Microbiota-directed therapies for atopic dermatitis: a three-tier framework for inflammation control, immune modulation, and microbiome restoration

Abstract

The pathogenesis of atopic dermatitis (AD) involves cutaneous barrier dysfunction, immune dysregulation, and microbiota imbalance. Although microbiota-directed therapeutic approaches have garnered increasing attention, current evidence is heterogeneous, encompassing probiotics, postbiotics, microbial metabolites, local microbial interventions, antimicrobials and ecological modulation strategies. This narrative review does not aim to provide clinical guidelines, but rather seeks to synthesise existing evidence within a three-tier conceptual framework. This framework classifies interventions into three categories: those primarily addressing local inflammation and barrier-associated microbiota dysbiosis; those exerting systemic immunomodulation via gut-derived microbial signals; and those modulating the skin microbiota ecology. It should be noted that human clinical data are still limited and heterogeneous, and much mechanistic insight is derived from preclinical research. By clarifying mechanistic layers and evidentiary gaps, this framework may facilitate future research to evaluate the sequencing or combination of anti-inflammatory therapy, barrier restoration, and microbiota ecological modulation.

1 Method

This narrative review was conducted by searching PubMed, Web of Science, and Google Scholar for literature using combinations of terms including ‘atopic dermatitis,’ ‘microbiome,’ ‘microbiota,’ ‘probiotics,’ ‘postbiotics,’ ‘faecal microbiota transplantation,’ ‘skin microbiome,’ ‘Staphylococcus aureus,’ ‘short-chain fatty acids,’ and ‘gut–skin axis.’ Priority was given to human studies, randomised controlled trials, and mechanistic studies with direct relevance to atopic dermatitis, as well as recent comprehensive reviews. Preclinical findings were included when they provided mechanistic insights but were interpreted separately from clinical evidence to avoid overgeneralization.

2 Introduction

Atopic dermatitis (AD) is a chronic, relapsing, and heterogeneous inflammatory skin disorder, the pathogenesis of which is associated with local cutaneous immune dysregulation, barrier dysfunction, the pruritus–scratch vicious cycle, and microbiota imbalance (1–3). Studies have demonstrated that the onset of AD is not only related to immune-mediated inflammation, but also involves changes in the skin barrier and cutaneous microbiota. Dysbiosis of the skin microbiota may exacerbate inflammation and barrier impairment, whilst inflammation and barrier disruption can in turn alter the skin microbiota (4–7).

Currently, a variety of microbiota-directed therapies (MDTs) are available, including probiotics, prebiotics, postbiotics, antimicrobial treatments, and faecal microbiota transplantation. These interventions are highly heterogeneous; due to differences in study populations, microbial strains, dosages, treatment durations, outcome measures, and levels of evidence, it is not possible to simply determine the effectiveness or ineffectiveness of these therapies. Thus, a framework is required to organise this heterogeneous body of evidence. To clearly summarise the evidence for various MDTs, we have collated key study designs, strain characteristics, and clinical research in Tables 1, 2.

In this review, we classify various MDTs according to their mechanisms of intervention: local inflammation control, systemic immunomodulation via the gut–skin axis, and modulation of the cutaneous microbiota. These three levels are not mutually exclusive therapeutic categories, but represent the primary mechanistic orientations, with different interventions potentially involving more than one level. MDTs are defined as interventions that specifically target the composition, function, metabolites, or host interaction capacity of the skin or gut microbiota.

This narrative review does not aim to provide clinical treatment guidelines, nor to offer an exhaustive inventory of current AD therapies. Instead, it synthesises the existing evidence on microbiota interventions and organises it into a three-tier conceptual framework: local inflammation control, systemic immunomodulation, and microbiota ecological modulation. This framework aims to elucidate mechanistic differences, identify translational research gaps, and support the design of future combination or sequential therapeutic strategies.

3 Regulation of local inflammation

Clinically, topical formulations are considered the first-line local intervention for the skin. For the topical treatment of AD, the following outlines how topical formulations may synergistically intervene through three distinct strategies: immunomodulation (such as via microbiota and their metabolites), rapid anti-inflammatory action (for example, through small-molecule metabolites), anti-infective effects, and barrier restoration (such as by antimicrobial peptide modulators). Kim et al. developed a mixture comprising inactivated cells and metabolites from three skin-derived Streptococcus strains, applied directly to keratinocytes and the skin microenvironment, specifically inhibiting Th2-type inflammatory effectors including interleukin-13 (IL-13), chemokine ligand 17 (CCL17), and CCL22, thereby achieving local immunomodulation. In patients with AD, after eight weeks, 41.5% exhibited near-complete clearance of lesions, and the intervention group showed an average reduction in transepidermal water loss (TEWL) of 3.95 g/h/m² (8). Due to the absence of pathway inhibitor validation, the blockade of Th2 cell activation can only be inferred from changes in biomarkers, which may be related to the suppression of direct dendritic cell–T cell interactions by Streptococcus fermentation products.

Although postbiotic emollients have demonstrated anti-inflammatory effects when applied locally to the skin, their efficacy only becomes fully apparent after eight weeks, and these emollients typically contain macromolecules with limited skin penetration. Propionate, a microbiota-derived metabolite, can be formulated into a highly permeable propionate cream, addressing the need for rapid-response therapy. Previous studies have shown that Staphylococcus aureus (S. aureus) induces the production of IL-33, which acts as a pro-inflammatory mediator, thereby exacerbating the inflammatory response in AD (9–11). Mechanistically, propionate rapidly inhibits histone deacetylase 2/3 (HDAC2/3) in keratinocytes, enhances Aryl hydrocarbon receptor (AhR) expression, and induces AhR nuclear translocation, suppressing IL-33 transcription. In vitro studies demonstrated that propionate reversed the increase in IL-33 expression in keratinocytes after 24 hours. Importantly, in patients with mild-to-moderate AD (n = 11), after two weeks, the regional SCORAD (SCORing Atopic Dermatitis) score was significantly reduced, with marked improvements in subjective pruritus and TEWL, showing a notably faster therapeutic response than postbiotic emollients (12).

It is noteworthy that, whilst propionate inhibits IL-33 production, it does not possess direct bactericidal activity. For cases where S. aureus colonisation may lead to skin infection and barrier disruption, MSF (fermentation extract of Lactobacillus plantarum K8) can synergistically target barrier defects by enhancing skin barrier function and reducing inflammation-driving factors. In vitro studies demonstrated that MSF rapidly induced the expression of human β-defensin-2/3 (HBD2/HBD3), with sustained high expression over 48 hours, thereby strengthening barrier protection. Additionally, in pre-treated in vitro inflammation models, MSF exerted negative feedback inhibition on p38/NF-κB phosphorylation, with significant reductions in tumour necrosis factor-α (TNF-α), IL-1β, IL-6, and IL-8 inflammatory mediators (13). MSF addresses gaps in antimicrobial defence and barrier restoration. However, the effect of MSF on HBD2/3 expression in human skin tissues has not been validated. The protective and antimicrobial roles of HBD-2 in the AD barrier have been previously reported in the literature (14, 15).

