Integrated Microbiome and metabolome analysis reveals Microbial-Metabolic interactions in psoriasis pathogenesis

Abstract

Background

Psoriasis is a chronic inflammatory skin disorder with unclear etiology. The roles of skin microbiome and metabolic dysregulation in psoriasis pathogenesis are not yet fully understood.

Methods

We conducted an integrated microbiome and untargeted metabolomic analyses on skin samples from 29 patients with psoriasis and 31 healthy controls. The skin microbiota was characterized using 16 S rRNA gene sequencing, and untargeted metabolomic profiling was performed using LC-MS/MS. Multivariate statistical analyses were used to identify differential microbes and metabolites, followed by correlation analyses to explore microbe-metabolite interactions.

Results

Psoriatic lesions exhibited significantly higher skin microbial alpha diversity compared to healthy controls. Principal component analysis revealed distinct microbial community structures between the two groups. At the genus level, Corynebacterium and Staphylococcus were significantly enriched in psoriatic lesions, while Cutibacterium was notably reduced. Metabolomic analysis identified 63 differential metabolites, with 39 upregulated and 24 downregulated in psoriatic lesions. These metabolites were primarily involved in lipid metabolism (particularly phospholipids and sphingolipids), amino acid metabolism, and inflammatory mediator pathways. Correlation analysis revealed significant associations between microbial alterations and metabolic dysregulation. Cutibacterium abundance was negatively correlated with inflammatory lipids and positively correlated with antioxidant metabolites, whereas Staphylococcus and Corynebacterium exhibited the opposite pattern. Notably, the abundance of Propionibacteriaceae strongly correlated with glutathione levels (r = 0.821, P < 0.001), indicating a potential role of microbiome-mediated oxidative stress in psoriasis.

Conclusions

This study highlights significant alterations in both the skin microbiome and metabolome in patients with psoriasis, revealing complex microbe-metabolite interaction networks. The findings suggest that microbial dysbiosis, particularly the decreased abundance of Cutibacterium and the increased abundance of Staphylococcus/Corynebacterium, may contribute to psoriasis pathogenesis by modulating lipid metabolism, inflammatory pathways, and oxidative stress responses.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12866-025-04395-5.

Introduction

Psoriasis is a chronic, immune-mediated inflammatory dermatosis affecting approximately 2–3% of the global population, clinically characterized by well-demarcated erythematous plaques with silvery scales, pathological keratinocyte hyperproliferation, and infiltration of inflammatory cells into the dermis and epidermis [1, 2]. The disease significantly impairs patients’ quality of life through physical discomfort, psychological distress, and social stigmatization. Despite significant advances in understanding psoriasis pathophysiology, particularly regarding the IL-23/IL-17 axis and the development of targeted biological therapies, the precise etiology and triggering factors remain incompletely understood [3, 4].

The pathogenesis of psoriasis involves a complex interplay between genetic predisposition, environmental triggers, immune dysregulation, and barrier dysfunction. Genome-wide association studies have identified multiple susceptibility loci associated with psoriasis [5], predominantly involving genes related to skin barrier function and immune regulation [6, 7]. Nevertheless, genetic factors alone cannot explain the variable disease manifestations, suggesting that additional environmental and microbial factors may play crucial roles in disease initiation and progression.

Increasing evidence suggests that the skin microbiome plays a crucial role in maintaining cutaneous homeostasis and may contribute to various inflammatory skin disorders [8, 9]. The human skin harbors diverse microbial communities that interact with host cells and influence immune responses, barrier function, and metabolic processes [10]. Recent studies have reported alterations in skin microbiome composition in psoriasis patients compared to healthy individuals [11], including changes in the abundance of Cutibacterium (formerly Propionibacterium), Staphylococcus, and Corynebacterium species [12]. However, the functional implications of these microbial alterations and their potential contributions to disease pathogenesis remain poorly characterized.

Concurrently, metabolic dysregulation has emerged as an important component of psoriasis pathophysiology. Metabolomics studies have identified significant alterations in various metabolic pathways in patients with psoriasis, including lipid metabolism, amino acid metabolism, energy metabolism, and oxidative stress responses [13, 14]. These metabolic changes may contribute to the inflammatory environment, impaired barrier function, and aberrant cell proliferation characteristic of psoriatic lesions [15, 16]. Sphingolipids and ceramides, essential components of the epidermal barrier, often show altered profiles in psoriatic skin, potentially contributing to barrier dysfunction [17, 18]. Similarly, dysregulation of arachidonic acid metabolism may lead to increased production of pro-inflammatory eicosanoids, further exacerbating the inflammatory cascade in psoriatic lesions.

