Nerve growth factor (NGF), a pivotal neurotrophic factor, operates through a circulatory mechanism involving multiple cellular sources, receptor-dependent signaling, and the reshaping of the immune microenvironment. The cholinergic-NGF signaling axis constitutes the molecular basis for neuro-immune interactions in gastrointestinal tumors. Specifically, acetylcholine (ACh) stimulates NGF expression via cholinergic signaling, while NGF promotes the growth of cholinergic nerve axons through its high-affinity receptor, TrkA (9, 10). NGF activates diverse signaling pathways by binding to both the high-affinity receptor TrkA and the low-affinity receptor p75NTR. The TrkA-mediated PI3K/Akt/mTOR pathway promotes neuronal survival, axon growth, and synaptic plasticity, with reduced NGF levels correlating with diminished axon outgrowth (11).

Conversely, the p75NTR receptor can trigger apoptosis or inhibit pathological neural hyperplasia by activating the RhoA/ROCK and JNK/c-Jun signaling cascades. Within the tumor microenvironment, the NGF-TrkA signaling axis drives tumor-associated neurogenesis (TAN). NGF secreted by tumor cells attracts infiltrating sensory nerve axons, which release calcitonin gene-related peptide (CGRP) to activate the RAMP1 receptor on tumor cells (12). This activation initiates PI3K-Akt/CaMK pathway-mediated phosphorylation of Rb protein and the release of E2F transcription factors, establishing a self-sustaining NGF-CGRP positive feedback loop. This mechanism also reshapes the immunosuppressive microenvironment by significantly boosting the IL-6 secretory capacity of cancer-associated fibroblasts (CAFs), thereby inhibiting cytotoxic T cell function (13, 14).

Furthermore, NGF promotes immune evasion by modulating immune cell functions: TrkA activation on macrophages drives M2 phenotype polarization and the secretion of IL-10 and TGF-β. NGF can also upregulate programmed cell death protein (PD-1) expression in CD8+ T cells via the β2-AR/cAMP signaling pathway, which inhibits mitochondrial oxidative phosphorylation and contributes to T cell exhaustion (15).

A separate investigation into apoptotic mechanisms revealed that NGF not only contributes to tumor progression but can also trigger neuronal apoptosis upon its deprivation. Interestingly, this apoptotic signal is transiently activated, permitting a window for neuronal recovery. This recovery depends on the expression of the anti-apoptotic protein Bcl-xL. Consequently, neurons exposed to apoptotic signals in the absence of NGF can restore survival via these transient signals, effectively undergoing a “reset” (16, 17). This finding implies that tumor cells may similarly exploit transient apoptotic signals for survival, providing a novel perspective for therapeutic intervention.

Within the stressful tumor microenvironment, matrix metalloproteinase 9 (MMP9) upregulation accelerates the maturation of brain-derived neurotrophic factor (BDNF) (18). BDNF demonstrates potential diagnostic value, given its abnormal expression in gastrointestinal tumors (19). Lactate produced by these tumor cells via aerobic glycolysis is transported into CAFs through monocarboxylate transporter 1 (MCT1). This lactate influx acidifies the intracellular environment of CAFs, which activates the NF-κB signaling pathway and stimulates substantial BDNF secretion. The CAF-derived BDNF then nourishes tumor cells, persistently activating their surface TrkB signaling to establish a positive feedback loop (20). This circuit enhances tumor cell invasiveness and metastatic potential while remodeling the local immunosuppressive microenvironment into a chemotherapy-resistant niche. These findings offer a mechanistic explanation for the low chemotherapy sensitivity observed in some gastrointestinal tumor patients (21). To verify the functional necessity of BDNF in drug resistance, researchers used CRISPR/Cas9 to generate a BDNF-knockout gastrointestinal tumor cell model for ex vivo drug sensitivity assays. The knockout cells exhibited significantly increased sensitivity to chemotherapeutic agents like anlotinib and showed elevated drug-induced accumulation of reactive oxygen species (ROS). This directly confirmed that BDNF antagonizes chemotherapy-induced cytotoxicity by regulating intracellular redox homeostasis and apoptotic signaling pathways (22). Another study demonstrated that the long non-coding RNA BDNF-AS recruits WDR5 to regulate FBXW7 transcription, thereby inhibiting FBXW7-mediated ubiquitination and degradation of VDAC3, maintaining mitochondrial membrane stability, and ultimately counteracting ferroptosis (23). This finding expands the known mechanisms by which BDNF and its regulatory network confer resistance to tumor cell death. However, key issues remain: current research focuses predominantly on the singular regulatory effects of the BDNF-TrkB pathway, while potential synergistic or antagonistic interactions between CAF-derived BDNF and other cells in the TME have not been systematically investigated. Although BDNF-knockout experiments confirm its necessity in drug resistance, clinical trials of small-molecule TrkB inhibitors in gastrointestinal tumors have shown limited efficacy. Despite these limitations, BDNF and its regulatory network represent potential targets for precision therapy in gastrointestinal tumors, providing a clear direction for subsequent mechanistic studies and drug development.