This tier of intervention is limited by its short duration, localised effect, and inability to achieve long-term immune regulation. Despite observed therapeutic efficacy, evidence is largely derived from short-term studies, and the sustained reduction in pathogenic burden and durability of barrier restoration require further confirmation. Within this framework, these three topical therapies may be suited as transitional interventions for the acute, subacute, and maintenance phases in patients with mild-to-moderate AD, rather than as independent long-term strategies.

4 Systemic immune modulation

In the following text, this category includes interventions whose primary mechanism of action is not directed at the affected skin, but which may influence atopic dermatitis via intestinal metabolites, systemic immune modulation, and immune regulatory circuits. The current evidence is derived primarily from preclinical studies, whilst human data are both limited and heterogeneous.

4.1 Gut-derived metabolites: SCFAs and immune regulation

Short-chain fatty acids (SCFAs) exert an influence on AD via the gut–skin axis, with studies demonstrating that faecal SCFA concentrations and the abundance of SCFA-producing gut microbiota are closely associated with AD (16, 17). Moreover, SCFAs have been shown to promote the differentiation of regulatory T cells (Treg) and enhance the expression of cutaneous barrier proteins, highlighting the importance of SCFAs in both immunomodulation and barrier protection (18, 19). Simultaneously, in other allergic diseases such as chronic spontaneous urticaria, a reduction in the principal SCFA-producing microbiota has also been observed (20).

In murine models, oral administration of Lactobacillus spp. for one week resulted in decreased markers of intestinal inflammation (IL-6, IL-1β, S100A8/A9), accompanied by reductions in IgE and Th2-type cytokines (IL-4, IL-13), alongside a trend towards improved epidermal barrier function. This preclinical study suggests that the alleviation of gut inflammation may be involved in the Lactobacillus-mediated systemic immunomodulation and improvement of skin barrier function; however, the causal relationship along the ‘inflammation–immunity–barrier’ axis and the applicability in human populations remain to be further established (21). Mechanistically, Kim, building on findings by Maslowski in a colitis model, hypothesised that oral probiotics modulate gut microbiota composition, increase SCFA production, and subsequently activate GPR43 to exert anti-inflammatory effects (21, 22). The study by Maslowski et al. demonstrated that the SCFA–GPR43 signalling axis plays a pivotal role in the regulation of intestinal inflammation, providing a mechanistic basis for immunomodulation by probiotics through their metabolites. Previous studies have confirmed, through multilayered experiments, that butyrate induces interleukin-22 (IL-22) production via GPR41 (rather than GPR43 or GPR109A), yet in vitro experiments have shown that SCFAs (acetate, propionate, and butyrate) can all enhance the generation of CD4⁺ T cells and IL-22 (23).

On this basis, another genetic model study found that prophylactic administration of the composite probiotic preparation IRT5 suppressed contact hypersensitivity (CHS). The specific involvement of GPR43 was further confirmed by the markedly diminished or absent protective effect in GPR43 knockout mice, as evidenced by increased ear thickness, failed Treg induction, and loss of IFN-γ suppression (24). These findings suggest that the SCFA–GPR43 axis may represent a key mechanism by which probiotics modulate T cell responses and ameliorate cutaneous inflammation. Specifically, SCFAs may promote the expansion of peripherally induced Treg cells (pTreg) via binding to GPR43 on T cell surfaces, thus participating in this immunoregulatory process. It should be noted that current evidence for GPR43-dependent mechanisms is predominantly derived from CHS models; although IRT5 intervention also demonstrated cutaneous anti-inflammatory effects in AD models (24), it remains to be clarified whether these effects are similarly dependent on the SCFA–GPR43 axis.

Although GPR43 knockout abolished the therapeutic effect of IRT5, single butyrate-producing strains did not exhibit efficacy, and given the complexity of multi-strain formulations, the possibility that other strains exert auxiliary effects via GPR43-independent pathways cannot be completely excluded. Future studies employing conditional knockout models for GPR43 are required to further validate this pathway.

4.2 FMT: promising but preclinical

Preclinical studies suggest that oral administration of probiotic preparations capable of elevating intestinal SCFA levels may help alleviate cutaneous inflammation. However, intestinal colonisation by these probiotics may be relatively transient, their influence on local microbiota composition and microbial metabolite concentrations may be limited, and intervention with a single metabolite (such as butyrate) may not consistently yield stable or significant therapeutic effects (24–26).

In murine models of AD, faecal microbiota transplantation (FMT) has been employed to transfer gut microbiota from healthy donors. Following FMT, recipient mice exhibited increased intestinal SCFA levels, accompanied by alterations in Th1/Th2-associated cytokine profiles, including upregulation of IL-12 and IFN-γ, as well as downregulation of IL-4, IL-5, and IL-13, suggesting that FMT may participate in modulating the immune imbalance associated with AD-like inflammation in this model (27). However, it should be noted that both IL-10 and IL-1β decreased post-FMT, yet IL-10 levels remained significantly higher than those in the donor group (p = 0.0007), indicating that immune status had not fully reverted to that of healthy donors. Furthermore, improvements in Th1/Th2-related indices were not accompanied by complete normalisation of the Treg axis, suggesting that the impact of FMT on immune tolerance-related pathways may remain insufficient in this animal model.

In AD animal models, FMT has demonstrated multi-pathway regulatory potential, evidenced by elevated intestinal SCFA levels, upregulation of programmed death ligand 1 (PD-L1), and a shift in Th1/Th2-related cytokines towards inflammation resolution (27). Following FMT, gut microbiota diversity recovered within one week and was maintained for up to eight weeks; concurrently, elevated SCFA levels, upregulated PD-L1 (CD274) expression, and a Th1/Th2 balance favouring inflammation alleviation were observed at eight weeks post-intervention (27).

At present, research into FMT for the treatment of atopic dermatitis relies primarily on data from animal models, which suggest that it has potential anti-inflammatory effects in terms of modulating the composition of the gut microbiota and immune responses. However, these preclinical findings have not yet been sufficiently validated in terms of clinical efficacy and safety. In clinical practice, particular attention must be paid to potential safety risks associated with antibiotic pre-treatment, such as secondary infections, antibiotic resistance and microbial imbalance (28). Therefore, future clinical studies should not only systematically evaluate the therapeutic efficacy of FMT but also strengthen safety monitoring to ensure that treatment risks are minimised and patient interests are safeguarded.

4.3 Early-life microbial programming and AD risk

The aforementioned studies suggest that a sole reliance on short-term immunomodulation may be insufficient to fully explain changes in immune homeostasis over the long-term course of AD. Given that existing research suggests that the development of atopic dermatitis may be linked to immune development in early life (29). In this context, whether the early microbiota-metabolite axis plays a role in regulating AD-related immune susceptibility has emerged as a promising area of research. Consequently, a number of animal studies have begun to explore the potential role of the early microbiome and its metabolites in the development of the immune system and in AD-related immune homeostasis.

