TYMP upregulation mediated by the hyperactivated IL-17/NF-κB1 axis promotes psoriasis through enhancing aberrant keratinization and neutrophil-mediated inflammation

Psoriasis is a common, chronic, and recurrent immune-mediated disorder with global prevalence, underscoring the need for novel biomarkers to improve diagnosis and treatment. In this study, differentially expressed genes (DEGs) in psoriatic tissues were comprehensively identified through integrated single-cell and bulk RNA-seq analyses. Thymidine phosphorylase (TYMP) emerged as one of the most significantly upregulated biomarkers in psoriasis. Multiplex immunohistochemistry (mIHC) and IHC assays jointly confirmed marked overexpression of TYMP in psoriatic keratinocytes. Mechanistically, we demonstrated that IL-17-mediated inflammatory signaling transcriptionally induces TYMP expression via NF-κB1. TYMP overexpression promotes keratinocyte proliferation and activates signaling pathways associated with keratinization and neutrophil degranulation in psoriasis. Immune infiltration analysis, blood routine tests, and ELISA verified that TYMP upregulation is closely correlated with neutrophil degranulation in psoriasis. In the imiquimod (IMQ)-induced psoriasis-like mouse model, pharmacological inhibition of TYMP by tipiracil partially reverses the pathological effects of TYMP overexpression, including accelerated keratinocyte proliferation, aberrant keratinization, and neutrophil-driven inflammation. In summary, TYMP is overexpressed in psoriasis due to hyperactivation of the IL-17/NF-κB1 signaling. Targeting TYMP by Tipiracil ameliorates psoriasis symptoms by suppressing abnormal keratinization and neutrophil degranulation, thereby highlighting its potential as a therapeutic target for psoriasis.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s13062-026-00734-2.

Highlights

Integrated analysis of single-cell and bulk RNA-seq data comprehensively identified biomarkers for psoriasis.

TYMP is transcriptionally induced by NF-κB1, and the overactivated inflammatory microenvironment leads to the overexpression of TYMP in psoriasis.

TYMP overexpression promotes abnormal keratinization and activates the neutrophil degranulation signaling pathway.

Targeting TYMP by TPI alleviates IMQ-induced psoriasis-like abnormal keratinization phenotype in the skin of mice.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13062-026-00734-2.

Introduction

Psoriasis is a chronic, immune–mediated systemic inflammatory condition characterized by well-demarcated red plaques coated with silvery scales [1]. The disease manifests in various clinical forms, with plaque psoriasis being the most prevalent, accounting for approximately 90% of all cases [2–4]. Globally, psoriasis affects around 2–3% of the population and is influenced by factors such as age, sex, geographic location, ethnicity, and environmental exposures [5–7]. Beyond its cutaneous manifestations, psoriasis is a systemic condition frequently accompanied by numerous comorbidities. These complications can significantly diminish patients’ quality of life, reduce their earning potential, and lead to psychological stress, financial hardship, mental health disorders, and even suicidal ideation [8–11]. In conclusion, psoriasis is a widespread and resource-intensive disease that imposes substantial personal, societal, and healthcare-related burdens worldwide [12–14].

The identification of psoriasis biomarkers contributes to early diagnosis, targeted therapy, and a deeper understanding of disease mechanisms. Although the precise mechanisms of psoriasis remains incompletely understood, accumulating evidence indicates that genetic factors, environmental triggerss, and dysregulation of the immune system collectively drive its onset and prorgression [15, 16]. Among the hallmark features of psoriasis are the inflammation and abnormal proliferation of keratinocytes [17].

The widespread adoption of high-throughput omics technologies—particularly next-generation sequencing (including bulk and single-cell RNA-seq) and mass spectrometry–based proteomics—has dramatically accelerated the discovery of biomarkers in diseases. Integrative multi-omics analysis—by combining genomic, transcriptomic, proteomic, and metabolomic data—enables a systems-level understanding of the complex molecular networks underlying disease, thereby facilitating the more accurate and comprehensive identification of biomarkers with diagnostic, prognostic, or therapeutic potential. In this study, we systematically identified key biomarkers of psoriasis by integrating single-cell RNA sequencing with our in-house bulk RNA-seq data.

As the most abundant cell type in the epidermis, keratinocytes play a central role in orchestrating cutaneous inflammatory responses by producing cytokines and chemokines and by interacting with other immune cells to amplify local inflammation in lesional skin [18, 19]. They are critically involved in both initiation and perpetuation of psoriasis, regulating innate immune response through the secretion of immunomodulatory molecules and promoting the activation of infiltrating immune cells [20]. The Interleukin-17 (IL-17) signaling pathway is a key driver of the autoimmune pathology in psoriasis [21]. Accordingly, anti-IL-17 therapies have emerged as highly effective treatments for psoriasis, demonstrating superior clinical efficacy compared to many other therapeutic approaches [22]. IL-17 is a key pro-inflammatory cytokine primarily produced by CD4(+) helper T cells (Th17 cells) and certain subsets of innate lymphoid cells [23]. Keratinocytes express surface receptors capable of sensing IL-17 and robustly respond to this signal [24]. IL-17-mediated inflammatory signaling promotes excessive keratinocyte proliferation, establishing a pathogenic feedback loop between keratinocyte hyperproliferation and IL-17–driven inflammation in psoriatic lesions [25].

Thymidine phosphorylase (TYMP), formerly known as endothelial cell growth factor-1 (ECGF1), has been identified as a key regulator of endothelial function [26]. Early studies linked TYMP to platelet-derived angiogenic activity; however, more recent finding have highlighted its dependence on thymidine catalysis by 2-deoxy-d-ribose, a process that actively drives endothelial cell behavior [27]. TYMP has been implicated in tumorigenesis, chemotherapy resistance, rheumatoid arthritis, and gastrointestinal malignancies such as gastric cancer [28–30]. Additionally, TYMP depletion contributes to insulin resistant lipoatrophic diabetes and metabolic remodeling and signaling pathways [31, 32].