Beyond neurotrophic factors, neuroimmunity involves numerous molecular signals. Netrin-1, a key axon guidance molecule, functions as a central regulator of neuroimmune crosstalk. It modulates both the immunosuppressive activity of MDSCs and the homeostasis of cancer stem cells (CSCs). In colorectal cancer, tumor-secreted Netrin-1 binds to the adenosine A2B receptor (A2BR) on MDSCs. This binding activates the cAMP/PKA pathway and induces cAMP-response element binding protein (CREB) phosphorylation, which potently enhances MDSC-mediated immunosuppression. Consequently, MDSCs inhibit CD8+ T cell proliferation and promote regulatory T cell (Treg) differentiation, thereby shaping an immunosuppressive tumor microenvironment (24, 25). Conversely, IL-4 and IL-13 secreted by TAMs upregulate neuronal Netrin-1 expression via STAT6 signaling, establishing a positive feedback loop that accelerates nerve fiber hyperplasia and tumor innervation. Furthermore, deleted in colorectal carcinoma (DCC) is a functional dependence receptor (DepR). The interaction between DCC and Netrin-1 can induce the phosphorylation of FAK, Rac1, and Cdc42, which plays a crucial role in axonal growth (26). Studies on oligodendrocyte death following brain injury have found that targeted inhibition of Netrin-1 expression or blockade of its interaction with DCC can restore the apoptotic capacity of tumor cells. This provides a molecular basis for developing precise therapeutic strategies for coloectal cancer (CRC) (27).

Adenosine (ADO), a key neuromodulator, is predominantly produced via the CD39-CD73 pathway, where extracellular ATP is hydrolyzed to AMP by CD39 and subsequently dephosphorylated to ADO by CD73. This enzymatic cascade is significantly upregulated in tumor cells, CAFs, and MDSCs, particularly under the hypoxic and oxidative stress conditions governed by the HIF-1α/NF-κB pathway (28). Within the TME, ADO influences tumor progression through diverse cellular and signaling mechanisms. ADO released from tumor-associated neurons, for instance, activates the A2B receptor on astrocytes, inducing metabolic reprogramming and increased lactate secretion that fuels aerobic glycolysis in cancer cells. Conversely, ADO binding to the A2A receptor on CD8⁺ T cells triggers the cAMP–PKA pathway, which directly inhibits T cell receptor (TCR) signaling and diminishes the release of interferon-γ (IFN-γ) and granzyme B, thereby attenuating antitumor immunity (29, 30). Preclinical evidence further indicates that A2A receptor overexpression in gastric and hepatocellular carcinomas promotes epithelial–mesenchymal transition (EMT) via the PI3K–AKT–mTOR axis, enhancing tumor invasion and metastasis. Meanwhile, ADO secreted by CAFs upregulates their own CD73 expression in a feedforward manner, creating a self-amplifying “ADO–CD73–ADO” circuit that exacerbates immunosuppression in the TME and contributes to chemotherapy resistance (31, 32). Current clinical research, however, is still predominantly focused on liver cancer and pancreatic ductal adenocarcinoma (PDAC), with a paucity of high-quality clinical data for other gastrointestinal malignancies. Most insights also remain heavily dependent on murine models. While molecules like CD39 and CD73 are well-studied, other critical regulators of ADO metabolism within the TME are not fully defined. Consequently, the regulatory landscape of ADO metabolism remains incomplete, hindering the development of precise therapeutic approaches (33). Given the reported toxicity of agents such as A2A receptor inhibitors, more tumor-selective and precise treatment strategies are therefore needed.