A study by Qi et al. on cross-feeding with Bifidobacterium and 2’-fucosyllactose (2’-FL) demonstrated that this intervention increases SCFA production in animal models. Furthermore, RNA sequencing of ileal Peyer’s patches (PPs) revealed that pathways related to retinol metabolism were activated. Concurrently, the abundance of tolerance-related cells, such as immature dendritic cells (DCs) and Tregs, within the PPs was also increased (30). Notably, Spearman analysis indicated a negative correlation between SCFA levels and AD inflammatory markers (IL-33, thymic stromal lymphopoietin [TSLP], etc.), suggesting that SCFAs may be involved in the modulation of AD-like inflammation via the gut microbiota. Qi et al. also observed that plasma retinoic acid (RA) levels correlated with ear skin thickening, IL-33, mast cells, TSLP, and final AD scores in the animal model (30).

Population cohort studies have also provided indirect support for the association between early-life microbiota changes and AD risk. The CHILD cohort study (n = 1115) showed that antibiotic exposure before one year of age was associated with an increased risk of AD at five years (OR=2.25, 95% CI 1.55-3.27), accompanied by a reduction in SCFA-producing microbiota at one year (31). However, this study primarily suggests an association between early antibiotic exposure, changes in SCFA-producing microbiota, and subsequent AD risk, without establishing that SCFAs directly mediate the development of immune tolerance or provide long-term protective effects. Furthermore, although SCFA and RA levels are correlated, direct evidence that SCFAs regulate the activity of retinol metabolism-related transcription factors in PPs is currently lacking. Thus, the role of the PPs–retinol metabolism pathway as a major mechanism by which the gut microbiota ameliorates AD-like inflammation requires further investigation.

In line with this, Zhou et al. reported in animal models that supplementation with Lactobacillus reuteri FN041 during late pregnancy and lactation could influence the gut microbiota of offspring through breast milk secretory immunoglobulin A (sIgA)-related pathways, activate the retinol metabolism pathway, upregulate alcohol dehydrogenase (ADH), retinal dehydrogenase (RALDH), and ALDH1A2 expression, and promote RA production (30, 32). This intervention also reduced Th2-associated cytokines (IL-4) and ear tissue levels of IL-33/TSLP. Compared with continuous intervention during late pregnancy and lactation, supplementation with the strain after weaning had a diminished effect on the suppression of Th2 inflammatory cytokines, suggesting the existence of an early window of greater susceptibility to microbiota interventions. However, this conclusion is mainly based on animal models and cannot yet be directly extrapolated to the prevention or long-term immune remodelling of human AD.

Further research indicates that neonatal skin CD301b+ DC2 cells can generate RA via RALDH2, participating in the differentiation of symbiotic microbiota-reactive Tregs and the establishment of cutaneous tolerance (33). Collectively, these findings suggest that early microbiota, SCFAs, and RA-related pathways may contribute to the regulation of cutaneous immune homeostasis. However, their causal roles, critical time windows, and long-term effects in the pathogenesis and progression of AD require further elucidation.

4.4 Probiotics/prebiotics and clinical heterogeneity

Randomised, double-blind trials in individuals at high risk of allergy have shown that supplementation with Lactobacillus rhamnosus GG (LGG) for six months after birth did not confer a preventive effect against infantile eczema (34).Similarly, the PREGRALL study, which implemented prebiotic intervention in pregnant women with a history of atopic disease starting from 20 weeks’ gestation, found no statistically significant difference in the prevalence of AD at one year of age between the intervention and placebo groups (20.2% vs 20.2%, OR = 1.01, P = 0.97) (35). In another study supplementing mothers with a history of atopic disease and their infants with Bifidobacterium bifidum PRL2010. There was no significant difference in AD incidence (4.17% vs 11.76%, P = 0.3700); however, amongst infants who developed severe AD, the probiotic group demonstrated a superior SCORAD improvement rate compared to placebo (66.91% vs 31.50%) (36).

In contrast, another randomised double-blind trial enrolling non-selected pregnant women found that perinatal supplementation (from four weeks before delivery to three months postpartum) with a multi-strain probiotic containing LGG, Lactobacillus acidophilus La-5, and Bifidobacterium animalis subsp. lactis Bb-12 resulted in a lower incidence of AD in offspring at two years of age (OR = 0.51) (37). This finding suggests that the efficacy of microbial interventions for AD prevention may be inconsistent across studies, potentially influenced by the characteristics of study populations, intervention protocols, and endpoints.

The differences observed amongst the above clinical trials may be attributable to several factors. Firstly, the timing of intervention may be critical, as postnatal supplementation and perinatal intervention target different developmental windows. Secondly, the effects of different strains or strain combinations may vary (38, 39). Some preclinical studies suggest that not all strains can replicate protective effects, and some may even exhibit divergent immunomodulatory actions depending on timing (24, 32). Thirdly, the strength of effect between single-strain and multi-strain interventions may differ, and whether there is a synergistic effect with combination preparations requires further investigation (24). In addition, individual baseline microbiota composition and host responses may also influence intervention outcomes; existing research indicates that, even when overall intervention effects are observed, those who develop AD post-intervention may possess distinct gut microbiota profiles (40). Lastly, heterogeneity in inclusion criteria, such as whether only individuals at high familial risk of allergy are enrolled, may also contribute to differences in study outcomes.

In summary, current clinical evidence regarding the prevention of AD with probiotics or prebiotics remains inconsistent. Future studies should further optimise research design with respect to timing of intervention, target populations, strain selection, and host stratification.

4.5 Molecular recognition-based interventions

Although MDTs have demonstrated some efficacy in animal models and select clinical studies, for patients with established AD or those who have missed the early intervention window, a sole reliance on overall microbiota modulation may be insufficient to achieve stable therapeutic effects. Accordingly, recent research has begun to shift from ‘microbiota composition modulation’ towards the identification of ‘microbiota–host interaction molecules’, with the aim of discovering more controllable and quantifiable therapeutic targets. The AhR is one of the most studied pathways in the regulation of cutaneous inflammation and barrier function. Topical AhR agonists have been shown in clinical trials to improve lesion severity and pruritus in AD (41).

Mechanistic studies indicate that active components of coal tar can activate AhR in keratinocytes. Once activated, AhR regulates Th2-related inflammatory responses via the AhR–nuclear factor erythroid 2-related factor 2 (NRF2)–signal transducer and activator of transcription 6 (STAT6) signalling axis. Furthermore, AhR is involved in the regulation of multiple skin barrier proteins (42). However, some exogenous AhR ligands may pose exposure-related toxicity or safety concerns, limiting their long-term application. Therefore, identifying endogenous or commensal microbiota-derived AhR ligands with greater physiological compatibility has become an important direction in the molecular research of MDTs. Metabolites produced by the gut microbiota can also interact with AhR (43, 44). Thus, the search for endogenous or symbiont-derived AhR ligands with improved physiological compatibility is a key strategy for enhancing the safety and precision of molecular therapies.

