Beyond glycemic control: How semaglutide reshapes intestinal neuroception via transforming growth factor-β/brain-derived neurotrophic factor signaling hubs

Abstract

The intestinal lumen harbors a dynamic sensory and responsive network composed of enteric neurons (expressing ion channels such as transient receptor potential vanilloid 1 and K⁺ channels), enteroendocrine cells, gut microbiota and neuroimmune cells. The enteric nervous system (ENS) - a “second brain” autonomously regulates gastrointestinal functions while closely communicating with the central nervous system via the gut-brain axis (GBA). Its dysfunction is closely associated with metabolic diseases such as type 2 diabetes mellitus and obesity, as well as functional gastrointestinal disorders characterized by visceral hypersensitivity. Semaglutide, a long-acting glucagon-like peptide-1 receptor (GLP-1R) agonist, has demonstrated remarkable efficacy in treating type 2 diabetes mellitus and obesity, extending far beyond glycemic control. The GBA and ENS are now recognized as key to explaining these effects. As the primary site of semaglutide action, the intestine hosts a complex neuro-endocrine-immune-microbial network whose remodeling remains incompletely elucidated. Emerging evidence suggests that the signaling hub formed by transforming growth factor-β (TGF-β) and brain-derived neurotrophic factor (BDNF) plays a central role in this process. This review summarizes recent advances in semaglutide-mediated regulation of intestinal neuroplasticity, inflammation, and metabolic homeostasis, with a focus on how semaglutide modulates the TGF-β/BDNF signaling hub to reshape intestinal sensory perception. Specifically, upon activation of GLP-1R, semaglutide phosphorylates and inhibits glycogen synthase kinase-3β via the phosphatidylinositol 3-kinase /protein kinase B pathway, thereby relieving its suppression of cyclic adenosine monophosphate-response element-binding protein and promoting BDNF transcription. Concurrently, it downregulates the pro-inflammatory and pro-fibrotic TGF-β and its downstream Smad2/3 phosphorylation, improving inflammation and barrier function. The elevated BDNF further exerts neurotrophic, anti-inflammatory, and neuroprotective effects. Additionally, semaglutide reshapes the gut microbiota, increasing the abundance of beneficial bacteria such as Akkermansia and Bacteroides, repairing dysbiosis, enhancing short-chain fatty acid production, upregulating BDNF, inhibiting neuroinflammation, and indirectly modulating transient receptor potential vanilloid 1 activity. In summary, semaglutide-induced remodeling of intestinal sensory perception involves ENS plasticity, neuro-immune crosstalk, and microbial metabolic regulation, with the TGF-β/BDNF axis serving as a central node in this intricate process. Elucidating this mechanism may provide a basis for expanding the therapeutic scope of GLP-1R agonists to neurological disorders and developing novel GBA-targeted drugs.

Key Words: Glucagon-like peptide-1 receptor agonist; Semaglutide; Transforming growth factor β; Brain-derived neurotrophic factor; Transient receptor potential vanilloid 1; Gut-brain axis

Core Tip: Semaglutide,a leading glucagon-like peptide-1 receptor agonist, reshapes gut-brain communication by bidirectionally modulating the transforming growth factor-β/brain-derived neurotrophic factor signaling axis in the gut. It promotes neuroprotective brain-derived neurotrophic factor via phosphatidylinositol 3-kinase/protein kinase B/cyclic adenosine monophosphate-response element-binding protein while suppressing profibrotic transforming growth factor-β/Smad signaling. Combined with beneficial shifts in gut microbiota and anti-inflammatory effects, this action remodels enteric neural perception, extending its therapeutic potential beyond metabolism to neuropsychiatric and neurodegenerative disorders.

INTRODUCTION

Traditionally, the primary action of glucagon-like peptide-1 receptor agonist (GLP-1RA) such as semaglutide has been understood as glucose-lowering[1]. However, emerging evidence reveals that these agents exert broader protective effects, particularly in mitigating multi-organ fibrosis. Central to these actions are two critical signaling hubs: The transforming growth factor-β (TGF-β) and brain-derived neurotrophic factor (BDNF) pathways. The TGF-β pathway is a key driver of fibrosis and inflammation, playing a pivotal role in chronic pathologies of organs such as the liver[2], kidney[3], and heart[4]. Meanwhile, BDNF is a crucial neurotrophic factor[5] influencing both central and enteric nervous system (ENS) function[6,7].