Several studies have reported upregulation of TYMP in psoriatic skin [33, 34]; however, the mechanisms driving its overexpression and its functional contribution to psoriasis pathogenesis remain poorly defined. In this study, we aimed to investigate the regulatory mechanism underlying TYMP overexpression in psoriasis and evaluate the therapeutic potential of targeting TYMP with Tipiracil in preclinical models.

Materials and methods

Clinical samples

The study protocol was approved by the Ethics Committee of Renmin Hospital of Hubei University of Medicine (SYSRMYY-020) and conducted in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants prior to enrollment. Tissue samples were immediately frozen in liquid nitrogen after resection and stored at − 80 °C until use. For blood routine analysis, 100 psoriasis patients and 20 healthy volunteers who visited the Department of Dermatology at the Renmin Hospital of Hubei University of Medicine from May 2023 to April 2025 were enrolled.

RNA sequencing of clinical samples

The clinical samples, including 6 normal skin tissues and 8 psoriatic skin lesion tissues, were transported on dry ice to Shanghai Lifegenes Technology Co., Ltd. for RNA sequencing. Briefly, 1.5 µg of total RNA per sample was used as input material for the library preparations. Sequencing libraries were generated using NEBNext Ultra RNA Library Prep Kit for Illumina (NEB, USA) following manufacturer’s recommendations and index codes were added to attribute sequences to each sample [35–39]. The RNA-seq data have been deposited in the Gene Expression Omnibus (GEO) database at NCBI under accession number GSE302831. Differentially expressed genes (DEGs) were identified using the DESeq2 R package, and those with an adjusted p-value < 0.05 and |log₂ fold change| ≥ 1 were considered statistically significant.

M5 pro-inflammation cytokines treatment

HaCaT cells were stimulated with five cytokines (M5, a mixture of 10 µg/L IL-1α, 10 µg/L IL-17, 10 µg/L IL-22, 10 µg/L TNF-α, and 10 µg/L Oncostatin M) to induce psoriatic inflammation-like inflammatory conditions (in vitro). The human keratinocyte cell line HaCaT was purchased from the Shanghai Cell Bank of Chinese Academy of Sciences. Cells were routinely cultured in DMEM medium added with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin (P/S) at 37 °C in a 5% CO₂ incubator. HaCaT cells were seeded into 6-well plates and grown overnight. The next day, when confluence reached approximately 50%, cells were treated with PBS or M5 cytokine mixture (Neobioscience Technology Co, Ltd). After 72 h of treatment, cells were harvested for mRNA extraction and subsequent RNA-seq analysis. The RNA-seq data have been deposited in the Gene Expression Omnibus (GEO) database at NCBI under accession number GSE302830.

IMQ mouse model and tipiracil treatment

Eight-week-old wild-type C57BL/6 mice were obtained from Hubei Sihui Biotechnology Co., Ltd. (Hubei, China). All experimental procedures were conducted using 8-week-old male C57BL/6 mice in strict accordance with the ethical standards set forth in the Declaration of Helsinki and were approved by the Institutional Animal Care and Use Committee of Renmin Hospital, Hubei University of Medicine (Ethical Approval No. 25008). The animals were housed under specific pathogen-free (SPF) conditions at the Laboratory Animal Center of Hubei University of Medicine (Shiyan, Hubei, China) and randomly assigned to one of three groups (n = 5 per group). To induce a psoriasis-like skin inflammation model, mice in the IMQ group received daily topical application of imiquimod (IMQ; Sichuan Mingxin Pharmaceutical Co., Ltd., Sichuan, China) on their shaved dorsal skin for seven consecutive days. In the treatment group, mice were administered the thymidine phosphorylase inhibitor tipiracil via oral gavage at a dose of 1 mg/kg body weight once daily, concomitantly with IMQ application. Control mice received an equivalent volume of Vaseline (Unilever, London, UK) applied topically to the shaved back for the same duration.

Single-cell analysis

The single-cell RNA-seq data in the GSE151177, GSE162183, GSE173706, GSE230842, and the GSE248121 datasets was available in the GEO database [40–44]. The 10x data matrixes were imported into Seurat V5.1 R package (https://satijalab.org/seurat) to perform data filtration, sample integration, gene normalization, dimension reduction and data visualization. The detail steps for sample integration were given below. The cell type was annotated using the CellMarker2 database. The detail information of single-cell analysis was given below. The 10x data matrixes were imported into Seurat V5.1 R package to perform data filtration, sample integration, gene normalization, dimension reduction and data visualization. Cells with low feature counts (< 200 or > 8000) and high percent of mitochondrial genes (> 20%) were removed. To remove the batch effect, which may affect the accuracy of single cell analysis, we applied the batch effect correction analysis by the Harmony v1.2 package based on the top 2000 variable genes with the default harmony parameter. Dimension reduction was done by Seurat “RunPCA” function. Then Uniform Manifold Approximation and Projection for Dimension Reduction (UMAP) was used to visualize single-cell clusters, by graph-based clustering the cells, employing the top 20 principal components with the largest variance (at resolution = 0.2 for all the merged samples). The cluster of different cell types was identified by the DISCO platform [45].

RNA isolation and quantitative RT-PCR

The qRT-PCR assay was performed as previously described [46–48]. Briefly, Total RNA was extracted using Trizol reagent (Invitrogen, USA). Reverse transcription was performed to obtain cDNA by using the PrimeScript^™ RT reagent Kit (Perfect Real Time, Takara). The qPCR protocol was using One Step TB Green PrimeScript^™ RT-PCR Kit II (Takara) according to the manufacturer’s instructions. The qPCR analysis was conducted on Bio-Rad CFX Manager 3.1 real-time PCR system. The primers used in this study were listed as below: TYMP_qF: GAGGGCAGCGCCCTAAG; TYMP _qR: TGATGTCCGCTTCGCTCAG; GAPDH-qF: TCACCAGGGCTGCTTTTA; GAPDH-qR: AAGGTCATCCCTGAGCTGAA; β-actin-qF, ATCGTCCACCGCAAATGCTTCTA, β-actin-qR, AGCCATGCCAATCTCATCTTGTT.

