Gut Microbiota Signaling Axes as Candidate Therapeutic Targets in Type 2 Diabetes Mellitus: Mechanistic Crosstalk and Translational Challenges.

Introduction

Diabetes mellitus represents a heterogeneous metabolic disorder characterized by persistent hyperglycemia, with two predominant subtypes including type 1 and type 2 diabetes mellitus. T2DM arises from the complex interaction between genetic susceptibility and multiple modifiable risk factors encompassing obesity, sedentary behavior, aging, unhealthy dietary patterns and environmental exposure, instead of a single independent pathogenic factor. Epidemiological statistics released by the International Diabetes Federation (IDF) demonstrate that T2DM constitutes the majority of adult diabetes cases worldwide.

The Global Burden of Disease Study 2021 predicts that the global prevalence of diabetes across all age groups will reach approximately 1.31 billion cases by 2050.1 The 11th edition of the IDF Diabetes Atlas further reports that global health expenditure attributable to diabetes surpassed USD 1 trillion in 2024.2 Diabetes and its chronic complications place a heavy socioeconomic and familial medical burden, accompanied by a rising incidence among young populations.3 Although multiple classes of hypoglycemic agents have been clinically applied, these drugs differ substantially in functional mechanisms, clinical indications and organ-protective effects. Detailed discussion regarding diabetes pharmacotherapy is not included in the scope of the present review, which focuses specifically on gut microbiota-derived signaling mechanisms underlying T2DM pathogenesis.

The gut microbiome is widely recognized as a functionally sophisticated virtual organ that modulates host metabolic balance and immune homeostasis.4 Accumulating evidence from preclinical animal experiments, in vitro mechanistic analyses and human observational cohort studies indicates that gut microbiota disturbance, defined as abnormal alterations in the composition and metabolic activity of intestinal microflora compared with healthy conditions, is closely associated with insulin resistance and metaflammation (metabolic inflammation, a chronic low-grade non-infectious inflammatory state driven by nutrient overload, which differs from systemic inflammation, intestinal inflammation and adipose tissue inflammation) in patients with T2DM. The association between gut microbiota disturbance and T2DM is bidirectional and multifactorial.5 Gut microbial community structure can be dynamically shaped by multiple host and environmental factors including dietary habits, obesity status, glycemic level, medication use, genetic background, ethnicity, age and comorbid conditions. Most existing human microbiome studies are observational in design, which fails to confirm definitive causal relationships.

Gut microbiota disturbance is therefore considered a crucial contributing factor rather than a well-established primary cause of T2DM.

Microbial metabolites and pattern recognition signaling pathways act as critical mediators bridging gut microbiota disturbance and host metabolic and immune dysfunction. Among the numerous identified microbial signaling cascades, four interactive and well-characterized signaling axes have attracted extensive research attention. The SCFAs–AMPK axis serves as a core energy-sensing regulatory hub.

The bile acid (BA)–FXR/TGR5 axis and tryptophan (Trp) –AhR axis synergistically maintain intestinal barrier integrity and immune homeostasis. The LPS–TLR4/NF-κB axis functions as a pivotal proinflammatory pathway that converges on the downstream NF-κB/NLRP3 inflammatory complex. The four axes are selected for discussion in this review for three key reasons.

These axes correspond to four major categories of microbial metabolites including short-chain fatty acids (SCFAs), bile acids, lipopolysaccharide endotoxins and Trp derivatives. They cover core functional modules involving energy sensing, metabolic regulation, intestinal barrier defense and inflammatory response. In addition, abundant evidence has verified the extensive molecular crosstalk among these signaling pathways.

Current review articles predominantly elaborate the individual biological functions of these microbial signaling pathways in T2DM, while systematic summaries regarding their interactive regulatory networks remain scarce. In particular, holistic investigations into the crosstalk between energy-sensing modules and immune barrier modules are insufficient. The spatiotemporal interaction patterns and relative regulatory magnitude of the four signaling axes also lack systematic characterization.

This narrative review integrates the latest research progress on four core interconnected gut microbial signaling axes, namely SCFAs–AMPK, BA–FXR/TGR5, LPS–TLR4/NF-κB and Trp-AhR. We propose a conceptual pathogenic framework in which concurrent dysregulation of multiple microbiota-mediated signaling axes drives T2DM progression. This review also summarizes the potential therapeutic targets of each pathway and their clinical translation prospects.

Notably, the proposed four-axis regulatory framework is predominantly supported by preclinical evidence. Comprehensive multi-omics data from human cohorts and clinical validation of multi-axis targeted intervention remain limited. Individual variations in gut microbial characteristics across different populations may also lead to heterogeneous therapeutic responses, which warrants further in-depth exploration.6

Methodology

This is a narrative review without a pre-registered systematic protocol, and we followed the SANRA reporting guideline to improve transparency. Literature was retrieved from PubMed, Web of Science, Scopus, and Google Scholar up to July 2026. To avoid omitting studies focusing on only one or two axes, separate search strings were built for each signaling axis plus one dedicated query for pathway crosstalk: (1) SCFAs-AMPK: (“gut microbiota” OR “intestinal microbiota”) AND (“type 2 diabetes mellitus” OR T2DM) AND (“short-chain fatty acids” OR SCFAs OR AMPK); (2) Bile acid-FXR/TGR5: (“gut microbiota” OR “intestinal microbiota”) AND (“type 2 diabetes mellitus” OR T2DM) AND (“bile acids” OR FXR OR TGR5); (3) LPS-TLR4/NF-κB: (“gut microbiota” OR “intestinal microbiota”) AND (“type 2 diabetes mellitus” OR T2DM) AND (“LPS” OR TLR4 OR NF-κB OR NLRP3); (4) Tryptophan-AhR: (“gut microbiota” OR “intestinal microbiota”) AND (“type 2 diabetes mellitus” OR T2DM) AND (“tryptophan” OR AhR); (5) Crosstalk: (“gut microbiota” OR “intestinal microbiota”) AND (“type 2 diabetes mellitus” OR T2DM) AND (“crosstalk” OR “network” OR “signaling interaction”).

Only English-language articles were included, which may introduce language bias, especially for natural-product-related preclinical work from non-English regions. Potential publication bias (under-reporting of negative/neutral results) should also be considered. Eligibility criteria: Included: human observational/clinical studies, animal in-vivo, in-vitro cellular studies addressing gut-microbiota-related signaling in T2DM; relevant studies of obesity, metabolic syndrome, NAFLD and diabetic complications were included only if they contained mechanistic links to T2DM-associated microbiota axes.

Excluded: pure type 1-diabetes work without T2DM-relevant comparison, papers lacking microbiota-related mechanistic readouts, case reports and conference abstracts without full-text data. Screening was performed via title-abstract followed by full-text evaluation; duplicates were removed by reference-management software. Disagreements on inclusion were resolved by group discussion.

Besides database outputs, we manually screened reference lists of included papers for additional relevant records; thus not all references derive directly from database searches. These four axes were selected as core hubs because each represents a distinct functional module (energy-sensing, bile-acid regulation, pro-inflammatory signaling, intestinal-barrier repair), receives microbiota-metabolite inputs and connects downstream to insulin resistance. Throughout the manuscript, we stratify evidence into human clinical/observational, animal, in-vitro, and hypothesis-driven inference, and discuss conflicting findings considering model, dosage and tissue-specific effects.

Since no prospective logging of retrieved/screened record numbers was performed at the initial search stage, exact quantitative flow-chart statistics are unavailable; this limitation is explicitly acknowledged.

The Interplay Between Four Interconnected Signaling Axes and Type 2 Diabetes Mellitus

Energy-Sensing Hubs: Synergistic Effects of the SCFAs–AMPK and BA–FXR/TGR5 Axes

The SCFAs–AMPK Signaling Axis

SCFAs represent volatile fatty acids primarily generated from the anaerobic bacterial fermentation of undigested dietary fibers, with acetate, propionate and butyrate identified as the predominant components. Multiple clinical investigations have demonstrated reduced gut microbial alpha diversity and markedly decreased abundance of butyrate-producing bacteria in patients with T2DM.7 Metagenome-wide association studies further validate moderate gut microbiota disturbance in T2DM individuals, which is characterized by the depletion of butyrate-producing microbial taxa and the proliferation of opportunistic pathogens.8 Notably, the reduced abundance of butyrate-producing bacteria does not universally lead to decreased total SCFA production in human cohorts, owing to potential functional compensation by other fermentative microbial species. SCFAs exert extensive and tissue-specific biological effects dependent on concentration, receptor activation and dietary status via sophisticated interactive regulatory networks, rather than relying on a single linear AMPK-centered signaling cascade.9

Under physiological conditions, typical butyrate-producing bacteria including Roseburia, Faecalibacterium and Eubacterium produce butyrate through dietary fiber fermentation. Butyrate modulates AMPK activity mainly through indirect regulatory mechanisms instead of direct protein activation. This metabolite binds to free fatty acid receptors FFAR2 (GPR43) and FFAR3 (GPR41) expressed on hepatocytes and intestinal epithelial cells, thereby triggering the activation of LKB1 and CaMKK2 kinases.

Butyrate also inhibits histone deacetylase (HDAC) activity to further elevate LKB1 expression. These upstream molecular cascades facilitate the phosphorylation of the AMPKα subunit at the Thr172 site and sustain stable hepatic p-AMPK levels, with additional contributions from altered cellular energy status and mitochondrial metabolic reprogramming.10,11

Although these mechanistic findings are predominantly derived from preclinical studies, human investigations have confirmed corresponding pathway alterations in prediabetic and T2DM conditions. Large-scale metagenomic cohort studies have verified the depletion of butyrate-producing microbes in T2DM patients, while fecal total SCFA concentrations vary across different populations due to microbial functional compensation.12 Observational clinical data indicate that circulating SCFA levels are positively correlated with insulin sensitivity, whereas fecal SCFA content fails to precisely reflect tissue-level AMPK activation status.13 Dietary fiber supplementation randomized controlled trials yield moderate glycemic improvement, and the therapeutic efficacy is largely determined by baseline gut microbial profiles.14

Luminal SCFAs primarily execute local regulatory functions in the intestinal epithelium, while circulating SCFAs mediate systematic biological effects in peripheral metabolic organs. The SCFA-induced secretion of intestinal L-cell-derived glucagon-like peptide-1 (GLP-1) is predominantly regulated by FFAR2 and FFAR3, which cooperate with multiple nutrient-sensing pathways. GPR119 does not serve as the core mediator of SCFA-stimulated GLP-1 release.

