” psoriasis patients and controls, PLS-DA were performed and identified 20 DMPs (in component 1) that strongly contributed to the discrimination between groups. Among these, the CpG site cg01877366 located in TRAPPC9, encoding for trafficking protein particle complex subunit 9, was of special interest. TRAPPC9 is involved in the activation of the transcription factor nuclear factor (NF)-κB through phosphorylation of the IκB kinase complex. A previous study investigating DNA methylation profiles in PsA patients identified TRAPPC9 as a candidate gene for the prediction of failure to respond to TNF-inhibitors (68). Moreover, we identified cg1120622, located in the REV3L gene, encoding a DNA polymerase ζ catalytic subunit. In mice, deletion of REV3L resulted in impaired wound healing and excessive proliferation of the epidermis (69). Single nucleotide polymorphisms (SNPs) in the intergenic region between REV3L and the neighboring gene KIAA1919 are associated with the development of rheumatoid arthritis (RA) in black South Africans (70). Furthermore, REV3L has recently been identified as a candidate for gene therapy for psoriasis and PsA using a genetics-dependent drug target prioritization approach (31). Another CpG site contributing to the separation of patients and controls was cg18925478 within the PHACTR2 gene encoding for the phosphatase and actin regulator 2 that has previously been linked with the development of inflammatory bowel disease (71). Furthermore, the PHACTR2 protein was suggested as a potential marker of disease exacerbation in systemic lupus erythematosus (SLE) (72).

When comparing controls to “all” psoriasis patients, almost two-thirds of the genes associated with DMPs were involved in or affected by type I IFN, type II IFN, or both, and DNA methylation scores considering IFN-associated genes allow prediction of disease. These findings support previous reports implicating type I and type II IFN dysregulation in psoriasis (73, 74) and PsA (75) that may be the target of therapeutic interventions, e.g. with Janus kinase (JAK) inhibitors that are now approved for the treatment for moderate to severe psoriasis and PsA patients who have had an inadequate response or intolerance to methotrexate or other DMARDs (76–78) (79). Because of difficulties predicting disease progression from psoriasis to PsA and diagnosing PsA in the absence of skin involvement, DNA methylation patterns were investigated as possible predictors of disease phenotypes and outcomes. Approximately 3,000 DMPs were identified between skin psoriasis and PsA patients. However, because all patients enrolled in this study had developed skin disease prior to arthritis, it was not possible to determine whether the observed differences in DNA methylation profiles were present at disease onset and therefore predict the development of PsA or whether they developed with disease progression. Prospective studies are required to determine whether DNA methylation profiling can identify psoriasis patients at risk of developing PsA. Indeed, skin psoriasis patients treated with biologic DMARDs may be at a decreased risk of developing PsA (80) when compared to patients treated with other, less “aggressive” treatments, which underscores the significance of early diagnosis and therapeutic intervention.

Investigating possible functional implications of DMPs between patients with skin psoriasis and PsA, GO terms related to cAMP-dependent protein kinase (PKA) activity were enriched. Cyclic AMP is a key second messenger of extracellular ligands that is involved in a wide range of cellular responses (81). The homeostasis of cAMP is primarily regulated by adenylate cyclases and phosphodiesterases (PDE), responsible for the conversion of adenosine triphosphate (ATP) into cAMP and the degradation of cAMP, respectively. High intracellular levels of cAMP trigger activation of PKA by binding to its regulatory region and mediating dissociation of the PKA catalytic subunit, which phosphorylates various target proteins. Furthermore, PKA phosphorylates and regulates the activity and expression of the cAMP-response element modulator (CREM) protein family, a group of transcription factors playing an important role in various cellular functions (82). Increased expression of CREMα in CD4⁺ T-cells has been linked with effector cytokine expression in psoriasis, PsA (83) and other autoimmune diseases, including SLE (84) and juvenile idiopathic arthritis (85). The involvement of the cAMP-PKA pathway in the pathogenesis of psoriasis and PsA is further supported by the efficacy of PDE-4 inhibitors, with the European Medicines Agency (EMA) and Food and Drug Administration (FDA) having approved apremilast for the treatment of these conditions (86–88). Inhibition of PDE4 leads to high levels of cAMP, which in turn, activates PKA resulting in phosphorylation of transcription factors, such as cAMP Response Element-binding protein (CREB) and activating transcription factor-1 (ATF-1). As a result, phosphorylated CREB and ATF-1 promote the expression of anti-inflammatory cytokines, while inhibiting NF-κB activity. These findings highlight the importance of the cAMP-PKA pathway in psoriasis and PsA.

Contiguous differentially methylated CpG sites can compose differentially methylated regions (DMRs), which may exert regulatory effects on various biological processes, including cell function, proliferation or ageing (89–91). DMRs are tissue-specific and may be associated with disease stages and outcomes in several autoimmune/inflammatory conditions (92), such as RA (93), SLE (94) and Sjögren’s syndrome (95). Since coordinated changes in DNA methylation in broader genomic regions may have greater downstream biological consequences and influence disease development and progression, we performed DMR analysis in the sub-cohorts of the study (96, 97). Among all DMRs identified here, two were shared between all comparisons (controls versus skin psoriasis versus PsA): GDF7 on chromosome 2 and PIGZ/piRNA on chromosome 3. The GDF7 gene encodes for a secreted ligand of the Transforming Growth Factor Beta (TGF-β) superfamily and enhances regulatory T-cell (Treg) function through the upregulation of Forkhead Box P3 (FOXP3) and Cytotoxic T-Lymphocyte Associated Protein 4 (CTLA4) (98). Reduced expression of GDF7 was observed in CD4⁺ T-cells from SLE patients, suggesting its link to impaired function of Treg cells (99). The PIGZ gene encodes for a mannosyl-transferase involved in glycosylphosphatidylinositol (GPI)-anchor biosynthesis. GPI-anchored membrane proteins have been reported to be down-regulated in psoriatic skin lesions (100) and metabolites in GPI-anchor biosynthesis are altered in psoriasis patients (101).