ELISA assay

The content of secreted S100A7 (RMK0080, ABclonal, China), S100A8 (MK3813A, Meikebio, China), S100A9 (MK0612MB, Meikebio, China), and MPO (MK0651B, Meikebio, China) were measured using a sandwich ELISA. Briefly, 96-well plates pre-coated with capture antibody were blocked with BSA, washed three times with PBST, and incubated with diluted serum or supernatant samples at 37 °C for 1 h. After washing, a biotinylated detection antibody was added, followed by incubation at 37 °C for 1 h. Plates were then washed thoroughly, and HRP-conjugated streptavidin was added for signal development. Following three final PBST washes, a chromogenic substrate solution was added to each well. The reaction was stopped promptly upon visible color development by adding stop solution. Absorbance was measured at 450 nm using a microplate reader.

Immunohistochemistry assay

Skin samples from patients with psoriasis and healthy controls were pathologically assessed. Paraffin sections of clinical samples were processed for Histopathology and Immunohistochemistry. The Images were acquired using Leica laser microdissection systems (Wetzlar, Germany). Sections were analyzed using ImageScope for image acquisition (3DHISTECH, Budapest, Hungary). For Immunohistochemistry, the sections were deparaffinized and stained with anti-TYMP (A1094, Abclonal, Wuhan, China).

mIHC assay

Multiplex immunohistochemistry (mIHC) was performed using mIF kits (Servicebio Technology, Wuhan, China) according to the manufacturer’s protocol. This four-plex immunofluorescence assay employs tyramide signal amplification (TSA) technology to simultaneously detect four proteins on a single tissue section [49].

Briefly, mIHC assay was performed on skin tissues from IMQ-induced psoriasis-like mouse models. The primary antibodies and corresponding TSA fluorophores used in this study are listed below. TYMP (yellow, A1094, Abclonal), S100A8 (red, GB11421, Servicebio), SPRR1B (orange, 11959-1-AP, Abclonal), KRT1 (green, GB111277, Servicebio), CitH3 (white, ab5103, Abcam, USA), and MPO (orange, GB11224, Servicebio).

Western blotting assay

The antibody of TYMP (A1094, Abclonal, Wuhan, China), NF-κB1 (A11160, ABclonal, Wuhan, China), S100A7 (A22271, ABclonal, Wuhan, China), S100A8 (A12018, ABclonal, Wuhan, China), S100A9 (A9842, ABclonal, Wuhan, China), and GAPDH (AC054, ABclonal, Wuhan, China) was purchased from Proteintech company (Wuhan, China). The HaCat cells were lysed in RIPA buffer added 1mM PMSF. Approximately 50–100 µg of total protein was electrophoresed through 10% SDS polyacrylamide gels and were then transferred to a PVDF membrane. After blocking with 5% skimmed milk at 4 °C for 1 h, the membrane was incubated with primary antibody at 4℃ overnights. The blots were then washed and incubated with horseradish peroxidase (HRP)-conjugated secondary antibody (1: 10,000, Earthox) for 1.5 h at room temperature. Detection was performed by using a SuperLumia ECL HRP Substrate Kit (Abbkine) and visualized using a Bio-Rad Imaging System (USA).

Statistical analysis

For differential gene expression analysis, the P values were estimated using Welch t’ test or Wilcoxon signed rank test. Spearman correlation analysis was used for the correlation test of the two groups of data. The other experiments were used unpaired t-test or one-way ANOVA test. For quantitative RT-PCR, a minimum of triplicates per group and repetition of at least three times was applied to achieve reproducibility. All tests with p values less than 0.05 were considered to be statistically significant.

Results

Integrated single-cell analysis in multiple psoriasis cohorts

The hyperproliferation of keratinocytes is a well-established hallmark of psoriasis. Compared to the normal, the genes significantly upregulated in psoriatic keratinocytes are potential biomarkers for the disease. To identify such markers, we integrated five independent single-cell RNA-seq datasets (including GSE151177, GSE162183, GSE173706, GSE230842, and GSE248121) from the GEO database [40–44]. This meta-analysis encompassed a total of 240,495 single cells from 75 samples, including 41 psoriatic tissues and 34 normal tissues (Fig. 1A).

Fig. 1.

Integrated single-cell analysis of psoriatic and healthy skin tissues. (A) The workflow for integrated single-cell analysis in multiple psoriatic cohorts. (B) The expression profile of each cell-type-specific biomarkers in different cell types were shown in the violin plot. (C) The UMAP plots displayed the expression pattern of cell-type-specific biomarkers in different cell types. (D) The UMAP plots of 14 different cell types between normal and psoriatic tissues. (E) The total proportion analysis of 14 different cell types between normal and psoriatic tissues. (F) The proportion analysis of 14 different cell types was normalized in per sample. The proportion differences of per cell type between normal and psoriatic tissues were calculated. *P < 0.05, **P < 0.01, ***P < 0.001

After clustering, we obtained 14 distinct cell subclusters and annotated them using the DISCO web tool [45]. Uniform Manifold Approximation and Projection (UMAP) visualization revealed the following cell types: CD4_T cell, venous EC cell, lymphatic EC cell, Monocyte, vascular smooth muscle cell (VSMC), fibroblast, secretory cell, glial cell, melanocyte, basal keratinocyte, spinous keratinocyte, granular keratinocyte, and cycling keratinocyte (Fig. 1B). Cell-type-specific marker genes are displayed in the violin plot (Fig. 1C) The embeddings and comparative proportions of the 14 cell types between psoriatic and normal tissues were shown in Fig. 1D and E, respectively. To minimize technical bias, the proportion of each cell type was normalized per sample. Cell proportion analysis revealed that spinous keratinocytes, fibroblasts, lymphatic ECs, and glial cells were significantly decreased in psoriatic tissues, whereas CD4⁺ T cells, monocytes, and cycling keratinocytes were markedly enriched (Fig. 1F).