Studies indicate that GLP-1RAs like semaglutide effectively suppress the TGF-β/Smad pathway, reducing the extent of organ fibrosis[2,3], which may represent a fundamental molecular mechanism underlying their multi-organ protective effects[1]. Furthermore, these agents demonstrate neuroprotective properties, such as alleviating neuroinflammation in models of brain injury[8]. Although direct evidence linking semaglutide to BDNF modulation in the gut remains limited, BDNF is known to be involved in energy metabolism and interacts with factors such as interleukin-6 (IL-6). Notably, glucagon-like peptide-1 (GLP-1)-based therapies can modulate AMP-activated protein kinase activity[1,2], and IL-6 in turn influences GLP-1 expression. This leads us to hypothesize the existence of an intricate regulatory network in the gut involving GLP-1 signaling, cytokines (e.g., IL-6), BDNF, and energy-sensing pathways (e.g., AMP-activated protein kinase). This network likely orchestrates enteroendocrine function, neural sensing, and gut-brain communication.

In summary, the pleiotropic effects of semaglutide are closely tied to its actions on the gut-brain axis (GBA) and the ENS, with the intestinal “TGF-β/BDNF signaling hub” playing a central role. Therefore, we believe that semaglutide may act on local glucagon-like peptide-1 receptor (GLP-1R) in the intestine to synergistically regulate TGF-β and BDNF signaling, thereby reshaping ENS function and gut-brain communication. This review will synthesize current evidence to explore the interplay between these two pathways, elucidating how semaglutide reshapes enteric neural perception, improves gut function, and coordinates systemic anti-fibrotic and neuroprotective responses. These insights may pave the way for novel therapeutic strategies in neurological disorders and future drug development (Figure 1).

Figure 1 Mechanism diagram. This mechanism diagram illustrates the core mechanism of semaglutide regulating the gut-brain axis: By activating the intestinal epithelial glucagon-like peptide-1 receptor, it inhibits glycogen synthase kinase-3β via the phosphatidylinositol 3-kinase/protein kinase B pathway to promote brain-derived neurotrophic factor transcription, while modulating the transforming growth factor-β/Smad pathway to improve inflammation and barrier function. Brain-derived neurotrophic factor exerts neuroprotective effects through tropomyosin receptor kinase B, remodeling enteric nervous system plasticity. In addition, the drug reshapes gut microbiota (e.g., Akkermansia, Bacteroides), increases short-chain fatty acids, inhibits harmful metabolites, indirectly modulates transient receptor potential vanilloid 1 and potassium channels, alleviates neuroinflammation, and connects to cognition via the vagal afferent pathway, achieving dual metabolic and neuroprotective effects. TRPV1: Transient receptor potential vanilloid 1; SCFA: Short-chain fatty acids; GPR43: G protein-coupled receptor 43; BDNF: Brain-derived neurotrophic factor; GLP-1R: Glucagon-like peptide 1 receptor; GLP-1: Glucagon-like peptide 1; PI3K/Akt: Phosphatidylinositol 3-kinase/protein kinase B; GSK-3: Glycogen synthase kinase 3; CREB: Cyclic adenosine monophosphate-response element-binding protein; TrkB: Tropomyosin receptor kinase B; TGF: Transforming growth factor; EMT: Epithelial-mesenchymal transition; EGC: Enteric glial cell; ENS: Enteric nervous system.

TGF-Β/BDNF: THE SIGNALING HUB IN THE GBA

The core roles of TGF-β and BDNF

TGF-β and BDNF are core signaling molecules of the GBA, collectively forming a hub that integrates neural, immune, and microbial signals. As a secreted polypeptide growth factor[9], TGF-β regulates enteric neuronal homeostasis, inhibits neuroinflammation[10,11], strengthens the intestinal mucosal barrier and gut microbiota balance, and mediates bidirectional gut-central nervous system communication via the ligand-receptor-Smad pathway[12]. It also serves as a key driver of organ fibrosis, with excessive activation exacerbating fibrosis in tissues such as the intestine, liver, and kidneys[13]. BDNF, a specific polypeptide neurotrophic factor, is widely expressed in various intestinal cells. It regulates neuronal survival, proliferation, and axon growth[14,15], exerts neuroprotective effects against inflammation/oxidative stress-induced damage[5], maintains intestinal barrier function, and mediates bidirectional GBA regulation[16,17]. An interaction exists between TGF-β and BDNF: BDNF may exert feedback regulation on TGF-β signaling by inhibiting Smad2/3 phosphorylation, and the two molecules collaboratively maintain intestinal homeostasis (Figure 1).