Moreover, emerging evidence suggests potential bidirectional interactions between the microbiome and host metabolism [19]. Skin commensals synthesize diverse bioactive molecules capable of modulating host cellular responses, including lipid metabolism and inflammatory pathways [20, 21]. Conversely, host-derived metabolites can shape the composition and function of the skin microbiome [22]. This bidirectional communication creates a complex ecosystem where perturbations in either component might disrupt skin homeostasis and contribute to disease states.

Recent technological advances in multi-omics approaches have facilitated more in-depth investigations of the skin ecosystem, allowing simultaneous analysis of microbiome composition and metabolic profiles [23]. These integrated approaches offer opportunities to uncover novel connections between microbial communities and metabolic pathways potentially involved in disease pathogenesis. However, comprehensive analyses integrating microbiome and metabolome data to explore microbe-metabolite interactions in psoriasis are still limited, representing a significant knowledge gap in our understanding of this complex disorder.

The skin, as the largest organ of the human body, represents a complex ecosystem where microorganisms, host cells, and metabolites interact continuously [10]. Disruptions in this delicate balance might contribute to the initiation and perpetuation of inflammatory skin disorders like psoriasis [24, 25]. Of particular interest is the role of commensal species such as Cutibacterium acnes, which has been shown to modulate skin lipid metabolism and possess anti-inflammatory properties [26, 27]. Alterations in the abundance or function of such beneficial commensals might contribute to the dysregulated inflammation and barrier dysfunction observed in psoriasis [28].

To address these knowledge gaps, we conducted an integrated analysis of skin microbiome and metabolome profiles in psoriasis patients and healthy controls. This integrated multi-omics approach could provide novel insights into the complex interplay between skin microbiota and host metabolism in psoriasis pathogenesis and potentially identify new therapeutic targets or biomarkers for this challenging skin disorder. Our findings reveal significant microbiome-metabolome interaction networks that likely play a pivotal role in the pathophysiology and offer new perspectives for understanding the complex etiology of psoriasis.

Materials and methods

Study participants and sample collection

The study was approved by the Institutional Ethics Committee of Shandong Provincial hospital, and all participants provided written informed consent before enrollment. We recruited 29 patients with plaque psoriasis (17 males and 12 females, mean age 42.3 ± 12.5 years) and 31 healthy volunteers (16 males and 15 females, mean age 38.7 ± 14.2 years) were recruited to achieve similar age and sex distributions as the psoriasis group to minimize confounding factors as controls (Table S1). Patients were diagnosed by dermatologists based on clinical manifestations and histopathological examinations when necessary. The inclusion criteria for patients were: (1) age between 18 and 65 years; (2) diagnosis of plaque psoriasis for at least 6 months; (3) Psoriasis Area and Severity Index (PASI) score ≥ 3; and (4) no systemic or topical anti-psoriatic treatments for at least 4 weeks before enrollment. Exclusion criteria for all participants included: (1) other concomitant dermatological diseases; (2) systemic antibiotic use within 3 months before sampling; (3) systemic immunosuppressive therapy within 3 months before sampling; (4) pregnancy or lactation; and (5) serious systemic diseases.

Skin samples were collected from the lesional skin at the dorsal aspects of extremities in patients and from comparable anatomical sites of healthy controls. Before sampling, the skin surface was gently cleaned with sterile saline solution. Skin microbiome samples were collected using sterile cotton swabs soaked in sterile saline solution by rubbing the skin area (5 cm × 5 cm) for 30 s. For metabolomic analysis, skin surface samples were collected from the same area using tape stripping method (D-Squame tapes, CuDerm Corporation, Dallas, TX, USA). Three consecutive tape strips were pooled for each participant. All samples were immediately frozen in liquid nitrogen and stored at −80 °C until analysis.

Skin Microbiome analysis

DNA extraction and 16 S rRNA gene sequencing

Microbial DNA was extracted from skin swabs using the DNeasy PowerSoil Kit (Qiagen, Hilden, Germany) according to the manufacturer’s instructions. The V3 and V4 regions of the 16 S rRNA genes were amplified by the primers 341 F (CCTAYGGGRBGCASCAG) and 806R (GGACTACNNGGGTATCTAAT). DNA concentration and purity were assessed using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA) and agarose gel electrophoresis. PCR reactions were performed in triplicate using the following conditions: initial denaturation at 95 °C for 3 min, followed by 30 cycles of denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 45 s, with a final extension at 72 °C for 10 min. PCR products were purified using the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA) and quantified using the QuantiFluor-ST system (Promega, Madison, WI, USA).

The purified amplicons were pooled in equimolar ratios and paired-end sequenced (2 × 250 bp) on an Illumina NovaSeq platform (Illumina, San Diego, CA, USA) according to standard protocols.