Indole-3-carbinol (I3C), a naturally occurring compound in cruciferous vegetables, is also a direct product of gut microbiota-mediated tryptophan metabolism (45, 46). Previous studies have reported that I3C exhibits anti-inflammatory effects in models of liver disease and general inflammation (45, 47–49). In AD-related research, Bifidobacterium longum CCFM1029 was found to produce I3C, which influences Th2 polarisation through the AhR pathway. Animal experiments demonstrated that oral administration of this strain increased faecal and serum I3C levels, accompanied by activation of cutaneous AhR signalling and reductions in inflammatory markers such as TSLP, IL-4, and IL-5. The AhR antagonist CH223191 was able to attenuate these effects, and exogenous I3C could partially mimic them, suggesting that the I3C–AhR axis may be involved in the modulation of AD-like inflammation by this strain. In a corresponding clinical intervention study, after eight weeks of treatment, 70.5% of participants (31/44) were classified as responders (defined by a decrease in SCORAD), who also exhibited increased serum I3C levels that were negatively correlated with SCORAD (r = –0.4652, p < 0.05); non-responders showed no significant changes in serum I3C (50). This finding suggests that serum I3C may be associated with treatment response and has potential value as a candidate biomarker. However, given that current evidence is limited to small sample sizes and the causal relationship between responders and non-responders remains unclear, whether serum I3C can serve as a reliable predictor of therapeutic efficacy requires further validation in larger and independent cohorts.

Beyond AhR-related pathways, some studies have focused on metabolite-mediated epigenetic regulation. Alpha-ketoglutarate (α-ketoglutarate, AKG), a key intermediate of the tricarboxylic acid (TCA) cycle, is also a cofactor for dioxygenases and is involved in DNA modification regulation (51–53). Studies have indicated that DNA demethylation can remodel the epigenetic status of FOXP3 to promote Treg differentiation, whilst histone methylation modifications (such as H3K9me3/H3K27me3) play significant roles in inflammation regulation (54–56). This mechanism provides an additional explanation for the effects of metabolites on immune cell differentiation. Nevertheless, evidence for AKG in AD is primarily from preclinical studies, and the specific epigenetic modifications at key gene loci such as FOXP3 remain to be fully elucidated.

In summary, molecular interventions in MDTs have provided mechanistic insights into the links between the ‘gut microbiota–metabolite–immune host–skin barrier’ axis. At present, these findings are best considered as candidate mechanisms and potential intervention directions with translational promise.

5 Microbial ecological regulation

Although AD is an inflammatory disease, cutaneous microbial dysbiosis is also a significant pathological feature, with increased colonisation by S. aureus being most characteristic and positively correlated with disease severity (57). Consequently, some MDTs focus not only on suppressing the overgrowth of S. aureus, but are also increasingly directed at supplementing or enhancing potentially protective commensal bacteria to restore the cutaneous microbial ecosystem in AD. This section primarily discusses therapeutic strategies that modulate skin microbial colonisation, impacting pathogenic bacterial burden, commensal composition, and local immune-inflammatory responses.

Studies have shown that targeting S. aureus alone provides limited clinical benefit for non-infectious AD patients (58), suggesting that increased S. aureus abundance may be more reflective of microbial imbalance and barrier dysfunction, rather than serving as a sole or direct therapeutic target. Therefore, interventions are gradually shifting from simply inhibiting S. aureus to selecting commensal strains with immunomodulatory properties. For instance, Staphylococcus cohnii 2A1(S. cohnii 2A1), derived from healthy skin, does not directly inhibit S. aureus growth in vitro, but mono-colonisation in germ-free mice alleviates skin inflammation, indicating it may modulate AD-like inflammation through mechanisms beyond direct antimicrobial activity (59). In Th2-type dermatitis models, S. cohnii 2A1 downregulated TSLP and Th2-related cytokines (IL-4, IL-13), without significantly increasing Foxp3⁺ regulatory T cells, suggesting a predominant effect on suppressing local Th2 inflammation rather than inducing Treg-mediated immune tolerance. Animal studies also showed symptom improvement within one week and maintenance of anti-inflammatory effects for three weeks without evident attenuation. Further research indicated this strain may modulate the cutaneous inflammatory milieu by promoting local glucocorticoid synthesis and activating anti-inflammatory genes such as TNFAIP3 and ATF3 (59). However, these mechanistic insights are primarily from animal models, and whether they explain the steroid-sparing effect observed in clinical MDT studies requires further validation (60, 61).

Compared to mono-strain colonisation, Myles and colleagues were the first to apply the concept of topical microbiota transplantation in a human trial (62). After 16 weeks of treatment, the paediatric cohort (n = 5) showed a significant reduction in the S. aureus/coagulase-negative staphylococci (CoNS) ratio (P < 0.05). In the overall cohort (n = 15), the SCORAD-50 response rate (>50% improvement) was 66.7% (10/15) and SCORAD-75 response rate was 40% (6/15), with clinical improvement persisting for at least four weeks—substantially exceeding the placebo effect (62).Importantly, no adverse events were reported in either adult or paediatric groups. Additionally, non-responders were associated with a persistent family history of AD into adulthood, suggesting that host genetic background may influence response to microbiota therapies, although this remains an exploratory finding.

To further evaluate the safety and anti-S. aureus efficacy of commensal-based interventions, a phase I randomised double-blind trial (NCT03151148) assessed the effects of Staphylococcus hominis A9 (ShA9) in AD patients (63). Mechanistic studies in animal models showed that ShA9 reduced mRNA expression of IL-4, IL-13, and IL-17a and decreased Th2/Th17 infiltration in S. aureus-colonised mice. Clinically, ShA9 rapidly expanded within one hour and reduced S. aureus burden on AD skin. By increasing colonisation with Staphylococcus epidermidis, a moderate inhibitory effect on S. aureus was observed within one week. Notably, in some patients, ShA9 did not completely eradicate antibiotic-resistant S. aureus strains but suppressed their virulence gene expression. While no significant improvement in Eczema Area and Severity Index (EASI) was observed in the overall analysis (P = 0.46), possibly due to rapid decline of viable bacteria at inflamed sites, safety was favourable (adverse event rate 55.6% vs 83.3%, P = 0.044), though the study duration was only one week and long-term efficacy was not assessed (63).

These clinical studies demonstrate the feasibility of modulating the cutaneous microbiota for AD, yet the inflammatory microenvironment and pathogen competition in lesional skin may still limit the delivery, survival, and colonisation of exogenous strains, presenting a key barrier to therapeutic durability. To address this, Liu et al. developed a living bacterial dressing loaded with Roseomonas mucosa (Hy@Rm), designed to enhance survival of R. mucosa in the AD skin microenvironment and improve competition with S. aureus (64). Both in vitro and in vivo studies showed that this composite material provided protection for R. mucosa: compared with free bacteria, the bacterial load in the dressing was significantly higher (approximately four orders of magnitude in vitro and 200-fold after 48 hours in vivo); Hy@Rm also facilitated R. mucosa penetration and colonisation of the epidermis. In animal models, after six weeks of treatment, robust R. mucosa colonisation was detected alongside suppression of S. aureus proliferation. Ecological modulation with Hy@Rm was associated with improved skin barrier function: at day 14, TEWL and skin hydration were significantly improved. After six weeks of treatment, the animal model showed that the thickness of the mouse epidermis had nearly returned to normal levels, and the mouse dermatitis score was approximately 3, which was markedly lower than that of the free-bacteria group (approximately 7) (64).