How TGF-β/BDNF regulates the ENS

In the ENS, TGF-β and BDNF form a critical bidirectional regulatory axis that jointly maintains intestinal homeostasis. Dysregulation of this signaling network is a core mechanism underlying functional gastrointestinal disorders. TGF-β is a multifunctional core cytokine in the ENS, primarily produced by neurons, glial cells, and resident immune cells (e.g., macrophages). Its roles in the ENS can be categorized into two pathways: The classical Smad signaling pathway, where TGF-β activates Smad2/3 to form a complex with Smad4, regulating gene expression and modulating non-transcriptional processes[18]. In intestinal neuroregulation, TGF-β secreted by M2 macrophages specifically upregulates the oxytocin signaling system (oxytocin/oxytocin receptor) by activating the neuronal Smad2/3 pathway and inhibiting paternally expressed gene 3 expression, a mechanism confirmed during disease recovery[19]. The non-classical signaling pathway: Signal transducer and activator of transcription 3 serves as the integration point for neuro-immune dialogue. M1 macrophages inhibit neuronal oxytocin signaling via the signal transducer and activator of transcription 3/nuclear factor kappa-light-chain-enhancer of activated B cells pathway, while TGF-β also participates in regulation through non-Smad pathways (e.g., Janus kinase), collectively influencing ENS homeostasis[20]. Additionally, ENS-derived TGF-β guides macrophages in the muscular layer to adopt a neuro-supporting phenotype, a niche critical for neuronal survival, function, and intestinal motility. ENS depletion or impaired TGF-β signaling leads to reduced supportive macrophages, neuronal loss, and motor dysfunction[21]. TGF-β and the ENS are closely interdependent and mutually regulated, forming a highly integrated functional system.

BDNF is a core signaling molecule in the ENS that regulates neuronal plasticity and function. It is primarily produced by neurons, glial cells, and neuro-supportive macrophages educated by TGF-β. Its key roles include: Binding to the tropomyosin receptor kinase B (TrkB) receptor on enteric neurons and enhancing neuronal excitability by inhibiting voltage-gated potassium channels, which forms the molecular basis for visceral hypersensitivity and pain amplification[22,23]; and stimulating secretomotor neurons to release vasoactive intestinal peptide and acetylcholine, thereby activating the cystic fibrosis transmembrane conductance regulator channel on intestinal epithelial cells, driving chloride ion and fluid secretion, and contributing to pathological diarrhea[5]. BDNF also serves as a critical mediator through which the gut microbiota influences ENS function: Beneficial metabolites such as short-chain fatty acids (SCFAs) can upregulate BDNF, whereas harmful metabolites like p-cresol sulfate downregulate its levels, thereby translating microbial signals into neural regulation. For example, Bifidobacterium bifidum CCFM1163 modulates the microbiota composition and SCFA levels to indirectly promote colonic BDNF expression[16,24]. In pathological conditions such as irritable bowel syndrome, abnormal enhancement of BDNF signaling directly leads to visceral hyperalgesia, secretory dysfunction, and forms a vicious cycle with symptoms like anxiety[25]. BDNF exhibits a “double-edged sword” characteristic in the ENS: It participates in physiological neural regulation under normal conditions, but drives visceral hypersensitivity and secretory dysfunction in disease. Its expression is precisely regulated by immune signals (e.g., TGF-β) and microbial metabolites, indicating that ENS dysfunction fundamentally results from disrupted dialogue among the immune system, microbiota, and nervous system.

Notably, in patients with depression, peripheral TGF-β levels are significantly positively correlated with BDNF levels, and TGF-β can upregulate BDNF expression[26], suggesting that the TGF-β/BDNF axis serves as a critical bridge connecting peripheral immune and central neurotrophic signaling. An increasing number of preclinical studies have provided direct evidence from molecular to systemic levels for the existence and function of this hub. At the molecular mechanism level, TGF-β exerts direct regulatory effects on BDNF. For instance, in astrocytes, the antidepressant effects of tetrahydrocurcumin depend on the activation of the TGF-β1/mothers against decapentaplegic homolog 3 pathway. When this pathway is blocked by the specific receptor inhibitor SB431542, its upregulatory effect on BDNF is completely abolished[27]. Additionally, the transcription factor Pax6 has been demonstrated to bind simultaneously to the promoter regions of both TGF-β and BDNF genes[28], suggesting that this hub may share a common upstream transcriptional regulatory program. At the cellular functional level, this hub integrates neuroprotective functions by modulating immune phenotypes. In depression models, pro-inflammatory M1 polarization of microglia is often accompanied by decreased TGF-β levels and reduced BDNF; whereas drugs such as piracetam can promote anti-inflammatory M2 polarization of microglia by upregulating signaling pathways like TGF-β, thereby significantly increasing BDNF expression[29]. This integrated pathway of “immunomodulation → neurotrophic output” is also supported by studies on the physiological ketone beta-hydroxybutyrate: Beta-hydroxybutyrate can simultaneously upregulate TGF-β and BDNF while improving microglial activation[30].

