Multi-omics characterization of the skin microbiota reveals the anti-aging roles of Stenotrophomonas maltophilia

Danni Guo; Yuanyuan Chen; Yahong Wu; Jingmin Cheng; Yawen Lin; Wenjie Lai; Wentao Ma; Hang Yang; Lianyi Han; Lan Ma; Haidong Jia; Xiao Liu

doi:10.1186/s40168-026-02433-6

. 2026 Jun 3;14:197. doi: 10.1186/s40168-026-02433-6

Multi-omics characterization of the skin microbiota reveals the anti-aging roles of Stenotrophomonas maltophilia

Danni Guo ^1,^#, Yuanyuan Chen ^2,^#, Yahong Wu ^1,^#, Jingmin Cheng ¹, Yawen Lin ¹, Wenjie Lai ³, Wentao Ma ¹, Hang Yang ¹, Lianyi Han ^4,⁵, Lan Ma ¹, Haidong Jia ^2,^✉, Xiao Liu ^1,^✉

PMCID: PMC13449706 PMID: 42237409

Abstract

Background

Shifts in the skin microbiome have shown a close link to chronological age. However, the contribution of the skin microbiome in skin-aging phenotypes remains unclear.

Results

To explore this, we performed phenotypic, metabolomic, metagenomic, and functional analyses on a cohort with divergent skin-aging phenotypes. Genome-scale metabolic models (GEMs) integrated with metabolomic analysis revealed that Stenotrophomonas maltophilia, enriched in the younger group (categorized by AI-predicted age and skin elasticity), utilizes the glutathione cycle to maintain redox homeostasis. Cellular experiments showed its metabolites enhanced GSH synthesis and alleviated oxidative-stress-induced phenotypic skin-aging by upregulating key genes in fibroblasts, including GCLM, PGD, SOD2, and NQO1. In addition, GEMs highlighted its potential in maintaining youthful skin phenotypes through the regulation of host metabolic pathways involving betaine, lysolecithin, and porphyrin. In parallel, Acinetobacter guillouiae was found to influence host melanin metabolism by degrading dopamine (DA) and 3-methoxytyramine (3-MT), offering potential therapeutic strategies for mitigating pigmentation.

Conclusions

Our findings highlight the dynamic interplay between skin microbiota and the host in phenotypic skin-aging, offering new insights for designing interventions to maintain youthful skin.

Download video file^{(43.8MB, mp4)}

Open in a new tab

Video Abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s40168-026-02433-6.

Keywords: Skin aging, Stenotrophomonas maltophilia, Multi-omics, Host-microbiome interactions, Skin microbiome

Background

The skin is the largest organ of the human body [1]. Beyond the gastrointestinal tract, it also harbors the second-most extensive microbial communities in the human host [2]. The human skin microbiome is a complex ecological system comprising bacteria, archaea, fungi, and viruses [3]. This microbiome interacts reciprocally with the host’s skin health, influencing conditions ranging from its general health to specific diseases and its aging state [4–6]. Resident skin microbes help maintain skin homeostasis and can modulate inflammatory responses or induce pathogenicity based on host environmental conditions [7–9].

Skin aging is a multifaceted biological process influenced by a myriad of intrinsic and extrinsic factors. Typically, with age progression, a gradual decline in cellular regeneration occurs, accompanied by diminished DNA repair capacities. Concurrently, there is a deterioration in the fundamental supporting structures of the skin, accelerating the process of skin aging [10]. In addition, environmental determinants, notably prolonged exposure to ultraviolet (UV) radiation, can amplify the generation of reactive oxygen species within cellular structures, thereby causing oxidative stress. This heightened stress damages cell membranes, destabilizing the skin’s integral collagen and elastin fibers, and compromising DNA integrity, consequently accelerating skin cell aging [11].

Physiological alterations brought about by skin aging manifest as increased wrinkling, loss of elasticity, and a decrease in water content in the stratum corneum, and reduced sebum production [12–15]. Skin aging also represents an ecological shift in the skin, impacting its resident microbial community. Several studies have unveiled a complex interplay between skin aging and the skin microbiome, including changes in microbial diversity and composition as age progresses [16–19]. These alterations extend beyond mere changes in microbial types and quantities, encompassing modifications in their metabolic pathways and functions. Importantly, variations within these microbial communities have been closely linked to critical indicators of skin health, including pigmentation, collagen synthesis, sebum production, and the moisturizing capability [20–22]. Specifically, changes in microbial abundance and metabolic activity influence these key physiological processes, thereby accelerating or decelerating the external manifestations of skin aging [23].

With growing research into the intricate dynamics of skin aging and its microbial influences, some studies have proposed that the phenotypes of skin aging and susceptibility might more accurately reflect the characteristics and changes in the skin microbiome than chronological age itself [24–26]. However, most of these studies relied on population-based cohorts with varied groups of chronological ages, in which ages may serve as a confounding factor. There is growing evidence that the age of 40s is considered the watershed of skin aging [27–29]. During this period, collagen production decreases, and menopause accelerates this by lowering estrogen levels. Slower cell turnover and increased oxidative stress make the skin more vulnerable to free-radical damage, speeding up the formation of wrinkles and dryness. Despite this recognized importance, studies specifically focusing on the relationship between microbial communities and skin-aging phenotypes within this pivotal age group remain scarce.

Thus, we conducted a comparative study on 103 healthy women at the age of 39–41 exhibiting divergent conditions of skin aging. Using metagenomic analysis, we identified skin bacterial markers associated with aging conditions. In addition, by constructing GEMs and implementing non-targeted metabolomics, we have elucidated the skin phenotypic aging-associated metabolic networks co-regulated by the skin microbiome and the host. Specifically, we identified and functionally validated the anti-aging role of skin commensal Stenotrophomonas maltophilia. The identified microbial markers and metabolites may help guide strategies to preserve youthful skin.

Results

Distinct features across phenotypic skin aging groups

In this study, 202 healthy Chinese women aged 39 to 41 were enrolled, and we utilized the artificial intelligence platform Megvii’s Face⁺⁺ to classify participants into two age groups [30]: 100 individuals with AI-estimated ages significantly below their chronological age (younger group) and another 102 with AI ages notably above (older group). Key skin traits were assessed across both groups. Using elasticity R2 as a principal metric, we selected the top 52 younger participants with the highest elasticity scores and the 51 older participants with the lowest scores for further analysis. We assess skin physiological traits and examine skin microbiome and metabolites separately using shotgun metagenomics and non-targeted LC–MS (Fig. 1a). Skin phenome analysis reveals that the younger group exhibits significant enhancements in skin elasticity index R2, sebum content, porphyrin optical density, porphyrin area percentage, and stratum corneum hydration, suggesting better moisturization and overall skin health compared to the older cohort (Fig. 1b, Table S1a–h). As anticipated, the older group exhibited more wrinkles. There were also notable differences in skin pigmentation between the groups, with the older group showing a significant reduction in Luminance (L*) and marked increases in redness (a*) and yellowness (b*), suggesting greater pigment accumulation with phenotypic skin aging. In addition, the lack of significant differences in transepidermal water loss (TEWL) and pH levels between younger and older groups is observed (Fig. 1b). Correlation analysis revealed that most skin phenotypes were strongly correlated with AI-estimated age, confirming its efficacy as a reliable indicator of phenotypic skin aging. Moreover, wrinkles appear to be an integrative manifestation of aging, reflecting diminished stratum corneum hydration and sebum levels, alongside increased pigment deposition which may be associated with elevated oxidative stress possibly due to sun damage [31]. These findings underscore the complex interrelationships among multiple indicators of skin aging phenotypes (Fig. 1c).

Fig. 1 — Overview of the study design and comparative skin-aging phenotypes in women of the same age group. a The overall workflow of this study. b Violin plots show the substantial differences in skin phenotypes between the younger and older groups, analyzed using the Wilcoxon test. AI age: estimated age using artificial intelligence; skin elasticity R2: Skin elasticity measured by the R2 parameter; Skin tone: L*: luminance; a*: skin red-green component; b*: skin yellow-blue component; porphyrin area percentage: percentage of skin area covered by porphyrin; WCSC: water content of the stratum corneum; TEWL (g/m²·h): Transepidermal Water Loss. c Heatmap of Spearman’s correlation among skin aging phenotypes. Purple hues indicate negative correlations, with the color intensity corresponding to the strength of the correlation. Significance levels are denoted as follows: *, adjusted P < 0.1; **, adjusted P < 0.05; ***, adjusted P < 0.01

Skin metagenomics reveals microbiome dynamics associated with phenotypic skin aging

From the metagenomic analysis, we identified 2419 microbiome species, with 1,464 species shared across the younger and older groups (Fig. S1a, Table S2a). Among the top 20 species, Cutibacterium acnes (C. acnes) is the most abundant resident skin microorganism, followed by Staphylococcus epidermidis and Staphylococcus hominis (S. hominis) (Fig. 2a). Our data reveal significantly lower Shannon diversity of microbial species in the older group (Fig. 2b, Table S2a, b). Previous studies, primarily focusing on chronological age, suggested that older individuals had higher microbial diversity in their skin [17, 18, 22, 23, 32]. Notably, it has also been reported that a decrease in C. acnes abundance in older individuals might lead to an increase in species diversity [19, 20]. In our samples, which were derived from women within the same age range, C. acnes abundance did not differ significantly between groups (adjusted P = 0.11, Log₂FoldChange = 0.39, Table S2c, d). We hypothesize that the lower microbial diversity observed in younger groups in prior studies might be due to increased C. acnes abundance. To verify our hypothesis, we analyzed a dataset of 742 samples, which previously categorized C-cutotypes as representative of young groups and M-cutotypes of old groups [32]. When removing the influence of C. acnes abundance on the microbial diversity between groups, the data also support a higher microbial diversity in the young group (Fig. S1b, Table S2e). Thus, our results provide additional evidence than those of earlier studies that a higher level of microbial diversity is associated with both greater ecological stability and youthful skin.

β-diversity analysis indicates significant differences in microbial species composition between groups (Fig. 2b). Differential analysis identifies 43 microbial species differentially enriched between groups (Fig. S1c, Table S2d). Notably, S. hominis, which was more prevalent in the older group, exhibits a significant negative correlation with sebum content (Fig. 2c). This observation aligns with the conclusions of previously published studies [33]. The association may reflect bidirectional interactions between host sebum levels and S. hominis abundance. Moreover, potential probiotics and opportunistic pathogens are more prevalent in the younger group (Fig. 2c). For instance, Lactococcus lactis and Janthinobacterium lividum were enriched, known for lipid regulation and antifungal activity [34–38]. Notably, S. maltophilia, though typically opportunistic [39–41], correlates positively with youthful traits such as hydration and elasticity (Fig. 2c), suggesting some pathogens may support skin youthfulness under healthy conditions.

Our findings reveal a phenotypic aging-associated metabolic shift in the skin microbiome (Fig. 2d–g, Table S2 f–h). In the older group, the skin microbiome shows higher enrichment of amino acid and aminoacyl-tRNA biosynthesis pathways (Fig. 2d), suggesting adaptation to nutrient-limited conditions via increased reliance on internal synthesis to support protein and energy needs [42, 43]. Conversely, the skin microbiome in the younger group exhibits enhanced amino acid metabolism (Fig. 2d), particularly in arginine and proline pathways, supporting immune modulation, oxidative stress resistance, and environmental resilience [44–50]. In addition, the older-group microbiome shows higher abundance of antibiotic resistance genes, particularly those related to cationic antimicrobial peptides (CAMP) and antifolate resistance pathways (Fig. 2e), suggesting adaptation to host immune pressures [51] and potentially contributing to Staphylococcus aureus susceptibility in aged skin [52, 53] (Fig. 2e). Conversely, the younger-group microbiome is enriched for pathways involved in the degradation of environmental pollutants, including dioxins [54–56], fluorobenzoates [57, 58], and styrenes [59], and in metabolizing harmful chemicals through pathways such as porphyrin metabolism (Fig. 2e). Moreover, the older-group microbiome exhibits an enrichment in nucleotide synthesis and degradation pathways, such as purine metabolism and folate biosynthesis, suggesting increased potential for DNA and RNA synthesis and repair (Fig. 2f). In contrast, pathways such as cell cycle, bacterial chemotaxis, flagellar assembly, and motor proteins are primarily enriched in younger skin microbiomes, consistent with higher growth-associated and motility-associated functional potential (Fig. 2f). These patterns suggest a shift from growth-oriented activity in younger skin to genomic maintenance in older skin. Furthermore, lipid and structural biosynthesis pathways show phenotypic aging-associated differences (Fig. 2g). The skin microbiome of the younger group favored unsaturated fatty acid and lipopolysaccharide biosynthesis, supporting membrane function, whereas the skin microbiome of the older group was enriched for fatty acid and cell wall biosynthesis, reflecting structural maintenance. To further dissect the microbial origins of these pathway-level differences, we quantify species-level contributions to each KEGG pathway by aggregating the relative contributions of taxa across the pathway’s constituent KOs (Fig. S1d, e, Table S2i, j). This pathway-centric decomposition showed that lipid and structural biosynthesis signals were supported by both group-differential taxa and taxa without significant between-group differences, consistent with coordinated, community-level functional shifts.

Skin metabolomics reveals characteristic metabolic signatures of phenotypic skin aging

Metabolomic profiling of skin revealed distinct metabolic patterns between age groups (Fig. S2a, b). The older group was enriched in metabolites linked to oxidative stress, chronic inflammation, and the accumulation of potentially toxic compounds. In contrast, the younger group shows higher levels of metabolites with moisturizing, anti-inflammatory, and anti-melanogenic functions (Fig. 3a, Table S3a–d).

In older skin, glutathione metabolism and arachidonic acid metabolism are significantly activated, with increased levels of oxidized glutathione and the inflammatory mediator PGD2-d4 (Fig. 3a), indicating a sustained oxidative and inflammatory state [31, 60, 61]. In addition, consistent with previous findings of antibiotic resistance gene enrichment in the aged skin microbiome, we observe elevated levels of antibiotic-associated metabolites such as rosamicin and trimethoprim (Fig. 3a), likely reflecting microbial imbalance. The accumulation of fungal toxins such as deoxynivalenol may further impair lipid synthesis and membrane integrity, compromising skin barrier function [62] (Fig. 3a). These combined alterations may contribute to phenotypic skin aging.

In pigmentation-related metabolism, the older group shows elevated levels of 3-MT (Fig. 3a), a product of tyrosine metabolism and implicated in melanin biosynthesis. Skin phenotype correlation analysis indicates its positive correlation with skin tone parameters a* and b*, and negative correlation with luminance (L*) (Fig. 3b). In contrast, the younger group is enriched in (R)-ar-turmerone and 3-hydroxycoumarin (Fig. 3a), known tyrosinase inhibitors with photoprotective potential, which may serve as functional ingredients for pigmentation control [63–65].

In addition, we identified two metabolites not previously associated with skin function but significantly enriched in the younger group. N-linoleoyl leucine, formed through an amide bond between linoleic acid and leucine, is notably elevated (Fig. 3a). As linoleic acid plays a key role in immune regulation, anti-inflammation, and maintaining skin barrier integrity [66–68], its derivative may similarly support skin protection through multiple pathways and represents a promising candidate for further investigation. We also observe enrichment of 2-hydroxy-5-(trifluoromethoxy)benzoic acid (Fig. 3a), a compound structurally similar to β-LHA (2-hydroxy-5-octanoylbenzoic acid), which is widely used for skin moisturization and anti-inflammatory applications [69–71]. While both share hydroxy and carboxylic acid groups, the trifluoromethoxy substitution in this compound enhances lipophilicity, chemical stability, and dermal permeability, potentially enabling deeper skin penetration and targeted anti-inflammatory, antibacterial, or antioxidant activity.

Predictive modeling and construction of genome-scale metabolic models (GEMs)

To assess whether skin commensals or metabolites can classify premature or delayed phenotypic skin aging, we built models using differential microbiome, metabolite, or combined features. Key features were selected using MeanDecreaseAccuracy, yielding 46, 205, and 133 features for the microbiome, metabolome, and combined models, respectively (Table S3e). The metabolomics model achieves strong predictive performance (AUC = 0.89), underscoring the importance of metabolic pathways in phenotypic skin aging (Fig. 3c, Table S3f). In contrast, the microbiome model shows limited performance (AUC = 0.72; specificity = 93.3%) (Fig. 3c), suggesting that microbial data alone may not capture the complexity of the phenotypic aging process. Combining microbiome and metabolomic features improves prediction (AUC = 0.90) (Fig. 3c), likely by capturing cross-domain interactions. To further explore this interplay, we constructed GEMs to investigate the functional roles of microbially derived metabolites in phenotypic skin aging.

We constructed GEMs for 56 strain-level metagenome-assembled genomes (MAGs) with ANI > 97.5%, covering 50 species (Fig. S3a, Table S4a). Phylogenetic trees were constructed for different strains within the same species and were compared against previously annotated strains (Fig. S3b). These MAGs accounted for ~ 78% of total clean reads with no significant differences between age groups (Fig. S3c, Table S4b), representing a substantial portion of the skin microbiome. MAG analysis revealed higher microbial diversity in the younger group, with S. maltophilia enriched and associated with youthful skin traits, while S. hominis was more abundant in the older group and negatively correlated with sebum production (Fig. S3c, d, Table S4c–e). These patterns are consistent with previous findings [33].