Identification of DEGs in psoriatic keratinocytes based on single-cell analysis

Herein, we focus on the four keratinocyte subtypes, spinous, granular, basal, and cycling keratinocytes. By comparing the transcriptomes of each subtype between psoriatic lesional and healthy skin tissues, we identified DEGs specific to each keratinocyte population in psoriasis (Fig. 2A-D). The full list of DEGs for each subtype of keratinocyte were shown in supplementary Table S1-S4. Intersection analysis revealed 2,027 genes significantly upregulated and 622 genes significantly downregulated across all four keratinocyte subtypes in psoriasis (Fig. 2E, F; supplementary Table S5). Given the large number of DEGs, we further narrowed our focus to the top 20 most upregulated genes in each subtype and identified a core set of five consistently overexpressed genes: TYMP, S100A8, S100A9, GJB2, and KRT6A (Fig. 2G). Notably, single-cell expression profiling showed that TYMP was markedly overexpressed not only in all keratinocyte subtypes but also in multiple other cell types within psoriatic lesions, including CD4⁺ T cells, monocytes, fibroblasts, endothelial cells, and vascular smooth muscle cells (VSMCs) (Fig. 2H, I).

Fig. 2.

Identification of the DEGs in different types of psoriatic keratinocytes. (A-D) Volcano plots display the log₂ fold changes (log₂FC) and P values for each gene from differential expression analysis comparing basal, cycling, granular, and spinous keratinocyte subpopulations in the integrated single-cell RNA-seq dataset. (E) The significant upregulated genes in all four types of psoriatic keratinocytes based on the integrated single-cell RNA-seq data was displayed in Venn plot. (F) The significant down-regulated genes in all four types of psoriatic keratinocytes based on the integrated single-cell RNA-seq data was displayed in Venn plot. (G) The top20 significantly upregulated genes in each type of keratinocyte were selected to take the intersection to identify the most significantly co-upregulated genes among the four types of keratinocytes. (H) The UMAP plots displayed the expression pattern of TYMP in normal and psoriatic tissues. (I) Analysis of differences in the expression levels of TYMP in each cell type between normal tissues and psoriatic tissues

Identification of DEGs in psoriasis based on bulk RNA-seq analysis

To identify potential biomarkers for psoriasis, we performed bulk RNA sequencing on our own psoriasis cohort, comprising 6 normal skin tissues and 8 lesional psoriatic tissues (Fig. 3A). Principal component analysis (PCA) clearly separated psoriatic and normal samples into two distinct clusters (Fig. 3B). Using the DESeq2 R package, we identified DEGs based on FPKM-normalized expression values (Fig. 3C, supplementary Table S6). According to the volcano plot, 3,991 genes were significantly downregulated and 3,741 genes were significantly upregulated in psoriasis (Fig. 3C). The GO/KEGG analysis showed that the upregulated genes in psoriasis were primarily enriched in IL-17 signaling, NF-κB signaling, Th1 and Th2 cell differentiation, and tuberculosis signaling (Fig. 3D). The top 50 DEGs (selected based on p-values) in psoriasis based on the GSE302831 cohort were displayed in the heatmap (Fig. 3E). Notably, TYMP was one of the most significant upregulated genes in psoriasis. Gene expression analysis of two independent, non-paired cohorts—comprising lesional psoriatic skin and healthy control skin (GSE302831 and GSE117405)—revealed that TYMP was significantly upregulated in psoriasis (Fig. 3F, G). Likewise, gene expression analysis of three independent paired cohorts—comprising lesional psoriatic skin and non-lesional skin from the same individuals (GSE34248, GSE13355, and GSE14905)—revealed that TYMP was significantly upregulated in psoriatic lesions (Fig. 3H). Furthermore, IHC assay in psoriatic cohort showed that TYMP protein was significantly overexpressed in the psoriatic tissues (Fig. 3I).

Fig. 3.

Identification of DEGs in psoriasis based on the bulk RNA-seq analysis. (A) We collected 8 psoriatic skin tissues and 6 healthy skin tissues for bulk RNA sequencing. (B) The PCA analysis based on the RNA-seq data of 14 samples. (C) The log₂ fold changes (log₂FC) and P values for each gene from DEG analysis based on the bulk RNA-seq data were shown in the volcano plot. (D) The GO/KEGG analysis showed the signaling pathways enriched by the upregulated genes in psoriasis. (E) The expression pattern of TOP 50 DEGs in psoriatic tissues was displayed in heatmap. (F, G) The gene expression analysis of TYMP in normal and psoriatic tissues in the GSE302831 and GSE117405 datasets. (H) TYMP expression in three independent paired cohorts comprising lesional psoriatic skin and non-lesional skin (GSE34248, GSE13355, and GSE14905). (I) Representative images of immunohistochemical staining of TYMP in normal skin tissues and psoriatic tissues. Statistical analysis revealed that the TYMP protein is significantly upregulated in psoriatic tissues. **P < 0.01, ***P < 0.001

TYMP expression is activated by inflammatory cytokines in keratinocytes

To investigate the underlying cause of TYMP overexpression in psoriasis, we performed KEGG pathway enrichment analysis using RNA-seq data from our GSE302831 psoriasis cohort. The results revealed significant hyperactivation of the IL-17 signaling pathway in lesional psoriatic skin (Fig. 4A). Furthermore, gene co-expression analysis based on the same dataset demonstrated a strong positive correlation between TYMP expression and the IL-17 signaling pathway in psoriatic lesions (Fig. 4B).

Fig. 4.