Of particular significance, this hub serves as the core functional module for bidirectional communication in the “microbiota-GBA”. In autism model mice, localized intestinal inflammation was associated with downregulation of TGF-β and BDNF expression in multiple brain regions. Notably, the use of anti-α4β7 monoclonal antibodies targeting only the gut to eliminate intestinal inflammation could remotely and selectively reverse the expression defect of this hub in the brain, thereby rescuing social behavior[31]. This strongly demonstrates that gut-derived immune signals can specifically regulate the central TGF-β/BDNF protective axis. Similarly, various interventions such as chicoric acid and aerobic exercise can synergistically upregulate TGF-β and BDNF in multiple organs including the gut, spleen, and brain, further highlighting their central role in systemic regulation[32].

In summary, the existing evidence has delineated a clear picture: TGF-β/BDNF constitutes a highly synergistic “immunoneurotrophic” hub that can be regulated at the metabolic, behavioral, and peripheral organ levels. It integrates microglial phenotypes and neuroplasticity at the central level[27,29], and mediates the influence of intestinal and systemic states on brain function at the systemic level[31,32]. Therefore, we hypothesize that the TGF-β/BDNF regulatory network in the ENS can be regarded as a fully functional “brain-gut interface” unit. However, directly elucidating the cell-specific co-localization, local bidirectional dialogue mechanisms, and direct regulation of this hub on intestinal motility, secretion, and other functions within the intrinsic neurons and glial networks of the intestinal ENS remains a critical issue requiring breakthroughs in future research.

SEMAGLUTIDE-INDUCED ACTIVATION OF THE TGF-β/BDNF SIGNALING CASCADE

Dual synergistic mechanisms targeting the pathological core

Semaglutide, as a long-acting GLP-1RA, activates the GLP-1R widely distributed on enteric neurons, enteroendocrine cells, and intestinal mucosal immune cells, initiating downstream signaling that precisely modulates the function and plasticity of the gut sensory neural circuitry. Its core mechanism involves the synergy of the following two pathways, which jointly target the pathological core of visceral hypersensitivity: Aberrant neural activity and the inflammatory microenvironment.

Upon binding to the GLP-1 receptor, semaglutide activates phosphatidylinositol 3-kinase (PI3K) and its downstream effector, protein kinase B (Akt). The phosphorylation of Akt subsequently leads to the phosphorylation of glycogen synthase kinase-3β (GSK-3β) at Ser9, thereby inhibiting its kinase activity. This pathway directly enhances enteric neuron function and stress resistance[33,34]. The inhibited GSK-3β can no longer suppress the transcriptional activity of cyclic adenosine monophosphate-response element-binding protein (CREB), allowing CREB to be phosphorylated at Ser133, translocate into the nucleus, and initiate the transcription of genes encoding neurotrophic factors, such as BDNF[35]. Within the ENS, the increased expression of BDNF is crucial for maintaining and repairing the survival, synaptic connectivity, and signal transduction of intrinsic sensory and interneurons. This pathway has been validated in studies of memory deficit disorders like sporadic Alzheimer’s disease[33], which elucidates how GLP-1RAs enhance synaptic plasticity and memory function by promoting CREB phosphorylation and BDNF expression. Similarly, in the gut, this pathway can enhance the stability and resilience of the enteric neuronal network, directly correcting aberrant electrical activity at the cellular level, thereby remodeling normal gut sensory thresholds and counteracting hypersensitivity states. Concurrently, BDNF is not only a neurotrophic factor but also possesses anti-inflammatory and neuroprotective properties. Elevated BDNF can further inhibit GSK-3β by reactivating the PI3K/Akt pathway through its TrkB receptor, forming a positive feedback loop that amplifies its own expression (Figure 1)[35].