This material provides a strain delivery platform for ecological modulation strategies, markedly enhancing the persistence of colonisation in AD skin and demonstrating translational potential. Key unresolved issues include the long-term survival of R. mucosa post-treatment, mechanisms of sustained efficacy, and the long-term biocompatibility and immunogenicity of the material’s degradation products.

It should be noted that although some microbial interventions have improved inflammatory or barrier-related indices, there is still a lack of direct evidence that the resultant low-inflammatory state can stably support subsequent combined or sequential microbiota therapies. Therefore, the integration of microbial interventions into long-term AD management remains best considered as a framework for future validation. Based on current research, a three-stage sequential intervention model may be proposed: the first stage focuses on acute local inflammation control to reduce pathogen burden and inflammatory activity; the second stage may explore combination with immunomodulatory therapy to improve host immune status; and the third stage emphasises maintenance of barrier and microbiota stability as a potential adjunct for long-term management (Figure 1).

Figure 1

Mechanistically, this framework has some theoretical basis. Firstly, reducing S. aureus-associated inflammatory burden may help decrease infection-related exacerbations or treatment interruptions (62, 63); Secondly, microbial interventions may stabilise the local immune microenvironment via modulation of inflammatory gene expression, potentially supporting the use of biologics or Janus kinase (JAK) inhibitors. However, whether this approach can reduce drug dosages or improve long-term response remains to be clinically validated. Similar combination strategies have been explored in other diseases, such as the superior efficacy and safety of rifaximin combined with probiotics versus monotherapy in irritable bowel syndrome (IBS) (65). Thirdly, if microbial interventions can reduce TEWL and improve barrier function, they may theoretically stabilise the local treatment environment and even affect topical drug delivery and utilisation.

Overall, this level of MDTs primarily focuses on cutaneous ecological modulation, aiming to reduce pathogenic burden, promote colonisation by protective commensals, and improve local inflammation and barrier function as adjuncts in AD therapy. Current preclinical and early clinical trials indicate feasibility and safety, but challenges remain for clinical translation, including insufficient persistence of exogenous strain colonisation, unclear long-term safety, substantial patient-to-patient variability in baseline microbiota, the need for optimised delivery systems, and lack of consensus on strain selection and quality control standards. Thus, ecological modulation strategies are best regarded as candidate adjuncts for long-term AD management, with their roles in inflammation control, barrier maintenance, and combined/sequential therapy requiring further clinical validation.

6 Conventional AD therapies as indirect microbiome modulators

An important consideration is that, although conventional therapies are not classified as MDTs, changes in the cutaneous microbiome are not exclusive to MDTs. Traditional anti-inflammatory and antimicrobial treatments can also indirectly remodel the microbial community by improving inflammation, barrier integrity, and the local skin microenvironment.

Existing studies have shown that dupilumab can significantly reduce the abundance of S. aureus in AD lesions, restore microbial diversity, and increase the number of commensal bacteria (66). Tralokinumab has demonstrated similar microbiome-improving effects to dupilumab (67). Anti-inflammatory treatments, including JAK inhibitors and biologics such as dupilumab, exert their effects by suppressing type 2 inflammation, thereby indirectly altering the microenvironment of lesional skin. Likewise, topical corticosteroids(TCS) can, through their anti-inflammatory and barrier-restorative actions, lead to significant remodelling and normalisation of the density, diversity, and community structure of the cutaneous microbiome in AD lesions (68). Therefore, the improvements in microbial diversity or reduction in S. aureus observed after conventional therapies may be interpreted as both a contributing factor to clinical improvement and a consequence of reduced inflammation and restored barrier function.

Overall, these findings complicate any simplistic causal explanation of dysbiosis in AD. Cutaneous microbial dysbiosis may be both a driver of inflammation and a consequence of barrier disruption and type 2 immune activation. Accordingly, future multidisciplinary therapeutic studies should account for background anti-inflammatory treatments and assess whether microbiota changes constitute a primary mechanism of therapy or are instead a secondary marker of disease control.

7 The evidence required for stratified MDTs to progress from proof of concept to clinical validation

The effectiveness of stratified MDTs should not be defined solely by the short-term efficacy of a single intervention compared to placebo. Instead, it should encompass several key aspects:

1) Head-to-head comparative studies. There is a need for direct comparisons of different MDT categories in similar patient populations, under unified endpoint frameworks. Such studies can clarify efficacy, safety, and reproducibility, thereby identifying the optimal therapeutic options for distinct biological subtypes. 2) Sequential therapy studies. It is necessary to verify whether stepwise or escalation treatment strategies are superior to fixed single regimens. Additionally, these studies should assess the impact on overall efficacy and risk exposure. 3) Accompanying biomarker research. Studies should integrate data on the microbiota, metabolites, and immune phenotypes. The aim is to establish multidimensional biomarker systems for patient selection, prediction of therapeutic response, monitoring, and mechanistic explanation. 4) Long-term follow-up studies. These should focus on evaluating the durability of efficacy, relapse rates, colonisation stability of exogenous strains or functional networks, and potential late-onset safety risks.

This core intervention logic is grounded in an overarching methodological framework (Figure 1), which ensures the clinical translatability of the model. In other words, the key question for stratified MDTs is not merely ‘Is a given microbiota therapy effective?’, but rather ‘Does an integrated treatment strategy—based on patient stratification and dynamic escalation—outperform traditional, non-stratified empirical therapies?’.

8 Summary

The three-tiered framework proposed in this article provides a clear conceptual structure for understanding the heterogeneity of microbiota-directed therapies in atopic dermatitis. The first tier focuses on local inflammation control within skin lesions. The second tier addresses the potential impact of the gut microbiota and its metabolites on systemic immune regulation. The third tier concerns longer-term restoration of the microbial ecosystem. The main contribution of this framework lies in its avoidance of grouping mechanistically distinct interventions—such as probiotics, topical microbial therapies, antimicrobial strategies, and faecal microbiota transplantation—into a single category. Instead, it distinguishes them according to their primary level of action. This distinction helps to explain inconsistencies in existing research findings and provides a theoretical basis for designing future combined or sequential therapeutic strategies.

Nevertheless, there remain clear limitations in the current evidence regarding microbiota-directed therapies for AD. Many mechanistic insights are derived primarily from animal models or in vitro experiments, which cannot be directly extrapolated to human disease. Existing clinical studies are also limited by small sample sizes, heterogeneity in intervention strains and dosages, differences in patient age and disease phenotype, and a lack of standardised outcome measures. Furthermore, changes in the cutaneous microbiome can also be observed during conventional treatments, complicating any simplistic causal interpretation of dysbiosis in AD. Cutaneous microbial dysbiosis may act both as a driver of inflammation and as a consequence of barrier disruption and type 2 inflammation. Therefore, therapies targeting the microbiome should be evaluated not only as independent interventions, but also in relation to traditional anti-inflammatory treatments that indirectly remodel the microbial ecosystem.