Microbiome data analysis

Raw sequencing reads were demultiplexed and quality-filtered using QIIME2 (version 2020.2). Briefly, reads with a quality score < 20, containing ambiguous bases, or with more than one mismatch in the primer sequence were removed. The filtered reads were merged using FLASH (version 1.2.11) with a minimum overlap of 10 bp and a maximum mismatch rate of 0.2. Chimeric sequences were identified and removed using USEARCH (version 11.0.667). Operational taxonomic units (OTUs) were clustered with a 97% similarity cutoff using UPARSE, and taxonomy was assigned using the Ribosomal Database Project (RDP) classifier against the SILVA database (release 138) with a confidence threshold of 0.7.

Alpha diversity metrics (Shannon and Simpson indices) were calculated using QIIME2. Beta diversity was assessed using principal component analysis (PCA) based on Bray-Curtis distances. Statistical significance of differences in alpha diversity between groups was determined using the Wilcoxon rank-sum test. Differential abundance analysis at the genus level was performed using the Linear Discriminant Analysis Effect Size (LEfSe) method (Kruskal-Wallis and pairwise Wilcoxon). The correlations between specific taxa and psoriasis severity (PASI score) were analyzed by using Spearman correlation tests.

Untargeted metabolomics analysis

Sample Preparation and LC-MS/MS analysis

Metabolites were extracted from tape strips using a methanol/water (4:1, v/v) solution containing internal standards. Briefly, tape strips were immersed in 1 mL extraction solution, vortexed vigorously for 1 min, and sonicated for 10 min in an ice-water bath. After centrifugation at 14,000 × g for 15 min at 4 °C, the supernatant was collected and dried under a gentle stream of nitrogen gas. The dried extracts were reconstituted in 100 µL of methanol/water (1:1, v/v) for LC-MS/MS analysis.

Metabolomic analysis was performed on a UHPLC system (Vanquish, Thermo Fisher Scientific) coupled with a Q Exactive HF-X mass spectrometer (Thermo Fisher Scientific). Samples were separated on a BEH C18 column (2.1 × 100 mm, 1.7 μm, Waters) using a gradient elution with mobile phase A (0.1% formic acid in water) and mobile phase B (0.1% formic acid in acetonitrile) at a flow rate of 0.3 mL/min. The gradient program was as follows: 0–1 min, 2% B; 1–9 min, 2–98% B; 9–12 min, 98% B; 12–12.1 min, 98 − 2% B; 12.1–15 min, 2% B.

The mass spectrometer was operated in both positive and negative ion modes with a spray voltage of 3.5 kV and − 2.8 kV, respectively. The capillary temperature was set at 320 °C. Full MS scans were acquired at a resolution of 60,000 with a mass range of m/z 100–1500. MS/MS fragmentation was performed using normalized collision energy of 20, 40, and 60 eV. Quality control (QC) samples, prepared by pooling equal volumes of all samples, were analyzed after every 10 samples to monitor system stability.

Metabolomic data analysis

Raw data were processed using Compound Discoverer 3.1 (Thermo Fisher Scientific) for peak detection, alignment, and quantification. The parameters were set as follows: mass tolerance, 5 ppm; retention time tolerance, 0.2 min; signal-to-noise ratio threshold, 3; minimum peak intensity, 5 × 10^4. Peak identification was performed by comparing mass spectra and retention times with standards in our in-house library and public databases, including HMDB, METLIN, and MassBank. For compounds without available standards, identification was based on accurate mass and MS/MS fragmentation patterns.

The normalized peak areas were log-transformed and Pareto-scaled before multivariate statistical analysis. Principal component analysis (PCA) and orthogonal partial least squares discriminant analysis (OPLS-DA) were performed using SIMCA (version 16.0.2, Sartorius Stedim Biotech). The quality of the OPLS-DA model was evaluated based on R^2X, R^2Y, and Q^2 values. Differential metabolites between patients and healthy controls were identified using a multi-criteria approach. First, an OPLS-DA model was built, and metabolites with a variable importance in projection (VIP) value > 1 were selected as potential candidates. Second, a Student’s t-test was performed for these candidates between the two groups. To account for multiple comparisons, the resulting P-values were adjusted using the Benjamini-Hochberg method to control the False Discovery Rate (FDR). Metabolites with both a VIP > 1 and an adjusted P-value (q-value) < 0.05 were considered significantly different. The correlations between metabolite alterations and PASI score were analyzed by using Spearman correlation tests. Pathway enrichment analysis was performed using MetaboAnalyst 5.0 based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) database. The pathway impact was calculated as the sum of the importance measures of the matched metabolites normalized by the sum of the importance measures of all metabolites in each pathway.