Simultaneously, semaglutide exerts important anti-fibrotic and anti-inflammatory effects by bidirectionally modulating the TGF-β signaling pathway, although the direction of its effect is context-dependent. In the intestine, TGF-β is a central mediator regulating immune cell phenotype, inflammatory responses, and tissue repair, and its dynamic balance directly influences the microenvironment surrounding sensory nerve endings. In studies of cerebral ischemia/reperfusion injury, semaglutide promotes the polarization of microglia towards the anti-inflammatory M2 phenotype, upregulates the anti-inflammatory factor TGF-β, and suppresses pro-inflammatory factors like tumor necrosis factor-alpha, thereby mitigating neuroinflammation[8]. However, in fibrotic disease research, its effect is opposite, significantly reducing the expression of TGF-β1 and its receptor, and inhibiting the downstream epithelial-mesenchymal transition process, thus effectively alleviating fibrosis progression (Table 1)[36].

Table 1 Core molecular mechanisms of transforming growth factor-β/brain-derived neurotrophic factor signaling cascade activation by semaglutide.

Pathway name	Key regulatory nodes and direction	Regulatory direction	Primary biological functions	Ref.
Semaglutide/GLP-1R/PI3K/Akt/GSK-3β/CREB/BDNF pathway	Semaglutide activates the GLP-1R, leading to enhanced PI3K/Akt signaling. This results in the inhibition of GSK-3β, which in turn promotes CREB activation and increases the transcription of BDNF	Positive activation	Promotes neuronal differentiation, synaptogenesis, and plasticity. Improves memory and cognitive function. BDNF, via TrkB receptor, initiates a positive feedback loop that reinforces PI3K/Akt signaling, creating a neuroprotective cycle	[33,35]
Bidirectional regulation of TGF-β signaling by semaglutide	In neuroinflammatory contexts, semaglutide elevates anti-inflammatory TGF-β levels and promotes microglial polarization toward the M2 phenotype. In fibrotic disease contexts, it downregulates the expression of TGF-β and its receptors, thereby inhibiting the epithelial-mesenchymal transition process	Bidirectional regulation	Anti-inflammatory: Attenuates neuroinflammation and promotes tissue repair in models like cerebral ischemia/reperfusion injury. Anti-fibrotic: Inhibits excessive extracellular matrix deposition and mitigates fibrosis progression in models of the liver	[8,36]

GLP-1R: Glucagon-like peptide-1 receptor; PI3K/Akt: Phosphatidylinositol 3-kinase/protein kinase B; GSK-3β: Glycogen synthase kinase-3β; CREB: Cyclic adenosine monophosphate-response element-binding; BDNF: Brain-derived neurotrophic factor; TrkB: Tropomyosin receptor kinase B; TGF-β: Transforming growth factor-β.

Evidence from in vitro and in vivo studies across systems

Evidence from both in vitro and in vivo experiments supports these mechanisms. In human neuroblastoma SH-SY5Y cells, treatment with GLP-1 significantly enhances the expression levels of α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors, N-methyl-D-aspartic acid receptors, and dopamine receptor D1. It also increases the expression of synaptic proteins such as synapsin 1, synaptophysin, and postsynaptic density protein 95, while reducing the expression of the proliferation marker vimentin. This effect can be blocked by the PI3K inhibitor LY294002, confirming the regulatory role of the PI3K/Akt signaling axis in this process[34]. This suggests that semaglutide may enhance the maturation and functional connectivity of enteric neuronal synapses via the same pathway, providing cytological basis for its ability to reshape gut neural perception and improve signal processing precision. In hepatic stellate cell models, components extracted from serum exosomes of semaglutide-treated diabetic patients significantly lower the expression of α-smooth muscle actin, phospho-Smad2, and connective tissue growth factor, directly demonstrating anti-fibrotic activity[37]. In a streptozotocin-induced mouse model of sporadic Alzheimer's disease, GLP-1RAs significantly downregulate hippocampal levels of phosphorylated extracellular signal-regulated kinase 1/2, elevate the expression of phosphorylated GSK-3β (p-GSK-3β, Ser9) and phosphorylated CREB protein (p-CREB, Ser133), and increase the content of BDNF. These changes subsequently reduce β-amyloid deposition and tau protein phosphorylation, exerting a neuroprotective effect[38]. This study demonstrated that GLP-1RAs can upregulate BDNF via the PI3K/Akt/GSK-3β/CREB pathway. This neuroprotective and plasticity-modulating mechanism is evolutionarily conserved, supporting the notion that semaglutide can activate the same pathway within the ENS[39], enhancing the resilience and function of intrinsic gut neurons, thereby stabilizing visceral sensory output. Similarly, in a high-fat diet-induced mouse model of diabetic cardiomyopathy, the GLP-1RA liraglutide effectively inhibits intracellular lipid droplet accumulation in cardiomyocytes, reduces the expression of TGF-β and type I/III collagen, lowers oxidative stress marker levels, and improves the degree of myocardial fibrosis and cardiac function (Table 2)[38]. In gut-related studies, analogous mechanisms are anticipated to alleviate interstitial fibrosis associated with chronic inflammation, creating a more favorable physicochemical environment for sensory nerves and indirectly correcting visceral hypersensitivity[38].