Future research should clarify which patients, at which disease stages, are most suitable for specific microbiota-directed interventions. Optimal study designs should combine clinical scoring, skin and gut microbiome analysis, metabolomics, immune phenotyping, and assessment of barrier function, with long-term follow-up. Such designs are needed to determine the durability of efficacy, the stability of colonisation by exogenous strains or functional networks, and the presence of delayed safety risks. There is also a need for systematic evaluation of the indirect effects of conventional treatments—such as dupilumab, JAK inhibitors, topical corticosteroids, and topical antimicrobials—on the skin microbiome. In summary, the value of the three-tiered framework lies not in providing immediate new clinical recommendations, but in offering a more systematic theoretical basis for interpreting existing evidence, guiding future trial design, and informing precision microbiota-directed therapeutic strategies.

References

• 1

  RaimondoALemboS. Atopic dermatitis: Epidemiology and clinical phenotypes. Dermatol Pract Concept. (2021) 11:e2021146. doi: 10.5826/dpc.1104a146

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 2

  TominagaMTakamoriK. Peripheral itch sensitization in atopic dermatitis. Allergology Int. (2022) 71:265–77. doi: 10.1016/j.alit.2022.04.003

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 3

  YangGSeokJKKangHCChoY-YLeeHSLeeJY. Skin barrier abnormalities and immune dysfunction in atopic dermatitis. IJMS. (2020) 21:2867. doi: 10.3390/ijms21082867

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 4

  ChiangC-CChengW-JDela CruzJRMSRavirajTWuN-LKorinekMet al. Neutrophils in atopic dermatitis. Clin Rev Allergy Immunol. (2024) 67:21–39. doi: 10.1007/s12016-024-09004-3

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 5

  CriadoPRMiotHABueno-FilhoRIanhezMCriadoRFJDe CastroCCS. Update on the pathogenesis of atopic dermatitis. Anais Brasileiros Dermatologia. (2024) 99:895–915. doi: 10.1016/j.abd.2024.06.001

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 6

  GatmaitanJGLeeJH. Challenges and future trends in atopic dermatitis. IJMS. (2023) 24:11380. doi: 10.3390/ijms241411380

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 7

  FyhrquistNYangYKarisolaPAleniusH. Endotypes of atopic dermatitis. J Allergy Clin Immunol. (2025) 156:24–40.e4. doi: 10.1016/j.jaci.2025.02.029

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 8

  KimMSKimHJKangSMHeoYMKangJRyuTKet al. Efficacy and safety of topical streptococcus postbiotic emollient in adolescents and adults with mild‐to‐moderate atopic dermatitis: A randomized, double‐blind, vehicle‐controlled trial. Allergy. (2024) 79:1612–6. doi: 10.1111/all.16077

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 9

  OngPY. Atopic dermatitis: Is innate or adaptive immunity in control? A clinical perspective. Front Immunol. (2022) 13:943640. doi: 10.3389/fimmu.2022.943640

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 10

  ŁacwikJKraikKLaskaJTotaMSędekŁGomułkaK. IL-31/33 axis in atopic dermatitis. IJMS. (2025) 26:10162. doi: 10.3390/ijms262010162

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 11

  Al KindiAWilliamsHMatsudaKAlkahtaniAMSavilleCBennettHet al. Staphylococcus aureus second immunoglobulin-binding protein drives atopic dermatitis via IL-33. J Allergy Clin Immunol. (2021) 147:1354–1368.e3. doi: 10.1016/j.jaci.2020.09.023

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 12

  QiuZZhuZLiuXChenBYinHGuCet al. A dysregulated sebum–microbial metabolite–IL-33 axis initiates skin inflammation in atopic dermatitis. J Exp Med. (2022) 219:e20212397. doi: 10.1084/jem.20212397

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 13

  NguyenA-TKimMKimY-EKimHLeeSLeeYet al. MSF enhances human antimicrobial peptide β-defensin (HBD2 and HBD3) expression and attenuates inflammation via the NF-κB and p38 signaling pathways. Molecules. (2023) 28:2744. doi: 10.3390/molecules28062744

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 14

  OngPYBoguniewiczMLeungDYM. Endogenous antimicrobial peptides and skin infections in atopic dermatitis. N Engl J Med. (2002) 347:1151–60. doi: 10.1056/NEJMoa021481

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 15

  ClausenM-LJungerstedJMAndersenPSSlotvedH-CKrogfeltKAAgnerT. Human β-defensin-2 as a marker for disease severity and skin barrier properties in atopic dermatitis. Br J Dermatol. (2013) 169:587–93. doi: 10.1111/bjd.12419

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 16

  FangZLiLZhaoJZhangHLeeY-KLuWet al. Bifidobacteria adolescentis regulated immune responses and gut microbial composition to alleviate DNFB-induced atopic dermatitis in mice. Eur J Nutr. (2020) 59:3069–81. doi: 10.1007/s00394-019-02145-8

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 17

  XiaoXHuXYaoJCaoWZouZWangLet al. The role of short-chain fatty acids in inflammatory skin diseases. Front Microbiol. (2023) 13:1083432. doi: 10.3389/fmicb.2022.1083432

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 18

  MohammadSKarimMIqbalSLeeJHMathiyalaganRKimYJet al. Atopic dermatitis: Pathophysiology, microbiota, and metabolome – A comprehensive review. Microbiol Res. (2024) 281:127595. doi: 10.1016/j.micres.2023.127595

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 19

  TrompetteAPernotJPerdijkOAlqahtaniRAADomingoJSCamacho-MuñozDet al. Gut-derived short-chain fatty acids modulate skin barrier integrity by promoting keratinocyte metabolism and differentiation. Mucosal Immunol. (2022) 15:908–26. doi: 10.1038/s41385-022-00524-9

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 20

  ZhuLJianXZhouBLiuRMuñozMSunWet al. Gut microbiota facilitate chronic spontaneous urticaria. Nat Commun. (2024) 15:112. doi: 10.1038/s41467-023-44373-x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 21

  KimM-JKimJ-YKangMWonM-HHongS-HHerY. Reduced fecal calprotectin and inflammation in a murine model of atopic dermatitis following probiotic treatment. Int J Mol Sci. (2020) 21:3968. doi: 10.3390/ijms21113968

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 22

  MaslowskiKMVieiraATNgAKranichJSierroFYuDet al. Regulation of inflammatory responses by gut microbiota and chemoattractant receptor GPR43. Nature. (2009) 461:1282–6. doi: 10.1038/nature08530

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 23

  YangWYuTHuangXBilottaAJXuLLuYet al. Intestinal microbiota-derived short-chain fatty acids regulation of immune cell IL-22 production and gut immunity. Nat Commun. (2020) 11:4457. doi: 10.1038/s41467-020-18262-6