GEMs were reconstructed for each MAG, comprising on average 2028 reactions and 1727 metabolites (Table S5a). Flux Balance Analysis (FBA), conducted to test model reliability, demonstrated good stoichiometric and flux consistency, with no abnormally high ATP production under either aerobic or anaerobic conditions, confirming the models’ reliability (Fig. S4a). Further, we weighted different MAGs by their relative abundance in samples and combined the GEMs of different samples from both older and younger groups to create a group-wide GEM. This combined model contained 4739 reactions and 3467 metabolites, with a flux consistency of 0.99 and stoichiometric consistency of 0.58, further confirming the reliability of the models (Table S5a). These validated models were subsequently used to investigate the functional metabolic capacities of specific microbes and explore metabolic differences associated with phenotypic skin aging.

S. maltophilia mitigates oxidative stress and supports youthful skin phenotypes

Oxidative stress is a pivotal factor in phenotypic skin aging [31]. Free radicals and reactive oxygen species (ROS) damage skin cells, leading to the breakdown of collagen and elastin, thereby accelerating aging [72]. Previous studies have shown that glutathione (GSH), crucial for the cellular antioxidant defense system, neutralizes these free radicals and peroxides using the ascorbate-glutathione cycle to maintain cellular redox balance, converting into oxidized glutathione (GSSG), which protects cells from oxidative harm [73, 74] (Fig. S4c, d).

Building on the findings of our study, which identifies a high accumulation of GSSG in the skin of older individuals (Fig. S4b), we focused on exploring the oxidative and reductive reactions of GSH within the skin microbiome. FBA analysis of the GEM reveals that the flux of GSH redox reactions was − 2.36 in the younger group and − 1.09 in the older group (Table S5b, c). These negative values indicate ongoing reduction processes of GSSG to GSH. The lower absolute value observed in the older group suggests reduced flux, consistent with higher levels of GSSG in the skin metabolism of the older group. This pattern implies that older individuals’ skin may undergo greater oxidative stress, as GSH is consumed to neutralize ROS, thereby increasing GSSG production. A larger absolute value of flux in the younger group also indicates a faster rate of GSSG reduction to GSH, suggesting that the skin microbiome of younger individuals may contribute to a stronger antioxidant recovery capacity and a more effective oxidative stress response, helping to maintain the balance of GSH/GSSG in the skin.

To identify key microbes and metabolites involved in the glutathione redox reaction, we applied shadow price analysis based on our constructed GEMs (Table S5d). Interestingly, S. maltophilia, which was more prevalent in the younger phenotypes, was predicted to utilize the glutathione cycle for antioxidant defense and maintenance of protein thiol homeostasis (Fig. 4a, b). Consistent with this, S. maltophilia was detected in 97.1% of samples (100 out of 103) with a mean relative abundance of 0.97%, ranking 14th among 2419 microbial species (top 0.58%), indicating that it constitutes a relatively abundant and stable member of the facial skin microbiome. It has been reported that S. maltophilia employs a variety of enzyme systems, including glutathione peroxidase, to combat oxidative stress from hydrogen peroxide, a type of ROS [75–77]. To determine whether the antioxidant defense capacity of S. maltophilia can mitigate cellular senescence in skin fibroblasts, skin primary fibroblasts were treated with H₂O₂ to simulate the oxidative stress which is typical of aged skin, followed by treatment with S. maltophilia culture supernatant (Fig. 4c). Results indicated that fibroblasts treated with S. maltophilia supernatant exhibited significantly increased levels of GSH and decreased levels of GSSG compared to the untreated control group. These effects are particularly evident in the H₂O₂-treated groups, where treatment with S. maltophilia supernatant leads to a markedly higher GSH/GSSG ratio (Fig. 4d, Table S5e). Given that senescent cells exhibit elevated β-galactosidase activity [78], analysis using SA-β-gal staining shows that fibroblasts treated with S. maltophilia supernatant displayed reduced senescence markers (Fig. 4d, Fig. S4e). This confirms the role of S. maltophilia in mitigating oxidative stress-induced cellular senescence.

Fig. 4 — *S. maltophilia* in antioxidant defense. a Shadow price analysis from FBA in glutathione redox reactions across different MAGs. This plot illustrates the shadow prices from FBA related to glutathione oxidation-reduction reactions within the GEMs. The x-axis represents shadow price values, while the y-axis lists different metabolites. Positive values indicate that an increase in the metabolite concentration promotes the forward direction of the metabolic reaction, and vice versa. b Antioxidant defense network and flux analysis in *S. maltophilia* GEM: the figure presents a network diagram of the glutathione (GSH)-dependent antioxidant defense mechanisms within the GEM of *S. maltophilia*. The diagram includes flux values depicting the dynamic metabolic interactions involving GSH, where it participates in reducing oxidized glutathione (GSSG) and combating oxidative stress via various enzymatic pathways, such as glutaredoxin. The fluxes are visualized in a color gradient representing their relative intensities, indicating the metabolic activity levels across different pathways. c Experimental workflow for investigating the effect of *S. maltophilia* supernatant on skin primary fibroblasts. *S. maltophilia* was first cultured on plates, followed by 48-h liquid culture. The bacterial culture was filtered through a 0.22-µm filter, and the resulting supernatant was used to stimulate skin primary fibroblasts, both with and without H₂O₂ treatment, for 24 h. The resulting samples were analyzed for GSH and GSSG levels, subjected to senescence-associated β-galactosidase (SA-β-Gal) staining for cellular senescence, and processed for RNA-seq analysis to assess gene expression changes. d Violin plots depicting the distribution of GSH, GSSG, GSH/GSSG ratio and the percentage of cells exhibiting SA-β-Gal activity in skin primary fibroblasts under different treatment conditions. Groups include control, control with *S. maltophilia*, H₂O₂ alone, and H₂O₂ with *S. maltophilia*. P values were calculated using T-test. Significance levels are indicated as follows: *, P < 0.05; **, P < 0.01; ***, P < 0.001. e Representative hematoxylin and eosin (H&E) staining of full-thickness human skin equivalents (T-Skin; EPISKIN) under the indicated conditions (control, H₂O₂ alone, and H₂O₂ with *S. maltophilia*). f Immunofluorescence staining of cytokeratin 10 (K10/KRT10; red) in skin equivalents from different treatment groups, with nuclei counterstained by DAPI (blue). g Heatmap of differentially expressed genes between H₂O₂ alone, and H₂O₂ treatment combined with *S. maltophilia* culture supernatant. The left panel shows the Z-score normalized expression levels of the differentially expressed genes. The right panel displays the KEGG pathway enrichment analysis results for these genes (adjusted P < 0.1). Bar lengths represent the significance of enrichment, expressed as -log₁₀(FDR), and are color-coded according to the treatment group

To evaluate whether the anti-senescence effects of S. maltophilia could be mediated indirectly via epidermal signaling, we established a keratinocyte-conditioned medium stimulation in primary skin fibroblasts. The results show that the medium stimulation could partially alleviate H₂O₂-induced senescence-associated changes, as indicated by an increased GSH/GSSG ratio and a reduced percentage of SA-β-gal-positive cells (Fig. S4f, g, Table S5e). Consistently, DCFH-DA staining reveals lower intracellular ROS levels in the treatment groups compared with the control group (Fig. S4h). We further validated these effects in a more physiologically relevant full-thickness human skin equivalent that preserves skin architecture and barrier context. Following H₂O₂-induced oxidative injury, S. maltophilia culture supernatant was topically applied to the epidermal surface. H&E staining shows that H₂O₂ exposure induces pronounced histological alterations, including a looser, vacuolated dermal matrix and a trend toward reduced epidermal thickness, whereas S. maltophilia treatment mitigates these changes (Fig. 4e). In parallel, immunofluorescence staining of the epidermal differentiation marker K10 indicates a recovery trend under S. maltophilia treatment in the H₂O₂-injured background (Fig. 4f). Together, these findings support that S. maltophilia-related interventions can modulate oxidative stress and senescence-associated phenotypes, potentially through keratinocyte-mediated signaling in a tissue-like context.

To explore how S. maltophilia mitigates cellular senescence, we performed RNA-seq on H₂O₂-treated cells (Table S5f). Glutathione metabolism was significantly enriched in the S. maltophilia conditioned group (Fig. 4g, Table S5g, h), with upregulation of GCLM, enhancing GSH synthesis via glutamate-cysteine ligase activity [79]. PGD was also elevated, promoting NADPH supply for GSH regeneration [80] (Fig. 4g, Fig. S4f). These changes supported oxidative stress defense, maintained fibroblast viability, and delayed senescence, confirming that S. maltophilia alleviates oxidative stress by promoting GSH metabolism. Antioxidant genes SOD2, NQO1, TXN, and FTH1 were significantly upregulated in the S. maltophilia group [81–84] (Fig. 4g). Notably, TXN synergizes with GSH to maintain protein thiol redox balance, consistent with shadow price analysis indicating that S. maltophilia supports thiol homeostasis (Fig. 4a, Fig. S4c). Meanwhile, the senescence-associated genes SERPINE1, CCND1, and GADD45B were downregulated [85] (Fig. 4g, Fig. S4i). In contrast, pathways related to cell growth, regeneration, and tissue repair were enriched (Fig. 4g, Fig. S4j). Notably, high expression of JUN, APCDD1, and SFRP2 activated Wnt signaling [86], while downregulation of NCOR2 and NOTCH3, and p53 signaling further promoted proliferation [85] (Fig. 4g). Additionally, suppression of ECM remodeling and cytoskeletal pathways limited excessive matrix deposition, supporting tissue elasticity and flexibility [87] (Fig. 4g, Fig. S4j, Table S5i).

Collectively, the functional assays, organotypic validation, and transcriptomic profiling are consistent with an antioxidative, anti-senescence signature associated with S. maltophilia in skin models. Although S. maltophilia is considered an opportunistic pathogen [39], it is significantly associated with multiple youthful skin phenotypes in our cohort (Fig. S3d), suggesting a context-dependent commensal role that may contribute to antioxidative capacity and attenuation of senescence-associated phenotypes under healthy skin conditions.

Potential roles of S. maltophilia in regulating betaine, lysolecithin, and porphyrin metabolism associated with phenotypic skin aging

In addition to its role in alleviating oxidative stress-induced skin senescence, S. maltophilia appears to participate in several key metabolic pathways associated with skin homeostasis. These include the biosynthesis of betaine and lysolecithin, which are known for their skin-moisturizing, photoprotective, and anti-inflammatory properties, as well as its contributions to porphyrin metabolism, a process crucial for maintaining skin redox balance and overall skin health (Fig. 5, Fig. S5).

Fig. 5 — Betaine and porphyrin biosynthesis pathways in skin microbiome and host co-metabolism. a FBA Shadow price analysis for betaine synthesis reactions across different MAGs. It highlights the biochemical value and metabolic impact of shifts in betaine-related compounds within microbial metabolism. The plot follows the same format as Fig. 4a, with the x-axis showing shadow price values and the y-axis listing metabolites. b Flux distribution diagram from the *S. maltophilia* GEM illustrating the betaine synthesis pathway. This diagram shows metabolic conversions, including the transformation of betaine aldehyde to betaine. The fluxes are visualized in a color gradient representing their relative intensities, indicating the metabolic activity levels across different pathways. c Violin plots illustrate significant differences in the concentrations of key porphyrin metabolism-related metabolites, such as precorrin-1 and S-Adenosyl-L-methionine, between older and younger groups. Notable differences are indicated, P < 0.05. d FBA Shadow price analysis for the conversion of Uroporphyrinogen III to Precorrin 2 in porphyrin metabolism. The plot follows the same format as Fig. 4a, with the x-axis showing shadow price values and the y-axis listing metabolites. e This figure presents the GEM flux analysis of porphyrin metabolism pathways facilitated by *S. maltophilia*. It depicts how the conversion of Uroporphyrinogen III involves complex interactions leading to the synthesis of siroheme, among other compounds, with flux values shown in a color gradient to represent the intensity of metabolic activities

Betaine and lysolecithin are known for their skin-moisturizing, photoprotective, and anti-inflammatory properties [88, 89]. The biosynthetic pathways of betaine and lysolecithin have been reported in earlier studies [90, 91] (Fig. S5a). Using GEMs, shadow price analysis of betaine and lysolecithin biosynthetic pathways identified S. maltophilia as a major microbial contributor involved in their production in the skin (Fig. 5a, Fig. S5b, Tables S6a, b). Furthermore, existing research suggests that lysolecithin may be a target or by-product of enzymatic activity in S. maltophilia [92, 93]. Specifically, elevated shadow prices for precursors such as choline and betaine aldehyde in S. maltophilia suggest that increased availability of these substrates could enhance betaine production (Fig. 5a). Similarly, accumulation of intermediates in the acyl-CoA and fatty acid β-oxidation pathways may promote lysolecithin synthesis (Fig. S5b, c). Exploring strategies to enhance the production of betaine and lysolecithin by S. maltophilia in the skin presents a potential direction for delaying phenotypic skin aging.

Moreover, our previous analyses indicated that the younger group exhibits higher porphyrin optical density and porphyrin area percentage, along with a high enrichment of microbes involved in porphyrin metabolism (Figs. 1b and 2e). The porphyrin metabolic pathway, in which uroporphyrinogen III is converted into precorrin-2 and subsequently processed to form cobalamin (vitamin B12), has been well characterized in previous studies [94, 95] (Fig. S5d). Through skin metabolomics analysis, we identified several metabolites involved in the porphyrin metabolic pathway (Fig. 5c, Fig. S5e). Shadow price analysis performed on GEMs for the reaction converting Uroporphyrinogen III and Precorrin 2 revealed microbes associated with porphyrin metabolism (Fig. 5d). A key bacterium affecting this process is C. acnes, which has been reported to undergo transcriptional and metabolic changes in healthy individuals upon oral vitamin B12 supplementation [96] (Fig. 5d, Table S6c). These changes include enhanced porphyrin production by C. acnes [96]. Elevated porphyrin accumulation could increase oxidative stress and pigment abnormalities in the skin [97]. However, adequate levels of vitamin B12 facilitate rapid skin cell renewal and repair [98], and provide antioxidant and photoprotective benefits [99]. Our data also underscore the significant role of S. maltophilia in the conversion of Uroporphyrinogen III to Precorrin 2 (Fig. 5d, e). Overall, the porphyrin-cobalamin module identified here suggests that microbial pathway routing may influence whether flux is directed toward vitamin B12 biosynthesis or toward the accumulation of intermediate porphyrins, providing a potentially actionable hypothesis for skin anti-aging research while acknowledging that causality remains to be established.

Skin microbiota involvement in melanin metabolism during phenotypic skin aging

The melanin synthesis pathway has been previously reported [100–102]. In this context, we delineated a co-metabolic pathway of tyrosine metabolism shared between the skin microbiome and the host that could contribute to melanin metabolism within the skin (Fig. 6a). Our data showed that skin microorganisms in the younger group were significantly enriched in the tyrosine metabolism pathway (Fig. 2d). Utilizing established knowledge, the melanin biosynthesis process principally involves the conversion of tyrosine to L-DOPA, followed by the oxidation of L-DOPA to dopaquinone, which is then polymerized into melanin (Fig. 6a). Although no significant differences were observed in the levels of L-tyrosine, dopaquinone, or melanin within the skin metabolome, notable enrichment of L-DOPA, dopamine and 3-MT in the older group was detected (Fig. 6b, Fig. S6a). Additionally, 3-MT exhibited a significant positive correlation with skin tone a* and b*, and a significant negative correlation with L* (Fig. 3b).

Fig. 6 — Tyrosine metabolism within the co-metabolic framework of skin microbiome and host. a Tyrosine metabolism in skin microbiome and host co-metabolism. This figure illustrates the shared metabolic pathway of tyrosine between the microbiome and host, highlighting its role in producing key metabolites and melanin. Left Module (Tyrosine to L-3,4-dihydroxyphenylalanine (L-DOPA)): tyrosine is hydroxylated to L-DOPA, a precursor for melanin and catecholamines, catalyzed by tyrosinase. L-DOPA is then oxidized to dopaquinone, which cyclizes into dopachrome and rearranges into melanin. Right module: L-DOPA is decarboxylated into dopamine (DA), which is metabolized into 3,4-Dihydroxyphenylacetaldehyde (DHPAA), 3,4-Dihydroxyphenylacetate (DOPAC), and Homovanillic acid (HVA). DA also forms 3-Methoxytyramine (3-MT), further processed into 3-Methoxy-4-hydroxyphenylacetaldehyde (HMA) and HVA. The diagram uses green dashed outlines to indicate skin metabolites that show no significant differences between groups, highlighting their consistent presence across different phenotypic aging conditions. Green solid outlines denote metabolites with higher abundance in the older group, indicating changes that may be associated with phenotypic skin aging. Blue outlines represent the reactions included in the GEM, emphasizing the biochemical interactions facilitated by the skin microbiome and host. b Violin plots depicting significant differences in key metabolites involved in tyrosine metabolism: dopamine (DA), L-DOPA, and 3-Methoxytyramine (3-MT) between older and younger groups in skin metabolomics. Notable differences are indicated, P < 0.05. c Tyrosine metabolism network and flux analysis in *K. aerogenes* GEM: this figure illustrates the pathways involved in tyrosine metabolism within the GEM of *K. aerogenes*. The network shows the conversion of DHPAA to DOPAC. Flux values are indicated by a color gradient, representing the intensity of each reaction within the metabolic pathways. d Tyrosine metabolism network and flux analysis in *A. guillouiae* GEM: this figure illustrates the pathways involved in tyrosine metabolism within the GEM of *A. guillouiae*. The network shows the conversion of HMA to HVA. Flux values are indicated by a color gradient, representing the intensity of each reaction within the metabolic pathway

GEM analysis demonstrated that the intermediate metabolic processes involving DA and 3-MT are sensitive to microbial reactions, particularly in the conversion pathways from DHPAA to DOPAC and from HMA to HVA (Fig. 6a, Fig. S6b,c, Tables S6d, e). Shadow-price analysis highlighted DA and 3-MT as key intermediates within these microbiome-involved metabolic routes. A key microbe identified in this process is Klebsiella aerogenes (K. aerogenes) (Fig. 6c, Fig. S6b, c). It has been reported to induce melanin production in fungi by providing melanin precursors [103], thus acting as a microbial source of melanin substrates. Interestingly, we also identified A. guillouiae, a microorganism enriched in the younger cohort, as a participant in this metabolic pathway, suggesting its potential capability in metabolizing melanin precursors (Fig. 6d, Figs. S3d and S6b, c). Accordingly, we view the microbiome as a parallel and actionable regulatory layer that may, depending on redox state and substrate availability, modulate DA/3-MT routing and potentially influence ROS-driven melanogenesis.