Table 2 In vitro and in vivo experimental evidence supporting the mechanisms of action of semaglutide.

Evidence type	Experimental model	Intervention/treatment	Key findings and outcomes	Core targets/pathways involved	Ref.
In vitro	Human neuroblastoma SH-SY5Y cells	Treatment with GLP-1	Upregulated synaptic protein and neuronal receptor expression; blocked by PI3K inhibitor LY294002	PI3K/Akt signaling axis	[34]
	Hepatic stellate cell line LX-2	Exosomes derived from serum of T2DM patients treated with semaglutide	Reduced expression of fibrotic markers	TGF-β/Smad signaling pathway	[37]
In vivo	Streptozotocin-induced mouse model of sporadic Alzheimer’s disease	GLP-1 receptor agonist treatment	Increased hippocampal p-GSK-3β (Ser9), p-CREB (Ser133), and BDNF levels; decreased Aβ deposition and tau phosphorylation	GSK-3β/CREB/BDNF neuroprotective pathway	[33]
	High-fat diet-induced mouse model of diabetic cardiomyopathy	GLP-1RA	Inhibited cardiomyocyte lipid accumulation; downregulated TGF-β1, collagen I/III, and oxidative stress; attenuated myocardial fibrosis and improved cardiac function	TGF-β signaling; oxidative stress; fibrosis	[38]

GLP-1R: Glucagon-like peptide-1; PI3K/Akt: Phosphatidylinositol 3-kinase/protein kinase B; T2DM: Type 2 diabetes mellitus; TGF-β: Transforming growth factor-β; p-GSK: Phosphorylated glycogen synthase kinase-3β; p-CREB: Phosphorylated cyclic adenosine monophosphate-response element-binding; Aβ: Amyloid-β; CREB: Cyclic adenosine monophosphate-response element-binding; BDNF: Brain-derived neurotrophic factor; GLP-1RA: Glucagon-like peptide-1 receptor agonist.

RESEARCH GAPS AND FUTURE DIRECTIONS

The regulatory mechanism of semaglutide on TGF-β/BDNF signaling axis and the controversy of ENS plasticity

Current evidence chain’s core gaps lie in the mechanistic gaps of semaglutide in regulating the TGF-β/BDNF signaling hub and controversies in ENS plasticity: The BDNF-TrkB signaling pathway has been confirmed as a core target for neuroprotection in neurodegenerative diseases (such as Parkinson’s disease and Alzheimer’s disease), and activation of the GLP-1 receptor can improve cognitive function and inhibit neuronal damage through this pathway. This conclusion has been verified in studies on GLP-1-related drugs such as sitagliptin[40,41]. However, as a widely used clinical GLP-1 receptor agonist, the interaction mechanism between semaglutide and the TGF-β/BDNF axis still has significant research gaps: Existing literature has not yet clarified the direct regulatory effect of semaglutide on the expression of TGF-β and BDNF in the ENS, nor is there evidence confirming whether TGF-β plays an upstream mediating role in the pathway by which semaglutide regulates BDNF expression[41]. More critically, the interaction mode (synergistic activation or negative feedback regulation) between TGF-β and BDNF in the ENS microenvironment has not been elucidated, and direct reports on the quadruple synergistic effect of semaglutide, TGF-β, BDNF, and ENS are lacking. This poses a core obstacle to deciphering the GBA regulatory network of semaglutide. At the same time, the plasticity of the adult ENS itself is controversial: Some studies suggest that the adult ENS still has neurogenic potential, but direct experimental evidence is insufficient, and the regulatory direction of TGF-β/BDNF on ENS neurogenesis and gliogenesis has not been unified (some studies indicate that BDNF promotes neural proliferation, while TGF-β may exert an inhibitory effect)[42,43]. This controversy over plasticity further complicates mechanistic research, leaving the core scientific question of whether semaglutide affects ENS plasticity through regulating the TGF-β/BDNF axis and thereby exerts neuroprotective effects unresolved to this day.