Integrated Microbiome-Metabolome analysis

To explore potential relationships between skin microbiome and metabolome, correlation analyses were performed between differential bacterial genera and differential metabolites. Spearman’s rank correlation analysis was conducted to assess monotonic relationships, while Pearson’s correlation analysis was employed to evaluate linear relationships. Significant correlations (|r| ≥ 0.6, P < 0.05) were visualized using heatmaps and network graphs generated with R (version 4.3.1).

Canonical correlation analysis (CCA) was performed to identify multivariate relationships between the microbiome and metabolome datasets using R (version 4.3.1). CCA plots were used to visualize the projection of microbes and metabolites onto the first two canonical dimensions, revealing clusters of closely associated microbes and metabolites.

Additionally, to assess the contribution of specific microbes to metabolite alterations, we employed a regression-based approach. For each differential metabolite, a multiple linear regression model was constructed with bacterial genera as predictors, and the relative contribution of each genus was calculated based on standardized regression coefficients.

Statistical analysis

Statistical analyses were performed using R (version 4.1.0). Continuous variables were expressed as mean ± standard deviation and compared using Student’s t-test or Wilcoxon rank-sum test depending on data distribution. Categorical variables were presented as frequencies and percentages and compared using the chi-square test or Fisher’s exact test as appropriate. For all statistical tests, a P-value or adjusted P-value (q-value) of < 0.05 was considered statistically significant. Specific details on multiple testing corrections for high-throughput data are provided in the relevant subsections.

Results

Altered skin Microbiome composition in patients with psoriasis

To investigate the relationship between psoriasis and skin microbiome, we analyzed skin microbial samples from 29 patients and 31 healthy controls.Patients with psoriasis had lower bacterial ace and chao1 indices than controls (Wilcoxon test unadjusted P < 0.05) (Fig. 1A and B). This finding suggests less species richness and evenness in the skin microbiota of patients with psoriasis, align with previous investigates [11].

Fig. 1.

Significant differences in skin microbiome between patients with psoriasis and healthy controls. A-B Alpha diversity measured by ace index (A) and chao1 index (B) showing significantly lower microbial diversity in patients with psoriasis compared to healthy controls (Wilcoxon test P<0.05). C Principal component analysis (PCA) displaying separation between patients with psoriasis (red dots) and healthy controls (blue dots). Ellipses represent 95% confidence intervals. D Relative abundance of major bacterial genera showing increased Corynebacterium and Staphylococcus, and decreased Cutibacterium in patients with psoriasis compared to healthy controls. E Simper differential contribution based on ASV（Amplicon Sequence Variant）

Principal component analysis (PCA) demonstrated distinct clustering patterns between psoriatic lesions and healthy controls, indicating significant differences in microbial community structures (Fig. 1C). The 95% confidence ellipses showed clear separation between psoriasis samples and healthy control samples.

At the genus level, we observed significant differences in the relative abundance of several major bacterial genera between the two groups (Fig. 1D). Corynebacterium and Staphylococcus were significantly enriched in psoriatic lesions compared to healthy controls (21.2% vs. 10.5% and 18.6% vs. 9.2%, respectively). In contrast, Cutibacterium was markedly depleted in psoriatic lesions (17.3% vs. 45.8%). Additionally, increased relative abundances of Lawsonella and Kocuria were observed in psoriatic lesions. We found significant correlations between PASI scores and several bacterial genera: Staphylococcus abundance (r = 0.58, P = 0.001), Cutibacterium abundance (r = −0.45, P = 0.014), and alpha diversity Shannon index (r = 0.41, P = 0.028) (Table S3). These differences suggest that psoriasis may be associated with specific alterations in microbial communities.

Distinct metabolomic profiles in patients with psoriasis

Untargeted metabolomic analysis revealed significant differences in metabolic profiles between patients with psoriasis and healthy controls. Both PCA and orthogonal partial least squares discriminant analysis (OPLS-DA) showed clear separation between the two groups (Fig. 2A-B), indicating systematic metabolic alterations in psoriasis.

Fig. 2.