Discussion

The relationship between skin aging and microbiome dynamics has been increasingly studied. Skin aging involves structural and physiological changes that correlate with shifts in microbial composition [22, 24, 32]. While prior studies focused on age-based population differences, our study targets individuals in their 40s, a critical watershed for skin aging, to investigate why some display significantly younger or older skin phenotypes. We aim to uncover microbial regulatory mechanisms underlying this variation and identify strategies to maintain youthful skin appearance through microbiome modulation.

Using a multi-omics approach integrating phenomics, metabolomics, metagenomics, and genome-scale metabolic modeling, we investigated phenotypic skin aging within the same age group. Phenomics confirmed reliable stratification into younger and older phenotypes [12, 13]. Notably, we found that younger skin phenotypes exhibited higher microbial diversity. This contrasts with earlier studies that reported lower diversity in youth, which we found was primarily due to the overwhelming abundance of C. acnes masking community richness [17, 18, 22, 23, 32]. Our findings suggest that enhanced microbial diversity supports skin homeostasis and a youthful appearance.

Our study also reveals that the skin microbiome shows phenotypic aging-associated differences in functional potential, with older-associated communities enriched for biosynthesis and genomically consistent with nutrient-limited, stressed conditions [104], while younger-associated microbiota are enriched for immune modulation and tissue renewal [51, 105, 106]. These microbial shifts co-occurred with metabolite differences: younger phenotypes are enriched in beneficial compounds with antioxidant and anti-inflammatory properties, such as methylcellulose [107], alnustone [108], and theobromine [109], whereas pro-aging toxins like deoxynivalenol [62] are more abundant in older phenotypes. Predictive modeling revealed that combining microbiome and metabolomic data enhanced phenotypic aging prediction, highlighting microbiome-metabolite interactions. This prompted GEM-based exploration of key microbial metabolites in phenotypic skin aging.

Although S. maltophilia is typically regarded as an opportunistic pathogen [39], our findings suggest that under commensal skin conditions it may contribute to redox homeostasis and skin health. This protective signature was further supported in a full-thickness human skin equivalent. GEM modeling and experimental validation showed that it enhances GSH synthesis and upregulates antioxidant genes such as GCLM and SOD2, improving fibroblast resistance to oxidative stress, a key driver of phenotypic skin aging [72–74]. Beyond redox regulation, S. maltophilia may support barrier integrity and anti-inflammatory states by producing beneficial metabolites such as betaine and lysolecithin [88, 89]. GEM further predicted contributions to cobalamin (vitamin B12) biosynthesis via porphyrin metabolism, which could help limit oxidative damage. These findings suggest a protective role for S. maltophilia in phenotypic skin aging, warranting further investigation of its dynamic functions in the skin microenvironment. Our study also suggests that K. aerogenes and A. guillouiae may participate in pigmentation-associated metabolic networks. The functional relevance of these associations remains to be validated.

Despite the strengths of our study, several limitations should be acknowledged. Although we controlled for age range and geographic background, we did not collect detailed information on individual sun exposure, skincare practices, vitamin B12 intake, hormonal status, or lifestyle variables such as sleep and stress, all of which are known to influence both skin physiology and the microbiome. Dermal collagen quality should also be assessed to establish a direct link between microbiome-metabolite features and collagen degradation. In addition, we did not pinpoint the specific S. maltophilia-derived metabolites that mediate the anti-senescence-associated effects under skin barrier conditions.

In summary, our integrative multi-omics analysis provides critical insights into the intricate crosstalk between the skin microbiome and metabolites during the phenotypic skin-aging process, revealing how microbial-host interactions shape a youthful skin ecosystem. S. maltophilia supports antioxidant defense and hydration, and A. guillouiae is implicated by GEM mapping in pigmentation-associated metabolism. Together, these findings highlight the therapeutic promise of targeting microbiome-metabolite crosstalk to preserve youthful skin function and appearance, especially around the transitional age of the 40s.

Materials and methods

Participants and study design

We initially recruited 202 healthy Chinese women aged 39 to 41 from the general population in Shanghai in August 2022. This specific age range was chosen to capture a well-defined transitional window, as there is growing evidence that the 40s represent a critical inflection point in skin aging [27–29], during which biological and structural changes become more pronounced. To further minimize confounding factors that could influence the skin microbiome, all participants were recruited from the same geographic region and were expected to share similar genetic backgrounds, dietary habits, and lifestyle patterns. Individuals with a history of skin diseases, recent anti-aging treatments, or antibiotic use within the past month were excluded. To maximize microbial skin load, all participants were instructed to avoid facial cleansing or the use of any skincare or cosmetic products within 12 h prior to sampling. Specifically, subjects were asked to perform their final facial wash using only tap water before 8:00 p.m. on the evening prior to sampling and to refrain from washing their face the following morning.

Ethics approval for this study was obtained from the Ethics Committee of Tsinghua Shenzhen International Graduate School (Approval No. [2022] 97). Written informed consent was obtained from all participants before skin microbiome sampling per institutional guidelines.

Skin phenotype assessment

We evaluated skin characteristics in 202 healthy women, focusing on ecological niche indicators and appearance. Traits measured included transepidermal water loss (TEWL), stratum corneum water content (WCSC), sebum production, surface pH, elasticity, and porphyrins (area percentage, average area, and optical density), as well as wrinkles and skin tone (L*, a*, b*). TEWL, reflecting skin barrier function, was gauged using a Vapometer® (Delfin Technologies Ltd). WCSC was determined via a Corneometer® CM 825 (EnviroDerm Services, UK), and sebum output was measured with a Sebumeter® SM815 (Courage & Khazaka electronic GmbH), noted in μg/cm². Skin pH was assessed using a Skin-pH-Meter PH 900 (Courage & Khazaka electronic GmbH). Elasticity, indicated by the R2 parameter (ratio of retraction to deformation), was analyzed with a Cutometer® dual MPA 580 (Courage & Khazaka electronic GmbH). Skin tone, porphyrin metrics, and porphyrin distribution were analyzed using ImageJ, based on VISIA-CR images (Canfield Scientific Inc). Wrinkle assessment was conducted by dermatologists. They examined various types including forehead lines, crow’s feet, glabellar lines, under-eye wrinkles, and nasolabial folds, with a composite score determining overall wrinkle severity.

Grouping criteria and sample collection procedures

To stratify participants based on phenotypic skin aging, we applied an artificial intelligence-based facial analysis platform (Megvii Face + +) [30], which estimated each participant’s apparent (AI-estimated) age. Based on this assessment, 100 women whose AI-estimated ages were significantly lower than their chronological ages were classified as the younger group, and 102 whose AI-estimated ages were significantly higher were classified as the older group. In addition to AI-based grouping, we used R2 (gross elasticity) as a complementary criterion to identify individuals with distinct biological aging profiles, as it reflects dermal properties that are not fully captured by AI-estimated age based on facial images (Megvii Face + + [30]). For downstream analysis, we selected a subset of 103 participants, including the 52 individuals from the younger group with the highest R2 scores and the 51 from the older group with the lowest R2 scores.

Skin microbiome samples were collected from the cheeks of participants in a temperature-controlled and humidity-controlled clinical setting (20 °C and 50% relative humidity). Study personnel wore sterile gloves for each sample collection to minimize environmental or handling contamination. To monitor potential contamination introduced during sampling and handling, field blank swabs were collected alongside the study samples: a sterile swab was opened and handled in the same room and for a similar duration as sample collection, but without contacting the skin or any surface, and was then placed into an identical collection tube and processed in parallel. The left cheek was sampled for metagenomic sequencing, and the right cheek for metabolomic analysis. Sterile medical cotton swabs pre-moistened with 0.15 M NaCl solution were used to swab each area 25 times, utilizing a systematic vertical and horizontal motion for thorough sampling. The swab heads were then detached and stored in sterilized 1.5 mL centrifuge tubes at − 80 °C.

DNA extraction and metagenomic sequencing

The experimental workflow encompassed several key stages: DNA extraction, DNA quality assessment, library preparation, library quantification, and sequencing. DNA extraction was carried out using the Novizan FastPure® Host Removal and Microbiome DNA Isolation Kit (item No. DC501), adhering strictly to the manufacturer’s protocol. This kit is specifically designed to efficiently lyse host cells and deplete host genomic DNA, while preserving microbial DNA from skin swabs. DNA concentration was determined using the Qubit fluorometer. Subsequent library preparation was conducted utilizing the Novizan VAHTS® Universal Plus DNA Library Prep Kit for MGI (item No. NDM617). Post-preparation, libraries were initially quantified with the Qubit fluorometer and then diluted to a concentration of approximately 5 ng/µl to optimize sequencing conditions. Each step underwent rigorous quality control to ensure reliability, and sequencing was performed on an MGI platform using the PE150 strategy to a depth of approximately 12 Gbp per sample.

Preparation of metabolomics samples and LC–MS measurement

All chemicals and solvents utilized were of analytical or HPLC grade. Water, methanol, acetonitrile, and formic acid were acquired from Thermo Fisher Scientific (Waltham, MA, USA). L-2-chlorophenylalanine was sourced from Shanghai Hengchuang Bio-technology Co., Ltd. (Shanghai, China), and chloroform was obtained from Titan Chemical Reagent Co., Ltd. (Shanghai, China). Each sample was treated with a total of 2 mL of pre-chilled methanol–water mixture (4:1, v/v), which was transferred in two portions to glass vials. The samples were then subjected to ultrasonication for 20 min in an ice-water bath. After extraction, samples were allowed to stand at −40 °C for 30 min before being centrifuged at 13,000 rpm for 10 min at 4 °C. One milliliter of the supernatant was then transferred to an LC–MS sample vial and dried. For reconstitution, 300 µL of methanol–water mixture (1:4, v/v) containing 2 µg/mL of L-2-chlorophenylalanine was added to each sample. Samples were vortexed for 30 s, ultrasonicated for 3 min in an ice-water bath, and subsequently rested at − 40 °C for 2 h. Post-resting, samples were centrifuged at 4 °C (13,000 rpm) for 10 min. A 150-µL aliquot of the supernatant was drawn with a crystal syringe, filtered through a 0.22-µm organic phase syringe filter, and transferred into LC vials. The vials were stored at − 80 °C until LC–MS analysis. Quality control (QC) samples were prepared by pooling equal volumes of the extracts from all samples. All reagents used in the extraction process were pre-chilled to − 20 °C prior to use.

Metabolomic profiling was performed using an ACQUITY UPLC I-Class system coupled with a VION IMS QTOF Mass Spectrometer (Waters Corporation, Milford, USA). The analyses were conducted in both ESI-positive and ESI-negative ion modes to enhance the detection spectrum. Chromatographic separation was achieved on an ACQUITY UPLC HSS T3 column (100 mm × 2.1 mm, 1.8 µm). The mobile phases consisted of water (A) and a mixture of acetonitrile/methanol (2:3 v/v) (B), both containing 0.1% formic acid. A linear gradient was employed as follows: 0 min, 5% B; 2 min, 5% B; 4 min, 30% B; 8 min, 50% B; 10 min, 80%B; 14 min, 100% B; 15 min, 100% B; 15.1 min, 5% B and 16 min, 5%B. The flow rate was maintained at 0.35 mL/min, and the column temperature was set at 45 °C. Sample injection volume was 2 µL, and samples were kept at 4 °C during the analysis.

Mass spectrometry was performed in full-scan mode, covering a mass-to-charge (m/z) range of 100 to 1200. Data acquisition was supplemented by MSE mode, where two independent scans with varying collision energies were alternated: a low-energy scan at 4 eV and a high-energy scan with a ramp from 20 to 45 eV for ion fragmentation. The spectrometer was operated with the following parameters: resolution (full-scan): 70,000, resolution (HCD MS/MS scans): 17,500, spray voltage: 3.8 kV for positive mode and − 3.0 kV for negative mode, sheath-gas flow rate: 35 Arb, aux-gas flow rate: 8 Arb, capillary temperature: 320 °C, argon (99.999%) as the collision gas, a scan time of 0.2 s, an interscan delay of 0.02 s, a capillary voltage of 2.5 kV, a cone voltage of 40 V, a source temperature of 115 °C, a desolvation-gas temperature of 450 °C, and a desolvation-gas flow of 900 L/h. Quality control samples were injected at regular intervals, after every 10 samples, to assess the repeatability and stability of the analytical conditions throughout the run.

LC–MS data processing and metabolomics analysis

LC–MS raw data were processed using Progenesis QI v2.3 software (Nonlinear Dynamics, Newcastle, UK). The analysis involved baseline filtering, peak identification, integration, retention time correction, peak alignment, and normalization. Normalization of metabolite intensities was conducted using the total-ion-current (TIC) method to correct for variations in signal intensities across samples. Peak detection and alignment were performed using the default apex-peak-picking algorithm in Progenesis QI. The analysis settings included a precursor ion mass tolerance of 5 ppm, a product ion mass tolerance of 10 ppm, and a minimum product ion threshold of 5%. Compound identification was performed based on accurate mass-to-charge ratio (m/z), secondary fragmentation patterns, and isotopic distributions. The identification databases utilized included the Human Metabolome Database (HMDB) [110], LipidMaps v2.3 [111], METLIN [112], EMDB [113], PMDB [114], and a proprietary database developed by Shanghai OE Biotech Co., Ltd., facilitating comprehensive qualitative analysis. Data refinement involved the exclusion of any peaks that exhibited a missing value (ion intensity = 0) in more than 50% of the samples within any group. To ensure analytical reproducibility, features with a relative standard deviation (RSD) > 30% in pooled QC samples were removed from subsequent analyses. Zero values were imputed using half of the minimum detected value to maintain data integrity. Compounds were screened based on their qualitative identification scores, and those scoring below 36 out of a possible 60 were excluded from further analysis to ensure data accuracy. Finally, a consolidated data matrix was created by integrating the positive-ionization and negative-ionization data results. In addition, data were analyzed using MetOrigin to identify enriched pathways.

Quality control and contaminant removal of metagenomic reads

Raw paired-end metagenomic reads were subjected to a comprehensive quality control (QC) pipeline using BBTools (v39.01). Initial reformatting was performed using ReformatReads to standardize sequence headers, convert all IUPAC ambiguity codes to “N” (iupacToN = t), enforce uppercase sequences (touppercase = t), and ensure consistent Phred + 33 quality encoding (qout = 33). Read pairing integrity was verified with verifypaired = t, and the output reads were saved as compressed FASTQ files. To eliminate PCR or optical duplicates, we applied Clumpify with duplicate removal enabled (dedupe = t), allowing a maximum of two mismatches per duplicate pair (dupesubs = 2). This step effectively collapsed redundant read clusters, yielding a deduplicated dataset for downstream analysis.

All quality control steps were executed within the Metagenome-Atlas v2.16.3 pipeline [115]. Adapter sequences and low-quality bases were eliminated from paired-end reads using bbduk.sh (BBTools v39.01) with: ktrim=r ref=adapters k=27 mink=8 hdist=1 qtrim=rl trimq=10 minlength=51 minbasefrequency=0.05 ecco=t qout=33. The adapter reference (adapters.fa) and PhiX (phiX174_virus.fa) were taken from the Metagenome-Atlas installed database (atlas2.16_db). Reads were then screened against a merged reference comprising masked human (host) and PhiX using bbsplit.sh (BBTools v39.01) with: local=t k=13 minhits=1 minratio=0.65 maxindel=20 ambiguous=best (paired and singleton reads processed separately). Reads mapping to any reference were removed; only non-host, non-PhiX reads were retained for downstream analyses. The host.fa file in the installed database corresponds to the hg19 masked index distributed via Zenodo (hg19_main_mask_ribo_animal_allplant_allfungus.fa.gz; https://zenodo.org/record/1208052).

Metagenomic data assembly

Assembly of metagenomic sequences was conducted using the Metagenome-Atlas v2.16.3 pipeline [115]. The process involved the ‘atlas run all’ command for streamlined execution. Initially, adapter sequences were trimmed, and human genomic contaminations were removed utilizing BBMap suite v39 [116]. Subsequently, reads underwent error correction and were merged prior to assembly with MEGAHIT v1.2.9 [117]. For binning, MetaBAT v2.15 [118] and MaxBin v2.2 [119] were employed, and the integrated results were processed using DAS Tool v1.1 [120]. The resultant MAGs, which achieved a minimum of 50% completeness and less than 10% contamination as assessed by CheckM v2 [121], were clustered based on 97.5% average nucleotide identity. 50 species-representative MAGs and 6 strain-level MAGs were then taxonomically annotated using GTDB-tk v2.1 against Genome Taxonomy Database release 214.0 [122]. The resulting phylogenetic tree was visualized using Interactive Tree Of Life (iTOL) v6 [123].