The enhanced GLP-1 signaling improves pancreatic β-cell function, increases glucose-stimulated insulin secretion and suppresses excessive appetite, thereby facilitating systemic glycemic stabilization.15 Under pathological conditions, gut microbiota disturbance contributes to the depletion of butyrate-producing microbes and insufficient SCFA synthesis. The subsequent impairment of FFAR2 and FFAR3 signaling further blunts AMPK activation, which promotes excessive hepatic gluconeogenesis and disrupts whole-body glucose homeostasis.16

Hepatic AMPK acts as a core regulator of systemic energy homeostasis, which maintains metabolic balance via the suppression of gluconeogenesis and promotion of fatty acid oxidation.17 AMPK activity is significantly inhibited under obese and T2DM pathological states, which impairs the endogenous metabolic protective mechanism.18 Activated AMPK upregulates PPARα expression and accelerates ACC phosphorylation, thereby reducing intracellular free fatty acid and triglyceride accumulation.19 By modulating the activity of key transcription factors including CRTC2, FOXO1 and PGC-1α, AMPK downregulates the expression of critical gluconeogenic enzymes PEPCK and G6Pase, and ultimately inhibits excessive hepatic glucose production.20–22

The BA–FXR/TGR5 Signaling Axis

Disordered BA metabolic homeostasis is closely implicated in the pathogenesis of multiple metabolic diseases, including obesity and T2DM.23 Farnesoid X receptor (FXR), encoded by theNR1H4 gene, is an essential nuclear receptor responsible for BA signal transduction. Together with the membrane G protein-coupled BA receptor 5 (TGR5, also termed GPR131), FXR forms the core signaling cascade that mediates BA-dependent metabolic regulation.24 In contrast, pregnane X receptor (PXR) and vitamin D receptor (VDR) can only be weakly activated by high-dose BA exposure and exert no regulatory effects on basal physiological BA homeostasis.24 FXR and TGR5 mediate distinct biological functions. FXR primarily mediates negative feedback inhibition of BA synthesis and suppresses hepatic gluconeogenesis, while TGR5 predominantly promotes intestinal hormone secretion, enhances systemic energy expenditure, exerts anti-inflammatory properties and improves peripheral insulin sensitivity.23,24

FXR is highly expressed in hepatic and intestinal tissues. The biological outcomes of FXR activation are context-dependent and vary across tissue types, ligand categories, intervention dosages and durations, disease models and gut microbial configurations. FXR signaling therefore does not confer consistent beneficial effects under all physiological and pathological conditions.

In the liver, hepatic BA-mediated FXR activation directly downregulates the key gluconeogenic enzymes PEPCK and G6Pase via the FXR–SHP signaling cascade, leading to reduced hepatic glucose output. Meanwhile, FXR activation represses the transcription of sterol regulatory element-binding protein 1c (SREBP-1c), which further decreases hepatic fatty acid and triglyceride synthesis.25,26 Intestinal FXR improves intestinal L-cell function, promotes the synthesis and secretion of GLP-1, enhances insulin secretion and counteracts the hyperglycemic effects of glucagon, making it a promising therapeutic target for T2DM treatment. Intestinally activated FXR induces the secretion of fibroblast growth factor 15 (FGF15) in mice and its human ortholog FGF19 into the portal venous circulation.

These endocrine factors bind to the hepatic FGFR4/β-Klotho complex and inhibit the transcription of BA synthesis rate-limiting enzymes CYP7A1 and CYP8B1 through the JNK pathway, forming a negative feedback loop to prevent excessive BA synthesis. A previous mouse study has demonstrated that endogenous intestinal FGF15 regulates BA and cholesterol metabolism without affecting systemic glucose or energy homeostasis under physiological conditions.27 Nevertheless, pharmacological or exogenous FGF15/19 supplementation moderately promotes hepatic glycogen deposition and suppresses hepatic gluconeogenesis in an insulin-independent manner, indicating distinct regulatory characteristics between endogenous and exogenous FGF15/19 signaling in a dose- and context-dependent manner.

Multiple layered FXR-dependent negative feedback circuits tightly modulate BA synthesis to sustain BA pool homeostasis. First, hepatic FXR activation upregulates SHP expression. SHP binds to liver receptor homologue 1 (LRH-1) and directly inhibits the transcription of CYP7A1 andCYP8B1.28 Second, intestinal-derived FGF19 acts on hepatic FGFR4 through endocrine circulation to repress the expression of BA synthetic genes.29,30 Emerging evidence has further identified MAFG as a downstream transcription factor of FXR.

MAFG directly binds to the promoter regions of CYP7A1 and CYP8B1 to amplify the negative feedback regulation of BA synthesis.31

As a transmembrane BA sensor, TGR5 is widely distributed in intestinal endocrine cells, pancreatic β-cells, brown adipose tissue, skeletal muscle and immune cells. TGR5 modulates systemic glucose metabolism primarily through intracellular cyclic adenosine monophosphate (cAMP) signaling.32 In intestinal L-cells, BA activates Gαs-coupled TGR5 and elevates intracellular cAMP concentrations, which subsequently triggers PKA and Epac signaling pathways to facilitate GLP-1 synthesis and secretion.33 In pancreatic β-cells, the TGR5-mediated cAMP/PKA signaling cascade effectively enhances glucose-dependent insulin secretion, improves the pancreatic islet’s glucose response, and maintains endocrine homeostasis in the islets.34 In brown adipose tissue, TGR5-mediated cAMP signaling upregulates type 2 iodothyronine deiodinase (D2), which catalyzes the conversion of thyroxine T4 to biologically active T3. This process further increases uncoupling protein 1 (UCP1) expression, promotes thermogenesis, reduces circulating lipid levels and improves systemic insulin sensitivity.35,36 Additionally, conjugated BA-activated TGR5 suppresses NLRP3 inflammasome activation and reduces the release of pro-inflammatory cytokines including IL-1β.37 At the same time, TGR5 activates the PI3K/Akt signaling pathway, promoting glucose uptake and utilization in skeletal muscle and adipose tissue, improving insulin resistance, and thereby alleviating the metaflammation associated with metabolic disorders.38

Several unresolved controversies and contradictory findings exist regarding the FXR/TGR5 regulatory network in metabolic disorders. Specific activation of hepatic FXR stably inhibits gluconeogenesis and improves glucose homeostasis, whereas sustained excessive activation of intestinal FXR exerts opposing effects by suppressing GLP-1 secretion and abrogating hypoglycemic benefits. No unified consensus has been established to explain the tissue-specific divergent functions of FXR across different preclinical animal models.

For TGR5 signaling, short-term TGR5 agonism significantly increases energy expenditure and mitigates metaflammation. In contrast, prolonged TGR5 stimulation induces receptor desensitization and impairs downstream cAMP-dependent signaling responsible for glycemic regulation and thermogenesis. Furthermore, the positive correlation between the BA–TGR5 axis and insulin sensitivity observed in animal studies cannot be consistently replicated in human clinical cohorts.

Such inconsistent research outcomes can be largely attributed to inter-individual heterogeneity in human populations. Variable BA composition, diverse gut microbial community profiles, and distinct host genetic backgrounds among individuals jointly alter the activation intensity and downstream transduction efficiency of the BA–FXR/TGR5 axis. In addition, differences in age, body mass index, dietary habits, disease severity and medication status further reshape BA metabolic patterns and receptor responsiveness, leading to variable metabolic regulatory effects of this signaling axis across different individuals.

These inter-individual variations fundamentally contribute to the inconsistent therapeutic efficacy of FXR or TGR5 monotherapy in clinical populations with metabolic disorders. Large-scale, stratified clinical studies are therefore required to clarify the population-specific regulatory characteristics of the BA–FXR/TGR5 axis and validate its clinical translational value.

Crosstalk Within Energy-Sensing Hubs

Figure 1 illustrates that the SCFAs–AMPK axis and BA–FXR/TGR5 axis collectively form an integrated energy-sensing network that modulates host metabolic homeostasis. These two pathways exhibit context-dependent regulatory interactions, with synergistic compensatory effects predominating under physiological conditions and negative feedback crosstalk emerging under pathological stress. SCFAs and secondary BAs initiate signaling via two distinct families of G protein-coupled receptors, namely FFAR2/FFAR3 and TGR5.

Their downstream signaling cascades converge on two core functional nodes, including AMPK phosphorylation and activation, as well as intestinal L-cell GLP-1 secretion. Under pathological conditions where one pathway is functionally compromised, the other axis may provide partial metabolic compensation to maintain systemic energy balance. Figure 1SCFAs-AMPK and bile acid-FXR/TGR5 crosstalk in T2DM.

Dietary fiber is fermented by the gut microbiota to produce short-chain fatty acids (SCFAs), which activate FFAR2/FFAR3 on L cells, triggering Ca2+ influx and the CaMKK2-LKB1-AMPK axis. Primary bile acids are converted into secondary bile acids by microbial bile salt hydrolase (BSH). Secondary bile acids exert dual effects: they rapidly activate the TGR5-cAMP-PKA pathway (fast pathway), while also mediating delayed transcriptional regulation via the nuclear receptor FXR (slow pathway).

PKA is proposed to enhance LKB1 activity. Notably, activated FXR is hypothesized to inhibit FFAR2 transcription, reducing SCFA sensitivity. Phosphorylated AMPK (p-AMPK) promotes GLP-1 secretion.

Circulating GLP-1 acts on pancreatic β cells to promote insulin release, and on peripheral tissues to upregulate GLUT4 and PGC-1α while inhibiting gluconeogenic genes (CRTC2, PEPCK, G6Pase). FGF15/19 induced by FXR provides additional cross-regulation. In T2DM, disruption of this balance leads to impaired GLP-1-dependent metabolic regulation.

Solid lines denote experimentally confirmed mechanisms, whereas dashed lines represent proposed or hypothesized crosstalk mechanisms in the context of T2DM (eg, PKA-mediated LKB1 enhancement, FXR-mediated FFAR2 transcriptional repression, and FGF15/19-mediated cross-regulation).The diagram illustrates SCFAs-AMPK and bile acid-FXR/TGR5 crosstalk in type 2 diabetes mellitus. Dietary fiber is fermented by gut microbiota to produce short-chain fatty acids, activating FFAR2 and FFAR3 on L-cells, leading to calcium influx and the CaMKK2-LKB1-AMPK axis. Primary bile acids are converted to secondary bile acids by bile salt hydrolase, activating TGR5-cAMP-PKA (fast pathway) and FXR (slow pathway).

FXR inhibits FFAR2 transcription, reducing SCFA sensitivity. Phosphorylated AMPK promotes GLP-1 secretion, affecting pancreatic beta cells to release insulin and peripheral tissues to upregulate GLUT4 and PGC-1 alpha while inhibiting gluconeogenic genes. FGF15/19 induced by FXR provides additional regulation.

Solid lines indicate confirmed mechanisms, dashed lines indicate hypothesized crosstalk. The pathways form an energy-sensing network modulating metabolic homeostasis, with compensatory effects under physiological conditions and negative feedback under stress.SCFAs-AMPK & bile acid-FXR/TGR5 in T2DM: L-cells, tissues, insulin pathways.Abbreviations: AMPK, AMP-activated protein kinase; BSH, bile salt hydrolase; cAMP, cyclic adenosine monophosphate; CaMKK2, calcium/calmodulin-dependent protein kinase 2; CRTC2, CREB-regulated transcription coactivator 2; FFAR2/3, free fatty acid receptor 2/3; FGF15/19, fibroblast growth factor 15/19; FXR, farnesoid X receptor; G6Pase, glucose-6-phosphatase; GLP-1, glucagon-like peptide-1; GLUT4, glucose transporter type 4; LKB1, liver kinase B1; PEPCK, phosphoenolpyruvate carboxykinase; PGC-1α, peroxisome proliferator-activated receptor gamma coactivator 1-alpha; PKA, protein kinase A; SCFAs, short-chain fatty acids; T2DM, type 2 diabetes mellitus; TGR5, G-protein-coupled bile acid receptor 1.

Pathological inter-pathway negative feedback represents a key regulatory mechanism underlying axis dysfunction. Excessive BA-mediated FXR activation enables FXR to bind the promoter region of the FFAR2 gene and suppress its transcriptional expression. This transcriptional inhibition reduces FFAR2 abundance and blunts the responsiveness of intestinal endocrine cells to SCFA-derived signals, thereby triggering trans-pathway suppression of energy-sensing signaling.39 In the context of T2DM-associated gut microbiota disturbance, depleted butyrate-producing microbes lead to insufficient basal SCFA signaling.