In conclusion, the bidirectional regulation of neuroimmunity in gastrointestinal tumors is driven by a diverse and intertwined molecular signaling network. These signals form multi-level interactions through transmembrane receptors, epigenetic modifications, and metabolic reprogramming. Together, they shape an immunosuppressive, neuroplastic, and metabolically adaptive tumor ecology (Figure 1). These signals also represent potent molecular targets for the treatment of gastrointestinal tumors, offering new directions for future therapies.

Recent research has further clarified the close association between gut nerves and immune cells (Figure 2). A chemogenetic screening study revealed that Trpv1+ nociceptive neurons activate intestinal Treg cell expression via the CGRP-RAMP1 signaling pathway, while Treg cells conversely promote the regeneration of these neurons (34). Perineural invasion (PNI) is a pivotal process in tumor infiltration and metastasis of gastrointestinal cancers. Zhao et al. identified that Schwann cells, the primary glial cells of the PNS, contribute to regulating NGF expression. Specifically, CRC cells downregulate von Hippel-Lindau (VHL) protein in Schwann cells via exosomal miR-21-5p. The downregulation of VHL protein leads to the stabilization of hypoxia-inducible factor 1α (HIF-1α), which subsequently upregulates the expression level of NGF, thereby forming a positive feedback loop that promotes cancer progression. In turn, NGF further stimulates the proliferation of Schwann cells, accelerating tumor infiltration and metastasis (35, 36). NGF then further stimulates Schwann cell proliferation, accelerating tumor infiltration and metastasis. These findings indicate that targeting NGF or exosomal miR-21-5p may have therapeutic potential. Within the tumor-glial niche, Schwann cells in the CRC microenvironment can activate the NF-κB pathway, inducing IL-8 secretion and creating a vicious cycle that advances cancer (37). Furthermore, TAMs can drive Schwann cells to secrete IL-33 via the secretion of basic fibroblast growth factor (bFGF), thereby promoting macrophage M2 polarization and facilitating neural invasion (38). Schwann cells also secrete CCL2 (monocyte chemoattractant protein-1), which binds to CCR2 receptors on cancer cells. This interaction promotes cancer cell migration along nerves and recruits immune cells, such as monocytes and memory T cells, to the tumor microenvironment, forming a pro-inflammatory network that fuels cancer growth. Blocking the CCL2-CCR2 signal has been found to significantly inhibit PNI (13).

Enteric glial cells (EGCs), a distinct intestinal glial population, are integral to the enteric nervous system and modulate the gastrointestinal tumor immune microenvironment through phenotypic plasticity and secretory functions (39). Experimental evidence indicates that depleting EGCs can slow cancer progression. EGCs secrete glial cell-derived neurotrophic factor (GDNF), which enhances cancer cell invasiveness by binding to GFRα1 receptors (40). Depleting EGCs experimentally slows tumor progression. These cells release GDNF, which binds GFRα1 receptors to increase cancer cell invasiveness. EGCs also secrete NGF and BDNF, stimulating neuronal branching and tumor-nerve interactions to accelerate perineural invasion (41).

Type 3 innate lymphoid cells (ILC3s), which are enriched in the intestinal mucosa and regulate the intestinal barrier, operate via a dual-signal mode mediated by β2-adrenergic receptors (β2-AR) and vasoactive intestinal peptide receptor 2 (VIPR2) (42). Sympathetic nerve-derived norepinephrine activates IL-22 transcription through β2-AR. In mouse models, Wang et al. demonstrated in vitro that IL-22 facilitates epithelial repair via signal transducer and activator of transcription 3 (STAT3) pathway (43). Notably, STAT3, a pathway commonly involved in CRC development, has been shown to act as a tumor growth factor in certain contexts, driving DNA damage and carcinogenesis under chronic inflammation (44). On the other hand, ILC3 directly senses intestinal neuronal signals through neuropeptide receptors such as VIPR2 and NMUR1. Experiments have shown that VIP, secreted by intestinal neurons, induces SHP-1 phosphorylation in CCR6+ ILC3 via VIPR2, inhibiting STAT3-mediated antimicrobial peptide expression. This promotes lipid absorption but elevates the risk of intestinal microbiota dysbiosis in obesity-related colorectal cancer models (45, 46). Further studies by Shao et al. revealed the key balancing role of the transcription factor FOXO1 in this process. Under homeostasis, FOXO1 promotes VIPR2 transcription while repressing expression of the adrenergic receptor ADRA2A, enabling ILC3s to preferentially respond to neuronal VIP signals and sustain IL-22-dependent barrier function (47). Further studies by Shao et al. revealed the key balancing role of the transcription factor FOXO1 in this process. Under homeostasis, FOXO1 promotes VIPR2 transcription while repressing expression of the adrenergic receptor ADRA2A, enabling ILC3s to preferentially respond to neuronal VIP signals and sustain IL-22-dependent barrier function (47). During chronic stress or within the tumor microenvironment, however, sustained sympathetic activation amplifies cAMP-PKA signaling, leading to FOXO1 phosphorylation and degradation. This shifts ILC3s toward an ADRA2A-dominated state, attenuates IL-22 secretion, and heightens ILC3 sensitivity to norepinephrine, thereby exacerbating cAMP accumulation and establishing a positive feedback loop.