Taxonomic identification and functional profiling of microbial communities

High-quality reads from each sample were mapped to a reference set of 2419 species-level microbial Operational Taxonomic Units (mOTUs) for taxonomic identification [124]. Functional profiling was conducted using HUMAnN v3.6 [125] with default parameters for all modules. Differentially enriched KEGG modules were determined based on their reporter scores. These scores were calculated from the Z-scores of individual KEGG Orthologs (KOs) by the “ReporterScore” R package v0.1.4 [126, 127]. A detection threshold was set at an absolute reporter score of 1.96 to identify modules showing significant differences in abundance.

Construction of Random Forest models for phenotypic skin aging prediction

To evaluate the predictive performance of metabolomics, microbiome, and integrated multi-omics features in predicting phenotypic skin aging, we constructed Random Forest models. All models were implemented using the randomForest package v4.7–1.2 [128] and reproducibility was ensured with a fixed random seed (set.seed(123)). The datasets were initially split into training and validation sets using a 7:3 ratio via stratified sampling with the createDataPartition() function, ensuring consistency of sample assignments across all three groups (metabolites, microbials, and combined omics) to maintain comparability across models. The original, unfiltered features consisted of 247 microbial species (selected based on frequency > 30% and relative abundance > 0.01), and 3033 core metabolites (selected based on frequency > 30% and relative abundance > 0.001%). Feature importance was evaluated using the MeanDecreaseAccuracy metric derived exclusively from the training dataset via the importance() function. Significant features for each data type: metabolites, microbials, and combined microbials + metabolites were independently selected using this metric. For each group, features with a MeanDecreaseAccuracy > 1 were retained for downstream analysis. A new Random Forest model was trained using these significant features, with mtry set to the square root of the feature count and ntree fixed at 200. Model performance on validation datasets was assessed using sensitivity, specificity, precision, recall, F1 score, and area under the ROC curve (AUC). The caret package v7.0–1 [129] was used for confusion matrix computation and performance evaluation, while the pROC package v1.18.5 [130] was employed for ROC curve generation and AUC calculation. Data visualization, including ROC curves, was conducted using ggplot2 v3.4.4 [131], ggpubr v0.6.0, and ggprism v1.0.5.

Genome-scale metabolic model (GEM) construction

In this study, GEMs were constructed based on metagenomic data from 103 samples, which produced 56 MAGs using the Metagenome-Atlas v2.16.3 [115] pipeline as mentioned above. The genome sequences of these MAGs were stored in.fna files for our GEM construction. We utilized gapseq v1.2 [132], a tool designed for the efficient and robust construction of metabolic pathways and the identification of transport proteins through homology-based methods. The “gapseq doall” process was instrumental in our approach, encompassing the identification of metabolic pathways (“find” function) and transport proteins (“find-transport” function), followed by the generation of preliminary draft models. These models were further refined by a thorough gap-filling process (“fill” function) to address metabolic inconsistencies. To generate individual sample-specific GEMs, we employed the “mergeTwoModels” function alongside our custom-developed function, “processDuplicateReactions”. The methodology involved weighting the flux bounds of reactions in each MAG GEM based on the relative abundance of the corresponding MAG in each sample. This process ensured that the flux bounds of the same reactions across different MAGs within a sample were aggregated based on their weighted values to establish the flux bounds in the sample’s GEM. Subsequently, these sample-specific GEMs were merged according to group classifications (older or younger), forming two distinct group-specific GEMs.

Metabolic network analysis

For analyzing and manipulating the reconstructed metabolic models, we employed the COBRA Toolbox (2024 release) [133], a MATLAB toolbox used for constraint-based modeling of metabolic networks. We performed linear and integer programming simulations using the Gurobi Optimizer (version 1003) [134]. Flux Balance Analysis (FBA) of each reaction was conducted using the “optimizeCbModel” function of the COBRA Toolbox, which optimizes metabolic networks to elucidate metabolic capabilities. We also assessed the impact of metabolites and microbes on reactions by computing shadow prices using “FBAsolution.f” and “FBAsolution.y” functions from the toolbox. For visualizing metabolic networks, the “createMetIntrcNetwork” function from the COBRA Toolbox was used to generate interactive network graphs, highlighting the interconnectivity and flux distributions within our metabolic models, thus providing a detailed view of the metabolic interactions.

Statistical analysis and data visualization

Differences in skin phenotypes, microbial species, and metabolites between older and younger groups were evaluated using Wilcoxon rank-sum tests to identify statistically significant variations. The p-values derived from these tests were adjusted for multiple comparisons employing the False Discovery Rate (FDR) correction method. Microbial differential abundance analysis was specifically performed on taxonomic profiles using the Wilcoxon rank-sum test, and the resulting p-values were adjusted for multiple comparisons using the Benjamini–Hochberg FDR method (adjusted P < 0.05 considered significant), and biological relevance (absolute log₂fold change > 1) was also required to define differential abundance. This adjustment was conducted using the “stats” package in R v4.1.1. In addition, correlations were assessed using Spearman’s rank correlation coefficient, implemented through the “psych” package v2.3.3 [135] in R, to explore relationships between variables across different groups. Python v3.8.8 and the Pandas library v1.2.4 [136] were utilized for compiling shadow price analysis results. The R packages ggplot2 v3.4.4 [131] and ggrepel v0.9.3 [137] facilitated the data visualization.

Bacterial growth conditions and supernatant preparation

The Stenotrophomonas maltophilia strain used in this study was obtained from the BeNa Culture Collection (BNCC) under preservation number BNCC185982.

The S. maltophilia strain was streaked onto Brain-Heart Infusion (BHI) agar plates and incubated overnight at 37 °C. The following day, individual colonies were inoculated into liquid BHI medium and cultured overnight at 37 °C with shaking at 220 rpm. Subsequently, cultures of the activated strain were transferred into fresh BHI medium. After 24 h of incubation, when cultures reached an OD₆₀₀ of 0.8–1.0, bacterial cells were pelleted by centrifugation, and the culture supernatants were filtered twice through 0.22 μm Spin-X centrifuge tube filters (Corning). The supernatants were then stored at − 80 °C for subsequent experiments. Notably, supernatants were also collected at 12 h and 72 h post-inoculation, respectively, to achieve the desired OD₆₀₀ values.

Cell culture

Primary human foreskin fibroblasts were purchased from ZQXZBio (catalog no. PRI-H-00078). Briefly, skin specimens were immersed in sterile normal saline and washed three times with phosphate-buffered saline (PBS). The specimens were then cut into 1 mm × 1 mm sections and placed in a 6-cm Petri dish, with the epidermis facing upward and the dermis downward. Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% (v/v) fetal bovine serum (FBS) and 1% (v/v) penicillin-streptomycin (all from Gibco, USA) was added to each dish. The culture medium was replaced every 2–3 days. After 7–14 days, dense fibroblast outgrowths were observed and subsequently passaged using 0.25% trypsin-EDTA. All experiments utilized primary fibroblasts within five passages. Cultures were maintained in a standard incubator at 37 °C with 5% CO₂.

Stimulation of fibroblasts with bacterial supernatants

Primary human skin fibroblasts were plated in 6 cm Petri dishes and grown until reaching approximately 80% confluence. Filtered supernatants from S. maltophilia cultures were diluted in DMEM at a ratio of 1:100 and used to stimulate fibroblasts for 24 h. Cells treated with BHI alone served as the control group.

Indirect stimulation of fibroblasts by bacterial supernatant via epidermal cells

A conditioned medium co-culture system was employed to investigate whether bacterial metabolites could indirectly affect fibroblast function through epidermal cells. Primary epidermal cells were seeded and cultured until reaching appropriate density, then stimulated for 24 h with the prepared filtered supernatant of S. maltophilia (diluted 1:100 in DMEM). After stimulation, the medium containing bacterial supernatant was completely removed, and the cells were gently washed three times with sterile PBS to eliminate any residual stimulants. Subsequently, fresh complete medium without any stimulants was added to continue culturing the epidermal cells for another 24 h. Following this incubation, the supernatant from the epidermal cell culture was collected. This conditioned medium was then used to treat human skin fibroblasts for 24 h to evaluate its indirect effects.

Quantification of GSH and GSSG levels in fibroblasts

GSH and GSSG levels in human skin fibroblasts were quantified using the S0053 Glutathione Assay Kit (Beyotime, China) following the manufacturer’s protocol. Cell samples were lysed to release intracellular glutathione. Total glutathione content, comprising both reduced (GSH) and oxidized (GSSG) forms, was measured by reducing GSSG to GSH using glutathione reductase. The resulting GSH reacted with 5,5′-dithiobis-(2-nitrobenzoic acid) (DTNB), forming the yellow compound 5-thio-2-nitrobenzoic acid (TNB). The absorbance of TNB was measured at 412 nm (A412) to quantify total glutathione. To specifically measure GSSG levels, a GSH masking reagent provided in the kit was used to selectively remove GSH from the samples before repeating the reduction and detection process. The concentration of reduced GSH was calculated by subtracting the GSSG concentration from the total glutathione content.

β-Galactosidase staining for cellular senescence detection

Cellular senescence was assessed using the β-Galactosidase Staining Kit (Cat # KTA3030) following the manufacturer’s protocol. Adherent cells were cultured in 24-well plates and subjected to the appropriate treatments. After treatment, the culture medium was aspirated, and the cells were washed twice with 1 × PBS. Next, 1 mL of 1 × fixative solution was added to each well, and the cells were fixed at room temperature for 15 min. Following fixation, the fixative solution was removed, and the cells were washed three times with 1 × PBS. Subsequently, 1 mL of staining working solution was added to each well, and the plates were incubated overnight in a CO₂-free incubator at 37 °C until the cells developed blue staining. To prevent cell drying during incubation, the plates were sealed with parafilm, and the staining duration was adjusted to achieve optimal staining, the cells were observed under a light microscope to evaluate β-galactosidase activity.

DCFH-DA staining for intracellular ROS detection

Intracellular ROS levels were measured using the fluorescent probe DCFH-DA (Cat # D837204). After treatment, adherent cells were washed with 1 × PBS and then incubated with DCFH-DA (5 μM) in serum-free medium at 37 °C in the dark for 30 min. Following incubation, the probe solution was removed and the cells were washed three times with 1 × PBS. Images were acquired under a fluorescence microscope at an excitation wavelength of 488 nm, and the fluorescence intensity was proportional to intracellular ROS levels.

Construction and treatment of full-thickness skin models

The T-Skin model (full-thickness skin model) was purchased from EPISKIN (Shanghai, China). The full-thickness skin model was established in an in vitro culture system to simulate skin tissue structure. The experiment was divided into four groups: control group, H₂O₂ damage group, bacterial treatment group (treated with S. maltophilia supernatant). Except for the control group, all other groups were treated with 200 µM H₂O₂ for 2 h to induce oxidative damage, after which the medium was replaced with fresh medium. For the bacterial treatment group, 30 µL/cm² of S. maltophilia supernatant was uniformly applied to the epidermal side. All models were further cultured under standard conditions (37 °C, 5% CO₂) for 24 h, after which tissues were collected for analysis.

Tissue sample collection and frozen section preparation

After treatment, all models were fixed with 4% paraformaldehyde for 30 min, followed by frozen sectioning. Tissues were embedded in OCT compound, rapidly frozen on dry ice, and sectioned at a thickness of 7 µm using a cryostat at − 20 °C. The sections were used for H&E staining and immunofluorescence staining, respectively.

H&E staining

Following methanol fixation, tissue sections were sequentially stained with hematoxylin and eosin to visualize tissue morphology. Sections were dehydrated with isopropanol and air-dried. Mayer’s hematoxylin was applied for 7 min, followed by rinsing with water and treatment with bluing buffer for 2 min to enhance nuclear staining. After further rinsing, sections were stained with eosin Y solution (in 0.45 M Tris-acetic acid buffer, pH 6.0) for 1 min to label the cytoplasm. Stained sections were observed directly under a light microscope.

Immunofluorescence staining

The sections were washed with PBS, fixed with 4% PFA for 30 min, and permeabilized with 0.3% Triton X-100 for 30 min. After blocking with 5% goat serum for 1.5 h at room temperature, the sections were incubated with the primary antibody, rabbit anti-Cytokeratin 10 (KRT10, 1:500; Proteintech, Wuhan, China), overnight at 4 °C. Then, an Alexa Fluor-conjugated goat anti-rabbit IgG secondary antibody (1:1000) was applied and incubated for 1 h at room temperature in the dark. Nuclei were counterstained with DAPI (5 µg/mL) for 10 min. Finally, the sections were mounted with an anti-fade mounting medium, and images were captured using a fluorescence microscope.

RNA sequencing

Total RNA was extracted from primary fibroblasts using the RNA Quick Extraction Kit (Magen, China) according to the manufacturer’s protocol. To remove DNA contamination, total RNA samples were treated with DNase I (NEB), and the reaction was terminated by adding 0.5 M EDTA, followed by incubation at 75 °C for 10 min and cooling on ice. The samples were then purified using RNA Clean beads (Vazyme). Enrichment and purification of mRNA were performed using mRNA Capture Beads (Vazyme). The enriched mRNA was incubated with 5 × HiScript II Buffer (Vazyme) and N6 primer (0.1 µg/µL) at 94 °C for 10 min. The fragmented RNA was combined with dNTP mix (10 mM, Vazyme), RNase Inhibitor (40 U/µL, Vazyme), and HiScript II Reverse Transcriptase (200 U/µL, Vazyme) for reverse transcription. The first-strand reaction products were added to a mix containing 5 × Second Strand Buffer (Invitrogen), dNTP mix (10 mM, Vazyme), RNase H (5 U/µL, Yeasen), DNA Polymerase I (5 U/µL, Takara), and nuclease-free water (Invitrogen). The reaction was carried out at 16 °C for 2 h, and the samples were purified using DNA Clean Beads (Vazyme). End repair and A-tailing were performed using T4 DNA Polymerase, Klenow DNA Polymerase, and T4 Polynucleotide Kinase (Vazyme). Details were added using Taq DNA Polymerase (Vazyme). Library amplification was performed using AD153 primers and 2 × KAPA HiFi HotStart ReadyMix (Vazyme). The final libraries were purified, quantified, and pooled for sequencing. Paired-end sequencing (PE150) was conducted on the MGI T7 platform. The integrity of RNA was assessed using the RNA Nano 6000 Assay Kit on the Agilent Bioanalyzer 2100 system (Agilent Technologies, CA, USA). The quality of libraries was also verified using the Agilent Bioanalyzer 2100 system.

Processing and analysis of RNA-seq data

Raw reads obtained from the high-throughput sequencing platform were processed to generate an expression matrix through the following steps. Quality control of the raw data was conducted using FastQC v0.12.1 [138] to assess sequencing quality. Low-quality reads and adapter sequences were removed using Trim Galore v0.6.10 [139], and non-coding RNA sequences were filtered out using SortMeRNA v4.3.6 [140]. The cleaned reads were aligned to the reference genome using STAR v2.7.10b [141], and gene-level read counts were quantified by mapping the alignment results to gene annotations with FeatureCounts v2.0.6 [142]. The final expression matrix served as the input for subsequent differential expression and downstream analyses. Differential expression analysis was performed using the DESeq2 R package v1.32.0 [143], identifying genes with an adjusted P < 0.05 and absolute Log₂FoldChange > 0.5 as differentially expressed. Gene Ontology (GO) and KEGG enrichment analyses were conducted using the clusterProfiler R package v4.0.5 [144], with KEGG terms considered significantly enriched at an adjusted P < 0.1.

Supplementary Information

40168_2026_2433_MOESM1_ESM.pdf^{(32.8MB, pdf)}

Supplementary Material 1. Figure S1 Metagenomic analysis reveals skin-aging-related microbial dynamics. Figure S2: Metabolomic analysis reveals skin-aging-related metabolites. Figure S3: MAGs analysis of age-related differences in skin microbiota. Figure S4: GEM quality assessment and experimental validation reveal antioxidant defense mechanisms. Figure S5: Shadow price analysis from FBA illuminates betaine, lysolecithin synthesis dynamics, and porphyrin metabolism dynamics. Figure S6: Shadow price analysis from FBA illuminates tyrosine metabolism dynamics.

40168_2026_2433_MOESM2_ESM.zip^{(26.4MB, zip)}

Supplementary Material 2. Table S1. Assessment of skin physiological traits, related to Fig. 1. Table S2. Metagenomic analysis of microbial species, diversity, and functional pathways, related to Fig. 2. Table S3. Comprehensive metabolomic analysis, differential core metabolites, and pathway enrichment related to Fig. 3. Table S4. Analysis of metagenome-assembled genomes (MAGs) across samples, related to Figure S3. Table S5. Genome-Scale Metabolic Modeling (GEM), experimental validation of the antioxidant and anti-aging roles of S. maltophilia, related to Fig. 4. Table S6. Comprehensive shadow price analysis of Betaine, Lysolecithin, Porphyrin, DHPAA, and HMA metabolism across MAGs in skin aging, related to Figs. 5 and 6.

Acknowledgements

We gratefully acknowledge the technical support from the Biopharmaceutical and Health Engineering Research Core Facility, Tsinghua Shenzhen International Graduate School, Tsinghua University. We specifically thank the platform for the assistance with frozen sectioning, slide-image analysis system and fluorescence microscope.