Combined with chronic FXR overactivation-induced FFAR2 downregulation, intestinal L-cells undergo dual inhibitory pressure from impaired SCFA and BA regulatory pathways. Under such dual dysfunctional conditions, single intervention strategies including SCFA precursor supplementation or intestinal-selective FXR monotherapy may yield suboptimal hypoglycemic effects. Combined multi-target approaches that restore the activity of both energy-sensing axes may provide a more promising therapeutic strategy for T2DM management.

Accumulating preclinical evidence indicates that SCFAs serve as upstream modulators of the BA–FXR/TGR5 signaling axis. SCFA enrichment can sustain the proliferation of intestinal bacterial taxa that express bile salt hydrolase (BSH). BSH catalyzes the deconjugation of primary conjugated BAs, which provides essential substrates for intestinal 7α-dehydroxylating bacteria and facilitates the bioconversion of primary BAs to secondary BAs.40–42 Dysregulation of this microbial metabolic cascade may concurrently impair SCFA–AMPK signaling and restrict secondary BA generation, which potentially suppresses TGR5 pathway activation and aggravates the overall dysfunction of intestinal energy-sensing networks in T2DM.

The SCFAs–AMPK and BA–FXR/TGR5 axes exhibit notable spatiotemporal decoupling in T2DM pathological states, characterized by temporal kinetic asynchrony and tissue-specific spatial heterogeneity. In terms of temporal regulation, SCFA-triggered FFAR2 signaling and BA-induced TGR5 signaling belong to rapid G protein-coupled receptor cascades, with TGR5-mediated GLP-1 secretion peaking within 20 minutes after activation.43 In contrast, FXR acts as a nuclear receptor that mediates slow, transcription-dependent regulation of target genes. In vitro studies in GLUTag enteroendocrine cells have verified that the selective FXR agonist GW4064 significantly downregulates FFAR2 mRNA expression at 24 hours, while sustained intervention over days to weeks is required to achieve stable FFAR2 repression in vivo.39 Chronic BA accumulation and persistent FXR overactivation in T2DM sustain this long-term transcriptional inhibition, rendering intestinal L-cells hyposensitive to SCFA stimulation even with exogenous high-dose SCFA supplementation.

Furthermore, preclinical observations demonstrate that intestinal SCFAs and secondary BAs display asynchronous diurnal oscillatory secretion patterns, which may exacerbate temporal mismatch in substrate availability and signal activation of the two energy-sensing axes.44 Notably, current evidence supporting spatiotemporal decoupling is predominantly derived from in vitro experiments and rodent models. Human gut microbial and BA rhythmic patterns exhibit substantial inter-individual variability, and asynchronous signaling dynamics between SCFA and BA pathways have not been consistently validated in human cohorts. Accordingly, the spatiotemporal decoupling hypothesis remains preclinical and requires further large-scale clinical verification.

Spatially, the crosstalk between the two axes presents distinct tissue-specific patterns. In colonic L-cells, activated FXR exerts unidirectional transcriptional repression on FFAR2, forming negative inter-pathway regulation. Conversely, brown adipose tissue predominantly expresses TGR5 with minimal FXR expression.

In this tissue, SCFAs and secondary BAs synergistically upregulate D2 and UCP1 expression to promote thermogenesis, constituting a positive metabolic regulatory circuit. Such tissue-specific spatiotemporal signal decoupling may partially explain the limited therapeutic efficacy of single-target interventions focusing solely on either the SCFA or BA receptor pathway in T2DM treatment.

Several unresolved questions and context-dependent regulatory characteristics remain in this dual-axis network. First, FXR-mediated FFAR2 repression is highly pathological condition-dependent. This inhibitory effect is prominent under high-fat diet-induced metabolic dysfunction, whereas the two axes primarily exert synergistic compensatory functions to maintain metabolic homeostasis under physiological conditions.

Second, SCFAs exert concentration-dependent bidirectional effects on BSH-harboring microbiota. Low-to-moderate SCFA concentrations tend to enrich BSH-positive microbial taxa and enhance secondary BA synthesis. However, supraphysiological SCFA levels may disrupt bacterial transmembrane pH homeostasis,45 alter BSH substrate metabolism, constrain the proliferation of BSH-expressing microbes, and potentially reduce endogenous TGR5 ligand availability.

Further clinical and mechanistic studies are warranted to clarify the concentration threshold and population specificity of these bidirectional effects.

Immune Homeostasis Regulatory Hub: Antagonistic Balance Between the LPS-TLR4 and Trp-AhR Pathways

The LPS-TLR4/NF-κB Axis

Emerging clinical evidence suggests that alterations in T2DM-related inflammatory markers, including hs-CRP and IL-6, occur prior to detectable perturbations in glucose homeostasis. This temporal sequence implies that metaflammation may serve as an early pathogenic driver during the initial stage of T2DM development.46 Consistent with this notion, previous mechanistic work has validated that sustained metaflammation constitutes a core upstream mechanism contributing to the onset and progression of obesity, insulin resistance and nonalcoholic fatty liver disease.47 T2DM-associated gut microbiota disturbance diminishes microbial SCFA production, which may downregulate the intestinal tight junction proteins occludin and ZO-1. This alteration compromises intestinal epithelial integrity, increases intestinal permeability and promotes the development of intestinal barrier dysfunction commonly referred to as leaky gut.48 Impaired intestinal barrier facilitates the translocation of Gram-negative bacteria-derived LPS into the systemic circulation, thereby inducing metabolic endotoxemia, sustaining chronic low-grade systemic inflammation and promoting the development of peripheral insulin resistance.49

High-fat dietary patterns further amplify this pathological cascade. High-fat nutrition accelerates intestinal LPS translocation via chylomicron-mediated transport and enriches circulating Gram-negative microbial populations, which enhances intestinal LPS synthesis and release.50 Notably, the detection and quantification of circulating LPS pose substantial methodological challenges. Circulating LPS measurements are susceptible to experimental contamination, assay variability, sample handling artifacts and lipoprotein-related interference, which may undermine the reliability of standalone LPS quantification.

Considering these technical limitations, multiple clinical observational studies have reported moderately elevated circulating LPS levels in obese and T2DM patients relative to healthy individuals. Surrogate biomarkers including LPS-binding protein (LBP) and soluble CD14 are therefore widely adopted to assist the comprehensive evaluation of metabolic endotoxemia status in human cohorts.51 Of note, available human intervention trials indicate that pharmacological TLR4 inhibition can effectively reduce systemic inflammatory marker levels, although such targeted inhibition fails to improve core metabolic endpoints including glycemic control and insulin sensitivity.52

Circulating LPS specifically activates the membrane pattern recognition receptor TLR4, which triggers the downstream MyD88-dependent signaling cascade. This process sequentially recruits IRAK and TRAF6, thereby promoting the phosphorylation and activation of TAK1. Activated TAK1 initiates two parallel proinflammatory signaling branches, namely the IKKβ/NF-κB pathway and the JNK pathway.

The IKKβ cascade mediates IκBα degradation and subsequent NF-κB nuclear translocation, which drives the transcriptional expression of multiple proinflammatory mediators including TNF-α, IL-6 and IL-1β. In parallel, the JNK pathway activates AP-1 to further potentiate systemic inflammatory responses. Both downstream signaling axes may indirectly induce serine phosphorylation of IRS-1, suppress functional tyrosine phosphorylation of IRS-1, and inhibit the PI3K/Akt signaling cascade.

These molecular disturbances impair GLUT4 membrane translocation and cellular glucose uptake, ultimately contributing to the development of peripheral insulin resistance.53–56 In adipose tissue, sustained TLR4 signaling blunts the insulin-mediated antilipolytic effect, exacerbates lipotoxic stress, and further aggravates systemic metabolic dysfunction.57 Moreover, persistent TLR4 activation enhances oxidative stress and promotes pancreatic β-cell injury via JNK-dependent signaling and subsequent NLRP3 inflammasome activation.57,58

Sustained TLR4 signaling activation also upregulates the basal transcription levels of NLRP3, pro-IL-1β and pro-IL-18. Combined with mitochondrial ROS accumulation59,60 and lysosomal functional impairment, these transcriptional alterations facilitate the secondary activation of the NLRP3 inflammasome. Under various particulate stimulus conditions, lysosomal rupture can serve as a trigger for NLRP3 inflammasome activation.61 IL-1β acts as a pivotal proinflammatory cytokine that initiates stress responses in pancreatic β-cells.62 Accumulating evidence indicates that IL-1β (often in synergy with IFN-γ) can upregulate iNOS expression to promote nitric oxide generation, which exacerbates endoplasmic reticulum stress and ultimately contributes to pancreatic β-cell damage and dysfunction.63

Since the initial proposal of the metabolic endotoxemia theory,64 extensive subsequent research has refined this pathogenic framework. Multiple studies have confirmed that individuals with obesity and prediabetic insulin resistance exhibit compromised intestinal barrier integrity and elevated circulating LPS levels at early disease stages, even when systemic glucose tolerance remains within the normal physiological range. Translocated LPS triggers persistent systemic and pancreatic low-grade inflammation through the TLR4/NF-κB/NLRP3 signaling cascade, which progressively aggravates insulin resistance and drives the transition from prediabetes to overt T2DM.65 Collectively, these findings support a plausible mechanistic model whereby metabolic endotoxemia may function as an upstream initiating factor of glucose metabolic disturbance, rather than merely a secondary consequence of sustained hyperglycemia.

Nevertheless, this causal hierarchy has not been consistently validated across all human clinical cohorts, and inter-study heterogeneity remains apparent. Accordingly, intestinal barrier integrity and metabolic endotoxemia represent promising early therapeutic targets for T2DM intervention. Targeted modulation of these inflammatory signaling mediators may provide feasible strategies to delay or halt T2DM disease progression.

The Trp-AhR Axis

Gut microbiota-dependent metabolism of dietary Trp is essential for the maintenance of intestinal barrier integrity and the modulation of systemic immune and metabolic homeostasis, primarily through the activation of aryl hydrocarbon receptor (AhR) signaling. AhR activation can be induced by multiple ligand categories, including microbiota-derived indole metabolites, host-generated kynurenines (KYN), dietary components, and synthetic agonists or antagonists. The biological outcomes of AhR signaling are highly tissue-specific and context-dependent, which contributes to its diverse regulatory roles in metabolic and immune physiology.

As an indispensable aromatic amino acid for human metabolism, dietary Trp can be degraded by gut microbial tryptophanase to generate a series of indole derivatives, such as indole-3-acetic acid, indole-3-propionic acid, indole-3-aldehyde, and indole-3-acrylic acid. These microbial metabolites serve as pivotal endogenous agonistic ligands for intestinal AhR.66 In parallel with microbial Trp catabolism, host cells metabolize Trp via the KYN pathway, which is catalyzed by indoleamine 2,3-dioxygenase (IDO) and Trp 2,3-dioxygenase (TDO). The microbial indole pathway and host kynurenine pathway therefore constitute two competing branches of systemic Trp metabolism.

Metaflammation in T2DM is accompanied by markedly enhanced IDO activity. Although both Trp metabolic branches can proceed simultaneously under pathological conditions, inflammatory stress tends to shift Trp metabolic flux toward the KYN pathway, potentially limiting the substrate availability for microbial indole synthesis. The dynamic balance between these two pathways is modulated by multiple regulatory factors, including substrate abundance, host IDO1/TDO2 enzymatic activity, gut microbial community configuration, systemic inflammatory tone, and tissue-specific metabolic capacity.