ILC3s also respond to tissue damage via the PGE2-EP2 axis (48). Following IL-1β activation, autocrine PGE2 signaling in ILC3 induces the expression of pro-heparin-binding EGF-like growth factor (pro-HB-EGF), which is then cleaved to release soluble HB-EGF. This soluble factor inhibits TNF-induced epithelial cell apoptosis through epidermal growth factor receptor (EGFR) signaling, preserving barrier integrity. However, in chronic inflammation, the depletion of ILC3 leads to decreased HB-EGF levels, exacerbating TNF-mediated intestinal epithelial damage, thus creating a vicious cycle linking inflammation, barrier disruption, and tumor progression.

ILC3 also produces inflammatory factors, including IL-17A and IL-17F (49). In maintaining ILC3 homeostasis, IL-17D regulates ILC3’s functional stability through CD93 (50). Recent studies have also discovered that intestinal GABAergic neurons can stabilize intestinal homeostasis by inhibiting the release of the inflammatory factor IL-17A through their secreted neurotransmitters (51). Another study reported a positive correlation between the transmembrane protein neuropilin-1 (NRP1) and ILC3s, suggesting NRP1 as a potential target for barrier maintenance through NF-κB–mediated regulation of IL-17A production (52). In summary, ILC3s play a dual role in gastrointestinal tumors by participating in a complex neuroimmune network and influencing intestinal microbiota. The gastrointestinal neuroimmune environment, one of the most neuronally dense sites, involves cellular mechanisms far more complex and diverse than those highlighted here. Deeper exploration of these mechanisms remains a promising direction for future research. The table below summarizes recent mechanistic studies on relevant immune cells (Table 1).

Helicobacter pylori (Hp) infection serves as a critical driver of gastrointestinal carcinogenesis, reshaping the TME through intricate neuroimmune interactions (72). The progression from infection to gastric atrophy, intestinal metaplasia, and gastric adenocarcinoma (GAC) involves H. pylori-mediated regulation of bidirectional nervous and immune system crosstalk. Early in infection, H. pylori activates the PI3K-AKT-β-catenin signaling axis in gastric epithelial cells, inducing NGF overexpression and promoting sensory nerve axon hyperplasia alongside TAN. Furthermore, CagA enhances ROS resistance in gastric cancer stem cells, suppresses autophagy, and promotes the survival of malignant clones (73).

Transient receptor potential (TRP) channels, which are widely distributed in the nervous system and function as pain receptors, are modulated by the gut microbiota, including H. pylori, to influence macrophage differentiation. Following H. pylori infection, TRPM2-deficient macrophages exhibit M1 polarization, which intensifies gastric inflammatory responses. In contrast, TRPM7 promotes M2 macrophage polarization, inhibiting apoptosis and enhancing tumor-promoting functions (74, 75). Through “reverse signaling,” H. pylori also activates the NF-κB pathway in dendritic cells via the TLR4/MyD88 axis, inducing PD-L1 expression and inhibiting CD8+ T cell function. Tregs, meanwhile, inhibit neuronal apoptosis by secreting IL-35, thereby maintaining the density of tumorigenic nerve fibers (76, 77). Another study revealed an enrichment of mucus-degrading bacteria, specifically Akkermansia spp. and Ruminococcus spp., at H. pylori infection sites. This enrichment, coupled with a significant reduction in goblet cells, constitutes a dual disruption that elevates cancer risk and exacerbates deterioration of the intestinal environment (78). In addition to its impact on the gastric microbiota, H. pylori also exerts certain effects on the colonic microbiota. This may be related to the impairment of the gastric barrier and symptoms of gastric hypochlorhydria. The weakening of the acidic environment drives a large number of bacteria to escape from the gastric environment and migrate to the colon. Furthermore, experiments have found that H. pylori can induce a significant proliferation of antigen-specific Th17 cells in the small intestinal and colonic mucosa, providing favorable regulation for the survival of pro-tumorigenic microbiota, which is closely associated with poorer prognosis in CRC patients (79). In summary, H. pylori infection mobilizes the neuro-immune system through various mechanisms and reshapes the gut microbiota. It induces pro-inflammatory cytokines or metabolites that damage the intestinal barrier, thereby triggering a pro-carcinogenic microenvironment.