TYMP expression was significantly induced by IL-17 signaling in keratinocytes. (A) KEGG analysis based on the bulk RNA-seq data showed that IL-17 signaling was activated in psoriasis. (B) The gene expression correlation analysis between TYMP and IL-17 signaling-related genes were shown. (C) The psoriasis-like keratinocytes were constructed by treatment of M5 pro-inflammatory cytokines in HaCaT cells. (D) The log₂ fold changes (log₂FC) and P values for each gene from DEG analysis between PBS group and M5 treatment group were shown in the volcano plot. (E) The expression pattern of most significant DEGs in psoriatic tissues was displayed in heatmap. (F) RNA-seq analysis showed that TYMP expression was induced by M5 treatment in HaCaT cells. (G-I) The qRT-PCR, western blotting assay verified that TYMP expression was significantly increased in HaCaT cells with M5 treatment. (J) The effects of individual or combined treatment with inflammatory cytokines—including TWEAK, TNF, and IL-17 A—on TYMP gene expression. (K) Changes in TYMP gene expression levels before and after two weeks of treatment with an IL-17 monoclonal antibody or placebo. **P < 0.01

Obviously, the activated IL-17 inflammation signaling plays a critical role in psoriasis progression. Thus, HaCaT cells were stimulated with five inflammation cytokines (M5) to induce psoriatic inflammation-like conditions. RNA sequencing studies were conducted in those HaCaT cells with or without psoriatic inflammation-like conditions (Fig. 4C). After DEG analysis, we noted that the five most significant upregulated in keratinocytes (S100A8, S100A9, KRT6A, GJB2, and TYMP) were all significantly upregulated by M5 treatment in HaCaT cells (Fig. 4D, E). As shown in the Integrative Genomics Viewer (IGV) software, TYMP transcripts was obviously induced by M5 treatment in HaCaT cells (Fig. 4F). The qRT-PCR and western blotting assays further confirmed that the mRNA and protein levels of TYMP were significantly upregulated by M5 pro-inflammation cytokines in HaCaT cells (Fig. 4G-I). Gupta et al. have published transcriptomic data (GSE171170) from neonatal human epidermal keratinocytes (nHEKs) treated with various inflammatory cytokines [50]. Through analysis of this transcriptomic dataset (GSE171170), we similarly found that treatment with IL17A, TNF, or TWEAK alone significantly upregulated TYMP expression (Fig. 4J). Krueger et al. reported transcriptomic profiles (GSE31652) from a clinical trial in which psoriasis patients were treated with an IL-17 monoclonal antibody or placebo, with samples collected before and two weeks after treatment [51]. Through reanalysis of this transcriptomic dataset (GSE31652), we found that IL-17 monoclonal antibody treatment significantly suppressed TYMP expression (Fig. 4K). Together, these results indicate that TYMP expression is induced by inflammatory cytokines such as IL-17 A.

TYMP was transcriptionally activated by NF-κB1 in keratinocytes

As shown in the Fig. 4A and B, NF-κB1, a master transcriptional regulator of the IL-17 signaling pathway, was significantly upregulated in psoriatic lesions. Thus, we speculated that TYMP gene might be transcriptionally induced by NF-κB1 in psoriasis.

To verify this possibility, we conducted a gene co-expression analysis between NF-κB1 and TYMP in cancerous tissues and normal tissues based on the TCGA database and the GTEx database. The results showed that NF-κB1 and TYMP were highly co-expressed in pan-cancer and pan-tissues (Fig. 5A, B). Then, we retrieved the 2-kb promoter region (− 2000 to + 1 bp relative to the transcription start site) of the TYMP gene using UCSC web tool (supplementary Table S7). Promoter analysis with JASPAR showed that TYMP gene promoter contains 2 putative NF-κB1 binding sites (Fig. 5C).

Fig. 5.

TYMP expression was transcriptionally induced by NF-ΚB1 in HaCaT cells. (A) Pan-cancer analysis of TYMP in different cancerous tissues from the TCGA cohorts using GEPIA web tool. (B) Pan-tissue analysis of TYMP in different normal human tissues from the GTEx cohort using GEPIA web tool. (C) The promoter analysis showed that TYMP gene promoter contains two putative NF-κB1 binding motifs. (D-F) The qRT-PCR and western blotting assays confirmed that TYMP expression was significantly induced by NF-κB1 overexpression in HaCaT cells. (G, H) The gene expression correlation analysis based on the RNA-seq data of GSE302831 showed that TYMP was positively correlated with the expression of IL-17 signaling pathway-related genes in psoriasis. (I) Dual-luciferase reporter assay in HaCaT cells showed that the induction of TYMP expression by NF-κB1 overexpression required the physical binding between NF-κB transcription factor and TYMP gene promoter. **P < 0.01

To functionally validate this prediction, we further constructed a HaCaT cell line with stable NF-κB1 overexpression via lentiviral transfection. qRT-PCR assays verified that NF-κB1 was successfully overexpressed in HaCaT cells (Fig. 5D). And TYMP expression was significantly increased in the HaCaT cell line with stable NF-κB1 overexpression (Fig. 5E). Consistently, the western blotting assays also indicated that TYMP protein level was increased in the HaCaT cell line with stable NF-κB1 overexpression (Fig. 5F). In addition, dual-luciferase reporter assays demonstrated that wild-type TYMP promoter activity was enhanced by NF-κB1 overexpression, whereas mutation of the NF-κB1 binding sites markedly attenuated this induction (Fig. 5G and H). Taken together, transcription factor NF-κB1 directly binds to the TYMP promoter and drives its transcriptional activation in keratinocytes, contributing to TYMP upregulation in psoriasis.