Such pathway imbalance may reduce the production of intestinal AhR-activating indole metabolites and attenuate AhR-mediated protective signaling cascades.67 Meanwhile, the circulating kynurenine-to-tryptophan ratio (KTR) is frequently elevated in T2DM conditions, and this metabolic ratio correlates closely with impaired glycemic control. Accumulating clinical evidence suggests that KTR may serve as a promising auxiliary biomarker and predictive indicator for T2DM progression.68

AhR is broadly expressed in intestinal epithelial cells and mucosal immune cells, and intestinal AhR signaling driven by microbial indole metabolites mediates local protective responses that are functionally distinct from AhR activation in the liver, pancreas, systemic immune tissues and circulating compartments. Microbe-derived indole derivatives selectively activate intestinal AhR, which promotes the activation and function of type 3 innate lymphoid cells (ILC3s) and facilitates IL-22 secretion. Elevated IL-22 signaling further upregulates intestinal tight junction proteins and enhances mucin-2 synthesis in intestinal epithelial cells.69 In addition, IL-22 augments the expression of antimicrobial peptides including RegIIIγ and β-defensins to strengthen intestinal mucosal defense.70 Collectively, these multimodal protective effects stabilize intestinal barrier structure and mucosal immune homeostasis, restrict excessive translocation of gut-derived LPS and other microbial endotoxins, and mitigate the upstream triggers of systemic low-grade inflammation.

The metabolic regulatory functions of the Trp–AhR axis are not limited to intestinal mucosal protection. This signaling cascade modulates systemic glucose and lipid homeostasis via gut–liver and gut–pancreas crosstalk axes. Previous mechanistic studies have demonstrated that dietary cereal protein supplementation enriches intestinal indole metabolites and enhances AhR–NLK/FOXO1 signaling, which suppresses aberrant hepatic gluconeogenesis and ameliorates systemic glucose metabolic dysfunction.71 Notably, AhR-mediated metabolic regulation exhibits prominent tissue specificity and context-dependent activation characteristics.

Physiological AhR activation in the intestinal microenvironment predominantly confers metabolic protective effects. In contrast, pathologically elevated AhR expression and activation in systemic circulation and pancreatic tissues are positively associated with HbA1c levels and body mass index. Excessive pancreatic AhR signaling may also downregulate core β-cell functional genes including INS and PDX-1,72 implying that sustained pathological AhR overactivation potentially contributes to pancreatic β-cell dysfunction and impaired glucose-stimulated insulin secretion.

Emerging preclinical evidence further indicates that gut-derived indole metabolites serve as key upstream regulators of systemic glycolipid metabolism. In addition to canonical AhR signaling activation, microbial indole derivatives remodel glycolipid metabolic profiles and improve peripheral insulin sensitivity through the modulation of the PXR pathway.73 Indole metabolites alleviate metaflammation and stabilize whole-body metabolic homeostasis via multiple complementary mechanisms, including intestinal barrier reinforcement, suppression of metabolic endotoxemia, and promotion of intestinal GLP-1 secretion.74,75 Taken together, the preservation of microbiota-dependent tryptophan-indole metabolism and the maintenance of intestinal protective AhR signaling represent promising therapeutic targets for early intervention and pathological modification in T2DM. Targeted modulation of this axis may provide novel insights for the precise prevention and comprehensive management of T2DM and its associated metabolic complications.

Crosstalk Between the LPS-TLR4/NF-κB Inflammatory Pathway and the Trp–AhR Repair Pathway

The proinflammatory LPS-TLR4/NF-κB pathway and the Trp–AhR/IL-22 barrier repair pathway do not function as independent regulatory modules. Centered on intestinal barrier integrity and systemic inflammatory tone, these two cascades form a dynamically antagonistic regulatory network, as illustrated in Figure 2. Under physiological conditions, the two pathways maintain a homeostatic balance to sustain intestinal immune and metabolic stability.

In the T2DM pathological microenvironment, this balanced state is disrupted, with excessive proinflammatory TLR4/NF-κB activation occurring concurrently with impaired AhR-mediated barrier repair capacity.56,69 This pattern of exaggerated inflammatory activation and insufficient mucosal defense may underpin the persistent progression of metaflammation. Figure 2A schematic diagram of gut inflammation and insulin resistance mediated by the crosstalk between Trp-AhR and LPS-TLR4/NF-κB pathways in T2DM. Gut dysbiosis increases LPS and decreases indole-producing bacteria.

LPS activates TLR4/NF-κB, releasing pro-inflammatory cytokines (TNF-α, IL-6, IL-1β). These cytokines inhibit SERT to elevate extracellular 5-HT, and activate IDO1 to shift Trp metabolism toward kynurenine, reducing protective AhR ligands. The disrupted bidirectional AhR-NF-κB antagonism leads to decreased IL-22 and barrier disruption.

Subsequent LPS leakage into circulation establishes a vicious cycle, driving IRS-1 serine phosphorylation, GLUT4 downregulation, insulin resistance, and β-cell apoptosis, ultimately progressing T2DM. Solid lines indicate experimentally confirmed mechanisms, while dashed lines indicate hypothesized crosstalk mechanisms.The diagram illustrates the crosstalk between tryptophan-AhR and LPS-TLR4/NF-kappa B pathways in T2DM gut inflammation and insulin resistance. Gut lumen microenvironment shows increased LPS and decreased tryptophan.

LPS activates TLR4, leading to NF-kappa B activation and release of cytokines IL-6, IL-1 beta and TNF-alpha. These cytokines inhibit SERT, increasing extracellular 5-HT and activating IDO1, shifting tryptophan metabolism to kynurenine, reducing AhR ligands. This disrupts AhR-NF-kappa B antagonism, decreasing IL-22 and causing barrier disruption.

Increased intestinal permeability allows LPS and inflammatory factors into circulation, leading to IRS-1 phosphorylation, PI3K/Akt inhibition, GLUT4 translocation reduction and beta-cell apoptosis. Solid lines indicate confirmed mechanisms, dashed lines indicate hypothesized crosstalk.Diagram of gut inflammation and insulin resistance via Trp-AhR and LPS-TLR4/NF-kappa B pathways in T2DM.Abbreviations: AhR, aryl hydrocarbon receptor; EC, enterochromaffin; GLUT4, glucose transporter type 4; IDO1, indoleamine 2,3-dioxygenase; LPS, lipopolysaccharide; NF-κB, nuclear factor kappa-B; SERT, serotonin transporter; T2DM, type 2 diabetes mellitus; TLR4, Toll-like receptor 4; Trp, tryptophan; 5-HT, 5-hydroxytryptamine.

Under physiological homeostasis, microbiota-derived indole metabolites activate intestinal AhR signaling, promote IL-22 secretion from intestinal immune cells, upregulate tight junction proteins including occludin and ZO-1,56 and constrain excessive NLRP3 inflammasome activation.76 These cooperative effects help sustain intestinal barrier structural and functional integrity, limit microbial endotoxin translocation, and restrain the overinitiation of systemic inflammatory cascades. In the context of T2DM-associated gut microbiota disturbance, Trp metabolic flux is substantially rewired. Microbial indole synthesis is relatively suppressed, while host KYN pathway metabolism is preferentially enhanced.

This metabolic shift reduces the availability of endogenous AhR agonistic ligands and compromises the basal mucosal protective function of intestinal AhR signaling.

It is worth noting that excessive activation of the LPS-TLR4/NF-κB inflammatory signaling pathway can further exacerbate functional defects in the Trp-AhR signaling axis, creating a self-amplifying pathological loop. Chronic inflammatory stimulation upregulates IDO1, driving the host’s Trp metabolism to be reprogrammed toward the KYN pathway. Competitive consumption of Trp by the host and the gut microbiota limits the supply of substrates required for the microbiota to synthesize AhR-activating indole metabolites,77 thereby weakening the mucosal protective effects of gut AhR mediated by microbiota-derived ligands.69 This pathological interaction continuously amplifies inflammatory signals, reduces mucosal repair substrates, and induces progressive damage to the intestinal barrier, ultimately exacerbating metabolic disorders and insulin resistance.78

Multilayered transcriptional crosstalk further modulates the dynamic antagonism between the two pathways. LPS-induced NF-κB (RelA/p50) activation may upregulate basal AhR expression at the transcriptional level.79 Nevertheless, the anti-inflammatory potency of AhR is largely dependent on adequate ligand binding. Upon ligand activation, AhR independently interacts with STAT1 and Sp1, which synergistically restrict macrophage inflammatory responses.

This molecular interaction attenuates LPS-triggered NF-κB nuclear translocation and activation, downregulates HDC expression, and reduces histamine synthesis and release, thereby mediating negative inflammatory feedback.80,81

Under T2DM-related ligand-insufficient pathological conditions, transcriptionally upregulated AhR remains largely ligand-free and functionally inactive. Deficient ligand-dependent AhR signaling fails to trigger downstream AhR–STAT1 and AhR–Sp1 anti-inflammatory cascades and loses transcriptional restraint over NF-κB, which potentially contributes to sustained NF-κB overactivation and subsequent upregulation of HDC and proinflammatory mediators. In addition, unliganded AhR may competitively sequester limited intracellular transcriptional coactivators, interfering with physiological transcriptional programs responsible for intestinal homeostasis and metabolic regulation and further amplifying inflammatory disturbance.

This dysregulated circuit facilitates persistent inflammatory stimulation via the TLR4/NF-κB axis, driving continuous secretion of TNF-α, IL-6 and IL-1β. These cumulative inflammatory stresses progressively disrupt intestinal barrier integrity, aggravate endotoxin translocation, and exacerbate NLRP3 inflammasome activation. Such persistent pathological disturbances may induce endoplasmic reticulum stress and pyroptotic injury in pancreatic β-cells, potentially driving gradual β-cell secretory dysfunction and ultimately exacerbating systemic glucose metabolic disorders in T2DM.

NF-κB/NLRP3: A Convergence Node of Multi-Axis Imbalance and Metabolic Injury

Although the four intestinal metabolic signaling axes (SCFAs–AMPK, BA–FXR/TGR5, LPS–TLR4 and Trp–AhR) exhibit distinct spatial distribution characteristics and dominant biological functions, their downstream regulatory cascades governing insulin sensitivity and metaflammation may ultimately converge on the core inflammatory hub NF-κB/NLRP3. As illustrated in Figure 3, this integrative regulatory pattern is not a simple linear superposition of individual pathway effects. Instead, it follows a nonlinear pathological logic, wherein concurrent dysfunction of multiple homeostatic pathways removes parallel anti-inflammatory constraints and synergistically drives excessive NF-κB/NLRP3 activation.

This mechanism partially underpins our proposed conceptual model, which explains the progressive amplification of metaflammation and the gradual breakdown of metabolic homeostasis during T2DM development. Notably, this NF-κB/NLRP3 convergence model is primarily synthesized from independent preclinical mechanistic observations. Comprehensive multi-omics human cohort evidence to systematically validate this central hub hypothesis remains insufficient.

It should be emphasized that the NF-κB/NLRP3 axis currently serves as a plausible mechanistic convergence node rather than a fully validated clinical therapeutic target. Figure 3Regulation of NF-κB and NLRP3 inflammasome by gut microbial metabolites in T2DM. Left: microbial metabolites exert anti-inflammatory effects.

Bile acids (BAs) activate TGR5-cAMP-PKA signaling; SCFAs activate the AMPK-NAMPT-NAD+-SIRT1 axis; and tryptophan metabolites (IPA, IAld) activate AhR. Right: LPS triggers the TLR4/MyD88/NF-κB cascade (Signal 1) to prime NLRP3 and pro-IL-1β. Danger signals (eg, ATP, ROS) provide signal 2, inducing NLRP3 inflammasome assembly, caspase-1 activation, and mature IL-1β release.