Another DMP separating “all” psoriasis patients from controls was DUSP22, encoding for a protein phosphatase responsible for activation of the c-Jun NH2-terminal kinase (JNK) signaling pathway that is dysregulated in psoriasis (102), SLE (103) and ankylosing spondylitis (104). DUSP22 contributes to the inactivation of Lymphocyte Cell-Specific Protein-Tyrosine Kinase (Lck) during the deactivation phase of T-cell receptor signaling, thereby suppressing T-cell-mediated immune responses and inflammation (105). In RA, promoter hypomethylation of DUSP22 associates with erosive disease (106).

Objective assessment and quantification of treatment responses can be challenging but is essential for disease monitoring in clinical practice and clinical trial settings (107, 108). Thus, we explored the impact of cytokine-blocking agents (IL-17 or TNF inhibitors) on DNA methylation patterns in “all” psoriasis patients. A total of 1,996 DMPs were identified between treatment naïve patients and those who received biological DMARDs, which compared to methylation in healthy controls. The observed shift in methylation patterns in response to treatment indicates a significant impact of these agents on the DNA methylation landscape. Remarkably, pathway analysis of DMPs associated with treatment delivered a significant enrichment of genes associated with glutathione (GSH) metabolism. GSH, the most abundant endogenous antioxidant, plays a role in mitigating oxidative stress and regulating immune function. It is crucial for effector T-cell functions through the regulation of metabolic activity and has been linked with the pathogenesis of several autoimmune diseases (109), including psoriasis (110). In a recent randomized clinical trial investigating the IL-17A inhibitor secukinumab, lesional skin transcriptomic profiles showed Glutathione-S-transferase α3 among the top 10 upregulated genes after 12 weeks of treatment (111). This suggests that GSH metabolism plays a key role in T-cell mediated autoimmune/inflammatory diseases that may be targeted directly by future therapeutic interventions.

Because all patients enrolled in this study with samples available before and after treatment initiation received IL-17 or TNF-inhibitors, we focused DNA methylation analyses on this pathway by calculating DNA methylation scores focused on DMPs in genes involved in the IL-17/TNF pathway. This approach delivered an inverse correlation between DNA methylation and skin disease activity, as measured by PASI scores. Comparable results were observed in a previous study investigating DNA methylation in CD8⁺ T-cells in psoriasis and PsA (14). This suggests that cytokine-blocking agents can modulate DNA methylation patterns and restores “normal” profiles. DNA methylation profiles may therefore represent a potential biomarker for treatment response, but validation in larger prospectively assembled cohort is required.

Notably, numbers and proportions of CD4⁺ T-cells and their subsets did not differ across psoriasis, PsA and healthy controls. Thus, the observed differences among cohorts reflect variations in DNA methylation profiles of CD4⁺ T-cells more than “simple” shifts between their subsets.

While this study offers valuable insights into molecular mechanisms involved in skin psoriasis and PsA, it has limitations. The sample size was limited due to the relative rarity of PsA, especially of treatment naïve patients recruited here. Thus, we were also unable to enroll participants from minority ethnic groups that may exhibit differential DNA methylation patterns. Despite the potential bias introduced by age differences between psoriasis and PsA patients (112), normalization techniques and correction using the ComBat function from the ChAMP package were applied to minimize the impact of age differences on DNA methylation. Because of lacking information regarding active joint counts and joint disease activity in the PsA patients, we were not able to correlate these with DNA methylation patterns. Furthermore, this study was not able to assess the impact of DNA methylation on gene transcription, as RNA sequencing was not performed. Validation of DNA methylation patterns as a tool for the prediction of disease progression from skin psoriasis to PsA requires larger unrelated cohorts. Furthermore, to gain deeper insights into the transition from skin psoriasis to PsA, it will be important to include a subgroup of patients with PsA without pre-existing skin involvement.

DNA methylation profiles in CD4⁺ T-cells discriminate psoriasis patients from healthy individuals and skin psoriasis from PsA patients. DNA methylation may represent a stable and reliable biomarker candidate for predicting disease progression from skin psoriasis to PsA, monitoring disease activity and evaluating treatment response. Future application clinical practice requires prospective validation in independent cohorts.

The datasets presented in this study can be found in online repository. The names of the repository/repositories and accession number(s) can be found below: GSE236694 and GSE236695 (Gene Expression Omnibus).

The studies involving human participants were reviewed and approved by Faculty of Medicine Carl Gustav Carus, TU Dresden, Dresden, Germany. The patients/participants provided their written informed consent to participate in this study.

VN and AC analyzed flow cytometric datasets, performed the bioinformatic analysis and wrote the first draft of the manuscript. SH, FS, and SR isolated immune cells and DNA. SA consented patients and collected clinical data and biospecimen. CH oversaw all experimental and analytic steps and revised the first draft of the manuscript. CH, SH, LM, and SA conceived the study. VN, AC, LM, SH, SA, and CH were involved in individual or all steps of data analysis and overall data interpretation. All authors read, commented and agreed to the final version of the manuscript and the authors’ list. All authors contributed to the article.