Authors’ contributions

X.L. and H.J. conceived the idea; X.L. and H.J. supervised the work; H.J., X.L., J.C. and Y.C. designed the experiments; D.G. wrote the manuscript; J.C., Y.C. and W.L. collected the samples.; J.C., W.M. and H.Y. performed the majority of the experiments; D.G. and J.C. analyzed the data; Y.W. and Y.L. conducted the functional experiments; X.L., L.H., and L.M. reviewed and edited the manuscript.

Funding

This work was supported by Scientific Research Start-up Funds (QD2021005N), Shenzhen Science and Technology Program (WDZC20220819134430002), Department of Chemical Engineering-iBHE Special Cooperation Joint Fund (DCE-iBHE-2025-2), and Shanghai Jahwa United Co., Ltd.

Data availability

The raw metagenomic sequencing data and RNA-seq data from this study are available in the NCBI BioProject database with accession number PRJNA1201908, and the raw metabolomics data have been deposited in the MetaboLights database with accession number MTBLS11994. Processed data matrices are provided in the Supplementary Tables. The code used in this study is available on GitHub (https://github.com/guodanni1/Skin_Aging) to ensure reproducibility.

Declarations

Consent for publication

Not applicable.

Competing interests

Haidong Jia and Yuanyuan Chen receive salaries from Shanghai Jahwa United Co., Ltd., and the other authors declare no competing interests.

Footnotes

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Danni Guo, Yuanyuan Chen and Yahong Wu contributed equally to this work.

Contributor Information

Haidong Jia, Email: jiahaidong@jahwa.com.cn.

Xiao Liu, Email: liuxiao@sz.tsinghua.edu.cn.