Metabolomic profiling reveals distinct metabolic signatures in psoriasis. (A) Principal component analysis (PCA) and (B) orthogonal partial least squares discriminant analysis (OPLS-DA) score plots showing clear separation between patients with psoriasis (orange) and healthy controls (blue). C Model validation parameters demonstrating good predictive ability (Q² >0.5). D S-plot identifying key differential metabolites. E Hierarchical clustering heatmap of differential metabolites, with functional classifications shown on the right. F Volcano plot showing 39 upregulated (red) and 24 downregulated (green) metabolites in psoriasis (VIP >1.0, unadjusted P < 0.05, and FDR-adjusted P < 0.05 (Benjamini-Hochberg correction)). G Relative abundance changes of major differential metabolites in patients with psoriasis compared to healthy controls

Using stringent statistical criteria (VIP > 1, P < 0.05), we identified 63 differential metabolites. Volcano plot analysis (Fig. 2F) showed that 39 metabolites were significantly upregulated and 24 were downregulated in patients compared to healthy controls (Table S2). These differential metabolites were primarily involved in:

1. (Lipid metabolism dysregulation: Phosphatidylcholines, lysophosphatidylcholines, ceramides, and sphingomyelins were significantly upregulated in psoriasis patients. Notably, the levels of phosphatidylcholines PC (16:0/18:1) and PC (18:0/20:4) were nearly twice as high in patients compared to healthy controls.

2. Amino acid metabolism abnormalities: Various amino acids and their metabolites showed significant alterations. Glutamate, arginine, and tryptophan metabolite levels were significantly elevated, while serine and glycine levels were decreased.

3. Energy metabolism changes: TCA cycle-related metabolites such as succinate, fumarate, and citrate were significantly increased in the psoriasis group.

4. Oxidative stress markers: Several oxidative stress-related metabolites showed significant alterations, with reduced glutathione (GSH) levels decreased and oxidized glutathione (GSSG) levels increased.

5. Inflammatory mediators: Multiple inflammation-related metabolites, including arachidonic acid and its metabolites such as prostaglandins and leukotrienes, were significantly elevated in psoriasis patients.

Among the 63 significantly altered metabolites, 12 showed significant correlations with PASI scores (|r| >0.4, P < 0.05). Inflammatory mediators showed positive correlations: arachidonic acid (r = 0.67, P < 0.001), leukotriene B4 (r = 0.59, P = 0.001), sphingosine-1-phosphate (r = 0.52, P = 0.004), and tryptophan (r = 0.48, P = 0.008). Skin barrier-related metabolites showed negative correlations: ceramide (d18:1/24:0) (r = −0.55, P = 0.002), cholesterol sulfate (r = −0.49, P = 0.007), and vitamin D3 (r = −0.44, P = 0.017). We have added these results as Table S3 (detailed correlation data). This analysis reveals that inflammatory lipid mediators and skin barrier-related metabolites are most strongly associated with disease severity, providing mechanistic insights into psoriasis pathogenesis.

Pathway enrichment analysis revealed that differential metabolites were significantly enriched in glycerophospholipid metabolism, sphingolipid metabolism, and arachidonic acid metabolism pathways (P < 0.01) Fig. 3. Additionally, purine metabolism, glutathione metabolism, and histidine metabolism pathways also showed significant enrichment.

Fig. 3.

Pathway analysis of differential metabolites in psoriasis. A KEGG pathway enrichment analysis bubble chart showing significantly enriched metabolic pathways. Glycerophospholipid metabolism, sphingolipid metabolism, and arachidonic acid metabolism were most significantly enriched (P < 0.01). B Pathway overall change analysis based on differential abundance score (DA Score). C KEGG pathway differential metabolite clustering heatmap displaying expression patterns across metabolic pathways in psoriasis and control groups. D Metabolite regulatory network analysis showing key regulatory metabolites (red nodes) with extensive connections to peripheral metabolites. E HMDB-based functional annotation and enrichment bubble chart. F Metabolome enrichment analysis (MSEA) results. G Analysis of associations between differential metabolites and other diseases

Correlation between skin Microbiome and metabolome in psoriasis

To explore potential interactions between skin microbiome and metabolome, we performed comprehensive correlation analyses between differential microbes and metabolites. Correlation network analysis (Fig. 4A) revealed multiple microbe-metabolite interaction patterns. Cutibacterium showed significant negative correlations with multiple phospholipid metabolites (r < −0.6, P < 0.05), particularly with phosphatidylcholine (PC) series compounds (Fig. 4B-C), suggesting that decreased Cutibacterium abundance might be associated with abnormal lipid metabolism [29] in psoriasis patients. In contrast, Staphylococcus and Corynebacterium exhibited significant positive correlations with various inflammation-related metabolites (such as arachidonic acid derivatives) and sphingolipid metabolites, indicating that the increase of these microbes might promote inflammatory responses and skin barrier dysfunction [30].

Fig. 4.