TYMP overexpression significantly promotes the proliferation of keratinocytes

To investigate the functional consequences of TYMP upregulation in keratinocytes, we established a TYMP-overexpressing HaCaT cell line using lentiviral transduction (Fig. 6A). TYMP overexpression was confirmed by the qRT-PCR (Fig. 6B) and RNA-seq analysis. We then assessed the impact of TYMP overexpression on HaCaT cell behavior. CCK-8 assays showed that TYMP overexpression significantly enhanced the growth rate of HaCaT cells (Fig. 6C). The cell colony formation and EdU staining assays indicated that TYMP overexpression significantly promoted keratinocyte proliferation (Fig. 6D, E). Cell cycle assay showed that overexpression of TYMP obviously decreased the G1-phase HaCaT cells, but increased the S-phase HaCaT cells (Fig. 6F). Tipiracil (hydrochloride), a TYMP inhibitor (TPI), exerts anti-angiogenic effects by inhibiting endothelial cell proliferation [52]. In addition, treatment with TPI (10 µM) effectively reversed the pro-proliferative effects induced by TYMP overexpression in HaCaT cells (Fig. 6G).

Fig. 6.

TYMP overexpression significantly promotes the proliferation and cell cycle progression of keratinocytes. (A) The TYMP overexpression cell lines were established in HaCaT cells using lentiviral transfection. Scale bar: 100 μm. (B) The overexpression efficiency of TYMP was determined using qRT-PCR assays. (C) TYMP overexpression significantly accelerated the growth rate of HaCaT cells. (D, E) The Edu and colony formation assays showed that TYMP overexpression significantly promoted the proliferation rate of HaCaT cells. Scale bar: 100 μm. (F) TYMP overexpression significantly promoted G1/S phase transition in HaCaT cells. (G) The TYMP inhibitor (TPI) treatment (10 µM) rescued the promotion effects on the proliferation of HaCaT cells by TYMP overexpression. *P < 0.05, **P < 0.01

TTYMP overexpression enhances neutrophil degranulation and aberrant keratinization in psoriasis

To explore the molecular mechanism by which TYMP promotes psoriasis progression, we performed transcriptome sequencing on TYMP-overexpressing HaCaT cells and their control counterparts, respectively (Fig. 7A). DEG analysis identified 225 genes significantly upregulated by TYMP overexpression (log₂FC > 1, p < 0.05, Fig. 7B). These DEGs were visualized in a heatmap, which revealed that TYMP-upregulated genes were predominantly enriched in the IL-17 signaling pathway, neutrophil degranulation pathway (red arrow), and epithelial keratinization pathway (Fig. 7C).

Fig. 7.

TYMP overexpression promotes abnormal keratinization in psoriatic tissues and neutrophils degranulation in patients’ blood. (A) RNA sequencing studies were performed in HaCaT cells with or without TYMP overexpression. (B) The differentially expressed genes (DEGs) after TYMP overexpression were shown in the heat map after RNA-seq analysis. (C) The Top 100 DEGs after TYMP overexpression were shown in the heat map. According to the GO/KEGG analysis, the upregulated DEGs were enriched in keratinization, IL-17 and neutrophils degranulation signaling. (D, E) The GSEA analysis based on the RNA-seq data showed that keratinization and neutrophils degranulation signaling pathways significantly activated in the TYMP overexpression HaCaT cells. (F) Immune infiltration analysis based on the ssGSEA convolution method was performed based on GSE302831 RNA-seq data. The results showed that the infiltration levels of 6 immune cell types were significantly upregulated, while those of 7 immune cell types were significantly downregulated in psoriatic tissues. TFH, Follicular Helper T cell; Tgd, γδT cell; iDC, immature dendritic cells; aDC, antigen-presenting dendritic cells, Tem, effector memory T cells. The relative level immune infiltration of each immune cell type in normal and psoriatic tissues was shown in the heatmap. (G) The correlation analysis between TYMP expression and neutrophils infiltration in psoriasis. (H) The blood routine analysis between healthy volunteers and psoriasis patients. (I-K) The concentration of inflammation factors (S100A9, S100A7, and S100A8) of IL-17 signaling pathway in the blood of healthy volunteers and psoriasis patients were detected using ELISA assays. (L, M) The impact of TYMP overexpression and TPI treatment on the secretion levels of S100A8 and S100A9 in HaCaT cells.**P < 0.01, ***P < 0.001

Notably, RNA-seq data confirmed that HaCaT cells—under both basal and TYMP-overexpressing conditions—did not express detectable levels of IL17A or IL17F transcripts, indicating that IL-17 is not produced by keratinocytes themselves. Gene Set Enrichment Analysis (GSEA) further corroborated significant activation of the neutrophil degranulation and epithelial keratinization pathways in TYMP-overexpressing HaCaT cells (Fig. 7D, E, Supplementary Table S8).

To validate these findings in a clinical context, we performed single-sample GSEA (ssGSEA) on bulk RNA-seq data from the GSE302831 psoriasis cohort to infer immune cell infiltration patterns. The ssGSEA algorithm quantified the enrichment scores of 24 immune cell types. Differential infiltration analysis revealed that 13 immune cell populations exhibited significant changes in psoriasis. Specifically, regulatory T cells (Treg cells), Th1 and Th2 cells, CD4 T cells, neutrophils, and antigen-presenting cells (aDC cells) showed significantly increased infiltration, whereas mast cells, macrophages, CD8 + T cells, effector memory T cells (Tem cells), plasmacytoid dendritic cells (pDC), NK CD56 bright cells, and total NK cells were significantly reduced in psoriasis (Fig. 7F). Among these, inflammation-associated subsets—including Th1, Th2, and neutrophils—were prominently elevated in psoriatic skin (Fig. 7F). Besides, TYMP expression was positively correlated with neutrophil infiltration levels in the GSE302831 cohort (Fig. 7G).