Pathological context: in T2DM, the protective effects of these metabolites are proposed to be weakened, driving excessive inflammation. Solid lines indicate experimentally confirmed mechanisms; dashed lines represent hypothesized or controversial crosstalk (eg, PKA-mediated NLRP3 inhibition, AhR-mediated NF-κB DNA-binding blockade).The diagram illustrates the regulation of NF-kappa B and NLRP3 inflammasome by gut microbial metabolites in type 2 diabetes mellitus. On the left, secondary bile acids activate TGR5, leading to cAMP and PKA signaling, which phosphorylates p65 at serine 276.

Short-chain fatty acids activate the AMPK-NAMPT-NAD-SIRT1 axis. Tryptophan metabolites IPA and IAld activate AhR. These pathways influence NF-kappa B (p65/p50) activity.

On the right, LPS triggers the TLR4/MyD88/NF-kappa B cascade, priming NLRP3 and pro-IL-1 beta. Danger signals like ATP and ROS induce NLRP3 inflammasome assembly, activating caspase-1 and releasing mature IL-1 beta, leading to pyroptosis. Solid lines indicate confirmed mechanisms, while dashed lines show hypothesized crosstalk.

Activation and inhibition are marked with arrows.NF-kappa B & NLRP3 regulation by gut microbes in T2DM, showing pathways & interactions.Abbreviations: AhR, aryl hydrocarbon receptor; AMPK, AMP-activated protein kinase; BAs, bile acids; IL-1β, interleukin-1β; LPS, lipopolysaccharide; NF-κB, nuclear factor kappa-B; NLRP3, NOD-like receptor family pyrin domain containing 3; PKA, protein kinase A; SCFAs, short-chain fatty acids; SIRT1, sirtuin 1; T2DM, type 2 diabetes mellitus; TLR4, Toll-like receptor 4.

Under physiological metabolic homeostasis, AMPK, TGR5 and AhR collectively form a triple parallel anti-inflammatory braking system that constrains excessive NF-κB activation, maintaining basal systemic inflammation at a physiologically quiescent state via distinct molecular mechanisms. AMPK promotes SIRT1-dependent deacetylation, which targets the NF-κB p65 subunit to suppress its nuclear translocation and transcriptional activity.82 Meanwhile, SIRT1 enhances LKB1 activity through deacetylation modification, which reciprocally potentiates AMPK phosphorylation and function, forming a protective positive AMPK–SIRT1 feedback loop to sustain intestinal immune and metabolic homeostasis.83 The BA receptor TGR5 exerts multi-layered anti-inflammatory effects dependent on the cAMP–PKA signaling cascade. Upon Gs-protein coupling, TGR5 activates adenylyl cyclase to elevate intracellular cAMP levels and trigger downstream PKA signaling.

Via CREB phosphorylation mechanisms, activated PKA efficiently restricts NF-κB transcriptional activity.84,85 TGR5 can regulate the translation of transcription factors via the mTOR pathway, thereby suppressing the inflammatory response in macrophages.86 As a key gut-protective signaling pathway, AhR can counteract NF‑κB-mediated inflammatory responses by regulating the stability of the RelA/p65 protein, thereby limiting the transcription of pro-inflammatory genes.87

In the T2DM pathological microenvironment, simultaneous dysregulation of the four major intestinal metabolic axes contributes to the functional failure of this triple anti-inflammatory braking system. Deficient SCFA production substantially impairs basal AMPK activity; insufficient secondary BA synthesis blunts TGR5-dependent anti-inflammatory signaling; gut microbiota disturbance-induced Trp metabolic reprogramming reduces microbial indole metabolite generation, which compromises AhR nuclear translocation and abolishes its downstream anti-inflammatory efficacy. Concurrent impairment of multiple inhibitory pathways markedly increases the susceptibility of NF-κB to proinflammatory stimuli derived from the LPS–TLR4 cascade.

Even low-level portal venous LPS translocation resulting from intestinal barrier leakage may trigger robust NF-κB nuclear translocation and sustained inflammatory activation, initiating cascading proinflammatory responses. From this conceptual perspective, coordinated multi-axis imbalance is presumed to serve as a critical pathological foundation driving inflammatory progression and metabolic injury in T2DM. Compared with single-pathway dysfunction, combined multi-axis dysregulation confers more profound metabolic and inflammatory disturbances, thereby possessing greater potential research value and clinical translational implications, although systematic quantitative comparative analyses between single-pathway and multi-axis disorders remain limited in current studies.

NF-κB and the NLRP3 inflammasome constitute a tightly interconnected upstream–downstream inflammatory amplification hub. Canonical NF-κB activation transcriptionally upregulates the expression of pro-IL-1β, pro-IL-18 and other inflammatory precursor proteins, providing essential molecular substrates for subsequent NLRP3 inflammasome assembly and activation.88,89 In parallel, excessive LPS-TLR4 signaling induces mitochondrial ROS accumulation and intracellular K+ efflux, which synergistically facilitate NLRP3 oligomerization, ASC recruitment, and caspase-1 proteolytic maturation, ultimately driving inflammasome activation.90 It is worth noting that the activity of the NLRP3 inflammasome can be finely regulated by phosphorylation modifications mediated by multiple intestinal metabolic pathways, which are key regulatory components of this integrated inflammatory hub.

Under physiological conditions, TGR5, AMPK and AhR constrain aberrant NLRP3 activation through distinct post-translational regulatory mechanisms. BA-induced TGR5–cAMP–PKA signaling promotes NLRP3 Ser291 phosphorylation and subsequent ubiquitination, thereby blocking inflammasome assembly and activation.91 AMPK directly phosphorylates key functional residues of NLRP3 and suppresses its ATPase activity, which fundamentally inhibits the formation of functional NLRP3 inflammasome complexes.92,93 Microbial indole-derived AhR agonists, including IPA and IAld, derived from intestinal Trp metabolism, can effectively restrain NLRP3 inflammasome assembly and mature IL-1β secretion.94–96 Of note, the Trp metabolite kynurenine (KYN) exerts dose-dependent bidirectional regulatory effects on inflammatory signaling. Physiologically low concentrations of KYN may weakly activate AhR to mediate mild anti-inflammatory responses.97 However, in the context of T2DM, sustained inflammatory stress drives massive Trp metabolic flux toward the KYN pathway, leading to the depletion of protective indole AhR ligands and abnormal KYN accumulation.

This metabolic switch abrogates the inhibitory effects on the NF-κB/NLRP3 inflammatory hub and further eliminates endogenous anti-inflammatory barriers. Concurrent dysfunction of multiple NLRP3 inhibitory pathways significantly lowers the threshold of inflammatory activation, rendering the inflammatory cascade prone to persistent and uncontrolled activation.

Sustained overactivation of the NF-κB/NLRP3 inflammatory hub translates upstream gut microbiota disturbance and intestinal barrier dysfunction into systemic metabolic injury, with preferential detrimental effects on pancreatic β-cells and adipose tissue. In the pancreatic microenvironment, elevated mature IL-1β induces excessive iNOS expression and massive NO accumulation. Combined with chronic glucolipotoxicity, these stressors trigger endoplasmic reticulum stress and mitochondrial dysfunction in pancreatic β-cells.98 NLRP3/caspase-1-dependent pyroptosis further contributes to progressive β-cell loss and impaired insulin secretory function.

In adipose tissue, IL-1β synergizes with TNF-α to promote the polarization of adipose tissue macrophages toward the proinflammatory M1 phenotype, stimulate the secretion of chemokines such as CCL2, and induce abundant peripheral monocyte infiltration, ultimately establishing a persistent local chronic inflammatory microenvironment.99 Collectively, based on existing preclinical evidence, the NF-κB/NLRP3 inflammatory hub is proposed as a core mechanistic switch that bridges gut microbiota disturbance, intestinal barrier damage, systemic insulin resistance and progressive pancreatic β-cell dysfunction in T2DM.

Therapeutic Strategies Targeting the Four Major Signaling Axes

The gut microbiota plays a crucial regulatory role in the progression of T2DM through the above four core signaling pathways. Dysregulation of these interconnected axes contributes to intestinal barrier impairment, persistent metaflammation, and progressive insulin resistance, with the NF-κB/NLRP3 inflammatory hub serving as the central downstream convergence node. Although microbiota-targeted interventions hold promising translational potential, current therapeutic evidence remains heterogeneous across different strategies, and preclinical mechanistic observations cannot be fully equivalent to consistent clinical benefits.

It should be emphasized that improvements in signaling molecules and inflammatory biomarkers observed in cellular or animal models represent only surrogate mechanistic endpoints, and cannot be equated with clinical benefits in patients with type 2 diabetes. Only outcomes from human clinical trials can demonstrate clinical therapeutic value. Representative single-agent interventions are graded in Tables 1–4 (Table 1 footnotes).

This section systematically classifies representative intervention approaches, covering small-molecule agents, probiotics/synbiotics, natural products, and dietary strategies, with stratified evaluation of current evidence grades and efficacy characteristics. By summarizing the pros and cons of existing regimens, this review further discusses the translational bottlenecks of single and combined interventions, aiming to provide evidence-based insights for the development of precise microbiota-targeted therapeutic strategies for T2DM.Table 1Interventions Targeting the SCFAs‑AMPK Signaling AxisInterventionEvidence Category & Experimental ModelPrimary Molecular TargetsKey Mechanistic InteractionBeneficial OutcomesAdverse/Off-Target EffectsLimitationsReferencesSodium butyratePreclinical evidence (DIO mice; L6 myotubes)AMPK, HDACButyrate activates AMPK and inhibits HDAC, promotes mitochondrial biogenesisImproves insulin sensitivity; reduces adiposity and fasting glucoseNot reportedNo human T2DM data; microbiota-independent effects remain unclear[100]Mulberry leaf polysaccharidesPreclinical evidence (C57BL/6J T2DM mice)BCAA-catabolic enzymesRemodels gut microbiota, lowers circulating BCAA levelsAmeliorates hyperglycaemia, insulin resistance and hepatic steatosisNot reportedActive compounds uncharacterized; lacks human data[101]Astragalus polysaccharidesIn-vitro or mechanistic evidence (T2DM patient faecal fermentation)Gut microbial metabolic pathwaysModulates gut microbiota to enhance SCFA productionElevates SCFA levels; reduces inflammatory metabolitesNot reportedOnly in-vitro fermentation data; no in-vivo validation[102]Prebiotics (resistant starch, inulin-type fructans, FOS, GOS, β-glucan)Clinical trial evidence (RCTs, T2DM/prediabetes); Preclinical evidenceFFAR2/FFAR3, GLP-1RFermented by gut microbiota to generate SCFAs; stimulates intestinal incretin secretionImproves fasting glucose, HOMA-IR, gut-barrier functionBloating, flatulence (dose-dependent)High inter-individual response variability; limited long-term safety evidence[15,103]Butyrate-producing probiotics (Faecalibacterium prausnitzii)Human observational evidence (T2D patients); Preclinical evidence (db/db mice)Hepatic lipogenic enzymes, pro-inflammatory mediatorsSuppresses hepatic lipogenesis and systemic inflammatory responsesReduces insulin resistance, serum lipids and hepatic steatosisNot reportedSmall-size animal studies; active components undefined[104]Notes: Established clinical therapy: Treatments with regulatory approval and demonstrated clinical benefit in T2DM. Clinical trial evidence: Interventions tested in humans but not established for routine diabetes treatment.