Fusobacterium nucleatum, a prominent tumor-promoting bacterium enriched in gastrointestinal cancers, has undergone numerous experimental investigations focusing on its mechanistic link with CRC (80, 81). Studies have shown that Fn can suppress METTL3’s m6A modification, promoting cancer metastasis (82). Fn also activates the TLR4/Myd88/NF-κB signaling pathway, transcriptionally upregulating miRNA-155-5p, which negatively regulates the expression of the DNA mismatch repair protein MSH6 (83, 84). This leads to increased genomic instability and impairs mismatch repair capability, ultimately promoting tumor cell immunogenic escape. Fn’s specific adhesin, FadA, binds to E-cadherin’s extracellular domain, activating the β-catenin signaling pathway and driving CRC cell proliferation and the release of inflammatory factors such as IL-8 and TNF-α (85). Additionally, Fn-derived LPS activates colon epithelial cell CCL2 secretion through TLR4 signaling, recruiting monocytes that differentiate into pro-inflammatory macrophages, which release IL-1β, inducing Th17 cell differentiation and IL-17A/F production (86, 87).

The IL-17 family of cytokines drives tumorigenesis in the immune system and enhances CAF IL-6 secretion indirectly regulated by IL-23 (60, 88, 89). This occurs through the activation of sensory neuron TRPV1 channels and the promotion of CGRP release, which has been shown to promote gastrointestinal tumor progression (12). Furthermore, LPS promotes BDNF secretion, enhancing tumor cell stemness and metastasis via TrkB (90, 91). The role of BDNF in promoting tumorigenic neurogenesis in gastrointestinal cancers has been previously discussed (92).

Moreover, short-chain fatty acids (SCFAs) secreted by Fn can regulate Th17 cell responses by activating the FFAR2 receptor on intestinal chromaffin cells (93, 94). The formic acid produced by Fn can also specifically promote the expansion of Th17 cells in an AhR signaling-dependent manner, thereby exacerbating intestinal barrier dysfunction and promoting the development of gastrointestinal tumors (95). Beyond mechanistic studies of Fn and CRC, Sen et al. identified iron overload as a significant factor contributing to CRC’s poor prognosis. Iron overload enhances the secretion of tumor-promoting chemokines such as CCL8 by inhibiting inhibitory phosphorylation of NF-κB p65 in macrophages, while TLR4/NF-κB signaling can be suppressed under iron deficiency conditions (96).

The gut microbiota directly influences neuronal excitability and synaptic plasticity by producing, consuming, or inducing key neuroactive molecules, thereby shaping a tumor microenvironment that can either promote or suppress tumor growth. In colorectal cancer stem cells, Zhu et al. demonstrated that the bacterial metabolite isovalerate (IVA) activates intestinal serotonergic neurons by inhibiting the chromatin-binding ability of the NuRD complex, which lifts its transcriptional repression of the tryptophan hydroxylase-2 (Tph2) gene promoter (97, 98). This process elevates 5-HT synthesis, and intestinal 5-HT subsequently activates the Wnt/β-catenin signaling pathway via its receptors to drive colorectal tumor growth and invasion. Furthermore, 5-HT receptor 5-Hydroxytryptamine receptor 2B (HTR2B) expression is significantly upregulated in gastric adenocarcinoma tissues, where it correlates positively with poor patient prognosis (99). Mechanistically, activated HTR2B complexes with the tyrosine kinase Fyn to directly phosphorylate the PI3K regulatory subunit p85, triggering the Akt/mTOR signaling cascade. This pathway remodels tumor cell metabolism by upregulating hypoxia-inducible factor HIF1α and the lipid transporter ABCD1, thereby inhibiting lipid peroxidation and ferroptosis (100).