To further assess the link between TYMP and systemic neutrophil activity, we analyzed peripheral blood routine data from 20 healthy volunteers and 40 psoriasis patients. Consistent with tissue findings, psoriasis patients exhibited significantly higher proportions of neutrophils and monocytes, along with a significant decrease in lymphocyte proportion (Fig. 7H). S100A8, S100A9, and S100A7 are key effector proteins downstream of IL-17 signaling and neutrophil degranulation. We therefore measured their serum levels in the same cohorts using enzyme-linked immunosorbent assay (ELISA). As shown in Fig. 7I-K, serum concentrations of S100A8, S100A9, and S100A7 were all significantly elevated in psoriasis patients compared to healthy controls. Additionally, in HaCaT cells, the TYMP-mediated promotion of S100A8 and S100A9 secretion was significantly reversed by TPI treatment (Fig. 7L, M). Taken together, TYMP upregulation contributes to neutrophil-mediated inflammation and aberrant keratinization in psoriasis.

Inhibition of TYMP by TPI alleviates IMQ-induced psoriasis-like inflammatory phenotypes in the skin

We first performed multiplex immunohistochemistry (mIHC) assays on human lesional psoriatic and non-lesional normal skin tissues (n = 8). In these assays, TYMP was labeled in yellow; S100A8, a well-established inflammatory marker in psoriasis, in red; SPRR1B in orange; and KRT1, a canonical keratinocyte differentiation marker, in green (Fig. 8A). Quantitative analysis revealed that the expression levels of TYMP, S100A8, SPRR1B, and KRT1 were all significantly elevated in psoriatic lesions compared to normal skin (Fig. 8B–E). Moreover, consistent with our single-cell RNA-seq findings, TYMP showed strong colocalization with the keratinocyte-associated proteins KRT1 and SPRR1B, supporting its keratinocyte-intrinsic role.

Fig. 8.

Inhibition of TYMP alleviates IMQ-induced psoriasis-like abnormal keratinization and the increase in plasma MPO levels. (A) The mIHC assay using S100A8 (red), TYMP (yellow), SPRR1B (orange), KRT1 (green), and DAPI (blue) was performed in our own cohort containing 5 paired of normal and psoriatic tissues. The keratinocytes (marker gene: KRT1) were labeled in green. (B-E) The IHC score of S100A8, KRT1, SPRR1B, and TYMP in normal and psoriatic tissues. (F) Photographs of the dorsal skin tissues from mice in the control group, the IMQ-alone treatment group, and the IMQ + TPI combined treatment group. (G) Representative HE staining images of the dorsal skin tissues from mice in the control group, the IMQ-alone treatment group, and the IMQ + TPI combined treatment group. (H, I) The IHC staining of TYMP in skin tissue of mice from the control group, the IMQ-alone treatment group, and the IMQ + TPI combined treatment group. (J) The serum levels of MPO in mice from the control group, the IMQ-alone treatment group, and the IMQ + TPI combined treatment group. (K) In the IMQ-induced psoriasis-like mouse model, NETs levels in skin tissues were assessed by colocalization of MPO and CitH3. *P < 0.05, **P < 0.01, ***P < 0.001

To evaluate the functional impact of TYMP inhibition in vivo, we employed the imiquimod (IMQ)-induced murine model of psoriasis-like dermatitis (Fig. 8F). Tipiracil hydrochloride (TPI), a specific TYMP inhibitor known to exert anti-angiogenic effects by suppressing endothelial cell proliferation [52], was administered to assess its therapeutic potential. Histological analysis of skin sections by H&E staining confirmed successful induction of psoriasis-like inflammation in the IMQ-treated group (Fig. 8G). Notably, TPI treatment (1 mg/kg) significantly attenuated IMQ-induced epidermal hyperplasia and reduced TYMP expression (Fig. 8G-I), indicating that pharmacological inhibition of TYMP ameliorates key pathological features of psoriatic skin inflammation.

Myeloperoxidase (MPO) is a key enzyme released by neutrophils during degranulation and is essential for the formation of neutrophil extracellular traps (NETs), where it promotes chromatin decondensation and microbial killing. We measured plasma MPO levels in mice using ELISA and found that serum MPO was markedly elevated in IMQ-treated mice, whereas TPI treatment significantly reversed this increase (Fig. 8J). Additionally, NETs were labeled by the colocalization of citrullinated histone H3 (CitH3) and MPO. We found that TYMP overexpression significantly increased NETs formation, and this effect was effectively reversed by TPI (Fig. 8K).

Discussion

In the last ten years, advancements in sequencing technologies have facilitated remarkable progress in understanding the molecular mechanisms underlying psoriasis [53]. Transcriptomic studies have been instrumental in identifying genes that are differentially expressed in psoriasis, shedding light on the disease’s pathological processes [54]. Epigenomic investigations have further elucidated how genetic, immunological, and environmental factors contribute to the dysregulated expression of key psoriasis-associated genes [55]. GWAS (Genome-wide association studies) have uncovered numerous genetic variants and susceptibility genes associated with psoriasis [56]. Additionally, single-cell sequencing has advanced our understanding of how the immune microenvironment influences disease progression [57]. These publicly available multi-omics datasets hold immense potential for psoriasis research, as their comprehensive integration and in-depth analysis remain largely unexplored.

The pathogenesis of psoriasis is driven by a complex interplay of immune dysregulation, genetic predisposition, and environmental influences [54]. A hallmark of the disease is the hyperproliferation of keratinocytes and chronic inflammation, which are well-established features of psoriasis [12, 58, 59]. Central to this process is the hyperactivation of T cells and the dysregulation of the IL-23/IL-17 axis, which are critical for driving keratinocyte hyperproliferation, angiogenesis, and persistent inflammation in psoriatic skin [17]. Dendritic cell (DC)-derived cytokines, such as TNF-α and IL-23, play a pivotal role in promoting Th17 cell differentiation and IL-17 secretion, further exacerbating psoriatic lesion development by amplifying keratinocyte proliferation and inflammatory responses [60]. Consequently, therapeutic agents targeting the IL-23 pathway, such as Guselkumab and Secukinumab, and those targeting IL-17, like Ixekizumab and Brodalumab, have demonstrated significant efficacy in treating psoriasis [17]. Despite these advancements, the precise mechanisms underlying psoriasis remain incompletely understood.