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 24

  KangH-JKimG-CLeeC-GParkSSharmaGVermaRet al. Probiotics-derived metabolite ameliorates skin allergy by promoting differentiation of FOXP3+ regulatory T cells. J Allergy Clin Immunol. (2021) 147:1517–21. doi: 10.1016/j.jaci.2020.11.040

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 25

  StecASikoraMMaciejewskaMParalusz-StecKMichalskaMSikorskaEet al. Bacterial metabolites: A link between gut microbiota and dermatological diseases. IJMS. (2023) 24:3494. doi: 10.3390/ijms24043494

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 26

  YouYLuHWangYWangBChenQShiY. Probiotic encapsulation strategies for controlled intestinal delivery and microbiome dysbiosis therapy. J Controlled Release. (2025) 387:114252. doi: 10.1016/j.jconrel.2025.114252

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 27

  KimJ-HKimKKimW. Gut microbiota restoration through fecal microbiota transplantation: a new atopic dermatitis therapy. Exp Mol Med. (2021) 53:907–16. doi: 10.1038/s12276-021-00627-6

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 28

  PetryVBessaGRPoziomczyckCSOliveiraCFDWeberMBBonamigoRRet al. Bacterial skin colonization and infections in patients with atopic dermatitis. Bras Dermatol. (2012) 87:729–34. doi: 10.1590/S0365-05962012000500010

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 29

  FanXZangTDaiJWuNHopeCBaiJet al. The associations of maternal and children’s gut microbiota with the development of atopic dermatitis for children aged 2 years. Front Immunol. (2022) 13:1038876. doi: 10.3389/fimmu.2022.1038876

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 30

  QiCLiZTuHSunFGuoWDiCet al. 2’-FL and cross-feeding bifidobacteria reshaped the gut microbiota of infants with atopic dermatitis ex vivo and prevented dermatitis in mice post-microbiota transplantation through retinol metabolism activation. Gut Microbes. (2025) 17:2474148. doi: 10.1080/19490976.2025.2474148

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 31

  HoskinsonCDaiDLYDel BelKLBeckerABMoraesTJMandhanePJet al. Delayed gut microbiota maturation in the first year of life is a hallmark of pediatric allergic disease. Nat Commun. (2023) 14:4785. doi: 10.1038/s41467-023-40336-4

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 32

  ZhouJXuGLiXTuHLiHChangHet al. Limosilactobacillus reuteri FN041 prevents atopic dermatitis in pup mice by remodeling the ileal microbiota and regulating gene expression in peyer’s patches after vertical transmission. Front Nutr. (2022) 9:987400. doi: 10.3389/fnut.2022.987400

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 33

  WeckelADhariwalaMOLyKTranVMOjewumiOTRiggsJBet al. Long term tolerance to skin commensals is established neonatally through a specialized dendritic cell subgroup. Immunity. (2023) 56:1239–1254.e7. doi: 10.1016/j.immuni.2023.03.008

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 34

  CabanaMDMcKeanMCaugheyABFongLLynchSWongAet al. Early probiotic supplementation for eczema and asthma prevention: A randomized controlled trial. Pediatrics. (2017) 140:e20163000. doi: 10.1542/peds.2016-3000

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 35

  BarbarotSAubertHBoivinMFoureauAMaruaniADroitcourtCet al. Evaluation of antenatal prebiotic intake in infants at risk of atopy: the double-blind PREGRALL randomized clinical trial. Br J Dermatol. (2025) 194(3):441–9. doi: 10.1093/bjd/ljaf414

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 36

  AnaniaCMatysVMarraSDe CanditiisDOliveroFCarraroCet al. Effect of supplementation with a specific probiotic (Bifidobacterium bifidum PRL2010) in pregnancy for the prevention of atopic dermatitis in children: Preliminary results of a randomized trial. Nutrients. (2025) 17:673. doi: 10.3390/nu17040673

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 37

  DotterudCKStorrøOJohnsenRØienT. Probiotics in pregnant women to prevent allergic disease: a randomized, double-blind trial: Probiotics in pregnant women to prevent allergic disease. Br J Dermatol. (2010) 163:616–23. doi: 10.1111/j.1365-2133.2010.09889.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 38

  RatherIAKimB-CLewL-CChaS-KLeeJHNamG-Jet al. Oral administration of live and dead cells of Lactobacillus sakei proBio65 alleviated atopic dermatitis in children and adolescents: a randomized, double-blind, and placebo-controlled study. Probiotics Antimicro Prot. (2021) 13:315–26. doi: 10.1007/s12602-020-09654-7

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 39

  FangZLuWZhaoJZhangHQianLWangQet al. Probiotics modulate the gut microbiota composition and immune responses in patients with atopic dermatitis: a pilot study. Eur J Nutr. (2020) 59:2119–30. doi: 10.1007/s00394-019-02061-x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 40

  AvershinaERubioRCLundgårdKMartinezGPColladoMCStorrøOet al. Effect of probiotics in prevention of atopic dermatitis is dependent on the intrinsic microbiota at early infancy. J Allergy Clin Immunol. (2017) 139:1399–1402.e8. doi: 10.1016/j.jaci.2016.09.056

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 41

  SilverbergJIEichenfieldLFHebertAASimpsonELStein GoldLBissonnetteRet al. Tapinarof cream 1% once daily: Significant efficacy in the treatment of moderate to severe atopic dermatitis in adults and children down to 2 years of age in the pivotal phase 3 ADORING trials. J Am Acad Dermatol. (2024) 91:457–65. doi: 10.1016/j.jaad.2024.05.023

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 42

  van den BogaardEHBergboerJGMVonk-BergersMvan Vlijmen-WillemsIMJJHatoSVvan der ValkPGMet al. Coal tar induces AHR-dependent skin barrier repair in atopic dermatitis. J Clin Invest. (2013) 123:917–27. doi: 10.1172/JCI65642

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 43

  ModouxMRolhionNLefevreJHOeuvrayCNádvorníkPIllesPet al. Butyrate acts through HDAC inhibition to enhance aryl hydrocarbon receptor activation by gut microbiota-derived ligands. Gut Microbes. (2022) 14:2105637. doi: 10.1080/19490976.2022.2105637

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 44

  LiBLiangXLiYWangRWeiYLiuQet al. Tryptophan catabolites from microbiota ameliorate immune-mediated hepatitis through activating aryl hydrocarbon receptor of T cells. Gut Microbes. (2025) 17:2557979. doi: 10.1080/19490976.2025.2557979

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 45

  WuYTaoQXieJLiuXZhouYWeiCet al. Indole-3-carbinol inhibits PD-L1-mediated immune evasion in hepatocellular carcinoma via suppressing NF-κB p105 ubiquitination. Phytomedicine. (2025) 141:156692. doi: 10.1016/j.phymed.2025.156692

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 46

  SunMMaNHeTJohnstonLJMaX. Tryptophan (Trp) modulates gut homeostasis via aryl hydrocarbon receptor (AhR). Crit Rev Food Sci Nutr. (2020) 60:1760–8. doi: 10.1080/10408398.2019.1598334