Human observational evidence: Associations between metabolites, microbiota, or biomarkers and T2DM, without interventional testing. Preclinical evidence: Findings obtained from animal models including T2DM, obesity, diabetic nephropathy, or NAFLD. In vitro or mechanistic evidence: Data derived from cell culture, receptor-binding assays, enzymatic assays, or other molecular-level studies.

Hypothesis-generating strategy: Conceptually proposed therapeutic approaches or combinatorial regimens without direct pre-clinical or clinical validation. The “Hypothesis-generating strategy” category is retained per reviewer-suggested classification framework. This category is not applied in the present manuscript, as we only discuss published single-agent interventions; multi-axis combinatorial approaches are mentioned only as prospective outlooks for future research.Abbreviations: AMPK, adenosine monophosphate-activated protein kinase; BCAA, branched-chain amino acid; db/db, leptin receptor-deficient mice; DIO, diet-induced obese; FFAR2/FFAR3, free fatty acid receptor 2/3; FOS, fructo-oligosaccharides; GOS, galacto-oligosaccharides; GLP-1R, glucagon-like peptide-1 receptor; HDAC, histone deacetylase; HOMA-IR, homeostasis model assessment of insulin resistance; L6, L6 skeletal muscle myotubes; RCT, randomized controlled trial; SCFA, short-chain fatty acid; T2DM, type 2 diabetes mellitus.Table 2Interventions Targeting the Bile‑Acid‑FXR/TGR5 Signaling AxisInterventionEvidence Category & Experimental ModelPrimary Molecular TargetsKey Mechanistic InteractionBeneficial OutcomesAdverse/Off-Target EffectsLimitationsReferencesOCA (INT-747)Clinical trial evidence (Phase II, human T2DM-NAFLD patients)FXR (agonist)Activates FXR, modulates endogenous bile-acid homeostasis; suppresses hepatic inflammation and fibrosisImproves insulin sensitivity; reduces hepatic steatosis and fibrosisDose-dependent constipation; LDL-C elevation; ALT/AST rise leading to discontinuationSmall sample size; short-term trial; long-term clinical validation lacking[105]GW4064 (systemic FXR agonist)Preclinical evidence (db/db mice)FXR (agonist)FXR activation inhibits hepatic gluconeogenesis, alleviates hepatic ER-stress and apoptosisReduces fasting glucose; improves glucose tolerance, insulin resistance and hepatic steatosisNot reportedIntraperitoneal administration; no oral formulation; lipids unchanged in short-term animal study[106]Fexaramine (intestine-selective FXR agonist)Preclinical evidence (DIO, db/db mice)Intestinal FXR, TGR5Selectively activates intestinal FXR-TGR5 axis; promotes BAT thermogenesisMitigates weight gain; improves glucose tolerance and insulin sensitivity; lowers inflammatory mediatorsMinimal systemic absorption; mild, reversible liver enzyme elevation reported for derivative MET409Only pre-clinical animal data; human trials absent; gut microbiota contributes to its in-vivo effects[107]INT-777 (TGR5-specific agonist)Preclinical evidence (db/db mice; in-vitro human/mouse islet cells)TGR5 (agonist)TGR5-dependent cAMP signaling enhances GLP-1 secretion and pancreatic β-cell functionImproves glucose tolerance, HOMA-IR and insulin secretionNot reportedIntraperitoneal delivery; no human data; long-term safety remains unevaluated[34]INT-767 (FXR/TGR5 dual agonist)Preclinical evidence (multiple diabetic/obese mouse models)FXR, TGR5Co-activates FXR and TGR5; regulates lipid oxidation, ER-stress and renal inflammatory responsesExerts renal protective effects; no consistent glucose-lowering effect in STZ-modelNot reportedFails to produce hypoglycaemic benefit in some T2DM animal models; lacks human data; long-term safety unknown[108]Notes: See Table 1 for definitions of evidence categories.Abbreviations: ALT, alanine aminotransferase; AST, aspartate aminotransferase; BAT, brown adipose tissue; cAMP, cyclic adenosine monophosphate; DIO, diet-induced obese; ER, endoplasmic reticulum; FXR, farnesoid X receptor; GLP-1, glucagon-like peptide-1; HOMA-IR, homeostasis model assessment of insulin resistance; LDL-C, low-density lipoprotein cholesterol; NAFLD, non-alcoholic fatty liver disease; STZ, streptozotocin; T2DM, type 2 diabetes mellitus; TGR5, G-protein-coupled bile acid receptor 5.Table 3Interventions Targeting the LPS‑TLR4/NF‑κB Signaling AxisInterventionEvidence Category & Experimental ModelPrimary Molecular TargetsKey Mechanistic InteractionBeneficial OutcomesAdverse/Off-Target EffectsLimitationsReferencesTLR4 antagonists (Eritoran/E5564, TAK-242)Preclinical evidence (rats with HFD/lipid infusion); Clinical trial evidence (human RCT)TLR4Blocks LPS-triggered TLR4-NF-κB inflammatory signaling.

Improves insulin resistance in rodents, yet fails in human trials.Ameliorates lipid-provoked insulin resistance in rats; no metabolic benefit observed in humansTransient injection-site phlebitisPoor rodent-to-human translation; small human sample size[52,109,110]Natural products (Puerarin, Celastrol)Preclinical evidence (HFD-IR/db/db mice and cell assays)NF-κB, ER-stress signalingSuppresses ER-stress and NF-κB-driven inflammation, restores insulin signaling.Reduces hyperglycaemia, insulin resistance and hepatic/renal injuryCelastrol: reduced food intake and body-weight loss; Puerarin: no obvious toxicityPreclinical-only data; long-term safety unevaluated[111,112]SGLT2 inhibitors (Empagliflozin, Dapagliflozin, Canagliflozin)Established clinical therapy (T2DM patients); Preclinical evidenceTLR4-NLRP3 inflammasome, macrophage polarizationInhibits TLR4-NLRP3-mediated inflammation; shifts macrophage toward anti-inflammatory M2 phenotype.Lowers systemic inflammatory markers; improves renal and cardiac functionMild pro-inflammatory effects observed in partial in-vitro assaysEvidence mainly derived from review summaries, limited original mechanistic data[113]GLP-1 receptor agonist (Liraglutide)Preclinical evidence (STZ-diabetic mice; rat mesangial cell studies)TLR4-MyD88-NF-κBDownregulates TLR4-related inflammatory cascades to alleviate glomerular inflammation and fibrosis.Mitigates proteinuria, glomerular hypertrophy and renal fibrosisNot reportedEffects partially independent of TLR4; STZ-T1DM model cannot fully recapitulate T2DM-DKD; no human mechanistic data[114]Notes: See Table 1 for definitions of evidence categories.Abbreviations: DKD, diabetic kidney disease; ER, endoplasmic reticulum; GLP-1, glucagon-like peptide-1; HFD, high-fat diet; IR, insulin resistance; LPS, lipopolysaccharide; MyD88, myeloid differentiation primary response 88; NF-κB, nuclear factor kappa-light-chain-enhancer of activated B cells; NLRP3, NOD-like receptor thermal protein domain-associated protein 3; RCT, randomized controlled trial; SGLT2, sodium-glucose cotransporter 2; STZ, streptozotocin; T1DM, type 1 diabetes mellitus; T2DM, type 2 diabetes mellitus; TLR4, toll-like receptor 4.Table 4Interventions Targeting the tryptophan‑AhR Signaling AxisInterventionEvidence Category & Experimental ModelPrimary Molecular TargetsKey Mechanistic InteractionBeneficial OutcomesAdverse/Off-Target EffectsLimitationsReferencesTryptophan / indole-based AhR ligands (Trp, I3C, IPA, IAA, IAld)Preclinical evidence (Akita/HFD/db/db mice); Human observational evidenceAhRActivates AhR, strengthens gut barrier, suppresses inflammation and enhances insulin sensitivity.Reduces glucose and insulin resistance; improves glucose toleranceCertain metabolites exert context-dependent pro-inflammatory effectsLimited human interventional evidence; clinical validation required[73,115,116]Dietary fiber & polysaccharides (wheat bran, AMP, HON)Preclinical evidence (diabetic mice); In-vitro mechanistic evidenceAhR, GLP-1Remodels gut microbiota to generate AhR-activating indole metabolites; improves gut barrier and GLP-1 release.Ameliorates hyperglycaemia, dyslipidaemia, hepatic steatosis and systemic inflammationNo obvious toxicity at high dosesPreclinical-only data; active constituents remain unclear[117–119]Probiotic (Parabacteroides distasonis)Preclinical evidence (T2D rats; Caco-2 cell assays)AhR, IL-22, TLR4-NF-κBProduces indoleacrylic acid to activate AhR-IL-22 axis; repairs intestinal barrier and dampens NF-κB-mediated inflammation.Improves insulin resistance, lipid profile and organ injuryNot reportedLacks human data; efficacy not confirmed in AhR-knockout models[69]Notes: See Table 1 for definitions of evidence categories.Abbreviations: AhR, aryl hydrocarbon receptor; AMP, AMP-activated protein kinase; HON, honokiol; I3C, indole-3-carbinol; IAA, indole-3-acetic acid; IAld, indole-3-aldehyde; IL-22, interleukin-22; IPA, indole-3-propionic acid; Trp, tryptophan; T2D, type 2 diabetes; HFD, high-fat diet; NF-κB, nuclear factor kappa-light-chain-enhancer of activated B cells; TLR4, toll-like receptor 4.

Targeting the SCFAs–AMPK Energy-Sensing Axis

Table 1 summarizes current intervention strategies targeting the SCFAs–AMPK axis, including relevant experimental models, molecular mechanisms, therapeutic outcomes, and translational limitations. Intestinal SCFA insufficiency and subsequent AMPK signaling impairment are closely associated with intestinal barrier disruption, metaflammation, and insulin resistance, rendering the SCFAs–AMPK axis a promising microbiota-related therapeutic target for T2DM. Both novel microbiota-modulating candidates and conventional antidiabetic agents, especially metformin, may exert metabolic benefits through the regulation of SCFA-dependent AMPK signaling.

Current targeting strategies encompass clinical medications, dietary prebiotics, probiotics, microbial metabolites, and natural compounds, with distinct disparities in evidence maturity and clinical transformation potential.

As a first-line clinical therapeutic for T2DM, metformin improves systemic metabolic homeostasis via dual modulation of host AMPK activity and gut microbial SCFA metabolism. It directly activates AMPK by elevating intracellular AMP/ATP ratios and indirectly remodels gut microbial composition to enrich SCFA-producing taxa, increase endogenous SCFA abundance, stabilize intestinal barrier integrity, and alleviate peripheral insulin resistance.120 Despite stable hypoglycemic outcomes validated in clinical trials, the microbiota-modulating effects of metformin are highly susceptible to individual gut microbial heterogeneity, which may contribute to variable therapeutic responses among different patients.121 In addition, interference by indigenous microbial metabolites may potentially attenuate AMPK activation, further limiting the consistency of metabolic improvement. In contrast, sodium butyrate supplementation may ameliorate glycolipid metabolic disorders and inflammatory phenotypes via AMPK activation and HDAC inhibition in multiple preclinical diabetic models, while robust clinical evidence supporting its efficacy in T2DM patients remains insufficient.100 Natural polysaccharides derived from mulberry leaves and astragalus are capable of optimizing gut microbial structure and enhancing intestinal SCFA production in animal models and in vitro fermentation systems, thereby relieving metabolic dysregulation.101,102 However, unclear active ingredients, undefined optimal clinical dosages, and insufficient human validation currently restrict their clinical application as exploratory adjuvant interventions.