Correlation analysis between skin microbiome and metabolome in psoriasis. A Microbiome-metabolome correlation network showing significant correlations (|r| ≥ 0.6, P < 0.05). Red lines indicate positive correlations, blue lines indicate negative correlations. B Hierarchical clustering heatmap of microbiome-metabolome correlations clearly separating psoriasis and control samples. C Correlation heatmap of top 20 differential metabolites and microbial genera. D Correlation scatter plot between Cutibacterium relative abundance and serine levels (r = -0.619, P < 0.001). E Chord diagram of Spearman correlations between key microbes and metabolites, showing distinct association patterns

More specifically, Cutibacterium showed strong negative correlations with ceramides, sphingomyelins, and phosphatidylcholines (r < −0.7), while Staphylococcus demonstrated strong positive correlations with arachidonic acid, prostaglandins, and leukotrienes (r > 0.7). Kocuria and Lawsonella exhibited significant positive correlations with various amino acid metabolites, suggesting their potential involvement in amino acid metabolism dysregulation in psoriasis patients.

Notably, we found a significant positive correlation between Propionibacteriaceae abundance and glutathione levels (r = 0.821, P < 0.001), as clearly illustrated in the correlation scatter plot (Fig. 5C). This finding supports the hypothesis that microbial dysbiosis, particularly decreased Propionibacteriaceae bacteria, might contribute to psoriasis pathogenesis by affecting the host’s antioxidant defense system.

Fig. 5.

Pearson correlation analysis revealing specific microbiome-metabolome association patterns. A Hierarchical clustering heatmap of Pearson correlations between microbes and metabolites. B Correlation heatmap of top 20 differential metabolites and microbial genera. Cutibacterium shows strong positive correlations with antioxidants, while Staphylococcus and Corynebacterium correlate positively with inflammatory mediators. C Correlation scatter plot between Propionibacteriaceae and glutathione (r = 0.61, P < 0.001). D Chord diagram of Pearson correlations between key microbes and metabolites

Additionally, we observed significant positive correlations between Corynebacterium tuberculostearicum abundance and specific phospholipid metabolites, particularly PE-NMe2, as well as metabolites involved in cysteine metabolism pathways. C. tuberculostearicum abundance also positively correlated with IL-23 pathway-related metabolites, suggesting its potential involvement in psoriasis pathogenesis through modulation of lipid metabolism and promotion of inflammatory responses.

CCA further confirmed these patterns, showing that Cutibacterium clustered with various antioxidant-related metabolites, while Staphylococcus and Corynebacterium clustered with inflammatory mediators and abnormal lipid metabolites in different regions of the plot. This multivariate analysis validated the overall correlation patterns between the microbiome and metabolome, suggesting that skin microbial dysbiosis in psoriasis is closely associated with coordinated dysregulation of multiple metabolic pathways.

Contribution analysis of specific metabolites, such as inosine (Fig. 6B), provided further mechanistic insights into how microbial changes might influence metabolite alterations. Cutibacterium showed the largest positive contribution to inosine levels, suggesting that decreased Cutibacterium abundance might be an important factor leading to reduced inosine levels in psoriasis patients. Conversely, Staphylococcus and Corynebacterium exhibited significant negative contributions, indicating that increased abundance of these microbes might inhibit inosine production or promote its metabolism.

Fig. 6.

Multivariate analysis revealing complex associations between microbiome and metabolome. A Canonical correlation analysis (CCA) scatter plot of differential microbes (blue points) and metabolites (orange points). Points located in the same region far from the origin and close to each other indicate high canonical correlation. B Contribution analysis of microbial species to inosine metabolism changes. Red bars indicate positive contributions, blue bars indicate negative contributions. Cutibacterium shows the largest positive contribution, while Staphylococcus and Corynebacterium show significant negative contributions

These findings collectively reveal intricate microbe-metabolite interaction networks in psoriasis and suggest that microbial dysbiosis may contribute to psoriasis pathogenesis through modulating host metabolic responses, particularly lipid metabolism, inflammatory pathways, and oxidative stress responses.

Discussion

In this study, we conducted an integrated analysis of skin microbiome and metabolome profiles in patients with psoriasis and healthy controls, revealing significant dysbiosis in microbial communities and dysregulation of key metabolic pathways. Our correlation analyses uncovered intricate microbe-metabolite interaction networks that may contribute to psoriasis pathogenesis, providing novel insights into the complex interplay between skin microbiota and host metabolism in this chronic inflammatory skin disorder.

Our finding that microbial alpha diversity was decreased in psoriatic lesions is noteworthy, as it appears to contrast with some studies that have reported increased or unchanged diversity in psoriasis. This discrepancy in the literature may be attributable to several factors, including differences in anatomical sampling sites, disease severity, patient populations, and molecular techniques. However, our results are in alignment with other key studies that have also observed less diversity in psoriatic skin [11]. A compelling biological explanation for this decreased diversity is imbalance between dominant microorganisms. This suggests that the health’ of the skin microbiome may be characterized more by its composition rather than by diversity alone.