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 47

  KangY-MKimH-MLeeJBaekJ-SLeeMAnH-J. Indole-3-carbinol alleviates allergic skin inflammation via periostin/thymic stromal lymphopoietin suppression in atopic dermatitis. Chin Med. (2024) 19:177. doi: 10.1186/s13020-024-01042-5

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 48

  TianCDengSYangMBaiBPanYXieGet al. Indole-3-carbinol and its main derivative 3,3′-diindolylmethane: Regulatory roles and therapeutic potential in liver diseases. BioMed Pharmacother. (2024) 180:117525. doi: 10.1016/j.biopha.2024.117525

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 49

  AmarakoonDLeeW-JTamiaGLeeS-H. Indole-3-carbinol: Occurrence, health-beneficial properties, and cellular/molecular mechanisms. Annu Rev Food Sci Technol. (2023) 14:347–66. doi: 10.1146/annurev-food-060721-025531

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 50

  FangZPanTLiLWangHZhuJZhangHet al. Bifidobacterium longum mediated tryptophan metabolism to improve atopic dermatitis via the gut-skin axis. Gut Microbes. (2022) 14:2044723. doi: 10.1080/19490976.2022.2044723

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 51

  Asadi ShahmirzadiAEdgarDLiaoC-YHsuY-MLucanicMAsadi ShahmirzadiAet al. Alpha-ketoglutarate, an endogenous metabolite, extends lifespan and compresses morbidity in aging mice. Cell Metab. (2020) 32:447–456.e6. doi: 10.1016/j.cmet.2020.08.004

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 52

  WuNYangMGaurUXuHYaoYLiD. Alpha-ketoglutarate: physiological functions and applications. Biomol Ther (Seoul). (2016) 24:1–8. doi: 10.4062/biomolther.2015.078

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 53

  TahilianiMKohKPShenYPastorWABandukwalaHBrudnoYet al. Conversion of 5-methylcytosine to 5-hydroxymethylcytosine in mammalian DNA by MLL partner TET1. Science. (2009) 324:930–5. doi: 10.1126/science.1170116

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 54

  ZhaoJKuiLHuangJDengJLiuLZhuCet al. Bifidobacterium animalis subsp. lactis BX-BC08 modulates gut microbiota and secretes alpha-ketoglutaric acid to alleviate MC903-induced atopic dermatitis. J Transl Med. (2025) 23:768. doi: 10.1186/s12967-025-06769-9

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 55

  MikamiNKawakamiRChenKYSugimotoAOhkuraNSakaguchiS. Epigenetic conversion of conventional T cells into regulatory T cells by CD28 signal deprivation. Proc Natl Acad Sci USA. (2020) 117:12258–68. doi: 10.1073/pnas.1922600117

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 56

  Arroyo-OlarteRDFlores-CastelánJCArmas-LópezLEscobedoGTerrazasLIÁvila-MorenoFet al. Targeted demethylation of FOXP3-TSDR enhances the suppressive capacity of STAT6-deficient inducible T regulatory cells. Inflammation. (2024) 47:2159–72. doi: 10.1007/s10753-024-02031-4

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 57

  KenneyHMYoshidaTBerdyshevECalatroniAGillSRSimpsonELet al. CERS1 is a biomarker of Staphylococcus aureus abundance and atopic dermatitis severity. J Allergy Clin Immunol. (2025) 155:479–90. doi: 10.1016/j.jaci.2024.09.017

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 58

  Bath-HextallFJBirnieAJRavenscroftJCWilliamsHC. Interventions to reduce Staphylococcus aureus in the management of atopic eczema: an updated Cochrane review: S. aureus and atopic eczema. Br J Dermatol. (2010) 163:12–26. doi: 10.1111/j.1365-2133.2010.09743.x

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 59

  ItoYSasakiTLiYTanoueTSugiuraYSkellyANet al. Staphylococcus cohnii is a potentially biotherapeutic skin commensal alleviating skin inflammation. Cell Rep. (2021) 35:109052. doi: 10.1016/j.celrep.2021.109052

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 60

  Navarro-LópezVRamírez-BoscáARamón-VidalDRuzafa-CostasBGenovés-MartínezSChenoll-CuadrosEet al. Effect of oral administration of a mixture of probiotic strains on SCORAD index and use of topical steroids in young patients with moderate atopic dermatitis. JAMA Dermatol. (2018) 154:37–43. doi: 10.1001/jamadermatol.2017.3647

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 61

  D’AuriaEPanelliSLunardonLPajoroMParadisoLBerettaSet al. Rice flour fermented with Lactobacillus paracasei CBA L74 in the treatment of atopic dermatitis in infants: a randomized, double-blind, placebo-controlled trial. Pharmacol Res. (2021) 163:105284. doi: 10.1016/j.phrs.2020.105284

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 62

  MylesIAEarlandNJAndersonEDMooreINKiehMDWilliamsKWet al. First-in-human topical microbiome transplantation with roseomonas mucosa for atopic dermatitis. JCI Insight. (2018) 3:e120608. doi: 10.1172/jci.insight.120608

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 63

  NakatsujiTHataTRTongYChengJYShafiqFButcherAMet al. Development of a human skin commensal microbe for bacteriotherapy of atopic dermatitis and use in a phase 1 randomized clinical trial. Nat Med. (2021) 27:700–9. doi: 10.1038/s41591-021-01256-2

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 64

  LiuXQinYDongLHanZLiuTTangYet al. Living symbiotic bacteria-involved skin dressing to combat indigenous pathogens for microbiome-based biotherapy toward atopic dermatitis. Bioact Mater. (2023) 21:253–66. doi: 10.1016/j.bioactmat.2022.08.019

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 65

  OhCKChungHHKimYJKimJB. Comparison of rifaximin monotherapy and rifaximin combined with probiotics in patients with irritable bowel syndrome: a randomized controlled trial. Nutrients. (2025) 17:763. doi: 10.3390/nu17050763

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 66

  SimpsonELSchlievertPMYoshidaTLussierSBoguniewiczMHataTet al. Rapid reduction in Staphylococcus aureus in atopic dermatitis subjects following dupilumab treatment. J Allergy Clin Immunol. (2023) 152:1179–95. doi: 10.1016/j.jaci.2023.05.026

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 67

  BeckLABieberTWeidingerSTauberMSaekiHIrvineADet al. Tralokinumab treatment improves the skin microbiota by increasing the microbial diversity in adults with moderate-to-severe atopic dermatitis: analysis of microbial diversity in ECZTRA 1, a randomized controlled trial. J Am Acad Dermatol. (2023) 88:816–23. doi: 10.1016/j.jaad.2022.11.047

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 68

  GonzalezMESchafferJVOrlowSJGaoZLiHAlekseyenkoAVet al. Cutaneous microbiome effects of fluticasone propionate cream and adjunctive bleach baths in childhood atopic dermatitis. J Am Acad Dermatol. (2016) 75:481–493.e8. doi: 10.1016/j.jaad.2016.04.066

  • Pubmed Abstract
  • CrossRef
  • Google Scholar