Prebiotic dietary fibers represent one of the most clinically mature non-pharmacological strategies targeting the SCFAs–AMPK axis. Accumulating randomized controlled trials have demonstrated that dietary fiber supplementation may improve glycemic control, insulin sensitivity, and intestinal barrier function in prediabetic and T2DM populations.15,103 Nevertheless, the metabolic effects of combined fiber and metformin intervention remain controversial. Several studies have reported potential synergistic metabolic benefits of specific fiber subtypes, whereas emerging evidence indicates that certain dietary fibers, such as psyllium, may interact with metformin and potentially attenuate its hypoglycemic efficacy under specific conditions, suggesting the necessity of individualized evaluation for combined medication regimens.122 Furthermore, dose-dependent gastrointestinal adverse reactions and interindividual variability driven by gut microbial diversity hinder standardized clinical promotion, and long-term safety assessment data are still limited.

Butyrate-producing probiotics have yielded favorable observational human evidence and preclinical results in lipid modulation and anti-inflammation, yet inadequate high-quality clinical trials and undefined functional components substantially limit their translational value.104

In summary, current SCFAs–AMPK-targeted interventions cover clinically validated drugs, maturing dietary regimens, and exploratory preclinical candidates with hierarchical evidence levels. Importantly, preclinical improvements in signaling activity and inflammatory biomarkers do not necessarily translate to reliable clinical T2DM remission. Even for clinically established metformin, its microbiota-dependent metabolic benefits are constrained by individual microbial variation.

The major translational bottleneck of SCFAs–AMPK-related therapies lies in the notable gap between solid preclinical mechanistic findings and insufficient high-grade clinical evidence, which restricts the widespread clinical application of novel microbiota-targeted strategies for T2DM treatment.

Targeting the BA–FXR/TGR5 Metabolic Sensing Axis

Table 2 systematically summarizes current intervention strategies targeting the BA–FXR/TGR5 axis, including stratified experimental evidence grades, molecular targets, core mechanisms, therapeutic outcomes, adverse events, and key translational limitations. Gut microbiota shapes intestinal BA profiles through deconjugation, dehydrogenation, and epimerization, thereby modulating the activation status of FXR and TGR5 to regulate systemic glycolipid metabolism and immune homeostasis. Hepatic and intestinal FXR signaling inhibits aberrant hepatic gluconeogenesis and mitigates lipotoxic injury, while intestinal TGR5 activation stimulates L-cell GLP-1 secretion to enhance glucose-dependent insulin release.

Collectively, these two receptor cascades constitute indispensable metabolic protective signaling networks in the gut–liver axis.

Tissue selectivity is a critical determinant of the efficacy, safety, and translational potential of synthetic FXR agonists. Obeticholic acid (OCA, INT-747), a clinically approved FXR agonist, is currently licensed for the treatment of primary biliary cholangitis. Phase II clinical trials have demonstrated that low-dose OCA may improve insulin sensitivity and hepatic steatosis in T2DM patients complicated with NAFLD, whereas such metabolic benefits are not consistently observed in patients with uncomplicated T2DM.105 Notably, OCA has not obtained regulatory approval for T2DM treatment, and existing human hepatic metabolic outcomes cannot fully validate the clinical applicability of FXR-targeted therapy for general T2DM populations.

Synthetic non-selective FXR agonists, including GW4064, and intestine-restricted agonists, such as fexaramine, have yielded favorable preclinical outcomes, including improved insulin sensitivity, reduced hyperglycemia, and alleviated hepatic steatosis in diabetic rodent models.106,107 However, these findings are limited to preclinical experimental systems and lack supportive human clinical evidence.

Systemic full FXR agonism is accompanied by notable translational challenges and safety concerns. Pruritus is a well-documented on-target adverse reaction in clinical application. In addition, excessive intestinal FXR activation may suppress intestinal L-cell GLP-1 secretion, a regulatory phenotype predominantly validated in cell experiments and rodent models, which cannot be directly extrapolated to human physiological and pathological conditions.39,123 To date, no FXR agonists have been approved for T2DM monotherapy, and their clinical application is currently restricted to adjuvant treatment for cholestatic liver diseases and NAFLD with concurrent hepatic injury.

Combination regimens of FXR agonists and GLP-1 receptor agonists have been proposed to maximize metabolic benefits, yet this synergistic strategy remains hypothesis-driven and lacks sufficient direct clinical validation.124

As another pivotal therapeutic target of the BA signaling axis, the selective TGR5 agonist INT-777 may ameliorate glucose intolerance and preserve pancreatic β-cell function in multiple in vitro and diabetic animal models.34 Nevertheless, the ubiquitous tissue distribution of TGR5 complicates the development of systemic TGR5-targeted drugs. Pruritus is a confirmed on-target adverse effect of partial systemic TGR5 agonists in clinical practice, while arrhythmia is currently regarded as a preclinical theoretical risk without definitive human adverse event evidence.125–127 Accordingly, the development of intestine-selective TGR5 agonists that minimize systemic off-target effects has become a priority direction for subsequent targeted drug research.128

Dual FXR/TGR5 agonists are designed to integrate the complementary protective functions of the two BA receptors. Preclinical studies have demonstrated that dual agonism may alleviate dyslipidemia, amplify GLP-1-mediated glycemic regulation, and suppress NF-κB/NLRP3-dependent inflammatory cascades, additionally exerting renoprotective effects in mouse models of diabetic nephropathy.108 However, inconsistent hypoglycemic efficacy across different animal strains and the complete absence of human clinical data severely limit their clinical translation. In contrast, BA sequestrants modulate systemic FXR signaling by disrupting enterohepatic BA circulation.

Supported by preliminary clinical trial evidence, these agents confer moderate glycolipid-improving effects with acceptable safety profiles, and are currently applied as adjuvant lipid-lowering therapies in dyslipidemic T2DM patients.

Importantly, mechanistic improvements and phenotypic metabolic benefits observed in cell and animal studies do not equate to validated clinical endpoints regarding glycemic control, insulin sensitivity, or organ protection in human T2DM. Even for candidate agents with available clinical data, their metabolic benefits tend to be confined to specific patient subgroups rather than general T2DM populations. Current combinatorial BA-targeted strategies face multiple translational bottlenecks, including potential pharmacokinetic incompatibility, suboptimal target engagement and dosage regimens, strict requirements for tissue-selective activation, interindividual gut microbial heterogeneity, and the absence of mature patient stratification frameworks.

In addition, unregulated excessive immune modulation and unintended gut microbial compositional shifts may bring potential safety risks. Further high-quality preclinical verification and well-stratified, large-scale human clinical trials are therefore essential to promote the precise clinical transformation of BA-targeted therapeutic strategies for T2DM.

Targeting the LPS–TLR4/NF-κB Inflammatory Axis

Table 3 summarizes representative therapeutic interventions targeting the LPS–TLR4/NF-κB inflammatory axis, systematically outlining stratified evidence grades, molecular targets, core mechanistic actions, metabolic and inflammatory outcomes, adverse events, and unresolved translational limitations. T2DM pathogenesis is closely coupled with progressive gut microbiota disturbance, which is typically characterized by the depletion of beneficial butyrate-producing commensal bacteria and the overgrowth of pro-inflammatory Gram-negative pathobionts. Such microbial dysregulation facilitates intestinal endotoxin (LPS) accumulation and translocation, driving sustained activation of the downstream TLR4/NF-κB inflammatory cascade.

Accumulating preclinical animal studies and human clinical observations have validated pathological overactivation of this axis in T2DM, wherein elevated intestinal TLR4 expression and enhanced peripheral NF-κB transcriptional activity are positively correlated with the severity of chronic metaflammation.129,130

Current interventions targeting the LPS–TLR4/NF-κB axis primarily encompass microbiota-remodeling strategies, natural product-derived agents, selective TLR4 antagonists, and conventional antidiabetic medications. Microbiota-oriented approaches, including probiotic and prebiotic supplementation, may reduce intestinal LPS burden and mitigate upstream TLR4-mediated inflammatory signaling via reshaping gut microbial composition.131 Selective TLR4 antagonists, such as Eritoran/E5564 and TAK-242, are capable of blocking LPS-triggered TLR4/NF-κB signaling and exert robust anti-inflammatory effects in preclinical rodent models. Although clinical evidence indicates that these antagonists can transiently reduce circulating pro-inflammatory mediators in human subjects, existing trials fail to identify significant improvements in glycemic control or systemic metabolic endpoints in T2DM patients, revealing a prominent translational gap between preclinical mechanistic efficacy and clinical therapeutic outcomes.52,109,110

Multiple natural products, including puerarin and celastrol, have been reported to suppress NF-κB-driven inflammatory responses and endoplasmic reticulum stress, thereby restoring impaired insulin signaling in HFD-induced and db/db diabetic mouse models and in vitro cellular systems.111,112 Preclinical findings have demonstrated their potential metabolic protective effects, including ameliorated hyperglycemia, improved insulin sensitivity, and alleviated hepatic and renal inflammatory injury. Nevertheless, these beneficial phenotypes are exclusively confined to preclinical experimental settings; rigorous human validation in T2DM populations remains absent, and the long-term safety and tolerability of these natural compounds have not been fully evaluated.

SGLT2 inhibitors and GLP-1 receptor agonists, such as liraglutide are first-line clinically approved therapeutics for T2DM management. Beyond their canonical hypoglycemic functions, emerging preclinical data suggest that these agents may constrain TLR4/NLRP3/NF-κB-related inflammatory cascades and confer renal protective effects.113,114 Notably, such anti-inflammatory activities represent secondary ancillary effects rather than their core therapeutic mechanisms. Furthermore, the extent to which these metabolic benefits rely on microbiota-dependent TLR4/NF-κB modulation remains poorly defined, and direct human mechanistic evidence supporting this regulatory pathway is still limited.

Notably, the TLR4/NF-κB axis exhibits context-dependent dual regulatory roles under physiological and pathological conditions. Moderate basal activation of this signaling cascade maintains intestinal innate immune homeostasis and host defense, whereas persistent excessive activation drives metaflammation, insulin resistance, and progressive diabetic complications. Accordingly, sustained non-selective systemic inhibition of this axis may disrupt physiological immune surveillance and host defensive function, theoretically increasing potential risks of infection and abnormal tissue proliferation.132 In addition, chronic inflammatory stress is not the dominant pathological driver in all T2DM patients; a subset of patients predominantly present with glycolipid metabolic disorders and pancreatic β-cell dysfunction without prominent inflammatory phenotypes.133 Therefore, single-agent blockade of the LPS–TLR4/NF-κB axis cannot achieve universal therapeutic efficacy, and targeted interventions may only yield tangible benefits for T2DM subgroups characterized by high inflammatory activity.

Consistent with the aforementioned therapeutic axes, preclinical signaling improvements and inflammatory biomarker remission do not necessarily translate to validated clinical glycemic, insulin-sensitizing, or organ-protective outcomes in human T2DM patients.

Future clinical translation of LPS–TLR4/NF-κB-targeted therapies faces multiple practical bottlenecks, including the demand for precise dosage optimization and specific target engagement strategies, the requirement for intestine-restricted targeted delivery to avoid systemic immune disturbance, substantial interindividual heterogeneity of gut microbial profiles, and the lack of standardized patient stratification systems for inflammatory phenotype classification. Additionally, excessive non-specific immune suppression may bring unneglectable safety risks. Multi-axis combined intervention strategies have been proposed to optimize therapeutic efficacy and reduce single-drug adverse effects, yet such combinatorial regimens remain hypothesis-driven.