Additionally, IL-33 can induce 5-HT secretion via the PLC-γ1-TRPA1 signaling pathway in neuroendocrine channels, offering a new perspective on intestinal homeostasis disorders (101). Through its receptors HTR2A and HTR3A, 5-HT activates prostaglandin E2 (PGE2) synthase (PTGES) in macrophages to promote PGE2 release. PGE2 then acts directly on intestinal stem cells (ISCs) via its receptors EP1 and EP4 to maintain their self-renewal and pluripotency (102). A separate study, however, reveals a distinct immunomodulatory mechanism for 5-HT, whereby tumor-infiltrating T cells employ TGM2 to “hijack” 5-HT molecules and convert them into anti-tumor weapons. In CD8+ T cells, 5-HT enhances glycolytic metabolism via TGM2-mediated GAPDH 5-HTylation, which supports CD8+ T cell proliferation, activation, and effector function by increasing ATP production and intermediate metabolite accumulation (103).

The enzyme TGM2 thus decouples the tumor-promoting and anti-tumor effects of 5-HT. While 5-HT drives malignancy in tumor cells through classical receptor signaling, TGM2-mediated GAPDH 5-HTylation reprograms glycolytic metabolism in CD8+ T cells to establish a “metabolic checkpoint,” transforming 5-HT from a tumor-promoting factor into an immunostimulant. This duality indicates that the biological effects of 5-HT depend critically on its cellular target and molecular modification mechanisms. Targeting TGM2 activation or HTR2B inhibition may therefore hold clinical translational significance. Currently, 5-HT₃ receptor antagonists such as ondansetron and granisetron are used to manage acute nausea and vomiting induced by radiotherapy and chemotherapy, while 5-HT₄ receptor agonists like mosapride treat gastrointestinal motility disorders (104). Nevertheless, the precise mechanisms of 5-HT in different cellular subpopulations within the tumor microenvironment remain unclear (105), and further detailed studies are required to determine whether its expression levels differ in metastatic lesions.

As a classical inhibitory neurotransmitter, Gamma-Aminobutyric Acid (GABA) is secreted not only by neurons but also originates from the intestinal flora, where it and its receptor subunits play crucial roles in gastrointestinal tumors. Research demonstrates that microbially derived GABA interacts with host GABA receptor signaling to synergistically regulate colon tumorigenesis (106). Regarding anticancer mechanisms, early work revealed that intestinal flora-derived GABA significantly inhibits colorectal cancer cell proliferation in vitro (107). Mechanistic studies indicate that intestinal GABAergic neurons secrete FMRFamide-like neuropeptides (FLP-6) to specifically suppress the transcription factors ZIP-10 and KLF-1, thereby activating the PMK-1/p38 MAPK signaling cascade (108).

GABA signaling directly influences ILC3 function. Recent work demonstrates that GABA markedly decreases insulin-like growth factor binding protein 7 (IGFBP7) transcription by downregulating expression of the LIP isoform of the transcription factor C/EBP-β. Acting as an autocrine signal, IGFBP7 then binds the IGF1 receptor (IGF1R) on ILC3s, thereby inhibiting their proliferation and limiting IL-17A production (51).

At the level of cancer-promoting mechanisms, studies such as those by Wei have identified abnormally high expression of the δ subunit GABA receptor (GABRD) in gastric cancer tissues. Systematic screening of downstream effectors confirmed that GABRD stabilizes cyclin D1 (CCND1) protein by inhibiting its ubiquitin-mediated degradation, thereby driving cell cycle progression and blocking p53-dependent apoptosis (109). Conversely, the GABA_B receptor subunit GABABR1 is expressed at low levels in colorectal cancer. Wang’s team found that its loss induces epithelial-mesenchymal transition (EMT) and enhances tumor cell migration and invasion by activating the Hippo/YAP1 signaling axis (110).