At the genus level, our findings of decreased Cutibacterium and increased Staphylococcus and Corynebacterium abundances in psoriatic lesions align with previous studies. Cutibacterium species, particularly C. acnes (formerly Propionibacterium acnes), are major components of the healthy skin microbiome and possess important immunomodulatory functions. They produce short-chain fatty acids and bacteriocins that can inhibit the growth of potential pathogens, modulate sebum composition, and maintain an acidic skin environment favorable for commensal over pathogenic bacteria. The significant reduction of Cutibacterium in psoriatic skin may compromise these protective functions, creating an environment more conducive to inflammation and colonization by potentially harmful microbes [31, 32].

In contrast, Staphylococcus species have been implicated in psoriasis pathogenesis through multiple mechanisms. These bacteria produce superantigens that can non-specifically activate T cells, leading to excessive cytokine production and inflammatory responses [33, 34]. Additionally, certain Staphylococcus strains produce δ-toxin, which can induce mast cell degranulation and contribute to skin inflammation. The increased abundance of Staphylococcus observed in our patients might, therefore, directly contribute to the inflammatory milieu characteristic of psoriatic lesions [35].

Corynebacterium is enriched in the lesions of psoriasis [11], where they are associated with expansion of γδT cells (γδT17), which produce IL-1β and IL-17 in the skin, contributing to skin inflammation. Moreover, IL-1β promotes keratinocyte secretion of chemokines and recruitment of peripheral γδT17 cells, amplifying the inflammatory cascade [36]. Interestingly, Corynebacterium tuberculostearicum (C. t.), a ubiquitous skin bacterium, may play a role in skin health and disease by upregulating inflammatory mediators through activation of the canonical nuclear factor-κB pathway [37], although its role in psoriasis remains to be explored.

Our metabolomic analysis revealed profound alterations in the skin metabolome of psoriasis patients, with 63 differential metabolites primarily involved in lipid metabolism, amino acid metabolism, energy metabolism, and inflammatory pathways. The upregulation of phospholipids, ceramides, and sphingomyelins in psoriatic lesions indicates significant disruption of epidermal barrier function. These complex lipids are essential components of the stratum corneum intercellular lipid matrix, which provides barrier protection and regulates skin permeability. Alterations in their composition can compromise barrier integrity, potentially facilitating microbial invasion and enhancing inflammatory responses [38, 39]. Additionally, various amino acids and TCA cycle-related metabolites showed significant alterations, indicating energy metabolism reprogramming potentially related to increased energy demands for cell proliferation and inflammatory responses [40].

The observed elevation of arachidonic acid and its metabolites, including prostaglandins and leukotrienes, aligns with the known inflammatory nature of psoriasis. These eicosanoids are potent pro-inflammatory mediators that can enhance vasodilation, increase vascular permeability, and promote recruitment of inflammatory cells to psoriatic lesions. Their upregulation might contribute to the persistent inflammation characteristic of psoriasis and potentially create a self-perpetuating inflammatory cycle [41].

Notably, our findings of decreased antioxidant metabolites, particularly reduced glutathione, in psoriatic lesions suggest compromised oxidative stress defense mechanisms. Oxidative stress has been implicated in psoriasis pathogenesis [42], as reactive oxygen species can damage cellular components, activate inflammatory signaling pathways, and promote keratinocyte proliferation. The observed reduction in antioxidant capacity might therefore exacerbate oxidative damage and contribute to disease progression.

The most novel aspect of our study is the comprehensive exploration of microbiome-metabolome interactions in psoriasis. Our correlation analyses revealed striking patterns of association between specific bacterial genera and metabolite classes. The strong negative correlations between Cutibacterium abundance and inflammatory lipids, coupled with positive correlations with antioxidant metabolites, suggest potential protective effects of this commensal genus. Cutibacterium species might directly or indirectly influence lipid metabolism [29], modulate inflammatory responses, and contribute to oxidative stress defense in healthy skin. Their depletion in psoriasis might therefore compromise these beneficial functions, contributing to disease pathogenesis.

Conversely, the positive correlations between Staphylococcus/Corynebacterium abundance and inflammatory mediators suggest potential detrimental roles of these genera in psoriasis. These bacteria might actively promote inflammatory responses through direct production of pro-inflammatory molecules or indirect modulation of host metabolic pathways [30]. The observed correlation patterns support a model where microbial dysbiosis, characterized by decreased Cutibacterium and increased Staphylococcus/Corynebacterium, contributes to metabolic dysregulation, particularly affecting lipid metabolism, inflammatory pathways, and oxidative stress responses.