Further systematic preclinical risk verification and well-designed, stratified large-scale clinical trials are urgently needed to advance the precise translational application of inflammatory axis-targeted therapies for T2DM.

Targeting the Trp–AhR Protective Metabolic Axis

Table 4 comprehensively summarizes current interventions targeting the intestinal Trp–AhR signaling axis, including stratified evidence grades, molecular targets, core mechanistic actions, metabolic and protective outcomes, potential adverse effects, and key translational constraints. In the context of T2DM, intestinal Trp metabolic flux is predominantly diverted from microbial indole synthesis toward the proinflammatory KYN pathway, leading to progressive depletion of gut-derived AhR-activating indole metabolites and impaired AhR-dependent intestinal anti-inflammatory defense. Restoration of Trp metabolic homeostasis and supplementation of AhR agonistic indole metabolites are therefore regarded as promising strategies to rescue the dysfunctional Trp–AhR axis.

Mechanistically, effective AhR activation reinforces intestinal barrier integrity, modulates intestinal macrophage polarization, and improves peripheral insulin signal transduction. Multiple indole-derived metabolites have yielded favorable mechanistic and metabolic protective phenotypes in diverse diabetic animal models.115,116 Although human observational studies support a correlative relationship between circulating indole-AhR metabolite abundance and systemic metabolic status, high-quality interventional clinical evidence validating their therapeutic efficacy in T2DM remains limited.

Exogenous Trp-derived AhR ligands, including Trp, I3C, IPA, IAA, and IAld, exert beneficial metabolic effects in preclinical T2DM models (HFD-fed, db/db, and Akita mice), including attenuated hyperglycemia and improved insulin sensitivity and glucose tolerance. Nevertheless, certain AhR-active metabolites exhibit context-dependent dual regulatory properties and may trigger off-target proinflammatory responses under specific pathological conditions. Of note, robust human interventional evidence for these metabolites is largely absent, and their clinical efficacy and safety in T2DM patient populations remain unvalidated.

Candidate molecules such as 5-HIAA and 5-methoxytryptamine confer hepaticprotective and anti-inflammatory effects in cellular and animal mechanistic studies,73,116 while these preclinical observations cannot be directly extrapolated to human T2DM pathophysiology and clinical outcomes.

Gut microbiota-oriented interventions represent another core strategy to reconstruct endogenous Trp–AhR homeostasis, mainly including dietary fibers, natural polysaccharides, and probiotic supplementation, with current evidence predominantly derived from preclinical experiments and in vitro mechanistic assays. Dietary fibers and natural polysaccharides remodel gut microbial community structure to promote the biosynthesis of indole-based AhR ligands, thereby enhancing intestinal barrier function and facilitating GLP-1 secretion. Preclinical studies have verified their potential to ameliorate hyperglycemia, dyslipidemia, hepatic steatosis, and systemic inflammation; however, the precise functional active components and optimal intervention dosages remain incompletely clarified.117–119 The probiotic strain Parabacteroides distasonis has been shown to trigger indoleacrylic acid-dependent AhR–IL-22 activation, restore intestinal barrier integrity, and alleviate NF-κB-mediated inflammatory injury in T2DM rat models and cellular experiments.

Its preclinical metabolic benefits cover improved glycolipid profiles and attenuated visceral and hepatic organ damage, whereas corresponding human clinical validation is still lacking, and definitive mechanistic verification based on AhR-knockout experimental systems remains insufficient.

Direct pharmacological modulation of AhR or Trp metabolic rate-limiting enzymes provides an alternative translational direction for Trp–AhR axis intervention. Tapinarof, a clinically approved AhR modulator for autoimmune diseases, possesses well-established clinical safety profiles, while its metabolic protective efficacy and therapeutic value in T2DM have not been clinically validated and are currently limited to exploratory preclinical mechanistic research. The candidate AhR ligand AGT-5 exerts protective immunomodulatory effects in type 1 diabetic animal models, yet its efficacy and applicability in T2DM remain untested.134 IDO and TDO are critical rate-limiting enzymes that drive Trp flux toward the KYN pathway.

Under chronic inflammatory conditions, elevated IDO/TDO activity consumes intestinal indole-AhR precursors and disrupts AhR-mediated immune and metabolic homeostasis. Indoximod is primarily regarded as an IDO inhibitor with complex pharmacology. Although IDO inhibition has been hypothesized to reverse Trp metabolic reprogramming and rescue microbial indole generation, direct experimental evidence supporting this metabolic redirection remains inadequate, and its immunometabolic effects cannot be attributed solely to IDO inhibition.

Complex immunometabolic side effects therefore cannot be excluded.

Consistent with the other three intestinal signaling axes, preclinical mechanistic improvements and phenotypic protective outcomes observed in cell culture and animal models do not equate to validated clinical glycemic, insulin-sensitizing, or organ-protective efficacy in human T2DM. Notably, AhR signaling exhibits strict context-dependent dual functions; either insufficient or excessive AhR activation may induce intestinal immune disturbance and metabolic dysregulation. Accordingly, single-agent intervention is difficult to fully restore Trp–AhR axis homeostasis.

Combinatorial regimens integrating AhR modulators, IDO inhibitors, and microbiota-reshaping strategies have been proposed to achieve precise fine-tuning of Trp metabolic balance, whereas such synergistic strategies remain hypothesis-driven and face multiple translational barriers. Further well-designed, stratified preclinical verification and standardized human clinical trials are essential to promote the clinical transformation of Trp–AhR-targeted therapeutic strategies for T2DM.

Conclusion and Future Prospects

This review summarizes the core mechanisms by which gut microbiota dysregulation mediates metabolic and immune dysfunction in the context of T2DM, establishing an integrated regulatory network centered on four pivotal intestinal signaling axes, including the SCFAs–AMPK energy-sensing axis, BA-FXR/TGR5 metabolic regulatory axis, LPS–TLR4/NF-κB proinflammatory axis, and Trp–AhR immune homeostatic axis. Based on stratified preclinical evidence regarding microbial metabolic sensing, immune homeostasis maintenance, and multi-signal convergent inflammatory injury, this review highlights that T2DM progression cannot be fully explained by the independent dysfunction of individual signaling pathways. Instead, the progressive deterioration of metabolic homeostasis may stem from disordered crosstalk and disrupted dynamic balance among multiple microbiota-dependent signaling cascades.

This network-based conceptual framework suggests that coordinated multi-axis imbalance confers stronger pathological implication in T2DM pathogenesis than isolated single-pathway defects, although direct comparative clinical evidence remains insufficient to validate this hypothesis.

From the perspective of metabolic regulation, the SCFAs–AMPK and BA-FXR/TGR5 axes jointly constitute the core microbiota-associated energy-sensing hub, whose synergistic crosstalk sustains systemic glycolipid homeostasis, energy metabolism, and mitochondrial function. Insufficient cooperative protective outputs from these two axes serve as a critical upstream driver of metabolic disruption and declining insulin sensitivity. In terms of immune-inflammatory modulation, the proinflammatory LPS–TLR4/NF-κB cascade and the barrier-protective, immune-stabilizing Trp–AhR axis form a dynamic antagonistic balance.

The breakdown of this bidirectional regulatory equilibrium is presumed to facilitate intestinal barrier impairment, microbial endotoxin translocation, and persistent low-grade metaflammation. Mechanistically, the NF-κB/NLRP3 inflammasome complex acts as a plausible downstream convergence node that integrates upstream metabolic and immune dysregulation signals, amplifies inflammatory cascades, and potentially accelerates insulin resistance and T2DM progression.

Collectively, this review refines two core regulatory patterns of the gut microbial signaling network during T2DM development: a physiological multi-axis synergistic mode dominated by AMPK/FXR/TGR5/AhR-mediated protective signaling, and a pathological multi-axis imbalance mode characterized by barrier disruption and NF-κB/NLRP3 inflammatory hyperactivation. This paradigm revises the previous single-pathway-centered understanding of T2DM pathogenesis, emphasizes the networked and interactive features of microbiota-derived metabolic injury, and further complements the upstream pathogenic framework of insulin resistance.

Despite accumulating preclinical advances in elucidating individual axis functions, several critical limitations remain in current research. First, existing studies predominantly focus on single-target and single-pathway mechanistic exploration, lacking systematic investigation of crosstalk and hierarchical interactions between energy metabolic and immune-inflammatory modules. The spatiotemporal regulatory patterns, upstream–downstream logical relationships, and functional weights of the four axes during T2DM progression remain poorly defined.

Second, as the core convergent hub of multi-axis dysfunction, the dynamic activation characteristics and signal-specific response mechanisms of the NF-κB/NLRP3 inflammasome to distinct microbial-derived metabolic and inflammatory stimuli require further in-depth clarification. Third, most current mechanistic conclusions are derived from in vitro experiments and rodent models; high-quality human cohort analysis and stratified clinical intervention evidence supporting this multi-axis network model are still lacking. Furthermore, interindividual gut microbial heterogeneity and patient phenotypic diversity contribute to variable therapeutic responses, which have not been adequately considered in existing translational studies.

In summary, dysregulated crosstalk among the four major gut microbial signaling axes is hypothesized to trigger metabolic-immune homeostasis collapse, representing a preclinically plausible upstream mechanism underlying progressive insulin resistance and T2DM development. Nevertheless, constrained by insufficient human validation and population heterogeneity, this network-centered pathogenic model requires further systematic clinical verification and cannot be universally extrapolated to all T2DM patient subgroups.

Notably, multi-axis combinatorial therapeutic strategies face multiple practical translational hurdles that should not be overlooked. First, pharmacokinetic interactions may occur between co-administered agents, altering drug absorption, enterohepatic circulation or metabolic clearance and changing effective target-site exposure. Second, target engagement and dose-setting for combinatorial regimens cannot be simply extrapolated from single-agent experiments.

Sub-optimal dosing may lead to insufficient target activation or undesired off-target effects. Third, tissue-selective action is critical: many metabolic-immune modulators exert opposing biological effects across liver, intestine, pancreas and immune cells; systemic non-selective delivery may compromise local beneficial functions and trigger organ-specific adverse outcomes. Fourth, substantial inter-individual gut microbiota variability modulates metabolite production and drug biotransformation, contributing to heterogeneous therapeutic responses.

Fifth, universal combinatorial regimens are unlikely to benefit all patients; patient phenotyping and stratification according to inflammatory status, intestinal barrier function and axis-specific dysfunction will be required to identify suitable candidate subgroups. Finally, combinatorial immune-metabolic modulation brings theoretical safety risks, including excessive immune suppression that may impair host immune surveillance, alongside unintended and potentially pathogenic reshaping of gut microbial communities. Collectively, these practical barriers highlight that combinatorial multi-target interventions remain predominantly hypothesis-driven, requiring rigorous pre-clinical risk assessment and well-stratified human clinical validation.

Future research should transcend the traditional single-target, single-pathway research paradigm and focus on the integrated multi-axis network composed of core nodes including AMPK, FXR, TGR5, TLR4, and AhR. Priority should be given to exploring the interactive mechanisms between energy-sensing and immune homeostatic hubs, deciphering the dynamic rules of synergistic multi-axis dysfunction during disease progression, and developing precise combinatorial intervention strategies targeting network imbalance rather than individual molecules. Such systematic investigations will provide more comprehensive theoretical support and translational insights for the development of novel microbiota-targeted preventive and therapeutic strategies for T2DM.