In the present study, we aim to identify novel psoriatic biomarkers using multi-omics data. Considering the characteristic of excessive proliferation of keratinocytes in psoriasis, we focused on identifying DEGs of keratinocytes between normal skin tissues and psoriatic skin tissues using single-cell datasets. Our integrated single-cell analysis identified four types of keratinocytes, including basal, cycling, granular, and spinous. DEG analysis showed that a total of 2027 genes were significantly upregulated in the 4 types of keratinocytes, and 622 genes were significantly downregulated in the 4 types of keratinocytes (Supplementary Tables S1-S5).

Notably, TYMP is one of the most significantly upregulated genes in psoriasis based on the single-cell and bulk RNA-seq analysis in multiple independent psoriatic cohorts. TYMP, also known as ECGF1 (endothelial cell growth factor), encodes an angiogenic factor which promotes angiogenesis in vivo and stimulates the in vitro growth of a variety of endothelial cells [61]. Several studies have reported that TYMP was upregulated in psoriasis [33, 34]. Consistently, our IHC and mIHC assays also confirmed that TYMP was significantly upregulated in keratinocytes of psoriatic skin lesion. However, the mechanisms underlying its increased expression and its precise role in psoriasis pathogenesis remain poorly understood.

In the present study, we confirmed that TYMP mRNA and protein levels were significantly upregulated in both the cellular model (M5 cytokine cocktail-induced psoriatic-like in vitro model) and the mice model (IMQ-induced psoriatic-like in vivo model). Both the M5 cytokine cocktail and IMQ can significantly activate inflammatory signaling pathways in cells. Thus, we speculated that TYMP expression was greatly induced by inflammatory signaling. RNA-seq analysis showed that the IL-17-mediated inflammatory signaling pathway is significantly activated in psoriasis-affected tissues (Fig. 4a). And TYMP exhibits high co-expression with the IL-17 signaling pathway-related genes (Fig. 4b), prompting us to hypothesize that TYMP gene expression could be induced by IL-17-mediated inflammatory signaling pathway. As expected, Treatment with IL-17 A alone significantly upregulates TYMP gene expression in human keratinocytes, and this upregulation can be completely blocked by an IL-17 monoclonal antibody (Fig. 4j, k). Most importantly, we demonstrated that TYMP expression is transcriptionally induced by NF-κB1, a downstream factor of the IL-17 signaling. In other words, owing to the abnormal activation of the IL-17/NF-κB1 signaling pathway, TYMP expression is significantly upregulated in psoriatic keratinocytes. Consistent with our view, Chapouly and team found that TYMP expression was induced by inflammatory factors such as IL-1β and IL-17 in astrocytes of multiple sclerosis, an inflammatory central nervous system condition [61].

Neutrophils are key effector cells of the innate immune system that play a critical role in host defense. Neutrophils can participate in and amplify cutaneous inflammatory responses through multiple mechanisms, including degranulation (releasing various effector molecules), formation of neutrophil extracellular traps (NETs), and secretion of cytokines and extracellular vesicles [62]. While T cells play a prominent role, polymorphonuclear neutrophils (PMN) are also significant players in the pathogenesis of psoriasis [63]. An increasing number of studies have shown that neutrophils play a pivotal role in the pathogenesis of psoriasis—particularly in subtypes such as generalized pustular psoriasis [64]. Aberrant or excessive degranulation and the formation of neutrophil extracellular traps (NETs) contribute significantly to sustained inflammation and epidermal damage in psoriasis [65–67], .

The molecular mechanism by which TYMP drives the progression of psoriasis has not been reported yet. In this study, we revealed that TYMP overexpression significantly activated the signaling pathways related to keratinization and neutrophil degranulation. Results from immune infiltration analysis, blood routine analysis, and ELISA indicated that neutrophil infiltration was increased in psoriasis and positively correlated with the expression level of TYMP. In the imiquimod (IMQ)-induced murine psoriasis model, we found that TYMP overexpression significantly promoted neutrophil degranulation and the formation of neutrophil extracellular traps (NETs), and this phenotype could be reversed by TPI treatment. Similarly, in HaCaT cells, TYMP overexpression markedly enhanced cell proliferation and secretion of S100A8/A9, which was also rescued by TPI treatment. Collectively, the above evidence suggests that TYMP may serve as a therapeutic target for psoriasis.

Conclusions

In psoriasis, TYMP expression is markedly upregulated due to aberrant activation of the IL-17 inflammatory signaling pathway. On one hand, TYMP promotes excessive proliferation and abnormal keratinization of keratinocytes; on the other hand, it induces aberrant neutrophil degranulation and enhances the formation of neutrophil extracellular traps (NETs), thereby amplifying inflammatory signaling and driving disease progression. Our study highlights TYMP as a promising therapeutic target for psoriasis, demonstrating that pharmacological inhibition of TYMP—such as with TPI—effectively alleviates disease-associated pathological progression.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

QSS and HYL conceived and designed the study. QSS and LDD wrote the paper. WJ, LDD, ZLT, ZZ, LQ, and CYY performed most of the experiments. WJ and LD carried out initial data analyses and performed partial experiments. All authors contributed to drafting the manuscript. All authors have read and approved the final submitted manuscript.

Funding

This study was supported by grants from the Hubei Provincial Natural Science Foundation (2025BCB065) and the Faculty Development Grants from Hubei University of Medicine (2024QDJZR009 to ZZ).

Data availability

RNA sequence data were deposited in GEO (accession number GSE302830, GSE302831, and GSE309584). All data generated or analyzed during this study are included in the manuscript and supporting files; source data files have been provided for all figures.

Declarations

Ethics approval and consent to participate

This study was approved by the Human Research Ethics Committee of Hubei University of Medicine (SYSRMYY-020), in accordance with the Declaration of Helsinki.

Consent for publication

All authors agreed on the manuscript.

Competing interests

The authors declare no competing interests.