The clinical bottlenecks of GLP-1RAs and the blind spot of interaction between the gut microbiota and ENS

Potential verification approaches primarily center on two core dimensions: The application of cutting-edge technologies in mechanism verification - providing feasible paths for addressing the aforementioned gaps - and the unresolved clinical translation bottlenecks of GLP-1RAs: The development of cutting-edge technologies such as single-cell and spatial omics, organoid culture, and meta-omics has provided feasible paths for filling the aforementioned mechanistic gaps. Single-cell RNA-seq technology has been successfully used to resolve the heterogeneity of ENS neuron subtypes and can accurately localize the molecular characteristics of specific functional neurons (such as TGF-β or BDNF-positive neurons)[44,45], while spatial transcriptomics technology can further clarify the spatial distribution of these neurons in intestinal tissues, providing a precise single-cell level tool for verifying the regulation of the local TGF-β/BDNF axis in the ENS by semaglutide. In addition, the co-culture system of intestinal organoids and ENS can simulate the interaction microenvironment between the mucosal barrier and nerves in vivo[46], which is expected to directly verify whether TGF-β-mediated BDNF expression is dependent on semaglutide activation; meta-omics technology can correlate the expression regulatory relationship between gut microbial metabolites (such as SCFAs) and TGF-β/BDNF by characterizing the metabolomic and transcriptomic profiles of the gut microbiota[46]. Real-time monitoring technologies such as voltage-sensitive dyes can also dynamically capture the activity changes of ENS neurons after semaglutide treatment, providing functional evidence for mechanism verification[47]. However, the clinical translation of GLP-1RAs in the treatment of neurological diseases still faces multiple bottlenecks: The low blood-brain barrier penetration rate limits the efficiency of drugs in regulating the central nervous system and local ENS signaling axes; the lack of efficacy-predictive biomarkers centered on TGF-β/BDNF makes precision medicine difficult to achieve; long-term safety data are insufficient, and large-sample clinical studies above phase 3 are scarce; at the same time, digestive dysfunction often associated with neurological disease patients may exacerbate the risk of gastrointestinal adverse reactions of GLP-1RAs[41,48]. These bottlenecks are intertwined with gaps in mechanistic research, and there is an urgent need to achieve breakthroughs through a closed-loop research of “technological innovation - mechanism elucidation - clinical translation”.