The strong positive correlation between Propionibacteriaceae abundance and glutathione levels is particularly noteworthy, suggesting a role of these bacteria in maintaining skin antioxidant capacity. While the role of oxidative stress in psoriasis is well-documented [42], Propionibacterium is known for its ability to produce propionic acid and free radical oxygenase (RoxP), mitigating oxidative stress, fortifying the skin barrier against external threats, and preventing skin inflammation [43, 44]. This direct quantitative link in the context of human skin disease is, to our knowledge, a novel observation. The mechanism underlying this association is likely multifaceted. Members of the Propionibacteriaceae family, including Cutibacterium species, may contribute to glutathione synthesis or recycling through currently unknown mechanisms. Alternatively, they may produce other antioxidant compounds that protect against oxidative stress. The depletion of these bacteria in psoriasis might therefore contribute to increased oxidative damage, which could exacerbate inflammation and tissue damage.

CCA and contribution analyses provided further insights into the complex microbe-metabolite interactions in psoriasis. The clustering of Cutibacterium with antioxidant-related metabolites and Staphylococcus/Corynebacterium with inflammatory mediators in CCA plots reinforces the distinct functional roles of these bacteria in skin homeostasis and psoriasis pathogenesis. The contribution analysis of inosine, showing positive contributions from Cutibacterium and negative contributions from Staphylococcus/Corynebacterium, suggests specific mechanisms through which microbial changes might influence metabolite alterations.

Collectively, these findings propose a mechanistic framework wherein skin microbial dysbiosis contributes to psoriasis pathogenesis through multiple mechanisms: depletion of beneficial commensals like Cutibacterium impairs their protective functions, including maintenance of lipid barrier integrity, modulation of inflammatory responses, and oxidative stress defense; expansion of potentially harmful bacteria like Staphylococcus promotes inflammation through production of pro-inflammatory molecules and modulation of host metabolic pathways; and altered microbial-host metabolic interactions disrupt normal skin physiology, creating a pro-inflammatory environment conducive to psoriasis development and progression.

Our study has several strengths, including the comprehensive assessment of both microbiome and metabolome profiles, the large sample size compared to previous studies, the careful selection of participants with strict inclusion/exclusion criteria, and the advanced analytical approaches used for integrated multi-omics analysis. However, several limitations should be acknowledged. First, the cross-sectional design precludes determination of causality—whether microbial and metabolic alterations are causes or consequences of psoriasis remains unclear. Second, we analyzed only genus-level microbial composition; strain-level differences might reveal additional important associations. Third, our sampling approach focused on skin surface microbiota and metabolites; analysis of deeper skin layers might provide complementary information. Fourth, while we identified numerous correlations between microbes and metabolites, the functional mechanisms underlying these associations require further investigation.

Future studies should address these limitations through longitudinal designs, strain-level microbial analysis, multi-site sampling, and experimental validation of key microbe-metabolite interactions. Specifically, gnotobiotic mouse models and in vitro skin models could be used to investigate the causal relationships between specific microbial changes and metabolic alterations. Additionally, intervention studies targeting the skin microbiome, such as probiotic applications or prebiotics to promote beneficial commensals, might provide therapeutic insights and further elucidate the role of microbiome-metabolome interactions in psoriasis.

In conclusion, our integrated microbiome-metabolome analysis reveals significant alterations in both skin microbiota and metabolic profiles in patients with psoriasis and uncovers intricate microbe-metabolite interaction networks that may contribute to disease pathogenesis. The findings suggest that microbial dysbiosis, particularly decreased Cutibacterium and increased Staphylococcus/Corynebacterium, may contribute to psoriasis through modulating lipid metabolism, inflammatory pathways, and oxidative stress responses. These insights enhance our understanding of psoriasis pathophysiology and might inform the development of novel microbiome-based therapeutic strategies for this challenging skin disorder.

Authors’ contributions

All authors contributed to the study conception and design. L.Z. and G.Z.C. designed and provided guidance of the study, reviewed, and re-vised the manuscripts. F.W. carried out data analysis and edited the chart. F.W. and X.J. prepared the manuscript. All authors reviewed the manuscript.

Clinical trial number

Not applicable.

Funding

This research was funded by the National Natural Science Foundation of China (grant no. 82370785 and 82172320), Shandong Provincial Natural Science Foundation (ZR2024MH220), and Shandong University Outstanding Young Scholars Program.

Data availability

The dataset analyzed during the current study is available in the National Omics Data Encyclopedia (NODE; https://www.biosino.org/node/index) with the accession number OEP00006243.

Declarations

Ethics approval and consent to participate

The studies involving human participants were in compliance with the Helsinki Declaration. The studies were reviewed and approved by the Institutional Ethics Committee of Shandong Provincial hospital (SWYX: NO. 2024-700). All participants provided written 649 informed consent before enrollment.

Competing interests

The authors declare no competing interests.