References

1.Oh J, Byrd AL, Deming C, Conlan S, Barnabas B, Blakesley R. Biogeography and individuality shape function in the human skin metagenome. Nature. 2014;514(7520):59–64. 10.1038/nature13786. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Grice EA, Segre JA. The skin microbiome. Nat Rev Microbiol. 2011;9(4):244–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
3.Kong HH. Skin microbiome: genomics-based insights into the diversity and role of skin microbes. Trends Mol Med. 2011;17(6):320–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Somerville DA. The normal flora of the skin in different age groups. Br J Dermatol. 1969;81(4):248–58. [DOI] [PubMed] [Google Scholar]
5.Leyden JJ, McGiley KJ, Mills OH, Kligman AM. Age-related changes in the resident bacterial flora of the human face. J Invest Dermatol. 1975;65(4):379–81. [DOI] [PubMed] [Google Scholar]
6.Oh J, Freeman AF, Park M, Sokolic R, Candotti F, Holland SM, et al. The altered landscape of the human skin microbiome in patients with primary immunodeficiencies. Genome Res. 2013;23(12):2103–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Kong HH, Oh J, Deming C, Conlan S, Grice EA, Beatson MA, et al. Temporal shifts in the skin microbiome associated with disease flares and treatment in children with atopic dermatitis. Genome Res. 2012;22(5):850–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Williams MR, Gallo RL. The role of the skin microbiome in atopic dermatitis. Curr Allergy Asthma Rep. 2015;15:1–10. [DOI] [PubMed] [Google Scholar]
9.Holmes AD. Potential role of microorganisms in the pathogenesis of rosacea. J Am Acad Dermatol. 2013;69(6):1025–32. [DOI] [PubMed] [Google Scholar]
10.Chaudhary M, Khan A, Gupta M. Skin ageing: pathophysiology and current market treatment approaches. Curr Aging Sci. 2020;13(1):22–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Amaro-Ortiz A, Yan B, D’Orazio JA. Ultraviolet radiation, aging and the skin: prevention of damage by topical cAMP manipulation. Molecules. 2014;19(5):6202–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Wilhelm K-P, Cua AB, Maibach HI. Skin aging: effect on transepidermal water loss, stratum corneum hydration, skin surface pH, and casual sebum content. Arch Dermatol. 1991;127(12):1806–9. [DOI] [PubMed] [Google Scholar]
13.Potts RO, Buras EM, Chrisman DA. Changes with age in the moisture content of human skin. J Investig Dermatol. 1984;82(1):97–100. 10.1111/1523-1747.ep12259203. [DOI] [PubMed] [Google Scholar]
14.Cotterill JA, Cunliffe WJ, Williamson B, Bulusu L. Age and sex variation in skin surface lipid composition and sebum excretion rate. Br J Dermatol. 1972;87(4):333–40. [DOI] [PubMed] [Google Scholar]
15.Pochi PE, Strauss JS, Downing DT. Age-related changes in sebaceous gland activity. J Investig Dermatol. 1979;73(1):108–11. [DOI] [PubMed] [Google Scholar]
16.Xia Jj, Zhong Q, Li Zm, Wei Qz, Jiang L, Duan C, et al. Culture dependent and independent approaches reveal the role of specific bacteria in human skin aging. iMetaOmics. 2024;1(2): e26. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Kim J-H, Son S-M, Park H, Kim BK, Choi IS, Kim H, et al. Taxonomic profiling of skin microbiome and correlation with clinical skin parameters in healthy Koreans. Sci Rep. 2021;11(1):16269. [DOI] [PMC free article] [PubMed] [Google Scholar]
18.Kim H-J, Oh HN, Park T, Kim H, Lee HG, An S. Aged related human skin microbiome and mycobiome in Korean women. Sci Rep. 2022;12(1):2351. 10.1038/s41598-022-06189-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Wu L, Zeng T, Deligios M, Milanesi L, Langille MGI, Zinellu A, et al. Age-related variation of bacterial and fungal communities in different body habitats across the young, elderly, and centenarians in Sardinia. mSphere. 2020;5(1):e00558-e619. [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Zhou W, Fleming E, Legendre G, Roux L, Latreille J, Gendronneau G. Skin microbiome attributes associate with biophysical skin ageing. Exp Dermatol. 2023. 10.1111/exd.14863. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Alkema W, Boekhorst J, Eijlander RT, Schnittger S, De Gruyter F, Lukovac S, et al. Charting host-microbe co-metabolism in skin aging and application to metagenomics data. PLoS One. 2021;16(11):e0258960. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Howard B, Bascom CC, Hu P, Binder RL, Fadayel G, Huggins TG. Aging-associated changes in the adult human skin microbiome and the host factors that affect skin microbiome composition. J Investig Dermatol. 2022;142(7):1934-46.e21. 10.1016/j.jid.2021.11.029. [DOI] [PubMed] [Google Scholar]
23.Kim M, Park T, Yun JI, Lim HW, Han NR, Lee ST. Investigation of age-related changes in the skin microbiota of Korean women. Microorganisms. 2020;8(10): 1581. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Larson PJ, Zhou W, Santiago A, Driscoll S, Fleming E, Voigt AY, et al. Associations of the skin, oral and gut microbiome with aging, frailty and infection risk reservoirs in older adults. Nat Aging. 2022;2(10):941–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Kim H-J, Kim JJ, Myeong NR, Kim T, Kim D, An S. Segregation of age-related skin microbiome characteristics by functionality. Sci Rep. 2019;9(1):16748. 10.1038/s41598-019-53266-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Shibagaki N, Suda W, Clavaud C, Bastien P, Takayasu L, Iioka E. Aging-related changes in the diversity of women’s skin microbiomes associated with oral bacteria. Sci Rep. 2017;7(1):10567. 10.1038/s41598-017-10834-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Reilly DM, Lozano J. Skin collagen through the lifestages: importance for skin health and beauty. Plast Aesthet Res. 2021;8: N-A. [Google Scholar]
28.Alexis AF, Obioha JO. Ethnicity and aging skin. J Drugs Dermatol. 2017;16(6):s77–80. [PubMed] [Google Scholar]
29.Shen X, Wang C, Zhou X, Zhou W, Hornburg D, Wu S. Nonlinear dynamics of multi-omics profiles during human aging. Nat Aging. 2024. 10.1038/s43587-024-00692-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Zhou E, Cao Z, Yin Q. Naive-deep face recognition: Touching the limit of LFW benchmark or not? arXiv preprint arXiv:1501.04690. 2015. https://arxiv.org/abs/1501.04690.
31.Papaccio F, Arino A, Caputo D′S, Bellei B. Focus on the contribution of oxidative stress in skin aging. Antioxidants Basel. 2022;11(6):1121. [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Li Z, Xia J, Jiang L, Tan Y, An Y, Zhu X, et al. Characterization of the human skin resistome and identification of two microbiota cutotypes. Microbiome. 2021;9:1–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Ahle CM, Stødkilde K, Poehlein A, Bömeke M, Streit WR, Wenck H, et al. Interference and co-existence of staphylococci and Cutibacterium acnes within the healthy human skin microbiome. Commun Biol. 2022;5(1):923. [DOI] [PMC free article] [PubMed] [Google Scholar]
34.Kimoto‐Nira H. New lactic acid bacteria for skin health via oral intake of heat‐killed or live cells. Anim Sci J. 2018;89(6):835–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Yoshida A, Aoki R, Kimoto-Nira H, Kobayashi M, Kawasumi T, Mizumachi K. Oral administration of live Lactococcus lactis C59 suppresses IgE antibody production in ovalbumin-sensitized mice via the regulation of interleukin-4 production. FEMS Immunol Med Microbiol. 2011;61(3):315–22. [DOI] [PubMed] [Google Scholar]
36.Kimoto H, Mizumachi K, Okamoto T, Kurisaki J. New Lactococcus strain with immunomodulatory activity: enhancement of Th1‐type immune response. Microbiol Immunol. 2004;48(2):75–82. [DOI] [PubMed] [Google Scholar]
37.Kimoto-Nira H, Aoki R, Sasaki K, Suzuki C, Mizumachi K. Oral intake of heat-killed cells of Lactococcus lactis strain H61 promotes skin health in women. J Nutr Sci. 2012;1:e18. [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Ramsey JP, Mercurio A, Holland JA, Harris RN, Minbiole KPC. The cutaneous bacterium Janthinobacterium lividum inhibits the growth of Trichophyton rubrum in vitro. Int J Dermatol. 2015;54(2):156–9. [DOI] [PubMed] [Google Scholar]
39.Brooke JS. Advances in the microbiology of Stenotrophomonas maltophilia. Clin Microbiol Rev. 2021;34(3):10–1128. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.An S-q, Berg G. Stenotrophomonas maltophilia. Trends Microbiol. 2018;26(7):637–8. [DOI] [PubMed] [Google Scholar]
41.Mukherjee P, Roy P. Genomic potential of Stenotrophomonas maltophilia in bioremediation with an assessment of its multifaceted role in our environment. Front Microbiol. 2016;7:159261. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Gonzalez JM, Aranda B. Microbial growth under limiting conditions-future perspectives. Microorganisms. 2023;11: 1641. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Merchant SS, Helmann JD. Elemental economy: microbial strategies for optimizing growth in the face of nutrient limitation. Adv Microb Physiol. 2012;60:91–210. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Nüse B, Holland T, Rauh M, Gerlach RG, Mattner J. L-arginine metabolism as pivotal interface of mutual host–microbe interactions in the gut. Gut Microbes. 2023;15(1): 2222961. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Roux P-F, Oddos T, Stamatas G. Deciphering the role of skin surface microbiome in skin health: an integrative multiomics approach reveals three distinct metabolite‒microbe clusters. J Invest Dermatol. 2022;142(2):469–79. [DOI] [PubMed] [Google Scholar]
46.Saini R, Badole SL, Zanwar AA. Arginine derived nitric oxide: Key to healthy skin. In: Bioactive dietary factors and plant extracts in dermatology. 2013. p. 73–82.
47.He J, Fang B, Shan S, Li Q. Mechanical stiffness promotes skin fibrosis through Piezo1-mediated arginine and proline metabolism. Cell Death Discov. 2023;9(1): 354. [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Patriarca EJ, Cermola F, D’Aniello C, Fico A, Guardiola O, De Cesare D, et al. The multifaceted roles of proline in cell behavior. Front Cell Dev Biol. 2021;9:728576. [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Solano F. Metabolism and functions of amino acids in the skin. In: Amino Acids in Nutrition and Health: Amino acids in systems function and health. 2020. p. 187–99. [DOI] [PubMed]
50.Bach TMH, Takagi H. Properties, metabolisms, and applications of L-proline analogues. Appl Microbiol Biotechnol. 2013;97:6623–34. [DOI] [PubMed] [Google Scholar]
51.Li Z, Bai X, Peng T, Yi X, Luo L, Yang J, et al. New insights into the skin microbial communities and skin aging. Front Microbiol. 2020;11: 565549. [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Chambers ES, Vukmanovic‐Stejic M. Skin barrier immunity and ageing. Immunology. 2020;160(2):116–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Prescott SL, Larcombe D-L, Logan AC, West C, Burks W, Caraballo L. The skin microbiome: impact of modern environments on skin ecology, barrier integrity, and systemic immune programming. World Allergy Organ J. 2017;10(1):16. [DOI] [PMC free article] [PubMed] [Google Scholar]
54.Rayan M, Sayed TS, Hussein OJ, Therachiyil L, Maayah ZH, Maccalli C, et al. Unlocking the secrets: exploring the influence of the aryl hydrocarbon receptor and microbiome on cancer development. Cell Mol Biol Lett. 2024;29(1):33. [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Dempsey JL, Cui JY. Microbiome is a functional modifier of P450 drug metabolism. Curr Pharmacol Rep. 2019;5(6):481–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
56.Lee HU, McPherson ZE, Tan B, Korecka A, Pettersson S. Host-microbiome interactions: the aryl hydrocarbon receptor and the central nervous system. J Mol Med. 2017;95:29–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Tiedt O, Mergelsberg M, Eisenreich W, Boll M. Promiscuous defluorinating enoyl-CoA hydratases/hydrolases allow for complete anaerobic degradation of 2-fluorobenzoate. Front Microbiol. 2017;8: 322966. [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Murphy CD, Quirke S, Balogun O. Degradation of fluorobiphenyl by Pseudomonas pseudoalcaligenes KF707. FEMS Microbiol Lett. 2008;286(1):45–9. [DOI] [PubMed] [Google Scholar]
59.Bhanot V, Pali S, Panwar J. Understanding the in silico aspects of bacterial catabolic cascade for styrene degradation. Proteins. 2023;91(4):532–41. [DOI] [PubMed] [Google Scholar]
60.Guan N, Li J, Shin H-d, Du G, Chen J, Liu L. Microbial response to environmental stresses: from fundamental mechanisms to practical applications. Appl Microbiol Biotechnol. 2017;101:3991–4008. [DOI] [PubMed] [Google Scholar]
61.Biernacki M, Conde T, Stasiewicz A, Surażyński A, Domingues MR, Domingues P. Restorative effect of microalgae Nannochloropsis oceanica lipid extract on phospholipid metabolism in keratinocytes exposed to UVB radiation. Int J Mol Sci. 2023;24:18 14323. [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Del Favero G, Janker L, Neuditschko B, Hohenbichler J, Kiss E, Woelflingseder L, et al. Exploring the dermotoxicity of the mycotoxin deoxynivalenol: combined morphologic and proteomic profiling of human epidermal cells reveals alteration of lipid biosynthesis machinery and membrane structural integrity relevant for skin barrier function. Arch Toxicol. 2021;95(6):2201–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
63.Firmansyah D, Sumiwi SA, Saptarini NM, Levita J. Curcuma longa extract inhibits the activity of mushroom tyrosinase and the growth of murine skin cancer B16F10 cells. J Herbmed Pharmacol. 2022;12(1):153–8. [Google Scholar]
64.Goenka S, Simon SR. Comparative study of curcumin and its hydrogenated metabolites, tetrahydrocurcumin, hexahydrocurcumin, and octahydrocurcumin, on melanogenesis in B16F10 and MNT-1 cells. Cosmetics. 2021;8(1): 4. [Google Scholar]
65.Riaz R, Batool S, Zucca P, Rescigno A, Peddio S, Saleem RSZ. Plants as a promising reservoir of tyrosinase inhibitors. Mini Rev Org Chem. 2021;18(6):808–28. [Google Scholar]
66.Guo L, Jin H. Research progress of metabolomics in psoriasis. Chin Med J. 2023;136(15):1805–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
67.Koussiouris J, Looby N, Anderson M, Kulasingam V, Chandran V. Metabolomics studies in psoriatic disease: a review. Metabolites. 2021;11(6): 375. [DOI] [PMC free article] [PubMed] [Google Scholar]
68.Sharma D, Arora S, dos Santos Rodrigues B, Lakkadwala S, Banerjee A, Singh J. Chitosan-based systems for gene delivery. In: Functional Chitosan: Drug Delivery and Biomedical Applications. 2019. p. 229–67.
69.Seité S, Rougier A. Lipohydroxy acid containing shampoo in the treatment of scalp seborrheic dermatitis. In: Handbook of hair in health and disease. Wageningen Academic; 2011. p. 466–76.
70.Uhoda E, Pierard C, Pierard G. Comedolysis by a lipohydroxyacid formulation in acne-prone subjects. European J Dermatol. 2003;13(1):65–8. [PubMed]
71.Avila-Camacho M, Montastier C, Piérard GE. Histometric assessment of the age-related skin response to 2-hydroxy-5-octanoyl benzoic acid. Skin Pharmacol Physiol. 1998;11(1):52–6. [DOI] [PubMed] [Google Scholar]
72.Russo E, Di Gloria L, Cerboneschi M, Smeazzetto S, Baruzzi GP, Romano F, et al. Facial skin microbiome: aging-related changes and exploratory functional associations with host genetic factors, a pilot study. Biomedicines. 2023;11(3): 684. [DOI] [PMC free article] [PubMed] [Google Scholar]
73.Homma T, Fujii J. Application of glutathione as anti-oxidative and anti-aging drugs. Curr Drug Metab. 2015;16(7):560–71. [DOI] [PubMed] [Google Scholar]
74.Foyer CH, Noctor G. Ascorbate and glutathione: the heart of the redox hub. Plant Physiol. 2011;155(1):2–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
75.Li L-H, Shih Y-L, Huang J-Y, Wu C-J, Huang Y-W, Huang H-H, et al. Protection from hydrogen peroxide stress relies mainly on AhpCF and KatA2 in Stenotrophomonas maltophilia. J Biomed Sci. 2020;27:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
76.Li L-H, Wu C-M, Lin Y-T, Pan S-Y, Yang T-C. Roles of FadRACB system in formaldehyde detoxification, oxidative stress alleviation and antibiotic susceptibility in Stenotrophomonas maltophilia. J Antimicrob Chemother. 2020;75(8):2101–9. [DOI] [PubMed] [Google Scholar]
77.Liao C-H, Ku R-H, Li L-H, Wu C-M, Yang T-C. Role of yceA-cybB-yceB operon in oxidative stress tolerance, swimming motility and antibiotic susceptibility of Stenotrophomonas maltophilia. J Antimicrob Chemother. 2023;78(8):1891–9. [DOI] [PubMed] [Google Scholar]
78.Dimri GP, Lee X, Basile G, Acosta M, Scott G, Roskelley C. A biomarker that identifies senescent human cells in culture and in aging skin in vivo. Proc Natl Acad Sci U S A. 1995;92(20):9363–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
79.Sikalidis AK, Mazor KM, Lee J-I, Roman HB, Hirschberger LL, Stipanuk MH. Upregulation of capacity for glutathione synthesis in response to amino acid deprivation: regulation of glutamate–cysteine ligase subunits. Amino Acids. 2014;46:1285–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
80.Kurata M, Suzuki M, Agar NS. Glutathione regeneration in mammalian erythrocytes. Comp Haematol Int. 2000;10:59–67. [Google Scholar]
81.Wiley CD, Campisi J. The metabolic roots of senescence: mechanisms and opportunities for intervention. Nat Metab. 2021;3(10):1290–301. [DOI] [PMC free article] [PubMed] [Google Scholar]
82.Nishida-Tamehiro K, Kimura A, Tsubata T, Takahashi S, Suzuki H. Antioxidative enzyme NAD (P) H quinone oxidoreductase 1 (NQO1) modulates the differentiation of Th17 cells by regulating ROS levels. PLoS One. 2022;17(7):e0272090. [DOI] [PMC free article] [PubMed] [Google Scholar]
83.Yu Y, Di Trapani G, Tonissen KF. Thioredoxin and glutathione systems: cancer cells’ defensive weapons against oxidative stress. In: Handbook of Oxidative Stress in Cancer: Mechanistic Aspects. Springer; 2022. p. 2407–20.
84.Scaramuzzino L, Lucchino V, Scalise S, Lo Conte M, Zannino C, Sacco A, et al. Uncovering the metabolic and stress responses of human embryonic stem cells to FTH1 gene silencing. Cells. 2021;10(9):2431. [DOI] [PMC free article] [PubMed] [Google Scholar]
85.Suryadevara V, Hudgins AD, Rajesh A, Pappalardo A, Karpova A, Dey AK, et al. SenNet recommendations for detecting senescent cells in different tissues. Nat Rev Mol Cell Biol. 2024. 10.1038/s41580-024-00738-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
86.Wang L, Tao Y, Zhang J, Wang X, Zhou S, He L, et al. APCDD1 as a co-receptor positively regulates Wnt5a/c-Jun non-canonical signaling pathway. Journal of Shanghai Jiaotong University (Science). 2019;24:510–6. [Google Scholar]
87.Herrera J, Henke CA, Bitterman PB. Extracellular matrix as a driver of progressive fibrosis. J Clin Invest. 2018;128(1):45–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
88.Boysen AK, Durham BP, Kumler W, Key RS, Heal KR, Carlson LT. Glycine betaine uptake and metabolism in marine microbial communities. Environ Microbiol. 2022;24(5):2380–403. [DOI] [PMC free article] [PubMed] [Google Scholar]
89.Hwang H, Chun H, Kim D, Shin M, Kim YS, In S. Lysophosphatidylcholine exerts an anti-skin photoaging effect via heat shock protein 70 induction. J Cosmet Dermatol. 2021;20(12):4060–7. [DOI] [PubMed] [Google Scholar]
90.Stapleton CM, Mashek DG, Wang S, Nagle CA, Cline GW, Thuillier P, et al. Lysophosphatidic acid activates peroxisome proliferator activated receptor-γ in CHO cells that over-express glycerol 3-phosphate acyltransferase-1. PLoS One. 2011;6(4):e18932. [DOI] [PMC free article] [PubMed] [Google Scholar]
91.Artegoitia VM, Middleton JL, Harte FM, Campagna SR, De Veth MJ. Choline and choline metabolite patterns and associations in blood and milk during lactation in dairy cows. PLoS One. 2014;9(8):e103412. [DOI] [PMC free article] [PubMed] [Google Scholar]
92.Lee H-R, Kwon S-Y, Choi S-A, Lee J-H, Lee H-S, Park J-B. Valorization of soy lecithin by enzyme cascade reactions including a phospholipase A2, a fatty acid double-bond hydratase, and/or a photoactivated decarboxylase. J Agric Food Chem. 2022;70(35):10818–25. [DOI] [PubMed] [Google Scholar]
93.Melani NB, Tambourgi EB, Silveira E. Lipases: from production to applications. Sep Purif Rev. 2020;49(2):143–58. [Google Scholar]
94.Stroupe ME, Leech HK, Daniels DS, Warren MJ, Getzoff ED. CysG structure reveals tetrapyrrole-binding features and novel regulation of siroheme biosynthesis. Nat Struct Mol Biol. 2003;10(12):1064–73. [DOI] [PubMed] [Google Scholar]
95.Warren MJ, Raux E, Schubert HL, Escalante-Semerena JC. The biosynthesis of adenosylcobalamin (vitamin B12). Nat Prod Rep. 2002;19(4):390–412. [DOI] [PubMed] [Google Scholar]
96.Kang D, Shi B, Erfe MC, Craft N, Li H. Vitamin B12 modulates the transcriptome of the skin microbiota in acne pathogenesis. Sci Transl Med. 2015;7:293 293ra103 293ra103. [DOI] [PMC free article] [PubMed] [Google Scholar]
97.Baier J, Maisch T, Maier M, Engel E, Landthaler M, Bäumler W. Singlet oxygen generation by UVA light exposure of endogenous photosensitizers. Biophys J. 2006;91(4):1452–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
98.Altun I, Kurutaş EB. Vitamin B complex and vitamin B12 levels after peripheral nerve injury. Neural Regen Res. 2016;11(5):842–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
99.Mori K, Ando I, Kukita A. Generalized hyperpigmentation of the skin due to vitamin B12 deficiency. J Dermatol. 2001;28(5):282–5. [DOI] [PubMed] [Google Scholar]
100.Muñoz P, Huenchuguala S, Paris I, Segura-Aguilar J. Dopamine oxidation and autophagy. Parkinson Dis. 2012;2012(1): 920953. [DOI] [PMC free article] [PubMed] [Google Scholar]
101.Agarwal P, Singh M, Singh J, Singh RP. Microbial tyrosinases: A novel enzyme, structural features, and applications. In: Applied microbiology and bioengineering. Elsevier; 2019. p. 3–19.
102.Muñoz P, Huenchuguala S, Paris I, Segura-Aguilar J. Dopamine oxidation and autophagy. Parkinsons Dis. 2012. 10.1155/2012/920953. [DOI] [PMC free article] [PubMed] [Google Scholar]
103.Frases S, Chaskes S, Dadachova E, Casadevall A. Induction by Klebsiella aerogenes of a melanin-like pigment in Cryptococcus neoformans. Appl Environ Microbiol. 2006;72(2):1542–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
104.Kourtis N, Tavernarakis N. Cellular stress response pathways and ageing: intricate molecular relationships. EMBO J. 2011;30(13):2520–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
105.Boyajian JL, Ghebretatios M, Schaly S, Islam P, Prakash S. Microbiome and human aging: probiotic and prebiotic potentials in longevity, skin health and cellular senescence. Nutrients. 2021;13(12): 4550. [DOI] [PMC free article] [PubMed] [Google Scholar]
106.Townsend EC, Kalan LR. The dynamic balance of the skin microbiome across the lifespan. Biochem Soc Trans. 2023;51(1):71–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
107.Tan W, Zhang J, Zhao X, Li Q, Dong F, Guo Z. Preparation and physicochemical properties of antioxidant chitosan ascorbate/methylcellulose composite films. Int J Biol Macromol. 2020;146:53–61. [DOI] [PubMed] [Google Scholar]
108.Rahaman MM, Rakib A, Mitra S, Tareq AM, Emran TB, Shahid-Ud-Daula AFM, et al. The genus Curcuma and inflammation: Overview of the pharmacological perspectives. Plants. 2020;10(1):63. [DOI] [PMC free article] [PubMed] [Google Scholar]
109.Scapagnini G, Davinelli S, Renzo Di L, Lorenzo De A, Olarte HH, Micali G. Cocoa bioactive compounds: Significance and potential for the maintenance of skin health. Nutrients. 2014;6(8):3202–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
110.Wishart DS, Guo A, Oler E, Wang F, Anjum A, Peters H. HMDB 5.0: The human metabolome database for 2022. Nucleic Acids Res. 2022;50(D1):D622–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
111.O’Donnell VB, Dennis EA, Wakelam MJO, Subramaniam S. LIPID MAPS: Serving the next generation of lipid researchers with tools, resources, data, and training. Sci Signal. 2019;12(563):eaaw2964. [DOI] [PubMed] [Google Scholar]
112.Guijas C, Montenegro-Burke JR, Domingo-Almenara X, Palermo A, Warth B, Hermann G. METLIN: A technology platform for identifying knowns and unknowns. Anal Chem. 2018;90(5):3156–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
113.Carvalho S, Leite J, Galdo-Álvarez S, Gonçalves OF. The emotional movie database (EMDB): A self-report and psychophysiological study. Appl Psychophysiol Biofeedback. 2012;37:279–94. [DOI] [PubMed] [Google Scholar]
114.Castrignano T, Meo De PDO, Cozzetto D, Talamo IG, Tramontano A. The PMDB protein model database. Nucleic Acids Res. 2006;34(suppl_1):D306–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
115.Kieser S, Brown J, Zdobnov EM, Trajkovski M, McCue LA. ATLAS: a Snakemake workflow for assembly, annotation, and genomic binning of metagenome sequence data. BMC Bioinformatics. 2020;21:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
116.Bushnell B. BBMap: a fast, accurate, splice-aware aligner. Berkeley, CA, USA: Technical report, Lawrence Berkeley National Laboratory; 2014. https://www.osti.gov/servlets/purl/1241166.
117.Li D, Luo R, Liu C-M, Leung C-M, Ting H-F, Sadakane K, et al. MEGAHIT v1. 0: a fast and scalable metagenome assembler driven by advanced methodologies and community practices. Methods. 2016;102:3–11. [DOI] [PubMed] [Google Scholar]
118.Kang DD, Li F, Kirton E, Thomas A, Egan R, An H, et al. MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies. PeerJ. 2019;7: e7359. [DOI] [PMC free article] [PubMed] [Google Scholar]
119.Wu Y-W, Simmons BA, Singer SW. MaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic datasets. Bioinformatics. 2016;32(4):605–7. [DOI] [PubMed] [Google Scholar]
120.Sieber CMK, Probst AJ, Sharrar A, Thomas BC, Hess M, Tringe SG, et al. Recovery of genomes from metagenomes via a dereplication, aggregation and scoring strategy. Nat Microbiol. 2018;3(7):836–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
121.Chklovski A, Parks DH, Woodcroft BJ, Tyson GW. CheckM2: a rapid, scalable and accurate tool for assessing microbial genome quality using machine learning. Nat Methods. 2023;20(8):1203–12. [DOI] [PubMed] [Google Scholar]
122.Chaumeil P-A, Mussig AJ, Hugenholtz P, Parks DH. GTDB-Tk v2: memory friendly classification with the genome taxonomy database. Bioinformatics. 2022;38(23):5315–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
123.Letunic I, Bork P. Interactive Tree of Life (iTOL) v6: recent updates to the phylogenetic tree display and annotation tool. Nucleic Acids Res. 2024. 10.1093/nar/gkae268. [DOI] [PMC free article] [PubMed] [Google Scholar]
124.Ruscheweyh H-J, Milanese A, Paoli L, Karcher N, Clayssen Q, Keller MI, et al. Cultivation-independent genomes greatly expand taxonomic-profiling capabilities of mOTUs across various environments. Microbiome. 2022;10(1):212. [DOI] [PMC free article] [PubMed] [Google Scholar]
125.Beghini F, McIver LJ, Blanco-Míguez A, Dubois L, Asnicar F, Maharjan S, et al. Integrating taxonomic, functional, and strain-level profiling of diverse microbial communities with bioBakery 3. Elife. 2021;10: e65088. [DOI] [PMC free article] [PubMed] [Google Scholar]
126.Kanehisa M. The KEGG database. Novartis Found Symp. 2002;247:91–101; discussion 101–103, 119–128, 244–252. https://onlinelibrary.wiley.com/doi/10.1002/0470857897.ch8. [PubMed]
127.Patil KR, Nielsen J. Uncovering transcriptional regulation of metabolism by using metabolic network topology. Proc Natl Acad Sci U S A. 2005;102(8):2685–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
128.Liaw A, Wiener M. Classification and regression by randomForest. R News. 2002;2(3):18–22. https://journal.r-project.org/articles/RN-2002-022/RN-2002-022.pdf.
129.Kuhn M. Building predictive models in R using the Caret package. J Stat Softw. 2008;28(1):26. [Google Scholar]
130.Robin X, Turck N, Hainard A, Tiberti N, Lisacek F, Sanchez J-C, et al. pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics. 2011;12:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
131.Wilkinson L, Wickham H. ggplot2: elegant graphics for data analysis. Oxford University Press; 2011. [Google Scholar]
132.Zimmermann J, Kaleta C, Waschina S. Gapseq: informed prediction of bacterial metabolic pathways and reconstruction of accurate metabolic models. Genome Biol. 2021;22:1–35. [DOI] [PMC free article] [PubMed] [Google Scholar]
133.Heirendt L, Arreckx S, Pfau T, Mendoza SN, Richelle A, Heinken A, et al. Creation and analysis of biochemical constraint-based models using the COBRA Toolbox v. 3.0. Nat Protoc. 2019;14(3):639–702. [DOI] [PMC free article] [PubMed] [Google Scholar]
134.Gurobi Optimization, LLC. Gurobi Optimizer Reference Manual. 2021. https://www.gurobi.com.
135.Revelle W. psych: Procedures for Personality and Psychological Research. Northwestern University, Evanston, Illinois. R package. 2015. https://CRAN.R-project.org/package=psych.
136.Reback J, McKinney W, Van Den Bossche J, Augspurger T, Cloud P, Klein A, et al. pandas-dev/pandas: Pandas 1.0.5. Zenodo. 2020. 10.5281/zenodo.3898987.
137.Slowikowski K. ggrepel: Automatically position non-overlapping text labels with ‘ggplot2’. R package version 0.9.1. 2021. https://CRAN.R-project.org/package=ggrepel.
138.Andrews S. FastQC: a quality control tool for high throughput sequence data. Babraham Bioinformatics. 2010. https://www.bioinformatics.babraham.ac.uk/projects/fastqc/.
139.Krueger F. Trim galore. A wrapper tool around Cutadapt and FastQC to consistently apply quality and adapter trimming to FastQ files. 2015;516(517):517. [Google Scholar]
140.Kopylova E, Noé L, Touzet H. SortMeRNA: fast and accurate filtering of ribosomal RNAs in metatranscriptomic data. Bioinformatics. 2012;28(24):3211–7. [DOI] [PubMed] [Google Scholar]
141.Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29(1):15–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
142.Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014;30(7):923–30. [DOI] [PubMed] [Google Scholar]
143.Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:1–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
144.Wu T, Hu E, Xu S, Chen M, Guo P, Dai Z. clusterProfiler 4.0: a universal enrichment tool for interpreting omics data. Innovation. 2021. 10.1016/j.xinn.2021.100141. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

40168_2026_2433_MOESM1_ESM.pdf^{(32.8MB, pdf)}

40168_2026_2433_MOESM2_ESM.zip^{(26.4MB, zip)}

Data Availability Statement

[CR1] 1.Oh J, Byrd AL, Deming C, Conlan S, Barnabas B, Blakesley R. Biogeography and individuality shape function in the human skin metagenome. Nature. 2014;514(7520):59–64. 10.1038/nature13786. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR2] 2.Grice EA, Segre JA. The skin microbiome. Nat Rev Microbiol. 2011;9(4):244–53. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR3] 3.Kong HH. Skin microbiome: genomics-based insights into the diversity and role of skin microbes. Trends Mol Med. 2011;17(6):320–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR4] 4.Somerville DA. The normal flora of the skin in different age groups. Br J Dermatol. 1969;81(4):248–58. [DOI] [PubMed] [Google Scholar]