Another key focus lies in the interaction blind spots and research prospects of the microbiota-ENS-TGF-β/BDNF axis: The interaction between the gut microbiota and ENS is a core link in GBA regulation. Microbial metabolites (such as SCFAs) can affect ENS function by regulating the integrity of the mucosal barrier, but it remains unclear whether this process involves the mediation of the TGF-β/BDNF axis[44,46]. Current research gaps predominantly center on three fundamental dimensions. First, the causal nexus between key bacterial genera (e.g., Akkermansia) and the expression of TGF-β/BDNF in the ENS remains unestablished, thus precluding the definitive determination of whether microbial dysbiosis modulates BDNF expression via the modulation of TGF-β signaling. Notably, the directionality, specificity, and contextual dependence of this hypothesized interaction have yet to be rigorously validated through mechanistic studies. Second, the molecular mechanisms orchestrating the crosstalk between microbial metabolites and the TGF-β/BDNF axis remain inadequately elucidated, with a conspicuous paucity of sequential experimental validation for the proposed pathway ‘microbial metabolite (e.g., SCFAs) → TGF-β receptor binding → BDNF transcriptional activation’. In particular, the molecular mediators, post-translational modifications, and tissue-specific regulatory checkpoints within this cascade remain incompletely delineated. Third, whether semaglutide-mediated modulation of the gut microbiota indirectly modulates ENS function via the TGF-β/BDNF axis - thereby completing the “drug-microbiota-signaling axis-nerve” regulatory cascade - remains unresolved. Furthermore, the extent to which this cascade contributes to the therapeutic efficacy of semaglutide in pathological contexts (e.g., gastrointestinal dysfunction, metabolic disorders) necessitates systematic investigation using integrative in vitro, in vivo, and translational models. Future studies can use technologies such as single-cell spatial omics, organoid co-culture, and combined meta-omics analysis to focus on the aforementioned interaction blind spots. To directly test the hypothesis that “SCFAs modulate TGF-β signaling to influence BDNF expression in the ENS”, we propose a tiered experimental strategy. First, in vitro mechanistic dissection should utilize primary enteric neuron-macrophage co-cultures isolated from mouse jejunum[21], treated with physiological SCFA concentrations (acetate 10 mmol/L, propionate 2 mmol/L, butyrate 1 mmol/L) that reflect colonic luminal levels[49,50]. G protein-coupled receptor 43 (GPR43) (free fatty acid receptor 2) knockdown via specific small interfering RNA[51,52] will verify receptor specificity [with quantitative polymerase chain reaction PCR (qPCR) confirming a knockdown efficiency ≥ 70%], while 10 μM TGF-β receptor 1 inhibitor SB525334[53,54] determines if SCFA effects on BDNF are TGF-β-dependent, with phosphorylated Smad2/3 and BDNF measured at 12, 24, and 48 hours by qPCR and Western blot[33]. Complementary organoid-ENS assembloids cultured with integrated neural crest cells can then be exposed to SCFAs with/without semaglutide (100 nM[55]) spatial transcriptomics combined with immunofluorescence co-localization will map BDNF (neuron-specific β-III tubulin co-localization) and TGF-β1 (macrophage-specific CD68 co-localization) expression along the crypt-villus axis[56], with expected SCFA-induced reduction of TGF-β1 in macrophage-rich regions and BDNF upregulation in neuronal clusters, alongside detection of intestinal barrier molecules (occludin, zonula occludens-1)[51]. Second, in vivo causality testing via germ-free mouse colonization with defined bacterial consortia - including low-SCFA (Proteobacteria-dominant[57]), high-butyrate (Anaerostipes rhamnosivorans + Akkermansia muciniphila + Eubacterium rectale[58]) and high-propionate (Parabacteroides goldsteinii[49]) communities - should be combined with semaglutide treatment (30 nmol/kg/day subcutaneous injection[59]) for 4 weeks, followed by qPCR, 16S ribosomal RNA sequencing, gas chromatography-mass spectrometry-based SCFA quantification[60], and whole-mount phosphorylated Smad2/3 immunostaining of myenteric plexus. To isolate neuronal GPR43 contribution, conditional GPR43 knockout mice (GPR43ΔENS) generated using Cre-loxP recombination[61] should undergo colorectal distension-visceromotor response testing for visceral sensitivity and dorsal root ganglion BDNF-TrkB signaling analysis by Western blot[56]. Third, feedback loop validation can be achieved by intestinal luminal in situ injection of enteric-specific adeno-associated virus-TGF-β1 (serotype adeno-associated virus 9, promoter villin 1/synapsin 1[62]) in the germ-free model; if SCFA-mediated BDNF upregulation is suppressed, this confirms TGF-β negative feedback. Epigenetic crosstalk should be mapped via 13C-acetate stable isotope-resolved metabolomics[58], tracking histone H3 lysine 9 acetylation incorporation at BDNF promoter and TGF-β1 enhancer regions by chromatin immunoprecipitation sequencing[63]. Parallel clinical translation requires a pilot trial (n = 30 type 2 diabetes mellitus patients, excluding Inflammatory bowel disease/irritable bowel syndrome cases[56]) measuring fecal SCFAs (gas chromatography-mass spectrometry[60]), plasma TGF-β1 (enzyme-linked immunosorbent assay[64]), and serum BDNF (electrochemiluminescence[65]) before/after 12 weeks semaglutide treatment, with Pearson correlation analysis testing whether SCFA changes (especially butyrate) predict TGF-β/BDNF shifts as response biomarkers. Meanwhile, develop clinical biomarkers centered on the TGF-β/BDNF axis, optimize the dosage form and administration regimen of GLP-1RAs to improve ENS targeting, and ultimately provide “mechanism - technology - clinical” trinity theoretical and practical support for the application of semaglutide in the treatment of neurological diseases.

CONCLUSION

Semaglutide, a prominent agent among GLP-1 receptor agonists, owes its remarkable glucose-lowering and metabolic benefits largely to its ability to reshape the GBA. Current evidence suggests its core mechanism involves widespread activation of GLP-1 receptors, which enhances BDNF-centered neurotrophic and anti-inflammatory pathways while suppressing pro-inflammatory TGF-β signaling. These dual actions converge on intestinal neuro-sensory units, fine-tuning vagal communication with the brain, fortifying the gut barrier, and mitigating neural inflammation in both the gut and the brain[66,67].

Unraveling this mechanism carries significant clinical implications. It broadens therapeutic potential and inspires next-generation drug discovery. Given its neuroprotective properties, semaglutide could be repurposed for neurological disorders like epilepsy and multiple sclerosis, moving beyond metabolic diseases. And, it paves the way for personalized medicine. Future therapy could be tailored by using biomarkers like gut microbiota or TGF-β/BDNF levels to predict a patient’s individual response[68,69].

In conclusion, deepening our understanding of how semaglutide modulates gut neural perception via the TGF-β/BDNF axis not only illuminates the dynamic dialogue between gut and brain but also unlocks new possibilities in addressing the global burden of metabolic and neurological diseases.