Inadequate sleep also influences tumorigenesis. Sleep deprivation elevates peripheral blood GABA levels, which promotes miR-223-3p expression in colon cancer cells. Through exosomes, miR-223-3p induces macrophage M2 polarization and IL-17 secretion to stimulate tumor cell proliferation and migration. Furthermore, miR-223-3p stabilizes cMYC and drives colon cancer metastasis by suppressing the E3 ubiquitin ligase CBLB, thereby reducing cMYC protein degradation (107). In summary, GABA acts as a concentration-dependent bidirectional regulator with complex biological effects in the gastrointestinal tumor microenvironment. It influences tumor progression through multiple mechanisms, including selective receptor subunit activation, rhythmic secretion, and epigenetic modification, underscoring its potential value as a therapeutic target. Future research should further analyze the spatiotemporal dynamics of the GABA signaling network and its interaction with tumor metabolic reprogramming to provide a theoretical foundation for developing precision therapies.

Beyond directly binding to receptors, microbial metabolites can also serve as metabolic substrates for epigenetic modifiers, fundamentally reprogramming the function and fate of immune cells within the tumor microenvironment. Research also found that specific gut microbiota regulate colonic sympathetic ganglia (CG-SMG) expression by producing SCFAs. This suppresses sympathetic nerve activation, and vagal afferent neurons sense microbiota metabolite changes, regulating sympathetic nerve activity. SCFAs activate G protein-coupled receptors, inhibiting NF-κB nuclear translocation and promoting cancer cell apoptosis (111). Butyrate, a specific SCFA, interacts with the GPR109A/HOPX axis to enhance the memory and cytotoxicity of CD8+ T cells, thereby improving the tumor immune microenvironment (112). In vitro analyses reveal a significant reduction in butyrate-producing bacteria in the microbiota of gastric cancer patients, and mouse model studies demonstrate that butyrate suppresses immunosuppressive factor production by downregulating STAT3 and NF-κB (113). Another study using a mouse model of bacterial depletion showed that SCFAs promote the recovery of intestinal neurons and participate in neurogenesis, indicating that an intact intestinal microbiota supports intestinal nerve generation (67). Consequently, microbiota dysbiosis can lower SCFA levels, which activates colonic sympathetic ganglia neurons, increases catecholamine release, inhibits gastrointestinal motility, and promotes tumor progression.

Tryptophan metabolites help maintain intestinal homeostasis through three classical pathways: the kynurenine (Kyn), indole derivative, and serotonin pathways (114). These three pathways have been extensively explored, with microorganisms playing a significant role (115). Tryptophan-metabolizing bacteria, such as Bacteroides and Lactobacillus, degrade tryptophan to produce metabolites like kynurenine, activating the aryl hydrocarbon receptor (AhR) and driving the formation of an immunosuppressive TME (116). A study found that the deubiquitinating enzyme USP14 enhances tryptophan metabolism and inhibits cytotoxic T-cell function by stabilizing the indoleamine 2,3-dioxygenase 1 (IDO1) protein (117). As IDO1 is an immunosuppressive enzyme, inhibiting USP14 in a colorectal cancer model reduced IDO1 expression, reversed T-cell exhaustion, and enhanced anti-PD-1 therapy efficacy without activating AhR (118). In this pathway, Fang et al. experimented with indole-3-carboxylic acid (ICA), a derivative of the probiotic Lactobacillus gallinarum. Their research revealed that ICA can downregulate IDO1 by competing with the aryl hydrocarbon receptor (AHR), reverse PD-1 efficacy, and reshape the tumor microenvironment (119). Another probiotic, Bifidobacterium breve, secretes indole-3-lactic acid (ILA). A recent experiment found that ILA inhibits the proinflammatory phenotype of macrophages through the AhR/AKT pathway, antagonizing colitis-associated cancer (120). While indole-3-lactic acid was previously known to inhibit IL-17 signaling to antagonize tumors, the protective effects in this mouse model were abolished by an AhR antagonist, confirming that bacterial metabolites directly regulate immune homeostasis through AhR (121). This has led to the discovery of various potential methods to promote cancer treatment through microbiota (122, 123). A previous study combined ginseng polysaccharides (GPs) with αPD-1 monoclonal antibodies, enhancing microbiota sensitivity and improving probiotic distribution (124). Wang et al. discovered that a diet rich in anthocyanins and dietary fiber can alter gastrointestinal microbiota distribution, enriching Bacteroides uniformis and Lactobacillus, thereby promoting ILA secretion. This potential high value of black rice diet may provide a new perspective for preventing intestinal tumors (125).