[CR5] 5.Leyden JJ, McGiley KJ, Mills OH, Kligman AM. Age-related changes in the resident bacterial flora of the human face. J Invest Dermatol. 1975;65(4):379–81. [DOI] [PubMed] [Google Scholar]

[CR6] 6.Oh J, Freeman AF, Park M, Sokolic R, Candotti F, Holland SM, et al. The altered landscape of the human skin microbiome in patients with primary immunodeficiencies. Genome Res. 2013;23(12):2103–14. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR7] 7.Kong HH, Oh J, Deming C, Conlan S, Grice EA, Beatson MA, et al. Temporal shifts in the skin microbiome associated with disease flares and treatment in children with atopic dermatitis. Genome Res. 2012;22(5):850–9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR8] 8.Williams MR, Gallo RL. The role of the skin microbiome in atopic dermatitis. Curr Allergy Asthma Rep. 2015;15:1–10. [DOI] [PubMed] [Google Scholar]

[CR9] 9.Holmes AD. Potential role of microorganisms in the pathogenesis of rosacea. J Am Acad Dermatol. 2013;69(6):1025–32. [DOI] [PubMed] [Google Scholar]

[CR10] 10.Chaudhary M, Khan A, Gupta M. Skin ageing: pathophysiology and current market treatment approaches. Curr Aging Sci. 2020;13(1):22–30. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR11] 11.Amaro-Ortiz A, Yan B, D’Orazio JA. Ultraviolet radiation, aging and the skin: prevention of damage by topical cAMP manipulation. Molecules. 2014;19(5):6202–19. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR12] 12.Wilhelm K-P, Cua AB, Maibach HI. Skin aging: effect on transepidermal water loss, stratum corneum hydration, skin surface pH, and casual sebum content. Arch Dermatol. 1991;127(12):1806–9. [DOI] [PubMed] [Google Scholar]

[CR13] 13.Potts RO, Buras EM, Chrisman DA. Changes with age in the moisture content of human skin. J Investig Dermatol. 1984;82(1):97–100. 10.1111/1523-1747.ep12259203. [DOI] [PubMed] [Google Scholar]

[CR14] 14.Cotterill JA, Cunliffe WJ, Williamson B, Bulusu L. Age and sex variation in skin surface lipid composition and sebum excretion rate. Br J Dermatol. 1972;87(4):333–40. [DOI] [PubMed] [Google Scholar]

[CR15] 15.Pochi PE, Strauss JS, Downing DT. Age-related changes in sebaceous gland activity. J Investig Dermatol. 1979;73(1):108–11. [DOI] [PubMed] [Google Scholar]

[CR16] 16.Xia Jj, Zhong Q, Li Zm, Wei Qz, Jiang L, Duan C, et al. Culture dependent and independent approaches reveal the role of specific bacteria in human skin aging. iMetaOmics. 2024;1(2): e26. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR17] 17.Kim J-H, Son S-M, Park H, Kim BK, Choi IS, Kim H, et al. Taxonomic profiling of skin microbiome and correlation with clinical skin parameters in healthy Koreans. Sci Rep. 2021;11(1):16269. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR18] 18.Kim H-J, Oh HN, Park T, Kim H, Lee HG, An S. Aged related human skin microbiome and mycobiome in Korean women. Sci Rep. 2022;12(1):2351. 10.1038/s41598-022-06189-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR19] 19.Wu L, Zeng T, Deligios M, Milanesi L, Langille MGI, Zinellu A, et al. Age-related variation of bacterial and fungal communities in different body habitats across the young, elderly, and centenarians in Sardinia. mSphere. 2020;5(1):e00558-e619. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR20] 20.Zhou W, Fleming E, Legendre G, Roux L, Latreille J, Gendronneau G. Skin microbiome attributes associate with biophysical skin ageing. Exp Dermatol. 2023. 10.1111/exd.14863. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR21] 21.Alkema W, Boekhorst J, Eijlander RT, Schnittger S, De Gruyter F, Lukovac S, et al. Charting host-microbe co-metabolism in skin aging and application to metagenomics data. PLoS One. 2021;16(11):e0258960. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR22] 22.Howard B, Bascom CC, Hu P, Binder RL, Fadayel G, Huggins TG. Aging-associated changes in the adult human skin microbiome and the host factors that affect skin microbiome composition. J Investig Dermatol. 2022;142(7):1934-46.e21. 10.1016/j.jid.2021.11.029. [DOI] [PubMed] [Google Scholar]

[CR23] 23.Kim M, Park T, Yun JI, Lim HW, Han NR, Lee ST. Investigation of age-related changes in the skin microbiota of Korean women. Microorganisms. 2020;8(10): 1581. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR24] 24.Larson PJ, Zhou W, Santiago A, Driscoll S, Fleming E, Voigt AY, et al. Associations of the skin, oral and gut microbiome with aging, frailty and infection risk reservoirs in older adults. Nat Aging. 2022;2(10):941–55. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR25] 25.Kim H-J, Kim JJ, Myeong NR, Kim T, Kim D, An S. Segregation of age-related skin microbiome characteristics by functionality. Sci Rep. 2019;9(1):16748. 10.1038/s41598-019-53266-3. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR26] 26.Shibagaki N, Suda W, Clavaud C, Bastien P, Takayasu L, Iioka E. Aging-related changes in the diversity of women’s skin microbiomes associated with oral bacteria. Sci Rep. 2017;7(1):10567. 10.1038/s41598-017-10834-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR27] 27.Reilly DM, Lozano J. Skin collagen through the lifestages: importance for skin health and beauty. Plast Aesthet Res. 2021;8: N-A. [Google Scholar]

[CR28] 28.Alexis AF, Obioha JO. Ethnicity and aging skin. J Drugs Dermatol. 2017;16(6):s77–80. [PubMed] [Google Scholar]

[CR29] 29.Shen X, Wang C, Zhou X, Zhou W, Hornburg D, Wu S. Nonlinear dynamics of multi-omics profiles during human aging. Nat Aging. 2024. 10.1038/s43587-024-00692-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR30] 30.Zhou E, Cao Z, Yin Q. Naive-deep face recognition: Touching the limit of LFW benchmark or not? arXiv preprint arXiv:1501.04690. 2015. https://arxiv.org/abs/1501.04690.

[CR31] 31.Papaccio F, Arino A, Caputo D′S, Bellei B. Focus on the contribution of oxidative stress in skin aging. Antioxidants Basel. 2022;11(6):1121. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR32] 32.Li Z, Xia J, Jiang L, Tan Y, An Y, Zhu X, et al. Characterization of the human skin resistome and identification of two microbiota cutotypes. Microbiome. 2021;9:1–18. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR33] 33.Ahle CM, Stødkilde K, Poehlein A, Bömeke M, Streit WR, Wenck H, et al. Interference and co-existence of staphylococci and Cutibacterium acnes within the healthy human skin microbiome. Commun Biol. 2022;5(1):923. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR34] 34.Kimoto‐Nira H. New lactic acid bacteria for skin health via oral intake of heat‐killed or live cells. Anim Sci J. 2018;89(6):835–42. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR35] 35.Yoshida A, Aoki R, Kimoto-Nira H, Kobayashi M, Kawasumi T, Mizumachi K. Oral administration of live Lactococcus lactis C59 suppresses IgE antibody production in ovalbumin-sensitized mice via the regulation of interleukin-4 production. FEMS Immunol Med Microbiol. 2011;61(3):315–22. [DOI] [PubMed] [Google Scholar]

[CR36] 36.Kimoto H, Mizumachi K, Okamoto T, Kurisaki J. New Lactococcus strain with immunomodulatory activity: enhancement of Th1‐type immune response. Microbiol Immunol. 2004;48(2):75–82. [DOI] [PubMed] [Google Scholar]

[CR37] 37.Kimoto-Nira H, Aoki R, Sasaki K, Suzuki C, Mizumachi K. Oral intake of heat-killed cells of Lactococcus lactis strain H61 promotes skin health in women. J Nutr Sci. 2012;1:e18. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR38] 38.Ramsey JP, Mercurio A, Holland JA, Harris RN, Minbiole KPC. The cutaneous bacterium Janthinobacterium lividum inhibits the growth of Trichophyton rubrum in vitro. Int J Dermatol. 2015;54(2):156–9. [DOI] [PubMed] [Google Scholar]

[CR39] 39.Brooke JS. Advances in the microbiology of Stenotrophomonas maltophilia. Clin Microbiol Rev. 2021;34(3):10–1128. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR40] 40.An S-q, Berg G. Stenotrophomonas maltophilia. Trends Microbiol. 2018;26(7):637–8. [DOI] [PubMed] [Google Scholar]

[CR41] 41.Mukherjee P, Roy P. Genomic potential of Stenotrophomonas maltophilia in bioremediation with an assessment of its multifaceted role in our environment. Front Microbiol. 2016;7:159261. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR42] 42.Gonzalez JM, Aranda B. Microbial growth under limiting conditions-future perspectives. Microorganisms. 2023;11: 1641. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR43] 43.Merchant SS, Helmann JD. Elemental economy: microbial strategies for optimizing growth in the face of nutrient limitation. Adv Microb Physiol. 2012;60:91–210. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR44] 44.Nüse B, Holland T, Rauh M, Gerlach RG, Mattner J. L-arginine metabolism as pivotal interface of mutual host–microbe interactions in the gut. Gut Microbes. 2023;15(1): 2222961. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR45] 45.Roux P-F, Oddos T, Stamatas G. Deciphering the role of skin surface microbiome in skin health: an integrative multiomics approach reveals three distinct metabolite‒microbe clusters. J Invest Dermatol. 2022;142(2):469–79. [DOI] [PubMed] [Google Scholar]

[CR46] 46.Saini R, Badole SL, Zanwar AA. Arginine derived nitric oxide: Key to healthy skin. In: Bioactive dietary factors and plant extracts in dermatology. 2013. p. 73–82.

[CR47] 47.He J, Fang B, Shan S, Li Q. Mechanical stiffness promotes skin fibrosis through Piezo1-mediated arginine and proline metabolism. Cell Death Discov. 2023;9(1): 354. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR48] 48.Patriarca EJ, Cermola F, D’Aniello C, Fico A, Guardiola O, De Cesare D, et al. The multifaceted roles of proline in cell behavior. Front Cell Dev Biol. 2021;9:728576. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR49] 49.Solano F. Metabolism and functions of amino acids in the skin. In: Amino Acids in Nutrition and Health: Amino acids in systems function and health. 2020. p. 187–99. [DOI] [PubMed]

[CR50] 50.Bach TMH, Takagi H. Properties, metabolisms, and applications of L-proline analogues. Appl Microbiol Biotechnol. 2013;97:6623–34. [DOI] [PubMed] [Google Scholar]

[CR51] 51.Li Z, Bai X, Peng T, Yi X, Luo L, Yang J, et al. New insights into the skin microbial communities and skin aging. Front Microbiol. 2020;11: 565549. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR52] 52.Chambers ES, Vukmanovic‐Stejic M. Skin barrier immunity and ageing. Immunology. 2020;160(2):116–25. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR53] 53.Prescott SL, Larcombe D-L, Logan AC, West C, Burks W, Caraballo L. The skin microbiome: impact of modern environments on skin ecology, barrier integrity, and systemic immune programming. World Allergy Organ J. 2017;10(1):16. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR54] 54.Rayan M, Sayed TS, Hussein OJ, Therachiyil L, Maayah ZH, Maccalli C, et al. Unlocking the secrets: exploring the influence of the aryl hydrocarbon receptor and microbiome on cancer development. Cell Mol Biol Lett. 2024;29(1):33. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR55] 55.Dempsey JL, Cui JY. Microbiome is a functional modifier of P450 drug metabolism. Curr Pharmacol Rep. 2019;5(6):481–90. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR56] 56.Lee HU, McPherson ZE, Tan B, Korecka A, Pettersson S. Host-microbiome interactions: the aryl hydrocarbon receptor and the central nervous system. J Mol Med. 2017;95:29–39. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR57] 57.Tiedt O, Mergelsberg M, Eisenreich W, Boll M. Promiscuous defluorinating enoyl-CoA hydratases/hydrolases allow for complete anaerobic degradation of 2-fluorobenzoate. Front Microbiol. 2017;8: 322966. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR58] 58.Murphy CD, Quirke S, Balogun O. Degradation of fluorobiphenyl by Pseudomonas pseudoalcaligenes KF707. FEMS Microbiol Lett. 2008;286(1):45–9. [DOI] [PubMed] [Google Scholar]

[CR59] 59.Bhanot V, Pali S, Panwar J. Understanding the in silico aspects of bacterial catabolic cascade for styrene degradation. Proteins. 2023;91(4):532–41. [DOI] [PubMed] [Google Scholar]

[CR60] 60.Guan N, Li J, Shin H-d, Du G, Chen J, Liu L. Microbial response to environmental stresses: from fundamental mechanisms to practical applications. Appl Microbiol Biotechnol. 2017;101:3991–4008. [DOI] [PubMed] [Google Scholar]

[CR61] 61.Biernacki M, Conde T, Stasiewicz A, Surażyński A, Domingues MR, Domingues P. Restorative effect of microalgae Nannochloropsis oceanica lipid extract on phospholipid metabolism in keratinocytes exposed to UVB radiation. Int J Mol Sci. 2023;24:18 14323. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR62] 62.Del Favero G, Janker L, Neuditschko B, Hohenbichler J, Kiss E, Woelflingseder L, et al. Exploring the dermotoxicity of the mycotoxin deoxynivalenol: combined morphologic and proteomic profiling of human epidermal cells reveals alteration of lipid biosynthesis machinery and membrane structural integrity relevant for skin barrier function. Arch Toxicol. 2021;95(6):2201–21. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR63] 63.Firmansyah D, Sumiwi SA, Saptarini NM, Levita J. Curcuma longa extract inhibits the activity of mushroom tyrosinase and the growth of murine skin cancer B16F10 cells. J Herbmed Pharmacol. 2022;12(1):153–8. [Google Scholar]

[CR64] 64.Goenka S, Simon SR. Comparative study of curcumin and its hydrogenated metabolites, tetrahydrocurcumin, hexahydrocurcumin, and octahydrocurcumin, on melanogenesis in B16F10 and MNT-1 cells. Cosmetics. 2021;8(1): 4. [Google Scholar]

[CR65] 65.Riaz R, Batool S, Zucca P, Rescigno A, Peddio S, Saleem RSZ. Plants as a promising reservoir of tyrosinase inhibitors. Mini Rev Org Chem. 2021;18(6):808–28. [Google Scholar]

[CR66] 66.Guo L, Jin H. Research progress of metabolomics in psoriasis. Chin Med J. 2023;136(15):1805–16. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR67] 67.Koussiouris J, Looby N, Anderson M, Kulasingam V, Chandran V. Metabolomics studies in psoriatic disease: a review. Metabolites. 2021;11(6): 375. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR68] 68.Sharma D, Arora S, dos Santos Rodrigues B, Lakkadwala S, Banerjee A, Singh J. Chitosan-based systems for gene delivery. In: Functional Chitosan: Drug Delivery and Biomedical Applications. 2019. p. 229–67.

[CR69] 69.Seité S, Rougier A. Lipohydroxy acid containing shampoo in the treatment of scalp seborrheic dermatitis. In: Handbook of hair in health and disease. Wageningen Academic; 2011. p. 466–76.

[CR70] 70.Uhoda E, Pierard C, Pierard G. Comedolysis by a lipohydroxyacid formulation in acne-prone subjects. European J Dermatol. 2003;13(1):65–8. [PubMed]

[CR71] 71.Avila-Camacho M, Montastier C, Piérard GE. Histometric assessment of the age-related skin response to 2-hydroxy-5-octanoyl benzoic acid. Skin Pharmacol Physiol. 1998;11(1):52–6. [DOI] [PubMed] [Google Scholar]

[CR72] 72.Russo E, Di Gloria L, Cerboneschi M, Smeazzetto S, Baruzzi GP, Romano F, et al. Facial skin microbiome: aging-related changes and exploratory functional associations with host genetic factors, a pilot study. Biomedicines. 2023;11(3): 684. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR73] 73.Homma T, Fujii J. Application of glutathione as anti-oxidative and anti-aging drugs. Curr Drug Metab. 2015;16(7):560–71. [DOI] [PubMed] [Google Scholar]

[CR74] 74.Foyer CH, Noctor G. Ascorbate and glutathione: the heart of the redox hub. Plant Physiol. 2011;155(1):2–18. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR75] 75.Li L-H, Shih Y-L, Huang J-Y, Wu C-J, Huang Y-W, Huang H-H, et al. Protection from hydrogen peroxide stress relies mainly on AhpCF and KatA2 in Stenotrophomonas maltophilia. J Biomed Sci. 2020;27:1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR76] 76.Li L-H, Wu C-M, Lin Y-T, Pan S-Y, Yang T-C. Roles of FadRACB system in formaldehyde detoxification, oxidative stress alleviation and antibiotic susceptibility in Stenotrophomonas maltophilia. J Antimicrob Chemother. 2020;75(8):2101–9. [DOI] [PubMed] [Google Scholar]

[CR77] 77.Liao C-H, Ku R-H, Li L-H, Wu C-M, Yang T-C. Role of yceA-cybB-yceB operon in oxidative stress tolerance, swimming motility and antibiotic susceptibility of Stenotrophomonas maltophilia. J Antimicrob Chemother. 2023;78(8):1891–9. [DOI] [PubMed] [Google Scholar]

[CR78] 78.Dimri GP, Lee X, Basile G, Acosta M, Scott G, Roskelley C. A biomarker that identifies senescent human cells in culture and in aging skin in vivo. Proc Natl Acad Sci U S A. 1995;92(20):9363–7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR79] 79.Sikalidis AK, Mazor KM, Lee J-I, Roman HB, Hirschberger LL, Stipanuk MH. Upregulation of capacity for glutathione synthesis in response to amino acid deprivation: regulation of glutamate–cysteine ligase subunits. Amino Acids. 2014;46:1285–96. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR80] 80.Kurata M, Suzuki M, Agar NS. Glutathione regeneration in mammalian erythrocytes. Comp Haematol Int. 2000;10:59–67. [Google Scholar]

[CR81] 81.Wiley CD, Campisi J. The metabolic roots of senescence: mechanisms and opportunities for intervention. Nat Metab. 2021;3(10):1290–301. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR82] 82.Nishida-Tamehiro K, Kimura A, Tsubata T, Takahashi S, Suzuki H. Antioxidative enzyme NAD (P) H quinone oxidoreductase 1 (NQO1) modulates the differentiation of Th17 cells by regulating ROS levels. PLoS One. 2022;17(7):e0272090. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR83] 83.Yu Y, Di Trapani G, Tonissen KF. Thioredoxin and glutathione systems: cancer cells’ defensive weapons against oxidative stress. In: Handbook of Oxidative Stress in Cancer: Mechanistic Aspects. Springer; 2022. p. 2407–20.

[CR84] 84.Scaramuzzino L, Lucchino V, Scalise S, Lo Conte M, Zannino C, Sacco A, et al. Uncovering the metabolic and stress responses of human embryonic stem cells to FTH1 gene silencing. Cells. 2021;10(9):2431. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR85] 85.Suryadevara V, Hudgins AD, Rajesh A, Pappalardo A, Karpova A, Dey AK, et al. SenNet recommendations for detecting senescent cells in different tissues. Nat Rev Mol Cell Biol. 2024. 10.1038/s41580-024-00738-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR86] 86.Wang L, Tao Y, Zhang J, Wang X, Zhou S, He L, et al. APCDD1 as a co-receptor positively regulates Wnt5a/c-Jun non-canonical signaling pathway. Journal of Shanghai Jiaotong University (Science). 2019;24:510–6. [Google Scholar]

[CR87] 87.Herrera J, Henke CA, Bitterman PB. Extracellular matrix as a driver of progressive fibrosis. J Clin Invest. 2018;128(1):45–53. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR88] 88.Boysen AK, Durham BP, Kumler W, Key RS, Heal KR, Carlson LT. Glycine betaine uptake and metabolism in marine microbial communities. Environ Microbiol. 2022;24(5):2380–403. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR89] 89.Hwang H, Chun H, Kim D, Shin M, Kim YS, In S. Lysophosphatidylcholine exerts an anti-skin photoaging effect via heat shock protein 70 induction. J Cosmet Dermatol. 2021;20(12):4060–7. [DOI] [PubMed] [Google Scholar]

[CR90] 90.Stapleton CM, Mashek DG, Wang S, Nagle CA, Cline GW, Thuillier P, et al. Lysophosphatidic acid activates peroxisome proliferator activated receptor-γ in CHO cells that over-express glycerol 3-phosphate acyltransferase-1. PLoS One. 2011;6(4):e18932. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR91] 91.Artegoitia VM, Middleton JL, Harte FM, Campagna SR, De Veth MJ. Choline and choline metabolite patterns and associations in blood and milk during lactation in dairy cows. PLoS One. 2014;9(8):e103412. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR92] 92.Lee H-R, Kwon S-Y, Choi S-A, Lee J-H, Lee H-S, Park J-B. Valorization of soy lecithin by enzyme cascade reactions including a phospholipase A2, a fatty acid double-bond hydratase, and/or a photoactivated decarboxylase. J Agric Food Chem. 2022;70(35):10818–25. [DOI] [PubMed] [Google Scholar]

[CR93] 93.Melani NB, Tambourgi EB, Silveira E. Lipases: from production to applications. Sep Purif Rev. 2020;49(2):143–58. [Google Scholar]

[CR94] 94.Stroupe ME, Leech HK, Daniels DS, Warren MJ, Getzoff ED. CysG structure reveals tetrapyrrole-binding features and novel regulation of siroheme biosynthesis. Nat Struct Mol Biol. 2003;10(12):1064–73. [DOI] [PubMed] [Google Scholar]

[CR95] 95.Warren MJ, Raux E, Schubert HL, Escalante-Semerena JC. The biosynthesis of adenosylcobalamin (vitamin B12). Nat Prod Rep. 2002;19(4):390–412. [DOI] [PubMed] [Google Scholar]

[CR96] 96.Kang D, Shi B, Erfe MC, Craft N, Li H. Vitamin B12 modulates the transcriptome of the skin microbiota in acne pathogenesis. Sci Transl Med. 2015;7:293 293ra103 293ra103. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR97] 97.Baier J, Maisch T, Maier M, Engel E, Landthaler M, Bäumler W. Singlet oxygen generation by UVA light exposure of endogenous photosensitizers. Biophys J. 2006;91(4):1452–9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR98] 98.Altun I, Kurutaş EB. Vitamin B complex and vitamin B12 levels after peripheral nerve injury. Neural Regen Res. 2016;11(5):842–5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR99] 99.Mori K, Ando I, Kukita A. Generalized hyperpigmentation of the skin due to vitamin B12 deficiency. J Dermatol. 2001;28(5):282–5. [DOI] [PubMed] [Google Scholar]

[CR100] 100.Muñoz P, Huenchuguala S, Paris I, Segura-Aguilar J. Dopamine oxidation and autophagy. Parkinson Dis. 2012;2012(1): 920953. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR101] 101.Agarwal P, Singh M, Singh J, Singh RP. Microbial tyrosinases: A novel enzyme, structural features, and applications. In: Applied microbiology and bioengineering. Elsevier; 2019. p. 3–19.

[CR102] 102.Muñoz P, Huenchuguala S, Paris I, Segura-Aguilar J. Dopamine oxidation and autophagy. Parkinsons Dis. 2012. 10.1155/2012/920953. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR103] 103.Frases S, Chaskes S, Dadachova E, Casadevall A. Induction by Klebsiella aerogenes of a melanin-like pigment in Cryptococcus neoformans. Appl Environ Microbiol. 2006;72(2):1542–50. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR104] 104.Kourtis N, Tavernarakis N. Cellular stress response pathways and ageing: intricate molecular relationships. EMBO J. 2011;30(13):2520–31. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR105] 105.Boyajian JL, Ghebretatios M, Schaly S, Islam P, Prakash S. Microbiome and human aging: probiotic and prebiotic potentials in longevity, skin health and cellular senescence. Nutrients. 2021;13(12): 4550. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR106] 106.Townsend EC, Kalan LR. The dynamic balance of the skin microbiome across the lifespan. Biochem Soc Trans. 2023;51(1):71–86. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR107] 107.Tan W, Zhang J, Zhao X, Li Q, Dong F, Guo Z. Preparation and physicochemical properties of antioxidant chitosan ascorbate/methylcellulose composite films. Int J Biol Macromol. 2020;146:53–61. [DOI] [PubMed] [Google Scholar]

[CR108] 108.Rahaman MM, Rakib A, Mitra S, Tareq AM, Emran TB, Shahid-Ud-Daula AFM, et al. The genus Curcuma and inflammation: Overview of the pharmacological perspectives. Plants. 2020;10(1):63. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR109] 109.Scapagnini G, Davinelli S, Renzo Di L, Lorenzo De A, Olarte HH, Micali G. Cocoa bioactive compounds: Significance and potential for the maintenance of skin health. Nutrients. 2014;6(8):3202–13. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR110] 110.Wishart DS, Guo A, Oler E, Wang F, Anjum A, Peters H. HMDB 5.0: The human metabolome database for 2022. Nucleic Acids Res. 2022;50(D1):D622–31. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR111] 111.O’Donnell VB, Dennis EA, Wakelam MJO, Subramaniam S. LIPID MAPS: Serving the next generation of lipid researchers with tools, resources, data, and training. Sci Signal. 2019;12(563):eaaw2964. [DOI] [PubMed] [Google Scholar]

[CR112] 112.Guijas C, Montenegro-Burke JR, Domingo-Almenara X, Palermo A, Warth B, Hermann G. METLIN: A technology platform for identifying knowns and unknowns. Anal Chem. 2018;90(5):3156–64. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR113] 113.Carvalho S, Leite J, Galdo-Álvarez S, Gonçalves OF. The emotional movie database (EMDB): A self-report and psychophysiological study. Appl Psychophysiol Biofeedback. 2012;37:279–94. [DOI] [PubMed] [Google Scholar]

[CR114] 114.Castrignano T, Meo De PDO, Cozzetto D, Talamo IG, Tramontano A. The PMDB protein model database. Nucleic Acids Res. 2006;34(suppl_1):D306–9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR115] 115.Kieser S, Brown J, Zdobnov EM, Trajkovski M, McCue LA. ATLAS: a Snakemake workflow for assembly, annotation, and genomic binning of metagenome sequence data. BMC Bioinformatics. 2020;21:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR116] 116.Bushnell B. BBMap: a fast, accurate, splice-aware aligner. Berkeley, CA, USA: Technical report, Lawrence Berkeley National Laboratory; 2014. https://www.osti.gov/servlets/purl/1241166.

[CR117] 117.Li D, Luo R, Liu C-M, Leung C-M, Ting H-F, Sadakane K, et al. MEGAHIT v1. 0: a fast and scalable metagenome assembler driven by advanced methodologies and community practices. Methods. 2016;102:3–11. [DOI] [PubMed] [Google Scholar]

[CR118] 118.Kang DD, Li F, Kirton E, Thomas A, Egan R, An H, et al. MetaBAT 2: an adaptive binning algorithm for robust and efficient genome reconstruction from metagenome assemblies. PeerJ. 2019;7: e7359. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR119] 119.Wu Y-W, Simmons BA, Singer SW. MaxBin 2.0: an automated binning algorithm to recover genomes from multiple metagenomic datasets. Bioinformatics. 2016;32(4):605–7. [DOI] [PubMed] [Google Scholar]

[CR120] 120.Sieber CMK, Probst AJ, Sharrar A, Thomas BC, Hess M, Tringe SG, et al. Recovery of genomes from metagenomes via a dereplication, aggregation and scoring strategy. Nat Microbiol. 2018;3(7):836–43. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR121] 121.Chklovski A, Parks DH, Woodcroft BJ, Tyson GW. CheckM2: a rapid, scalable and accurate tool for assessing microbial genome quality using machine learning. Nat Methods. 2023;20(8):1203–12. [DOI] [PubMed] [Google Scholar]

[CR122] 122.Chaumeil P-A, Mussig AJ, Hugenholtz P, Parks DH. GTDB-Tk v2: memory friendly classification with the genome taxonomy database. Bioinformatics. 2022;38(23):5315–6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR123] 123.Letunic I, Bork P. Interactive Tree of Life (iTOL) v6: recent updates to the phylogenetic tree display and annotation tool. Nucleic Acids Res. 2024. 10.1093/nar/gkae268. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR124] 124.Ruscheweyh H-J, Milanese A, Paoli L, Karcher N, Clayssen Q, Keller MI, et al. Cultivation-independent genomes greatly expand taxonomic-profiling capabilities of mOTUs across various environments. Microbiome. 2022;10(1):212. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR125] 125.Beghini F, McIver LJ, Blanco-Míguez A, Dubois L, Asnicar F, Maharjan S, et al. Integrating taxonomic, functional, and strain-level profiling of diverse microbial communities with bioBakery 3. Elife. 2021;10: e65088. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR126] 126.Kanehisa M. The KEGG database. Novartis Found Symp. 2002;247:91–101; discussion 101–103, 119–128, 244–252. https://onlinelibrary.wiley.com/doi/10.1002/0470857897.ch8. [PubMed]

[CR127] 127.Patil KR, Nielsen J. Uncovering transcriptional regulation of metabolism by using metabolic network topology. Proc Natl Acad Sci U S A. 2005;102(8):2685–9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR128] 128.Liaw A, Wiener M. Classification and regression by randomForest. R News. 2002;2(3):18–22. https://journal.r-project.org/articles/RN-2002-022/RN-2002-022.pdf.

[CR129] 129.Kuhn M. Building predictive models in R using the Caret package. J Stat Softw. 2008;28(1):26. [Google Scholar]

[CR130] 130.Robin X, Turck N, Hainard A, Tiberti N, Lisacek F, Sanchez J-C, et al. pROC: an open-source package for R and S+ to analyze and compare ROC curves. BMC Bioinformatics. 2011;12:1–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR131] 131.Wilkinson L, Wickham H. ggplot2: elegant graphics for data analysis. Oxford University Press; 2011. [Google Scholar]

[CR132] 132.Zimmermann J, Kaleta C, Waschina S. Gapseq: informed prediction of bacterial metabolic pathways and reconstruction of accurate metabolic models. Genome Biol. 2021;22:1–35. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR133] 133.Heirendt L, Arreckx S, Pfau T, Mendoza SN, Richelle A, Heinken A, et al. Creation and analysis of biochemical constraint-based models using the COBRA Toolbox v. 3.0. Nat Protoc. 2019;14(3):639–702. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR134] 134.Gurobi Optimization, LLC. Gurobi Optimizer Reference Manual. 2021. https://www.gurobi.com.

[CR135] 135.Revelle W. psych: Procedures for Personality and Psychological Research. Northwestern University, Evanston, Illinois. R package. 2015. https://CRAN.R-project.org/package=psych.

[CR136] 136.Reback J, McKinney W, Van Den Bossche J, Augspurger T, Cloud P, Klein A, et al. pandas-dev/pandas: Pandas 1.0.5. Zenodo. 2020. 10.5281/zenodo.3898987.

[CR137] 137.Slowikowski K. ggrepel: Automatically position non-overlapping text labels with ‘ggplot2’. R package version 0.9.1. 2021. https://CRAN.R-project.org/package=ggrepel.

[CR138] 138.Andrews S. FastQC: a quality control tool for high throughput sequence data. Babraham Bioinformatics. 2010. https://www.bioinformatics.babraham.ac.uk/projects/fastqc/.

[CR139] 139.Krueger F. Trim galore. A wrapper tool around Cutadapt and FastQC to consistently apply quality and adapter trimming to FastQ files. 2015;516(517):517. [Google Scholar]

[CR140] 140.Kopylova E, Noé L, Touzet H. SortMeRNA: fast and accurate filtering of ribosomal RNAs in metatranscriptomic data. Bioinformatics. 2012;28(24):3211–7. [DOI] [PubMed] [Google Scholar]

[CR141] 141.Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29(1):15–21. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR142] 142.Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Bioinformatics. 2014;30(7):923–30. [DOI] [PubMed] [Google Scholar]

[CR143] 143.Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol. 2014;15:1–21. [DOI] [PMC free article] [PubMed] [Google Scholar]

[CR144] 144.Wu T, Hu E, Xu S, Chen M, Guo P, Dai Z. clusterProfiler 4.0: a universal enrichment tool for interpreting omics data. Innovation. 2021. 10.1016/j.xinn.2021.100141. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Multi-omics characterization of the skin microbiota reveals the anti-aging roles of Stenotrophomonas maltophilia

Danni Guo

Yuanyuan Chen

Yahong Wu

Jingmin Cheng

Yawen Lin

Wenjie Lai

Wentao Ma

Hang Yang

Lianyi Han

Lan Ma

Haidong Jia

Xiao Liu

Abstract

Background

Results

Conclusions

Supplementary Information

Background

Results

Distinct features across phenotypic skin aging groups

Fig. 1.

Skin metagenomics reveals microbiome dynamics associated with phenotypic skin aging

Fig. 2.

Skin metabolomics reveals characteristic metabolic signatures of phenotypic skin aging

Fig. 3.

Predictive modeling and construction of genome-scale metabolic models (GEMs)

S. maltophilia mitigates oxidative stress and supports youthful skin phenotypes

Fig. 4.

Potential roles of S. maltophilia in regulating betaine, lysolecithin, and porphyrin metabolism associated with phenotypic skin aging

Fig. 5.

Skin microbiota involvement in melanin metabolism during phenotypic skin aging

Fig. 6.

Discussion

Materials and methods

Participants and study design

Skin phenotype assessment

Grouping criteria and sample collection procedures

DNA extraction and metagenomic sequencing

Preparation of metabolomics samples and LC–MS measurement

LC–MS data processing and metabolomics analysis

Quality control and contaminant removal of metagenomic reads

Metagenomic data assembly

Taxonomic identification and functional profiling of microbial communities

Construction of Random Forest models for phenotypic skin aging prediction

Genome-scale metabolic model (GEM) construction

Metabolic network analysis

Statistical analysis and data visualization

Bacterial growth conditions and supernatant preparation

Cell culture

Stimulation of fibroblasts with bacterial supernatants

Indirect stimulation of fibroblasts by bacterial supernatant via epidermal cells

Quantification of GSH and GSSG levels in fibroblasts

β-Galactosidase staining for cellular senescence detection

DCFH-DA staining for intracellular ROS detection

Construction and treatment of full-thickness skin models

Tissue sample collection and frozen section preparation

H&E staining

Immunofluorescence staining

RNA sequencing

Processing and analysis of RNA-seq data

Supplementary Information

Acknowledgements

Authors’ contributions

Funding

Data availability

Declarations

Consent for publication

Competing interests

Footnotes

Contributor Information

References

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles